Journal of Undergraduate Life Sciences Volume 7 Issue 1 Spring 2013 by Journal of Undergraduate Life Sciences

JULS Journal of Undergraduate Life Sciences

1953 to 2013

Celebrating 60 Years of DNA

Volume 7 • Issue 1 • Spring 2013

Sponsors The Journal of Undergraduate Life Sciences would like to thank the following sponsors for supporting the production of this year’s issue.

Laboratory Medicine and Pathobiology Cells & Systems Biology Ecology & Evolutionary Biology Human Biology Trinity College University of Toronto Students’ Union

Cell & Systems Biology

ECOLOGY & EVOLUTIONARY BIOLOGY

ISSN 1911-8899

| | |

University of Toronto Journal of Undergraduate Life Sciences (Print) juls.sa.utoronto.ca jps.library.utoronto.ca/index.php/juls

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Volume 7 • No. 1 • Spring 2013

Contents 5

JULS

Letter From the Editors

News 6 The role of the ion transporter KCC2 in excitatory synaptic transmission

Elisse Magnuson and Olivia Herzog

Th22: A new cell type implicated in the mucosal immunopathogenesis of HIV

Aria Shakeri

Preparation of extra-cellular matrix scaffolds via growth factor immobilization for whole organ bioengineering Imindu Liyanage

Research Articles 10 Generation of more efficient lentiviral

constructs with modified U1snRNA for anti-HIV-1 gene therapy

Matt Douglas-Vail, Maria Malm, Alan Cochrane, Juan Carlos Zúñiga-Pflücker

An ab initio computational study of a model decarboxylated schiff base of asparagine as a precursor in carcinogenic acrylamide formation

Frank Y. Gao, Amro Dodin, Harun Mustafa, Natalie J. Galant, Imre G. Csizmadia

Sex Comb rotation in Drosophila melanogaster: Changes in tissue length and cell extrusion

Tahsin Khan, Larry Zhang, Juan Nicolas Malagon

Cell dynamics of Sex Comb morphogenesis in Drosophila melanogaster

Yunlong Liang, Ellen W. Larsen, Juan Nicolas Malagon

Relationship between population density, individual fitness, and blackspot infections in the Yellow Perch Perca flavescens

Elaine Y. Luo

Sequence-based typing of clinical isolates of Mycobacterium xenopi isolated from Ontario, Canada

Functionalized surface enhanced raman scattering gold nanoparticles: Size correlation of optical and spectroscopic properties and stabilities in solutions

Qi Jing G. Sun, Gilbert C. Walker

Using a D2-dopamine receptor chimera to assay genes essential to psychoactive drug response in Baker’s Yeast Zelun Zhang, Alexander Tigert, Marinella Gebbia, Corey Nislow

Review Articles 62 Keeping an eye on vision: Molecular genetics and evolution of visual pigment proteins

Benedict Darren, Belinda Chan

The “weakness” in single gene disorder research: The role of signaling pathways and vasoregulation in the dystrophin-glycoprotein complex (DGC) in Duchenne Muscular Dystrophy

Nicole Fogel

The mammalian unfolded protein response: A potential source of novel therapeutic targets in the hypoxic tumour fraction

Wenwan Lu, Darrell Desveaux

The mammalian unfolded protein response: A potential source of novel therapeutic targets in the hypoxic tumour fraction

Brandon Sit

Adoptive cell therapy: CD4+ and CD8+ T cells and the cells that educate them Nothando Z. D. Swan

Interviews 86 Dr. Thomas Jessell 88 Dr. Jeffrey Ravetch 90 Dr. Michael Young

Samia Mirza, David J. Farrell, Theodore K. Marras, Jennifer Ma, Daniel Liu, David C. Alexander, Julianne V. Kus, Frances B. Jamieson

JULS Spring 2013 Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

Journal of Undergraduate Life Sciences University of Toronto Staff List Editors-in-Chief

Ishita Aggarwal

Alainna Jamal

Alina Guna

Bellal Jubran

Senior Editors

Jacquie Lu

Glenda Sun

Members

Review Board

Anastasia Bosc Howard Chow Katie Dunlop Danica Estavillo Laureen Hachem Buddhi Hatharaliyadda

Faculty

Jingwei Chen

Jin Suk Park

Managers

Dr. Martha Brown Dr. Scott Browning Dr. Amy Caudy Dr. Jennifer Gommerman Dr. Tony Harris

Johnny Huang Julian Hung Bernard Ma Alissa Nicolucci Judy Tran Roman Zyla Dr. William Ju Dr. Martin Krkosek Dr. Sian Patterson Dr. Carol Schwartz

Associates

Layout

Gina Hou Deanna Lue Jennifer Su Kenneth Ting Priscilla Yung

Cover Design

Lucy Chau

Ellen Wong

Lucy Chau

Managers

News Editor

External Affairs Manager

Bahar Behrouzi

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Benedict Darren

JULS

Journal of Undergraduate Life Sciences University of Toronto Call for Submissions The University of Toronto Journal of Undergraduate Life Sciences (JULS) is always looking for submissions that showcase the research achievements of undergraduate life science students. We welcome manuscripts in the form of Research Articles or Reviews. Submissions must come from University of Toronto undergraduate students or undergraduate students outside of U of T who have conducted research for at least three months under the supervision of a faculty member at U of T. Research articles should present original research and address an area of the life sciences. Mini-reviews should focus on a specific scientific topic of interest or related to the research work of the author. Research articles should be between 2,000-3,000 words and mini-reviews between 1,500-2,000 words. All works must not have been previously submitted or published in another undergraduate journal. The deadline for submissions for each issue will appear on the JULS website at http://juls.library.utoronto.ca.

Contact Us JULS is always looking for contributions from writers, artists, designers, and editors. Please contact JULS at juls@utoronto.ca if you are interested in joining the JULS team, or have any questions regarding any matter pertaining to the journal, or visit our websites: http://juls.sa.utoronto.ca or http://jps. library.utoronto.ca/index.php/juls.

Letter from the Editors Dear reader, It is our pleasure to present you with the 2013 issue of the University of Toronto Journal of Undergraduate Life Sciences (JULS). This year, we continue to feature the work of undergraduate students who have conducted research in a variety of disciplines in the life sciences. Promoting the innovative work of undergraduate students has always been and remains central to the vision of JULS. The publication of this issue would not have been possible without the dedication and expertise of our staff and executive team, and for this we thank them. We would also like to thank the authors not only for their quality submissions, but for their perseverance and patience. Sixty years ago, a one-page article published in the pages of the prestigious British journal Nature revolutionized science. In it, Drs. James Watson and Francis Crick described the molecular structure of the DNA double helix. Nine years later, Watson and Crick shared the 1962 Nobel Prize in Physiology or Medicine with Dr. Maurice Wilkins for their work. Understanding DNA’s structure immediately solved one of biology’s greatest mysteries: how genetic instructions are transferred from one generation to the next. In doing so, their discovery opened the door to what has become one of the hottest areas of medical research: molecular genetics. The 2013 issue of JULS aims to celebrate the 60th anniversary of the publication of Watson and Crick’s iconic paper and the 50th anniversary of the Nobel Prize for DNA structure. As a result, the research articles published in this year’s issue of JULS celebrate the fields of cell and molecular biology and genetics. We highly encourage you to embark on a journey of scientific discovery by participating in the research opportunities that the University of Toronto has to offer. We hope that you will consider contributing to JULS in the future. Sincerely,

Alainna Juliette Jamal and Ishita Aggarwal Co-Editors-in-Chief, 2012-2013

NOTE: All articles in this issue as well as supplementary information are freely available online at http://juls.library.utoronto.ca/. If you would like to join the JULS team, submit an article or have any comments or suggestions, please feel free to contact us as at juls@ utoronto.ca.

Acknowledgements We thank Trinity College for their support in the creation of this year’s issue of JULS. © 2013 University of Toronto Journal of Undergraduate Life Sciences (JULS). Authors retain all rights to their work published in JULS except the right to publish in another undergraduate journal. The content and/or opinions expressed in this publication are expressively those of the authors and do not necessarily reflect the views of the University of Toronto or JULS.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

NEWS@UOFT

The role of the ion transporter KCC2 in excitatory synaptic transmission Elisse Magnuson and Olivia Herzog KCC2 is a protein of the K+-Cl- co-transporter family, notable for being the only protein of its family that can work under isotonic conditions and functioning exclusively in the central nervous system [1]. By extruding Cl- from the cell, it establishes a low intracellular Cl- concentration, which is vital for synaptic inhibitory transmission. Its ability to extrude Cl- under isotonic conditions is conferred to a protein domain located in its C-terminal, termed the ISO domain [2]. The Woodin Lab under Dr. Melanie Woodin at the University of Toronto is currently investigating the possible importance of KCC2 and the ISO domain at excitatory synapses. Previous research has shown that KCC2 may play a novel and important role in excitatory transmission in addition to its known role in inhibitory transmission. KCC2 proteins have been found in clusters near excitatory synapses, such as glutamatergic synapses, on the dendritic spine heads of hippocampal neurons [3]. The interaction between the C-terminal in KCC2 and the protein 4.1N, a cytoskeleton-interacting protein that helps to form and maintain the structural shape of organelles and cells, appears to play a major role in spine morphogenesis. When inhibiting or disrupting the interaction between KCC2 and the 4.1N protein, unusual spine maturation and a reduction in active excitatory glutamatergic synapses occur. A previous study determined that the effect of KCC2 on dendritic spines was not a result of its ion co-transporter function, but rather a result of a structural role [3], as expressing both the full-length KCC2 and nonfunctional KCC2 were capable of restoring normal spine morphology and some excitatory transmission in KCC2-knockout cells. However, KCC2’s ability to function isotonically means that it is possible it has a functional role in the dendritic spine, thereby increasing the efficacy of glutamatergic excitatory signaling. The Woodin Lab hypothesizes that the protein interaction may not be the only impact KCC2 has on dendritic spine maturation due to its unique ability to function isotonically. The Woodin Lab is now looking at the possibility of KCC2 having a functional role in excitatory transmission by examining the expression of GluR1 receptors, which are found predominately in excitatory synapses on dendritic spines, under different chemical conditions. Full-length KCC2 and non-functional KCC2 (lacking the ISO domain) are separately overexpressed in rat hippocampal neurons. GluR1 is stained by immunofluorescence and examined using a confocal microscope. A decrease in GluR1 stain-

ing in the cells expressing nonfunctional KCC2 would suggest the possibility of a functional role for KCC2 in the dendritic spines, while finding similar expression would suggest that KCC2 plays only a structural role. In the experiment, cells are cultured in one of three treatments: 1) baseline conditions, 2) TTX, which depresses neuronal activity, or 3) KCl, which activates neuronal activity. This investigates whether differences in expression are only apparent under exaggerated conditions. The study of this topic may eventually be useful in the research and treatment of neurological disorders, as we come to better understand the many functions of KCC2. The study of learning and memory may also be furthered, as dendritic spine morphology and plasticity have been linked to these topics.

References

1. Fiumelli H and Woodin MA. Role of activity-dependent regulation of neuronal chloride homeostasis in development. Current Opinion in Neurobiology 2007;17(1):81-86. 2. Acton B, Mahadevan V, Mercado A, Uvarov P, Ding Y, Pressey J, et al. Hyperpolarizing GABAergic Transmission Requires the KCC2 C-Terminal ISO Domain. The Journal of Neuroscience 2012;32(25):8746-8751. 3. Li H, Khirug S, Cai C, Ludwig A, Blaesse P, Kolikova J, Afzalov R, et al. KCC2 interacts with the dendritic cytoskeleton to promote spine development. Neuron 2007;56(6):1019-1033.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

News@UofT

JULS JULS

NEWS@UOFT

Th22: A new cell type implicated in the mucosal immunopathogenesis of HIV Aria Shakeri The earliest stage of human immunodeficiency virus (HIV) infection is characterized by “flu-like” symptoms, followed by entry into the chronic “asymptomatic” stage. Although the main hallmarks of HIV-associated clinical pathology, such as CD4+ T-cell depletion and chronic inflammation, are established in the chronic stage of infection, it is assumed that many of the events that drive this pathogenesis are initiated at earlier time points. Gutassociated lymphoid tissue (GALT) is an early mucosal target and a persistent viral reservoir, as well as a site of severe CD4+ T-cell depletion. Macrophages and CD4+ T-cells are the main GALT cell types targeted in early infection. HIV infection is also associated with increased permeability of the intestinal epithelium and chronic HIV infection causes a structural transformation of the intestinal mucosa, manifested by impaired epithelial cell maturation and villous atrophy. In a recent study, Kim et al. [1] implicated a new cell type in potentially explaining the link between mucosal immune compromise and intestinal barrier disruption. Th22 cells are a CD4+ T-cell subset that produces IL-22, which is a cytokine expressed by various immune cell types and acts on nonhematopoietic cells. It is involved in innate immune recognition and the pathological promotion of inflammation. IL-22 can also promote epithelial cell proliferation, survival, and repair in barrier tissues. This is achieved by inducing the expression of genes encoding prosurvival and proliferative factors. By measuring various immunological parameters in blood and sigmoid mucosal biopsies, Kim et al. [1] found that Th22 cells constitute a major source of mucosal IL-22 in uninfected individuals and are the only major IL-22-producing cell type to be significantly lower in an untreated HIV-infected individual, compared to an uninfected individual. They then investigated the intestinal epithelial damage manifested by an increase in the pore-forming tight junction protein, claudin-2, and a decrease in the seal-forming tight junction protein, ZO-1. These criteria were met only by the chronically infected patient specimens. Furthermore, some plasma markers of microbial translocation are increased in all HIV-infected individuals, while lipopolysaccharide (LPS, a component of bacterial cell walls) is only detected in chronically infected cases. Kim et al. [1] also assayed the reparative effects of IL-22 on the intestinal epithelium by using the reduction in transepithelial resistance across an in vitro

enterocyte monolayer system as a measure of epithelial damage. They found that the exogenously administered IL-22 abrogates HIV or TNF-induced epithelial damage. These results suggest that gut IL-22 production has a major role in the maintenance of the intestinal epithelium, hence the preferential depletion of Th22 cells is a significant step in HIV mucosal immunopathogenesis. These findings explain the observation that co-occurrence of both intestinal epithelial damage and the disruption of mucosal T-cell homeostasis is required for microbial translocation and systemic immune activation. HIVinduced disruption of the epithelial barrier leads to an enhanced translocation of antigens associated with luminal microbes. Over time, this antigen influx will elicit sustained immune activation, which is the driving force of progressive immune failure in chronic HIV infection. Hence, Th22 cells may be an appropriate focus for novel HIV therapeutics.

References

1. Kim, C. J., A. Nazli, O. L. Rojas, D. Chege, Z. Alidina, S. Huibner, S. Mujib, E. Benko, C. Kovacs, L. Y. Shin, A. Grin, G. Kandel, M. Loutfy, M. Ostrowski, J. L. Gommerman, C. Kaushic, and R. Kaul. A role for mucosal IL-22 production and Th22 cells in HIV-associated mucosal immunopathogenesis. Mucosal Immunology, 2012. 5 (6):670-680.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

NEWS@UOFT

Preparation of extra-cellular matrix scaffolds via growth factor immobilization for whole organ bioengineering Imindu Liyanage A flagrant inadequacy of modern medicine can be found in the treatment of end-stage organ failure. Presently, its sole definitive remedy is an orthotropic transplantation - usually with organs obtained from deceased volunteers; yet this treatment is facing a crippling shortage of donors with waitlists rapidly escalating beyond control [1]. Moreover, all transplantations of this nature risk chronic immune rejection, and often necessitate a life-long regime of immunosuppressant therapy, which is known to induce a host of deleterious side effects, including increased patient morbidity [1]. The logical solution is to fabricate patient–specific organs, and indeed whole-organ bioengineering has become a focus in the field of regenerative medicine. However, fundamental complications arise because the regeneration of an organ entails more than just the production of its parenchymal cells (organ tissue); for even if this were achieved, the resulting mass of cells would be more reminiscent of a tumour, rather than an organ itself. This typically occurs as a functioning organ mounts its parenchyma onto a protein based scaffold, referred to as the extracellular matrix (ECM), which both defines and shapes its structure [1]. Yet research has identified that in addition to its simple structural role, the ECM houses a biochemical landscape of factors, which either mechanically or chemically influences the functionality and fate of cells [4]. Moreover, the scaffold itself is constantly changing in response to metabolic activity and micro-environmental conditions [1]. Ultimately, all these factors must be considered when fabricating a whole organ, which regrettably is beyond the scope of current science. Instead, elementary scaffolds that consist of hydrogels or porous collagens are being employed to examine the proliferation and development of cells [4]. However, this very proliferation generates a fundamental problem: as these systems grow in size, cells at the interior of the formations face deficits of oxygen and other vital nutrients due to the physical limitations of diffusion distances [3]. In complete organs, a vasculature delivering nutrients to these cells would develop; yet these synthetic systems are often incapable of such a development, and thus a necrotic core forms – limiting cell survival. Yet, in a recent study, published in Acta Biomaterialia, Shoichet and colleagues demonstrate that manipulations in growth factors could be employed to direct cellular development within the scaffold. Specifically, a concentration gradient of vascular

endothelial growth factors (VEGF - 165) was shown to guide endothelial cell growth into the scaffold, such that interior necrosis was diminished. These protein growth factors were first activated with N-hydroxysuccinimide and immobilized to the scaffold via differing methods to form a gradient that extended radically, with the greatest concentrations occurring at the interior of the scaffolds. This generated an approximate 2ng/ml/mm immobilized gradient, which ultimately exhibited greater migration of cells into the scaffolds, and not proliferation, in comparison to the uniformly immobilized VEGF – 165 and VEGF-free controls, hence reducing the aforementioned necrosis. Furthermore, these modified hydrogels may also promote the ingrowth of vasculature, securing future cell viability as well [3]. This research will provide a basis for future studies of in vitro scaffolds, and ultimately aid in the fabrication of whole-organ extra cellular matrices; which themselves will become the very foundations on which whole organs will be constructed.

References

1. Badylak S, Taylor D, Uygun K. Whole-organ tissue engineering: Decellularization and recellularization of three-dimensional matrix scaffolds. Annual Review of Biomedical Engineering 2011; 13:27-53. 2. Badylak S, Faulk D, Fox I, Fukumitsu K, Gramignoli R, Jiang H, Komori J, Lagasse E, Medberry C, Nagaya M, et al. A whole-organ regenerative medicine approach for liver replacement. Tissue Engineering, Part C: Methods 2011; 17(6):677-686. 3. Odedra D, Chiu LLY, Shoichet M, Radisic M. Endothelial cells guided by immobilized gradients of vascular endothelial growth factor on porous collagen scaffolds. Acta Biomaterialia 2011; 7(8):3027-35. 4. Owen S and Shoichet M. Design of three-dimensional biomimetic scaffolds. Journal of Biomedical Materials Research 2010; 94A(4):1321.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Thinking about Graduate Studies in the Life Sciences, Biomedical Research, Rehabilitation Sciences, or Community Health? Visit the Graduate and Life Sciences Education (GLSE) and the 15 departmental websites. Talk to our departmental graduate coordinators.

IT IS NEVER T O O E A R LY T O BE INFORMED ! For more information visit: http://www.facmed.utoronto.ca/programs/graduate.htm

JULS

RESEARCH

Generation of more efficient lentiviral constructs with modified U1snRNA for anti-HIV-1 gene therapy Matt Douglas-Vail 1, Maria Malm 2, Alan Cochrane 3, Juan Carlos Zúñiga-Pflücker 2 Human Biology Program, University of Toronto; email: matt.douglas.vail@utoronto.ca Department of Immunology, University of Toronto; Sunnybrook Research Institute, Toronto Ontario 3 Department of Molecular Genetics, University of Toronto 1 2

Abstract

The difficulty in treating drug-resistant strains of HIV-1 has highlighted the need for the development of novel inhibitory methods to complement highly active antiretroviral therapy (HAART). To address this issue, lentiviral vectors were used to transduce human CD34+CD38-/lo hematopoietic stem cells (HSCs) with modified U1snRNAs that suppress HIV-1 RNA processing. Engraftment of transduced HSCs has the potential to generate HIV-1 resistance, as cells normally targeted by HIV-1, CD4+ T cells, dendritic cells and macrophages, are reconstituted by the modified HSCs [1]. These modified U1snRNA constructs contain a 5’ 10-nucleotide substitution that binds to highly conserved 3’ terminal exon regions of HIV-1. This binding prevents 3’ end formation of pre-mRNA, reducing HIV-1 protein expression and viral replication [2]. However, a limiting factor of this method of gene therapy is the difficulty in obtaining a suitable number of transduced CD34+CD38-/lo HSCs for transplantation. Here we show that anti-HIV-1 snRNA lentiviral vectors can be produced with sufficient titer to transduce up to 37% of cord blood-derived CD34- cells. Three protocols were used to produce lentiviral particles (vectors) carrying different U1snRNA constructs, as well as a GFP reporter gene. The titers of progeny viral vectors were compared in 293T (human embryonic kidney) cells and in cord blood-derived CD34- cells. Dr. Alan Cochrane kindly provided the transfer plasmids, encoding the constructs to be package within the lentivral particles. The preparation of vector particles produced using the serum-free expression media/ultrafiltration protocol transduced 36.9% of cord-blood derived CD34- cells, compared to viruses prepared by the original protocol, which transduced 7.61% of these cells.

Introduction

In 2009, the Joint United Nations Programme on HIV-1 and AIDS reported that there were 1.8 million AIDS-related deaths and 33.3 million people worldwide living with HIV-1, the causative agent of AIDS [3]. The entry of HIV-1 into host immune cells is facilitated by the expression of the gp120 protein that protrudes from the virus envelope and binds to lymphocytes, macrophages, or T cells expressing the protein CD4 [4]. The viral core, containing two copies of single stranded 10 kilobase RNA, reverse transcriptase, integrase, viral proteins, and two tRNAs, is released into the cytoplasm where reverse transcriptase converts the viral RNA into double-stranded DNA (dsDNA) [5]. Once generated, the dsDNA is inserted into the host genome where it serves as a template for the transcription of viral mRNAs. This mechanism allows viral mRNA and protein to be synthesized by host machinery [6]. Newly synthesized molecules of envelope protein (gp120) are inserted into the plasma membrane of the infected cell. Newly made genome and virion proteins assemble into particles that acquire a viral envelope by budding through the modified patches of plasma membrane. During this process, the virion protease, which is packaged inside the virions, cleaves internal virion proteins (Gag and Gag-Pol) to generate infectious particles [7].

There are two distinct phases to HIV infection. The acute stage of HIV-1 infection lasts from the first detection of HIV-1 RNA to seroconversion, typically three to four weeks after initial infection [8]. During this initial stage, viral reservoirs are established and CD4+ memory T cells are depleted by Fas-mediated apoptosis and lytic infection. During the second stage, chronic infection, there is significant immune activation, which generates new targets for HIV-1 infection. Cytotoxic T lymphocytes, antibodies, and attrition eliminate infected lymphocytes [6]. Short-lived T cells, characterized by limited regeneration potential replace the depleted T cells [9]. The hallmarks of chronic infection include depletion of CD4+ T cells, clonal exhaustion of T cells, a decrease in CD4+ and CD8+ halflife, abnormal T cell trafficking, and loss of memory T cell pools [1]. The current standard treatment for HIV-1 is the use of antiretroviral drugs. Highly Active Antiretroviral Therapy (HAART) suppresses HIV-1 replication and also partially restores immune function [10]. The most common regimens of HAART include two nucleoside reverse transcriptase inhibitors (NRTIs) and a nonnucleoside reverse transcriptase inhibitor (NNRTI) or a protease inhibitor (PI). NRTIs inhibit reverse transcription by preventing HIV-1 DNA chain elongation. NNRTIs inhibit reverse transcription

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Generation of more efficient lentiviral constructs with modified U1snRNA for anti-HIV-1 gene therapy

Promoter

LTR

Promoter

U1snRNA

GFP

LTR

Ellis Transgene

Transfection

Packaging Cell RNA Nucleus

Reverse Transcriptase

Viral RNA RNA-DNA hybrid

Co-Transfection VSV-G

Gag/pol

Tat

Double-Stranded DNA

REV

Integration

Stable Cell line HeLa Cells Or Cord Blood

Translation Transcription GFP Reporter

Genome

U1snRNA

Cytoplasm

Nucleus

Figure 1: Production of cells expressing anti-HIV-1 U1snRNAs. DNA encoding the modified U1snRNA along with a GFP reporter gene was transfected into 293T cells along with plasmids encoding proteins required for virus generation. 293T cells then produce virions, which can be used to infect cord blood-derived HSCs to express anti-HIV-1 U1snRNAs as well as the GFP reporter [12].

by binding reverse transcriptase causing a conformational change that inactivates it [11]. PIs disrupt the cleavage of Gag and GagPol viral proteins during budding [11]. HAART alone, however, is unable to completely rid the body of HIV-1 as the virus remains dormant in reservoirs. These restrictions have placed a demand for the development of new treatments, namely stem cell based therapies [2]. Hematopoietic stem cell transplantations (HSCTs) are able to reconstitute populations of immune cells that have been destroyed by HIV-1. These stem cells may also be genetically modified to prevent the infection of specific target cells. HSCT has the potential to restore total functional immunity to an individual infected by HIV-1 as the cells normally targeted by HIV-1, CD4+ T cells, dendritic cells and macrophages, are reconstituted by the modified HSCs [10]. The use of modified U1snRNAs in HSCT for the treatment of HIV-1 has shown much promise. U1snRNA is a small nuclear RNA responsible for encoding the human U1 small nuclear ribonucleoprotein particles (snRNPs). Wild type U1snRNP plays a role in the assembly and initiation of the spliceosome [2]. During mRNA maturation, the spliceosome is positioned at the appropriate intron and subsequently excises the introns from the pre-mRNA. By altering the first 10 nucleotides of the U1snRNA transcript, creating

an anti-HIV U1snRNA, the U1snRNP can be targeted to the 3’ terminal exon of the gene of interest. Sajic et al. [12] demonstrated that by modifying U1snRNAs to target highly conserved regions of the terminal exon they were able to disrupt pre-mRNA formation by inhibiting the HIV-1 cleavage of pre-mRNA as well as inhibiting the addition of the poly(A) tail required for the mRNA to be exported from the nucleus [2]. In addition, Sajic et al. [12] used lentiviral vectors to transduce human CD34+CD38-/lo HSCs with anti-HIV U1snRNAs (Figure 1). Engraftment of these transduced HSCs in immune deficient NOD/SCID/γcnull humanized mice generated HIV-1 resistance [2]. Sajic et al. [12] used an HIV-1 challenge of HSC-derived T cells from these mice to demonstrate that their modified U1 a-HIV-1 #1 and modified U1 a-HIV-1 #5 snRNAS reduced HIV-1 protein production by 76.6% and 70.65% respectively based on a p24 ELISA [12]. One caveat of this approach, however, is that no thymus engraftment was observed in the mice injected with U1-wt snRNA or the U1 a-HIV-1 #1 snRNA [12]. It was also unclear if the transduced cells affect the ability of the mice to generate progenitor T cells in the bone marrow. These issues necessitated further study of the off target effects of the transduction and transplantation of HSCs modified by these anti-HIV snRNAs. To acquire a suitable num-

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Generation of more efficient lentiviral constructs with modified U1snRNA for anti-HIV-1 gene therapy

ber of transduced CD34+CD38-/lo HSCs for transplantation, high titer U1snRNA lentiviruses were required. The goal of this study was to develop more efficient antiHIV-1 snRNA lentiviral vectors for the purpose of HSC transduction. Three protocols for lentiviral production were tested. Two expression plasmids, modified U1 a-HIV-1 #1 (U1-1) and modified U1 a-HIV-1 #5 (U1-5) were used. Each construct also contained a reporter gene for GFP. The same envelope and packaging vectors were used for each method tested. The titer of each method was determined by the number of cells expressing GFP by transducing 293T cells with the vector. Cord blood-derived CD34- cells were then transduced to test lentiviral efficacy.

Materials and Methods

Day 1

Envelope Vector VSV-G

Packaging Vector(s)

Transfer Vector

Gag/Pol/Rev/Tat

cDNA/shRNA

293 T Cells

Transfect Target Cells

Day 2

Generation of Lentivirus via Calcium Phosphate Transfection of 293T Cells

Freestyle media

fresh media fresh media 293T cells, stored at -159°C, were thawed in a 37°C water bath. Cells were gently transferred to a 10 ml Falcon tube (Becton Dickinson, Cat. 352070) containing 10 ml DMEM (Thermo LENTI - X Ultrafiltration Ultrafiltration Scientific Cat. 3008.01) and centrifuged for 5 concentrator minutes at 1500 rpm (Allegra X-22R Centrifuge, Beckman Coulter). The supernatant was discarded and cells were resuspended in 12 mL of DMEM. A 3 ml volume of the cell suspension Harvest Lentivirus Harvest Lentivirus was added to four 10 cm dishes, each containHarvest Lentivirus ing 7 ml of DMEM. Cells were cultured in an incubator for 24 hours at 37°C (Forma Series II Water Jacketed CO2 Incubator, Thermo Scientific). Cells were between 60%-80% confluent and evenly distributed prior to transfection. Infect Target Cells All liquids were at room temperature and all 293T/ CB cells volumes below were measured per 10 cm plate of 293T cells. The following plasmids were added to a 50 mL Falcon tube: 3 µg GagPol, 3 µg Rev, 3 µg Tat, 1 µg VSV-G and 10 µg modified U1snRNA (U1-wt, U1 a-HIV-1 #1 or U1 a-HIV-1 #5). 450 µl of sterile TE (Invitrogen, Cat. 12090-015) and 50 µl of 2.5 M CaCl2 (Fischer Scientific, Cat. C79-500) were added Figure 2: Overview of methods for the generation of lentiviral constructs. Lenti X method: Day 1-Lentivirus was generated by Calcium Phosphate transfection of 293T cells. to the tube and mixed by vortexing. 2x BES solution Day 2-Fresh media placed on cells. Day 3-Cells harvested by Lenti-X Concentrator. (Sigma-Aldrich, B4554-25G) was added drop-wise Ultrafiltration Method: Day 1-Lentivirus was generated by Calcium Phosphate transfection of 293T to the tube as a pipette aid was used to continuously cells. Day 2-Fresh media placed on cells. Day 3-Cells harvested by Ultrafiltration. blow air through the plasmid-TE-CaCl2 mix. The Freestyle media + Ultrafiltration method: Day 1-Lentivirus was generated by Calcium Phosphate transfectube was incubated at room temperature for 20 tion of 293T cells. Day 2-Media on cells replaced with Freestyle media. Day 3-Cells harvested by Ultrafiltration. minutes, during which 100 µl of chloroquine (Sigma-Aldrich, Cat. C-6628) was Harvesting Lentivirus Particles added to each plate of 293T cells containing 10 ml DMEM to increase transMaking sure to detach all visible cells from the plate, cells were transfection efficiency. The plasmid-TE-CaCl2-BES mix was added drop wise while ferred to a 50 ml Falcon tube and centrifuged at 1,700 rpm for 10 minutes swirling the plate of 293T cells. The transfected cells were incubated overnight at room temperature (Allegra X-22R Centrifuge, Beckman Coulter). The at 37°C. The media on the cells was changed 16 hours following transfection supernatant was filtered into a fresh 50 ml tube using a 0.45 mm filter and was replaced with either pre-warmed DMEM or pre-warmed serum-free (Becton Dickinson, Cat. 352340). The lentivirus particles were concentratGibco® FreeStyle 293T Expression Medium (Life Technologies, Cat. 12338-018). ed with Lenti-X Concentrator (Clontech, Cat. 631231) or by Ultrafiltration The transfected cells were cultured at 37°C for an additional 24 hours. in a Vivaspin 20 Concentrator (Sartorius Stedim Biotech, Cat. V52041)

Day 3

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Generation of more efficient lentiviral constructs with modified U1snRNA for anti-HIV-1 gene therapy

4000

3000

1/10

DAPI-A

SSC-A

2000

1000

1/20

63.9

0 87.3

1000

2000 FSC-A

3000

4000

1000

2000 FSC-A

3000

4000

# Cells

120

98.7

98.8 # Cells

0.65

# Cells

0 0 0

10 FITC-A

0 0

FSC-A, DAPI-A subset 9801.fcs Event Count: 2790

102

103 FITC-A

104

105

FSC-A, DAPI-A subset 9803.fcs Event Count: 1725

1/40

102

103 FITC-A

104

105

FSC-A, DAPI-A subset 9805.fcs Event Count: 2400

1/80

1/160

40 40

# Cells

0 0

10 FITC-A

FSC-A, DAPI-A subset 9807.fcs Event Count: 2615

0 0

10 FITC-A

FSC-A, DAPI-A subset 9809.fcs Event Count: 2653

1/320

1/640

120

80 # Cells

47.3

10 FITC-A

FSC-A, DAPI-A subset 9811.fcs Event Count: 2935

100

60 # Cells

66.4

1/1280

17.6

40 20

0 0

10 FITC-A

FSC-A, DAPI-A subset 9813.fcs Event Count: 3089

120

0 0

10 FITC-A

FSC-A, DAPI-A subset 9815.fcs Event Count: 2994

10 FITC-A

FSC-A, DAPI-A subset 9817.fcs Event Count: 3056

1/5120

1/2560

1/10240

120 60 90 6.24

4.55 # Cells

90 # Cells

# Cells

10.4

was aspirated and the lentivirus pellet was resuspended in 250 µl of PBS (Thermo Scientific, Cat. AXD34935) per 10 ml of filtered supernatant. The lentivirus suspension was aliquoted and stored at -80°C until use. For concentration by Ultrafiltration, up to 40 ml of filtered supernatant was transferred to a VivaSpin 20 Concentrator. The tube was centrifuged at 4,000 rpm for 45 minutes for every 20 ml of supernatant as specified by the manufacturer (Avanti J-20 XP Centrifuge, Beckman Coulter). 10 ml of PBS was then added to the tube and centrifuged at 4,000 rpm for 45 minutes. A total of 5 ml of PBS was added to the tube. The tube was centrifuged for 4,000 rpm for 30 minutes to wash the virus particles. The lentivirus particles were eluted from the membrane in 250 µl of PBS per 10 ml of filtered supernatant. The lentivirus suspension was aliquoted and stored at -80°C until use.

Generation of Retronectin Coated Plates

28.8 # Cells

# Cells

85.6 # Cells

95.9

Research Articles

490 µl of PBS and 100 µl of 1 mg/ml Retronectin (Takara, Cat. T100B) were added to each well of a 12-well polystyrene plate (Becton Dickinson, Cat. 351143). The plate was incubated for 1.5 hours at room temperature. The PBS and Retronectin solution was removed and each well was blocked with 500 µl of 2% BSA (Calbiochem, Cat. 2980) in PBS. The plate was incubated for 30 minutes at room temperature. The 2% BSA in PBS was removed and the wells were rinsed with PBS. The plate was stored at 4°C until use.

Transduction of 293T cells

3x104 293T cells per well were plated in 500 µl DMEM media in a FSC-A, DAPI-A subset FSC-A, DAPI-A subset FSC-A, DAPI-A subset 12-well Retronectin-coated plate. 9819.fcs 9821.fcs 9823.fcs Event Count: 2684 Event Count: 2709 Event Count: 1471 10 µl of polybrene (Sigma Aldrich, Figure 3: Transduction of 293T cells with the U1 a-HIV-1 #1 construct generated by Ultrafiltration. Flow cyto- Cat. 107689) was added to 10 ml of metric analysis for the expression of GFP, the reporter of the U1snRNA #1 construct. 3x104 293T cells per well DMEM media, 50 µl of this dilution were cultured at 37°C for 48 hours with serial two-fold dilutions of lentiviral suspension. One well received no was added to each well containing virus served as a negative control. Cells were gated DAPI low/negative, FITC high (Graph, top left). 293T cells. The cells v6.3.3) were incubated 12:39 PM Page 1 of 1 (FlowJo For concentration with Lenti-X concentrator, 3.3 ml of Lenti-X at 37°C for two hours. Eleven 1:2 serial dilutions of the lentivirus suspenConcentrator was added for every 10 ml of supernatant in the 50 ml tube sion were prepared. 50 µl of each dilution was added drop wise while and mixed by inverting the tube several times. The tube was incubated at swirling to one well of the 12-well plate. As a negative control, one well 4°C for 30 minutes. The tube was centrifuged at 3,000 rpm at 4°C for 45 received no lentiviral suspension. The cells were incubated at 37°C for minutes (Allegra X-22R Centrifuge, Beckman Coulter). The supernatant 48 hours. Cells were harvested by trypsinization using Gibco Trypsin 0

102

103 FITC-A

104

105

102

103 FITC-A

104

105

102

103 FITC-A

104

105

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

aug14.jo Research Articles

Generation of more efficient lentiviral constructs with modified U1snRNA for anti-HIV-1 gene therapy

0 uL virus

6 uL virus

50 40 0.45

8.92 # Cells

# Cells

0 0

102

103 FITC-A

104

105

FSC-A, DAPI-A subset 12752.fcs Event Count: 1790

102

103 FITC-A

104

105

FSC-A, DAPI-A subset 12780.fcs Event Count: 1783

Layout

1.5x105 CD 34- cells per well were plated in 500 µl aMEM media in a 24-well retronectin-coated plate and incubated at 37°C for 2 hours. 6 µl, 12.5 µl, 25 µl, 50 µl or 100 µl of lentiviral suspension was added to each well of the plate. As a negative control, one well received no lentiviral suspension. The cells were then incubated at 37°C for 48 hours. Cells were harvested by trypsinization and centrifuged at 1500 rpm (Allergra X-22R Centrifuge, Beckman Coulter) at room temperature for 50 minutes. The transduced cells were resuspended in 300 ul of 1:50,000 DAPI- FACS buffer and analyzed by flow cytometry. Cells were gated DAPI low/negative, FITC-A high.

Results

Calculation of Lentiviral Titer Via Transduction of 293T Cells

To determine Infection units/militer (IU/ml) the efficiency of each method of lentiviral

12 uL virus

40 9.43

15 # Cells

# Cells

0 0

102

103 FITC-A

104

105

FSC-A, DAPI-A subset 12782.fcs Event Count: 1781

102

103 FITC-A

FSC-A, DAPI-A subset 12784.fcs Event Count: 1813

(Proportion of GFP expressing cellsthe × Cells plated) of 293T cells using the 3 shows, after 48 hours, transduction = × Dilution Factor (Volume virus supernatant) U1snRNA #1 construct generated by Ultrafiltration. The propor-

tion of GFP expressing cells at each dilution was used to calculate the titer with the following equation:

preparation, the viral titer for each method was calculated. Figure

25 uL virus

50 uL virus

Sample titer calculation of U1 a-HIV-1 #1 construct generated by Ultrafiltration with a dilution Infection units/militer (IU/ml) factor of 1:10 of the virus supernatant: (Proportion of GFP expressing cells × Cells plated) = × Dilution Factor supernatant) = (0.987 GFP-expressing cells(Volume x 30,000virus cells)/(0.01 mL virus suspension) x dilution factor 10 = 104

105

100 uL virus

26.9

Sample titer calculation of U1 a-HIV-1 #1 construct generated by

2.96 x 107 IU/mL Ultrafiltration with a dilution factor of 1:10 of the virus supernatant: Sample titer calculation of U1 a-HIV-1 #1 construct generated by Ultrafiltration with a di (0.987 GFP-expressing cells × 30,000 cells) × Dilution Factor 10 factor=of 1:10 of(0.01 the virus supernatant: ml virus suspension)

! = (0.987 GFP-expressing = 2.96×10 IU/ml cells x 30,000 cells)/(0.01 mL virus suspension) x dilution facto

2.96 x 107 IU/mL

36.9

Transduction of Cord Blood-Derived CD34- Cells It was important to comparecells the×different lentivirus protocols (0.987 GFP-expressing 30,000 cells) Lenti-X (Control) Lenti-X = × Dilution for their ability to transduce cord blood-dervied CD34 cells. Factor 293T 10 (0.01 ml virus suspension) 20 U1 a-HIV #1 U1 a-HIV #1 20 cells%are useful indicator cells, however, CD34 cells % areGFPmore repGFP! IU/ml = of 2.96×10 resentative the target cells ofDilution interest. Dilution expressing expressing 0 0 Factor Factor cells lentiviral IU/uL cells HSCs IU/uL These vectors were developed to transduce 0 10 10 10 10 0 10 10 10 10 No virushowever CD34 0 - cells provide 0 No virus 0.1test the 0 FITC-A FITC-A a cost effective surrogate to 7 7 FSC-A, DAPI-A subset FSC-A, DAPI-A subset 10constructs’ 97.6 2.9 x 10 10 94 2.8 - x 10 12788.fcs 12786.fcs efficiencies. Figure 4 illustrates the proportion of CD34 7 Event Count: 1969 Event Count: 2016 Lenti-X (Control) Lenti-X 20 89.6 5.3 x 10 20 81.5 4.9 x 107 cells transduced using different amounts of U1 a-HIV-1 #5 virus 7 Figure 4: Transduction of cord blood-derived CD34- cells with the U1 a-U1 a-HIV #1 U1 a-HIV #1 40 75.6 9.0 x 10 40 50.1 6.0 x 107 8 HIV-1 #5 construct generated by Ultrafiltration + Freestyle media. Flow 80produced46.3 with media and concentrated by Ultrafiltration. % Freestyle GFP1.1 x 10 80 30.8 % GFP7.4 x 107 8/14/12 3:32 PM for the expression of GFP (FITC), Page the 1 ofreporter 1 (FlowJo v6.3.3) 8 cytometric analysis of theDilution Dilution expressing expressing These23.1 results suggest 160 1.1 x 10 that virus stocks 160 prepared with 18.1Lenti-X 8.6 x 107 8 7 U1 a-HIV #5 construct. 1.5x105 CD 34- cells per well were cultured at 37°CFactor IU/uL Factor cells 1.0to cells 320concentrator 10.8 appear x 10 320 5.13 have a lower titer in both 293T indicator4.9 x 107 8 for 48 hours with lentiviral suspension. One well received no virus as a640 No virus 6.69 0 Noproportion virus 3.65 - x 10 1.30x 10 640the 7.0 cells and in CD34- cells. Figure 5 illustrates of CD340.1 negative control. Cells were gated DAPI low/negative, FITC high. 10 3.48 97.6 10 1.57 94 x 107 2. 1280 1.3 x 108 2.9 x 107 1280 6.0 cells transduced using different amounts of U1 aHIV-1 #5 virus 7 20 2.19 89.6 20 1.51 81.5 2560 1.7 x 108 5.3 x 10 2560 1.1 x 108 4. 8 (Life Technologies, Cat. 15090-046) and centrifuged at 1500 rpm with Freestyle media by 40 1.33 75.6 9.0 xand 107 concentrated 40 Ultrafiltration 50.1 5120produced 2.0 x 10 5120 0.96 1.5 x 108 6. 8 8 (Allegra X-22R Centrifuge, Beckman Coulter) at room temperature for with the same with normal 80 0.98 46.3 1.1 x produce 10 80 culture 30.8 10240compared 3.0 x 10 virus 10240 1.28 me3.9 x 108 7. 8 8 160 23.1 1.1 x 10 160 18.1 5 minutes. The transduced cells were resuspended in 300 ul of 1:50,000 dium Average and concentrated by Lenti-X concentrator. Average 1.4 x 10 1.1 x 108 8. # Cells

# Cells

320 10.8 1.0 x 108 320 5.13 4. dilution of 4’-6’Diamidino-2-phenylindole DAPI (Sigma-Aldrich, Cat. 8 640 6.69 1.3 x 10 640 3.65 7. Ultrafiltration Ultra +Freestyle media D8417) in FACS buffer and analyzed by flow cytometry. Cells were gated Discussion 8 1280 3.48 1.3 xU1 10a-HIV 1280 1.57 6. U1 a-HIV #1 #5 DAPI low/negative, FITC-A positive. The data shows that1.7 lentiviral stocks 2560 concentrated 1.51 by 2560 2.19 x 108 1. % GFP% GFP8 Ultrafiltration had1.33higher2.0titer those concentrated with 5120 xDilution 10than 5120 0.96 1. Dilution expressing expressing Transduction of Cord Blood-Derived CD34- Cells Lenti-X Both3.0titers after Ultrafiltration 10240 concentrator. 0.98IU/uL xFactor 108 calculated 10240 1.28IU/uL 3. Factor cells cells x 108 No after Human umbilical cord blood samples were purified as previously than the titer calculated Nodevirusappear higher 0.65 Average 0 1.4 virus Lenti-X concentration. 0.71 Average 0 1. scribed [2]. Cord blood-derived CD34 cells stored at -80°C were thawed Viruses produced by the Ultrafiltration/Freestyle media protocol Ultra +Freestyle media in a 37°C water bath. Cells were transferred to a 15 ml tube (BeckmanUltrafiltration achieved a transduction rate of 36.9%, compared to viruses generU1 a-HIV #1 U1 a-HIV #5 Dickinson, Cat. 352096) containing 10 ml aMEM media (Thermo ated by Lenti-X concentrator, which achieved a transduction rate % GFP% GFPScientific, Cat. SV30010) and centrifuged for 5 minutes at 1600 rpmDilution of 7.61% in cord blood-derived CD34- cells. Dilution expressing expressing (Allergra X-22R Centrifuge, Beckman Coulter). The supernatant was dis-Factor These highercells titer virus IU/uL preparations will be economical cells in Factor carded and the cells were resuspended in 12 ml of aMEM media. terms of the cost 0.65 associated with0 acquiringNorare . No virus virusCD34+CD38-/lo 0.71

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Generation of more efficient lentiviral constructs with modified U1snRNA for anti-HIV-1 gene therapy

40 40!

36.9 36.9!

% Cells Transduced Transduced! % Cells

35 35! 30 30!

26.9! 26.9

25 25! 20! 20

15 15!

15! 15 10! 10

9.42 9.42!

5! 5

0 0!

00!

1.9! 1.9 12.5! 12.5

2.5 2.5! 25 25!

6.26 6.26! 50! 50

7.61 7.61!

100! 100

Volume virus supernatant Volume virus supernatant (uL) (uL)! "Lenti X Virus” Virus"! “Lenti X "Ultraﬁltration media Virus” Virus"! “Ultrafiltration ++ Freestyle Freestyle media

Figure 5: Comparison of transduction efficiency of Lenti-X U1 a-HIV-1 #5 virus and Ultrafiltration + Freestyle media U1 a-HIV-1 #5 virus in CB-derived CD34- cells. Flow cytometric analysis used to measure GFP (FITC) expression, the reporter for the U1 a-HIV-1 #5 construct. 1.5x105 CB 34- cells per well were cultured at 37°C for 48 hours with lentiviral supernatant. One well received no virus as a negative control.

They are also advantageous in reducing the time associated with obtaining a sufficient number of cells for HSC transplantation. This work will facilitate further research required to compare the suppression of HIV protein synthesis between the higher and lower titer virus particles. Higher titer virus preparations will also facilitate the further study of the off target effects of transplantation of HSCs modified by anti-HIV U1snRNAs. Studying the off target effects of transplanting modified HSCs is essential when developing a gene therapy intended for human use. As shown in Table 1, there was no significant difference in the titer of the a-HIV #1 construct and the a-HIV #5 construct, both produced by Ultrafiltration. This illustrates that the protocols described here should be applicable to any lentiviral particles regardless of the constructs packaged inside the vectors. These protocols can now be used to generate higher titer virus using a variety of lentiviral particles. It should be possible to obtain a sufficient number of HSCs for transplantation in less time and at less cost.

Conclusion

Both the Ultrafiltration and the Ultrafiltration + Freestyle media protocols achieved titers four to five fold higher than titers obtained with the Lenti-X protocol. The Ultrafiltration + Freestyle media protocol is an effective method to produce lentiviral vectors with titers sufficient to transduce 37% of cord-blood derived CD34cells. These protocols should facilitate the generation of higher titer lentivirus regardless of the constructs packaged inside them.

Acknowledgements

I would like to thank Gisele Knowles at the Centre for Flow Cytometry & Microscopy for the training and assistance throughout my project. I would like to thank my supervisor, Maria Malm, PhD. for her guidance throughout the summer as well as all of the members of the Zúñiga-Pflücker lab for their encouragement. Finally, I would like to sincerely thank Juan Carlos ZúñigaPflücker, PhD for the opportunity to work in his lab.

