Scientia - Winter 2017

Scientia

A Journal by The Triple Helix at The University of Chicago

Winter 2017

Front Cover Photo: The schematic of an idealized neuron processing binary code along its dendritic branches.

Attribution: Nicolas P. Rougier, 2005. Released under the GNU General Public License.

Scientia 5 7 10 12 14

Inquiry: A Glimpse into a Psychologist's Psyche: Picking the Brain of UChicago's Dr. Norman Samara Singh

Research Advice:

Tips and Trick to Make You a Research Whiz Clara Sava-Segal and Nilanjana Ray

Abstracts: Health and Social Sciences Abstracts

Calculations of MODY Risk in Younger Adult Diabetic Patients in Nigeria The Effect of Three Separate SNPs on Neural Activity of Patients Diagnosed with Prosopagnosia: an fMRI Study A Quantitative Assessment of Telomere Length and Mortality in Arsenic-exposed Bangladeshi Populations Associations between Gut Microbiota and Neurological Disorders in the American Gut Data Integration of Behavioral Health and Type 2 Diabetes Care at Community Health Centers in the Midwest Adapting a Mobile Cessation (mCessation) Intervention for a Pacific Islander Context: The Guam Experience Investigating Social and Infrastructural Parameters Concerning Toxoplasmosis in Panama

Life Sciences Abstracts

Flightless-I and Daam1 are Proteins with Opposing Functions that Interact at the Barbed End of Actin Filaments Dynamics of Asterias forbesi and A. vulgaris microbial communities in relation to sea star wasting disease (SSWD) of the Northeast coast of the U.S. Probing the Kinetic Isotope Effect and pH Sensitivity of the Cyanobacterial Circadian Clock through In Vitro Reconstitution TEL-mediated Dysregulation is Highly Sensitive to Changes in its Polymerization

Produced by The Triple Helix at the University of Chicago Layout and Design by Helena Zhang, Production Director Cover Letter written by Jeremy Chang, Co-Editor in Chief Scientia Board: Clara Sava-Segal, Jeremy Chang, Nilanjana Ray, Abhinav Ranjan, Quang Tran

Winter 2017

16 19 21 24 25 28 2

Potentiation of D1 Dopamine Receptor Signaling Pathways Using Positive Allosteric Modulators (PAMs) CRISPR-Cas9 as a Novel DNA-Damage Tool for Studying Senescence Assessing Sequence Conservation and Phylogenetic Relationships Amongst X-Family DNA Polymerases Signaling Centers That Pattern Developing Cerebral Cortex Are Conserved between Mice and Ferrets The Effects of Amyloid Beta (Aβ) on Mitochondrial Function in Hippocampal Neurons Observed in vitro Lipid Dynamics to Confirm Novel Model for CNS Myelinogenesis The Effect of CEACAM-7 on Interactions Between the Small Intestinal Epithelium and Patient-Derived Strains of B. fragilis Using Intestinal Epithelial Cell Line Monolayers as an in vitro Model TBX5-microRNA Negative Feedback Loop Maintains Atrial Rhythm Transcriptional Co-Activators YAP and TAZ Drive Melanoma Progression and Survival Through Differing and Overlapping Pathways Variation in Male Genital Morphology Within and Between Populations of Poecilia mexicana during Ecological Speciation Measuring the Growth of Shewanella oneidensis MR-1 to Understand Nanoparticle Toxicity Characterizing the Acoustic Behavior of Chilean Blue Whales (Balaenoptera musculus) using DTAG Accelerometers Morphological Variation in Himalayan Birds Implies Biogeographical Rules can be Explained by Biotic Interactions Efficacy of Telavancin in Mouse Skin Infection Model In vitro studies of Penelope-like retroelements in the genome of the bdelloid rotifer Adineta vaga Generalized Category Encoding in Posterior Parietal Cortex Across Two Categorization Tasks