Research Articles

Table 1: Viral titer for each method of lentiviral production tested. Lenti-X (Control) U1 a-HIV #1 % GFPDilution expressing Factor cells No virus 0 10 97.6 20 89.6 40 75.6 80 46.3 160 23.1 320 10.8 640 6.69 1280 3.48 2560 2.19 5120 1.33 10240 0.98 Average Ultrafiltration U1 a-HIV #1 % GFPDilution expressing Factor cells No virus 0.65 10 98.7 20 98.8 40 95.9 80 85.6 160 66.4 320 47.3 640 28.8 1280 17.6 2560 10.4 5120 6.24 10240 4.55 Average

Lenti-X U1 a-HIV #1

% GFPexpressing cells 0.1 94 81.5 50.1 30.8 18.1 5.13 3.65 1.57 1.51 0.96 1.28

IU/uL 0 2.9 x 107 5.3 x 107 9.0 x 107 1.1 x 108 1.1 x 108 1.0 x 108 1.3 x 108 1.3 x 108 1.7 x 108 2.0 x 108 3.0 x 108 1.4 x 108

Dilution Factor No virus 10 20 40 80 160 320 640 1280 2560 5120 10240 Average

IU/uL 0 2.9 x 107 5.9 x 107 1.1 x 108 2.0 x 108 3.0 x 108 4.2 x 108 5.2 x 108 6.7 x 108 8.0 x 108 9.6 x 108 1.3 x 109 5.4 x 108

Ultra +Freestyle media U1 a-HIV #5 % GFPDilution expressing Factor cells No virus 0.71 10 98.1 20 91.7 40 75.2 80 44 160 31.5 320 26.5 640 15.3 1280 15.7 2560 10.4 5120 6.14 10240 6.1 Average

IU/uL 0 2.8 x 107 4.9 x 107 6.0 x 107 7.4 x 107 8.6 x 107 4.9 x 107 7.0 x 107 6.0 x 107 1.1 x 108 1.5 x 108 3.9 x 108 1.1 x 108

IU/uL 0 2.9 x 107 5.5 x 107 9 x 108 1 x 108 1.5 x 108 2.5 x 108 2.9 x 108 6.0 x 108 7.9 x 108 9.4 x 108 1.8 x 109 4.6 x 108

References

1. Douek DC, Picker LJ, Koup RA. T cell dynamics in HIV-1 infection. Annu Rev Immunol 2003;21:265-304. 2. Sajic R, Lee K, Asai K, Sakac D, Branch DR, Upton C, et al. Use of modified U1 snRNAs to inhibit HIV-1 replication. Nucleic Acids Res 2007;35(1):247-255. 3. UNAIDS. Global report: UNAIDS report on the global AIDS epidemic 2010. 2011; Available at: http://www.unaids.org/documents/20101123_GlobalReport_em.pdf. Accessed 10/10, 2012. 4. Yan N, Lieberman J. Gaining a foothold: how HIV avoids innate immune recognition. Curr Opin Immunol 2011 Feb;23(1):21-28. 5. Margolis L, Shattock R. Selective transmission of CCR5-utilizing HIV-1: the ‘gatekeeper’ problem resolved? Nat Rev Microbiol 2006 Apr;4(4):312-317. 6. Sleasman JW, Goodenow MM. 13. HIV-1 infection. J Allergy Clin Immunol 2003 Feb;111(2 Suppl):S582-92. 7. McMichael AJ, Rowland-Jones SL. Cellular immune responses to HIV. Nature 2001 Apr 19;410(6831):980-987. 8. Mogensen TH, Melchjorsen J, Larsen CS, Paludan SR. Innate immune recognition and activation during HIV infection. Retrovirology 2010 Jun 22;7:54. 9. Grossman Z, Meier-Schellersheim M, Paul WE, Picker LJ. Pathogenesis of HIV infection: what the virus spares is as important as what it destroys. Nat Med 2006 Mar;12(3):289-295. 10. Kitchen SG, Zack JA. Stem cell-based approaches to treating HIV infection. Curr Opin HIV AIDS 2011 Jan;6(1):68-73. 11. Abdool Karim SS, Naidoo K, Grobler A, Padayatchi N, Baxter C, Gray A, et al. Timing of initiation of antiretroviral drugs during tuberculosis therapy. N Engl J Med 2010 Feb 25;362(8):697-706. 12. Sajic R. Use of Modified UqsnRNAs to Inhibit HIV-1 Replication. Toronto, Canada: Department of Molecular Genetics, University of Toronto; 2011

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

RESEARCH

An ab initio computational study of a model decarboxylated schiff base of asparagine as a precursor in carcinogenic acrylamide formation Frank Y. Gao1*, Amro Dodin1*, Harun Mustafa1, Natalie J. Galant1,2, Imre G. Csizmadia1 Department of Chemistry, University of Toronto, Toronto, Ontario, Canada, M5S 3H6 Department of Chemical Informatics, University of Szeged, H-6725 Szeged, Hungary * Authors contributed equally to content 1 2

Abstract

Ab initio computations at the B3LYP/6-31G(d,p) level of theory have been performed on a model decarboxylated Schiff base of asparagine. Potential Energy Surfaces (PES) scans across two dihedrals, ψ and ϕ, of the endo and exo isomers of a sample molecule were performed. Stabilization is seen in the [s, s] conformer of the exo isomer due to hydrogen bonding between opposing carbonyl and amine groups, whereas the endo isomer sees rapid destabilization in the same conformation due to steric strain between the carbonyl oxygen and the terminal methylene group. However, there was no notable free energy difference between the minimum energy conformers of the endo and exo isomers, although the exo isomer was slightly energetically favored. The work presented here lays the groundwork for further studies into decarboxylated Schiff Bases from reactions with various sugars. This in turn allows for further investigation of mechanisms of formation, providing a basis for comparison of proposed pathways. *Please note that [a, b] represents a specific conformation based on values of the dihedral angles ψ and ϕ. The letter s stands for syn (dihedral between 0° and ± 30°), a for anti (between ± 150° and 180°), g+ for gauche-plus (between 30° and 90°) and g- for gauche-minus (between -30° and -90°)

Introduction

Acrylamide is a carcinogen that has been proposed to be positively associated with renal [1], oral [2], and esophageal [3] cancers in recent years. The identification of acrylamide in many cooked starch-rich foods by Swedish health authorities in 2002 [4] caused significant public health concerns and led to a series of studies investigating its formation in these systems. Several studies using isotope labeled compounds determined that the acrylamide backbone originates from free asparagine [5, 6]. Furthermore, it was determined that the formation of acrylamide from the free amino acid proceeds through the action of reducing groups in the Maillard Reaction [5–7]. This process could theoretically occur through the thermal decarboxylation of free asparagine in isolation, but in reality, was shown to give very low yields due to a competing intramolecular cyclization reaction [8]. The Maillard Reaction, often referred to as ‘non-enzymatic browning’ [9], is a complex network of reactions that occurs during the heating of food. It is responsible for giving cooked foods their pleasant odour and flavour. In addition, it has been shown that the Maillard reaction also occurs in living organisms [10, 11], thereby raising concerns that the acrylamide may form in vivo through the same mechanism. Several investigations of the acrylamide formation mechanism

of in these systems have been conducted, proposing multiple possible pathways and intermediates [12]. However, the mechanisms proposed by Zyzak et al [6] and Yaylayan et al [8] are considered the best supported [13]. Interestingly, many of these studies agree on the initial steps of the mechanism, starting with the condensation of asparagine with a reducing sugar, followed by the dehydration of the resulting intermediate to give a Schiff base. The Schiff base is then decarboxylated to give a decarboxylated Schiff base of asparagine, also referred to as an azomenthine ylide [12]. This ylide acts as the central intermediate in these mechanisms, providing the last common point before the proposed pathways diverge. Mechanistically, the key difference between the proposals lies in the treatment of the decarboxylated Schiff Base. Yaylayan et al suggest that the Schiff base is decarboxylated before undergoing an Amadori rearrangement reaction to form a decarboxylated Amadori product. A final elimination reaction then yields acrylamide. In contrast, Zyzak et al suggest two alternative pathways for acrylamide formation. The first entails the hydrolysis of the decarboxylated Schiff base to form 3-aminopropionamide, with a subsequent elimination of ammonia to form acrylamide. Alternatively, the decarboxylated Schiff base can form acrylamide directly through the elimination of an imine. All of these pathways highlight the significance of the nature of the reducing group in determining the intermediates and efficacy of the reaction [12].

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

An ab initio Computational Study of a Model Decarboxylated Schiff base of Asparagine as a Precursor in Carcinogenic Acrylamide Formation

O NH₂

Reducing Sugar

COOH

OH H₂N O Asparagine

H₂O

H₂N

N OH

COOH O

H₂N

COOH O

Schiff base

CO₂ N⁺

HO H₂N

N⁺ –

..–

H₂N

H NH

+ R

NH₂

OH O

Acrylamide

Decarboxylated Schiff base

Ib O

H₂O

NH₂

+ R

NH₂ + R

NH₂

NH O

OH O

+ NH₃ NH₂

NH₂

Acrylamide

H₂N O Figure 1: Formation pathways of acrylamide by processes Ia and Ib proposed by Zyzak et al [6] and process II proposed by Yaylayan et al [8].

In addition to their significance in the formation of acryl- Methods and Materials amide in heated starch-rich systems, decarboxylated Schiff bases Through the use of ab initio quantum computations, we have conare present in an analogous reaction which has been described ducted an in silica study on a model (Figure 2) of the decarboxylated Schiff for all α amino acids and for some organic amines as early as base of asparagine in order to elucidate its properties and the geometries of 1968 [14]. This provides further understanding on the role of the the thermally preferred conformers of both its endo and exo isomers. The decarboxylated Schiff bases. decarboxylated base is the intermediate resulting from the reaction of free For this reaction, there is a lack of experimental evidence to asparagine with formaldehyde (Figure 2). differentiate between the relative contribution of the mechanisms A range of conformers of the endo and exo isomers of the decarboxylproposed by Zyzak et al [6] and Yaylayan et al [8, 13] to the over- ated Schiff base of asparagine with formaldehyde was generated by varying all reaction, although evidence has been provided that the nature the central dihedrals, ψ (O1C2-C3C4) and φ (C2C3-C4N5), by 30° increments of the reducing group has a strong influence on this [12]. Hence, from -180° to +180°. Each conformer of the model was structurally opan improved understanding of the thermodynamic differences timized and then subjected to thermodynamic analysis. Computations between the decarboxylated Schiff bases that result from reactions were performed using the Gaussian09 A.1 (G09) software package [15]. with common reducing sugars would assist in understanding the Structural optimization and thermochemical analysis was performed by mechanism through which acrylamide formation is promoted or DFT methods [16], using the Becke, three parameter, Lee, Yang, and Parr inhibited by different reducing groups. This can be used to elucidate (B3LYP) [17–19] method at the 6-31+G(d,p) level [20, 21]. Heats and the conditions that the mechanisms theoretically O₁ O₁ H₁₂ H₁₂ H₁₅ prefer. These straight chain substituted decarboxylated Schiff bases are fairly complex with many H₈ C₂ C₄ H₁₃ C₂ C₄ C₆ dihedral angles and hydrogen-bonding groups. H₈ C₃ C₃ H₁₄ N₅ N₅ N₇ N₇ As such, knowledge of their common backbone is ₊ ₊ φ φψ ψ essential to their investigation. This study aims to H₁₀ H₁₁ analyze this key segment through molecular comH₉ H₁₀ H₁₁ C₆ H₉ H₁₃ putations of the simple model described above in endo H₁₅ exo H₁₄ order to provide a foundation for further study of Figure 2: Dihedral angles (φ,ψ) of the endo and exo isomers in the [s, a] conformer of the its real-world analogue. model decarboxylated Schiff base of asparagine and formaldehyde. Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

An ab initio Computational Study of a Model Decarboxylated Schiff base of Asparagine as a Precursor in Carcinogenic Acrylamide Formation

Table 1: The thermodynamic functional changes of Gibbs free energy (ΔG) (kJ/mol) in the decarboxylated Schiff base of asparagine and formaldehyde. Conformers [ψ, φ] [a, a] [a, s] [a, g+] [s, a] [s, s] [s, g+] [g+, a] [g+, s] [g+, g+] [g+, g-]

exo isomer (kJ/mol) 33.27 56.65 8.38 23.57 0.00 19.03 33.82 16.32 32.75 5.03

endo isomer (kJ/mol) 21.13 63.63 5.76 0.375 57.97 20.30 15.13 44.67 17.48 20.87

entropies for each conformer were given in Hartrees per mole and calories per mole kelvin, respectively. Data from the partially optimized conformations (with the central dihedrals frozen) was used to plot one dimensional cross-sections of the semi-rigid scan of the PES, while results from the fully optimized run were used to locate the lowest energy conformer for both the endo and isomer. The thermodynamic functional changes of thermally corrected energy (ΔU), enthalpy (ΔH), Gibbs free energy (ΔG), and entropy (ΔS), were calculated from the computed U, H, G, and S.

Results and Discussion

The lowest energy conformers for the endo and exo isomers were found to be the [s, a] and [s, s] conformers, respectively (Figure 3). The PES for the exo isomer shows a large depression surrounding the [s, s] conformer, the global minimum, (Figure 4) with local minima present at distorted [a, g+] and [a, g- ] conformer, with a relative free energy of 8.38 kJ mol-1. The least stable conformer of the exo isomer was the [a, s] conformer. In contrast, the PES for the endo isomer shows the greatest stabilization surrounding the [s, a] conformer, with local minima at a distorted [a, g+] and distorted [a, g-] conformer, with a relative free energy of 5.76 kJ mol-1. Most notably, the endo isomer contains a broad region of instability when φ is in the syn conformation, a region which includes the global energy maximum of the [a, s] conformation and a local energy maximum at the [s, s] conformation (Figure 4). Both PESs were symmetrical about the origin as a consequence of the internal mirror plane of the molecule. The free energy difference between the lowest energy conformers of the endo and exo isomers was calculated to be 0.375 kJ mol-1 (Table 1, Figure 5).

Figure 3: Minimum Energy Conformers of the exo: [s, s] and endo: [s, a] isomers of the Decarboxylated Schiff Base of asparagine and formaldehyde.

The minimum energy conformers of the two isomers provide an interesting contrast between the structures and intramolecular forces at play in the pair of isomers. The exo isomer’s stability over the endo isomer is likely the result of hydrogen bonding between O1 and H13 in the exo isomer’s minimum energy conformer. The same hydrogen bonding is impossible in the endo isomer as N5 will block any approach between O1 and H1. It is also important to note that the N5-C6 bond length is significantly longer in the two lowest energy conformations of the endo isomer ([s, a] and [a, g+]) than in the two lowest energy conformations of the exo isomer ([s, s] and [g+, g-]). In fact, the N5-C6 bond length for the endo isomer is greater than that of the exo isomer by more than 0.1 Å. This suggests that the endo isomer will more often adopt conformations with longer and thus weaker N5-C6 bonds. This may create a notable difference in the pathways taken by different isomers, particularly since mechanism Ib depends on the hydrolysis of this bond (Figure 1, Table 2). Another interesting observation is the presence of a structure similar to a six-membered ring in the exo. This similarity perhaps assists in the stabilization through the hydrogen bonding between the hydrogen of the double-bonded nitrogen (N5) and the carbonyl oxygen (O1) (Figure 2). This property may facilitate the exchange of protons from the amine to the oxygen of the carbonyl compound during pH changes. As such, this may lead to significant structural dependence on pH. In addition, this property is most likely the cause of the depression in the PES of the exo isomer surrounding the global minimum. The notable lack of proximity between hydrogen bonding groups in the endo isomer, however, makes this structural pH-dependence phenomenon unlikely.

Conformers [ψ, φ] [a, a] [a, s] [a, g+] [s, a] [s, s] [s, g+] [g+, a] [g+, s] [g+, g+] [g+, g-]

C4-N5 Bond Length (Å) exo endo 1.3374 1.3385 1.3388 1.3334 1.3429 1.3480 1.3347 1.3329 1.3384 1.3457 1.3418 1.3380 1.3377 1.3356 1.3376 1.3429 1.3385 1.3386 1.3425 1.3424

N5-C6 Bond Length (Å) exo endo 1.3381 1.3392 1.3403 1.3574 1.3429 1.3342 1.3400 1.3431 1.3312 1.3345 1.3334 1.3383 1.3389 1.3413 1.3351 1.3379 1.3364 1.3385 1.3306 1.3358

ΔG (kJ/mol)

Table 2: C4-N5 and N5-C6 bond lengths for the conformers of the exo and endo isomers of the decarboxylated Schiff base of asparagine and formaldehyde as computed at the B3LYP level of theory with the 6–31G(d,p) basis set.

Figure 4: Single variable potential energy curves of the endo and exo isomers about their minima, [s, s] and [s, a] respectively.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

An ab initio Computational Study of a Model Decarboxylated Schiff base of Asparagine as a Precursor in Carcinogenic Acrylamide Formation

Research Articles

its endo and exo isomers. Information was obtained for the relative stability of various conformers and scans of PES were performed. The minimum energy conformer for the endo and exo isomers were the [s, s] and [s, a] conformers, respectively. The [s, a] exo conformer was found to be the most stable of the two isomers, however there was little energetic difference between them. Additionally, the N5C6 bond in the secondary amine group of the preferred conformer of the endo isomer was found to be longer than its counterpart in the preferred conformer of the exo isomer, which may suggest that the different isomers take different pathways. Furthermore, it was proposed that hydrogen-bonding may be significant in influencing the optimized structure and its stability.

Acknowledgments

The authors would like to thank David H. Setiadi from the Global Institute of Molecular and Materials Science (www.giocomms.org) for his assistance. In addition, they would also like to extend their thanks to the Research Opportunity Program at the University of Toronto for providing the opportunity to conduct this research.

References

Figure 5: Relative free Energies of the conformers of the exo and endo isomers of the decarboxylated Schiff base of asparagine and formaldehyde.

Furthermore, the very small difference in the stability of the endo and exo isomers would suggest that there is very little thermodynamic preference for either isomer. Therefore, both forms must be considered in modeling and developing a mechanism for the reaction, as the variations in their properties may result in differences in the reactivity of the two isomers in the formation of acrylamide. As for the relevance of the model base conformation studied here to the backbone of biologically relevant decarboxylated bases, differences will be inevitable, especially due to the presence of hydrogen bonding groups on the R group. Nonetheless, the results of our computations will establish a starting point for studies of these systems. The use of the B3LYP basis set allows for effective determination of structure without the introduction of experimental and systematic errors that abound in the investigation of the complex matrices of food samples, hence making computational investigation a highly effective tool in studying these systems. Although the results of this study are not directly comparable to current experimental data, they provide a valuable basis from which further computations can be conducted to predict the properties and behavior of real world systems. It is hoped that these efforts will provide key insights into the mechanism of acrylamide formation and assist future efforts in mitigating such a process.

Conclusion

Molecular computations, including conformational and thermodynamic functional analysis, were performed at the B3LYP level of theory with a basis set of 6-31G+(d,p) on a model of the decarboxylated Schiff Base from asparagine and formaldehyde in both

1. Hogervost JG, Schouten LJ, Konings EJ, Goldbohm RA, ven der Brandt, P. A. Dietary acrylamide intake and the risk of renall cell, bladder and prostate cancer. Am J Clin Nutr 2008;87:1428-1438. 2. Schouten LJ, Hogervost JG, Konings EJ, Goldbohm RA, van den Brandt, P. A. Dietary acrylamide intake and the risk of head-neck and thyroid cancers: Results from the Netherlands cohort study. Am J Epidemiol 2009;170:873-884. 3. Lin YL, Lagergren J, Lu YX. Dietary acrylamide intake and risk of esophageal cancer in a population-based case-control study in Sweden. Int J Cancer 2011;128:676-681. 4. Tareke E, Rydberg P, Karlsson P, Erickson S, Tornqvist M. Analysis of acrylamide, a carcinogen formed in heated foodstuffs. J Agric Food Chem 2002;50:4998-5006. 5. Stadler RH, Blank I, Varga N, Robert F, Hau J, Guy PA, et al. Acrylamide from Maillard reaction products. Nature 2002;419(6906):449-450. 6. Zyzak DV, Sanders RA, Stojanovic M, Tallmadge DH, Eberhart BL, Ewald DK, et al. Acrylamide formation mechanism in heated foods. J Agric Food Chem 2003;51(16):4782-4787. 7. Mottram DS, Wedzicha BL, Dodson AT. Food chemistry: acrylamide is formed in the Maillard reaction. Nature 2002;419(6906):448-449. 8. Yaylayan VA, Wnorowski A, Locas CP. Why asparagine needs carbohydrates to generate acrylamide. J Agric Food Chem 2003;51(6):1753-1757. 9. Hodge JE. Dehydrated foods, chemistry of browning reactions in model systems. J Agric Food Chem 1953;1(15):928-943. 10. Frye EB, Degenhardt TP, Thorpe SR, Baynes JW. Role of the Maillard reaction in aging of tissue proteins. J Biol Chem 1998;273(14):18714-18719. 11. Rérat A, Calmes R, Vaissade P, Finot PA. Nutritional and metabolic consequences of the early Maillard reaction of heat treated milk in the pig. Eur J Nutr 2002;41(1):1-11. 12. Lineback DR, Coughlin JR, Stadler RH. Acrylamide in Foods: A Review of the Science and Future Considerations. Annual Review of Food Science and Technology 2012;3:15-35. 13. Stadler RH, Scholz G. Acrylamide: an update on current knowledge in analysis, levels in food, mechanisms of formation, and potential strategies of control. Nutr Rev 2008;62(12):449-467. 14. Al-Sayyab A, Lawson A. Schiff bases. Part I. Thermal decarboxylation of α-amino-acids in the presence of ketones. J.Chem.Soc.C 1968:406-410. 15. Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, et al. Gaussian 09, Revision A.1. Gaussian, Inc., Wallingford CT, 2009. 16. Hohenberg P, Kohn W. Inhomogeneous electron gas. Physical Review 1964;136(3B):B864. (17) Becke AD. Density‐functional thermochemistry. III. The role of exact exchange. J Chem Phys 1993;98:5648. (18) Lee C, Yang W, Parr RG. Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Physical Review B 1988;37(2):785. (19) Miehlich B, Savin A, Stoll H, Preuss H. Results obtained with the correlation energy density functionals of Becke and Lee, Yang and Parr. Chemical Physics Letters 1989;157(3):200-206. (20) Hehre WJ, Ditchfield R, Pople JA. Self—consistent molecular orbital methods. XII. Further extensions of gaussian—type basis sets for use in molecular orbital studies of organic molecules. J Chem Phys 1972;56:2257. (21) Hariharan PC, Pople JA. The influence of polarization functions on molecular orbital hydrogenation energies. Theoretical Chemistry Accounts: Theory, Computation, and Modeling (Theoretica Chimica Acta) 1973;28(3):213-222.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

RESEARCH

Sex Comb rotation in Drosophila melanogaster: Changes in tissue length and cell extrusion Tahsin Khan, Larry Zhang, Juan Nicolas Malagon Department of Cell and Molecular Biology, University of Toronto,

Abstract

The remodeling of epithelial tissues is an important process during development. One model of such morphogenesis is the process of sex comb arrangement and rotation in the fruit fly Drosophila melanogaster. Unlike the simpler homogeneous-sheet epithelial systems studied to date, this process involves the reorientation of sex combs, rows of leg bristles, by an approximately 90° rotation along epithelial cells during pupal development. Epithelial changes in length influence the resulting position of sex combs. This study investigated the connection between sex bristles and morphological response in epithelial tissue. Genetic perturbations were introduced to produce sex combs of varying lengths and subsequent effects on epithelial tissue was monitored across developmental stages. Our results indicate that tissue above sex combs always exhibits an increase in cell number between landmark bristles independent of sex comb length. However, spatial and temporal patterns of epithelial cell extrusion dramatically change based on changes in sex comb length. Our study provides key early findings that implicate sex comb rotation variations as a variable affecting the process of epithelial development in D.melanogaster.

Introduction

Tissue elongation is an essential developmental process that occurs in different biological systems. Multiple examples of tissue elongation have been studied including the egg chamber, germ band, legs, wings and trachea [1-4]. These works have mainly focused on the early stages of development and undisturbed epithelia where cell neighbours interchange in a relatively uniform manner. However, cases of elongating epithelial systems at late developmental stages with more complex cellular processes remain poorly understood. In order to study more complex cases of tissue elongation, we have chosen to examine sex combs which are specific to males and indicative of late development in the fly system. In males, there is a row of developing leg bristles that have been shown to act as a barrier to epithelial cell rearrangement. Sex comb rotation is a mid-pupal stage process (23 hours to 36 hours after pupation), which involves heterogeneous types of cellular rearrangements and multiple cellular processes (Figure 1a) [5]. In addition, sex comb rotation displays another novelty in the study of tissue elongation, the presence of bristles embedded in an epithelial sheet. Sex combs are a sexually dimorphic row of bristles located in the first leg segment of D. melanogaster males [6]. During development, this group of bristles rotate from a horizontal position to vertical position of up to 90o, while the surrounding epithelial cells undergo dramatic changes in apical size. However, the cellular processes underlying sex comb rotation remain unknown. The sex comb is a rapidly evolving male-specific trait that is used to grasp the females’ abdomen and genitalia and spread their wings prior to copulation.

In order to determine the cellular mechanisms underlying sex comb rotation, previous studies have used artificial selection and transgenic lines to perturb aspects of the rotation [5]. For example, the normal number of sex comb teeth (9-11 bristles) can be increased (12-13 bristles) or reduced (5-6 bristles). As a result of these preliminary studies, the cellular dynamics of the rotating sex comb seem to display an interesting paradox—the emergence of similar changes at the tissue level based on different spatial and temporal cell behaviors (Figure 1.b-c). In this work, we studied how modifying the number of sex comb teeth would affect tissue elongation. Our results show that sex comb of different lengths produce similar changes in neighbouring apical tissue shape during rotation (Figure 1b). On the contrary, the spatial and temporal patterns of cell extrusion differ dramatically between samples (Figure 1c). This data raises the question whether the development of the sex comb is a representative system to study self-organization in epithelia.

Materials & Methods

Fly Stocks Mutants and artificial selection were used to change the number of sex comb teeth as indicated in Table 1. The artificially selected lines for high and low number of sex comb teeth were developed by Ahuja and Singh [9], while the mutant babPR72 was generated by Godt et al. [10].

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Sex Comb Rotation in Drosophila Melanogaster: Changes in Epithelial Length and Extrusion

Figure 1: Cellular events during sex comb rotation. a) Diagram of sex comb rotation and cellular processes involved. Rectangular boxes display the location of the cellular process in the area surrounding the sex comb. Sex combs rotate from posterior to anterior region of the first tarsal segment, during which several cellular processes occur in the surrounding epithelium. b) Changes in apical tissue shape during rotation. Hexagons in b) and c) represent epithelial cells surrounding the sex comb teeth. In b), red hexagons represent cells in the distal region, while blue hexagons represent cells in proximal region. In c), color-coded hexagons represent a standard technique to study how cells re-organize in a tissue during development. In the initial stages, horizontal lines of cells are labeled and followed through time. If lines are conserved, it indicates that cells share the same cell neighbors during development. Mixed color lines indicate that tissues interchange cell neighbors during development. At the tissue level, similar changes occur between samples; for example, both proximal (blue hexagons in b) and distal regions (red hexagons in b) display tissue elongation. c) In contrast, at the cellular level, cellular processes display a high variation between samples; for example, cells interchange neighbors showing different patterns between samples. In the initial stage, cells in the horizontal line exchange neighbors and display different color combinations in the final stage, (Modified from Malagon 2013). Table 1: Fly stocks studied Lines Wild type female Wild type male Low line male High line male bric à brac [PR72] (bab[PR72]) male

Genetic Background

Artificial selection Artificial selection Mutation (Ectopic sex comb in 2nd tarsal segment)

Number of sex comb teeth 0 9-10 5-6 12-13

5-7

Generation of Fusion Constructs To visualize cells, the ubi-DEcad:GFP construct generated by Oda and Tsukita was used. Ubi-DEcad is a membrane protein found on the surface of epithelial cells [13]. UbiDEcad:GFP con¬structs were transfected into cell lines with high and low bristle teeth expression (high and low cell lines generated by colleagues Ahuja and Malagon) [5,7]. Standard genetic crosses were performed in order to introduce the UbiDEcadherin:GFP construct into the mutant babPR72 [5]. Following the protocol described by Atallah [6], four-dimensional live imaging was conducted for the different treatments studied: wild type male and female, the mutant babPR72, and high and low lines.

Research Articles

Figure 2: Cell extrusion during sex comb rotation. Sex comb is indicated by yellow bracket. The location of landmark bristles is shown by the orange circles. Three bristles were chosen: the most distal sex comb tooth (A), sensillum campaniform (B), closest chemosensory bristle to the sex comb (C) Extruding cells are numbered and outlined in yellow. Purple lines represent boundaries in the region proximal to the sex comb.

Quantification of Cells Between Landmark Bristles Quantification of the number of cells between vertical and horizontal landmark bristles was performed as indicated in Figure 2. The bristles chosen as landmarks were the same as those described by Wang [8]. Except for the high-line, representative data from the wild type, mutant and artificially selected lines was taken from lines express¬ing a low number of sex comb teeth. For comparison, data for female D. melanogaster were used from Wang [8]. Changes in the number of epithelial cells were assumed to correspond to changes in tissue length. Monitoring Cell Extrusion Images of the high line D. melanogaster were obtained and cells were manually labelled using Image J software (Figure 2) [2, 3] (Software can be found at : http://rsbweb.nih.gov/ij/). Cells were analyzed from 23 to 36 hours after pupariation (AP). Cells that disappear during this time period were outlined in yellow. The sequences in which cells disappear in a given movie were also labelled with the corresponding numbers. Also in Figure 2, Each D. melanogaster leg was divided into 4 sections in order to determine the region of cell extrusion for each cell. Statistics Errors bars represent standard deviation. Comparison of number of cell extrusion were made using ANOVA, while post hoc comparisons were performed using tukey’s test.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Sex Comb Rotation in Drosophila Melanogaster: Changes in Epithelial Length and Extrusion

Figure 3: Comparison of narrowing and elongation of tissues between rotating and non-rotating sex combs. (a) Elongation of tissue in rotating sex comb. (b) The region of tissue elongation measured in number of cells. (c) Elongation in non-rotating rows of bristles. (d) Narrowing of tissue in rotating sex comb. (e) The region of tissue narrowing measured in number of cells. The red arrows indicate the region that was measured by counting the number of cells, which resulted in an observation that the tissue in that region narrowed with time from 23h to 36h AP. (f) Narrowing of tissue in non-rotating rows of bristles. The sex of the wild type is male. Error bars represent one standard deviation about the mean. Rotating sex combs show narrowing and while in non-rotating regions, the number of cells remain the same or only slightly change. Data for female D. melanogaster was taken from Wang (7).

Journal of Undergraduate Life Sciences â&#x20AC;˘ Volume 7 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Spring 2013

Sex Comb Rotation in Drosophila Melanogaster: Changes in Epithelial Length and Extrusion

Figure 4: Comparison of spatial pattern of cell extrusion between wild type and high line. A) Number of cell extrusion in the wild type and High line. In the wild type, the majority of the cells extrude from the epithelium in the top part of the proximal region. In contrast, in the high line the number of cells extruding greatly varies between samples. The vertical bars represent one standard deviation about the mean. Brackets indicate statistical differences (Tukey test) (***<0.001, **<0.05). B) Schematic of the region proximal to sex comb. The proximal region was further divided in three regions: region above the sex comb (green polygon), region 1 (blue polygon) and region 2 (red polygon). The region proximal to the sex comb is divided into top (region 1) and bottom (region 2). We divided because we wanted to see whether the cell extrusion take place homogeneous in the sex comb region. Above the sex comb, region close to the most distal transverse row. The stars represent cells and the blue arrows indicate sex comb rotation. The black arrows indicate cells shifting upwards due to sex comb rotation, which occasionally leads to extrusion of a subset of these cells.

Results

Epithelial cell numbers significantly change only in the case of sex comb rotation As previously shown by Atallah et al and illustrated in Figure 1b, wild type sex combs result in a change in the apical tissue shape between vertical and horizontal landmark bristles. To test whether this change in tissue shape and size is dependent on sex comb length, we studied the cell dynamics of artifically selected lines for a higher and lower number of sex comb teeth relative to the wild type. Four-dimensinonal live imaging showed that that only rotating sex combs (wild type, high line, and low line) display a reduction in number in cells between horizontal landmark bristles (horizontal tissue shortening) and an increase in number of cells between vertical landmark bristles (vertical tissue lengthening).

Research Articles

Figure 5: Non predictable sequence of extrusion events: Sequence of individual cell extrusion events as indicated by the numbers in the high line. The yellow polygons outline the cells that extrude within 23h to 36h AP. The blue arrows highlight the sequence of cell extrusion events, illustrating the lack of any clearly noticeable patterns.

The results are summarized in Figure 3 a,b,d. On the other hand, the non-rotating row of bristles (females and the ectopic sex comb in the mutant babPR72), the number of cells between landmark bristles changed to a much lesser extent (Figure 3 c,e,f). Vertical error bars represent +/-1 standard deviation about the mean. Alterations in sex comb length result in more variability in the time, position, and frequency of cell extrusion during sex comb rotation Malagon [6] observed that proximal to the sex comb, cells extrude from the epithelium during the rotation. Approximately 10 cells extrude in tissue section with most of them in the top section of the proximal region (Figure 4). To test whether this process is independent of the sex comb length and whether the number of cell extruding and spatial pattern is conserved, we studied this cellular process in the high line. We focused on both the spatial and temporal aspects of the cellular dynamics in this system. Time lapse movies from 23 to 36 AP showed that the high line also displays cell extrusion during rotation. However, unlike wild type, the cell extrusion in the high line had an increased variation in the number of cells extruded (data not shown). The overall observed range of cell extrusion in the high line was 5-31 cells. The initial location of epithelial cells is not a determinant of their potential for extrusion To test whether cell extrusion in the high line displays an organized spatial pattern, the sequence and location of cell extrusion events was tracked during rotation. Live imaging showed that the

Journal of Undergraduate Life Sciences â&#x20AC;˘ Volume 7 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Spring 2013

Research Articles

Sex Comb Rotation in Drosophila Melanogaster: Changes in Epithelial Length and Extrusion

extrusion potential of a certain cell was not dependent on its initial location within the epithelial sheet (Figure 5). In addition, although cells in the high line demonstrated extrusion only towards the proximal region (Figure 4A and B), the variation of extrusion within this region is higher in the high line than in the wild type. Cell extrusion is enhanced during the slow stage of sex comb rotation To test whether the high line displays a temporal pattern of cell extrusion, time lapse movies were analyzed. Figure 6 illustrates that each movie seems to have highly variable temporal patterns of cell extrusion. In some cases, four cells can extrude from the epithelium in one hour (movie 2 and 4), while in other movies only one cell extrudes every 2 hours (movies 3 and 5). Nonetheless, it is important to note that previous studies (6) have showed that based on the degree of rotation, the change in position can be divided into two stages. A rapid rotation of the sex comb occurs between 23-28 hours AP, whereas a slow rotation takes place between 28-36 hours AP. Confocal images illustrate that although the cell extrusion is three times higher on average in the slow stage of rotation than in the rapid stage of rotation (Figure 7). These values have a high margin of error in our study.

Discussion

Changes in apical tissue shape between rotating sex combs Rotating sex combs (high line, low line and wild type) always tend to display narrowing and elongation of the proximal region, linking these morphological processes to sex comb rotation. This result is consistent with a lower degree of narrowing and elongation in non-rotating region seen in females and the ectopic sex comb of the mutant babPR72. Although the number of cells changes be¬tween the regions studied this result is insufficient to make definite conclusions. Future studies should examine changes in individual apical cell area and shape. We recognize that the narrowing and elongation in number of cells could have been due to the methods used to measure them.

Figure 6. Variable pattern of cell extrusion in the high line: A comparison of temporal patterns of cell extrusion between high line movies. Each line represents the number of cells that extrude from the epithelium with respect to time. Data was recorded from 23h to 36h AP.

Since the tissue lengths were measured in number of cells, the author acknowledges that this didn’t take into account cell size heterogeneity. Therefore, it is possible for there to be the same number of cells but a different tissue length, due to varying cell diameter, which would not have been accounted for by the methods used. A more accurate method would have been to quantify the distance between landmark bristles rather than only counting the number of cells. Cell extrusion in D. melanogaster high line In the high line, the cells tended to only extrude towards the proximal region. This general pattern was observed in other rotat¬ing sex combs as well (Low line and wild type). While epithelial cell extrusion and the degree of interchange of wild type cell neighbors showed low variation between samples (Malagon, unpublished), these cell processes show a high degree variation between samples in the high line. For example: 12 cells extruded in movie 3, and 31 cells extruded in movie 1 in the high line, while in the wild type line, 0-4 cells would typically extruded at any given time among the different movies analyzed. The low degree of variation in the wild type and high degree of variation in the high line should be fur¬ther rationalized using alternative experimental approaches. First, understanding the reason for the variation in cellular interchange and extrusion despite similar end results between the various cell lines is an important phenomenon to understand. The coupling of cellular inter¬change and extrusion is also evident, and future experiments could provide a rationale for this. Does sex comb rotation show signs of self-organization? The emergence of ordered systems based on stochastic pat¬terns appear in multiple disciplines from computer science to ecology [11,12]. In the wild type flies, different cellular processes, including cell extrusion and the type and degree of cell rearrange¬ments exhibit a stochastic pattern of development [5]. In contrast, the final outcome of such cellular processes is orga-

Figure 7: Average number of cells extruding in the rapid (23h-28h AP) and slow stages (28h-36h AP) of rotation in the high line. Error bars represent one standard deviation about the mean.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Sex Comb Rotation in Drosophila Melanogaster: Changes in Epithelial Length and Extrusion

Research Articles

nized (Figure 1). This raises the possibility that the sex comb rotation system could be an example of self organization despite its seemingly sto¬chastic patterning features.

Conclusion

The findings discussed in the high line are consistent with those patterns found in wild type. In the high line, the apical shape of the tissue changes in a similar way within samples. However, at the cellular level, although cell extrusion also takes place in the same region, the sequence of extrusion events and spatial and temporal patterns of extrusion are different between the high line movies. Future studies are necessary to examine whether there is a connection between the cellular dynamics of sex comb rotation in addition to the process of self-organization and its possible implications in development.

Acknowledgements

We would like to thank Dr. Ellen Larsen for her tremendous support and constructive remarks throughout the course of this study and for providing the opportunity to undertake this research project in her lab.

References

1. Haigo, S.L. and Bilder, D. Global Tissue Revolutions in a Morphogenetic Movement Controlling Elongation. Science. 2011 Feb; 331(6020): 1071-1074. 2. Blankenship T, Backovic S, Sanny J., Weitz O, and Zallen J. Multicellular Rosette Formation Links Planar Cell Polarity to Tissue Morphogenesis. Dev Cell. 2006 Oct; 11(4): 459–470. 3. Bertet, C., Sulak, L. and Lecuit, T. Myosin-dependent Junction Re-modelling Controls Planar Cell Intercalation and Axis Elongation. Nature. 2004 Jun; 429(6992): 667-671. 4. Keller, R. Mechanisms of elongation in embryogenesis. Development. 2006 Jun; 133(12): 2291-2302. 5. Malagon et al. Sex comb in motion: cellular processes involved in the sex comb rotation in Drosophila melangoster. In preparation. 6. Atallah J. The Development and Evolution of Complex Patterns: The Drosophila Sex Comb as A Model System [PhD Thesis]. Toronto, ON: University of Toronto; 2009. 7. Wang M. Cellular dynamics involved in the narrowing of the first tarsal segment of the first leg of Drosophila melanogaster females used as a control for male sex comb rearrangements. University of Toronto JULS. 2012; 6(1): 42-46. 8. Ahuja and Malagon et al. Bridging the gaps between population genetics and evo-devo. In preparation. 9. Sivapatham G. Understanding the Cell Dynamics of Artificially Selected Sex Combs with a Low Number of Sex Comb Teeth in the Male Drosophila melanogaster. Research Opportunity Program. 2011. 10. Hung J. Effect of artificial selection on cell dynamics: Drosophila melanogaster with high number of sex comb teeth. Research Opportunity Program. 2011. 11. Oda H., Tsukita S. Real-time imaging of cell-cell adherens junctions reveals that Drosophila mesoderm invagination begins with two phases of apical constriction of cells. J Cell Sci. 2001 Feb; 114(3): 493-50. 12. Kauffman, S. The Origins of Order: Self-organization and Selection in Evolution. New York: Oxford University Press, USA; 1993. 13. Goodwin, B. How the Leopard Changed Its Spots: The Evolution of Complexity. Princeton: Princeton University Press; 2001.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

RESEARCH

Cell dynamics of Sex Comb morphogenesis in Drosophila melanogaster Yunlong Liang, Ellen W. Larsen, Juan Nicolas Malagon Department of Cell & Systems Biology, University of Toronto

Abstract

The sex comb (SC) of Drosophila melanogaster is a linear arrangement of bristles found on the basitarsus of male forelegs, and has long been considered an excellent model for studying evolutionary developmental biology. Although many of its developmental regulators have been discovered, how they are translated into cell dynamics and eventually SC morphology still remain unclear. Previous studies have demonstrated the dramatic remodeling of tarsal epithelium proximal and distal to SC during SC development displays a variety of cellular processes characteristic of systems modulated by surface mechanics. To explore the possibility of modulation of surface tension as the mechanism underlying SC ontogeny, we examined the effect of varying SC length on the cell dynamics of epithelial cells in the presumptive SC field. Confocal time series of SC morphogenesis in four lines of flies with increasingly longer SC were obtained and analyzed with ImageJ. Apical cell surface dimensions were measured and cell positions relative to landmark bristles were tracked. Our results demonstrated that changes in epithelial cell size and shape are closely correlated with the size of the SC being rotated. Furthermore, we showed that changes in cell position within the epithelium over time appeared to be random, and that cell intercalation is perhaps not actively contributing to SC rotation in contrast to what was previously believed. In summary, our findings provide evidence of association between epithelial remodeling and the size of SC. This study paves the way for future experiments in investigating the modulation of surface tension as the mechanism underlying SC morphogenesis.

Introduction

Sex comb (SC) is a linear arrangement of modified bristles found on prothoracic legs of a subclade of Drosophila species [1]. It is a sexually dimorphic trait that occurs only in males and is crucial for courtship and mating behaviours [1]. Not only does it exhibit incredible morphological diversity among species in its size, orientation, and teeth morphology, it has also been shown that even identical morphologies in closely related species can be developed by different cellular processes [2]. SC has long piqued the interest of evolutionary developmental biologist due to its potential as a model for addressing important questions such as how do developmental processes affect the evolution of traits [3]. It is an excellent model for both independent modifications of shared ancestral states as well as independent ontogeny of similar morphological structures in separate lineages [2]. These in conjunction with its rapid pace of evolution make it a powerful comparative tool for studying the genetic and developmental changes underlying convergent and divergent evolution [3]. However, before this model can be utilized widely to address the questions of evolutionary developmental biology, detailed understanding of its development must be established. This paper will focus on the cellular processes of SC formation in Drosophila melanogaster due to the large amount of knowledge already available on the model organism and the availability of well-established methods [1]. Since novel forms are frequently

produced by novel regulations of conserved mechanisms [2], our findings from this study may be widely applicable to SC development in other species as well. SC in D.melanogaster originates from the most distal Transverse Bristle Rows and has been shown to develop by male specific morphogenesis [4]. During this process, SC rotates anteriorly from its initial transverse position parallel to the anterior-posterior axis (AP-axis) of the leg to its final longitudinal position parallel to the proximal-distal axis (PD-axis) of the leg [5]. A lot of research has been done in the past to identify key signaling pathways involved in the regulation of this process. It has been shown that a HOX gene, Sex combs reduced (Scr), and a sex determination gene, doublesex (dsx), play a central role in coordinating regulatory inputs and specifying the SC morphology, such as position, size, orientation and degree of rotation [6]. Some of the regulatory genes involved in the process include the leg patterning genes such as dachshund (dac), bric a brac (bab), Wingless (Wg), and the leg segmentation regulators such as the Notch signaling pathway [1, 7-9]. However, how this information is translated into changes in cell dynamics in the SC region and eventually leading to the adoption of the longitudinal position of SC remains poorly explored. One possible mechanism through which this may be done is through the modulation of surface tension. Intercellular surface tension has been shown to influence shape of cells within a tissue, and is dependent on cell adhesion and cortical tension

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Cell dynamics of sex comb morphogenesis in Drosophila melanogaster

Figure 1: Visual representation of landmark bristles and key regions under study. Shown above is the confocal micrograph of the 1st Tarsal Segment (TS1) of a male D. melanogaster expressing ubiDEcad::GFP 36hrs after pupariation (AP). Landmark bristles are represented by white circles; the Sex Comb is represented by white cross-marks; and the key regions are represented by red boxes. The Proximal Region (red box 1) refers to the tarsal epithelium surrounded by five landmark bristles (white circles): Longitudinal Bristle, Chemosensory Bristle, most proximal SC Bristle, most distal SC Bristle and Campaniform Sensilum (for motion detection & balance enhancement) [1]. The Distal Region (red box 2) is the tarsal epithelium surrounded by three landmark bristles (white circles) and a right angle: Central Bristle (the bristle that originated in the same transverse bristle row as the SC but got left behind during SC rotation) [1], most proximal SC Bristle, and most distal SC Bristle. At the beginning of SC morphogenesis when SC originates in its transverse position, the DR is distal to SC and the PR is proximal, hence their names. However at the end of the morphogenetic event, the DR appears posterior and PR appears anterior due to remodeling of the epithelium during the event.

[10]. Tissue surface tension, on the other hand, has been shown to explain global geometry of the tissue, and is dependent on cell adhesion [10]. Both have been used to explain the cell dynamics of many well-known morphogenetic events such as mesoderm invagination, convergent extension, dorsal closure, and retina development in D.melanogaster [10]. During the remodeling of the tarsal epithelium through SC rotation, epithelial cells seem to display a variety of behaviours similar to those occurring in systems experiencing complex surface mechanics, suggesting similar mechanisms may be at work. Malagon et al. have shown that regions of the presumptive SC field undergo considerable changes in area during SC rotation [11], while Tanaka et al. observed notable changes in cell shapes in the same regions [2]. Moreover, Atallah et al. have observed intercalation of cells similar to those seen during germ-band extension [4], suggesting a possible functional correlation of the process in both systems. In order to evaluate association between surface tension and SC rotation, we studied three cell behaviours that are under the modu-

Research Articles

lation of cell surface mechanics [10]: changes in cell size, changes in shape, and changes in position. Under the assumption that greater extent of changes would be required to generate greater motor forces, and eventually rotating a larger SC, we examined correlation between changes in cell size as well as cell shape and the length of SC. We also examined correlation in patterns of cell rearrangement among wild type (WT) individuals. The regions studied were parts of the tarsal epithelium immediately proximal and distal to the SC at the beginning of the rotation, termed the Proximal Region (PR) and the Distal Region (DR). The vertices of these regions were defined by land mark bristles to ensure consistency across all lines of flies studied (Figure 1). The PR is outlined by the Longitudinal Bristle, Chemosensory Bristle, Campaniform Sensilum, the most proximal SC bristle, and the most distal SC bristle; while DR is outlined by the same SC bristles, in addition to Central Bristle, and a right angle. Note that by the end of rotation, the PR becomes anterior to SC while the DR becomes posterior. The PR and DR in females refer to the regions proximal and distal to the most distal transverse bristle row homologous to SC, and are defined by a similar set of landmark bristles. To monitor changes in cell dimensions and positions, confocal time series of the tarsal epithelium during SC development were created. The movies were analyzed with ImageJ to acquire measurements of apical surfaces of cells, and statistical analyses were done on the results using MedCalc, and G*Power. Our results demonstrated a strong correlation between changes in cell size, as well as cell shape, in the cellular neighbourhood of SC and the number of teeth in SC. Furthermore, the exchange of neighbours in the same regions seemed to vary largely among WT individuals and appeared stochastic. Although it is difficult to predict the role of cell rearrangement in SC rotation at the moment, our analysis do show that cell intercalation is perhaps not actively driving SC rotation in contrast to what was previously believed. Our results suggest possible cause and effect relationship between epithelial remodeling and SC rotation. These also provide support for future investigation of modulation of surface tension as a mechanism behind SC morphogenesis.

Materials and Methods Fly stocks

Four lines of flies with varying length of SC expressing ubiDEcad::GFP were studied (Table 1): WT-Female, WT-Male, Low Line and High Line. WT-Female was used as a control; the female region homologous to the male SC region as defined by a similar set of landmark bristles was studied. DR and PR in females are proximal and distal to the most distal transverse bristles which are homologous to SC in males. Low and High were lines of flies with abnormally low and high number of SC bristles compared to WT-Male, and were created through divergent artificial selection [12]. Table 1: Lines of flies studied and their respective length of SC. We used four lines of flies with increasingly longer SC to probe relationships between cell behaviours and SC rotation. Line WT-Female Low WT-Male High

Length of SC (# of bristles) 0 [11] 4-5 [12] 9-10 [11] 12-13 [12]

Region of leg imaged TS1 TS1 TS1 TS1

Journal of Undergraduate Life Sciences â&#x20AC;˘ Volume 7 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Spring 2013

Research Articles

Cell dynamics of sex comb morphogenesis in Drosophila melanogaster

Confocal time-lapsed series Confocal movies documenting epithelial cell dynamics of the TS1 during SC development between 23-36 hrs AP were obtained from works previously done by others with permission [4, 11]. The time interval between the acquisition of z-stacks were 30 minutes for WT-Female, Low Line, and High Line; and 20 minutes for WT-Male [4, 11].

Measuring changes in cell size and cell shape The confocal movies were analyzed with ImageJ frame by frame. First, epithelial cell boundaries in the region of interest were manually outlined with white coloured lines using the brush tool. Second, the image was converted into 2-bit using the built-in “Make Binary” function and cropped to remove unnecessary regions. Third, the image was further edited using the brush tool and paint bucket tool to remove noise and blemishes until each cell was represented with a contoured patch of solid black sitting against a solid white background. Lastly, the edited image was processed with the built-in “Analyze Particles” subroutine to measure cell surface area, height and width. At least three samples were studied for each line of flies. To analyze change in cell size, cells were divided into two categories according to surface area: <9 μm2 were considered “Small”; while ≥9 μm2 were considered “Large”. The proportions of cells belonging to each size category (p(cell size)) at 23 hrs and 36 hrs in both DR and PR were calculated for each sample. To analyze change in cell shape, cells were divided into three categories according to height to width ratio: 0.5<H/W<1.5 was considered square; H/W<0.5 was considered horizontal rectangle, H/W>1.5 was considered vertical rectangle. The proportion of cells belonging to each shape category (p(cell shape)) at, again, 23-36 hrs in both DR and PR were calculated.

Measuring changes in cell position (cell rearrangement) The confocal movies were also analyzed with ImageJ frame by frame. Linear arrays of cells parallel to the AP-axis in regions of interest were manually labeled and followed from 23 hrs to 28 hrs AP. Cells were classified into six types of rearrangements based on local behaviours: one line of cells becoming two were considered intercalation along the PD-axis; two lines of cells becoming one were considered intercalation along the AP-axis; a group of minimum three cells rotating in unison clockwise (CW) or counter-clockwise (CCW) were considered CW rotation or CCW rotation respectively; cells not changing position were considered static; and lastly, cells that were disappearing from epithelium were considered as extrusion. Only WT flies were investigated for cell rearrangement. Three samples were studied.

Statistical analysis All statistical tests were done using MedCalc unless otherwise specified. To assess the correlation between SC length and cell size, Spearman’s coefficient of rank correlation (ρ) was calculated for the proportion of cells that got larger (Δp(Large Cell)) between 23-36 hrs AP and the length of SC (# of bristles) in each of the lines studied. The sample size required to yield statistically significant result for each rho was obtained using the Correlation Coefficient Sampling function to assess the significance of correlation. Furthermore, two sample t-tests were performed to analyze statistical significance of the difference inΔp(Large Cell) between all four lines of flies studied. Similar procedure was taken to assess relationship between SC length and cell shape, except that ρ was calculated for the proportion of cells that got more elongated along the PD-axis (Δp(Rectangular Cell)) between 23-36hrs AP. Lastly, for the cell rearrangement data, Chi-Square Test of Independence (χ2) was done

using Excel to assess whether the three samples of WT-Males studied differ from one another. Minimum required sample size was calculated using G*Power’s “A priori Power Analysis” function for “Goodness-offit tests: Contingency tables”.

Results

The tarsal epithelium surrounding the SC displays dramatic remodeling during SC morphogenesis [11]. To explore the possibility of regulation of surface tension being the mechanism underlying SC rotation, we studied several cellular processes often modulated by surface tension, and attempted to determine their relationship with SC rotation. Changes in area of apical surface of cells strongly correlate with SC length To determine the relationship between changes in cell size and SC rotation, we quantified the changes in apical area of epithelial cells during SC rotation in four lines of flies with increasingly larger SC. Under the assumption that cell surface area is generally indicative of cell size, our results showed that cells in the DR consistently increased in size while cells in PR consistently reduced in size for males. At 23 hrs the DR began with a large proportion of small cells and almost no large cells (Figure 2A), while at 36 hrs there was a reduction in the proportion of small cells and an increase in the proportion of large cells (Figure 2B), indicating that cells had become larger over time. In DR, the change in the proportion of large cells between 23 hrs to 36 hrs AP positively correlated with the number of SC teeth (ρ= 1.00, P<0.0001, α=0.05, β=0.20, Minimum Sample Size = 4). In other words, more cells became larger in the DR when the SC being rotated was longer (Figure 2C). The opposite trend was observed for the PR. At 23 hrs the PR began with more large cells in lines with longer SC, and more small cells in lines with shorter SC. Females did not seem to follow the trend (Figure 2D). At 36 hrs the PR ended with much more small cells and the proportion of small cells was roughly similar in different lines of flies. Females, again, did not seem to follow the trend (Figure 2E). In contrast to DR, the change in the proportion of large cells in PR negatively correlated with the number of SC teeth (ρ= 1.00, P<0.0001, α=0.05, β=0.20, Minimum Sample Size = 4). In other words more cells became smaller in the PR when the SC being rotated was longer (Figure 2F). It is also interesting to note the difference between males and females. While all males showed an increase in cell size in the PR, females showed a decrease. The patterns for the two sexes were also opposite in the PR. A comparison was drawn between cell size heat maps of a WT individual at the beginning (Figure 3B) and end (Figure 3C) of the rotation to visually present the extent of change. Notice how the cells in DR dramatically enlarged, while the cells in PR dwindled. Changes in shape of apical surface of cells strongly correlate with SC length To determine the relationship between changes in cell shape and SC rotation, we quantified changes in apical shape of epithelial cells during SC rotation in four lines of flies with increasingly larger SC. Our results showed that cells in both DR and PR elongated along the PD-axis of the leg during the course of rotation for males. At 23 hrs, there were a lot of square cells in the DR, the proportion of square cells are higher in lines of flies

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Cell dynamics of sex comb morphogenesis in Drosophila melanogaster

Research Articles

Figure 2: Effect of varying SC length on the apical surface area of epithelial cells in DR and PR [11]. The x-axis depicts the lines of flies arranged by their corresponding number of SC teeth from the smallest to the largest; y-axis represents the proportion of cells belonging to different size categories. (A,B) The proportion of “small” (red) and “large” (blue) cells in the DR at 23 hrs and 36 hrs AP respectively. (C) Change in the proportion of “small” and “large” cells from 23 hrs to 36 hrs AP. Notice the dramatic increase in cell size. Also notice how the proportion of cells that got larger positively correlates with the length of SC. (D-F) Same as above except for PR. The cells dramatically decreased in size, the extent of reduction positively correlates with the length of SC.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Cell dynamics of sex comb morphogenesis in Drosophila melanogaster

Figure 3: A visual representation of the changes in the apical surface area of epithelial cells in DR and PR [11]. (A) Color code for cells belonging to different size categories (B) Sample heat map of cell size in the DR and PR of a WT-Male at 23 hrs AP. The array of cross marks represents the position of SC. Red boxes outline the DR and PR. (C) Sample heat map of cell size in the DR and PR of the same WT-Male at 36 hrs AP. Both diagrams are shown at the same scale. There is considerable increase in the number of large cells (green, blue) in the DR and smaller cells (orange, yellow) in the PR.

with longer SC (Figure 4A). The female flies, as expected, did not follow the trend. At 36 hrs, a large increase in the proportion of vertical cells was observed and all four lines of flies ended up with similar proportion of both square and vertical cells (Figure 4B). This suggested that a lot of cells in the DR had elongated along the PD-axis during SC rotation. In DR, the change in the proportion of vertical cells between 23hrs to 36hrs AP positively correlated with the number of SC teeth (ρ= 1.00, P<0.0001, α=0.05, β=0.20, Minimum Sample Size = 4). In other words, more cells in DR elongated along the PD-axis when the SC being rotated was longer (Figure 4C). A similar trend was observed for the PR. At 23 hrs, the PR began with a large excess of square cells, the proportion of square cells were higher in lines with longer SC (Figure 4D). At 36 hrs, the proportion of vertical cells dramatically increased, while the proportion of square cells decreased. All four lines of flies ended with a similar proportion of both square and vertical cells (Figure 4E). Similar to DR, the change in the proportion of vertical cells in PR also positively correlated with the number of SC teeth (ρ= 1.00, P<0.0001, α=0.05, β=0.20, Minimum Sample Size = 4). In other words, more cells in PR elongated along the PD-axis in lines of flies with longer SC, similar to DR (Figure 4F). Furthermore, it is also notable that the change in the proportion of vertical cells in the DR was no longer opposite between the two sexes, suggesting that the magnitude of change in the DR is perhaps not as large as in that the PR. Finally, a comparison was drawn between the cell shape heat maps of a WT male at the beginning (Figure 5B) and end (Figure 5C) of the rotation to depict the extent of changes. Notice how a large number of cells elongated along the PD-axis in both the DR and the PR. Patterns of cell rearrangement vary largely among individuals and are perhaps stochastic In order to determine the relationship between the exchange of cell neighbours and SC rotation, we manually tracked the movement of individual epithelial cells near SC during SC rotation. Due to time limitations, we only compared the patterns of cell rearrangement among individuals of WT flies and we observed the following. First, cell rearrangement in DR and PR showed sig-

nificant variations with no associations between individuals. (For DR, χ2=0.86, DF=10, P=0.9999, Effect Size=1.27, α=0.05, β=0.20, Minimum Sample Size=11; For PR,χ2= 0.55, DF=10, P=0.9999, Effect Size=1.40, α=0.05, β=0.20, Minimum Sample Size=9) This was the case in terms of both the proportion of cells undergoing each type of rearrangement as shown in Figure 6, as well as the spatial distribution of those cells as shown in Figure 7. For instance, in the DR, clockwise rotation (Rotation-CW) was completely missing in Pupa 1, but was the dominant type of rearrangement in Pupa 2 and 3 (Figure 6A). Meanwhile in the PR, intercalation along the AP-axis (Intercalation-AP) was very prevalent in Pupa 3 but was present only to a small extent in Pupa 1 and Pupa 2 (Figure 6B). Moreover, when considering the spatial distribution of cells undergoing different rearrangements, no consistent patterns were observed either. For example, CW rotations occupied the entire anterior section in Pupa 1 but was only restricted to the distal region in Pupa 2 and to the proximal section in Pupa 3 (Figure 7B, D, F). Furthermore, the diversity of cell rearrangement in the DR was considerably smaller than in the PR. In the DR, only three types of rearrangements were present and more cells remained static (Figure 6A). While in the PR, all five types of rearrangements were observed and fewer cells remained static (Figure 6B).

Discussion

Association between changes in cell size and SC length suggests the interplay between cell surface area, intercellular surface tension, and tissue surface tension plays an important role in SC rotation In contrast to liquid where surface tension is constant, cell surface tension has been shown to closely correlate with cell surface area [10]. Therefore, if surface tension is indeed fundamental behind SC morphogenesis, we might be able to observe some kind of association between changes in cell surface area and SC rotation. Our results demonstrated a strong correlation between changes in cell size during SC rotation and SC length. While cells generally increased in size in the DR, and reduced in size in the PR, the extent of change was larger in lines with longer SC, and smaller in lines with shorter SC. Such correlation suggests that changes in cell size may play an important role in SC rotation. Moreover, the

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Cell dynamics of sex comb morphogenesis in Drosophila melanogaster

Research Articles

Figure 4: Effect of varying SC length on the apical surface shape of epithelial cells in DR and PR [11]. The x-axis depicts the lines of flies arranged by their number of SC teeth from the smallest to the largest; y-axis represents the proportion of cells belonging to different shape categories. (A,B) Proportion of cells belonging to “square” (green), “rectangle - horizontal” (red), and “rectangle - vertical” shape categories at 23 hrs and 36 hrs AP respectively. (C) Change in the proportion of cells belonging to the three different shape categories between 23 hrs to 36 hrs AP. Notice the dramatic increase in the proportion of vertical rectangles in DR. Also notice how the increase in the proportion of vertical rectangles positively correlates with the length of SC. (D-F) Same as A-C except for PR. Again, cells elongated along the PD-axis, the proportion of cells that did so positively correlate with the length of SC.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Cell dynamics of sex comb morphogenesis in Drosophila melanogaster

Figure 5: A visual representation of the changes in the apical surface shape of epithelial cells in DR and PR [11]. (A) Color code for cells belonging to different shape categories (B) Sample heat map of cell shape in the DR and PR of a WT-Male at 23 hrs AP. The array of cross marks represents the position of SC. The red boxes represent the boundaries of DR and PR. (C) Sample heat map of cell shape in the DR and PR of the same WT-Male at 36 hrs AP. Both diagrams are shown at the same scale. Notice how both DR and PR start off with a lot of squares (green) and end up with a lot of vertical rectangles (blue), indicating systematic elongation in the regions.

extents of changes were in opposite direction in males and females in both DR and PR, suggesting the changes are sexually dimorphic, providing a further piece of evidence of association between cell size changes and SC rotation. However, it remains unclear as to whether these changes are the cause of the rotation or simply a result of it based on our findings alone. Changes in apical surface area of cells have been shown to produce the necessary motor force needed to drive morphogenetic event such as mesoderm invagination in D.melanogaster [10]. The dramatic expansion and reduction in the DR and PR could allude to a similar role played by cell size changes in the SC system. If cell size changes are indeed the cause, three predictions could be made. First, the correlation could be a result of the differences in the amount of tissue tension necessary for rotating SCs of varying length, based on the assumption that greater changes could produce greater forces. Second, since changes in cell surface area could significantly impact the area of intercellular contact, which in turn influences intercellular and tissue surface tension [10], we speculate surface tension could be acting in synergy on both cellular and tissue level to drive SC rotation. Third, based on the models of SC rotation put forth by Malagon et al. [4] and Atallah et al. [11], we speculate that the role of PR and DR may alternate during the course of the event. Previous studies have demonstrated that larger contact area between cells as a result of better adhesion strengthens tissue surface tension [10] while smaller contact area as a result of less favorable adhesion weakens tissue tension [10]. Based on this observation, we speculate the constriction in PR may be needed for driving the rotation at the beginning due to initially larger cell sizes in the region producing stronger tissue surface tension. At this stage PR may be exerting a “pull” on the SC. As time progresses the cells in PR get smaller and the force exerted by the region diminishes, while cells in DR get larger and begin to exert stronger forces on the SC instead. At this stage the expansion in DR might be able to produce a “push” and take over the role of driving SC rota-

tion. On the other hand, if cell size changes are the result of SC rotation, they may be explained as passive response to the motor forces applied on the epithelium during SC rotation. Association between changes in cell shape and SC length suggests cell elongation, likely regulated by intercellular surface tension, plays an important role in SC rotation Spatial patterns of cell shape are regulated by intercellular surface tension [10]. Cells in tissues have been shown to adopt similar shapes as bubbles in 2D foam during morphogenetic events that span medium to long timescale [13]. This is due to the fact that both cells and bubbles share the similar tendency to optimize packing and minimize surface energy [13]. Similarly, the global shape of cell aggregates has been found to be under the regulation of tissue surface tension[14], evident in the formation of spheres by cell aggregates and the spontaneous sorting of mixed groups of cells [10]. Given the close correlation between cell shape and surface tension, if morphogenesis is indeed achieved by regulation of surface tension, we should be able to observe certain kind of association between changes in cell shape and SC rotation. Our results demonstrated that cells in both DR and PR elongated along the PD-axis during SC rotation. The extent of change was positively correlated with the length of SC, in both DR and PR, suggesting a possible role played by changes in cell shape during SC morphogenesis. Moreover, the proportion of cells elongating along the PD axis changed in opposite directions in PR of males and females, suggesting cell elongation in the region is male specific and adds a further piece of evidence supporting the association between cell shape changes and SC rotation. In contrast, the pattern of shape change observed in the DR was similar for both sexes – cells in the DR of both males and females elongated along the PD-axis. This is either due to an artifact of our data limited by its small sample size, or could suggest that the elongation of cells in the DR is intrinsic to proper leg development, such as the formation of joint, and is therefore

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Cell dynamics of sex comb morphogenesis in Drosophila melanogaster

Research Articles

Figure 6: Quantification of epithelial cell rearrangement among three different WT-Males. The x-axis represents different types of cell rearrangement, while the y-axis represents the proportion of cells partaking in each type of rearrangement. (A) Proportion of cells undergoing each type of cell rearrangement in the DR of three WT-Males between 23-28 hrs AP. The black bars represent data from pupa 1, grey bars from pupa 2, and white bars from pupa 3. (B) Same as (A) except for the PR. The proportion of cells doing each type of rearrangement showed significant variation from individual to individual and lacked any patterns. It should be also noted that cell intercalations along the AP-axis and cell extrusions were only present in the PR.

Figure 7: Spatial pattern of epithelial cell rearrangement in 3 different WT-Males [11]. (A) Color code of cells belonging to different categories of rearrangement. (B-G) Fate maps of cells of three WT-Males (column 1-3) at about 28 hrs AP in the PR (top row) and DR (bottom row). The array of cross-marks represent the position of SC. The red boxes represent the boundaries of DR and PR. Notice the significant variations in the spatial pattern of rearrangement between individuals. Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Cell dynamics of sex comb morphogenesis in Drosophila melanogaster

present to a certain degree in both males and females. Although the trend where more cells elongated in lines with longer SC still existed in DR, but it was present to a lesser extent, suggesting DR is probably less involved in SC rotation than PR. Again, due to the nature of the study, although we were able to demonstrate the correlation between shape changes and SC rotation, we were unable to determine whether cell shape changes are the cause or the effect of SC rotation. We speculate it is perhaps the latter because changes in cell shape regulated by intercellular surface tension have been known to drive morphogenesis and pattern formation in a variety of systems, including the development of retina in D.melanogaster [10]. Lastly, to further characterize cell shape changes, further experiments could attempt to identify the polygonal shapes taken by the cells. Epithelium regulated by intercellular surface tension has been known to favour the formation of three-fold vertices where three edges of the neighbouring cells meet at 120 degree, and result in the formation of hexagonal networks [13]. Therefore, observation of similar patterns in the tarsal epithelium could provide further evidence of surface tension at work. The lack of association between patterns of cell rearrangement among individuals demonstrates that it is perhaps stochastic in nature, and that cell intercalation is likely not driving SC rotation as previously anticipated Cell rearrangement has been known to cause changes in overall tissue geometry, particularly in systems where tissue mechanics has been used to explain morphogenesis [10]. For instance, cell intercalation has been shown to produce the motor forces needed to drive germ-band extension, and to results from cell shape changes regulated by intercellular surface tension [10, 15]. To investigate the role of cell rearrangement in SC rotation, we followed changes in cell positions during the initial stage of rotation (23 to 28 hrs AP). Due to the time consuming nature of this procedure, instead of examining extent of changes in lines of flies with varying SC length, we observed variations among WT individuals. Consistent patterns of cell intercalation were observed among individuals for germ-band extension. If cell rearrangement is also driving SC rotation, we might be able to observe similar consistency for the SC system as well. The classification of each type of cell rearrangement is based on changes in cell position relative to their neighbours and landmark bristles. One of the limitations of this method is that the groups of cells assessed for their types of rearrangement are chosen arbitrarily at the discretion of the experimenter and the result is heavily dependent on subjective choices. Another limitation is that manual tracking does not permit analysis of large sample sizes due to the amount of work involved. However, this may be resolved in the future by tagging nuclear envelopes with fluorescent proteins and then tracking and analyzing the position of each nucleus using automated image processing software [11]. Not only would this allow batch processing of large amounts of samples, but this would also eliminate the necessity for arbitrary assignment of cells into groups for more precise assessment. Our interpretation of the results below should be considered in light of these limitations. The most prominent feature of cell rearrangement in the DR and PR is the large variation among individuals in terms of the

number of cells undergoing each type of rearrangement and the spatial distribution of these cells. This is the opposite of what was seen for cell size and cell shape, where overall changes remained largely similar on the level of individual organisms, despite differences on the level of individual cells. However, although this result suggests that cell rearrangement is probably stochastic, it is difficult to say whether rearrangement is actively contributing to SC rotation or not. It has been shown that some degree of disorder could almost always be seen on finer scales during tissue reorganization, reflecting a limit to the extent of tight genetic regulation [10]. The reason that systems could usually produce consistent morphologies is due to the robustness of the developmental process, which enables the systems to tolerate a certain extent of noise [10]. Furthermore, examples of self-organization where apparently random behaviours consistently result in similar outcomes have been found in both the development of mouse optic cup [16] and zebra fish heart [17]. In these systems, a set of genes specify parameters for the system, then simply allow spontaneous interactions between components of the system to achieve the desired phenotype [18]. While the behaviour of each component may seem random, the overall outcome is always consistent because the components act within the confinement of parameters [18]. On the other hand, instead of actively contributing to SC rotation, cell rearrangement may also be a passive response to the other processes at work, such as the changes in cell size and cell shape mentioned previously, or to the rotation of SC, which applies mechanical force on the tarsal epithelium. Similar haphazard behaviours have been found in fly notum where apparently random delamination occurs in response to mechanical stress caused by over-crowding to ensure proper packing of epithelial cells [19]. Another interesting observation from our results is the higher diversity in the types of rearrangement in PR when compared to DR. In the PR, fewer cells remained static and all five types of rearrangement were observed. While in the DR, more cells remained static and only three types of rearrangement were present. Under the assumption that cell rearrangement is stochastic, one possible cause of this disparity in diversity is that shrinking cells in PR reduces contact area at cell boundaries and therefore weakens cell-cell adhesion. This would in turn allow for easier exchange of neighbours and lead to more varied patterns of rearrangement. Furthermore, since PR reduces in area during SC rotation it is likely that it will experience more mechanical pressure from the SC and thus require more reposition of cells to accommodate the constriction in space. Finally, previous studies done by Atallah et al. [1, 4] have observed cell intercalations in the PR, and suggested that it might play a similar role as the intercalations during germ-band extension. In order to evaluate the functional similarity of cell intercalation in both systems, we compared the formation and resolution of multicellular rosettes in the PR to that of the germ-band extension. In germ-band extension, rosettes form and resolve in directional fashion under the regulation of planar polarity proteins allowing for tissue elongation [20]. However, during SC rotation the progression between configurations (T1, T2, and T3) is bidirectional resulting in rosettes resolving in multiple directions and effectively preventing net elongation of tissue in any particular direction. This shows that cell intercalation is playing a different role during SC rotation.

Journal of Undergraduate Life Sciences â&#x20AC;˘ Volume 7 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Spring 2013

Cell dynamics of sex comb morphogenesis in Drosophila melanogaster

Research Articles

Conclusion

In short, our result demonstrates that change in cell size and cell shape closely correlate with the length of SC, suggesting a possible association between surface tension and SC rotation. Our study provides important evidence of cell dynamics that warrant further investigation of the role of surface mechanics, as well as the remodeling of cytoskeleton and intercellular junctions in SC development. To confirm the association between epithelial remodeling and SC rotation, further studies characterizing epithelial changes in lines of flies with varying length of non-rotating SCs could be done in comparison with our study. To elucidate the cause and effect relationship between epithelial remodeling and SC rotation, future studies could attempt to selectively reducing the number of cells in the DR and PR using approaches such as laser ablation to observe the effect on SC rotation [11]. Finally, to study the association between surface tension and SC rotation, further experiments could focus on measuring temporal and spatial patterns of intercellular surface tension using approaches such laser-induced cell fusion, or tissue surface tension using tensiometers. Alternatively, experiments could also focus on characterizing the activity of molecular regulators of cytoskeleton and intercellular adhesion in the tarsal epithelium, since surface tensions are ultimately controlled by cortical tension and cell-cell adhesion.

References

1. Atallah J. The development and evolution of complex patterns: the Drosophila sex comb as a model system [doctoral thesis]. Toronto: University of Toronto; 2008. 2. Tanaka K, Barmina O, Kopp A. Distinct developmental mechanisms underlie the evolutionary diversification of Drosophila sex combs. Proceedings of the National Academy of Sciences of the United States of America. 2009 Mar 24;106(12):4764-4769. 3. True JR. Combing evolution. Evol Dev. 2008 July;10(4):400-402. 4. Atallah J, Liu NH, Dennis P, Hon A, Larsen EW. Developmental constraints and convergent evolution in Drosophila sex comb formation. Evol Dev. 2009 Mar;11(2):205-218. 5. Tokunaga C. Cell lineage and differentiation on the male foreleg of Drosophila melanogaster. Dev Biol. 1962 June;4(3):489-516. 6. Barmina O, Kopp A. Sex-specific expression of a HOX gene associated with rapid morphological evolution. Dev Biol. 2007 Nov 15;311(2): 277-286. 7. Kopp A. Drosophila sex combs as a model of evolutionary innovations. Evol Dev. 2011 Nov;13(6):504-522. 8. Godt D, Couderc JL, Cramton SE, Laski FA. Pattern formation in the limbs of Drosophila: bric a brac is expressed in both a gradient and a wave-like pattern and is required for specification and proper segmentation of the tarsus. Development. 1993 Nov;119(3):799-812. 9. De Celis JF, Tyler DM, de Celis J, Bray SJ. Notch signalling mediates segmentation of the Drosophila leg. Development. 1998 Dec;125(23):4617-4626. 10. Lecuit T, Lenne P. Cell surface mechanics and the control of cell shape, tissue patterns, and morphogenesis. Nat Rev Mol Cell Biol. 2007 Aug;8(8):633-644. 11. Malagon J. Sex comb in motion: Cellular processes involved in the sex comb rotation in Drosophila melanogaster [doctoral thesis]. Toronto: University of Toronto; 2013. 12. Ahuja A, Singh RS. Variation and Evolution of Male Sex Combs in Drosophila: Nature of Selection Response and Theories of Genetic Variation for Sexual Traits. Genetics. 2008 May;179(1):503-9. 13. Plateau J. Statique Expérimentale et Théorique des Liquides Soumis aux Seules Forces Moléculaires. Paris: Gauthier-Villars; 1873. 14. Thomson DW. On Growth and Form. New York: Cambridge University Press; 1961. 15. Bertet C, Sulak L. Myosin-dependent junction remodeling controls planar cell intercalation and axis elongation. Nature. 2004 June;429(6992):667-671. 16. Eiraku M, Takata N, Ishibashi H, Kawada M, Sakakura E, Okuda S. Self-organizing optic-cup morphogenesis in three-dimensional culture. Nature. 2011 Apr 07;472(7341):51-6. 17. Hove JR, Koster RW, Forouhar AS, Acevedo-Bolton G. Intracardiac fluid forces are an essential epigenetic factor for embryonic cardiogenesis. Nature. 2003 Jan 09;421(6919):172-7. 18. Larsen EW, Atallah J. Epigenesis, preformation and the Humpty Dumpty problem. In: Hallgrímsson B, Hall BK, editors. Epigenetics: linking genotype and phenotype in development and evolution. Berkeley: University of California Press; 2011. p. 103-115. 19. Marinari E, Mehonic A, Curran S, Gale J, Duke T, Baum B. Live-cell delamination counterbalances epithelial growth to limit tissue overcrowding. Nature. 2012 Apr 26;484(7395):542-545. 20. Blankenship JT, Backovic S, Stephanie T, Sanny JP, Weitz O, Zallen J. Multicellular rosette formation links planar cell polarity to tissue morphogenesis. Dev Cell. 2006 Oct;11(4):459-470.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

RESEARCH

Relationship between population density, individual fitness, and blackspot infections in the Yellow Perch Perca flavescens Elaine Y. Luo Department of Ecology and Evolutionary Biology, University of Toronto.

Abstract

Population density, disease, and individual fitness are central themes in the study of population ecology. However, the fitness costs of increased population density and blackspot parasitism remain unclear. This study addresses the effects of population density and parasitism on the fitness of the yellow perch Perca flavescens. Fifteen populations of yellow perch were examined for blackspots, a common parasitic nematode, measured, weighed in Lake Opeongo, Algonquin National Park. Blackspot parasitic load did not appear to correlate with either population density or with fitness, while fitness was correlated with population density. These findings suggest that rather than directly through blackspot parasitism, individual fitness appears to be influenced by factors associated with increased population density. Our correlational study elucidates the relationship between these three population parameters and explores other mechanisms to which parasitism can influence the fitness of juvenile P. flavescens.

Introduction

The parasitic fauna of the yellow perch Perca flavescens is mostly dominated by several species of larval trematodes, more commonly referred to as blackspot parasites [1]. This parasite is ubiquitous in freshwater communities and can be commonly found in several species of fish. For example, in the pumpkinseed Lepomis gibbosus (commonly found in Ryan Lake, Ontario), blackspot parasites were found in nearly 100% of age-1 and older fish [2]. Other studies have described a similarly high prevalence [3, 4, 5, 6]. The high prevalence of this parasite in littoral fish, such as the juvenile yellow perch, make it an ideal model of parasitism to study concurrently with population dynamics in relationship to fitness. These blackspots can be caused by several species in various families of digenetic trematodes, including the diplostomoids Crassiphiala spp., Uvulifer spp. and Neascus spp. [7]. Their life cycle is complex, involving a series of at least three hosts and fish serving as the second intermediate host [7]. Usually, fish are infected with the metacercariae, which is passed on to the intestinal mucosa of certain piscivorous birds [7]. There, adult worms ultimately produce eggs, which are shed in the host feces and hatch to release miracidia [7]. The miracidium enters the snail, their first intermediate host, and produce free-swimming cercariae in the water column [7]. Infection occurs when a fish encounters cercariae that penetrate its integument and develop into encysted metacercariae, where the worm undergoes development before it matures in the appropriate avian [7]. Encysted metacercariae of some species can survive more than 4.5 years in a fish [1]. The characteristic blackspot (Figure. 1) is the result of the fish forming a capsule of connective tissue containing melanophores around the encysted metacercaria [8].

The high prevalence of parasitism in fish has resulted in many studies concerning fitness costs of parasitism. Various accounts of parasitism decreasing fish fitness exist in current literature [9, 10, 11]. For example, G. xenomas, A. brevis and R. acus infections were shown to reduce the growth and visceral fat reserves of juvenile yellow perch in a nutrient-poor Canadian Shield lake similar to our study site [9]. The juvenile bluegills Lepomis macrochirus, the sheepshead minnow Cyprinodon variegates and the roach Rutilus rutilus have all been shown to incur increased winter mortality from parasitic trematodes [12, 13, 14]. Low food supplies during long winters in northern latitudes, such as those at our study site in southern Ontario, require juvenile fish to have a greater reliance on energy stores for survival than fish in locations with shorter, warmer winters [15]. This greater reliance on energy stores is hypothesised to result in the observered winter-related mortality of heavily parasitized juvenile fish due to the competing energetic costs of parasitism and survival [4]. Nonetheless, the effect of the blackspot parasite on populations of juvenile fish remains unclear. In the laboratory, heavy infections have been shown to be fatal or cause a reduction in growth [1, 5, 16]. In natural systems, results are more variable with some studies showing deleterious effects from heavy infections [5, 17], and others showing little or no effect from the parasite [18, 19]. Indeed, blackspot parasite studies in relation to population density are uncommon. Little evidence exists as to whether the host immune system, which may be inhibited by the stress of increasing population density, plays a role in regulating infection rates. This study attempts to address these gaps in understanding of the blackspot parasite by examining the relationships between popula-

Journal of Undergraduate Life Sciences â&#x20AC;˘ Volume 7 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Spring 2013

Relationship between population density, individual fitness, and blackspot infections in the yellow perch Perca flavescens

Research Articles

Figure 1: Heavily infested juvenile yellow perch with 58 blackspot infections on a measuring board.

tion density, individual fitness, and parasitic infections in yellow perch. First, if the blackspot parasites rely on weakened hosts for successful infections, does blackspot prevalence correlate with an increasing population density (assuming population density results in increased resource competition and thus weakened immune systems)? Further, if parasitic infections come at a cost to fitness, does the infection intensity negatively correlate with individual fitness? Finally, if blackspot prevalence is not influenced by population density nor affects fitness, is individual fitness affected by varying population density?

Materials and Methods

I attempted to address these questions by studying the yellow perch populations in Lake Opeongo at Algonquin National Park, Canada. In this study, population density is represented by catch per unit effort (fish per trap per day), which is the standard in fisheries research for measuring abundance and density [20]. Parasitism is measured by the proportion of infected individuals, which measures infection prevalence, and average number of blackspots per fish, which measures infection intensity. Individual fitness is represented by weight residuals, which is the actual weight of the fish minus the expected weight of the fish at that given length. In other words, fish with a positive weight residual are heavier than expected, while those with a negative weight residual are lighter than expected. Weight residuals can be accurate measures of fitness for this stage of juvenile fish since they have high mortality rate, especially throughout the winter months, and we can assume that juveniles with higher energy stores, or higher weight residuals, would have a higher overwinter survival rate. These variables can be accurately measured in the field and can provide good proxies of environmental parameters of interest. Fish were sampled in sites throughout South Arm and Sproule Bay of Lake Opeongo, 45°42’33”N 78°22’05”W in August 2012 (Figure 2). The sites selected were distant, isolated areas in which localized, nontransferable populations of juvenile yellow perch can be found due to their non-migratory nature. These sites yielded a range of varying population densities. The shallower Sproule Bay lacks any coldwater predators of the juvenile yellow perch, such as lake trout, and yielded relatively high population densities. On the other hand, South Arm has a deeper basin, allowing for the formation of a sizeable hypolimnion and provides a habitat for coldwater predatory fish of juvenile yellow perch. Four overnight minnow traps, which were baited with equal amounts of dog food (Selection brand), and left in for around 20 to 26 hours, were set at one to two meter depths for each of the 17 sample sites and sampled twice in three consecutive days. Of the 17 sites, 15 yielded a total of 101 yellow perch fish that were measured using a fish board (Figure 1), weighed using a flat table scale, and counted for blackspots. Data analysis was performed in R and in Excel. A MANOVA was performed to determine whether infection prevalence or intensity was associated with population density. Then, a regressional analysis was

Figure 2: Map of sampling locations throughout South Arm (deep basin across the north region of the map) and Sproule Bay (shallower section towards the southern bay) of Lake Opeongo.

performed to determine if individual weight residuals correlated with number of infections. Finally, an ANOVA was performed to determine whether weight residuals was associated with population density.

Results

Summary of sampled fish size, weight and parasitic features A total of 158 fish were caught from all 17 sample sites. The yellow perch consistently stood out as the most common species, comprising 101 out of 158 fish caught. Two sites out of the 17 total sites did not yield yellow perch. The sampled yellow perch specimens consisted of juvenile fish ranging from 37 to 177 millimetres in length (34 to 172 millimetres in fork length) and 0.4 to 87.0 grams in weight. Prevalence of blackspot infections was high, with blackspots found in 57.4% (58 out of 101) of the juvenile yellow perch caught. The intensity of infection ranged from 1 to 58 spots per fish. Larger fish tended to be more heavily infested than smaller fish (Figure 3; R2 = 0.196 ; P<0.001), presumably due to increased temporal exposure to parasites and larger available surface area for cercariae attachment [7].

Figure 3: Larger, older fish tend to be more heavily infested than smaller, younger fish (P<0.001). Each data point represents an individual fish.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Weight'residual'(actual'?'expected)' '

Research Articles

Figure 6: Infection intensity (number of spots) does not affect the weight of fish. Here, the weight residuals are calculated by pooled length and weight from allyellow subpopulations. Each data point represents an individual Relationship between population density, individual fitness, and blackspotdata infections in the perch Perca flavescens fish. 2.5! 2!

1.5! 1!

R²!=!2.7E005!

0.5! 0!

00.5!

10!

20!

01!

30!

40!

50!

60!

70!

Number'of'spots'

Weight'residual'(actual'?'expected)'by' Weight'residual'(actual'?'expected)'by' site' site'

Figure 6: Infection intensity (number of spots) does not affect the weight of fish (P=NS). Here, the weight residuals are calculated by pooled length and weight data from all subpopulations. Each data point repreFigure 4: Proportion of infected individuals does not correlate with pop- Figure Infection intensity sents7:an individual fish. (number of spots) does not affect the weight of fish. ulation density (P=NS). Each data point represents one subpopulation. Here, the weight residuals are calculated by specific length and weight 2! data from each subpopulation to account for natural variations in weight from varying environmental conditions. Each data point 1.5! represents an individual fish. 1!

R²!=!0.00013!

2! 0.5!

1.5! 0!

1! 0! +0.5! 0.5! +1!

0! +1.5! 0! 00.5!

10!

20!

10!

20!

30!

40!

30! 40! Number'of'spots'

50!

60! 70! R²!=!0.00013! 60!

70!

Figure 7: Infection intensity (number of spots) does not affect the

01! of fish (P=NS). Here, the weight residuals are calculated by specific weight Figure 8: Population density negatively correlates with averaged weight residuals of that is the actual weight for of the fish length and weight datapopulation, from each which subpopulation to account natural 01.5! Figure 5: Infection intensity (in average number of blackspots per fish) variations Number'of'spots' minus expected at environmental that length (calculated by regressional in the weight fromweight varying conditions. Each data does not correlate with population density (P=NS). Each data point rep- point represents analysis of an weight-to-length graph of all populations in Fig. 8. Equation individual fish. resents one subpopulation. for expected weight = 3.2295*length in mm - 12.903). Each data point represents a subpopulation.

Population density does not correlate with the occurrence or intensity blackspot infections ! An increase in population density does not correlate with an increased proportion of infected individuals (Figure 4) nor with an increase in the occurrence of high intensity infections (Figure 5; MANOVA: F1,12=0.517; P=0.610). Blackspot infection is not significantly related to individual fitness Blackspot infection intensity does not seem to significantly correlate with individual fitness (Figure 6 and 7). Here, individual fish are plotted with weight residuals against number of blackspots. The two figures represent different methods of measuring fitness via weight residuals. In Figure 6, weight residuals are all calculated using one regression equation through all sampled fish, which is a pooled population of all subpopulations. This tests whether fish weight is affected by the parasite in absolute relation to the whole population. In Figure 7, weight residuals are calculated using specific regression equations made from each individual subpopulation. This tests whether fish weight is affected by the parasite relative to the subpopulation. The lack of relationship between weight and parasite infection is evident whether variations in weight from different habitats are taken into account or not, as linear regressions indicate P-values of 0.928 for both figures. Fish with extreme parasitic loads tend to have similar weight residuals in comparison to fish with little or no parasitic loads. These results suggest that a higher parasitic load does not seem to emaciate their hosts nor deteriorate the health of the individual fish.

Population density negatively correlates with individual fitness Population density negatively correlates with individual fit18! ness. In Figure 8, averaged weight residuals of various populations are plotted against population density in average fish/trap/day. All populations in the South arm were pooled together to obtain a more accurate measure of average weight residuals and population density. This pooling allowed for an accurate comparison between population groups and fit the assumption for homoskedasticity. Results show a significant decrease in weight residuals as population density increases (F1,154=3.875; P=0.050). In other words, fish that come from higher population densities tend to weigh less than those from lower population densities. This effect appears to be independent of fish length (regression of residuals vs. length: t154=0.607; P=0.545) and parasitic load (addressed in part II). In Figure 9, the same subpopulations are compared on a weight-to-length graph. Although some variation for weight at a given length remains, some subpopulations (for example, S1) tend to contain consistently heavier fish than other subpopulations (for example, SA1). Relating back to Figure 8, these heavier-weighing subpopulations tend to be lower19!in population density.

Discussion

Population density and blackspot infections Contrary to our expectations, population density did not seem to increase parasitism and disease load. This suggests that either an increased population density does not decrease immune system function or that parasitism is mediated more via frequency of encounter rather than host immune response. Considering the

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Relationship between population density, individual fitness, and blackspot infections in the yellow perch Perca flavescens

Research Articles

Weight'residuals'(actual'?'expected)'

0.5!

0.4! 0.3! 0.2!

R²!=!0.5624!

0.1! 0!

+0.1!

0.5!

1.5!

2.5!

3.5!

+0.2! +0.3! +0.4!

Density'(:ish/trap/day)'

Figure 8: Population density negatively correlates with averaged weight residuals of that population (P=0.05), which is the actual weight of the fish minus the expected weight at that length (calculated by regressional analysis of weight-to-length graph of all populations in Figure 8. Equation for expected weight = 3.2295*length in mm - 12.903). Each data point represents a subpopulation.

ubiquity of the blackspot parasite in freshwater lakes such as Lake Opeongo, it may be for the former reason. Yellow perch is a social fish that prefers to live in shoals, so increased population density Figure Weight to length incur graph additional of fish sampled, by colour into night9: not necessarily stressseparated to the individual fish sampled subpopulations. Three subpopulations were omitted due to [9]. This thatregressional shoaling fish, such as the yellowusing perch, may low suggests densities since analysis is more accurate a larger be exempt idea stress with increasing popusamplefrom size. the Here, it isthat evident thatincreases some subpopulations (for example, consistently contained higher-weighing fish than others at a certain lationS1) density [5]. If so, other species of shoaling fish should foldata point represents an individual fish. low a length. similarEach result in the lack of relationship between population density and parasite prevalence or infection intensity.

Blackspot infections and individual fitness It is also interesting to note that in this study, parasitic load did 20! not seem to decrease fitness, despite various accounts for parasite increasing fish energetic demands and mortality [9, 10, 11]. Larval trematodes, such as the blackspot parasite, cannot synthesize lipids and must derive energy from its hosts [11]. Previous studies have reported that high infection intensities of the blackspot parasitic trematode Uvulifer ambloplitis increase oxygen consumption and lipid metabolism in juvenile bluegill fish. This increased energetic demand from parasitism has led to an estimated 10–20% population morality during their first winter [5]. However, a decrease in fitness may become more evident in the harsher winters since parasitism may create taxing energetic demands in heavily infested individuals, while more abundant resources in the summer may offset minute increases in energetic costs of parasitism [11]. Perhaps an effect of parasitic load on weight will become evident should this study be repeated during the winter, whereby resource and food limitation would incur a more evident fitness cost of parasitism to juvenile yellow perch. Alternatively, the parasite may decrease fitness through other behavioural or cognitive means such as increasing lethargy or decreasing predation avoidance, which may not be evident using physiological weight measurements alone. Diplostomum spp. are found within the vitreous humour of the eye and can hinder visual acuity, while Neascus spp. are found within the muscles and can hinder muscle function [11]. Thus, parasitism may indirectly

Figure 9: Weight to length graph of fish sampled, separated by colour into sampled subpopulations. Three subpopulations were omitted due to low densities since regressional analysis is more accurate using a larger sample size. Here, it is evident that some subpopulations (for example, S1) consistently contained higher-weighing fish than others at a certain length. Each data point represents an individual fish.

decrease the effectiveness of antipredatory behaviour. In addition, taxed lipid stores in the spring due to the parasite can make juvenile fish more vulnerable to predation [11]. This is especially plausible since it confers a transmission advantage to the larval trematodes [11]. Several species of piscivorous birds could take advantage of juvenile bluegill prey in poor condition due to high larval trematode abundances [11]. For example, our study site Lake Opeongo contains several species of piscivorous birds (for example, the belted kingfisher, Megaceryle alcyon or the great blue heron, Ardea herodias) that can serve as definitive hosts for P. minimum, Neascus sp., and Diplostomum sp. [21]. Parasite-induced host mortality caused by these subjective behavioural measurements is difficult to demonstrate in field studies, since predated fish would not be accounted for in samples of live fish. Population density and individual fitness Despite the lack of relationship between fitness and parasitism, fitness shows a clear decreasing correlation with an increasing population density even during the resource-abundant summer months. Higher density areas tend to have the lowest-weighing fish after having accounted for length, which suggests that there is a fitness cost of increased intraspecific competition. Food and resource limitation would be a likely cause of this relationship since high-density populations may be more prone to resource limitation. In instances of high population densities of conspecifics (which is the case here since juvenile yellow perch alone consisted almost two-thirds of our total sampled fish), individuals may mitigate the effects of intraspecific competition by shifting behaviours to utilize alternative resources not used by conspecific competitors. These alternate resources – for example, a substituted, more diversified diet – may not represent an optimal resource range and can thereby lead to decreased individual fitness in high-density populations. This suggests that factors associated with increased population density – for example, resource limitation – may be a larger factor in mediating fitness and weight than parasitic infections alone.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Relationship between population density, individual fitness, and blackspot infections in the yellow perch Perca flavescens

Finally, it is still unclear whether the fish immune response mediates parasitic infection rates in conspecifics. There seems to be no relationship between population densities, which may increase susceptibility to disease from stress incurred by increased intraspecific competition, and both infection rate and intensity. It would be interesting to further explore these relationships by isolating confounding variables through a population density manipulation experiment to establish causation between density and disease.

Acknowledgements

I would like to thank Professor Helene Cyr for providing an opportunity to participate in this field course, Victoria College and the Department of Ecology and Evolutionary Biology at the University of Toronto for their generous awards, without which this field course would not have been financially feasible, and the wonderful crew at the Harkness Laboratories of Fisheries Research.

References

1. Hoffman, G. L., and R. E. Putz. 1965. The black-spot (Uvulifer ambloplitis: Trematoda: Strigeoidea) of centrarchid fishes. Transactions of the American Fisheries Society. 94:143–152. 2. Cone, D. K. and Anderson, R. C. 1977. Parasites of pumpkinseed (Lepomis gibbosus L.) from Ryan Lake, Algonquin Park, Ontario. Canadian Journal of Zoology. 55:1410–1423. 3. Evans, H. E., Mackiewicz, J. S. 1958 The incidence and location of metacercarial cysts (Trematoda: Strigeida) on 35 species of central New York fishes. Journal of Parasitology. 44:231–235. 4. Lemly, A. D. 1996. Winter stress syndrome: An important consideration for hazard assessment of aquatic pollutants. Ecotoxicology and Environmental Safety. 34:223–227. 5. Lemly, A. D., and G. W. Esch. 1984. Effects of the trematode Uvulifer ambloplitis on juvenile bluegill sunfish, Lepomis macrochirus—ecological implications. Journal of Parasitology. 70:475–492. 6. Ferrara, A. M., Cook, S. B. 1998. Comparison of black-spot disease metapopulations in the central stonerollers of two warm-water streams. Journal of Freshwater Ecology. 13:299–305. 7. Quist, M. C., Bower, M. R., Hubert, W. A. 2007 Infection by a black spot-causing species of

Uvulifer and associated opercular alterations in fishes from a high-desert stream in Wyoming. Diseases of Aquatic Organisms. 78:129-136. 8. Berra, T. M., Au, R. 1978. Incidence of black spot disease in fishes in Cedar Fork Creek, Ohio. Ohio Journal of Sciences. 78:318–322 9. Johnson, M. and Dick, T. 2001. Parasite effects on the survival, growth, and reproductive potential of yellow perch Perca flavenscens in Canadian Shield lakes. Canadian Journal of Zoology. 79(11): 1980-1992 10. Marcogliese, D and Pietrock, M. 2011. Combined effects of parasites and contaminants on animal health: parasites do matter. Trends in Parasitology. 27(3): 123-130 11. Pracheil, B. and Muzzall, P. 2010. Population Dynamics of Larval Trematodes in Juvenile Bluegills from Three Lakes II, Michigan, and the Potential for Overwinter Parasite-Induced Host Mortality. Transactions of the American Fisheries Society. 139:652–659 12. Fischer, S. A., and W. E. Kelso. 1990. Parasite fauna development in juvenile bluegill and largemouth bass. Transactions of the American Fisheries Society. 119:877–884. 13. Coleman, F. C., and J. Travis. 1998. Phenology of recruitment and infection patterns of Ascocotyle pachycystis, a digenean parasite in the sheepshead minnow, Cyprinidon variegates. Environmental Biology of Fishes. 51:87–96. 14. Knopf, K., A. Krieger, and F. Holker. 2007. Parasite community and mortality of overwintering young-of- the-year roach (Rutilus rutilus). Journal of Parasitology. 93:985–991. 15. Wohlschlag, D. E., and R. O. Juliano. 1959. Seasonal changes in bluegill metabolism. Limnology and Oceanography. 4:195–209. 16. Hunter, G. W. III, and Hunter, W. S. 1938. Studies on host reactions to larval parasites. I. The effect on weight. Journal of Parasitology. 24:477–481. 17. Harrison, E. J. and Hadley W. F. 1982. Possible effects of black-spot disease on northern pike. Transatlantic American Fisheries Society. 111:106–109. 18. Baker, S. C. and Bulow, F. J. 1985. Effects of black-spot disease on the condition of stonerollers Campostoma anomalum. American Midland Naturalist. 114:198–199. 19. Vinikour, W. S. 1977, Incidence of Neascus rhinichthysi (Trematoda: Diplostomatidae) on longnose dace, Rhinichthys cataractae (Pisces: Cyprinidae), related to fish size and capture location. Transatlantic American Fisheries Society. 106:83–88. 20. Harley, S., Myers, R., and Dunn, A. 2001. Is catch-per-unit-effort proportional to abundance? Canadian Journal of Fisheries and Aquatic Sciences. 58(9): 1760-1772 21. Hoffman, G. L. 1999. Parasites of North American freshwater fishes. Cornell University Press, Ithaca, New York.