Physical Sciences Abstracts

Hydrogen Diffusion in Olivine Phenocrysts from Episode 1 of the 1959 Kilauea Iki eruption Precision Photometry on the SDSS Photometric System Improved Estimation of Atmospheric Particulate Matter in Three U.S. Regions Using Remote Sensing with Public Health Applications Machine Learning at the Energy Frontier: A Signal Processing Approach Towards Finding Dark Matter & Natural Supersymmetry Neutrino Beam Optimization An Investigation of Thermotropic Blue Phase and Nematic Liquid Crystals Optical Resonance Imaging: Sub-diffraction-limited Imaging on an Ultrafast Timescale Interaction of Ultrafast Terahertz Electromagnetic Pulses with Metamaterials A Type-theoretic Foundation for Unsupervised Learning Surface Modification of Ion Conducting Ceramic Particles

Scientia

31

Purification of Multiply Kinked Nanowire by Configuration Novel Carbonized Gelatin-Laponite System with Previously Unreported Thermal Stability for Applications in Bone Repair

32

Scientia Original Abstracts

33 43 49 54

Vagabond Viruses: Variations in Human Endogenous Retroviral Sequences of HCP5 and Susceptibility to Exogenous Viruses Alterting Lentiviral Tropism: Design and Implications of Targeted Drug-Delivery System

In Depths: Going for The Goal: The Effects of Removing Preparatory Information on The Fast and Unconscious Shira Eisenberg

The Effect of Electromagnetic Fields on the Arion distinctus in the Presence of Various Wavelengths of Light Sofia Garrick

Statistical Microlensing and Gravitationally Lensed Black Holes Patrick Halkett

ERBB2 Germline Mutations Predispose to Cancer and Schizophrenia Hyesan Lee

3

Winter 2017

About Scientia Dear Reader, The front cover of this year’s Winter Edition depicts the schematic of a neuron processing binary code along its dendritic branches. Neurons represent the basic unit of the nervous system, and their activities result in the emergent property known as cognition, which encompasses decision-making, problem-solving, and comprehension. One of our Full-Length articles this issue investigates the role of preparatory cues on the human-machine interface, exploring whether humans can effectively process computational cues to make predictions in ways that are analogous to more naturalistic paradigms, such as human-human interactions (see article “Going for the Goal…” for more).

Scientia aims to celebrate the vibrant University of Chicago research community by featuring work from our very own undergraduates. With this goal in mind, our Winter Edition offers an expanded selection of four Full-Length articles this year, as opposed to the customary two articles. In addition to discussing the human-machine interface, our article line-up also examines the electromagnetic sensing abilities of slugs, gravitational lensing as a means to study black hole phenomena, and the effects of hereditary mutations in the ERBB2 gene that contribute to increased risk for schizophrenia and cancer. Our Edition also includes an Inquiry with faculty member Dr. Greg Norman, as well as the very best abstracts from the 3rd Annual University of Chicago Undergraduate Research Symposium in collaboration with Phoenix Biology and UCISTEM. Scientia is always looking to broaden our scope to new and novel research fields. If you are completing a research project and want to see it in print, or if there is a professor performing eye-opening investigations that you want to be promulgated, consider writing for us!

We encourage all interested writers to contact a member of our team, listed in the back. In the meantime, please enjoy this Winter Edition of Scientia, presented by The Triple Helix.

Yours, Jeremy Chang Co-Editor-in-Chief, Scientia

4

Scientia

Inquiry A Glimpse into a Psychologist's Psyche: Picking the Brain of UChicago's Dr. Norman Samara Singh

Ask twenty-year-old Greg Norman what was on his mind at the time, and you probably would have learned about Ohio State sports teamsdefinitely not about his interest in pursuing a PhD in psychology or a professorship at the University of Chicago. Dr. Norman was a student athlete at Ohio State University, competing as a short-distance sprinter on the track team, until he was injured his sophomore year. No longer spending his time at the track and soon running out of time taking “joke classes”, Norman realized he had to redirect this newly unused energy towards a more academic, pre-professional track.

Channeling his passion for sports and competitive spirit into his academics, Norman decided to obtain a PhD, seeking to achieve the best in whatever task he undertook. He used the momentum of starting a totally new path to instigate his love for psychology.

Starting his psychology major as a junior and having to play catchup, Norman opened himself up to many incredible lab opportunities. As Dr. Norman put it, it took “unbelievably fortunate events of working in a couple of labs [he] just randomly picked” to allow him to work with pioneers in the field of neuroscience, such as Dr. Gary Bernstein and now University of Chicago’s Dr. John T. Cacioppo.