Consider taking your next course outside Lake Ontario

Discover what ﬁeld courses are available to University of Toronto students • • • •

University of Toronto ﬁeld courses Ontario Universities Program in Field Biology Research Opportunities Funding & Scholarships

Department of Ecology & Evolutionary Biology Field School of Biology 40

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

http://fb.eeb.utoronto.ca

Research Articles

JULS

RESEARCH

Sequence-based typing of clinical isolates of Mycobacterium xenopi isolated from Ontario, Canada Samia Mirza1, David J. Farrell2,3, Theodore K. Marras4, Jennifer Ma3, Daniel Liu1, David C. Alexander3, Julianne V. Kus2,3, Frances B. Jamieson2,3 University of Toronto Department of Laboratory Medicine and Pathobiology, University of Toronto 3 Public Health Ontario Laboratories 4 Department of Medicine, University of Toronto 1 2

Abstract

Mycobacterium xenopi is an environmental pathogen responsible for pulmonary infections and tuberculosis-like disease typically found in patients with previous pulmonary disease. A sequence typing (ST) scheme was developed to classify and analyze sixteen isolates obtained from a prospective clinical study and subsequently extended to include five additional clinical isolates. Twenty housekeeping genes were sequenced and inspected for single nucleotide polymorphisms (SNPs). Seven targets including atpD, fusA1, glnA1, pheT, secA1, topA, and the internal transcribed spacer (ITS) region were selected for continuing ST analysis, using the type strain ATCC 19250T as reference. Three isolates contained SNPs in glnA1 and pheT while the remaining 13 isolates displayed SNPs in topA. Consistency was observed with topA SNPs exclusively present in ITS subgroups B, C, and BC whereas glnA1 and pheT SNPs were only evident in ITS A. Among the sixteen clinical isolates obtained from the prospective study, strain types (ST) ST4 and ST7 were most prevalent. Two patients with pulmonary symptoms, each with two isolates collected three months apart, displayed identical sequences in all target genes suggesting persistence of the strain causing infection rather than reinfection. We have described seven SNPs in six targets (atpD, secA1, fusA1, glnA1, pheT and topA) and ITS, which provide a typing tool to analyze relatedness between strains and a foundation for a genotypic phylogenetic comparison. This typing scheme represents a basic tool that facilitates the comparison of M. xenopi strains to investigate the source and transmission dynamics between patients and their environment.

Introduction

Mycobacterium xenopi, a species of nontuberculous mycobacteria (NTM), is the third most common Mycobacterium species isolated at the Public Health Ontario (PHO) Laboratories. Previous analysis of NTM isolates demonstrated increasing prevalence of M. xenopi between 1997 and 2007 (Tuberculosis and Mycobacteriology Laboratory, Public Health Laboratory) [1]. An annual 7.3% (P = 0.0005) increase in M. xenopi isolation prevalence was observed in Ontario, with an average increase of 8.9% (P<0.0001) of isolation prevalence for all NTM isolates analysed [1]. M. xenopi is a slow-growing [1] organism found typically in soil and water from natural or engineered sources. With optimal growth temperatures in the warmer 42-45°C range, domestic hot water is an ideal environment supporting the growth of this organism. Transmission to humans is thought to occur via inhalation of aerosols from environmental sources, which differentiates M. xenopi from forms of pulmonary Mycobacterium infection causing tuberculosis [3, 4]. It is an opportunistic pathogen that primarily

causes prolonged pulmonary disease, where a radiologically evident tuberculosis-like pulmonary disease develops [5]. It can also cause extrapulmonary disease (e.g. in the lymph nodes and pleura) in immunocompromised individuals [6]. M. xenopi is most prevalent in individuals with significant smoking history and/or underlying respiratory diseases such as chronic obstructive pulmonary disease (COPD), asthma, and emphysema, and in persons with immunocompromised conditions such as hematologic malignancies [7, 8]. Despite the increasing prevalence of M. xenopi, little is known about its genetic diversity and the impact of DNA polymorphisms on its environmental persistence or clinical presentation. For many bacterial species, DNA sequencing and multi-locus sequence typing (MLST) based-analysis of housekeeping genes has allowed for the development of powerful phylogenetic tools which can be used to study a number of clinical questions, such as the persistence of a strain or acquisition of a new strain, as well as public health and epidemiological questions [9], such as the source of the pathogen and potential environmental niche; these are currently unanswered questions for M. xenopi. The goals of

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Sequence-based typing of clinical isolates of Mycobacterium xenopi isolated from Ontario, Canada

this study were to identify potential genetic targets that could be used for strain typing (ST) to assess the genetic diversity of M. xenopi. Additionally, this tool will allow future investigations to establish whether the prevalence of certain clinical M. xenopi strains is restricted to the type of host infected and how particular strains may affect the clinical presentations of disease.

Materials and Methods Clinical Information

Sixteen isolates of M. xenopi were identified from patients who had consented to participate in a clinical study currently being conducted at the University Health Network, Toronto and PHO Laboratories. Five additional isolates were convenience samples obtained from the PHO collection of clinical strains. Clinical information was collected from 16 patients enrolled in the prospective study, during routine clinical visits. Clinical measures such as demographic information, prior lung disease, chest radiologic findings, diagnostic criteria for NTM disease [5], time span of disease, medication information, and pulmonary function study results were also recorded. Respiratory specimens were collected from patients and processed at the PHO laboratories for the isolation and detection of Mycobacterium species. M. xenopi isolated from these patient specimens were used in the study.

Bacterial Strains

All isolates of M. xenopi were grown on solid Löwenstein-Jensen (LJ) slants, Middlebrook 7H10 agar plates and in 7H9 liquid broth (MGIT 960, Becton-Dickenson and Company, Franklin Lakes, NJ). Cultures were grown at 42°C for a period of 4-6 weeks until sufficient growth was observed as dense yellow/orange colonies covering LJ slants. Aliquots were archived at -80°C for future use. DNA extraction was performed using cells from these frozen stocks where additional DNA analysis was required.

Polymerase Chain Reaction (PCR) Amplification

Twenty housekeeping genes (ftsK, gcvB, hsp65, tpi, rpoB, infB, leuS, glnA1, fusA1, secA1, pheT, topA, gyrB, alaS, atpD, gyrB, dnaK, gap, ffh, fum) and the 16S-23S rRNA internal transcribed spacer (ITS) region were selected for sequence analysis.

Primers were designed to generate PCR amplicons of approximately 500 base pairs to optimize sequencing analysis. 20 μL PCR reactions were run in 96-well plates using the HotStarTaq Plus Master Mix Kit (QIAGEN, Cat. No. 203645). Each reaction was run with 1 μL of each primer (10 μM), 10 μL of 2x HotStartMM , 2 μL of Q solution, 2 μL of 10x CL buffer, 3 μL of nuclease-free water, and 1 μL of template DNA. Template DNA was prepared using a crude cell lysis method. Briefly, bacteria were resuspended in 200 μL of TE buffer (Tris-EDTA), vortexed to disrupt clumped cells, then lysed by heating at 95°C for 20 minutes. Samples were centrifuged for 1 minute at 3500 RPM to discard cellular debris. The DNA-containing supernatant was stored at -20°C. For all gene targets, PCR amplification was performed under the following conditions: 1 cycle of initial denaturation at 95°C for 5 minutes, 35 cycles of denatur¬ation at 94°C for 45 seconds, annealing at 57°C for 45 seconds, extension at 72°C for 1 minute, and 1 cycle of final extension at 72°C for 10 minutes; the reactions were then stored at 4°C. Upon completion, 2.5 μL of PCR product was subjected to electrophoresis on a 1.5% agarose gel with 2.0 μL of ethidium bromide to visualize the amplicons. DNA ladders were used (100 bp and 1 kb+) in order to determine size and approximate concentration of each amplicon.

Amplicon Labeling and Sanger DNA Sequencing Labelled amplicons for Sanger DNA Sequencing were prepared using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Life Technologies) following manufacturer’s instructions. The forward and reverse strands of each PCR amplicon were sequenced separately, using 10 μM F-primer and R-primer concentrations. The following labeling reaction cycle was run: 1 cycle at 96°C for 1 minute, 35 cycles at 96°C for 10 seconds, 50°C for 5 sec and 60°C for 4 minutes. Prior to sequencing, labeled amplicons were purified using recommended EDTA/Ethanol methods (Applied Biosystems) and DNA was resuspended in a final volume of 10 μL nuclease free water. Sequences were determined by capillary electrophoresis using an AB 3730x1 DNA Analyzer (Applied Biosystems, Life Technologies, Foster City, CA, USA).

Polymorphic Site 1 ATCC19250T PHO 08 PHO 34 PHO 04* PHO 05* PHO 25 PHO 22 # PHO 28 # PHO 30 PHO 31 PHO 33 PHO 38 PHO 14 PHOL001 PHO 35 PHO 19 PHOL002 PHO 12 PHO 29 PHOL003 PHO 37 PHOL004

C C C C A CGC CGGG T CGA T . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

T C C CAGT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

T C A CG T A A CGGCGGCGC CGAGGCGC T . . . . . T . . . . . . . . . . . . . . . . . . . . . . . . . T . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . C. . . . . . . . . . . . . . . . . . . . . . . . . C. . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . T . . . . . . . . . . . . . . . . . . . . . . . . . T . . . . . . . . . . . . . . . . . . . . . . . . . T . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . T . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . C. . . . . . . . . . . . . . . . . . . . . . . . . T . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . . . . . . . C . . . . . . . . . . . . . . . . . . . .

Polymorphic Site 2 T . . . . . . . . . . . . . . . . . . . . .

T GC CGGC CG CGGCG T . . . . . . . .G. . . . . . . . . . . . . .G. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . . . . . . . . . .G C . . . . . . . . . . . . . .G. . . . . . . . . . . . . .G. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . . . . . . . . . .G. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . . . . . . . . . .G. . . . . . . . . . . . . .A. . . . . . . . . . . . . .A. . . . . .

T . . . . . . . . . . . . . . . . . . . . .

T GGGC T CGGCG T AGA T C T . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

T . . . . . . . . . . . . . . . . . . . . .

Figure 1: Alignment of a portion of topA highlighting two polymorphic sites. Sequence alignment of the housekeeping target topA (DNA topoisomerase I) using MEGA 5.05 software [10]. SNPs were seen at two sites. The T/G allele was conserved in 8 strains, including the M.xenopi Type strain ATCC 19250T. The C/A allele was only evident in strains lacking an ITS ‘A’ alleles (e.g. ITS subgroups BB, CC and BC). Strains marked with ‘*’ and ‘#’ are sequence isolates, received three months apart, from Patient 1 and 2, respectively.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

PHO 04 PHO 05 PHO 22 PHO 25 PHO 28 PHO 30 PHO 31 PHO 33 PHO 35 PHO 29 PHO 14 PHO 12 PHO 37 PHO 19 PHO 08 PHO 34 PHO 38

ST4

ST6

ST5 ST8 ST3 ST7

Figure 2: Phylogenetic tree depicting relationships between SNPs. A phylogenetic tree was developed through comparing SNP patterns seen in six housekeeping genes atpD, fusA1, glnA1, pheT, secA1, topA and the internal transcribed spacer (ITS) region in clinical strains of nontuberculous M. xenopi. The ITS region was analyzed due to the low level of evolutionary pressure acting on the non-functional RNA, providing a significant measure of variance. From the phylogenetic comparison, three main clusters were originally formed, with ST7 differentiating completely from the other two clusters. ST7 and ST4 were least closely related to the Type strain, ATCC 19250T, and ST3 was most closely related to ATCC 19250T. PHO = samples isolated at Public Health Laboratories.

Molecular Analysis DNA sequence alignments were used to identify single nucleotide polymorphisms (SNPs) in twenty target genes (MEGA 5.05, M5 Alignment Explorer, Tempe, Arizona). In Figure 1, alignment of topA is presented. Six housekeeping genes – atpD (ATP synthetase beta chain), fusA1 (elongation factor G), glnA1 (glutamine synthetase), pheT (phenylalanyl-tRNA synthetase), secA1 (preprotein translocase), topA (DNA topoisomerase I) – and the ITS region were selected to develop ST analysis through the presence of single nucleotide polymorphisms (SNPs). All clinical strains were compared to the type strain ATCC 19250T. Further analysis, including phylogenetic comparisons (Figure 2), of sequenced strains and the ST groups was performed using Bionumeric software (Applied Maths, Austin, TX, US). Patients enrolled in the prospective clinical study (Table 1) were grouped under ST groups formed through phylogenetic comparison (Figure 2). This was done in order to determine if there was a clinical correlation between certain ST designation and disease presentation.

Results

Research Articles

100

Sequence-based typing of clinical isolates of Mycobacterium xenopi isolated from Ontario, Canada

Analysis of PCR Products and Selection of Targets for Sequence Typing Successful PCR of housekeeping genes was determined through visualization of bands of expected sizes via electrophoresis. PCR reactions were deemed unsuccessful if negative, weakly stained, or multiple bands were seen on gel electrophoresis. For instance, infB repeatedly displayed multiple bands when amplicons were separated through gel electrophoresis. Consequently, infB was the only target that failed to produce interpretable sequences in any of the clinical strains through capillary electrophoresis, as no distinguished peaks were observed on the chromatograms of this target, and it was not included in further analysis. Target sequences from all 21 clinical strains were aligned and only genes in which SNPs were detected were used for the development of ST analysis system. After deriving sequences from capillary electrophoresis of amplified genes and performing target specific strain alignment (MEGA 5.05, M5 Alignment Explorer, Tempe, Arizona),

Figure 3: Chromatograms of mixed region of ITS. Chromatograms derived through capillary electrophoresis from ABI 3730x1 DNA Analyzer (Applied Biosystems, Life Technologies, Foster City, CA, USA). A) The 237-276 base pair chromatogram section from the reverse ITS sequence of isolate PHO08; no mixed peaks were detected, and the isolate was designated as ITS A. B) The 237-276 base pair chromatogram section from the reverse ITS sequence of the isolate PHO37; several mixed peaks were evident, and ITS was determined to be ITS BC. Both chromatograms were compared to known ITS sequences [11], in order to derive ITS designation.

we observed no polymorphisms in the housekeeping genes, ftsK, gcvB, hsp65, tpi, rpoB, infB, leuS, gyrB, alaS, gyrB, dnaK, gap, ffh and fum, and therefore, these targets were not included in further ST analysis. Six housekeeping genes – atpD (ATP synthetase beta chain), fusA1 (elongation factor G), glnA1 (glutamine synthetase), pheT (phenylalanyl-tRNA synthetase), secA1 (preprotein translocase), topA (DNA topoisomerase I) – and the ITS region were selected for development of the M. xenopi MLST method. Target-specific Alignment and further ITS Discrimination SNP patterns in target genes glnA1, pheT and topA provided a way to differentiate clinical strains, with 3 out of the 16 clinical strains carrying SNPs in glnA1 and pheT and the remaining 13 depicting SNPs in topA. Five subsequent routine clinical strains were analyzed and additional SNPs were observed in target genes atpD, fusA1, and secA1. This allowed for a greater degree of discrimination and a more precise phylogenetic comparison when used in combination with glnA1, pheT, and topA (Figure 2). Interestingly, the SNP pattern seen in secA1 occurred in all 21 clinical strains and only the type strain ATCC 19250T differed at this target. Unlike other slow growing mycobacteria, the M. xenopi genome contains two rRNA operons and thus, two ITS sequences. To date, three ITS alleles (A, B, and C) have been identified [10] through detection of interspecific polymorphisms to further distinguish Mycobacterium species, which traditionally differentiate based on 16S rRNA sequencing. Individual strains can exhibit homozygous (AA, BB, CC) or heterozygous (AB, AC, BC) combinations of ITS alleles. DNA sequences from strains with homozygous alleles exhibited clean chromatograms with no mixed peaks (Figure 3). In contrast, chromatograms from strains with heterozygous ITS

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Sequence-based typing of clinical isolates of Mycobacterium xenopi isolated from Ontario, Canada

Table 1: Single nucleotide polymorphism (SNP) distribution in housekeeping gene targets. SNP distribution in atpD, fusA1, glnA1, pheT and topA housekeeping targets in comparison with internally transcribed spacer (ITS) region subgroups for clinical strains of Mycobacterium xenopi. Only topA contained SNPs at two different sites. The glnA1 ‘C’ and pheT ‘T’ polymorphisms were only observed in strains belonging to ITS ‘AA’ subgroup. Sequence types (ST) were derived from phylogenetic comparison (Figure 2) through measures of relatedness. PHO = samples isolated at Public Health Ontario laboratories, * = isolates from Patient 1, # = isolates from Patient 2 (Refer to Figure 2 for phylogenetic comparison), & = additional routine clinical strains analyzed. Strain

topA atpD secA1 fusA1 glnA1 pheT site 1

site 2

ITS

ATCC 19250T PHO14& PHO19 PHO04* PHO05* PHO22# PHO28# PHO25 PHO30 PHO31 PHO33 PHO35 PHOL004& PHO12 PHO29 PHO08 PHO34 PHO38 PHOL002& PHOL003& PHO37 PHOL001&

g g g g g g g g g g g g g g g g g g g g g a

g g a a a a a a a a a a a a a g g g g g g g

AA AB BB BC BC BC BC BC BC BC BC BC BC CC BB AA AA AA AA AA BC AA

1 2 3 4 4 4 4 4 4 4 4 4 4 5 6 7 7 7 7 7 8 N/A

t c c c c c c c c c c c c c c c c c c c c c

a a g a a a a a a a a a a a a a a a a a c a

t t t t t t t t t t t t t t t c c c c c t c

g g g g g g g g g g g g g g g t t t t t g t

t t c c c c c c c c c c c c c t t t t t c t

alleles exhibited mixed peaks (Figure 3b). The ITS profile (i.e. AB, AC or BC) was assigned using the distinct patterns of mixed peaks which resulted from the combination of heterozygous alleles. An association between the detection of SNP patterns in certain ITS groups was observed, as SNPs patterns seen in glnA1 and pheT occurred in strains only belonging to ITS group AA, and topA SNP patterns were only seen in ITS groups BB, CC, and BC. Phylogenetic Analysis of Sequence Types Clinical strains were categorized into 8 different STs based on the sequences aligned at atpD, fusA1, secA1, glnA1, pheT, topA and ITS, with the majority of patients categorized into the ST4 and ST7 strain designations (Table 1). From the phylogenetic tree in Figure 2, three main clusters exist, with ST7 differentiating completely from the other two clusters. A gradual evolution of strains appears from ST8, up to ST4, with ST8 being most closely related to the type strain ATCC 19250T, and ST4 being least closely related. Clinical Correlation Table 1 and Figure 2 also highlight a close relationship between isolates from two patients (PHO5 and PHO22), both ST4. Of further interest is that isolates from these patients collected 3 months later appeared to have identical sequences in all target genes, suggesting persistence of the strain causing the infection rather than acquisition of a new

strain. Fibrocavitary disease, a more destructive stage or type of infection was observed in four of seven (57%) of patients with ST4. Nodular bronchiectasis, a less destructive type of infection was observed in two of three (67%) of patients with infection from ST7 strains.

Discussion

Sequencing-based genotyping techniques permit classification of bacterial pathogens and provide insight into population structure, and are a valuable tool for monitoring disease transmission. We have demonstrated that the housekeeping genes atpD, glnA1, pheT, fusA1, secA1, and topA and ITS are useful targets to use as a preliminary typing tool to analyze relatedness between clinical isolates of M. xenopi. This initial typing has provided a strong foundation for phylogenetic comparisons (Figure 2). Nine of out twenty-one clinical isolates belonged to ST4, making it the most prevalent genotype in this limited study. It is tempting to speculate that strains of ST4 may be more suited to the human host in comparison to other strain types, or may come from a similar source, due to the similar clinical disease in the patients with these isolates and the high degree of differentiation from the type strain ATCC 19250T. However, this limited sample size is too small to permit this conclusion. Additional analysis of strains using sequencing typing tools is needed in order to study the relationship between certain strains of M. xenopi and factors driving the persistence of infection by ST4 in these hosts. Persistence of M. xenopi infection was also observed in patients 1 and 2 (Figure 2, Table 1), using ST analysis, as all sequences at all targets (Table 1) were identical, suggesting a persistence of the same strain rather than replacement with a new strain. Interestingly, we observed that some specific strain types have been associated with particular forms of pulmonary disease described radiographically: ST7 with nodular bronchiectasis and ST4 with fibrocavitary disease. Unfortunately our sample was too small to rigorously assess associations between strain types and clinical disease presentations. Further studies are required to determine the consistency of our results and to study associations between strain types with host factors and clinical outcomes. This preliminary typing scheme can be used for comparing any number of clinical and/or environmental strains of M. xenopi in order to investigate the source and transmission dynamics of this opportunistic pathogen in and between patients and their environment. In addition, this typing tool can also be advantageous in highlighting any virulence mechanisms driving the greater prevalence of certain strain types of M. xenopi.

Conclusion

The housekeeping genes atpD, glnA1, pheT, secA1, fusA1, topA and the 16S-23S rRNA ITS region provided an efficient MLST tool for creating a phylogenetic comparison between clinical isolates of M. xenopi. This typing scheme helps build a foundation for future clinical studies into the strain specific effects on methods of transmission, the effect of host factors on prevalence and the effect of strains on determining clinical presentations of disease. Future directions would involve analyzing a greater number of clinical strains in order to graph a more accurate phylogenetic comparison. Moreover, clinical isolates from various geographical locations can give greater insight into the impact of environmental conditions on the prevalence and type of M. xenopi infections.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Sequence-based typing of clinical isolates of Mycobacterium xenopi isolated from Ontario, Canada

Research Articles

Acknowledgements

The authors would like to gratefully acknowledge the members of the Genomics Core Laboratory at Public Health Ontario Laboratories.

References

1. Al Houqani M, Jamieson F, Chedore P, Mehta M, May K, Marras TK. Isolation prevalence of pulmonary nontuberculous mycobacteria in Ontario in 2007. Can Respir J 2011 JAN-FEB;18(1):19-24. 2. Al Jarad N, Demertzis P, Jones D, Barnes N, Rudd R, Gaya H, et al. Comparison of characteristics of patients and treatment outcome for pulmonary non tuberculous mycobacterial infection and pulmonary tuberculosis. Thorax 1996 FEB; 51(2):137-139. 3. Huard R, Lazzarini L, Butler W, van Soolingen D, Ho J. PCR-based method to differentiate the subspecies of the Mycobacterium tuberculosis complex on the basis of genomic deletions. J Clin Microbiol 2003 APR;41(4):1637-1650. 4. Van Ingen J, Boeree MJ, de Lange WCM, Hoefsloot W, Bendien SA, Magis-Escurra C, et al. Mycobacterium xenopi clinical relevance and determinants, the Netherlands. Emerg Infect Dis 2008 MAR;14(3):385-389. 5. Griffith DE, Girard W, Wallace R. Clinical-Features of Pulmonary-Disease Caused by Rapidly Growing Mycobacteria - an Analysis of 154 Patients. Am Rev Respir Dis 1993 MAY;147(5):1271-1278. 6. Jones B, Young S, Antoniskis D, Davidson P, Kramer F, Barnes P. Relationship of the Manifestations of Tuberculosis to Cd4 Cell Counts in Patients with Human-ImmunodeficiencyVirus Infection. Am Rev Respir Dis 1993 NOV;148(5):1292-1297. 7. Huang C, Tsai Y, Wu H, Wang J, Yu C, Lee L, et al. Impact of non-tuberculous mycobacteria on pulmonary function decline in chronic obstructive pulmonary disease. Int J Tuberc Lung Dis 2012 APR;16(4):539-545. 8. Marras TK, Mehta M, Chedore P, May K, Al Houqani M, Jamieson F. Nontuberculous Mycobacterial Lung Infections in Ontario, Canada: Clinical and Microbiological Characteristics. Lung 2010 AUG;188(4):289-299. 9. Motiwala A, Li L, Kapur V, Sreevatsan S. Current understanding of the genetic diversity of Mycobacterium avium subsp paratuberculosis. Microb Infect 2006 APR;8(5):1406-1418. 10. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, and Kumar S. MEGA5: Molecular Evolutionary Genetics Analysis using Maximum Likelihood, Evolutionary Distance, and Maximum Parsimony Methods. Molecular Biology and Evolution 2011;28: 2731-2739. 11. Roth A, Fischer M, Hamid M, Michalke S, Ludwig W, Mauch H. Differentiation of phylo¬genetically related slowly growing mycobacteria based on 16S-23S rRNA gene internal transcribed spacer sequences. J Clin Microbiol 1998 JAN;36(1):139-147.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

RESEARCH

Functionalized surface enhanced raman scattering gold nanoparticles: Size correlation of optical and spectroscopic properties and stabilities in solutions Qi Jing G. Sun, Gilbert C. Walker Department of Chemistry, University of Toronto

Abstract

Gold nanoparticles can be used as surface enhanced Raman scattering (SERS) substrates to amplify the Raman scattering intensities of molecules adsorbed to their surfaces via localized surface plasmon resonance (LSPR). The utility of SERS probes in application depends on the design, therefore it is crucial to understand the impact of nanoparticle diameter on SERS scattering intensity, ease of functionalization, and longevity. Although the optical and spectroscopic properties of gold nanoparticles have been extensively investigated for chemical, bioanalytical, and biomedical applications, the study of the size correlation with such properties for nanoparticles in solutions had rather been limited. This study reports the functionalization of gold nanoparticles of various sizes, ranging from 30 to 90 nm, for use as SERS probes. With the intention of studying monomeric particles, we show that the surface plasmon resonance band and SERS intensities red-shifted and increased with particle size, respectively. Further evaluation revealed that the functionalized particles remained stably dispersed, with no detectable change in SERS signal intensities for up to seven days of storage in biologically-relevant solutions. These findings serve as a promising basis for the further development and exploitation of diagnostically relevant and therapeutically important SERS probes.

Introduction

There is currently a crucial need in biophotonics research for the optimization of the stability, biocompatibility, sensitivity, specificity and information content provided by the optical labels. An emerging optical technology exploited extensively for biomedical and therapeutic applications is Raman scattering-based imaging. Although traditional fluorescence-based sensing and quantum dots have been predominantly utilized in optical studies, the two strategies exhibit very limited detection capacities and sensitivities [1, 2]. Surface enhanced Raman scattering (SERS)-based detection schemes offer several unique advantages over the conventional approaches. Similar to fluorescence, Raman scattering produces distinct spectra over a range of wavelengths from monochromic light, in accordance to the vibrational mode of the molecule [3]. On average, Raman emissions peaks have been reported to be 30 times narrower in width in comparison with those of fluorescence of quantum dots [3-5]. The sharp Raman vibrational bands greatly reduce spectral overlap and nonspecific molecular signals even for labels with similar chemical structures, which offers an immense potential for multiplexing [1-6]. Raman scattering does not involve electrons that are not excited in the process; the reporter molecules do not undergo electronic excitation and are inherently photostable upon repeated and prolonged illumination [1-4]. Recent interest in

the functionalization of metallic nanoparticles with biomolecules has led to the development of novel analytic tools based on Surface Enhanced Raman Scattering [7]. While Raman scattering signals are incredibly weak, as typically only one in 104 photons Raman scatter, SERS can exponentially amplify the intensity of the Raman signal [3]. It was discovered that a Raman reporter in close proximity to a plasmonic surface enhances the Raman scattering intensity by a factor of ~103 [3, 8]. In addition, these reporters continuously emit strong signals even when used in conjunction with traditional pathology stains [9]. As a result of the high signal enhancement due to the proximity of nanostrctured metal surfaces, SERS has been increasingly exploited to target trace amounts of biologically relevant molecules in a number of studies [1, 4, 10]. Currently, SERS was found to be the only sensitive technique capable of detecting a single molecule and examining the chemical structure simultaneously [10]. Gold nanoparticles can be used as surface enhanced Raman scattering (SERS) substrates to amplify the Raman scattering intensities of molecules adsorbed to their surfaces via localized surface plasmon resonance (LSPR) [3, 11]. Despite continuous investigation on the optical and spectroscopic properties of gold nanoparticles for its wide range of applications, the knowledge in the particle size correlation with these properties in solution had

Journal of Undergraduate Life Sciences â&#x20AC;˘ Volume 7 â&#x20AC;˘ Issue 1 â&#x20AC;˘ Spring 2013

Functionalized Surface Enhanced Raman Scattering Gold Nanoparticles: Size Correlation of Optical and Spectroscopic Properties and Stabilities in Solutions

Research Articles

Figure 1: Schematic illustration of the SERS nanoparticle preparation. Raman tag malachite green isothiocyanate (MGITC) was physio-adsorbed onto the gold surface via electrostatic interaction. The gold nanoparticles were further stabilized with a layer of thiol-terminated polyethyleneglycol (mPEG-SH).

rather been limited [11]. Recent developments have indicated that the enhancement of SERS intensity can be tuned by altering the nanoparticle size, the nanoparticle-substrate distance, the composition of SERS-active complex and its surrounding environment, and the assembled structure of nanoparticle-aggregates [12, 13]. In addition, the tuning of surface plasmon resonance (SPR) bands has been demonstrated across the visible and near-infrared range of the electromagnetic spectrum [14]. Studies of optical properties of metallic nanoparticles have also revealed that the SP band correlates with the particle size and shape [11, 13, 15]. The applicability of SERS probes essentially depends on its assembly, making it critical to understand the relation of particle diameter with SERS scattering intensity, ease of functionalization, and longevity. The aim of this study is to produce and characterize functionalized SERS gold nanoparticles and study the effect of their sizes on surface plasmon (SP) and SERS mechanisms. We report the successful functionalization of gold nanoparticles of different sizes, ranging from 30-90 nm, for use as SERs probes. Solution studies are the most relevant to the biological and therapeutic applications of SERS probes. We examine and monitor the SERS spectra, stabilities and aqueous solubility of the functionalized gold nanoparticles over time in different biologically-relevant solutions. With the intention of studying monomeric particles, we demonstrate the SERS particle biocompatibility in solution suspensions.

Materials and Methods

in Figure 1, the SERS activity was imparted to the nanoparticles by electrostatically adsorbing the positively-charged ionic reporter molecules to the negatively-charged citrate-coated gold surface. The MGITC-gold nanoparticles and mPEG-SH suspensions were mixed in a 5:2 volume ratio. Coating the particles with a stabilizing layer of mPEG-SH renders the nanoparticles monodispersed in solution and maintains the SERS enhancement. Excess mPEG-SH and unbound dye molecules were removed from suspension by centrifugation. All samples of the functionalized particles were resuspended to an equal concentration and stored at 4°C in 18.2 MΩ•cm water, 1xPBS-1%BSA (pH 7.4), and Tris Buffer (pH 7.4).

Raman Spectroscopy An inverted microscope (Nikon TE2000) was used to focus the cw 632.8 nm HeNe (15 mW) laser beam onto the sample in an episcopic configuration. The laser beam was collimated before entrance into the optics of the objective (S Plan Fluor ELWD 40X, NA 0.6). Rayleigh scattering from the sample at the wavelength of the laser line was blocked from entering the monochromator by a notch filter (λ > 645 nm). An achromatic doublet lens (f 6.6) focused the Raman scattered light on to the monochromator slit for spectral separation. The monochromator was connected to a charge-coupled device detector (PIXIS BR 400; Princeton Instruments, Acton, Massachusettes) with 1340×400 pixel array that was Peletier cooled to -75°C. Spectra were collected using WinSpec/32 software, with integration times ranging from 5-30 seconds. All spectra were expressed as absolute Raman intensity in counts on the y-axis.

Reagents

UV-Visible Spectroscopy

Purified water (MilliQ, 18 mOhms) was obtained from Barnstead EASYpure II Ultra Pure Water System. Citrate-stabilized gold nanoparticles (30 nm, 60 nm, and 80 nm) were purchased from Ted Pella Inc. (Redding, CA, USA). Citrate-stabilized 90nm gold nanoparticles particles were purchased from cYtodiagnostics. According to the manufacturers, the particles are provided at concentrations of 2.0×1011, 2.6×1010, 1.1×1010 and 5.37×109 particles/mL, respectively. All the particles were used without further processing, except for the 90 nm gold nanoparticles, where the stock solution was washed once prior to use. Malachite green isothiocyanate (MGITC) from Invitrogen was used without further purification. mPEG-SH (MW 5 kDa) from Rapp polymers was employed without further processing.

Stock gold colloid of all sizes and purified functionalized particles were placed in a 1 cm-path-length cuvette. A HP 8452A Diode Array Spectrophotometer was used to collect the absorption spectra of the particles using 18.2 MΩ•cm water as reference. Spectra were normalized by shifting the baseline to zero and scaling the data linearly, setting the maximum value for absorbance as one. The spectra of two samples from the stability examination experimental section were smoothed using Igor Pro software.

Preparation of SERS Nanoparticles The functionalization was accomplished by physically mixing the aqueous Raman-active dye, malachite green isothiocyanate (MGITC), with the gold colloid solution for 15 minutes. As schematically illustrated

Dynamic Light Scattering DLS was used to evaluate the hydrodynamic sizes of the functionalized SERS gold nanoparticles. DLS measurements were performed using a DynaPro/Protein solutions DLS machine (Wyatt Technologies Corporation, Santa Barbara, CA, USA). The data were processed using Dynamics software version 6.7.1. The particle dilution and laser power were adjusted for the optimal signal. Typically, particle concentrations were diluted by 1:25 compared to stock concentrations. The thermal stator was set to 25°C.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Results

Functionalized Surface Enhanced Raman Scattering Gold Nanoparticles: Size Correlation of Optical and Spectroscopic Properties and Stabilities in Solutions

PEG Functionalization of Gold Nanoparticles The UV-vis absorption spectra confirmed the monodispersity of the functionalized SERS gold nanoparticles (Figure 2). The presence of the bulky PEG layer on the gold surface can be shown by the typical 2 nm red-shift in the main localized surface plasmon resonance (LSPR) absorption wavelength [1, 9], for the functionalized particles compared to the stock. Red-shifts in the LSPR absorption wavelength were observed in the UV-vis spectra for all the gold nanoparticles. No absorption peak was detected in the longer wavelength range from 600 to 700 nm. The DLS measurements indicated an average of 10 nm increase in the hydrodynamic radius of the particles upon pegylation. The amount of increase in the radii affirmed the functionalization and monodispersity of the SERS gold nanoparticles.

Raman Scattering by the Functionalized Particles The successful syntheses of Raman-active nanoparticles of 30 nm, 60 nm, 80 nm and 90 nm are demonstrated in Figure 2. No detectable Raman spectrum was obtained from illuminating the dissolved MGITC alone in aqueous solution, using the same laser power and integration time (data not shown). Conversely, the SERS signal of MGITC-labeled gold nanoparticles was detectable and intense, as the gold particles provided significant enhancement in the scattered signal from the Raman reporter. Greater magnitude of enhancement can be achieved with nanoparticle multimers or aggregates. The resultant SERS signals were reproducible, with monodispersed particles being the primary species contributing to the SERS spectra shown in Figure 2.

Figure 2: Characterization of functionalized SERS nanoparticles of different sizes: (A) 30 nm; (B) 60 nm; (C) 80 nm; and (D) 90 nm malachite green isothiocyanate (MGITC) pegylated gold nanoparticles. Column (I) SERS spectra of single MGITC-PEG-coated nanoparticles showing strong SERS spectra with consistent peak positions characteristic of MGITC. (II) Normalized UV-vis absorption spectra of stock (gray, solid), MGITC-PEG-coated particles (black, solid), and free MGITC in aqueous solution (grey, dashed). LSPR absorption peaks for stock particles and for MGITC-PEG-coated particles were labelled as shown. (III) DLS histograms of the hydrodynamic radius for stock particles (shaded) and functionalized particles (outline).

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

2.00E+09! Properties and Stabilities in Solutions Functionalized Surface Enhanced Raman Scattering Gold Nanoparticles: Size Correlation of Optical and Spectroscopic 1.80E+09!

30nm(PEG4MG4Au( 60nm(PEG4MG4Au( 80nm(PEG4MG4Au( 90nm(PEG4MG4Au(

Research Articles

Intensity((counts)(

Abs(

Intensity((counts)(

Particle Size Correlation of 2.00E+09! 1.60E+09! SERS Properties and SP Bands 30nm(PEG4MG4Au( 1.80E+09! 60nm(PEG4MG4Au( The high monodispersity 1.40E+09! 80nm(PEG4MG4Au( of the functionalized nanopar1.60E+09! 90nm(PEG4MG4Au( 1.20E+09! ticles allowed for a parallel 1.40E+09! correlation of the particle size 1.00E+09! with the intensities of surface 1.20E+09! 8.01E+08! enhanced Raman scattering (SERS) and surface plasmon 1.00E+09! 6.01E+08! resonance (SPR) band. When 8.01E+08! normalized by the total surface 4.01E+08! area available for dye adsorp6.01E+08! 2.01E+08! tion, the maximum Raman intensity increased with particle 4.01E+08! 1.00E+06! sizes; they were found to be 900! 1000! 1100! 1200! 1300! 1400! 2.01E+08! Raman(Shi3((cm41)( 3.83×107, 3.74×108, 6.15×108 Figure 3. Compiled SERS Spectra of malachite green isothiocyanate ( and 1.92×109 counts for 30, 60, 1.00E+06! of1300! different 1400! sizes in aqueous solution, accord 80, and 90 nm gold nanopar900! 1000! 1100! nanoparticles 1200! 1500! 1600! normalized 1700! 2 41)( Raman(Shi3((cm solution, unit in counts per second per unit surface area in m . ticles, respectively (Figure 3). Figure 3: 3. Compiled SERSSERS SpectraSpectra of malachite green isothiocyanate (MGITC) from functionalized nanoparCompiled of malachite green isothiocyanate (MGITC) fromgold functionalized gold In addition, the SERS signal Figure ticles of different sizes in aqueous solution, normalized according to total surface area of particles in area 1 mL of of particles nanoparticles of different sizes in aqueous solution, normalized according to total surface for NPs greater than 60 nm in solution, unit in counts per second per unit surface area in m2. 2 diameter was clearly detectable. solution, unit in counts per second per unit surface area in m . Figure 4 shows accompanying increases in the maximum absor1! 30nm(f4Au( bance wavelength of the SP band from 524 to 558 nm with increases 60nm(f4Au( in sizes of the nanoparticles from 30 to 90 nm. As the particle size 0.8! 80nm(f4Au( increases, the red-shift is also associated with a small broadening 90nm(f4Au( 1! region. For the 30 nm par30nm(f4Au( of the SP band in the longer wavelength 0.6! 60nm(f4Au( ticles that showed a λmax at 524 nm, the absorbance of the SP band at 0.8! 80nm(f4Au( 0.4! wavelengths greater than 630 nm was minimal. In contrast to the 30 90nm(f4Au( nm gold particles, particles larger than 0.6!60 nm displayed absorbance 0.2! detectable in the wavelength range over 630 nm. The SERS intensity for the main 0.4! peak positions of MGITC, in0! 400! 450! 500! 550! 600! 650! 700! 750! 800! cluding 929 cm-1, 1183 cm-1, 1231 cm-1, 1305 cm-1, 1591 cm-1, 1591 0.2! -1 -1 Wavelength((nm)( cm , and 1618 cm were plotted against the particle size and the Figure 4: Normalized optical UV-visUV-vis absorption spectra for functionalmaximum absorbance wavelength of the Figure 4. Normalized optical absorption spectra for functional 0! SP bands for each particle ized gold nanoparticles of different sizes in aqueous solution. Thegold surf 400! 450! 500! 550! 600! 650! 700! 750! 800! (Figure 5). The correlation between SERS intensity and particle size aqueous solution. The spectra show a gradual red shift of the spectra show a gradual red shift of the gold surface plasmon band as the Wavelength((nm)( corresponds to that for SERS intensity and maximum absorbance particle of of thetheSPR of30, 30,60,60, 80, an increases. The respective λmax values size increases. The respective λmax values SPR band band of wavelength of SP. The maximum absorbance wavelength of SP Figure 4. Normalized optical UV-vis absorption spectra for functionalized gold nanoparticles of different 536nm, 556nm, and 572nm, respectively. 80, and 90 nm gold nanoparticles were 526 nm, 536 nm, 556 nm, and 572 band at 572 nm showed the highest enhancement Raman scat-show aqueous solution.inThe spectra a gradual red shift of the gold surface plasmon band as the particle siz nm, respectively. tering for each of the peaks investigated here.The respective λmax values of the SPR band of 30, 60, 80, and 90nm gold nanoparticles were 52 increases. the results for the addition of PEG to particle surfaces have 536nm, 556nm, and 572nm, respectively. Colloidal Stability in Biologically-Relevant Solutions been widely reported [1, 16]. In addition to the 2 nm red-shift, The stability and robustness of the functionalized particles the formation of the PEG layer on the SERS gold nanoparticle was explored by monitoring the SERS signals and UV-vis spec- surface can be affirmed by the 10 nm average increase in the tra over time. In Figure 6, the Raman spectra of the MGITC- particle hydrodynamic radius. The PEG layer kept the assembled pegylated-gold nanoparticles showed no significant sign of signal nanoparticles stable in solution and protected them from agchange, even after seven days of storage at 4°C in each of the three gregation, most likely via steric shielding. The stability of the biologically-relevant solutions, including water, PBS/BSA, and functionalized gold nanoparticles demonstrated from both the Tris buffer. In addition, it was found that the pegylated particles UV-vis and DLS measurement were consistent, confirming that remained monodispersed and resisted aggregation effectively for the particles remained isolated as monomers in aqueous soluup to seven days in all the solutions. Stability over longer periods tion. The association of the Raman-active molecules with the of time was not tested. metal particle was demonstrated, even subsequent to multiple rounds of purification. The SERS spectra of the functionalized Discussion gold nanoparticles all displayed consistent distinctive peaks corEvidence demonstrating the successful assemblage of func- responding to the vibrational modes of MGITC, which has been tionalized SERS gold nanoparticles of 30, 60, 80, and 90 nm in reported previously in literature [1, 9, 16, 17]. The anchorage size has been shown. The characteristic 2 nm red-shift in the of! the reporter molecules on the gold surface is believed to be SP peak of the particles is in agreement with literature, where mediated via the sulfur atom in the high-affinity isothiocyanate

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

! Research Articles

!!!!

Functionalized Surface Enhanced Raman Scattering Gold Nanoparticles: Size Correlation of Optical and Spectroscopic Properties and Stabilities in Solutions

15000

9 2 9 c m -1

In te n sity (c o u n ts )

1 1 8 3 c m -1 1 2 3 1 c m -1

10000

1 3 0 5 c m -1 1 3 7 3 c m -1 1 5 9 1 c m -1

5000

1 6 1 8 c m -1

0 40

P a r t ic le D ia m e t e r ( n m )

Figure 5: Plots of SERs intensity vs particle size and surface plasmon maximum absorption wavelengths (inset); only the main representative Raman peaks of MGITC were used, each are labelled by its respective symbol as shown.

Figure 5. Plots of SERs intensity vs. particle size and surface plasmon maximum absorption wavelengths (inset); only the main representative Raman MGITC werenanoparticle used, each are labelled by its respective symbolwill as provide group (-N=C=S) [18, 19]. The intense SERS peaks signalsofindicate that surface is necessary. Such evaluation the MGITC shown. molecules were protected and not displaced by the further insight into the observed pattern in the correlation beadsorption of the thiol-PEG. Such finding is in agreement with Qian et al. (2008), and supports the notion that the reporter and thiol-PEG molecules interact with the gold surface at noncompetitive sites [20]. It has been reported that the surface structure of isothiocyanate derivatives is “locked” in its π-conjugated form [19]. Hence, both the electronic structure and electrostatic interactions are expected to have contributed to the strong dye adsorption at the gold surface. In the size correlation analysis, we showed an increase in magnitude of Raman signal enhancement with particle size. Since the SERS signal for functionalized gold nanoparticles 60 nm or great was clearly detectable, there may be a critical particle size in solution beyond which the interaction between particles becomes cooperative in producing the SERS effect [11]. The wavelengths of the SP bands are dependent on the particle size. The correlation between particle size and the maximum absorbance wavelength of SP support the validity of using the SPR wavelength as one measure for particle sizes [11]. Also, the SP bands showed intensified absorbance in the longer-wavelength region (>630 nm) in close proximity to the laser wavelength (632.8 nm), especially for particles no less than 60 nm in size. Such findings collectively support that the SERS effect of the nanoparitcles can be mainly attributed to the effective coupling of the wavelength of the laser with the SP band, which needs to be further verified by investigating the correlation of the SERS spectra with the laser wavelength. The surface chemistry aspects are beyond the scope of this study and were not assessed for the assembled gold nanoparticles. As a result, the concentration or number of the Raman reporter and PEG molecules adsorbed onto the gold nanoparticle were unknown. A systematic quantitative characterization on the occupancy and identification of the chemical species on the gold

tween SERS intensity and particle size. Ultimately, the utility and the effectiveness of the nanoparticle composites are dependent on the ability to assemble the surface species in a well-controlled and predictable manner. To evaluate the colloidal stability, the functionalized SERS gold nanoparticles were subjected to treatments in biologicallyrelevant solutions, including deionised water, PBS/BSA and Tris buffer. SERS spectroscopy was used to detect changes in the Raman scattering intensity of MGITC for the functionalized SERS gold nanoparticles. Precipitation of the particles or dissociation of the Raman reporter molecules from the particles would result in a reduction in Raman intensity. The lack of detectable changes in SERS signals indicated the maintenance of particle stabilities over a period of seven days upon storage in the three solutions. The minor changes in the signal to noise ratio in Tris buffer can be explained by protonation and orientation changes of the reporter molecule relative to the gold surface. Furthermore, the stable monomeric dispersion of the assembled SERS particles was supported by the absence of a second SP absorbance peak in the red wavelength region (>600 nm), because a peak at beyond 600 nm would signify the generation of particle aggregates in solution. The relative red-shift of UV-vis bands observed for the 90 nm SERS nanoparticles can be explained by the formation of the PEG layer and weak interaction between the surface molecules at the gold surface. Such minor red-shifts alone are not sufficient to indicate a generation of gold particle aggregates in the solution. Altogether, the findings from SERS spectroscopy and UV-vis measurements demonstrated the capability for the SERS particles to withstand physiological environments. However, an assessment on the robustness of the functionalized SERS particles under different ionic or pH conditions is deferred to a future investigation.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Functionalized Surface Enhanced Raman Scattering Gold Nanoparticles: Size Correlation of Optical and Spectroscopic Properties and Stabilities in Solutions

(I)

(II)

(A)

! 1.0

0.8!

Abs(

Abs

30nm_Au t_day 1 t_day 3 t_day 7

0.8 0.6

0.6!

0.4

0.4!

0.2

0.2! 0! 400!

0.0 400

500

600 Wavelength (nm)

700

800

90nm!Au! t_day!1! t_day!3! t_day!7!

450!

500!

550!

600!

650!

Wavelength((nm)(

700!

750!

800!

(B)

1.0

Abs

'30nm Au' t_day 1 t_day 3 t_day 7

0.8 0.6

0.8! 0.6!

0.4

0.4!

0.2

0.2!

0.0 400

90nm!Au! t_day!1! t_day!3! t_day!7!

Abs(

500

600 Wavelength (nm)

700

800

0! 400!

450!

500!

550!

600!

650!

Wavelength((nm)(

700!

750!

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

800!