After receiving his bachelor’s degree and PhD from Ohio State University, Dr. Norman worked

in the well known Dr. Cacioppo’s lab for his postdoctoral research. Having worked with animal models in graduate school, but not wanting to be pegged as the “animal dude”, Dr. Norman worked with Cacioppo on a project for the army. Norman cites his time working with Cacioppo as some of the most intense times, having “learned twenty years’ worth of information in two years” because “that’s how John was, always five steps ahead of everyone else”.

For his post-doctoral research in Cacioppo’s lab, Dr. Norman, along with others, did research on how to increase soldiers’ resilience prior to going off to war, with hopes of soldiers becoming less likely to develop mental health issues after their tours. The researchers designed interventions and coping mechanisms to try and increase the soldiers’ awareness of their social skills and surroundings.

Currently, Dr. Norman’s lab at the University of Chicago is interested in psychophysiology and how people react to stressful environments. His lab works with a school in Indiana that specializes with children who either have social problems or have been enrolled there by School Social Workers. Dr. Norman’s lab has observed how children react to their adverse environments for almost three years now. They utilize physiobands that record heart rates, teacher ratings on the kids’ behaviors, and

5

Winter 2017 other sources of information on the children’s home lives. Still gathering much data, Norman’s lab hopes to gain more insight on the relationships between immune function and physiological behavior states; for example, how one’s immunity can be affected by stress symptoms, like increased heart rate, sweating, lack of appetite, etc.

In addition to his research, Dr. Norman is an Associate Professor, his first faculty position at the University of Chicago. Joining the faculty in 2012, Norman had no idea what to expect, but shared that “teaching at a place like this is just unfair. It’s too easy because everyone is so smart.” Teaching both graduate and undergraduate classes, Dr. Norman’s favorite class to teach is Neurobiology of Stress. This class brings in a broad, diverse range of students and is his particular favorite considering its applicability with U of C Students. “Stress is a tricky thing; for the most part, stress isn’t bad. It fuels our fires, but it needs to be controlled in a happy medium.”

Dr. Norman finished off our interview considering how “it almost feels obligatory to complain about something,” but he is unable to since he’s enjoyed his time thus far at the University of Chicago. Dr. Norman plans to keep enjoying his time teaching and researching as he continues to claw his way to tenure.

About the Author Samara Singh is a second year student at the University of Chicago majoring in biochemistry.

6

Scientia

Research Advice

Tips and Trick to Make You a Research Whiz Clara Sava-Segal and Nilanjana Ray

As a part of the larger Triple Helix division, Scientia focuses on Chicago-based research. This includes staff and student-conducted research in all fields within the social, and physical sciences. However, our newest venture aims to help undergraduate students discover the research options on campus.

UChicago is a university with a vast number of opportunities for students of all backgrounds to try; research is no exception.

--------Second Year Biological Sciences Major, Managing little about. The Institute of Molecular Engineering Editor of Scientia, and Research Assistant in Huang had a helpful webpage that showcased the works/ Lab in the IME, Nilanjana Ray, shares her experiences: projects of each of the research professors. Being a biology major, I was more eager to learn about Background: the bio-focused molecular engineering research.

As a first year considering the biological Therefore, at the start of winter quarter, I reached sciences major, I knew I wanted to continue doing out to a handful of professors in this field. It wasn’t research as I had done in high school. But, I honestly easy to get a response. Therefore, I also used the was not sure where to begin. Using Handshake UCIHP listhost and the Student Employment and other resources, I began looking at available page to find interesting positions in the UChicago opportunities. I realized I did not want to work in a Hospital, as well. Eventually, a professor responded typical biology lab, since the research did not entail with interest. what I wanted to do. Since the research I did in high school involved physical sciences and energy, I felt After the Response: limited not having done wet lab work outside my To reassess, I read more on his research and courses—the limited experience was valuable and decided I was actually interested in pursuing the enjoyable, nonetheless, and I felt more inclined to offer. I met with a postdoc to discuss my previous do this type of lab work within research. background and my research interests, specifically in this lab. I was then able to meet with the PI and Reaching Out: discuss, once again, what my plans were. From this, After speaking with an upperclassman about I was able to begin considering topics for projects his previous experiences, I began to research the new I could work on in the lab. I started as a research molecular engineering department - a field I knew assistant and am still currently working there.

7

Winter 2017 The Position: I attend weekly meetings, which includes journal club, where the lab can collaborate to discuss progress and future project ideas based on presentations of papers. I’ve honestly learned a lot about wet lab techniques and biological concepts that I would later learn in my actual bio classes at the College, and even beyond, from the papers I’ve read and work I’ve done so far.

Summer Work: I continued working in my lab this past summer but had an unpaid position. With the help of Career Advancement, I was able to turn my job into a Metcalf internship and was subsidized by the University’s funding. Not only was I able to stay in Chicago and enjoy my summer, but also I was also able to get a sense of working full-time in a lab and could focus on my research without classes. I’ll admit, working in a lab closely with a good group of people is a great bonding experience, and I’ve even gotten to spend time with my lab group beyond just doing research together. I’ve gained extremely helpful and supportive mentors, and I wouldn’t trade my experience.

-----Tips to Apply: *Indicates a tip that was applied in the narrative discussed above 1. Don’t be afraid to reach out, communicate and ask for help. There are plenty of resources on campus. * Professors: Reach out to the head of the department you’re interested in or a professor or TA that you have. Undoubtedly, they are doing their own research and know what types of work their colleagues are doing as well, in addition to who might have openings in their lab. Even if there aren’t open positions, talking to different PIs about their research can help you get a sense of what you may be interested in.

Career Advancement advisors and your academic advisors are great resources and middle grounds between you and your future research position.

List hosts: Career Advancement and its different divisions, like UCISTEM and UCIHP, as well as major list hosts send out weekly emails containing tons of opportunities to join labs

8

and research programs. It never hurts to apply to something that may look even a remotely interesting! * Student Employment Page: Oftentimes professors have their lab managers post about positions directly there * 2. Expand beyond campus responses: Reach out to students. * Upperclassmen: Talk to upperclassmen that are involved in research! Sometimes it helps to get advice from those that know your struggle best.

You may even learn of some cool opportunities your fellow peers have had that you may have never heard of. * Scientia and the rest of the Triple Helix branches offer some insight into the research your peers are doing. You may find some inspiration from these experiences.

RSOs: There are RSOs that also help you find cool programs or openings, so make sure you take advantage. GeneHackers is an example of one that has members perform their own research to compete at an international conference annually. Those in charge of these RSOs can often guide you as well.

3. Explore on your own: Department Websites.

Each department at the university has helpful information on their websites that contain links to the pages of the different faculty research groups. Lab sites often give insight on both the past works of professors and how to contact them. Sometimes, professors will note whether they are looking for students to join. Read up on their research to have a heads up when you meet with them to identify specific topics you would potentially like to work on*.

4. Don’t feel obligated to stick to research only within your major. * Some departments do have certain requirements if you choose to write a thesis for honors, but any major can do research in any field they’re interested in. If you are thinking of writing a thesis, explore your options early; talk to the department head or even professors you would be interested in having as an advisor.

5. Look for summer research opportunities!

They don’t always have to be on the UChicago campus. You would be surprised by how many summer research programs there are on various campuses throughout the country. Some

Scientia universities might have research opportunities 7. You are never bound to a lab: there is no shortage that UChicago does not cover. Likewise, many of research positions. official research programs do the hard work for This doesn’t mean that you should lab-hop and you by providing pay, housing, and setting you switch every month, but feel free to be flexible. up with a direct professor and research project This is your time to discover interests. Give it a to complete by the end of the summer. shot, and if it doesn’t suit you, don’t feel pressure Just a couple of examples: UCLA has SPUR, to keep that to yourself. Take the time to find an NYU’s Medical School has SURP, Mayo Clinic in area that you are passionate about - you could Minnesota has a fellowship, as does Memorial propose your own project from some of the work Sloan-Kettering Cancer Center in New York. you do.