! Articles Research

Functionalized Surface Enhanced Raman Scattering Gold Nanoparticles: Size Correlation of Optical and Spectroscopic Properties and Stabilities in Solutions

90nm Au t_day 1 t_day 3 t_day 7

0.8

0.6

Abs

1.0

30nm Au t_day 1 t_day 3 t_day 7

0.6

0.4

0.2

0.2 0.0

0.0 400

500

600

Wavelength (nm)

700

800

400

500

600

Wavelength (nm)

700

800

Figure 6: Raman and UV-Vis spectra of the functionalized (I) 30 and (II) 90nm gold nanoparticles, recorded a = on the day of assembly, and b = after Figure 6. Raman UV-Vis spectra of the functionalized (I) 30 page); and (II) nanoparticles, a = sign on of 3, and c = 7 days of storage at 4oCand in (A) Water (previous page); (B) PBS/BSA (previous and90nm (C) Trisgold Buffer. The data showrecorded no significant Raman signalthe change the particles. measured with theatequivalent integration are offset and on the y-axis Water;time, (B) and PBS/BSA; (C) Trisfor clarity. dayorofaggregation assembly,ofand b = afterAll3,spectra and c were = 7 days of storage 4oC in (A)

Buffer. The data show no significant sign of Raman signal change or aggregation particles. All important spectra wereSERS Conclusion diagnostically-relevant andof the therapeutically

measured with the equivalent integration time, and are offset y-axis for clarity. The assembly and functionalization of nanoparticles presented probes.onInthe order to optimize effectiveness of the SERS nanotags, the in this study( constitutes a platform for controlling and tuning the synthesis parameter for monomeric SERS particles with maximal surface properties and SERS spectra for gold nanoparticle. More Raman signal can be balanced in accordance with the toxicity and ( specifically, we report the successful functionalization of SERS cellular responses of different nanoparticle sizes. Further studies gold nanoparticles from 30 nm to 90 nm in size, where maximal will focus on the application of the SERS nanoparticles in cancer Raman signals were obtained from monomers in solution. Raman cells detection, which should involve the following three key steps: spectroscopy indicated that these particles had been successfully first, the formulation of a robust and controlled approach for the labeled by the malachite green isothiocyanate (MGITC) chromo- assemblage of nanoparticle-bioconjugate, second, the extension of phore. The data provided from UV-visible spectrometry and dy- the SERS nanotags in biomedical applications both in vitro and in namic light scattering were in agreement, indicating that the SERS vivo, and third, the implementation of a multiplex strategy to fully nanoparticles dispersed as monomers in aqueous solution due to maximize the detection capacities. the presence of thiolated (poly)ethylene glycol (PEG) attached on the gold surface. From the size correlation analysis with the optical Acknowledgements and spectroscopic properties of the gold nanoparticles, our results This work was funded by Biopsys, the Natural Sciences and indicated that both SERS intensity and maximum absorbance Engineering Research Council of Canada Strategic Network for wavelength of SP increase with particle size. Furthermore, the Bioplasmonic Systems and the Richard Ivey Foundation. The aufunctionalized gold nanoparticles 30 nm and 90 nm in size showed thors would like to thank Christina MacLaughlin for all her guidretention of stable Raman signal intensities and resistance to ag- ance and advice throughout the course of the research project, and gregation for a period of seven days when stored in the particular Shell Ip for helpful discussions and technical support with Igor Pro. biologically-relevant solutions used. There is currently a critical need for earlier, higher speci- References ficity and certainty detection methods in cancer prognosis. 1. MacLaughlin CM, Parker EPK, Walker GC, Wang C. Evaluation of SERS labeling of CD20 on CLL cells using optical microscopy and fluorescence flow cytometry. Nanomedicine: NBM, Nanoparticles offer high surface areas, allowing for conjugation 2013. 9:55-64. to many biological agents [20]. The significant signal enhance- 2. Kneipp J, Kneipp H, Rajadurai A, Redmond RW, Kneipp K. Optical probing and imaging of live ment of Raman scattering signals and the intrinsic optical prop- cells using SERS labels. J. Raman Spectrosc., 2009. 40:1-5. 3. Kneipp K. Surface-enhanced Raman scattering. American Institute of Physics, 2007. erties of SERS probes offers immense potential in enabling high 60(11):40-46. molecular sensitivity, specificity, and for multiplexing. The above 4. Zavaleta CL, Smith BR, Walton I, Doering W, Davis G, Shojaei B, Nata MJ, Gambhir SS. findings serve as a promising basis for the further exploitation of Multiplexed imaging of surface enhanced Raman scattering nanotags in living mice using

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Functionalized Surface Enhanced Raman Scattering Gold Nanoparticles: Size Correlation of Optical and Spectroscopic Properties and Stabilities in Solutions

Research Articles

noninvasive Raman spectroscopy. PNAS, 2009. 106(32): 13511-13516. 5. Lutz B, Dentinger C, Sun L, Nguyen L, Zhang J, Chmura AJ, Allen A, Chan S, Knudsen B. Raman Nanoparticle Probes for Antibody-based Protein Detection in Tissues. Journal of Histochemistry and Cytochemistry, 2008. 56(4):371-379. 6. Gellner M, Kompe K, Schlucker S. Multiplexing with SERS labels suing mixed SAMs of Raman reporter molecules. Anal Bioanal Chem, 2009. 394:1839-1844. 7. Zheng YB, Kiraly B, Weiss PS, Huang TJ. Molecule plasmonics for biology and nanomedicine. Nanomedicine, 2012. 7(5):751-770. 8. Zhu Z, Zhu T, Liu Z. Raman scattering enhancement contributed from individual gold nanoparticles and interparticle coupling. Nanotechnology, 2004. 15(3):357-364. 9. Nguyen CT, Nguyen JT, Rutledge S, Zhang J, Wang C, Walker GC. Detection of chronic lymphocytic leukemia cell surface marker using surface enhanced Raman scattering gold nanoparticles. Cancer Letters, 2010. 292:91-97. 10. Sabur A, Havel M, Gogotsi Y. SERS intensity optimization by controlling the size and shape of faceted gold nanoparticles. J. Raman Spectrosc., 2008. 39(1):61-67. 11. Njoki PN, Lim IS, Mott D, Park H, Khan B, Mishra S, Sujakumar R, Luo J, Zhong C. Size Correlation of Optical and Spectroscopic Properties for Gold Nanoparticles. J. Phys. Chem., 2007. 111(40):14664-14669. 12. Stewart A, Zheng S, McCourt MR, Bell SEJ. Controlling Assembly of Mixed Thiol Monolayers on Silver Nanoparticles to Tune Their Surface Properties. ACS Nano., 2012. 6(5):3718-3726. 13. Jain PK, El-Sayed IH, El-Sayed MA. Au nanoparticles target cancer. Nanotoday, 2007. 2(1):18-28. 14. Hu M, Chen J, Li ZY, Au L, Harland GV, Li X, Marquez M, Xia Y. Gold nanostructures: engineering their plasmonic properties for biomedical applications. Chem. Soc. Rev., 2006. 35(11):1084-1094. 15. Hao E, Schatz GC, Hupp JT. Synthesis and Optical Properties of Anisotropic Metal Nanoparticles. Journal of Fluorescence. 2004. 14(4):331-341. 16. Ip S, MacLaughlin CM, Gunari N, Walker GC. Phospholipid Membrane Encapsulation of Nanoparticles for Surface-Enhanced Raman Scattering. Langmuir, 2011. 27(11):7024-7033. 17. Pettinger B, Krischer K. Comparison of Cross-Sections for Absorption and surfaceenhanced resonance Raman Scattering for Rhodamine 6G at Coagulated Silver Colloids. J. Electron Spectrosc. Relat. Phenom., 1987. 45:133-142. 18. Ansari DO. Raman-encoded Nanoparticles for Biomolecular Detection and Cancer Diagnostics. Atlanta: Georgia Institute of Technology; 2008. 19. Qian X, Emory SR, Nie S. Anchoring Molecular Chromophores to Colloidal Gold Nanocrystals: Surface-Enhanced Raman Edidence for Strong Electronic Couping and Irreversible Structural Locking. J Am Chem Soc., 2012. 134(4):2000-2003. 20. Qian X, Peng X, Ansari DO, Yin-Goen Q, Chen GZ, Shin DM, Yang L, Young AN, Wang MD, Nie S. In-vivo tumor targeting and spectroscopic detection with surface-enhanced Raman nanoparticle tags. Nature Biotechnology, 2008. 26(1):83-90.

Undergraduate research in the Human Biology Program Human Biology is a collabora1ve undertaking of the Faculty of Arts & Science and the Faculty of Medicine. Our mul1disciplinary undergraduate programs integrate topics from the biological, medical and social sciences, as well as elements of the humani1es. A major goal of each of our programs is to encourage analy1cal thinking and independent discovery. To facilitate achieving this goal, we oﬀer various research project courses in Human Biology, Neuroscience, and Global Health during the regular academic year as well as the summer.

www.hmb.utoronto.ca

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

RESEARCH

Using a D2-dopamine receptor chimera to assay genes essential to psychoactive drug response in Baker’s Yeast Zelun Zhang1*, Alexander Tigert2*, Marinella Gebbia3, Corey Nislow3 University of Waterloo Queen’s University 3 Department of Molecular Genetics, University of Toronto * Authors contributed equally to content 1 2

Abstract

A powerful approach to understanding gene-drug interactions is to express human cell membrane receptors in baker’s yeast (S. cerevisiae) because of the experimental tractability of yeast and homology between human and yeast proteins. S. cerevisiae individuals were engineered to express the ligand binding domain of the human D2-dopamine receptor as a fusion on the amino terminus of the yeast mating receptor in order to quantify the efficacy of various psychoactive drugs. In addition, an examination into the side effects of drug treatments was conducted through a genome-wide screen. Experimental and control strains were mated with the Yeast Knockout Collection using Synthetic Genetic Array technology, and pools of deletion mutants expressing the construct were then treated with 55 serotonin uptake inhibitor and α2-adrenoceptor drugs to quantitatively measure receptor response via growth assays. Nine drug treatments were then selected to determine their effects on pools of individual mutant strains – PCR barcodes were amplified from treated pools of the D2-expressing and empty vector cultures and hybridized to microarrays. Gene set enrichment analysis of these data identified genes involved in mitotic progression as required for wild-type survival in (S)-MCPG; genes involved in phosphatidylinositol biosynthesis were required for survival in the presence of D2. Both of these results present aberrant interactions that require additional research. It is expected that the method from this study will provide a platform for further investigation into the direct and off-target effects of other psychoactive drugs, in addition to the further characterization of D2 and other human receptors.

Introduction

Modern advancements in chemical genomics have allowed researchers to streamline the process of drug discovery, yet the primary focus in early stages of development is still driven by a “one-drug, one-target” philosophy. Constructing massive libraries of potential compounds and conducting high-throughput genome-wide screens against selected targets is the technique of choice to achieve this goal [1-4]. While this method provides the ability to survey a multitude of drugs for a specific target, it does not take into account the complexity that exists between drugs and proteins in vivo. Testing for the off-target effects of a drug is usually delayed until the drug has progressed into preclinical mouse models or even further [5], by which time large monetary sums and research efforts have already been invested. Often, the potentially serious side-effects of a drug are not identified until this in vivo testing is performed [6-9]. The shortcomings of this approach can be seen in research for treatments of neurodegenerative diseases and depression, where achieving agonistic or antagonistic effects on a specific neuroreceptor target is the primary objective of drug development. One of the primary targets for neurological diseases is the D2-dopamine receptor. This

G-protein coupled receptor (GPCR) is involved in normal neural signaling; however upregulation and abnormal structure can result in diseases such as Parkinson’s disease, schizophrenia, and depression [10, 11]. Baker’s yeast, or S. cerevisiae, is a well-characterized yeast and useful eukaryotic model organism. Being a eukaryote, it possesses the membranous organelles necessary for the expression and function of a multitude of human genes that bacteria often cannot. Despite yeast not sharing all of the complexities of human cells, approximately 31-45% of all genes have strong human protein homologs [12]. This homology can be used to relate results from genome-wide screens in yeast back to human systems. Another powerful feature of S. cerevisiae is its diploid-haploid duality. In times of environmental stress, cells can sporulate, creating haploid spores with a mating type of either MATa or MATα [13]. Pheromones from the opposite “sex” bind to mating receptors and trigger a well-characterized mating pathway [14]. In our construct, when a ligand binds to the ligand binding domain of D2, the yeast mating transduction pathway is activated. This leads to a cascade of protein kinase activity resulting in the transcription of Fus1, which prepares the cell for mating and the activation of the

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Using a D2-Dopamine Receptor Chimera to Assay Genes Essential to Psychoactive Drug Response in Baker’s Yeast

CDC28/Cln complex which arrests the cell cycle in G1 [14]. This leads to a decrease in the growth of the overall culture, which can be measured using a spectrophotometer. The yeast mating process can also be exploited through Synthetic Genetic Array (SGA) – a recently developed novel technique. In this process, two strains of haploid yeast containing resistance markers are mated and through a series of selection steps, strains containing the desired genotype can be isolated [15]. The potential of SGA can be fully realized with the use of the “Yeast Knockout Collection” as genes or plasmids can be shuffled into this unique library in a parallel manner. For this collection, 96% of annotated Saccharomyces cerevisiae genes were systematically disrupted in an attempt to discover each gene’s effect on phenotype [16]. In each strain, one unique ORF (open reading frame) is replaced with the KanMX selectable marker by way of homologous recombination [17]. The collection contains approximately 5000 ORF deletions of genes non-essential to basic yeast growth and function [16]. Each strain’s KanMX cassette contains unique 20bp DNA “barcodes” [18]. These barcodes allow for the creation of pools containing all deletion mutants, growth of these pools in parallel in various conditions, and microarray hybridization of the pool’s PCR-amplified barcodes to determine how each strain fares when subjected to a selected condition. Strains with severe fitness defects can be determined to be missing a gene essential for survival to the specific treatment. Using these techniques, a procedure was devised to identify drugs that had the most direct effect on a selected receptor and the gene deletions that were implicated in the response to each drug. To demonstrate proof of concept for this method, a collection of serotonin uptake inhibitors and α2-adrenoceptor ligands were screened for their direct and off-target effects on our D2-dopamine receptor construct. The construct makes use of D2’s ligand binding domain and the Yeast Mating Pathway to quantitatively assay the agonistic or antagonistic effects of the drugs in these collections. Using SGA, the Yeast Knockout Collection was utilized to express this construct allowing for parallel growth of all strains in drug conditions. The relative abundance of a strain’s barcode after growth reflects the abundance of that strain and allows genes essential for response to drugs to be identified.

Materials and Methods

Research Articles

Figure 1: Immunofluorescence microscopy confirming dopamine receptor expression and localization in transformed S. cerevisae. The nuclei of the cells were stained blue with DAPI, while the construct’s reporter module is green as a result of the activity of the secondary antibody. In the top two images, where immunofluorescence of the empty vectors strains is displayed, diffuse cytoplasmic staining can be observed indicating background fluorescence and no specific D2 expression. In contrast, a concentration of green fluorescence is observed on the plasma membrane of the D2-2 and D2-3 samples, indicating correct D2 expression. [Magnification 630x] cultures were transformed with this construct in the following manner: D2-2 and D2-3 were 2 identical clones expressing the D2 ligand binding domain while EV-1 and EV-2 lacked this insert and acted as empty vector controls. Immunofluorescence microscopy was performed on all strains to confirm D2 expression and proper membrane localization in the D2-2 and D2-3 strains (Figure 1). A rabbit anti-VP16 polyclonal antibody (SigmaAldrich) followed by an Alexa-Fluor 488® conjugated donkey anti-rabbit IgG antibody (Invitrogen) was used in immunofluorescence to confirm D2 expression and localization. Cells were then imaged using a Zeiss Axiovert Microscope with a 63x objective oil-immersion lens. The Yeast Knockout Collection plates were also converted into 1536 density by an automated colony pinning robot (BioMatrix, S and P Robotics, Inc.).

Synthetic Genetic Array (SGA)

Strains

The Synthetic Genetic Array (SGA) protocol used both the query strain Y7092 (MATα bni1D::natR can1Δ:: MFA1pr-HIS3 lyp1Δ ura3Δ0 leu2Δ0 his3Δ1 met15Δ0), which, during mating, crossed the D2 construct plasmids into the Yeast Knockout Collection strain (MATa xxxΔ::kanR leu2Δ0 met15Δ0 ura3Δ0). SGA was performed on 1536 density assay microplates containing 4 replicates of 384 strains. Both strains were obtained from the lab of Dr. Charles Boone at the University of Toronto [17, 19]. The two strains contained antibiotic resistance markers that allowed for the selection of double mutants during the SGA process. Two cultures of Y7092 were transformed with the empty vector pCCW-STE (Dualsystems Biotech), while another two strains were transformed with vectors containing the cDNA for the ligand binding domain of the D2 receptor. These vectors expressed this domain as an amino-terminal fusion on the reporter module Cub-LexA-VP16 [20, 21]. Since the ligand binding domain was coupled to the yeast mating pathway, activation could be quantitatively assayed in these chimeric collections by reading growth as an output. Four

The SGA method [15] was used to create a collection of double mutants that would express the D2 and empty vectors in all of the deletion strains of the DMA (Figure 2). Prior to the first mating step, lawns of all four strains were created by spreading 10 mL of overnight culture onto SD plates lacking leucine and incubating for 3 days at 30°C. The media composition for every step thereafter was modified to select for the pCCW-STE plasmid in every selection step by preparing the required media without leucine. The entire process was carried out using the BioMatrix Colony Processing Robot (S&P Robotics) and completed in quadruplicate for all four strains. Due to identical behavior from the biological replicates EV1 and EV2 as well as D2-2 and D2-3, EV1 and D2-2 were selected to proceed with pooling. Both sets of final plates contained a copy of the DMA in quadruplicate with one collection expressing D2, while the other acted as a negative control. These collections were independently pooled and normalized for later genomic screens with selected drugs. DMSO was added to a final concentration of 7% (v/v) for storage at -80°C.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Using a D2-Dopamine Receptor Chimera to Assay Genes Essential to Psychoactive Drug Response in Baker’s Yeast

Figure 2: Schematic overview of the construction of the chimeric D2 – yeast knockout collections. In order to create two collections of the Deletion Mutant Array (DMA), one expressing the D2-ligand binding domain and one that does not, Synthetic Genetic Array was utilized to create double mutants. Lawns of EV1 and D2-2 were grown and repinned into 1536 format. These plates were then mated on YPD with the entire DMA. Through a number of sporulation and selection steps, which are described in detail in Tong and Boone [15], the final plate sets were created to contain mutants expressing the gene deletions in addition to the D2 or control construct. This provided the platform of custom collections pools for completing parallel screens.

Drug Screens To select drugs for further genetic analysis, collections of serotonin uptake inhibitors and α2-adrenoceptor ligands from Tocris Bioscience (Appendix 1) were systematically screened against all four strains to determine their potency in affecting cell growth both with and without D2 expression. A 96-well microplate was inoculated with samples of all 4 strains at an OD600 of 0.0625. In total, 55 drugs were added, each in 2% DMSO and at a final drug concentration of 200 μM. Plates were then grown for 16 hours at 30°C in a shaking spectrophotometer (Tecan). The OD600 of every well was taken every 15 minutes and growth curves were generated using ACCESS [22]. Based on their growth curves, the drugs were divided into four main categories – cultures that grew more than one standard deviation (calculated relative to each plate) than the average for more than two trials, cultures that grew less than one standard deviation, ones that grew either more than one standard deviation in one trial and less in another or ones that exhibited non logistic-shaped curves (based on less than 0.5 value of R2 when fit with curve). Drugs that were chosen for gene-drug interaction testing were randomly selected from the drug tests where the culture grew more than one SD and less than one SD from the average, with both categories having an equal number of chosen drugs.

Gene-Drug Interaction Analysis Using the genetically-barcoded Yeast KnockOut Collection, samples of the EV1 and D2-2 pool created by SGA were grown in the presence of identified drugs (LY 367385, DL-TBOA, ATPA, MPEP HCl, (S)-MCPG, Prazosin HCl, Guanfacine HCl, UK14,304, LY 341495) to determine gene-drug interactions [23]. Individual wells of a 48 well microplate were inoculated at an OD600 of 0.0625 and drugs were added to a final volume of 200 μM in a total volume of 700 μL per well. 2% DMSO (final concentration) was used as a control. Cultures were again grown in a shaking spectrophotometer (Tecan), this time for 200 reads to allow for fitness defects of individual strains to be detected. Genomic DNA was extracted using reagents from Zymo Research followed by a standard chloroform extraction [24]. The barcodes were then amplified by PCR and hybridized to the corresponding TAG4 array (Affymetrix) [25]. After washing and staining,

arrays were scanned and fluorescence values obtained for each of the DNA probes on the array. Values were quantile normalized to remove background fluorescence and omit outliers [26]. Log2 ratios were then calculated as a measure of a particular strain’s fitness when exposed to drug with D2 expression compared to without D2 expression [26]. Gene lists compiled from each of the arrays were enriched using a standard hypergeometric test. P-values were calculated and genes with p<0.0001 were noted. Gene ontology categories containing 20 or more genes marked as significant were included in our schematic diagrams. Gene Ontology (GO) maps were then generated using GOstats, with the minimum threshold of 20 significant genes, for GO categories to be included in the schematics [27].

Results

Drug Assay (Figure 3) Of the 55 serotonin uptake inhibitor and α2-adrenoceptor drugs tested, the exhaustive list of drugs that had a significant delay in the growth curve includes MPEP HCl, (S)-MCPG, and prazosin HCl, Cirazoline HCl, BMY 7378 dihydrochloride, guanfacine HCl, UK 14 304 and (S)-3,5-DHPG. These drugs all exhibited a less than one standard deviation of difference over the plate control more than once. An exhaustive list of drug treatments that had increased growth of at least one standard deviation in the growth curve compared to the control was compiled – they include LY 367385, LY 341495, SYM 2081, DL-TBOA, NKH 447, and ATPA. Furthermore, drugs that both grew greater than the control in one trial and less on the other or had growth that did not resemble a standard logistic-shaped growth curve were also noted. Exhaustively, these included BRL 44408 maleate, rauwolscine HCl, isoproterenol HCl, sotalol HCl, kainic acid, NBQZ disodium salt, cAMPS-Rp thriethylammonium salt, 8-bromo-cAMP sodium salt, DCG IV, L-AP4, and forskolin. Drugs that did not exhibit any significant changes from the control include, exhaustively, 6-nitroquipazine maleate, cilostamide, zardaverine, milrinone, (R)-(-)-Roilpram, ICI 118 551 HCl, Ro 20-1724q, (R)-(-)-α methylhistamine dihydrobromide, iodophenpropit dihydrobromide, thioperamide maleate,

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Using a D2-Dopamine Receptor Chimera to Assay Genes Essential to Psychoactive Drug Response in Baker’s Yeast

Research Articles

Figure 3. Growth curves of drug treated cultures (Left: EV1, Right: D2-2). Each window shows the growth curve results of one 96 well microplate grown for 16 hours. These growth curves represent the increase in OD600 measured at 15 minute intervals. The red line is the curve of the plate average, while the black line is the curve for that specific well. Drug-treated cultures of all four cultures—EV1, EV2, D2-2, D2-3—were analyzed by growth curves in duplicate. Left: EV1, “empty vector” control strain lacking D2 expression. Right: D2-2, experimental strain expressing D2-2 construct. For comparison, the growth curves for 200µM prazosin HCl are shown from both strains.

L-quisqualic acid, BRL B37344 sodium salt, (2R, 4R)-APDC, DCG IV, EGLU, A61603 hydrobromide, L-(-)-threo-3-hydroxyaspartic acid, L-trans-2,4,-PDC, dihydrokainic acid, SYM 2081, ARC 239 dihydrochloride, UK 14 304, cyclothiazide, GYKI 52466 HCl, CNQZ disodium salt, and CGP 20712A dihydrochloride. Microarray Analysis (Figure 4) Individual gene deletions that caused a strain to grow significantly more with the D2-receptor versus the empty vector control were identified. For example, as a result of the (S)-MCPG drug treatment, GDB1 (responsible for glycogen catabolism), FKH2 (responsible for pseudohyphal growth), and AK1 (mediates response to stress) were amongst the several thousand individual genes with the lowest log2 ratios. Gene deletions that cause a strain to grow significantly less with the D2-receptor versus the empty vector control (high log2 ratio) were also identified - for example, results from the (S)-MCPG drug treatment highlighted BAT2 (allows branched chain family amino acid biosynthesis), RPT4 (essential in ubiquitin-dependent protein catabolism), and SAT4 (used by cells in G1/S transition of mitotic cell cycle) to be implicated in the cell’s response. Complete gene lists for tested antagonists and agonists were compiled for further reference, and available upon request at the Hip Hop Lab; it should be noted these exhaustive lists are not directly useful to the purpose of the study, as it is focused on identifying entire cell processes that are affected by these drugs. Gene Ontology Mapping (Figure 5) The GO maps created from the microarray data corroborate the results into a more succinct and visual format. The empty vector schematics (left panels) reflect the core cellular response to drug, whereas the D2 schematics (right panels) reflect the response

in the presence of the D2-receptor. Only GO maps for (S)-MCPG and prazosin HCl were compiled as sample tests. The enrichments for (S)-MCPG display that the cell manifests a growth defect when genes involved in mitotic progression are deleted in the presence of the D2-receptor. For the D2 experiment, (S)-MCPG treatment results in growth defects for mutant strains with genes involved in cellular response to nutrients. A large fraction of genes that comprise these GO categories code for other amine-based receptors. In the experiments involving Prazosin, it can be seen that Prazosin has relatively benign side-effects, with only genes required for basic morphogenesis being enriched. In the case of Prazosin for cells with the D2-receptor, the very specific GO term “phosphatidylinositol biosynthetic process” scored significantly as required for proper response with the D2-receptor.

Discussion

In interpretation of our drug assay results, a treated culture that has a significant delay in the growth curve compared to the control in more than two trials indicates that the drug has activated the mating pathway; therefore the receptor was activated to a greater extent than in the strain with the vector control, and we would classify these drugs as potential agonists of the D2-receptor. Alternatively, a treated culture that has increased growth in the growth curve compared to the control means the opposite; therefore these drugs were classified as antagonists of the receptor. For gene list and hybridization results, gene deletions that cause a strain to grow significantly less in the presence of the D2-receptor versus the empty vector control were identified. In these cases, the mating pathway is activated, signifying that the D2-receptor binds with high affinity to the drug. The compiled GO

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

ORF ID YDR283C::CHR4_6 YBR236C::CHR2_4 YDR180W::CHR4_5 YDR454C::CHR4_8 YDR488C::CHR4_8 YDL117W::CHR4_2 YDR310C::CHR4_6 YDR216W::CHR4_5 YJR139C::CHR00_12 YLR134W::CHR12_3 YDR308C::CHR4_6 YDR156W::CHR4_5 YDR206W::CHR4_5 YDL174C::CHR4_2 YNL189W::CHR14_2 YBR267W::CHR2_4 YHR202W::CHR8_3

Using a D2-Dopamine Receptor Chimera to Assay Genes Essential to Psychoactive Drug Response in Baker’s Yeast

LY367385 DL-TBOA MPEP (S)-MCPG Prazosin 200μM 200μM 200μM 200μM 200μM Gene ID 0.602 0 0.569 0.11 7.232 GCN2 0.128 -0.223 -0.072 0.411 6.15 ABD1 0.068 -0.598 -0.004 -0.321 6.103 SCC2 3.046 -0.171 0.449 2.868 6.098 GUK1 -0.599 0.071 -0.208 -0.818 5.989 PAC11 0.313 -0.043 0.153 -0.145 5.846 CYK3 1.574 0.821 1.481 0.164 5.716 SUM1 2.784 1.64 2.691 1.194 5.709 ADR1 -0.913 0.655 -0.161 0.533 5.613 HOM6 1.528 3.023 2.197 1.251 5.569 PDC5 0.013 -0.595 -0.063 0.322 5.532 SRB7 -0.742 2.427 -0.263 -0.134 5.501 RPA14 -0.312 -0.015 0.328 -0.121 5.483 EBS1 -0.034 2.583 -0.324 1.8 5.255 DLD1 -1.158 -1.255 1.724 -0.657 5.068 SRP1 -0.364 -0.104 0.077 -0.275 5.044 REI1 2.673 5.417 5.82 2.54 4.946 YHR202W

GO_process protein amino acid phosphorylation mRNA capping mitotic sister chromatic cohesion GMP metabolism nuclear migration, microtubule-mediated cytokinesis telomere maintenance transcription methionine metabolism pyruvate metabolism transcription from RNA polymerase II promoter telomere maintenance telomere maintenance aerobic respiration nucleocytoplasmic transport ribosomal large subunit biogenesis biological process unknown

Figure 4: Workflow for Analysis of Microarray Data. Upon completion of microarray hybridization and staining, every microarray was scanned to determine the intensity of each probe’s fluorescence. After normalization, gene lists containing log2 ratios and designated Gene Ontology categories were then compiled. Scatter plots graphing each strain’s log2 ratio were generated to visually represent sensitive strains. The microarray scan and data shown are from DL-TBOA – a serotonin uptake inhibitor.

maps for the trials with D2 expression suggest that these categories of genes are involved in the cell’s response to the binding of drug to D2. For example, our results for the (S)-MCPG treatment of the D2 strain brought the significance of other amine-based receptors to our attention (Figure 5). This result strongly suggests that dopa-

mine and related catecholamines that interact with the dopamine receptor interact with these other membrane receptors. This also provides evidence that (S)-MCPG interacts with dopamine receptors in addition to the ones it has previously been identified with such as the metabotropic glutamate receptor (mGlu) [28]. In ad-

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Using a D2-Dopamine Receptor Chimera to Assay Genes Essential to Psychoactive Drug Response in Baker’s Yeast

Research Articles

disorders as a common evaluation on the efficacy of current; newly developed drugs can thereby potentially be established. Its quantitative comparability can theoretically provide greater insight into the function of any receptor by quantifying its activation to various external conditions and treatments. The method for doing so would involve taking the ligand-binding domain of any other human receptor and placing it within the construct detailed in this experiment, and undergoing the identical steps while replacing the treatment with the experimental variable in question, such as a different variety of drugs. Coupling of Figure 5. Representative Gene Ontology Maps for (S)-MCPG and Prazosin. These GO maps represent the categories human GPCR’s to the yeast of genes coding for cell processes that are involved with the cell’s response to drug. From these results, we observe mating pathway has also been that strains with gene deletions that code for essential generic cell processes such as Cellular Response to Nutrient accomplished with the use of and cell division processes such as Regulation of Spindle Pole Body Separation face a significant fitness defect. G-protein chimeras; however dition, genes in the category of phosphatidylinositol biosynthesis only a few have been accomhave never been previously identified to interact with dopamine plished [32]. With the refinement of molecular techniques, researchreceptors [29]. This interaction thus warrants further investigation, ers have also discerned minute differences between the diverse possibly relating lipid bilayer composition to the activity of trans- classes of dopamine receptors. With different activity and expression membrane receptors. levels in various organs, results of pharmacological assays on differIn the trials using the empty vector collection, gene knockouts ent classes of dopamine receptors can tailor drugs to specific types. that manifested growth defects indicate that gene is important in Research into D1-specific ligands [33], for example, complements basal cell tolerance to psychoactive drugs as opposed to effects seen this research and opens up the possibility for a collective database of from the presence of the D2-receptor. The requirement of genes for ligands and their activity on various classes of receptors. mitotic progression in (S)-MCPG treatment seems fitting as in their absence, cells would be at a disadvantage. The data also suggests Conclusion that Prazosin has relatively benign side-effects, with only genes The experiment has demonstrated proof of concept for a new required for morphogenesis being enriched (Figure 5). This GO method of testing drug treatments and gene-drug interactions category is a common cellular response to drugs. So in the absence in the presence of human receptors. Since almost 45% of all curof D2, the effects of Prazosin are generally mild and non-specific, rently marketed therapeutics target cell surface receptors [5], this with the category of cell morphogenesis meeting those criteria. could have implications in the development of pharmaceutical By combining a drug assay with a genome-wide screen, the drugs. Entire collections of common anti-hypertensives such as direct effects on the D2-dopamine receptor-mediated and non- prazosin HCl and antidepressants such as (S)-MCPG were efreceptor mediated effects on the cell caused by diverse psychoac- ficiently assayed for their activity on the D2-dopamine receptor. tive drugs were determined. This allows for the quantitative com- Strong agonists and antagonists of D2 were then screened for parison of receptor activation and cell response, thereby indicating their interactions with other cellular processes. Interesting rethe degree of efficacy the drugs will have upon D2 as a therapeutic sults include the implication of prazosin in phosphatidylinositol target. The use of the pheromone pathway as described by Groß et biosynthesis, highlighting a possible avenue of further research. al. can be effectively utilized as a biosensor such that the readout (S)-MCPG treatments also interacting with other amine-based of ligand binding is a decrease in growth [30]. In addition, express- receptors during dopamine sensing is another side-effect that was ing GPCRs in yeast membranes and coupling them to the yeast uncovered. These results suggest that many drugs, perhaps even pheromone pathway has been used previously to determine mo- those approved for use, could have unknown side effects that are lecular interactions and ligand activity [31]. Our method provides detrimental or beneficial for other therapeutic targets. a unique application of such a ligand binding assay with the extenIdeally, the methods of this experiment will spur further study sion of a genome-wide interaction study. due to the emphasis of two key qualities – efficiency of testing, and Thus, this experiment provides the possibility for improvement quantitative comparability (Refer to Figure 3 and Figure 5). This in the development of antidepressants and drugs for neurological allows for relative comparisons of drug efficacy and provides the Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Research Articles

Using a D2-Dopamine Receptor Chimera to Assay Genes Essential to Psychoactive Drug Response in Baker’s Yeast

potential to streamline drug development even more by adding an extra pre-clinical step that would be able to screen out ineffective drugs before expensive further testing. In addition, the process has been demonstrated to be time-efficient – thousands of genes and hundreds of drugs can be tested within a reasonable timeframe of less than one year. Again, this concept can be extended beyond the D2-dopamine receptor, as the construct is hypothesized to work for any human receptor by ligating the ligand-binding domain of the target receptor instead of the D2-receptor into the construct. Thus, the results shown are a versatile proof of concept through having demonstrated specific drug and gene testing for the D2-dopamine receptor. It is hoped that this method can be adapted to assay drug candidates against a variety therapeutic targets while maintaining high-throughput capability and in vivo relevance.

Acknowledgements

Zelun Zhang and Alexander Tigert thank Dr. Marinella Gebbia and Dr. Corey Nislow for their continued mentorship over the duration of the project and the entire Hip Hop Lab for their support. Lastly, the pair also wishes to thank Dr. Danielle Gauci of Northern Secondary School for her guidance throughout the research process and manuscript formulation, and for providing much inspiration to conduct this research project.

brane protein complexes using the split-ubiquitin membrane yeast two-hybrid (MYTH) system. Methods Mol Biol, 2009. 548:247. 21. Wong V, Stagljar I. Map of the pCCW-STE plasmid. 2011 Oct 25;Email. 22. Proctor M, Urbanus ML, Fung EL, Jaramillo DF, Davis RW, Nislow C, et al., The automated cell: compound and environment screening system (ACCESS) for chemogenomic screening. Methods Mol Biol, 2011. 759:239. 23. Smith AM, Durbic T, Oh J, Urbanus M, Proctor M, Heisler LE, et al., Competitive genomic screens of barcoded yeast libraries. J Vis Exp, 2011.(54). 24. Zymo Research. YeaStar Genomic DNA Kit. 2011; Available at: http://www.zymoresearch. com/downloads/dl/file/id/13/d2002i.pdf. Accessed Jan 28, 2011. 25. Pierce SE, Fung EL, Jaramillo DF, Chu AM, Davis RW, Nislow C, et al., A unique and universal molecular barcode array. Nat Methods, 2006. 3(8):601. 26. Ericson E, Gebbia M, Heisler LE, Wildenhain J, Tyers M, Giaever G, et al., Off-target effects of psychoactive drugs revealed by genome-wide assays in yeast. PLoS Genet, 2008. 4(8). 27. Falcon S, Gentleman R, Using GOstats to test gene lists for GO term association. Bioinformatics, 2007. 23(2):257. 28. Hayashi Y, Sekiyama N, Nakanishi S, Jane DE, Sunter DC, Birse EF, et al., Analysis of agonist and antagonist activities of phenylglycine derivatives for different cloned metabotropic glutamate receptor subtypes. J Neurosci, 1994. 14(5 Pt 2):3370. 29. Gardocki ME, Jani N, Lopes JM, Phosphatidylinositol biosynthesis: biochemistry and regulation. Biochim Biophys Acta, 2005. 1735(2):89. 30. Groß A, Rödel G, Ostermann K, Application of the yeast pheromone system for controlled cell–cell communication and signal amplification. Lett Appl Microbiol, 2011. 52(5):521. 31. Price LA, Kajkowski EM, Hadcock JR, Ozenberger BA, Pausch MH, Functional coupling of a mammalian somatostatin receptor to the yeast pheromone response pathway. Mol Cell Biol, 1995. 15(11):6188. 32. Brown AJ, Dyos SL, Whiteway MS, White JHM, Watson MAEA, Marzioch M et al., Functional coupling of mammalian receptors to the yeast mating pathway using novel yeast/mammalian G protein α-subunit chimeras. Yeast, 2000. 16(1):11. 33. Neumeyer JL, Kula NS, Bergman J, Baldessarini RJ, Receptor affinities of dopamine D1 receptor-selective novel phenylbenzazepines. Eur J Pharmacol, 2003. 474(2-3):137.

References

1. Broach JR, Thorner J, High-throughput screening for drug discovery. Nature, 1996. 384(6604):14. 2. Bailey SN, Wu RZ, Sabatini DM, Applications of transfected cell microarrays in high-throughput drug discovery. Drug Discov Today, 2002. 7(18):S113. 3. Eldridge GR, Vervoort HC, Lee CM, Cremin PA, Williams CT, Hart SM, et al., High-throughput method for the production and analysis of large natural product libraries for drug discovery. Anal Chem, 2002. 74(16):3963. 4. Astle JM, Simpson LS, Huang Y, Reddy MM, Wilson R, Connell S, et al., Seamless bead to microarray screening: rapid identification of the highest affinity protein ligands from large combinatorial libraries. Chem Biol, 2010. 17(1):38-45. 5. Bleicher KH, Böhm HJ, Müller K, Alanine AI, Hit and lead generation: beyond high-throughput screening. Nat Rev Drug Discov, 2003. 2(5):369. 6. Wong ML, Licinio J, From monoamines to genomic targets: a paradigm shift for drug discovery in depression. Nat Rev Drug Discov, 2004. 3(2):136. 7. Waldmeier P, Bozyczko-Coyne D, Williams M, Vaught JL, Recent clinical failures in Parkinson’s disease with apoptosis inhibitors underline the need for a paradigm shift in drug discovery for neurodegenerative diseases. Biochem Pharmacol, 2006. 72(10):1197. 8. Knowles J, Gromo G, A guide to drug discovery: Target selection in drug discovery. Nat Rev Drug Discov, 2003. 2(1):63. 9. Brown D, Superti-Furga G, Rediscovering the sweet spot in drug discovery. Drug Discov Today, 2003. 8(23):1067. 10. Seeman P, Niznik HB, Dopamine receptors and transporters in Parkinson’s disease and schizophrenia. FASEB J, 1990. 4(10):2737. 11. Missale C, Nash SR, Robinson SW, Jaber M, Caron MG, Dopamine receptors: from structure to function. Physiol Rev, 1998. 78(1):189. 12. Botstein D, Chervitz SA, Cherry M, Yeast as a Model Organism. Science, 1997. 277(5330):1259. 13. Herskowitz I, Life Cycle of the Budding Yeast Saccharomyces cerevisiae. Microbiol Rev, 1988. 52(4):536. 14. Bardwell L, A walk-through of the yeast mating pheromone response pathway. Peptides, 2004. 25(9):1465. 15. Tong A, Boone C, Synthetic Genetic Array (SGA) Analysis in Saccharomyces cerevisiae. In: Xiao W, editor. Yeast Protocols. 2nd ed. Totowa, NJ: The Humana Press Inc.; 2006. p. 171-192. 16. Giaever G, Chu AM, Ni L, Connelly C, Riles L, Véronneau S, et al., Functional profiling of theSaccharomyces cerevisiae genome. Nature, 2002. 418(6896):387. 17. Winzeler EA, Shoemaker DD, Astromoff A, Liang H, Anderson K, Andre B, et al., Functional characterization of the S. cerevisiae genome by gene deletion and parallel analysis. Science, 1999. 285(5429):901. 18. Wach A, Brachat A, Pöhlmann R, Philippsen P, New heterologous modules for classical or PCR-based gene disruptions inSaccharomyces cerevisiae. Yeast, 1994. 10(13):1793. 19. Tong A, Boone C, High-Throughput Strain Construction and Systematic Synthetic Lethal Screening inSaccharomyces cerevisiae. In: Stansfield I, Stark M, editors. Yeast Gene Analysis. 2nd ed.: Elsevier Ltd.; 2007. p. 369-386. 20. Kittanakom S, Chuk M, Wong V, Snyder J, Edmonds D, Lydakis A, et al., Analysis of mem-

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Using a D2-Dopamine Receptor Chimera to Assay Genes Essential to Psychoactive Drug Response in Baker’s Yeast

Research Articles

Appendix 1

List of Drug Treatments Used in Screens Drug Name 6-nitroquipazine maleate Cilostamide Zardaverine Milrinone (R)-(-)-Rolipram Ro 20-1724q Toc drug (R)-(-)-α-methylhistamine dihydrobromide Iodophenpropit dihydrobromide Thioperamide maleate (S)-3,5-DHPG L-Quisqualic acid (S)-MCPG BRL B37344, sodium salt (2R, 4R)-APDC DCG IV LY 341495 EGLU A 61603 hydrobromide Cirazoline hydrochloride BMY 7378 dihydrochloride Prazosin hydrochloride Guanfacine hydrochloride UK 14,304 ARC 239 dihydrochloride Forskolin KT 5720 NKH 477 DL-TBOA LY 341495 (S)-AMPA Cyclothiazide GYKI 52466 hydrochloride CNQX disodium salt CGP 20712A dihydrochloride ICI 118,551 hydrochloride LY 367385 MPEP hydrochloride Isoproterenol hydrochloride BRL 44408 maleate Rauwolscine hydrochloride Sotalol hydrochloride Kainic acid ATPA SYM 2081 NBQX disodium salt cAMPS-Rp, triethylammonium salt 8-Bromo-cAMP, sodium salt Unknown L-(-)-threo-3-Hydroxyaspartic acid L-trans-2,4-PDC Dihydrokainic acid SYM 2081 DCG IV L-AP4

S = serotonin uptake inhibitors A = α2-adrenoceptor drugs

Vial # 2S 6S 7S 8S 9S 10S 12S 13S 14S 15S 16S 17S 18S 19S 21S 22S 23S 24S 26A 27A 28A 29A 30A 31A 32A 36S 37S 38S 41S 42A 43A 44A 45A 46S 48A 49A 51S 52S 53S 55A 56A 58A 59S 60S 61S 63A 64S 65S 68A 69S 70S 71S 72S 74A 75A

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

REVIEWS

Keeping an eye on vision: Molecular genetics and evolution of visual pigment proteins Benedict Darren1, Belinda Chang2,3 Department of Laboratory Medicine and Pathobiology, University of Toronto Department of Ecology and Evolutionary Biology, University of Toronto 3 Department of Cell and Systems Biology, University of Toronto 1 2

Abstract

Vision is intimately related to the ability and means with which an organism interacts with its environment. On a molecular level, the absorption of photons by visual pigment complexes in the retina, known as opsins, initiates a downstream cascade which culminates in the sensory impulse signaling the perception of light. Resolution of the rhodopsin crystal structure and subsequent biochemical assays have revealed key residues essential to the basic function of these light-sensitive proteins. In recent years, a body of research has surmounted to fully catalog the range of spectral sensitivities observed in opsins across a variety of vertebrate taxa. This review addresses the fundamental molecular mechanisms of spectral tuning which have been uncovered by sequence analyses and mutagenesis studies. Moreover, the upstream forces which drive these shifts in sensitivity are considered, particularly spectral heterogeneity across space, and developmental changes across time. These insights into the molecular genetics and evolution of visual pigments constitute an improved understanding of a protein’s capacity to develop novel and specialized functions.

Introduction

The visual transduction cascade begins when photons are absorbed by opsins, protein moieties bound to light-sensitive molecules called chromophores. Opsins, or visual pigments, are G-protein coupled receptors (GPCRs) that span the membrane of the outer segment of photoreceptors in the retina [1]. Typical of GPCRs, opsins are comprised of seven α-helical transmembrane domains, an extracellular domain, and an intracellular domain coupled to a G protein (Figure 1A). Each opsin covalently binds a chromophore, commonly the vitamin A aldehyde 11-cis-retinal, at a specific lysine residue through a protonated Schiff base linkage [1]. Opsins expressed in rod photoreceptors are referred to as rhodopsins while those in cones are simply called cone opsins [2]. In both rhodopsins and cone opsins, light induces isomerization of 11-cis-retinal to all-trans-retinal [Figure 1B], activating the G protein and triggering downstream effectors which hyperpolarize the cell, eventually culminating in the visual impulse [2]. Despite this shared activating mechanism, however, spectral sensitivities differ between opsins, and each type is photoactivated at a wavelength of maximal absorbance, called λmax [3]. Interestingly, λmax values can vary across the electromagnetic spectrum from 355 nm, near ultraviolet, to 570 nm, close to the end of the red region [3]. A focus in the field of vision research is how this wide range of maximal absorbance values is achieved in opsins. Given that all visual pigments contain 11-cis-retinal or a derivative, spectral sensitivity is determined largely by the interaction between the chromophore and amino acid residues in the chromophore

Figure 1: Tertiary structure of opsins and the retinal chromophore adapted from [1]. (A) Crystal structure of bovine rhodopsin showing seven helical transmembrane domains (H1-H7) characteristic of GPCRs, and an intracellular helix (H8). The amino terminus (N) is situated in the extracellular space while the carboxy terminus (C) is intracellular. Retinal (yellow) is bound to H7. (B) Light-induced isomerization of 11-cisretinal to all-trans-retinal which initiates the visual transduction cascade. Adapted from Terakita (2005) and Menon et al. (2001).

binding pocket [4]. In vertebrates, four groups of cone opsins have been classified based on their spectral properties: the SWS1 group with λmax between 355-440 nm (violet-ultraviolet sensitive),

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Keeping an eye on vision: Molecular geneticsand evolution of visual pigment proteins

the SWS2 group with λmax between 410490 nm (blue-violet sensitive), the RH2 group with λmax between 480-535 nm (green sensitive), and the LWS group with λmax between 490-570 nm (redgreen sensitive) [3]. Additionally, there is a single class of rhodopsin called RH1, maximally sensitive at around 500 nm. The RH2 cone opsin group derives its name from its close homology and spectral overlap with the RH1 group [4]. Note that there are spectrally distinct members within each opsin class. This review shall address the tuning mechanisms through which varying spectral sensitivities in visual pigments are attained in different animal species. Furthermore, the evolutionary history and processes behind these differences shall be considered, particularly as driven by environmental photic pressures. These two considerations are based on a working hypothesis that animals have evolved their visual systems to best correspond with attributes of their photic environment. As such, light intensity and spectral availability within the animal’s habitat act as selective pressures which drive adaptation, either by primary sequence changes in the opsin protein or by differential expression of the different classes of opsin genes.

Review Articles

Figure 2: Phylogenetic tree modeling the evolution of the five vertebrate opsin gene classes. Representative organisms from each major vertebrate group are included: human (mammals), green anole and gecko (reptiles), chicken (birds), clawed frog (amphibians), and zebrafish (fish). An early gene duplication diverged longer- (LWS) and shorter-wave cone opsins. The initial division was followed by two further gene duplications to produce three spectral classes of cones (SWS1, SWS2, RH2) within the shorter-wave family. The third gene duplication led to the evolution of rhodopsins (RH1) from the RH2 family. Gene duplications within the LWS class produced red (L)- and green (M)-sensitive isoforms, conferring trichromatic vision in primates, including humans. Adapted from Bowmaker (2008).

Evolutionary History of Opsin Genes

Gene duplication is the proposed evolutionary event through which visual pigment classes arose, producing an additional copy of a given opsin gene that is free from selective pressures [5]. As a result, mutations in the gene copy have no deleterious consequences to the individual, allowing for the emergence of a novel opsin gene that may potentially confer new function and spectral sensitivity [6]. The other gene copy retains its original function as constrained by purifying selection [6]. There has been evidence to show that retrotransposable elements, which transpose themselves via RNA intermediates, may have played a role in opsin gene duplications [7]. Thus, functional divergence of the five opsin classes is founded on a series of gene duplications, followed by nucleotide substitutions leading to neofunctionalization in the gene duplicate. The extent of gene duplications correspond to the number of opsin classes present in the organism, with the coexistence of more than one opsin class giving potential for colour vision (Figure 2). Microspectrophotometric studies of lampreys, considered to be extant representatives of ancient vertebrates, reveal that at least two opsin classes were present early in vertebrate evolution [5]. The results suggest that a gene duplication occurred around 500 MYA to produce a longer-wave opsin (homologous to LWS) and a shorter-wave opsin, prior to the divergence of major vertebrate classes [4]. This initial duplication was succeeded by four more

duplications in the shorter-wave class to give rise to the SWS1, SWS2, RH2, and RH1 opsin classes, all appearing by about 450 MYA [8]. Most extant reptiles, birds, and teleost fish possess rhodopsin and a representative of each of the four cone opsin classes, giving potential for tetrachromatic vision [5]. In contrast to these lower vertebrates, placental mammals have colour vision reduced to dichromacy, mediated by the LWS and SWS1 opsin classes. Humans and Old World primates are unique amongst mammals in possessing trichromatic colour vision [5]. This trichromacy is conferred by three classes of cone cells each containing the SWS1 opsin (violet-sensitive) or either of two variants of the LWS opsin, L (red-sensitive) and M (green-sensitive). Sequence comparisons between the L and M opsin genes in Old World species [9] suggest that they arose from the duplication of one ancestral opsin gene, demonstrating similar organization into head-to-tail tandem arrays [8] flanked upstream by a locus control region critical for the exclusive expression of either gene in a given photoreceptor [10, 11]. Furthermore, sequence analysis suggests that the 5’ region of the L gene represents the ancestral sequence and that duplication of the L opsin gene in humans and Old World primates was limited only to a short sequence upstream of the translation start site [7]. Notably, sequence alignments of the different opsin classes reveal conservation of functionally important amino acid sites. In particular, the counterion residue is necessary for the stabilization of the protonated Schiff base within the hydrophobic chromo-

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

Keeping an eye on vision: Molecular geneticsand evolution of visual pigment proteins

Figure 3: Opsin secondary structure highlighting residues critical for function and sites implicated in spectral tuning, adapted from [4]. Lysine 296 (orange) binds the chromophore through a protonated Schiff base linkage while glutamic acid 113 (blue) provides the counterion for the positive charge (inset) in the folded state. Key sites involved in spectral tuning are coloured according to opsin class: SWS1 in violet, LWS in red, and RH2 in green. Adapted from Bowmaker (2008).

phore-binding pocket of the opsin molecule [12] (Figure 3). The counterion, which is a negatively charged amino acid proximal to Schiff base linkage, has been suggested to be residues 113 and 181, with residue 181 being a conserved Glu across most opsin classes and residue 113 being variable [13]. However, vertebrate opsins generally have Glu at both 113 and 181 [13]. Mutations at E113 in bovine rhodopsin (RH1 opsin) sensitize it to UV light, indicating the absence of a protonated Schiff base at physiological pH [14]. The deprotonated opsin requires more energy to isomerize the retinal chromophore, resulting in the marked blue-shift in λmax. E181 mutants, on the other hand, have similar λmax to the wild type rhodopsin [12]. These findings suggest that E113, rather than E181, serves as the primary counterion in vertebrate opsins (Figure 3). However, spectroscopic studies have given evidence that E181 serves as an important secondary counterion during the photoactivation of rhodopsin, specifically during the formation of the first photointermediate, Meta I [15]. It has been shown that E181Q mutants have a destabilized Schiff base linkage to the chromophore at Meta I, effectively halting the photoactivation pathway.

Spectral Tuning by Sequence Substitution

Sequence substitutions prove to be an important mechanism in diversifying opsins within a single class. In SWS1 cone opsins, the shift from a maximal sensitivity in the ultraviolet to longer wavelengths in the violet region of the visible spectrum represents one of the key evolutionary phenomena that have punctuated the evolution of vertebrate visual systems [16]. This shift may be related to selective pressures on protection of the retina from the adverse properties of UV light and improvement of images projected onto the retina.

The mechanism of tuning between UV and violet light was first asserted by sequence comparisons between the violet-sensitive (VS) opsin of the cow, Bos taurus, and the UV-sensitive (UVS) opsin of the goldfish, Carassius auratus, which found that a Tyr residue at site 86 may be responsible for the red shift of the UVS opsin (17). Site-directed mutagenesis has shown that a Y86F substitution in the bovine VS opsin was sufficient to shift its λmax to the UV (at 363 nm) and conversely, a F86Y mutation in the goldfish UVS opsin shifted its λmax to the violet (at 413 nm) [17]. Interestingly, in the guinea pig, Cavia porcellus, a Val residue is present at site 86 instead of Tyr in its VS opsin [18]. Regardless, a V86Y substitution sufficiently red-shifts its λmax and a V86F substitution sufficiently blue-shifts its λmax, confirming the roles of Tyr and Phe in spectral tuning between the violet and the UV (Figure 3). The mechanism of tuning the VS pigment in primates would seem to be a caveat to the proposed roles Phe86 and Tyr86 in determining short wavelength sensitivities. Phe86, which typically confers UV sensitivity, was found in the VS pigment of the aye aye, Daubentonia madagascaiensis, a prosimian primate [19]. Additionally, spectral analysis reveals substitution of other residues (Cys, Ser, and Val) at site 86 also conferred violet sensitivity, thereby indicating that substitution at site 86 is not the primary mechanism through which primates tune their VS opsins. A P93T substitution, however, shifted the aye aye VS opsin λmax to the UV, suggesting that violet tuning in primates may be dependent on Pro93 rather than Tyr86 as in other mammals [19] (Figure 3). In the case of mammalian LWS opsins, spectral tuning was primarily investigated in primates. As aforementioned, gene duplication has given rise to two variants of the LWS opsin in Old World primates, the longer-wavelength (L) opsin with λmax at 563 nm and the middle-wavelength (M) opsin with λmax at 535 nm [8, 9]. The spectral shifts between the L and M opsins can be accounted for by differences at three sites, namely 164, 261 and 269 [20] (Figure 3). In the L opsin, polar residues Ser, Tyr and Thr respectively occupy these sites, while in the M opsin, non-polar residues Ala, Phe and Ala are present. Sequence substitutions may also be responsible for differences across opsin classes. A noteworthy change is that which diversified rhodopsin from the four classes of cone opsins. Cone opsins are marked by a faster pigment regeneration rate from the purified opsin moiety and 11-cis-retinal compared to rhodopsin [21]. Further mutagenesis experiments have found that sequence substitutions at Glu122 and Ile189 in rhodopsin are responsible for the difference in opsin regeneration rates (Figure 3). Experiments in chicken (Gallus gallus) rhodopsin show that the highly conserved Glu122, when substituted with Ile, the corresponding residue in cone opsins, conferred the mutant rhodopsin an accelerated regeneration rate. Moreover, when Pro189 in chicken RH2 cone opsin is mutated to its rhodopsin residue, Ile, the mutant RH2 opsin had a decelerated regeneration rate. For both substitutions, the reverse procedure produced the opposite effect. RH2 double mutants with mutations to conserved rhodopsin residues (Q122E and P189I) were also found to have a slowed decay rate of their photointermediates and a blue-shifted λmax, both of which distinguish rhodopsin from the RH2 opsin [22]. The chemical backings underlying the discussed shifts in the λmax of opsins (Table 1 and Figure 3) are still unclear and further work must be done in order to characterize the relationship between

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

Keeping an eye on vision: Molecular geneticsand evolution of visual pigment proteins

Table 1. Summary of amino acid sites and residues important in spectral tuning within opsin classes. Residue for Spectral Shift Red-shift Blue-shift

Species Bos Taurus, Carassius auratus, Cavia porcellus Daubentonia madagascaiensis Homo sapiens, Saimiri sciureus, Saguimis fuscicollis

Opsin Class

Key site(s)

SWS1

Tyr

Phe

17, 18

SWS1

RH2

Pro Ser Tyr Thr Gln Pro

--Ala Phe Ala Glu Ile

Gallus gallus

93 164 261 269 122 189

LWS

spectral shifts and the electric properties of specific residues. Though there has been success in identifying amino acid residues important for defining spectral sensitivity in opsins, there is still difficulty in predicting λmax shifts due to specific substitutions.

Spectral Shifts and Opsin Gene Expression

A complementary mechanism of tuning opsin photosensitivity has been found in addition to sequence mutations: differential opsin gene expression [23]. In particular, real-time quantitative PCR (qPCR) analysis of the RNA complements of cone opsin genes in several cichlid species have revealed that these changes are responsible for the larger spectral shifts between visual pigments [24]. Sequence comparisons between cone opsins of cichlid species from Lake Malawi with varying visual pigment sensitivities show that each corresponding opsin gene had nearly identical sequences [24]. Thus, changes in opsin sequence are not the primary means through which Malawi cichlids vary their spectral sensitivities. Rather, each of the species studied expresses a different subset of the seven available opsin genes in cichlids, with two spectral variants of the SWS2 gene and three of the RH2 gene [25]. For instance, microspectrophotometry shows that the Metriaclima zebra visual system is UV-sensitive, with λmax of the predominant visual pigments at 368 nm (SWS1), 488 nm (RH2B), and 533 nm (RH2Aα) [25]. Dimidiochromis compressiceps has a red-shifted visual system, with wavelength sensitivities at 447 nm (SWS2A), 536 nm (RH2Aα), and 569 nm (LWS). qPCR results confirm that M. zebra obtains its blue-shifted vision by downregulating its expression of the LWS and SWS2A genes and make use of the SWS1 and RH2 genes instead [23]. In contrast, D. compressiceps has a reduced expression of the SWS1 and SWS2A genes and a greater expression of the LWS and SWS2A genes [23]. As such, a change in the opsin gene expression in a given photoreceptor results in a visual pigment with a different spectral sensitivity. However, the upstream forces which direct this differential opsin gene expression have yet to be determined. Some evidence has accumulated to contend that changes in gene expression may follow a developmental pattern. A switch in opsin gene expression was initially reported in the eel [26], Anguilla anguilla, which undergoes metamorphosis from the freshwater “yellow eel” to the sexually mature oceanic “silver eel” [27]. Remarkably, in the eel retina, this physiological and ecological metamorphosis is paralleled by a molecular transition. Typical in freshwater species, the immature yellow eel expresses porphyropsin, a rhodopsin variant which binds 3-dehydroretinal instead of 11-cis-retinal as the chromophore, which red-shifts the λmax of the pigment [28]. During development, the visual pigment expressed in the rod cells change from the freshwater porphyropsin, with λmax at 525 nm, to

Reference(s)

20 22

a typically marine rhodopsin with λmax at 480 nm, incorporating not only a change in the chromophore, but also in rod opsin gene expression in favour of a shorter wavelength-sensitive opsin [29]. Recent research has also shown a similar blue-shift in the greensensitive RH2 cone opsin of A. anguilla, with λmax shifting from 550 nm to 525 nm [28]. These spectral shifts indicate a morphogenetic pattern to rod opsin gene expression in the eel associated with the photic environments at each developmental stage. Furthermore, qPCR analysis suggests that the visual system of the Nile tilapia, Oreochromis niloticus, changes dynamically throughout its morphogenesis [30]. The gene set progresses from short-wavelength SWS1 and RH2B genes in the larvae to the longerwavelength SWS2B in the juvenile set, and finally to the even more red-shifted SWS2A and LWS genes in the adult set. Thus, the tilapia has a temporally-dependent developmental program which shifts its visual-system toward longer-wavelengths as the larva matures. These results suggest that sensitivity at shorter wavelengths may be important for foraging in the early life stages of the tilapia, while longer wavelength sensitivity can confer advantage in turbid riverine environments, given the typically murky African river [30]. Again, it is still unclear how this differential opsin gene expression is actually regulated, whether under the control of transcriptional factors, or by mutations in promoter regions upstream of opsin genes. There is evidence of that hormonal modulation may play a role in changing opsin gene expression. A comparative qPCR analysis between thyroid hormone (TH)induced and natural development in rainbow trout, Oncorhynchus mykiss, show that the SWS1 gene is strongly downregulated under TH treatment [31]. Importantly, this TH-induced decrease in SWS1 expression precedes the documented loss of UV-sensitive cones in the retina of trout juveniles under similar TH induction [32], suggesting that TH can modulate opsin gene expression and ultimately shape the visual system of the rainbow trout. In addition to the identification of more hormone candidates, the protein effectors which receive the hormonal signal and relay it to alter expression of opsin genes must be defined in order to fully characterize the molecular interactions which mediate this change.

Environmental Drive to Spectral Shifts

The question still remains with regard to the effect of varying photic environments on the visual system of a given organism. It is argued that the array of visual pigments expressed and the λmax of those pigments are optimally attuned to the available spectral range of light in the organism’s habitat. A notable study of the cottoid fish in Lake Baikal demonstrates a clear correlation between water depth and the subset of opsin classes expressed [33]. Light is progressively attenuated as it passes through a body of water, especially at longer

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

Keeping an eye on vision: Molecular geneticsand evolution of visual pigment proteins

and shorter wavelengths, such that down-welling light is reduced to a narrow spectrum between 470-480 nm in the blue-green [34]. Correspondingly, only the surface-dwelling cottoid species were found to have three spectrally distinct classes of cone opsins, presumably two isoforms of the RH2 opsin and one LWS (with λmax around 450, 525, and 546 nm, respectively). In contrast, species found in deeper water possessed only two middle-wave cone opsin classes (with λmax at 450 and 521 nm), thus reflecting loss of long wavelength-sensitivity [33]. This reduction of opsin complement to the middle wavelength-sensitive class parallels the attenuation of light in water toward the middle region of the spectrum. Moreover, sequencing and microspectrophotometric studies on the opsin genes of the Comoran coelacanth, Latimeria chalumnae, reveal that they too have dispensed of their SWS1, SWS2, and LWS genes while retaining their RH1 and RH2 pigments [35]. Since the coelacanth lives at depth of about 200 m, light is attenuated such that colours spread about 480 nm is available [34]. When compared to the λmax of most RH1 and RH2 opsins (around 500 nm), both the coelacanth’s RH1 and RH2 opsins have blueshifted λmax values (at 478 and 485 nm respectively). These shifts were achieved through independent sets of double mutations in each pigment: E122Q/A292S for RH1 and E122Q/M207L for RH2 [35]. It appears then that the blue-shifts in the opsins of the coelacanth served as an adaptive mechanism to tune its vision to the available light spectrum within its unique habitat. Note, however, that this spectral tuning comes with the loss of colour vision in the coelacanth as it has only one class of cone opsin. The observations in the cottoid fish of Lake Baikal and the Comoran coelacanth seem to argue in favour of the evolutionary hypothesis that differing photic environments act as a drive for visual adaptation. There are, however, a few counterpoints to be made against this prediction. In Antarctic waters, the attenuation of light at short and long wavelengths is further amplified by the snow and sea ice cover, effectively limiting light transmission to around 500 nm with lower intensity [37]. Yet, in several notothenioid fish species living below the sea ice of the Antarctic, UV-sensitive SWS1 and blue-sensitive SWS2 cone opsins were detected in addition to the RH2 cone opsin and RH1 rod opsin [36]. Given a UV-deprived photic environment, the motivation behind the retention of the UV-sensitive visual pigment is unclear for the notothenioid fish. More recently, the visual pigments of the elephant shark, Callorhincus milii, have been identified, taking advantage of the shark’s sequenced genome [38]. Genes encoding the RH1 rod opsin, RH2, and two variants of the LWS cone opsins were detected. Particularly, the LWS isoforms are analogous to those found in Old World primates, with one more sensitive in the green (λmax at 499 nm) and the other in the red (λmax at 548 nm). Thus, with the inclusion of the RH2 opsin, the elephant shark possesses three spectrally distinct classes of cone opsins, giving potential for trichromatic vision. It has therefore retained a function lost in the Comoran coelacanth, which resides at a similar depth [38]. Hence, in both the notothenioid fish and the elephant shark, their visual system, based on their opsin complements, does not seem to correspond to their photic environments. It appears then that for deep-water vision, optimal spectral tuning necessitates retention of only the middle-wave opsins due to the attenuation of light, in a tradeoff with colour vision which requires more than one spectral class of cone opsin.

Despite these exceptions, there is still a pervasive trend associating opsin expression with the spectral environment of an organism. Additional research is required to solidify this relationship, especially in non-aquatic systems. Candidate terrestrial organisms include those with fossorial habits, for instance snakes (suborder Serpentes), since an environment devoid of light favors sensory modalities in place of or in addition to vision such as chemoreception and infrared sensing, as observed in extant snake species [39].

Conclusion and Future Directions

The combination of molecular techniques and accessibility to a wide range of animal species has allowed the identification and isolation of opsin genes, producing interesting observations regarding the evolution of visual pigments. Mutations at several key sites in the opsin protein moiety fine tune the maximal absorbance of the given opsin, while differential opsin gene expression generate large-scale changes in the visual system of the organism. At this point, there is still difficulty in associating λmax shifts with the biochemical properties of specific substitutions. A phylogenetic approach may provide some resolution to this problem: a comparative analysis of opsin sequences can be used to construct a phylogenetic tree and identify non-conservative amino acid substitutions in independently evolved opsins which give rise to similar spectral shifts [40]. This approach can be further facilitated by increasing the resolution of the crystal structure of rhodopsin and its photointermediates [41], allowing the identification of amino acids nearby the chromophore which are likely to exert some effect on the spectral sensitivity of the molecule. Additionally, there are still disparities with regard to how a particular photic environment can drive these visual modifications in an organism. Though function and environment have yet to be reconciled, our current understanding of opsin genotype and phenotype has surely brought such a prospect within reach.

References

1. Terakita A. The opsins. Genome Biol. 2005; 6(3): 213. 2. Palczewski K. Chemistry and biology of vision. J Biol Chem. 2011; 287(3): 1612-9. 3. Yokoyama S. Molecular evolution of vertebrate visual pigments. Prog Retin Eye Res. 2000; 19(4): 385-419. 4. Bowmaker JK. Evolution of vertebrate visual pigments. Vision Res. 2008; 48(20): 2022-41. 5. Bowmaker JK. Evolution of colour vision in vertebrates. Eye; 12 (Pt 3b): 541-7. 6. Hunt DM, Dulai KS, Cowing JA, Julliot C, Mollon JD, Bowmaker JK, Li WH, Hewett-Emmett D. Molecular Evolution of trichromacy in primates. Vision Res. 1998; 38(21): 3299-3306. 7. Dulai KS, Dornum MV, Mollon JD, Hunt DM. The evolution of trichromatic colour vision by opsin gene duplication in New World and Old World primates. Genome Res. 1999; 9(7): 629-38. 8. Ibbotson RE, Hunt DM, Bowmaker JK, Mollon JD. Sequence divergence and copy number of the middle- and long-wave photopigment genes in Old World monkeys. Proc Biol Sci. 1992; 247(1319): 145-54. 9. Nathans J, Thomas D, Hogness DS. Molecular genetics of human color vision: The genes encoding blue, green, and red pigments. Science 1986; 232(4747): 193-203. 10. Wang Y, Macke JP, Merbs SL, Zack DJ, Klaunberg B, Bennet J, Gearhart J, Nathans J. A locus control region adjacent to the human red and green visual pigment genes. Neuron 1992; 9(3): 429-440. 11. Nathans J, Davenport CM, Maumenee IH, Lewis RA, Heijtmancik JF, Litt M, Lovrien E, Weleber R, Bachynski B, Zwas F et al. Molecular genetics of human blue cone monochromacy. Science 1989; 245: 831-8. 12. Menon ST, Han M, Sakmar TP. Rhodopsin: Structural basis of molecular physiology. Physiol Rev. 2001; 81: 1659-88. 13. Shichida Y, Matsuyama T. Evolution of opsins and phototransduction. Phil Trans R Soc B. 2009; 364(1531): 2881-95. 14. Sakmar TP, Franke RR, Khorana HG. Glutamic acid-133 serves as the retinylidene Schiff base counterion in bovine rhodopsin. Proc Natl Acad Sci USA. 1989; 86(21): 8309-13. 15. Yan ECY, Kazmi MA, Ganim Z, Hou JM, Pan D, Chang BSW, Sakmar TP, Mathies RA.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Keeping an eye on vision: Molecular geneticsand evolution of visual pigment proteins

31. Veldhoen K, Allison WT, Veldhoen N, Anholt BR, Helbing CC, Hawryshyn CW. Spatiotemporal characterization of retinal opsin gene expression during thyroid hormone-induced and natural development of rainbow trout. Vis Neurosci. 2006; 23(2): 169-79. 32. Allison WT, Dann DG, Helvik JV, Bradley C, Moyer HD, Hawryshyn CW. Ontogeny of ultraviolet-sensitive cones in the retina of rainbow trout (Oncorhynchus mykiss). J Comp Neurol. 2003; 461(3): 294-306. 33. Bowmaker JK, Govardovskii VI, Shukolyukov SA, Zueva LV, Hunt DM, Sideleva VG, Smirnova OG. Visual pigments and the photic environment: The cottoid fish of Lake Baikal. Vision Res. 1994; 34(5): 591-605. 34. Jerlov NG. Marine Optics. Amsterdam: Elsevier Scientific; 1976. 35. Yokoyama S, Zhang H, Radlwimmer B, Blow NS. Adaptive evolution of the Comoran coelacanth (Latimeria chalumnae). Proc Natl Acad Sci USA. 1999; 96(11): 6279-84. 36. Pointer MA, Cheng CH, Bowmaker JK, Parry JW, Soto N, Jeffery G, Cowing JA, Hunt DM. Adaptations to an extreme environment: retinal organisation and spectral properties of photoreceptors in Antarctic notothenioid fish. J Exp Biol. 2005; 208(Pt 12): 2363-76. 37. Perovich DK, Longacre J, Barber DG, Maffione RA, Cota GF, Mobley CD, Gow AJ, Onstott RG, Grenfell TC, Pegau WS et al. Field observations of the electromagnetic properties of first-year sea ice. IEEE T Geosci Remote. 1998; 36(5): 1705-15. 38. Davies WL, Carvalho LS, Tay BH, Brenner S, Hunt DM, Venkatesh B. Into the blue: gene duplication and loss underlie color vision adaptations in a deep-sea chimaera, the elephant shark Callorhincus milii. Genome Res. 2009; 19(3): 415-26. 39. Sillman AJ, Govardovskii VI, Rohlich P, Southard JA, Loew ER. The photoreceptors and visual pigments of the garter snake (Thamnophis sirtalis): A microspectrophotometric, scanning electron microspcope and immunological study. J Comp Physiol. 1997; 181(2): 89-101. 40. Chang BSW, Crandall KA, Carulli JP, Hartl DL. Opsin phylogeny and evolution: a model for blue shifts in wavelength regulation. Mol Phylogenet Evol. 1995; 4(1): 31-43. 41. Palczewski K, Kumasaka T, Hori, T, Behnke CA, Motoshima H, Fox BA, Le Trong I, Teller DC, Okada T, Stenkamp R, Yamamoto M, Miyano M. Crystal structure of rhodopsin: A G proteincoupled receptor. Science 2000; 289(5480): 739-45.

LMP

Counterion switch in the photoactivation of the G protein-coupled receptor rhodopsin. Proc Natl Acad Sci USA. 2003; 100(16): 9262-7. 16. Hunt DM, Carvalho LS, Cowing JA, Parry JW, Wilkie SE, Davies WL, Bowmaker JK. Spectral tuning of shortwave-sensitive visual pigments in vertebrates. Photochem Photobiol. 2007; 83(2): 303-10. 17. Cowing JA, Poopalasundaram S, Wilkie SE, Robinson PR, Bowmaker JK, Hunt DM. The molecular mechanism for the spectral shifts between vertebrate ultraviolet- and violetsensitive cone visual pigments. Biochem J. 2002; 367(Pt 1): 129-35. 18. Parry JW, Poopalasundaram S, Bowmaker JK, Hunt DM. A novel amino acid substitution is responsible for spectral tuning in a rodent violet-sensitive visual pigment. Biochemistry 2004; 43(25): 8014-20. 19. Carvalho LS, Davies WL, Robinson PR, Hunt DM. Spectral tuning and evolution of primate short-wavelength-sensitive visual pigments. Proc Biol Sci. 2012; 279(1727): 387-93. 20. Neitz M, Neitz J, Jacobs GH. Spectral tuning of pigments underlying red-green color vision. Science 1991; 252(5008): 971-4. 21. Imai H, Kuwayama S, Onishi A, Morizumi T, Chisaka O, Shichida Y. Molecular properties of rod and cone visual pigments from purified chicken cone pigments to mouse rhodopsin in situ. Photochem Photobiol Sci. 2005; 4(9): 667-74. 22. Kuwayama S, Imai H, Hirano T, Terakita A, Shichida A. Conserved proline residue at position 189 in cone visual pigments as a determinant of molecular properties different from rhodopsin. Biochemistry 2002; 41(51): 15245-52. 23. Carleton KL. Cichlid fish visual systems: Mechanisms of spectral tuning. Integr Zool. 2009; 4(1): 75-86. 24. Carleton KL, Kocher TD. Cone opsin genes of African cichlid fishes: Tuning spectral sensitivity by differential gene expression. Mol Biol Evol. 2001; 18(8): 1540-50. 25. Parry JW, Carleton KL, Spady T, Carboo A, Hunt DM, Bowmaker JK. MMix and match color vision: tuning spectral sensitivity by differential opsin gene expression in Lake Malawi cichlids. Curr Biol. 2005; 15(19): 1734-9. 26. Carlisle DB, Denton EJ. On the metamorphosis of the visual pigments of Anguilla anguilla. J Mar Biol Assoc UK. 1959; 38(1): 97-102. 27. Berry L, Brookes D, Walker B. The problem of the migration of the European Eel (Anguilla anguilla). Science Progress 1972; 60: 465-85. 28. Bowmaker JK, Semo M, Hunt DM, Jeffery G. Eel visual pigments revisited: The fate of retinal cones during metamorphosis. Vis Neurosci. 2008; 25(3): 249-55. 29. Hope AJ, Partridge JC, Hayes PK. Switch in rod opsin gene expression in the European eel, Anguilla anguilla. Proc Biol Sci. 1998; 265(1399): 869-74. 30. Carleton KL, Spady T, Streelman JT, Kidd MR, McFarland WN, Loew ER. Visual sensitivities tuned by heterochronic shifts in opsin gene expression. BMC Biol. 2008; 6: 22.

Review Articles

INVESTIGATING DISEASE. IMPACTING HEALTHCARE.

Boundless Opportunity Thinking about pursuing a career in research? At the centre of innovative research and clinical practice, LMP offers customized training with unparalleled access to expert faculty, diverse training sites and a stunning array of research opportunities. Join LMP today and become a catalyst for better health care tomorrow.

For more information visit

www.lmp.utoronto.ca

www.lmp.utoronto.ca Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

REVIEWS

The “weakness” in single gene disorder research: The role of signaling pathways and vasoregulation in the dystrophinglycoprotein complex (DGC) in Duchenne Muscular Dystrophy Nicole Fogel Department of Cell and Molecular Biology and Neuroscience, University of Toronto, Toronto, Canada

Abstract

Duchenne muscular dystrophy (DMD) is the severe form of a group of muscle degenerative diseases caused by mutations in the DMD gene on the X chromosome. The DMD gene is approximately 2.4 Mb in size and encodes the muscle structural support protein dystrophin. Dystrophin is a component of the dystrophin-glycoprotein complex (DGC), which assembles at the plasma membrane of muscle and non-muscle tissues. Although the role of dystrophin and the DGC has been the focus of therapeutic studies, there is currently very little knowledge about the cell signaling events that occur within the complex or about its role in vasoconstriction. Characteristics of DMD, such as muscle weakness and impaired cognition, are thought to arise from deletions and duplications in the DMD gene that result mainly in the loss of dystrophin. Loss of dystrophin and upsetting of the DGC induces a change in the downstream cascade of events: improper placement of SLO-1 channels, decrease in nitric oxide (NO), increase in phosphatidylinositol 3-kinase (PI3K/Akt) and nuclear factor-kappa B (NF-κB) signaling, increase in creatine kinase (CK) levels, and abnormal vasoconstriction. These events bring about symptoms of the disorder. This review will compare important studies that have contributed to the current understanding of the molecular basis of signaling mechanisms that occur within the DGC. Specifically, it will address the changes that occur in signaling pathways from the loss of dystrophin that lead to decreased regulation of vasoconstriction and muscle stability, thereby promoting disease. By providing a more integrated approach to analyzing cell signaling pathways, this review hopes to encourage the continuation of these types of studies and promote the understanding of single gene disorders. In this way, novel pharmacological and genetic counseling therapies can be established, and health policies can be revised for incapacitating diseases like DMD.

Introduction

Duchenne muscular dystrophy (DMD) is a severe, x-linked recessive disease characterized by muscle degeneration and impaired cognition [1, 2]. It is estimated that 1 in 3500 males worldwide are affected by this disorder every year, of which 25% die from heart failure [1, 9]. The disease is mainly caused by mutations in the DMD gene on the X chromosome - 70% of cases are due to deletions and duplications of this gene, while 30% are due to point mutations [1, 2]. The DMD gene normally encodes a skeletal muscle structural support protein called dystrophin [2], which acts, with its associated proteins, as a component of the dystrophin-glycoprotein complex (DGC) at the plasma membrane of muscle and non-muscle tissues. The DGC includes many proteins; a constant component of it is dystrophin, whose loss largely accounts for DMD symptoms though it only makes up a small portion of muscle tissue [1, 2].

Dystrophin and its compensating homologue, utrophin, are both critical and largely responsible for many of the DGC’s functions and cohesion. This complex is involved in regulating interactions between the cytoskeleton, the cell membrane, and the extracellular matrix (ECM). It also contains signaling pathways responsible for regulating the organization of ion channels and postsynaptic receptors when motor neurons contact muscle fibers. Dystrophin plays a key role in preventing sarcolemmal ruptures by binding to the plasma membrane of the muscle fiber and supporting the stresses on the muscle during contraction [3]. Without this essential protein, muscles will be unable to support and adapt to mechanical stresses, leading to the stiffness and limited muscle control that characterizes DMD disorders. Specifically, loss of dystrophin results in disruptions in the connections between the actin cytoskeleton and the ECM, causing the sarcolemma to become less stable with each contraction [2, 3]. Microfissures or

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

The “weakness” in single gene disorder research: the role of signaling pathways and vasoregulation in the dystrophin-glycoprotein complex (DGC) in Duchenne muscular dystrophy

small cracks can appear on the destabilized plasma membrane of skeletal muscle fibers, causing calcium influx into the cells [3]. As a result, calcium-dependent proteases are aberrantly activated, leading to cell death and increased protein degradation. Muscle fibers then degenerate, leading to muscle fatigue or weakness. Loss of dystrophin also results in increased muscle excitability or skeletal muscle hypertrophy, and a corresponding reduction in adipose stores. In a normal process of muscle contraction, calcium release from the muscle sarcoplasmic reticulum causes myosin to bind to actin, inducing a conformational change that increases the overlap between two muscle filaments, thus creating contraction [3]. Improper calcium regulation due to microfissures in the sarcolemma, stated above, additionally results in unregulated muscle contractions, leading to cell death due to muscle fiber degeneration. In this way, DMD disorders are associated with limited control and greater fatigue of muscles with lack of resting time. This review will compare important studies that have contributed to the current knowledge of the molecular basis of signaling mechanisms that occur within the DGC (Figure 1) that lead to decreased regulation of vasoconstriction and muscle stability from dystrophin loss.

Review Articles

that the C. elegans mutants showed incorrect localization, observed by a 75-90% reduction in SLO-1 expression compared to wildtype (WT) expression. Results also indicated an increase in muscle excitability, measured by action potentials of SLO-1 channels. AChR and chemical synaptic transmission expression was unaffected. Therefore, it can be concluded that the DGC and dystrophin are involved in controlling muscle excitability through SLO-1 channels and are less involved in AChR alteration and calcium influx. Loss of dystrophin affects the localization of SLO-1 ion channels which leads to improper muscle control [2, 5]. However, it is important to note that while the genetic abnormalities of mouse mdx and C. elegans models closely resemble those of humans, they present with milder symptoms in studies than those associated with DMD in humans due to the decreased size and force of their muscles. For this reason, DMD mutations in some components of the DGC are still unlinked to human DMD [5]. Sancar et al. were one of the few authors to study loss of dystrophin and muscle excitability through signaling pathways other than the PI3K/Akt and NF-κB signaling pathways. Understanding the interaction of dystrophin with SLO-1 provides insight into how multiple signaling pathways in the DGC are integrated, particularly how SLO-1 interacts with the more commonly studied signaling pathways.

Signaling pathways in response to dystrophin loss

Figure 1: The dystrophin-glycoprotein complex (DGC). The proteins dystrophin and utrophin are components of the dystrophin-glycoprotein complex (DGC), which assembles at the plasma membrane of muscle and nonmuscle tissues. Loss of dystrophin is associated with Duchenne muscular dystrophy (DMD). The interactions between the signaling pathways and proteins of the DGC can be visualized. [2]

The DGC controls muscle excitability via SLO-1 channels

Although the DGC is known to be involved in muscle cell support and maintenance of acetylcholine receptors (AChRs) [2], Sancar et al. [5] examined the mechanisms of its cell signaling events that are not well known. Using DGC-lacking mutants of C. elegans DGC homologues, the authors tested the expression of the DGC’s chemical synaptic transmissions and ion channel localization, specifically SLO-1, a postsynaptic calcium-gated potassium channel. Muscle excitability was tested in relation to the expression of these chemical transmissions and channel localization. Results indicated

NO signals in the heart In addition to SLO-1 localization, loss of dystrophin also affects major signaling pathways including NO, NF-κB, and PI3K/ Akt [3, 6, 7, 8]. Under normal dystrophin levels, neuronal nitric oxide synthase (nNOS) is recruited to the sarcolemma in skeletal muscle and allows for the production of nitric oxide (NO), which plays a key role in inducing vasodilation [6]. The balance between vasodilation and vasoconstriction is essential for proper muscle function. Without vasodilation control, increased vasoconstriction can result, leading to muscle fatigue. This concept will be discussed in more detail under the heading “Dystrophin’s role in regulating vasoconstriction via CK levels in VSMC”. Individuals with DMD experience abnormal vascular regulation [10]. Often overlooked by studies involving skeletal muscle function, cardiac muscle failure is also prevalent in individuals with DMD [8]. It is important to note that by around twenty years of age, individuals with DMD have an increased risk of dying from cardiac failure. It has been shown that approximately 95% of these individuals develop dilated cardiomyopathy and 25% will die from heart failure [9]. In order to better understand the role of nNOS in the heart, Ramachandran et al. [6] compared the cardiomyocytes and hearts of mdx and mdx:utrophin (-/-) (mdx:utr) double knockout (DKO) mice to WT mice using immunohistochemistry and PCR techniques. Mdx (X-chromosome-linked muscular dystrophy) mice lack dystrophin and serve as an animal model or homolog that partially resembles human DMD [3]; utrophin is a homologue of dystrophin. According to their findings, a lack of dystrophin disrupted nNOS recruitment to skeletal muscle sarcolemma, thereby reducing NO production to less than 20% in the mdx mice cardiomyocytes and to only 10% in mdx:utr mice [6]. A disruption in the nNOS signal pathway may be a contributing factor to the muscle weakness that characterizes DMD as it induces abnormal cardiac function and vasoregulation. Additionally, it was reported that nNOS function was de-

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

The “weakness” in single gene disorder research: the role of signaling pathways and vasoregulation in the dystrophin-glycoprotein complex (DGC) in Duchenne muscular dystrophy

creased in mdx and double knockout mice, but expression levels of nNOS decreased solely in mdx:utr mice [6]. This indicates that utrophin, which is up-regulated in the absence of dystrophin in sarcolemma, is up-regulated in cardiomyocytes and is responsible for maintaining nNOS levels, but is not responsible for its function. Other steps in the signaling pathway were studied such as soluble guanylyl cyclase activity and L-arginine transporter activity. Results indicated that levels of soluble guanylyl cyclase activity in mutants paralleled control levels, whereas levels of L-arginine transporter activity increased compared to control [6]. Therefore, both dystrophin and utrophin are important contributors to DMD pathology, though dystrophin has a greater role in the regulation of events. Also, a recent study used a promising gene therapy, exon skipping, to increase dystrophin and up-regulate nNOS signaling pathways [11]. However, the authors failed to determine exactly how much of a role dystrophin plays in the regulation of nNOS signaling events with utrophin compensation. NF-kB signals in the heart To further study cardiac muscle effects in DMD, Delfin et al. [8] synthesized a peptide sequence called Nemo Binding Domain (NBD) to target and inhibit NF-κB signaling. Diluted NBD was injected into mice three times weekly until eight weeks of age. NF-κB signaling is thought to increase in response to dystrophin loss, and is responsible for regulating immune responses and anti-apoptotic function [7]. Inhibition of NF-κB results in reduced inflammation, enhanced muscle fiber regeneration, and improved cardiac contractile function in the diaphragm of mdx mice [8]. Utilizing a DKO for both utrophin and dystrophin with the NBD treatment compared to vehicle-treated, or control, DKO mice, the authors measured force development in cardiac muscles. It was discovered by using techniques such as Electrophoretic Mobility Shift Assay (EMSA), Western blotting, and stimulation techniques that the cardiac muscle force doubled, the frequency-dependent behavior of muscles improved, and the force to β-adrenergic response was greatly restored in the DKO mice with the NBD treatment. The NF-κB signaling pathway, with loss of dystrophin, may play a role in down-regulating Ca2+ channels, which is the basis behind NBD and its functions. The authors maintain that NBD may improve contractile function by improving excitations-contraction coupling and handling of calcium [8]. This study provides hope that by blocking certain signaling pathways, muscle function can be restored and excitation can be regulated, aiding in the creation of novel therapeutic techniques. However, this study does not compare the effects of NBD on the NF-κB signaling pathway to other signaling pathways involved in DMD. PI3K/Akt and NF-κB in skeletal muscle The PI3K/Akt and NF-kB signaling pathways are activated by mechanical stimulation [7]. PI3K/Akt is thought to play a role in cell viability and maintenance of skeletal muscle mass, while NF-κB is involved in anti-apoptotic and immune function [7, 8]. Dogra et al. [7] showed that both signaling pathways are actually up-regulated in dystrophin-deficient mice. Mdx mice were compared to WT mice in pre- and necrotic states in muscle fibers using immunoprecipitation (IP), Western blot, and EMSA techniques. Mechanical stretch was induced or passively applied (ex vivo) for 15 minutes and the tension generated by stretching was monitored using a force trans-

ducer. Results indicated that in the diaphragm muscle, activation of Akt increased from 4 fold in WT to 8 fold in mdx mice. There was also an increase in PI3K activation from 3 fold in WT to 6 fold in mdx mice. Specifically with Akt activation, there was an increase in phosphorylation of its downstream targets [7]. With the introduction of a PI3K inhibitor, Akt activation and the induced activation of NF-κB were blocked in both normal and mdx mice. While the loss of Akt is expected due to its position downstream of PI3K, the induced activation of NF-κB is surprising. Passive mechanical stretch reduced the interaction of NF-κB and histone deacetylase (HDAC) [7]. There is a symbiotic relationship among the different components of the signaling pathways. Interestingly, up-regulation of both signaling pathways is induced by mechanical stretch, not just loss of dystrophin. This suggests that the mechanisms of DMD might be more complicated than previously thought. It is speculated that up-regulation of both signaling pathways might be a survival and coping mechanism, or what Kim et al. [3] call a “compensatory response” to loss of dystrophin or mechanical stress [7]. However, this study demonstrates that there is still much to learn about these signaling events, their purpose, and wide ranging functions. This study effectively integrates the changes that occur in two main signaling pathways in response to loss of dystrophin. Kim et al. [3] further utilized Akt as a potential therapy. Studies were completed in mdx double transgenic (DTG) mice in comparison to single transgenic (STG) mice such that STG mice could not activate Akt. Introducing Akt treatment in DTG mice after the onset of disease (> 6 weeks) led to increased formation of adhesion complexes involving integrin and utrophin to the sarcolemma. Levels in albumin, which regulates osmotic pressure in blood, and in nNOS increased compared to STG mice (Figure 2a). Overall Akt levels were increased by 3 fold and phosphorylated Akt (P-Akt) levels were increased by greater than 20-fold in DTG muscles compared to STG muscles [3]. The authors successfully managed to incorporate utrophin and other signaling pathways, like nNOS levels, into their analyses; however, future studies should include effects on the heart muscle to provide therapies for the high incidence of heart failure associated with DMD. Muscle function is generally measured by force and resistance to fatigue, demonstrated through grip tests and in vitro contraction measurements of the muscle [3]. In mdx DTG mice, muscle function levels improved after Akt treatment, however they still did not reach WT levels. Muscle fatigue decreased dramatically in both mdx STG and DTG mice after repeated electrophysiological stimulation, from 100% fatigue in 5 minutes to 40% fatigue after about 2.5 minutes of stimulation (Figure 2b). Muscles of WT mice that were injured using cardiotoxin experienced faster regeneration with Akt treatment. This research demonstrates the potential of understanding the powerful signaling pathways associated with the DGC in order to create novel therapeutic approaches to counteract the symptoms of DMD.

Dystrophin’s role in regulating vasoconstriction via CK levels in VSMC

Ito et al. [10] examined the lesser-known functions of dystrophin in vascular smooth muscle cell (VSMC) vasoconstriction. The authors used atypical mdx mice types, comparing transgenic mdx mice that expressed dystrophin only in smooth muscle to non-transgenic mdx mice which were inbred to express a disease similar to DMD in skeletal and smooth muscle, although not as

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

The “weakness” in single gene disorder research: the role of signaling pathways and vasoregulation in the dystrophin-glycoprotein complex (DGC) in Duchenne muscular dystrophy

severe. They measured vasoconstriction using norepinephrine (NE) injections and measured CK levels. In humans with DMD, blood vessels are constricted and the enzyme CK converts ATP to ADP very rapidly in the brain and in skeletal muscle. Both of these factors contribute to muscle fatigue or weakness and cognitive impairment, which are characteristic of the disease. With continuous fatigue, tension in the muscles decreases, and proper control of contraction, relaxation, and restoration are compromised as CK cannot effectively regulate ATP to ADP conversions. In mouse models, results of Ito et al. revealed that when dystrophin was expressed, the transgenic mdx mice had reduced CK levels and restored or normalized vasoconstriction in the hindlimbs, in comparison to the non-transgenic mdx mice [10]. It can be concluded from this study that dystrophin plays a necessary role in regulating vasoconstriction and CK levels, specifically in VSMC [10]. The authors did not mention if increased vascular resistance also resulted from increased vasoconstriction in mdx mice.

Additional roles of DGC and mental retardation

Components of the DGC interact via similar signaling pathways. For example, the glucocorticoid-induced protein myocilin, which is normally associated with glaucoma was found to interact

Review Articles

Figure 2: Akt signaling effects on muscle stability and function. a) Studies were completed in mdx double transgenic (DTG) mice in comparison to single transgenic (STG) mice such that STG mice could not activate Akt. Introducing Akt treatment in DTG mice after the onset of disease (> 6 weeks) led to increased formation of adhesion complexes involving integrin and utrophin to the sarcolemma. Akt and phosphorylated Akt levels increased in DTG muscles compared to STG muscles. b) The PI3K/Akt and NF-kB pro-survival signaling pathways are activated by mechanical stimulation. Kim et al. [3] found that in mdx DTG mice, muscle function levels improved after Akt treatment, however they still did not reach WT levels. Muscle fatigue decreased dramatically in both mdx STG and DTG mice after repeated electrophysiological stimulation, from 100% fatigue in 5 minutes to 40% fatigue after about 2.5 minutes of stimulation. Using Akt signaling could prove to be a beneficial therapy for DMD.

with syntrophin, a component of the DGC, in tissues other than the eye [12]. This interaction increased nNOS levels, phosphorylated Akt signaling, reduced muscle atrophy markers by up to 88%, and even increased muscle fibers in transgenic mice by 36% compared to controls. Understanding interactions of the DGC, specifically the role of dystrophin in cognitive impairments, could help to establish therapies not only for muscle weakness, but also for cognitive characteristics of DMD.

Conclusion

Evidently, muscular degeneration and impaired cognition in Duchenne muscular dystrophy (DMD) result from mutations in the DMD gene that give rise to the loss of dystrophin, a component of the DGC [1, 2]. This causes activation or disruption of a number of signaling pathways, improper displacement of SLO-1 channels, and abnormal vasoconstriction [3, 4, 5, 6, 7, 8, 10]. These changes greatly contribute to the decreased regulation of muscle excitability, which is characteristic of the disease, and leads to muscle fatigue [2]. While the studies examined in this review elucidated the mechanisms of different cell signaling events in the DGC, many failed to integrate the interactions of multiple signaling pathways. In addition, some

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

The “weakness” in single gene disorder research: the role of signaling pathways and vasoregulation in the dystrophin-glycoprotein complex (DGC) in Duchenne muscular dystrophy

studies did not incorporate other components besides dystrophin, such as its homologue, utrophin, into their analyses for comparison purposes. Future studies are needed to examine the relationship between different types of signaling pathways in the DGC and dystrophin’s role in VSMC in cardiovascular and skeletal pathologies that are attributed to loss of the signaling mechanisms [2,10]. In addition, other members of the DGC should be further examined, as other proteins can have specific interactions with DGC components [12]. Examining the combined interactions among DGC, cell signaling pathways, vasoconstriction, and SLO-1 channels would be beneficial for future studies of DMD, specifically as they relate to total body muscle stability and cognitive impairments. In each of these studies, however, it is necessary to remember that using mouse mdx or C. elegans models might not relate to the symptoms and effects observed in human DMD [5]. Effective studies will allow for a better understanding of the genetics and cell signaling events of single gene diseases. In this way, the gaps or weaknesses that remain in larger studies of single gene disorder research can slowly be eliminated to ultimately allow new therapies in pharmacology and genetic counseling to be established, and health policies to be revised.

Acknowledgements

I would like to thank the Department of Cell and Molecular Biology for sparking my interest in gene therapy research. I would also like to thank the Department of Neuroscience at the University of Toronto for providing me with the freedom to take a wide breadth of interesting courses that relate to genetics and disease research. The author declares no conflicts at this time.

References

1. Eason J. Prenatal diagnosis of single gene disorders. Obstet Gynecol Reprod Med. 2010; 20: 155-160. 2. Perronnet C, Vaillend C. Dystrophins, utrophins, and associated scaffolding complexes: role in mammalian brain and implications for therapeutic strategies. J Biomed Biotechnol. 2010; 2010: 849426. 3. Kim M, Kay D, Rudra R, Chen B, Hsu N, Izumiya Y, et al. Myogenic Akt signaling attenuates muscular degeneration, promotes myofiber regeneration and improves muscle function in dystrophin-deficient mdx mice. Hum Mol Genet 2011; 20 (7): 1324-1338. 4. Sarkis J, Vie V, Winder S, Renault A, Le Rumeur E, Hubert J-F. Resisting sarcolemmal rupture: dystrophin repeats increase membrane-actin stiffness. FASEB J 2012; doi:10.1096/fj.12-208967. 5. Sancar F, Touroutine D, Gao S, Oh HJ, Gendrel M, Bessereau JL, et al. The dystrophin-associated protein complex maintains muscle excitability by regulating Ca2+-dependent K+ (BK) channel localization. J Biol Chem. 2011; 286 (38): 33501-33510. 6. Ramachandran J, Schneider J, Crassous P-A, Zheng R, Gonzalez J, Xie L-H. Nitric Oxide signaling pathway in Duchenne muscular dystrophy mice: upregulation of L-arginine transporters. Biochem J 2012; doi:10.1042/BJ20120787. 7. Dogra C, Changotra H, Wergedal J, Kumar A. Regulation of Phosphatidylinositol 3-Kinase (PI3K)/Akt and Nuclear Factor-Kappa B signaling pathways in dystrophin-deficient skeletal muscle in response to mechanical stretch. J Cell Physiol 2006; 208: 575-585. 8. Delfin D, Xu Y, Peterson J, Guttridge D, Rafael-Fortney J, Janssen P. Improvement of cardiac contractile function by peptide-based inhibition of NF-kB in the utrophin/dystrophin-deficient murine model of muscular dystrophy. J Transl Med 2011; 9 (68): doi: 10.1186/1479-5876-9-68. 9. Nigro G, Comi LI, Politano L, Bain RJI. The incidence and evolution of cardiomyopathy in Duchenne muscular dystrophy. Int J Cardiol 1990; 26 (3) : 271-277. 10. Ito K, Kimura S, Ozasa S, Matsukura M, Ikezawa M, Yoshioka K, et al. Smooth muscle-specific dystrophin expression improves aberrant vasoregulation in mdx mice. Hum Mol Genet. 2006; 15 (14): 2266-2275. 11. Cazzella V, Martone J, Pinnaro C, Santini T, Twayana S, Sthandier O, et al. Exon 45 skipping through U1-snRNA antisense molecules recovers the Dys-nNOS pathway and muscle differentiation in human DMD myoblasts. Mol. Ther. 2012; doi: 10.1038/mt.2012.178. 12. Joe M, Kee C, Tomarev S. Myocilin interacts with syntrophins and is member of dystrophinassociated protein complex. J Biol Chem 2012; 287 (16) : 13216-13227.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

REVIEWS

The coevolutionary arms race between pathogen and plant host Wenwan Lu 1, Darrell Desveaux 2 Cells and Systems Biology, University of Toronto Email: luwenwan@hotmail.com 2 Email: darrell.desveaux@utoronto.ca 1

Abstract

The coevolutionary arms race between pathogen and plant host has prompted diverse pathogen virulence strategies and has modified plant surveillance and defense systems. Different models have been postulated to explain the pathogen-host interaction over evolutionary history. A classic bacterial pathogen, Pseudomonas syringae , utilizes the Type III secretion system to translocate its effectors into the plant cell to enhance bacterial virulence. During evolutionary time, the plant host evolved Resistance protein (R protein) in order to trigger a Hypersensitive Response (HR) to terminate the pathogenic virulence function.The bacterial pathogen retaliates by evolving new effectors to evade host innate immune response. This paper examines three different models of the host-pathogen system by discussing the roles of Type III secretion system effectors (T3SEs), host targets and R proteins in the plant surveillance system. Understanding the specific host-pathogen interaction is crucial in improving agriculture, food, and biofuel production [1].

Introduction

The coevolutionary arms race between pathogen and plant host has prompted diverse pathogen virulence strategies and modified plant surveillance and defense systems. The plant host must defend itself against pathogenic virulence in order to survive, while the pathogen has to evade or suppress host immune response in order to proliferate. Pseudomonas syringae is a Gram-negative, rod-shaped bacterial pathogen, used as a model for the study of specific host-pathogen interactions. Since mid-1980, P. syringae has demonstrated to be an effective genetically tractable pathogen in over 50 plant pathovars, such as soybean, tomato, and bean [2]. P. syringae enters plant tissues through wounds or natural openings such as stomata and hydathodes, and then continues to multiply in intercellular spaces. Once in the host tissue, P. syringae acquires plant resources, such as water and nutrients, and/or bypasses the plant innate immune response to promote bacterial growth in the apoplast [3-9]. The primary virulence mechanism employed by P. syringae is the Type III Secretion System (T3SS). Different pathogens have convergently evolved the T3SS to permit pathogen proliferation by secreting and translocating Type III Secreted Effector (T3SE) proteins directly into host tissues [10-11]. A single strain of P. syringae can deliver more than 30 effectors into plant hosts. Some effectors contribute to host cell structure modification, while some effectors play roles in acquiring nutrient, water and other resources for pathogen proliferation. Other effectors may function in the suppression of plant innate immune system. Different from vertebrates, plants do not have mobile immune cells. Instead, plants must effectively launch a self-reactive, selftolerant immune response [12]. The first layer of plant immunity involves Pathogen-Associated Molecular Patterns (PAMPs). The

plant immune system is triggered when transmembrane Pattern Recognition Receptors (PRRs) recognize PAMPs and subsequently trigger the PAMP-Triggered Immunity (PTI) to prevent further infection by the pathogen. There are different hallmarks of PTI, including alkalization, oxidative burst, and callose deposition [1314]. The successful pathogen responds by evolving or acquiring effectors to suppress or evade PAMP-Triggered Immunity (PTI), resulting in Effector Triggered Susceptibility (ETS) [15]. The pathogen imposes a strong selection pressure on the plant host by utilizing T3SS to evade the plant immune system. The plant host retaliates by evolving Resistance (R) proteins to recognize specific T3SEs, and subsequently activates plant Effector Triggered Immunity (ETI) to thwart pathogen infection, which is often accompanied by a localized and programmed cell death termed the Hypersensitive Response (HR) [15]. Consequently, it triggers another round of the arms race. For example, a common family of Type III secreted effectors found in plant and animals is the HopZ/ YopJ family. HopZ1a is predicted to be most similar to the ancestral HopZ allele. P. syringae carrying HopZ1a triggers a host immune response in Arabidopsis thaliana, rice, sesame and soybean and Nicotiana benthamiana [5, 16]. However, the closely related effector HopZ2 has evolved to evade and suppress host immune response by having virulence function on the plant host Arabidopsis thaliana [16]. Natural selection drives the coevolutionary arms race between the pathogen and the plant by evolving new effector genes and additional corresponding R genes, respectively [17]. The host R protein is essential to the plant’s surveillance system for pathogen detection. In most cases, R proteins are defined as having nucleotide-binding and leucine-rich repeat (NB-LRR) domains, while their N-terminal domains are structurally and

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

The coevolutionary arms race between pathogen and plant host

activity in the plant cell. Direct interaction between pathogen avirulence proteins and plant resistance proteins has been demonstrated using yeast two-hybrid assays for flax rust fungus. In order to avoid recognition by the host NB-LRR protein, the geologically diverse rust flax fungus AvrL567 gene undergoes sequence diversification instead of loss of virulence function. Both the pathogen Avr gene and the plant NB-LRR protein undergo diversifying selection, which suggests a direct gene-specific arms race [26]. Dodds et al did not present the first demonstration of direct Avr-R interaction. However, their experiments well-illustrate the coevolutioanry arms race between pathogen Avr gene and plant R protein by demonstrating the correlation between sequence diversification and protein-protein interaction under the pressure of diversifying selection.

The Guard Model

Figure 1: Two different conditions when T3SEs enter into the plant cell wall. The plant hosts elicit a effector-triggered susceptibility and subsequently get diseased when the interaction among R proteins, host targets, and effectors is not recognized. Otherwise, the effector-triggered immunity is elicited and plant hosts terminate/reduce the bacterial growth by triggering a localized cell death–hypersensitive response.

functionally diverse. The N-terminal domains are usually composed of two categories: coiled-coil (CC) domain or a TIR domain. R proteins from the CC-NBS-LRR category, such as RPS2, RPM1, and RPS5, require NDR1 protein’s membrane localization ability [18-20]. R proteins from the TIR-NBS-LRR class, such as RPS4 from P.syringe and RPP2 from Hyaloperonospora Arabidopsis, requires EDS1 and PAD4 to elicit a ETI response [21-23]. In the model plant organism Arabidopsis thaliana, ~170 putative R genes are present in the genome [24]. The gene-for-gene interaction first discovered by H.H. Flor marked an important milestone in our understanding of the pathogen-host system. Flor et al worked on rust (Melamposora lini) of flax (Linum usitatissimum) on the molecular level and illustrated that a resistance (R) gene in the host can recognize an avirulence (avr) gene in the pathogen [25]. Also, in the absence of recognition by the R gene, the avr gene might function as a virulence gene to suppress the plant defense system [11, 17]. Even though dozens of R-avr gene interactions have been characterized, Avr proteins are now being referred to as effectors to stress their intrinsic virulence function. Different models have been postulated to explain how the host plant utilizes the R protein to recognize and interact with T3SEs.

The Receptor-Ligand Model

The receptor-ligand model postulates that plant host R proteins interact directly with the T3SEs. Therefore, host R proteins terminate/ reduce the effectors’ virulence function by recognizing the effectors’

The guard hypothesis has been proposed to explain the lack of direct interaction between the effector and the R protein. The guard model was originally postulated to explain the indirect interaction between P. syringae avirulence protein Pto and tomato resisitance proteins Prf through the host target protein Pto [27]. Later, multiple indirect interactions between R proteins and effector have been demonstrated, and the significance of host targets acting as a guardee protein is emphasized in the guard hypothesis. It postulates that effector proteins originally act as virulence factors on host targets. Therefore, the host target plays a role in promoting pathogen fitness in the absence of host recognition. Later in the evolutionary time, the plant host evolves an R protein acting as a guardee that indirectly monitors the effector’s activity on a host target and triggers the host surveillance system [15,16,28]. Also, one host target may be targeted by more than one independently evolved effector and/or more than one host R protein. For example, Arabidopsis RPM1- interacting protein (RIN4) acts as a host target guarding the activities of different effector proteins, providing a classic example for the guard model. Three different effectors (AvrB, AvrRpm1 and AvrRpt2) target the Arabidopsis RPM1- interacting protein (RIN4). In the meantime, two R proteins (RPM1 and RPS2) monitor the effector activity through the guarded effector protein RIN4. In the absence of the host R protein RPM1 and RPS2, the guarded effector proteins RIN4 is indispensible for AvrRpm1 and AvrRpt2’s virulence function [29-31].

The Decoy Model

In the guard hypothesis, the effector proteins modify the host target (guardee) protein and promote its own virulence by modifying the guardee protein in the absence of host resistance protein. In contrast to the guard hypothesis, the decoy model proposes that host decoy proteins, with which target effectors interact, silence the effectors’ modification ability and thus limit the pathogen fitness in the absence of host R protein. The host decoy protein’s only function is to mimic the guardee host target, and to allow ETI response to trigger with the R protein Prf recognition [9]. The major difference between the guard and decoy models is the function of the host targets. In the absence of the host R protein Prf, T3SE’s manipulation of the decoy proteins provides no benefit to the pathogen. In contrast, in the guard model, effectors’ manipulation of the guardee host target promotes pathogen fitness in the absence of the R protein. P. syringae AvrPto is a kinase inhibitor that targets Pto in tomato and Arabidopsis, in order to suppress host innate

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

The coevolutionary arms race between pathogen and plant host

immune system and subsequently promote its own virulence function. AvrPto has been demonstrated to disrupt the function of FLS2 and EF2R and specifically target FLS2 in the host cell in Arabidopsis. It is interesting to note that Pto acting as a host target also actively binds with FLS2 in competition with AvrPto, therefore dramatically reduce the virulence function of AvrPto [32-33]. RIN4 may also be a decoy host target for different effectors. Both AvrRpm1 and AvrRpt2 also appear to target other hosts by promoting the pathogen fitness in rin4 knockout plants in comparison to ecotype plants. No evidence has been proposed whether RIN4 fits the guard model or decoy model at this stage. In the absence of the host R protein, whether RIN4 promotes or silences/reduces the pathogen fitness is the key to distinguish the difference between the guarded effector protein and the decoy effector protein. Although many R-Avr gene pairs have been well documented, the ligand-receptor model has been under intense debate over the

Review Articles

past decade. The diversifying selection acting on the flax rust fungus AvrL567 genes, have suggested gene-specific coevolutioy arms race between the effector genes and their corresponding R protein. However, evidences supporting direct physical Avr-R interaction are still lacking. An alternative model, the guard hypothesis, was proposed suggesting indirect interaction between effector proteins and the host R protein through host targets. In the absence of the host R protein, the host guard protein evolves to have lower affinity towards the effector to evade ETI suppression. Meanwhile, in the presence of host R protein, the host guard protein is also under pressure to have high affinity towards the effector to enhance R protein recognition. In the guard model, the host target (guardee) undergoes opposing selection pressures, therefore forming an evolutionary unstable condition for the guardee protein. As suggested by Renier et al, the decoy model solves the evolutionary unstable situation the host target may be facing [9]. The decoy model proposes that the only function of decoy protein is to mimic a target protein. In this case, the decoy proteins undergo a selection pressure to continuously increase their binding affinity towards the effector proteins. As a result, the decoy proteins specialize in silencing the virulence function of effector proteins under uniform evolutionary selection.

Challenges and Opportunities for the Future

Figure 2: Different models illustrate the interaction between T3SEs, host targets, and R proteins on the plant surveillance system. (A) The receptor-ligand model, in the presence of R protein, Effector-Triggered Immunity (ETI) is induced and therefore terminate/reduce the pathogen virulence. (B) The guard model illustrates two scenarios. In the presence of R protein, R protein recognizes effector activity through host target modification and ETI is subsequently induced. In guard model, with the absence of R protein, effectors modify the host target (guardee) and enhance pathogen fitness. (C) The decoy model postulates that ETI is triggered in the presence of R protein by recognizing the interaction between effectors and the host decoy target. In the absence of R protein, no enhancement to pathogen fitness is observed due to the modification of host decoy target by effectors.

The receptor-ligand model only involves the direct interaction between the host R protein and the effector protein. The receptor-ligand model can be instantly excluded once the plant host target is identified. It remains to be a challenge to discriminate the difference between the guard model and the decoy model. An intermediate stage may be present, in which a guard protein may evolve to a decoy host target. In addition, multiple effectors presented in the plant cell may produce confounding results for genetic analysis. During infection, a pathogen strain, such as P. syringae strain, can deliver more than 30 T3SEs into a plant host. As a result, functional redundancy is likely to take place. In that case, we cannot investigate the function of a specific T3SE because several other T3SEs might together contribute their virulence function to the infection of Arabidopsis. An alternative approach is to investigate the function of a specific effector by delivering the effector protein using a transgenic plant. Transgenic plants were created to have transgenes with specific effector sequence enclosed in a vector. In this way, we can exclude the interference of other effectors and only focus on the study of the function of one overexpressed effector. However, the transgene might interrupt a gene in the host, which in turn might affect a pathogen-related phenotype. One strategy to deal with this situation is to use multiple independent lines of the same effector and confirm the pathogen-related phenotypes. Understanding the

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

The coevolutionary arms race between pathogen and plant host

specific host-pathogen interaction on a molecular basis is crucial to the improvement of agriculture, food, and biofuel production. By determining the evolutionary history of the host NB-LRR genes and the pathogen effectors, we can apply the current understanding of the Type III secretion system to effective disease control in agriculture in the future.

Acknowledgments

This study was funded by the Cell and System Biology Department CSB497 program. The author would like to thank Professor Darrell Desveaux for giving me this opportunity to work in his lab and Dr. Jennifer Lewis for her encouragement and support. The author also acknowledges Timothy Lo, Fengyi Cao, Brenden Hurley, Michael Wilton, and Karl. J. Screiber for their thoughtful and constructive comments on this paper.

References

1. Hirano SS, Upper CD. Bacteria in the Leaf Ecosystem with Emphasis on Pseudomonas syringae---a Pathogen, Ice Nucleus, and Epiphyte.Microbiology and Molecular Biology Reviews, 2000. 64 (3): 624–53. 2. Keen NT. Gene-For-Gene Complementarity in Plant-Pathogen Interactions. annual review of genetics, 1990.24: 447-463. 3. Grant SR, Fisher EJ, Chang JH., Mole BM, Dangl JL. Subterfuge and manipulation: type III effector proteins of phytopathogenic bacteria. Annu Rev Microbiol, 2006. 60: 425–449. 4. Lewis JD, Desveaux D and Guttman DS. The targeting of plant cellular systems by injected type III effector proteins. Semin. Cell Dev. Biol. 2009. 20: 1055–1063. 5. Ma WB, Dong FFT, Stavrinides J and Guttman DS. Type III effector diversification via both pathoadaptation and horizontal transfer in response to a coevolutionary arms race. PLoS Genet. 2006. 2: e209. 6. Mudgett MB. New insights to the function of phytopathogenic bacterial type III effectors in plants. Annu Rev Plant Biol, 2005. 56: 509–531. 7. Mukherjee S, Keitany G, Li Y, Wang Y, Ball HL, et al. Yersinia YopJ acetylates and inhibits kinase activation by blocking phosphorylation. Science, 2006. 312: 1211–1214. 8. Zhou JM, Chai J. Plant pathogenic bacterial type III effectors subdue host responses. Curr Opin Microbiol, 2008. 11: 179–185. 9. Van der Hoorn RAL, Kamoun S. From guard to decoy: A new model for perception of plant pathogen effectors. Plant Cell, 2008. 20: 2009 –2017. 10. Chang JH, Goel AK, Grant SR., and Dangl JL. Wake of the flood: ascribing functions to the wave of type III effector proteins of phytopathogenic bacteria. Curr. Opin. Microbiol, 2004. 7: 11–16. 11. Katagiri F, Thilmony R, and He SY. The Arabidopsis thaliana Pseudomonas syringae interaction, p. 1–35. In C. R. Somerville and E. M. Meyerowitz (ed.), The Arabidopsis book. American Society of Plant Biologists, 2002. Rockville, MD. 12. Spoel SH, and Dong X. How do plants achieve immunity? Defence without specialized immune cells. Nature Rev. Immunol, 2012. 12: 89–100. 13. Felix G, Duran JD, Volko S, Boller T. Plants have a sensitive perception system for the most conserved domain of bacterial ﬂagellin. Plant J, 1999. 18: 265–276. 14. Gomez-Gomez L and Boller T. FLS2: an LRR receptor-like kinase involved in the perception of the bacterial elicitor flagellin in Arabidopsis. Mol. Cell, 2000. 5:1003–1011. 15. Jones JD, and Dangl JL. The plant immune system. Nature , 2006. 444: 323–329. 16. Lewis JD, Abada W, Ma WB, Guttman DS. and Desveaux D. The HopZ family of Pseudomonas syringae type III effectors require myristoylation for virulence and avirulence functions in Arabidopsis thaliana. J. Bacteriol, 2008.190: 2880–2891. 17. Dawkins R., Krebs JR. Arms races between and within species. Proc R Soc Lond B Biol Sci, 1979. 205: 489–511. 18. Century KS, Holub EB, Staskawic BJ. NDR1, a locus of Arabidopsis thaliana that is required for disease resistance to both a bacterial and a fungal pathogen. Proc Natl Acad Sci U S A , 2005. 92:6597-6601 19. Century KS, Shapiro AD, Repetti PP, Dahlbeck D, Holub E, et al. NDR1, a pathogen-induced component required for Arabidopsis disease resistance. Science, 1997. 278:1963-1965. 20. Coppinger P, Repetti PP, Day B, Dahlbeck D, Mehlert A, et al. Overexpression of the plasma membrane-localized NDR1 protein results in enhanced bacterial disease resistance in Arabidopsis thaliana. Plant J, 2004. 40:225-237. 21. Aarts N, Metz M, Holub E, Staskawicz BJ, Daniels MJ, et al. Different requirements for EDS1 and NDR1 by disease resistance genes define at least two R gene-mediated signaling pathways in Arabidopsis. Proc Natl Acad Sci U S A, 1999. 95:10306 – 10311. 22. Falk A, Feys BJ, Frost LN, Jones JDG, Daniels MJ, et al. EDS1, an essential component of R gene-mediated disease resistance in Arabidopsis has homology to eukaryotic lipases. Proc Natl Acad Sci U S A, 1999. 96:3292-3297. 23. Feys BJ, Moisan LJ, Newman MA, Parker JE. Direct interaction between the Arabidopsis disease resistance signaling proteins, EDS1 and PAD4. EMBO J, 2001. 20: 5400-5411

24. Lewis JD, Wu R, Guttman DS. and Desveaux D. Allele-specific virulence attenuation of the Pseudomonas syringae HopZ1a type III effector via the Arabidopsis ZAR1 resistance protein. PLoS Genet, 2010. 6, e1000894. 25. Flor HH. Current status of the gene-for-gene concept. Annu. Rev. of Phytopathol, 1971. 9: 275-296. 26. Dodds PN. Lawrence GJ, Catanzariti AM, The T, Wang CI., Ayliffe MA., Kobe B, and Ellis JG. Direct protein interaction underlies gene-for-gene specificity and coevolution of the flax resistance genes and flax rust avirulence genes. Proc. Natl. Acad. Sci. USA, 2006. 103:8888-8893 27. Van der BiezeN EA, and Jones JDG. Plant disease-resistance proteins and the gene-for-gene concept. Trends Plant Sci, 1998. 23:454-456. 28. Collier SM, Moffett P. NB-LRR work a “bait and and switch” on pathogens. Trends Plant Sci, 2009. 14: 512-529. 29. Mackey D, Holt BF, Wiig A, Dangl JL . RIN4 interacts with Pseudomonas syringae type III effector molecules and is required for RPM10mediated resistance in resistance. Cell, 2002. 108:743-758. 30. Macky D, Belkharir Y, Alonso JM, Ecker JR, Dangle JL. Arabidopsis RIN4 is a target of the type III virulence effector AvrRpt2 and modulates RPS2-mediated resistance. Cell, 2003. 112:379-389. 31. Axtell MJ, Staskawicz JB. Initiation of RPS2-specified disease resistance in Arabidopsis is coupled to the AvrRpt2-directed elimination of RIN4. Cell, 2003. 112:369-377. 32. Zhou JM and Chai J. Plant pathogenic bacterial type III effectors subdue host responses. Curr. Opin. Microbiol, 2008. 11:179-185. 33. Zipfel C and Rathjen JP. Plant Immunity: AvrPto targets the frontline. Curr. Biol, 2008.18: R218- R220.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

REVIEWS

The mammalian unfolded protein response: A potential source of novel therapeutic targets in the hypoxic tumour fraction Brandon Sit University of Toronto, brandon.sit@mail.utoronto.ca

Abstract

Low oxygenation (hypoxia) in solid cancers contributes to tumour resistance to standard treatment and correlates strongly with negative patient prognoses. Hypoxia activates the mammalian unfolded protein response (UPR), an ER-localized signaling program that regulates transcription and translation in response to an accumulation of misfolded polypeptides. In cancer, UPR signaling is known to be active and important in at least enabling cells to survive hypoxia. The UPR is also hypothesized to play a role in carcinogenesis. The UPR signals through three ER transmembrane sensors, each with their own respective downstream pathways. Depending on the stress input, UPR signaling may result in one of two outcomes – cell adaptation to stress, or apoptosis. While each separate pathway is well studied, few efforts have been made to integrate them to assess their relative importance in contributing to either outcome. This review provides an overview of the current understanding of UPR signaling under hypoxia and highlights an as-yet incompletely understood question of whether the UPR can be artificially manipulated to preferentially signal towards one output, potentially acting as a novel therapeutic target in the hypoxic tumour fraction.

Introduction

The heterogeneity of solid cancers is a well-known contributor to tumour resistance to drug and radiation treatment and negative patient outcomes. This variation can be observed within a single malignancy not only between tumour cells at the genetic and epigenetic levels [1], but at the level of the microenvironment. The microenvironment of a solid tumour exhibits a marked heterogeneity in many factors such as cell composition, pH, nutrient availability and vascularization [2]. Each cell displays a distinct phenotype, which increases the likelihood of a tumour responding to a singular treatment in a highly non-uniform manner. Combinatorial therapies generally work more effectively, however dosages often become toxic to the patient before the full regimen can be completed.

Consequences of hypoxia for tumour biology

Tumour hypoxia (poor oxygenation) is one of the defining features of the heterogenous solid tumour microenvironment. Cancer cells generally experience one of two classes of hypoxia – chronic hypoxia, where insufficient vasculature and oxygen perfusion lead to oxygen-deprived regions, and acute hypoxia, which stems from random fluctuations in blood flow [3-5]. The former is commonly attributed to the high basal proliferation rate of cancer cells, which often outpaces the rate at which tumour angiogenesis

occurs. Hypoxic regions can define more than 70% of a tumour and are strongly linked to negative tumour traits such as increased metastatic potential and resistance to apoptosis linked to conventional chemo- and radiotherapies [6-9]. Radiation resistance arises from the fact that molecular oxygen (required to produce toxic oxygen species in radiated cells) is unavailable in this population. The hypoxic tumour fraction is also typically located far away from blood vessels, limiting the amount of a chemotherapeutic agent that can reach it. It appears that acute hypoxia is more important in creating aggressive tumour phenotypes, although this may vary by tumour subtype [5]. As a result of this oxygen deprivation, tumour cells can be differentiated from normal cells, thus highlighting hypoxic regions and their unique biology as possible targets for hypoxia-selective drugs. As shown in Figure 1, hypoxia modulates three major pathways in tumour cells, each of which present multiple targets for treatment. The most intensively studied hypoxia response mechanism is the stabilization and activation of the hypoxia inducible factors (HIFs). The HIFs are a family of transcription factors that are rapidly ubiquitinated and proteasomally degraded under normoxic conditions. This does not occur when there is a lack of cytosolic oxygen, allowing them to dimerize and begin upregulating a variety of hypoxia resistance genes such as VEGF, which promotes angiogenesis [10, 11]. The HIFs are also strongly implicated in a

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

The mammalian unfolded protein response: a potential source of novel therapeutic targets in the hypoxic tumour fraction

UPR signaling under hypoxia

IRE1 is the only conserved UPR sensor between yeast and mammals, and as such has been heavily studied on a structural and functional level [20,21]. First identified in yeast in 1993 and mammals in 1998, the most important role of IRE1 during the UPR is to catalyze the splicing of immature transcripts coding for X-box binding protein 1 (XBP1), a transcription factor that directs the upregulation of a subset of known ER stress response genes such as Bip, Derp and Erdj4 [22-25]. XBP1 has Figure 1: Hypoxia modulates tumour phenotypes and leads to more aggressive malignancies. also been implicated in the differentiation of Hypoxia activates signaling pathways downstream of the hypoxia-inducible factors (HIFs), the plasma cell lineages [26,27] and it has been mammalian target of rapamycin (mTOR) and the unfolded protein response (UPR), leading to in- suggested that IRE1 is more important for creased potential for metastasis and negative patient responses to conventional therapy. building ER secretory capacity than directing metabolic switch that focuses on oxygen-independent glycolysis a downstream stress response [28]. rather than oxygen-requiring oxidative phosphorylation. The Similar to IRE1, PERK is a kinase which autophosphorylates second pathway is through the suppression of the mammalian and homomultimerizes during its activation. The main function target of rapamycin (mTOR), a central kinase involved in cell of PERK is to initiate the global attenuation of protein translation, growth, proliferation and survival. mTOR forms two complexes which reduces the workload of the ER and hence ER stress. This is – mTORC1 and mTORC2, of which mTORC1 is implicated in the carried out through the phosphorylation of eukaryotic initiation hypoxic response [12]. Under hypoxia, mTOR cannot phosphory- factor 2, eIF2a. eIF2a is required for the initiation of cap-depenlate its primary targets p70-S6 kinase 1 (S6K1) and eIF4E-binding dent translation, the mechanism by which most mRNAs in the cell protein 1 (4EBP1), both of which are critical to the initiation of are processed. This phosphorylation and translational attenuation global protein translation [13]. It is thought that such translational is well characterized under hypoxia [29]. PERK is known to be repression would help the cell to reduce its energy requirement essential for cell survival under hypoxic stress and has also been and hence oxygen consumption. shown to be embryonic lethal in a mouse embryonic fibroblast The third hypoxia-modulated pathway, which forms the focus (MEF) model. PERK also signals downstream of p-eIF2a through of this review, is the unfolded protein response (UPR). The UPR is the transcriptional upregulation of activating transcription factor 4 activated in situations where endoplasmic reticulum (ER) homeo- (ATF4), which upregulates ER stress response genes, notably those stasis is perturbed and an accumulation of unfolded or misfolded encoding chaperones [30]. ATF4-/- MEFs are significantly more proteins occurs in the ER lumen [14]. This “ER stress” can be sensitive to hypoxia than their ATF4+/+ counterparts. It has been caused by a multitude of factors, one of which is severe hypoxia. proposed that PERK-ATF4 activity is responsible for the majorWith regard to cell survival under ER stress, it is accepted that ity of UPR-associated transcriptional activity based on microarray unfolded proteins form toxic, insoluble aggregates that are lethal evidence in HeLa cells [31]. to the cell. The mechanism by which hypoxia leads to ER stress ATF6 is part of a class of AP-1 transcription factors and is the remains undefined although the general consensus centres on least well studied of the three UPR stress sensors. Under unstressed the requirement of oxygen for the formation of disulfide bridges conditions, ATF6 is localized to the ER membrane as a population during protein maturation. It is known that extremely low levels of monomers, dimers and oligomers [31, 32]. When ER homeostasis of hypoxia (or anoxia, 0% O2) robustly activate the UPR [13]. The is disrupted, ATF6 is cleaved by the Site 1 and Site 2 proteases and UPR consists of three unique protein stress sensors localized to translocates to the nucleus to serve as a highly active transcription the endoplasmic reticulum (ER) membrane – inositol requiring factor. A recent study suggests ATF6 may be involved with Mediator enzyme 1 (IRE1), PKR-like RNA kinase (PERK) and activating recruitment to target gene promoters [33]. Cleaved ATF6 influences transcription factor 6 (ATF6). Each initiates its own distinctive ER the transcription of genes such as BiP, XBP1 and SERCA2 which all stress response pathway to regulate transcription and translation. contribute to the ER stress response [31, 34]. Interestingly, ATF6 activApart from cancer, the UPR has been implicated in metabolic dis- ity is synergistic with the activity of ATF4 and spliced XBP1 however orders such as Type II diabetes and obesity [15]. this interaction has not been fully explored [32, 35]. The generally accepted mechanism for UPR activation involves the binding of binding immunoglobulin protein (BiP) to UPR activity can signal for adaptation to stress or each of the three sensors, which represses their activity. Under apoptosis conditions of ER stress, this bound BiP is detached from each senWithin the last decade, a deeper understanding of the signalsor and moved to the ER to help fold misfolded proteins, freeing ing outcomes of the UPR has been achieved through several studeach UPR effector to activate its own downstream pathway [16-18]. ies that highlight the fact that the UPR does more than induce an Recent studies have shown that this binding activity may merely be adaptive response. In periods of prolonged or irreversible ER stress, a correlation and not a direct cause-effect interaction – an inter- the UPR may instead activate an apoptotic program, as illustrated esting paper showed that unfolded proteins may directly bind and in Figure 2. This can be achieved through any of the three signaling activate IRE1, however this remains incompletely understood [19]. arms of the response.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

The mammalian unfolded protein response: a potential source of novel therapeutic targets in the hypoxic tumour fraction

Review Articles

duration of activation [51, 52]. This concept of a “switch” between adaptation and apoptosis has only recently been seriously investigated, and as such is relatively poorly defined – several recent reviews on the subject have suggested much further research is needed [15, 53-55].

The UPR is a potential therapeutic target in solid cancers

Given the near-ubiquitous activation of the UPR and its importance to cell survival in solid malignancies, interest in small molecule inhibitors as modulators of the UPR and hence novel cancer therapeutics has grown significantly, mostly based on evidence obtained through genetic disruption of the response [56]. Drugs targeting the UPR can be currently classified into two broad categories – those increasing the ER stress in the cell (driving the UPR through to apoptosis) and those inhibiting UPR adaptation pathways. An excellent Figure 2: Hypoxic signaling through the UPR can either lead to stress adaptation or apoptosis. Through a currently undefined mechanism, UPR signaling can lead to two distinct out- example of the former is bortezomib (Velcade), comes for the cell – adaptation to a stressor or apoptosis. Each pathway is generally identified an FDA-approved proteasome inhibitor for late by upregulation of a specific set of genes as indicated, although both sets may be expressed stage multiple myeloma that blocks ERAD and insimultaneously. creases the speed and severity of unfolded protein IRE1 has been observed to modulate apoptosis under UPR buildup in the ER lumen [57, 58]. Several other activation by two mechanisms – through downstream phos- compounds such as sorafenib, paraquat, and ursolic acid have been phorylation of Jun n-terminal kinase (JNK) and activation of shown to increase ER stress and, in the case of paraquat and ursolic IRE1-dependent decay of RNAs (RIDD). In a key study published acid, to activate the IRE1-JNK apoptosis pathway [59-62]. To this in 2000, Urano et al. showed that IRE1 associated with TRAF2 end, an interesting approach utilizing peptides to inhibit the activ(TNF receptor-associated factor 2) to cause the phosphorylation ity of BiP and push the UPR through to apoptosis in a prostate of JNK, which is strongly linked to apoptosis [36-41]. In addition cancer model has been investigated [63]. to association with TRAF2, IRE1 can affect apoptosis through the A recent number of papers have identified small molecule degradation of RNAs through RIDD. RIDD occurs when IRE1 inhibitors of IRE1 and PERK, 2 of the 3 UPR stress transducers. A is hyperactive (i.e. active for a prolonged period of time) and is a majority of the IRE1 inhibitors identified so far are aldehydes that non-specific method of mRNA degradation [42,43]. While RIDD inhibit IRE1 endogenous RNase activity (i.e. XBP1 splicing and could present a survival mechanism (by decreasing translational RIDD), and not IRE1 kinase activity [28, 64-68]. Specific residues load) it is thought of more as a means to apoptosis through the within IRE1 such as Lys599 have been identified as important for degradation of key survival transcripts such as those encoding BiP endonuclease compared to kinase activity [28]. In the context of or XBP1. Interestingly, a recent study illustrated a putative link cancer, IRE1 inhibitors have been shown to exhibit anticancer from mTORC1 to the IRE1-JNK pathway, suggesting that different activity in vitro and in vivo in two separate studies of multiple hypoxia response pathways may be interrelated [44]. myeloma [64, 65]. Given that IRE1 signaling still remains incomPERK is connected to apoptosis through the ATF4-dependent pletely understood, the majority of these molecules are largely used upregulation of CHOP (C/EBP homologous protein), which is as research tools and not as potential therapeutics [69]. considered a proapoptotic transcription factor [45-49]. In a pivotal As a result of PERK lacking the nuclease activity seen with 2004 study, CHOP was shown to induce GADD34, a phosphatase IRE1, inhibitors are generally targeted to the ATP-binding region that specifically dephosphorylates p-eIF2a in a negative feedback important for its kinase activity. In a landmark 2011 study, Wang loop [49]. This returns the ER to a pre-stressed state and promotes and colleagues reported the first pharmacophore model of PERK renewed protein synthesis, which in a continuously stressed sys- to be built, specifically identifying Met7 and Asp144 as required tem is a proapoptotic action. contacts for specific inhibitor binding to the protein [70]. Several As the least studied stress sensor, ATF6 has not been linked other inhibitors of PERK kinase activity are in development as solidly to apoptosis although an isolated study suggests it may me- specific anticancer agents although none have progressed past diate apoptosis through its own unique pathway [50]. mouse model investigation. Given that the UPR activates all three arms differentially and The above options are directed at inhibiting adaptive UPR concomitantly, it has been difficult to elucidate the specific effect of pathways in cancer cells – however, small molecule activators each arm on cell fate. Two recent studies by Lin et al. clearly show targeted at apoptotic UPR pathways instead could present another divergent cell fates under stress when either IRE1 or PERK sig- potential treatment modality. For example, in yeast it has been naling was sustained artificially, suggesting each arm contributes discovered that certain flavonols (flavonoids characterized by a differently to the decision between life and death depending on its 3-hydroxyflavone backbone) activate IRE1 activity [71]. If this can Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

The mammalian unfolded protein response: a potential source of novel therapeutic targets in the hypoxic tumour fraction

be translated into a mammalian model, then it is reasonable to speculate that artificially prolonged IRE1 activation may promote apoptosis instead of adaptation in response to ER stress.

Conclusion

The unfolded protein response is activated and important in the hypoxic fraction of solid human tumours. It is important to note that although this review focused on the hypoxia-induced UPR, the UPR can be activated by a host of other factors (i.e. disruption of Ca2+ homeostasis or changes in ER lumen oxidation state) through different mechanisms. No specific study exists to compare modes of activating the UPR, but given the exceptional precision of the response, it would be reasonable to expect that a different mode of activation may lead to different downstream outcomes. Regardless of cause, the UPR is highly active in cancer cells and can be used to differentiate tumour phenotypes from normal ones. Thus, there is much potential for targeted therapies to be developed against this response to selectively kill tumour cells, possibly targeting the hypoxic fraction in the process.

References

1. Gerlinger M, Rowan AJ, Horswell S, Larkin J, Endesfelder D, Gronroos E, et al. Intratumor Heterogeneity and Branched Evolution Revealed by Multiregion Sequencing. N Engl J Med 2012 03/08; 2012/07;366(10):883-892. 2. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell 2011 Mar 4;144(5):646-674. 3. Powathil G, Kohandel M, Milosevic M, Sivaloganathan S. Modeling the spatial distribution of chronic tumor hypoxia: implications for experimental and clinical studies. Computational and mathematical methods in medicine 2012;2012:410602. 4. Wang K, Yorke E, Nehmeh SA, Humm JL, Ling CC. Modeling acute and chronic hypoxia using serial images of 18F-FMISO PET. Med Phys 2009 Oct;36(10):4400-4408. 5. Bayer C, Vaupel P. Acute versus chronic hypoxia in tumors : Controversial data concerning time frames and biological consequences. Strahlentherapie und Onkologie : Organ der Deutschen Rontgengesellschaft [et al] 2012. 6. Vaupel P, Mayer A. Hypoxia in cancer: significance and impact on clinical outcome. Cancer Metastasis Rev 2007 Jun;26(2):225-239. 7. Yokoi K, Fidler IJ. Hypoxia increases resistance of human pancreatic cancer cells to apoptosis induced by gemcitabine. Clin Cancer Res 2004 Apr 1;10(7):2299-2306. 8. Walsh S, Gill C, O’Neill A, Fitzpatrick JM, Watson RW. Hypoxia increases normal prostate epithelial cell resistance to receptor-mediated apoptosis via AKT activation. Int J Cancer 2009 Apr 15;124(8):1871-1878. 9. Park S, Billiar T, Seol D. Hypoxia inhibition of apoptosis induced by tumor necrosis factor-related apoptosis-inducing ligand (TRAIL). Biochem Biophys Res Commun 2002;291(1):150-153. 10. Appelhoff R, Tian Y, Raval R, Turley H, Harris A, Pugh C, et al. Differential function of the prolyl hydroxylases PHD1, PHD2, and PHD3 in the regulation of hypoxia-inducible factor. The Journal of biological chemistry 2004;279(37):38458-38523. 11. Carmeliet P, Dor Y, Herbert J, Fukumura D, Brusselmans K, Dewerchin M, et al. Role of HIF-1alpha in hypoxia-mediated apoptosis, cell proliferation and tumour angiogenesis. Nature 1998;394(6692):485-575. 12. Brugarolas J, Lei K, Hurley RL, Manning BD, Reiling JH, Hafen E, et al. Regulation of mTOR function in response to hypoxia by REDD1 and the TSC1/TSC2 tumor suppressor complex. Genes Dev 2004 Dec 1;18(23):2893-2904. 13. Wouters BG, Koritzinsky M. Hypoxia signalling through mTOR and the unfolded protein response in cancer. Nat Rev Cancer 2008 Nov;8(11):851-864. 14. Ron D, Walter P. Signal integration in the endoplasmic reticulum unfolded protein response. Nature reviews Molecular cell biology 2007;8(7):519-548. 15. Hetz C. The unfolded protein response: controlling cell fate decisions under ER stress and beyond. Nature reviews Molecular cell biology 2012;13(2):89-191. 16. Pincus D, Chevalier M, AragÃ³n T, van Anken E, Vidal S, El-Samad H, et al. BiP binding to the ER-stress sensor Ire1 tunes the homeostatic behavior of the unfolded protein response. PLoS biology 2010;8(7). 17. Sou S, Ilieva K, Polizzi K. Binding of human BiP to the ER stress transducers IRE1 and PERK requires ATP. Biochem Biophys Res Commun 2012;420(2):473-481. 18. McKibbin C, Mares A, Piacenti M, Williams H, Roboti P, Puumalainen M, et al. Inhibition of protein translocation at the endoplasmic reticulum promotes activation of the unfolded protein response. Biochem J 2011. 19. Gardner B, Walter P. Unfolded proteins are Ire1-activating ligands that directly induce the unfolded protein response. Science (New York, N Y ) 2011;333(6051):1891-1895. 20. Korennykh A, Korostelev A, Egea P, Finer-Moore J, Stroud R, Zhang C, et al. Structural and functional basis for RNA cleavage by Ire1. BMC biology 2011;9:47.

21. Ron D, Hubbard S. How IRE1 reacts to ER stress. Cell 2008;132(1):24-30. 22. Cox JS, Shamu CE, Walter P. Transcriptional induction of genes encoding endoplasmic reticulum resident proteins requires a transmembrane protein kinase. Cell 1993 Jun 18;73(6):1197-1206. 23. Lee A, Iwakoshi N, Glimcher L. XBP-1 regulates a subset of endoplasmic reticulum resident chaperone genes in the unfolded protein response. Mol Cell Biol 2003;23(21):7448-7507. 24. Dong M, Bridges J, Apsley K, Xu Y, Weaver T. ERdj4 and ERdj5 are required for endoplasmic reticulum-associated protein degradation of misfolded surfactant protein C. Mol Biol Cell 2008;19(6):2620-2650. 25. Lai C, Otero J, Hendershot L, Snapp E. ERdj4 protein is a soluble endoplasmic reticulum (ER) DnaJ family protein that interacts with ER-associated degradation machinery. The Journal of biological chemistry 2012;287(11):7969-8047. 26. Xu G, Liu K, Anderson J, Patrene K, Lentzsch S, Roodman GD, et al. Expression of XBP1s in bone marrow stromal cells is critical for myeloma cell growth and osteoclast formation. Blood 2012 May 3;119(18):4205-4214. 27. Taubenheim N, Tarlinton DM, Crawford S, Corcoran LM, Hodgkin PD, Nutt SL. High Rate of Antibody Secretion Is not Integral to Plasma Cell Differentiation as Revealed by XBP-1 Deficiency. J Immunol 2012 Oct 1;189(7):3328-3338. 28. Cross BCS, Bond PJ, Sadowski PG, Jha BK, Zak J, Goodman JM, et al. The molecular basis for selective inhibition of unconventional mRNA splicing by an IRE1-binding small molecule. Proceedings of the National Academy of Sciences 2012 April 10;109(15):E869-E878. 29. Fels D, Koumenis C. The PERK/eIF2alpha/ATF4 module of the UPR in hypoxia resistance and tumor growth. Cancer biology & therapy 2006;5(7):723-731. 30. Bull VH, Thiede B. Proteome analysis of tunicamycin-induced ER stress. Electrophoresis 2012 Jul;33(12):1814-1823. 31. Okada T, Yoshida H, Akazawa R, Negishi M, Mori K. Distinct roles of activating transcription factor 6 (ATF6) and double-stranded RNA-activated protein kinase-like endoplasmic reticulum kinase (PERK) in transcription during the mammalian unfolded protein response. Biochem J 2002;366:585-679. 32. Adachi Y, Yamamoto K, Okada T, Yoshida H, Harada A, Mori K. ATF6 is a transcription factor specializing in the regulation of quality control proteins in the endoplasmic reticulum. Cell Struct Funct 2008;33(1):75-164. 33. Sela D, Chen L, Martin-Brown S, Washburn MP, Florens L, Conaway JW, et al. Endoplasmic Reticulum Stress-responsive Transcription Factor ATF6α Directs Recruitment of the Mediator of RNA Polymerase II Transcription and Multiple Histone Acetyltransferase Complexes. Journal of Biological Chemistry 2012 June 29;287(27):23035-23045. 34. Li M, Baumeister P, Roy B, Phan T, Foti D, Luo S, et al. ATF6 as a transcription activator of the endoplasmic reticulum stress element: thapsigargin stress-induced changes and synergistic interactions with NF-Y and YY1. Mol Cell Biol 2000;20(14):5096-5202. 35. Teske B, Wek S, Bunpo P, Cundiff J, McClintick J, Anthony T, et al. The eIF2 kinase PERK and the integrated stress response facilitate activation of ATF6 during endoplasmic reticulum stress. Mol Biol Cell 2011;22(22):4390-4795. 36. Urano F, Wang X, Bertolotti A, Zhang Y, Chung P, Harding H, et al. Coupling of stress in the ER to activation of JNK protein kinases by transmembrane protein kinase IRE1. Science (New York, N Y ) 2000;287(5453):664-670. 37. Dhanasekaran D, Reddy E. JNK signaling in apoptosis. Oncogene 2008;27(48):6245-6296. 38. Liu J, Lin A. Role of JNK activation in apoptosis: a double-edged sword. Cell Res 2005;15(1):36-78. 39. Luo D, He Y, Zhang H, Yu L, Chen H, Xu Z, et al. AIP1 is critical in transducing IRE1mediated endoplasmic reticulum stress response. The Journal of biological chemistry 2008;283(18):11905-11917. 40. Tournier C, Hess P, Yang D, Xu J, Turner T, Nimnual A, et al. Requirement of JNK for stressinduced activation of the cytochrome c-mediated death pathway. Science (New York, N Y ) 2000;288(5467):870-874. 41. Bossy-Wetzel E, Bakiri L, Yaniv M. Induction of apoptosis by the transcription factor c-Jun. EMBO J 1997;16(7):1695-2404. 42. Hollien J, Lin J, Li H, Stevens N, Walter P, Weissman J. Regulated Ire1-dependent decay of messenger RNAs in mammalian cells. J Cell Biol 2009;186(3):323-354. 43. Hollien J, Weissman J. Decay of endoplasmic reticulum-localized mRNAs during the unfolded protein response. Science (New York, N Y ) 2006;313(5783):104-111. 44. Kato H, Nakajima S, Saito Y, Takahashi S, Katoh R, Kitamura M. mTORC1 serves ER stress-triggered apoptosis via selective activation of the IRE1-JNK pathway. Cell Death Differ 2012;19(2):310-330. 45. Song B, Scheuner D, Ron D, Pennathur S, Kaufman RJ. Chop deletion reduces oxidative stress, improves Î² cell function, and promotes cell survival in multiple mouse models of diabetes. J Clin Invest 2008 10/01;118(10):3378-3389. 46. Yamaguchi H, Wang H. CHOP is involved in endoplasmic reticulum stress-induced apoptosis by enhancing DR5 expression in human carcinoma cells. The Journal of biological chemistry 2004;279(44):45495-45997. 47. Pino SC, O’Sullivan-Murphy B, Lidstone EA, Yang C, Lipson KL, Jurczyk A, et al. CHOP Mediates Endoplasmic Reticulum Stress-Induced Apoptosis in Gimap5-Deficient T Cells. PLoS ONE 2009 05/08;4(5):5468. 48. McCullough K, Martindale J, Klotz L, Aw T, Holbrook N. Gadd153 sensitizes cells to endoplasmic reticulum stress by down-regulating Bcl2 and perturbing the cellular redox state. Mol Cell Biol 2001;21(4):1249-1308. 49. Marciniak SJ, Yun CY, Oyadomari S, Novoa I, Zhang Y, Jungreis R, et al. CHOP induces death by promoting protein synthesis and oxidation in the stressed endoplasmic reticulum. Genes &

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

The mammalian unfolded protein response: a potential source of novel therapeutic targets in the hypoxic tumour fraction

Review Articles

Development 2004 December 15;18(24):3066-3077. 50. Morishima N, Nakanishi K, Nakano A. Activating transcription factor-6 (ATF6) mediates apoptosis with reduction of myeloid cell leukemia sequence 1 (Mcl-1) protein via induction of WW domain binding protein 1. J Biol Chem 2011 Oct 7;286(40):35227-35235. 51. Lin JH, Li H, Yasumura D, Cohen HR, Zhang C, Panning B, et al. IRE1 signaling affects cell fate during the unfolded protein response. Science 2007 Nov 9;318(5852):944-949. 52. Lin JH, Li H, Zhang Y, Ron D, Walter P. Divergent Effects of PERK and IRE1 Signaling on Cell Viability. PLoS ONE 2009 01/12;4(1):4170. 53. Tabas I, Ron D. Integrating the mechanisms of apoptosis induced by endoplasmic reticulum stress. Nat Cell Biol 2011. 54. Treglia A, Turco S, Ulianich L, Ausiello P, Lofrumento D, Nicolardi G, et al. Cell fate following ER stress: just a matter of “quo ante” recovery or death? Histol Histopathol 2012;27(1):1-13. 55. Fribley A, Zhang... K. Regulation of apoptosis by the unfolded protein response. Methods Mol Biol 2009. 56. Michallet A, Mondiere P, Taillardet M, Leverrier Y, Genestier L, Defrance T. Compromising the Unfolded Protein Response Induces Autophagy-Mediated Cell Death in Multiple Myeloma Cells. PLoS ONE 2011 10/18;6(10):25820. 57. Fels D, Ye J, Segan A, Kridel S, Spiotto M, Olson M, et al. Preferential cytotoxicity of bortezomib toward hypoxic tumor cells via overactivation of endoplasmic reticulum stress pathways. Cancer Res 2008;68(22):9323-9353. 58. Nawrocki S, Carew J, Pino M, Highshaw R, Dunner K, Huang P, et al. Bortezomib sensitizes pancreatic cancer cells to endoplasmic reticulum stress-mediated apoptosis. Cancer Res 2005;65(24):11658-11724. 59. Yi P, Higa A, Taouji S, Bexiga MG, Marza E, Arma D, et al. Sorafenib-mediated targeting of the AAA+ ATPase p97/VCP leads to disruption of the secretory pathway, endoplasmic reticulum stress and hepatocellular cancer cell death. Mol Cancer Ther 2012 Oct 5. 60. Yang W, Tiffany-Castiglioni E, Koh HC, Son IH. Paraquat activates the IRE1/ASK1/JNK cascade associated with apoptosis in human neuroblastoma SH-SY5Y cells. Toxicol Lett 2009 Dec 15;191(2-3):203-210. 61. Niso-Santano M, Bravo-San Pedro JM, Gomez-Sanchez R, Climent V, Soler G, Fuentes JM, et al. ASK1 overexpression accelerates paraquat-induced autophagy via endoplasmic reticulum stress. Toxicol Sci 2011 Jan;119(1):156-168. 62. Zheng QY, Li PP, Jin FS, Yao C, Zhang GH, Zang T, et al. Ursolic acid induces ER stress response to activate ASK1-JNK signaling and induce apoptosis in human bladder cancer T24 cells. Cell Signal 2012 Sep 18. 63. Maddalo D, Neeb A, Jehle K, Schmitz K, Muhle-Goll C, Shatkina L, et al. A Peptidic Unconjugated GRP78/BiP Ligand Modulates the Unfolded Protein Response and Induces Prostate Cancer Cell Death. PLoS One 2012;7(10):e45690. 64. Papandreou I, Denko N, Olson M, Van Melckebeke H, Lust S, Tam A, et al. Identification of an Ire1alpha endonuclease specific inhibitor with cytotoxic activity against human multiple myeloma. Blood 2011;117(4):1311-1315. 65. Mimura N, Fulciniti M, Gorgun G, Tai Y, Cirstea D, Santo L, et al. Blockade of XBP1 splicing by inhibition of IRE1Î± is a promising therapeutic option in multiple myeloma. Blood 2012;119(24):5772-5853. 66. Tashiro E, Hironiwa N, Kitagawa M, Futamura Y, Suzuki S, Nishio M, et al. Trierixin, a novel Inhibitor of ER stress-induced XBP1 activation from Streptomyces sp. 1. Taxonomy, fermentation, isolation and biological activities. J Antibiot 2007;60(9):547-600. 67. Volkmann K, Lucas J, Vuga D, Wang X, Brumm D, Stiles C, et al. Potent and selective inhibitors of the inositol-requiring enzyme 1 endoribonuclease. The Journal of biological chemistry 2011;286(14):12743-12798. 68. Fribley A, Cruz P, Miller J, Callaghan M, Cai P, Narula N, et al. Complementary cell-based high-throughput screens identify novel modulators of the unfolded protein response. Journal of biomolecular screening 2011;16(8):825-860. 69. Wang L, Perera BG, Hari SB, Bhhatarai B, Backes BJ, Seeliger MA, et al. Divergent allosteric control of the IRE1alpha endoribonuclease using kinase inhibitors. Nat Chem Biol 2012 Oct 21. 70. Wang H, Blais J, Ron D, Cardozo T. Structural determinants of PERK inhibitor potency and selectivity. Chem Biol Drug Des 2010 Dec;76(6):480-495. 71. Wiseman R, Zhang Y, Lee K, Harding H, Haynes C, Price J, et al. Flavonol activation defines an unanticipated ligand-binding site in the kinase-RNase domain of IRE1. Mol Cell 2010;38(2):291-595.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

REVIEWS

Adoptive cell therapy: CD4+ and CD8+ T cells and the cells that educate them Nothando Z. D. Swan University of Toronto

Abstract

In adoptive cell therapy (ACT), patients’ own tumor-specific T cells are selected and expanded in vitro so that they can be readministered with increased antitumor activity. Several studies have reported objective clinical responses in cancer patients who have received such autologous cell transfer. This review discusses the key differences in antigen recognition between CD4+ and CD8+ T cells. It then outlines the advances and clinical implementations of these cells in ACT as well as the utility of artificial antigenpresenting cells in tumor-specific T cell education and expansion.

Introduction

Several immune-based therapies exist to treat cancer. Researchers have developed both innate and adaptive immune approaches to enhance patients’ defenses against malignancies [1]. Adaptive immunity involves B and T cells that express receptors for specific immunogenic substances called antigens. While B cells produce antibodies that arrest extracellular invaders, T cells themselves can eliminate threats within the body’s tissues, and they can also help other cells do so. Adoptive cell therapy (ACT) uses the specificity of T cell-mediated immunity to target cancer cells that display immunogenic peptides [2]. Researched and developed since the 1980s, this approach expands autologous T cells that recognize tumor-associated antigen (TAA) in vitro. This in vitro expansion allows scientists to culture large numbers of tumor-specific T cells outside of the tumor microenvironment, which is replete with immunosuppressive cells and molecules [3]. Then, these autologous tumorspecific cells are intravenously infused into patients suffering from the targeted cancer type. Clinical success has been reported in studies using ACT to treat melanoma, chronic lymphocytic leukemia, non-Hodgkin lymphoma, and other types of cancer. Interestingly, both CD4+ and CD8+ T cells present unique and effective tumor-fighting responses, and these responses have been studied both separately and together. Given the importance of the appropriate in vitro education and expansion of these tumor-fighting cells, artificial antigen-presenting cells (aAPCs) as T cell primers are also being studied.

CD4+ T cells versus CD8+ T cells in clinical translation

While CD4+ and CD8+ T cells recognize antigen processed by different pathways, their overarching mechanism of antigen recognition is similar. In contrast to B cells that can recognize the three dimensional conformation of small chemicals and all classes of macromolecules, classical CD4+ and CD8+ T cells recognize

antigens derived from proteins [4]. What is more, these protein antigens must be processed into linear peptide sequences and loaded onto a class I or class II major histocompatibility complex molecule (MHC I or MHC II, respectively). Thus, in order for a T cell to recognize an antigen and then mediate a response against it, this antigen must be peptidic and presented on an MHC molecule in order for a T cell to recognize it and mediate a response against it. The preponderance of clinical efficacy using ACT lies within the realm of CD8+ T cells. Most examples of ACT using CD4+ T cells encompass work done with transgenic murine models, which do not perfectly mimic human processes. Furthermore, in these studies, specific CD4+ T subtypes have been selected because certain subtypes, namely regulatory T cells, can in fact impair ACT results [5]. One of the main factors that make CD4+ T cells a less obvious choice for ACT is that they recognize antigens associated with MHC class II molecules (Table 1). MHC II is expressed by professional APCs (dendritic cells, B cells, and macrophages) but very rarely by tumor cells [6]. This renders CD4+ T cell tumor cytotoxicity largely dependent on activation by APCs that must take up tumor protein, process it, and then present it on their cell surface on an MHC II molecule. In contrast, most nucleated cells, including cancerous ones, express MHC class I molecules that can directly interact with CD8+ T cells. In this way, CD8+ T cells can directly recognize and respond to these antigens displayed by tumor cells on MHC I. Importantly, CD4+ T cells are also capable of direct tumor recognition when tumor cells naturally present antigenic peptides via MHC II or when tumor MHC II expression is induced by IFN-γ [6, 7]. For example, in the recently reported ACT clinical trial where tumor-specific CD4+ T cell clones were transferred, a patient suffering from metastatic melanoma exhibited a complete response. Two months after receiving autologous NY-ESO-1-specific CD4+ T cells, positron emission tomography (PET) and computed tomography (CT) scans showed complete resolution of pulmo-

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Adoptive cell therapy: CD4+ and CD8+ T cells and the cells that educate them

Table 1. CD4+ and CD8+ T cells recognize antigen processed by different pathways. CD8+ T cells are a more obvious choice for ACT because they recognize antigen in the context of MHC I, which is more commonly expressed by tumors than is MHC II. Can recognize and respond to

Examples of tumor peptide epitopes

CD4+ T cell 10 – 30 amino acid tumor peptide sequences presented on MHC II [4] NY-ESO-1 peptide presented on MHC II: SLLMWITQCFLPVF [8]

CD8+ T cell 8 – 11 amino acid tumor peptide sequences presented on MHC I [4] MART1 peptide sequence presented on MHC I: ELAGIGILTV [38]

nary and nodal metastases with no regression two years later [8]. Importantly, the authors acknowledge the likelihood of the involvement of other cells in these tumor regressions, since not all of the patient’s melanoma cells expressed the NY-ESO-1 epitope recognized by the CD4+ T cells.

CD8+ T cells in ACT

Significant clinical progress has been made using CD8+ T cells in ACT, particularly in the treatment of melanoma. Melanoma, a particularly fatal skin cancer, is highly immunogenic [3]. In the late 1980s, scientists showed conclusively that tumors and peripheral blood samples of some patients suffering from this disease contained lymphocytes that could recognize melanoma tumor cells in vitro [9]. Since then, research has focused on techniques that harbor the antitumor effects of these tumor-reactive cells, which include CD8+ T cells among others [6]. One such technique involves the expansion and infusion of a patients’ own tumor-infiltrating lymphocytes (TILs). A tumor biopsy is cut into small samples and incubated with high levels of interleukin-2 (IL-2), resulting in the expansion of anti-tumor T cells. In the past, once large enough numbers of T cells were expanded, they were indiscriminately infused into the patient [10]. However in more recent trials, only those cells that displayed antitumor activity were infused [11-13]. Autologous TIL treatments for the majority of patients who experienced objective clinical responses predominantly contained CD8+ T cells. Much of the success of these later trials has been attributed to the ability to manipulate patients before autologous cell transfer. Pre-infusion lymphodepletion through chemotherapy with or without total-body irradiation creates a lymphopenic environment [11, 12]. Such environments can increase transferred cells’ chances of survival by eliminating immunosuppressive regulatory T cells [5] and by eliminating competition for cytokines such as interleukin-7 (IL-7) and interleukin-15 (IL-15) [14]. Clinical trials combining autologous TIL infusions with lymphodepletion have yielded objective clinical responses ranging between 49% and 72% of patients [11, 12]. While studies with TILs suggest that CD8+ T cells can mediate antitumor effects, they do not rule out the contributory roles of other lymphocytes present in the infusions. In contrast, work with CD8+ T cell clones has definitively shown that CD8+ T cells specific for TAAs can halt cancer progression [15]. In one study, disease stabilization was reported in 50% of patients receiving autologous CD8+ T cell clones primed by their own dendritic cells [16]. Though low dose administration of IL-2 increased transferred T cell persistence, loss of transferred cells and downregulation of antigen presentation by tumor cells presented drawbacks.

Review Articles

Recently, rather than screening patients’ T cells to select those demonstrating antitumor functionality, researchers have induced this cytotoxicity by genetically modifying CD8+ T cells to express tumor antigen-specific T-cell receptors (TCRs) [17, 18]. In this technique, genes for tumor-specific TCRs are transduced into T cells isolated from tissue or peripheral blood samples, enabling them to recognize MART1, gp100, or other TAAs. In one clinical trial, tumor shrinkage following the transfer of transgenic TCR-T cells was seen in 13% of patients [18]. Similarly, chimeric antigen receptors (CARs) can be introduced into T cells. CARs are unique protein fusions made from the extracellular variable region of a tumor-specific antibody attached to the constant region of a TCR [2]. This combination of extracellular and intracellular domains provides an alternate route for TAA binding and recognition that is independent of both TCR variable regions and major histocompatibility complex molecules. In a study using CAR transgenic CD8+ T cells to treat patients with neuroblastoma, the transgenic cells were short-lived and only a partial response was observed in one patient [19]. However more recently, scientists using a CAR-T cell construct observed two complete and one partial response in three patients suffering from chronic lymphocytic leukemia [20]. Importantly, this construct included two intracellular costimulatory domains that may have allowed the transferred cells to persist for at least 6 months in all patients. Finally, questions remain about how best to prime CD8+ T cells in vitro for persistence and cytotoxicity in vivo. For instance, although IL-2 has long been used to support T cell cultures, IL-15 and IL-21 have been purported to produce more potent effector CD8+ T cells [21]. Researchers have also suggested that central or early memory CD8+ T cells that are better able to travel to secondary lymphoid organs for reactivation display stronger in vivo antitumor effects than do more-differentiated effector CD8+ T cells [22, 23].

CD4+ T cells in ACT

CD4+ T cells mediate effective responses to threat by supporting and directing both innate and adaptive immune cells. In fact, CD8+ T cell-mediated tumor cytotoxicity is often augmented by the presence of CD4+ T cells. For example in mice, subcutaneous, pulmonary, and intracranial tumors were all most effectively treated when both CD4+ and CD8+ antigen-specific T cells were adoptively transferred [24]. In this study, intratumoral CD8+ T cell proliferation was dependent on CD4+ T cell presence, indicating that CD4+ T cells helped to maintain cytotoxic CD8+ T cell numbers in vivo. In another murine model, investigators tracked the in vivo fate of adoptively transferred CD8+ T cells and found that they trafficked to tumors and persisted in higher numbers when simultaneously infused with CD4+ T cells [25]. Besides providing help, CD4+ T cells may also be necessary for effective ACT in murine models. In one study, a 1:1 ratio of erbB2+ antigen-specific CD4+ and CD8+ T cells resulted in complete survival rates of mice with lung metastases, whereas no antitumor effect was measured when CD8+ T cells were infused alone [26]. Interestingly, mice that received both cell types survived not one, but two tumor cell inoculations, demonstrating the induction of prolonged antitumor T cell memory and function. Although tumor-specific CD8+ memory T cells do have potent cytotoxic effects, CD4+ T cells may be necessary to reactivate these effector functions. Several scientific groups have shown this crucial role for

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Review Articles

Adoptive cell therapy: CD4+ and CD8+ T cells and the cells that educate them

CD4+ T cells in secondary TAA encounters [27-29]. Importantly, CD4+ T cells show antitumor effects that are independent of CD8+ T cells. B16 mouse melanomas were eradicated by CD4+ T cells that were polarized towards the Th17 subset [30] and by CD4+ T cells that differentiated into the Th1 subset when transferred into mice in another experiment [7]. Furthermore, another study showed that CD4+ T cells eradicated 6 different tumors that CD8+ T cells could not [31]. Still other studies have shown that where the direct cytotoxic effect of CD8+ T cells fails, CD4+ T cells can recruit natural killer cells and eosinophils among other immune cells to indirectly eliminate malignancies [31, 32]. For example, through an IFN-γdependent mechanism, tumor-specific CD4+ T cells transferred to tumor-bearing mice activated macrophages found within those tumors. The activated macrophages then released tumoricidal nitric oxide and reactive oxygen species [33].

Artificial antigen-presenting cells in ACT

The appropriate in vitro education of antitumor CD4+ and CD8+ T cells is conceivably the most crucial step in ACT. In order for the therapy to yield clinical responses, T cells displaying antitumor effects in vitro must also display these effects in vivo. Artificial antigen-presenting cells (aAPCs) may offer quicker, more efficient ways of priming T cells for in vivo efficacy than using autologous professional antigen-presenting cells, which require specific manufacturing for each patient. Cell-based aAPCs are derived from human, insect, or mouse cell lines. They are engineered to engage TCRs while providing the costimulatory signals (such as CD80 and CD86) that are required to make responsive T cells [34]. In particular, aAPCs derived from the K562 human erythroleukemic cell line have seen recent success [35-38]. Butler and colleagues have designed K562-derived aAPC systems that have successfully expanded both CD4+ and CD8+ T cells [35, 36]. Furthermore, they have tested a cell-based aAPC system and shown that it educated CD8+ T cells that could mediate persistent clinical responses in patients suffering from metastatic melanoma [37]. Other researchers have used a K562-derived aAPC system to successfully expand TILs from ovarian cancer tumors [38]. Bead-based aAPCs provide a viable, acellular method for the in vitro education of T cells. The plastic microspheres are coated with MHC-peptide dimers or tetramers that readily present a selected immunogenic peptide to CD4+ or CD8+ T cells [34]. Though capable of priming antigen-specific lymphocytes [39, 40], beadbased aAPCs do not perfectly mimic the cell-cell synapse found between T cells and professional APCs or cell-based aAPCs [34]. However, researchers have found that an MHC-peptide tetramer bound to beads by a flexible linker provided the best CD4+ T cell stimulation, perhaps partially due to increased fluidity at the aAPC-T cell interface [41].

Conclusions

Research in ACT has approached cancer treatment from several angles. The role of CD4+ and CD8+ T cells has been highlighted in varying degrees from study to study, revealing the unique attributes and advantages of each cell type. While dichotomies between CD4+ and CD8+ T cells are easily drawn, researchers acknowledge the contributory effects of each. At present, although professional

APCs play crucial roles in educating antitumor T cells, aAPCs may offer faster, ready-to-go methods of immune cell activation, which may further advance our ability to provide patients with personalized and effective cancer treatment.

Acknowledgements

Many thanks to Dr. Marcus Butler for guiding my research and writing, for reading and re-reading this review, and for correcting my scientific misreports. Thanks is also due to Dr. Bill Ju, who has encouraged my scientific writing. Glossary Adoptive cell therapy. The in vitro selection and expansion of patients’ own tumor-specific/genetically-modified T cells followed by their re-infusion, aimed to mediate in vivo antitumor effects. Antigen. A substance, often a peptide sequence, which elicits an immune response when bound to an antibody or T cell receptor. Artificial antigen-presenting cells (aAPCs). Cellular or acellular structures engineered to substitute professional APCs in the antigen-specific activation of T cells. Autologous. Derived from the patient (as opposed to an outside donor). Allogeneic. Derived from an outside donor. Antigen processing pathway. The process through which proteins are lysed into peptide sequences and loaded onto MHC-I/II molecules so that T cells can recognize them. Costimulatory molecules. Molecules such as CD80 and CD86 that are found on APCs and signal to T cells, providing them with input necessary for T cell activation. Cytokine. Secreted proteins such as interleukins and IFN-γ that mediate various immune system functions, such as inflammation and immune cell migration. Expansion. The process through which selected/specific T cells increase in number. Immunogenic. Able to elicit an immune response. Lymphopenic. An environment with abnormally low numbers of lymphocytes. Major histocompatibility complex molecules (MHC-I/II). A diverse group of cell surface proteins that have the unique ability to present antigens in a format that is “readable” to T cells. Also known as human leukocyte antigens (HLA-I/II) when found in humans (as opposed to in mice). Professional antigen-presenting cells. Dendritic cells, macrophages and B lymphocytes: the immune system cells naturally responsible for processing antigens and presenting them to T cells. T cell receptors (TCRs). Proteins found on the surface of T cells that bind to MHC-antigen complexes, initiating the process of T cell activation.

References

1. Topalian SL, Weiner GJ, Pardoll DM. Cancer immunotherapy comes of age. J Clin Oncol 2011;29(36):4828-4836. 2. Brenner MK, Heslop HE. Adoptive T cell therapy of cancer. Curr Opin Immunol 2010;22(2):251-257. 3. Rosenberg SA, Dudley ME. Adoptive cell therapy for the treatment of patients with metastatic melanoma. Curr Opin Immunol 2009;21(2):233-240. 4. Abbas AK, Lichtman AH. Basic Immunology. 3rd ed. Philadelphia: Saunders; 2011. 5. Antony PA, Piccirillo CA, Akpinarli A, Finkelstein SE, Speiss PJ, Surman DR, et al. CD8+ T cell immunity against a tumor/self-antigen is augmented by CD4+ T helper cells and hindered by naturally occurring T regulatory cells. J Immunol 2005;174(5):2591-2601. 6. Muranski P, Restifo NP. Adoptive immunotherapy of cancer using CD4+ T cells. Curr Opin Immunol 2009;21(2):200-208. 7. Xie Y, Akpinarli A, Maris C, Hipkiss EL, Lane M, Kwon EM, et al. Naive tumor-specific CD4+ T cells differentiated in vivo eradicate established melanoma. J Exp Med 2010;207(3),651-667. 8. Hunder NN, Wallen H, Cao J, Hendricks DW, Reilly JZ, Rodmyre R, et al. Treatment of metastatic melanoma with autologous CD4+ T cells against NY-ESO-1. N Engl J Med 2008;358(25):2698-2703. 9. Muul LM, Spiess PJ, Director EP, Rosenberg SA. Identification of specific cytolytic immune responses against autologous tumor in humans bearing malignant melanoma. J Immunol 1987;138(3):989-995. 10. Rosenberg SA, Yannelli JR, Yang JC, Topalian SL, Schwartzentruber DJ, Webser JS, et al. Treatment of patients with metastatic melanoma using autologous tumor-infiltrating lymphocytes and interleukin-2. J Natl Cancer Inst 1994;86(15):1159-1166. 11. Dudley ME, Wunderlich JR, Yang JC, Sherry RM, Topalian SL, Restifo NP, et al. Adoptive cell transfer therapy following non-myeloablative but lymphodepleting chemotherapy for the treatment of patients with refractory metastatic melanoma. J Clin Oncol 2005;23(10):2346-2357.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Adoptive cell therapy: CD4+ and CD8+ T cells and the cells that educate them

12. Dudley ME, Yang JC, Sherry R, Hughes MS, Royal R, Kammula U, et al. Adoptive cell therapy for patients with metastatic melanoma: evaluation of intensive myeloablative chemoradiation preparative regimens. J Clin Oncol 2008;26(32):5233-5239. 13. Dudley ME, Wunderlich JR, Robbins PF, Yang JC, Hwu P, Schwartzentruber DJ, et al. Cancer regression and autoimmunity in patients after clonal repopulation with antitumor lymphocytes. Science 2002;298(5594):850-854. 14. Gattinoni L, Finkelstein SE, Klebanoff CA, Antony PA, Palmer DC, Spiess PJ, et al. Removal of homeostatic cytokine sinks by lymphodepletion enhances the efficacy of adoptively transferred tumor-specific CD8+ T cells. J Exp Med 2005;202(7):907-912. 15. Yee C. Adoptive therapy using antigen-specific T cell clones. Cancer J 2010;16(4):367-373. 16. Yee C, Thompson JA, Byrd D, Riddell SR, Roche P, Celis E, et al. Adoptive T cell therapy using antigen-specific CD8+ T cell clones for the treatment of patients with metastatic melanoma: In vivo persistence, migration, and antitumor effect of transferred T cells. Proc Natl Acad Sci 2002;99(25):16168-16173. 17. Hinrichs CS, Borman ZA, Cassard L, Gattinoni L, Spolski R, Yu Z, et al. Adoptively transferred effector cells derived from naïve rather than central memory CD8+ T cells mediate superior antitumor immunity. Proc Natl Acad Sci 2009;106(41):17469-17474. 18. Morgan RA, Dudley ME, Wunderlich JR, Hughes MS, Yang JC, Sherry RM, et al. Cancer Regression in patients after transfer of genetically engineered lymphocytes. Science 2006;314(5796):126-129. 19. Park JR, DiGiusto DL, Slovak M, Wright C, Naranjo A, Wagner J, et al. Adoptive transfer of chimeric antigen receptor re-directed cytolytics T lymphocyte clones in patients with neuroblastoma. Molecular Therapy 2007;15(4):825-833. 20. Kalos M, Levine BL, Porter DL, Katz S, Gripp, SA, Bagg A, et al. T cells with chimeric antigen receptors have potent antitumor effects and can establish memory in patients with advanced leukemia. Sci Transl Med 2011;3:95ra73. 21. Li Y, Liu S, Hernandez J, Vence L, Hwu P, Radvanyi L. MART-1-specific melanoma tumorinfiltrating lymphocytes maintaining CD28 expression have improved survival and expansion capability following antigenic restimulation in vitro. J Immunol 2010;184(1):452-465. 22. Gattinoni L, Klebanoff CA, Palmer DC, Wrzesinski C, Kerstann K, Yu Z, et. al. Acquisition of full effector function in vitro paradoxically impairs the in vivo antitumor efficacy of adoptively transferred CD8+ T cells. J Clin Invest 2005;115(6):1616-1626. 23. Klebanoff CA, Gattinoni L, Torabi-Parizi P, Kerstann K, Cardones AR, Finkelstein SE, et al. Central memory self/tumor-reactive CD8+ T cells confer superior antitumor immunity compared with effector memory T cells. Proc Natl Acad Sci 2005;102(27):9571-9576. 24. Wang L, Shu S, Disis ML, Plautz GE. Adoptive transfer of tumor-primed, in vitro-activated, CD4+ T effector cells (TEs) combined with CD8+ TEs provides intratumoral TE proliferation and synergistic antitumor response. Blood 2007;109(11):4865-4876. 25. Marzo A, Kinnear BF, Lake RA, Frelinger JJ, Collins EJ, Robinson BWS, et al. Tumor-specific CD4+ T cells have a major “post-licensing” role in CTL mediated anti-tumor immunity. J Immunol 2000;165(11):6047-6055. 26. Moeller M, Haynes NM, Kershaw MH, Jackson JT, Teng MWL, Street SE, et al. Adoptive transfer of gene-engineered CD4+ helper T cells induces potent primary and secondary tumor rejection. Blood 2005;106(9):2995-3003. 27. Gao FG, Khammanivong V, Liu WJ, Leggatt GR, Frazer IH, Fernando GFP. Antigen-specific CD4+ T-cell help is required to activate a memory CD8+ T cell to a fully functional tumor killer cell. Cancer Res 2002;62(22):6438-6441. 28. Shedlock DJ, Shen H. Requirement for CD4 T cell help in generating functional CD8 T cell memory. Science 2003;300(5617):337-339. 29. Sun JC, Bevan MJ. Defective CD8 T cell memory following acute infection without CD4 T cell help. Science 2003;300(5617):339-342. 30. Muranski P, Boni A, Antony PA, Cassard L, Irvine KR, Kaiser A, et al. Tumor-specific Th17polarized cells eradicated large established melanoma. Blood 2008;112(2):362-373. 31. Perez-Diez A, Joncker NT, Choi K, Chan WFN, Anderson CA, Lantz O, et al. CD4 cells can be more efficient at tumor rejection than CD8 cells. Blood 2007;109(12):5346-5354. 32. Mattes J, Hulett M, Xie W, Hogan S, Rothenberg ME, Foster P, et al. Immunotherapy of cytotoxic T cell-resistance tumors by T helper 2 cells: an eotaxin and STAT6-dependent process. J Exp Med 2003; 197(3):387-393. 33. Corthay A, Skovseth DK, Lundin KU, Røsjø E, Omholt H, Hofgaard PO, et al. Primary antitumor immune response mediated by CD4+ T cells. Immunity 2005;22(3):371-383. 34. Turtle CJ, Riddell SR. Artificial Antigen-Presenting cells for use in adoptive immunotherapy. Cancer J 2010;16(4):374-381. 35. Butler MO, Ansén S, Tanaka M, Imataki O, Berezovskaya A, Mooney MM, et al. A panel of human cell-based artificial APC enables the expansion of long-lived antigen-specific CD4+ T cells restricted by prevalent HLA-DR alleles. Int Immunol 2010;22(11):863-873. 36. Butler MO, Lee J, Ansén S, Neuberg D, Hodi FS, Murray AP, et al. Long-lived antitumor CD8+ lymphocytes for adoptive immunity generated using an artificial-antigen presenting cell. Clin Cancer Res 2007;13(6):1857-1867. 37. Butler MO, Friedlander P, Milstein MI, Mooney MM, Metzler G, Murray AP, et al. Establishment of antitumor memory in humans using in vitro-educated CD8+ T cells. Sci Transl Med 2011;3(8):80ra34. 38. Ye Q, Loisiou M, Levine BL, Suhoski MM, Riley JL, June CH, et al. Engineered artificial antigen presenting cells facilitate direct and efficient expansion of tumor infiltrating lymphocytes. J Transl Med 2011;9(1):131. 39. Shen C, Zhang J, Xia L, Meng F, Xie W. Induction of tumor antigen-specific cytotoxic T cell responses in naïve mice by latex microspheres-based artificial antigen-presenting cell constructs. Cell Immunol 2007;247(1):28-35.

Review Articles

40. Caserta S, Alessi P, Guarnerio J, Basso V, Mondino A. Synthetic CD4+ T cell targeted antigenpresenting cells elicit protective antitumor responses. Cancer Res 2008;68:3010-3018. 41. Maus MV, Riley JL, Kwok WW, Nepom GT, June CH. HLA tetramer-based artificial antigenpresenting cells for stimulation of CD4+ T cells. Clin Immunol 2003;106:16-22.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

INTERVIEWS

Dr. Thomas Jessell Many of our readers will recognize Thomas Jessell as co-editor of ‘Principles of Neural Science’, a textbook that has remained a classic tome of neuroscience knowledge for over thirty years. Professor Jessell completed his PhD in neuropharmacology at Cambridge University in 1977. He later completed a postdoctoral fellowship at Harvard Medical School where he became Assistant Professor in the Department of Neurology starting in 1981. Today, Dr. Jessell is faculty at Columbia University where he conducts research at Howard Hughes Medical Institute and teaches both biochemistry and molecular biophysics. As a 2012 Gairdner Laureate, Dr. Jessell presented an overview of his groundbreaking work delineating the complex genetic and molecular pathways responsible for the development of spinal cord sensory and motor circuits. This work has led to many other awards and honours in recognition of Dr. Jessell’s contributions to the field of neuroscience, including the 2008 Kavli Prize in Neuroscience and his position as a Fellow of the Royal Society. Interviewed by Amirah I. Momen

“Science, in the end, is all about reductionism. No matter how complicated the process, don’t get overwhelmed by the labyrinthine complexity of it.” AM: You started off in pharmacology for your undergrad. How did you get to where you are and what was your journey like? TJ: Wobbly, in a word. You’ve got to remember that back in

the old days when molecular biology hadn’t had an impact on neural science in the way that everybody takes for granted today, the most effective way of interrogating the nervous system was actually through pharmacology using the specificity of drug action. So, when I was an ‘aimless’ undergraduate I wasn’t sure what I wanted to do, but somehow the brain seemed interesting and I had an undergraduate lecturer who started to tell me about the ways that drugs acted on the brain. This was in London in the sort of early 1970s and he was a fantastic lecturer; he engaged [students] and described the specificity of behavioral effects when you administer drugs to the brain. Outside of the college in London, on The King’s Road, was this sort of hotbed of London life at that time. You only had to walk down The King’s Road to see the effects of drugs on the brain on a daily basis and so the combination of being interested in pharmacological intervention in the nervous system and the real life consequences of it got me interested. When I went on to do a PhD, I chose a laboratory in Cambridge, or, was lucky enough to be accepted into a lab that was run by a guy called Leslie Iverson who was a fantastic neuropharmacologist, or neuroschemist, that was trying to use biochemistry and pharmacology to understand the way the brain works [and] to understand issues like addiction and behavioral disorders. That introduced me to an international world of science that was just fantastic!

AM: You mentioned being an ‘aimless undergrad’-TJ: Aren’t all undergraduates aimless, except in Toronto!? AM: (Laughs) This is my question though, because now we have more specialized undergraduate programs like ‘Neuroscience’ or ‘Laboratory Medicine and Pathobiology’. Do you think that it’s more beneficial to approach research with an interdisciplinary background or a more basic background? TJ: I think one of the great things about neural science–which ranges from atomic resolution structural analyses to human cognition and psychology–is that there are a lot of ways of approaching a common problem and you want people from diverse backgrounds: from engineering, from chemistry, from physics, from psychology, from biology and from quantitative sciences like applied math and computer science; you want all of these people thinking about problems and taking their individual perspective. What you have to remember is that we don’t understand anything yet, despite all this brave talk. The fundamental questions that drive one’s interest in the field are unsolved. And so it’s a fantastic, fantastic opportunity and even though it’s complicated, there is great scope for individuals coming in and thinking differently about the problem, learning from what’s gone before (but not necessarily being overly swayed by it) and then deciding to do something and making a big, big impact.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Dr. Thomas Jessell

AM: You came into the field at a time when all of the genetic and molecular pathways you later worked to delineate were largely uncharted territory. How do you begin to approach such a complicated puzzle?

“...The most important thing is: don’t go with trends or with conventional wisdom.” TJ: One step at a time. Don’t try and do it all at once. Science, in the end, is all about reductionism. No matter how complicated the process, don’t get overwhelmed by the labyrinthine complexity of it. Think about what you can do next that makes a difference on the timescale of a year, or two, or three years, at the onset. What question, little question, can I pose and address, and then hopefully solve? Draw some conclusion and inference from that and then go on to the next. Sometimes you’ll be surprised and sometimes you start something that seems relatively routine...but the great experiments are the ones that give you a result that is quite different from what you expected and then you pay attention, you keep your eyes open. You get a new result, it changes the way you think about things, and suddenly there are new questions and new paths to follow. So don’t get overwhelmed by the big picture. Think reductionist even though the problems are super complicated and in that way you can make progress. AM: There are, ultimately, therapeutic implications for your work regarding the understanding of embryonic stem cells and cell differentiation in the treatment of neurodegenerative disease and spinal cord injury. With this in mind, how would you like to see your legacy unfold and how long do you think it will take before relevant therapies are discovered?

Research Interviews Articles

TJ: Yes, I mean, you know, there’s a fine line between being completely objective (and I think humans can’t be completely objective) and having a passion about an idea or a way of approaching things. I think the most important thing is that it doesn’t matter how tightly held your initial starting hypothesis is, if there is decisive evidence against that idea you’ve got to listen to that, realize that you were wrong, throw that idea out and come up with a better idea. The people who I think do a disservice to science are those who hold onto the idea in the face of overwhelming contrarian evidence. You’ve got to be prepared to change, and change, and change and you always have to be sensitive to the changing world of information that you’re living in. You have to pay attention to data that says, ‘Yeah, no matter how clever this idea was, it’s not actually tenable or sustainable in the face of the growing evidence.’ You then come up with a better idea and you constantly recalibrate. AM: Lastly, do you have any advice for our undergraduates about what makes a good scientist and how they should approach the next 50 years of their life? TJ: One, go into science; two, go into neuroscience; three, think very hard about what you want to do, and the most important thing is: don’t go with trends or with conventional wisdom. Think about the problem, decide how you can make a difference and then pursue that. And don’t worry if, initially, the relevance of what you’re doing (to large problems, or clinical problems, or basic problems in neural science) isn’t immediately evident. Follow your intuition but be very, very ready to admit defeat when things don’t look good....and don’t lose heart! It’s a great profession, I think, because you are asking questions that no one else in the history of the universe has asked. So, follow your intuition and be prepared to change and be enthusiastic! AM: And read everything?

TJ: You can’t read everything so decide what to read and talk TJ: Let’s address that again a little bit because what motivated to people because they can synthesize information in a great way.

me in science is not the clinical implications but, rather, the native curiosity to understand how something works and to address it and hopefully get a little insight into it. If there are translational clinical benefits that emerge, so much the better, but as a basic scientists it’s the curiosity of how nature works that drives you, not the desire to cure human disease. In some cases, because we work on motor neurons and there are lots of motor neuron diseases, there are, gratifyingly, implications of what we’ve done; the stem cell stuff being a good example. But we don’t do that. Even though we discovered how to make motor neurons from stem cells, our role is to act as a cheerleader in this and encourage other people to use these basic advances to a more clinical end. I feel that’s what I like doing and our greatest contribution is always going to be: to pick apart basic mechanism and then tell people, try and introduce to people, why understanding basic mechanism should be interesting even if you have interests in the clinical nature of the problem. It doesn’t mean that in the end one doesn’t want to contribute to the curing of a disease or more effective therapies, but you have to realize your limitations and divide the labors amongst people.

AM: Maybe I’ll ask one more question since your protégé [Dr. Marc Tessier-Lavigne] just walked in and since Dr. Eric Kandel was here before. How important are mentors in the field of neuroscience, or any science? TJ: Well, it’s very important because you can’t do everything yourself and, whether they’re protégés or mentors, they are smart people in science and you want to surround yourself with smart people. You want to surround yourself with people that are smarter than you are, ideally, because then you can talk and get ideas and integrate and you’ll have a different perspective. So, one of the great things about being at Columbia is that you’ve got Eric, and you’ve got Richard Axel and last night we heard Charles Zuker. These are all people who are passionate about science [and who] will tell you their opinions, even if it’s not what you want to hear. You can listen to that, you can re-evaluate and it’s an exchange. And then you try and convey that way of thinking to people that you come into contact with; younger people, like you!

AM: Thank you so much for your time, Dr. Jessell and conAM: And it’s probably a less bias way to approach things. gratulations on your award. We can’t wait to see what you do next!

You come closer to the truth if you don’t set out with that goal in mind, right?

TJ: Thank you. I have no idea...which is usually the case.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

INTERVIEWS

Dr. Jeffrey Ravetch Jerry V. Ravetch investigates the cellular and molecular mechanisms underlying the generation of antibody specificity and the effect of that specificity on cellular responses. Dr. Ravetch graduated from Yale University in 1973 and completed his Ph.D. in 1978 from Rockefeller University. He received his M.D. from Cornell Medical School in 1979. Dr. Ravetch joined the faculty of Memorial Sloan-Kettering Cancer Center and in 1984 and was appointed professor at Rockefeller in 1996. He has received numerous awards, including the Canada Gairdner International Award, the Burroughs-Wellcome Scholar Award, the Pew Scholar Award, the Boyer Award, the NIH Merit Award, and the Lee C. Howley.

Interview conducted by Benedict Darren

“. . . sometimes mentoring means knowing when to walk away and let the student figure out for his or herself what the right experiment is, what the right questions are.” BD: First question, after graduating, you completed an MD/ Ph.D program jointly at Cornell University. As an undergraduate, what inspired you to pursue both research and medicine at the same time? JR: Good question. So my interests in undergraduate studies were mostly in chemistry and physics, and I found myself interested in expanding the areas of work into biological systems. So after completing my chemistry and physics training, and in molecular biochemistry and biophysics, I thought that the best way to learn biology would be to go to medical school. So my interest was never really in medicine but much more in human biology. I wanted to understand how biological systems operated, and what was known about the health and sickness situations. So that was the genesis of my interests—I never really thought I’d want to practice medicine and I thought I’d like to learn about human diseases. BD: Right, and I guess with these transitions between different

subject areas, you must have had good mentors along the way. So how critical do you think it is to have good mentorship when taking on a career in the life sciences?

JR: Well, of course it’s very valuable. I was very fortunate as an undergraduate to have a tremendous mentor in chemistry, Donald Crothers, who was the first to introduce me to science in a serious

way. I didn’t really have the background before coming to Yale. I learnt from that what laboratory research was all about, watching how his lab functioned. When I got to Rockefeller, people like Norton Zinder and Peter Model were role models as well as mentors. Mentoring is a complex process, sometimes mentoring means knowing when to walk away and let the student figure out for his or herself what the right experiment is, what the right questions are. But the most important thing I thought to instill in my students is not so much whether you master techniques, but whether you learn how to ask the right questions, and how to have a sense of when a problem is deep enough and rich enough to give you a satisfying result.

BD: And following that, when did you first find that you were really starting to ask the right questions, the questions which you were interested in? JR: It’s an ongoing process; we’re always trying to ask the right questions. I guess when I started my graduate work, I remember posing a variety of different problems to Norton as to things I might want to work on. He showed his interest or lack thereof in certain questions, and I started to understand what questions he found interesting and that began the shaping of my questions. I think it really reached a more developed stage during my postdoc years with Phil Leder, transitioning from bacterial genetics to mammalian biology with someone who is really an expert at

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Dr. Jeffrey Ravetch

knowing and having a good nose for a good problem. That was how it started, and when I started my own laboratory I tried to take on some of those traits. Some of the problems we asked I think were the right ones. Others didn’t develop quite as well as I’d thought, and I’ve walked away from some of those. So there’s always a constant checking and moving up then back.

“Don’t be persuaded by everyone saying that’s not important, or that’s not interesting, or that’s been done already.” BD: That’s a good point. So, you work a lot on antibodies and at the time it was first proposed, your model which implicated inhibitory Fc receptors in anti-inflammatory responses obviously generated its share of controversy. What challenges were presented from the scientific community and how did you overcome them?

Research Interviews Articles

BD: That’s true. And as you alluded to it just now, there is some clinical potential for incorporating antibody glycosylation patterns and Fc receptors. What arenas of treatment are you exploring currently? JR: Well it’s actually moved beyond potential, it’s now actually in clinical studies. The manipulation of an Fc to enhance its ability to kill a tumour cell is now a widespread approach that’s being used in many different companies for developing new anti-tumour drugs. The development of small molecule inhibitors that will prevent Fc receptors from becoming activated in autoimmune diseases is well into late-stage clinical testing. So those are actually accepted therapeutic approaches. The manipulation of Fc by changed glycosylation is an accepted modality now for application. So I think the areas we’re pushing ahead on are the anti-inflammatory areas and seeing to what extent the antibody Fc components are important in infectious diseases, and neutralization of viruses and bacterial toxins. That’s an area we’re now focusing on.

BD: As a last question, a lot of our readers are interested in JR: I’d step back here even further. When we proposed that the embarking on a career similar to yours. What personal advice would

Fc was required for the in vivo biological activity of neutralizing antibodies, of anti-tumour antibodies, and pathogenic antibodies, it was dismissed as being an incidental finding. That had two effects really: one, it left the field to me alone—and that was a good thing, so I worked in isolation. But it meant that I was certainly not in the mainstream of immunology, and I was kind of in a peripheral side. And that was fine, because I liked the idea of being left alone to do my own work. It probably took longer to have the ideas become incorporated into mainstream thinking—what could have taken 10 years took 20 years. But I think that’s always the situation when you’re coming into a field that is perceived as being well-established and settled, and you’re introducing ideas that people find to be outside the mainstream. So it really becomes your obligation to convince your colleagues that you’re right, that it’s worth considering these different approaches. The sialylation example is still highly controversial, and it’s still taking it’s time to work into mainstream thinking. I think the ideas of Fcs as being required for cytotoxicity and pathogenicity is now accepted, but we’re in a constant state of evolution. The penetration among scientists is high, but the penetration in the medical community is very low. Medicine will take another decade before it begins to incorporate ideas that are now becoming established in the scientific community.

you give our readers to be successful?

JR: Well I think success is, of course, a combination of luck and hard work. But I think the single lesson I’ve taken away is that you have to be self-motivated, you have to work on the problem because you find it interesting, and persist in it if you’re convinced that it’s the right problem. It requires a certain level of being thickskinned, and impervious to the outside world. Ironically, it means ignoring the literature as often as we do, because if you’re convinced you’re right, you have to go with your instinct. Don’t be persuaded by everyone saying that’s not important, or that’s not interesting, or that’s been done already. BD: That’s great advice. On behalf the JULS staff, I’d like to thank you for your time and wish you the best of luck on your future endeavours.

BD: I’d assume that’s always the case with clinical applications. JR: Not just the applications, but the teaching too. Seeing how medical schools are teaching courses on inflammation and the role of antibodies in inflammation, they are unfortunately 10, 20 years behind the literature. The textbooks are incorporating the ideas that are now really emerging, and it’ll probably take another 10 years before drugs are on the clinical practice arena that are manipulating these pathways and then that would change, I think, the way of approaching things. So it’s highly conservative— scientists tend to be conservative, clinicians tend to be even more conservative, so that explains these transitions.

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

JULS

INTERVIEWS

Dr. Michael Young Michael Young’s work is focused on neuromuscular development and the genetics of behaviour, particularly circadian rhythms. Dr. Young received his Ph.D. in genetics from the University of Texas in 1975. He completed postdoctoral work in Stanford and was an investigator at the Howard Hughes Medical Institute. Dr. Young is a member of National Academy of Sciences and a Fellow of the American Academy of Microbiology. He has received various awards, including the 2009 Neuroscience Prize of the Peter and Patricia Gruber Foundation and the 2011 Louisa Gross Horwitz Prize of Columbia University.

Interview conducted by Alina Guna

“...I didn’t even know that plants could move. Why these rhythms and so forth?” AG: So you did your undergraduate degree in biology and then you went on to genetics and then biochemistry. Is it safe to say you’ve always wanted to end up in biology? MY: Yeah, yeah. I have been interested in biology from very early in childhood. I grew up in South Florida, Miami. AG: I grew up in Miami too! MY: So you know. It was even crazier when I was a kid because everybody’s backyard is a jungle...a parrot jungle, a monkey jungle, the zoos and the botanical gardens and everything. Things are always escaping. You drive down the street, even now and you see iguanas, four foot feet iguanas. And kids would have these strange pets like a sloth or an ocelot or Caymans, these alligators. So with all this going on you’ve just got...for me you were always picking something up or wondering whether you should pick something up. And so there was a sort of a naturalist basis to my interest and then of course as I got into high school and later in college I was taking courses that embellished on that. A lot of biology but then, you also need to know chemistry; in particular biochemistry. And you become more and more dissatisfied with

superficial answers. For example as a kid there used to be this plant that would open its blooms at night and I was so fascinated because I didn’t even know that plants could move. Why these rhythms and so forth? But you could learn a bit about it and see descriptions but you could never get much past observations. The original interest has everything to with how animals behave, how they move. I was interested in plants and animals both, it evolved into more complex questions as I got older.

AG: So how did your interest evolve to Drosophila, which you have been working with for about what, 30 years? MY: You know it’s interesting because if I hadn’t grown up in Southern Florida, for example and been exposed to all this stuff who knows what I would have ended up interested in. If I had grown up in the Silicon Valley it could have gone in a very different direction. The interest in Drosophila came in a sort of accidental way. I was taking a course in genetics at Austin University in Texas as a senior undergraduate and I had this terrific professor who really drew me in with what you can do with genetics. He was available for summer research projects and I just walked in one day and said ‘What are you doing? I liked your course’ and he told me about

Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

Dr. Michael Young

the kinds of questions they were asking in this lab. They were really good questions. They weren’t about behaviour but they were about how chromosomes are put together, how genes are organized, how they’re packed into chromosomes. One of the things that interested me first was he was asking the question about how many genes there were. How many genes does it take to build something like a fly? Because there are huge amounts of DNA in

“Ultimately you have to say, is this really interesting to me? Is the problem growing in interest?” a fly and yet the best available genetic methods to assess functions were making it look like E. coli, 5,000 genes or so. Even today there are like 13, 000 genes. So I was interested in what kind of regulatory biology could make such a small number of genes do such different things comparing E. coli to a fruit fly. Drosophila had this wonderful genetics, you could build almost anything. But by the end of the time that I was a graduate student I had sort of run out of ways to approach these questions. Michael Rosbash talked about recombinant DNA technology falling from the heavens at the right time and that is really the case. We had sort of done all that we could without touching molecules but then this really made it possible to isolate a gene that governs a biological clock and look at it in chemical detail, manipulate it and modify it. So again, if I hadn’t had grown up in Miami, would I have this naturalist kind of background? If I’d been a few years earlier I might have had to take up something else just because there were not adequate approached for the big problems I wanted to work on.

Interviews

is resistance that makes sense and there is resistance that doesn’t make sense. You have to sit back and say look, I think I know how to do this and I know what it is I want to ask. Over the course of doing that if you see progress and you continue to see a way forward then you have to make your own judgements. You also get different kinds of advice from different people. At the same time as the committee by neighbour who was also a faculty member would run down every day to see how this project was going, ‘have you fixed though arrhythmic flies yet with your cloned DNA?’ So, I’m sure it would be tough if you only had negative feedback but you’re always going to get a mix of feedback. Ultimately you have to say, is this really interesting to me? Is the problem growing in interest? This was at first an interest in behaviour but it’s become an interest in biology. We have rhythmic bodies and the loss of those rhythms have huge consequences. Recent experiments in mice have shown that if you knock these out in just one tissue in the pancreas, even if all the other clocks in the body are fine you get diabetes mellitus. If you knock it out in just the liver you cause the mouse to become obese and you have severe regulatory problems with energy storage. We didn’t anticipate any of that. We expected these genes to be active in the head and would at best explain how they were controlling this feature of behaviour. But things like that keep drawing you off and keep pushing you forward. You are always looking for experimental feedback or feedback from discussions but ultimately…don’t give up too easily. You may find that the problem that interests you now will interest you more later and that’s a really good sign that you should keep at it.

AG: One final question, obviously a lot of people who read this journal are aspiring scientists. What advice would you give for when times are looking rough and.... MY: Yeah, well I can tell you a story. When I first moved to Rockefeller where I was an assistant professor and my first graduate student wanted to work on per. I would tell him the story: ‘Look I know where this thing is all we need to do is clone this DNA and we can test it by moving these gene into flies that don’t have clocks,’ and so forth and so on. He was all fired up and we went into his faculty advisory committee, his first faculty advisory committee that students have to defend their project. There was a senior behavioural biologist faculty that was a member of that committee, long ago retired...and genetics had been used like this before...and he said ‘Let’s say you get this gene, what can you learn about behaviour from a gene? And even if you do learn something about controlling this behaviour in the eye, what’s that going to tell you about how we sleep and how our rhythms and biology work?’ Those were hard questions to answer because you don’t know yet! These are tools and you want to apply the tools. Obviously we went ahead and the argument you have to make at times like that is you have to do the work. We’re not just looking for excuses to have tied hands. So let’s look and if we’re not making progress, if we get to a point where we can’t see a way forward then OK we will have to concede some of these. But how can you know if you don’t go ahead and push forward? So, back to this question, there Journal of Undergraduate Life Sciences • Volume 7 • Issue 1 • Spring 2013

nd s

Fro m: Tre

ll Ce

6 ; 6 : 2 6 7-73 y, 19 9 l og B io