8. Any step counts.

Don’t forget to keep your resume up to date as Research isn’t all about immediately becoming you begin applying – any experience helps! an assistant on an important project. Start small 6. Worried about finding a research position but if you need to. Try to be as involved as you not getting paid for it? can, especially through weekly lab meetings The University has a plethora of grants and and journal club*; you could learn valuable new funding opportunities* if your lab cannot research that you could apply to creating your subsidize your time with them, even for summer own project. Stay engaged, and don’t let the lab opportunities. Each department may have some work get monotonous. grants, but there are different programs you can also apply to that are often mentioned in list host emails.

For instance, there are Metcalf Internships and Summer Action Grants readily available for students to use. A useful tool is the Center for Scholarly Advancement that has a tool to get you paired up with fellowships throughout the entire world for your field!

Career Advancement

UCISTEM UCIHP Metcalf Center for Scholarly Advancement

Summer Programs Outside of UChicago:

Summer Action Grants

Summer Programs through UChicago:

UCLA SPUR

Argonne National Laboratory

Amgen Scholars

Research Experiences for Undergraduates (REU) Fellowship

Stanford Summer Research Tufts University Sackler School Summer Research Program UCSD SURF Program U. of Buffalo SURE Penn State University College of Medicine Research Internship Program Northwestern Feinberg School of Medicine CURE Mayo Clinic

Mellon Mays Undergraduate Fellowship Physical and Chemical Biology Training Program (PCBio) Beckman Scholars Program in Molecular Sciences BCSD Summer Fellowship Program VIGRE at the University of Chicago Leadership Alliance PRISM

Medical College of Georgia STAR National Institute of Health Summer Internship

9

Winter 2017

Abstracts Health and Social Sciences Calculations of MODY Risk in Younger Adult Diabetic Patients in Nigeria Adeoluwa Ayoola University College Hospital, Ibadan, Nigeria Diabetes is a heterogeneous disease that affects a growing number of people around the world. Maturity-onset Diabetes of the Young (MODY), a subtype of monogenic diabetes, is a rare form of diabetes that occurs in younger adults, usually before the age of 25 years. There is very sparse data on occurrences of MODY in African natives. However, MODY has been reported among African Americans. Given the high occurrence of atypical diabetes in Africa and the lack of resources in African hospitals to properly diagnose a disease like MODY, it is possible that there are young African adults who have MODY. In order to determine the possibility of occurrence of MODY in Africans, with the goal of improving precision of diabetes treatment in Africa, this research project was undertaken. Nigerian diabetic patients were diagnosed based on ADA/WHO clinical criteria at University College Hospital, Ibadan, Nigeria before the age of 40 years. Of the 134 younger adult patients with onset of diabetes (before 40 years) that were extracted from a clinic registry compiled in 2010, 19 had the complete data required for analysis using our risk assessment tool. The risk assessment tool used was the online MODY risk calculator run by the University of Exeter. One major finding from the study is that 5 out of 19 (23%) of the younger adult patients measured had significant risk for MODY. Significant risk was defined as risk higher than 20%, the threshold for further testing recommended by the assessment tool used. Additionally, females appear to have higher risk calculations than males. Further studies will be required to confirm their diagnosis. From the study thus far, it is concluded that there is a possibility of misdiagnosed MODY cases among patients with existing clinical diagnosis of diabetes within these Nigerian subjects. Therefore, the continued rigorous study of MODY and diabetes of all forms is necessary in the African population in order to address the need for improved diagnostic measures and precision of treatment in Africa.

10

Scientia

The Effect of Three Separate SNPs on Neural Activity of Patients Diagnosed with Prosopagnosia: an fMRI Study Tyler Chan Peking University, School of Life Sciences, IDG McGovern Institute for Brain Research Prosopagnosia is a facial recognition disease diagnosed when a patient is unable to recognize and remember another human face. Our lack of knowledge of causes, risk factors, treatments, therapies or even rigorous methods of diagnosis for this cognitive disorder make an inquiry into the neural and genetic basis of prosopagnosia worthy of further investigation. Based on genotypes and results of several facial recognition behaviour tests, candidate gene quantitative association was used on a general population and three diseased families, revealing three separate single nucleotide polymorphisms (SNP) on a gene found to be associated with facial recognition ability. We wanted to see if people carrying different genotypes for the gene would perform differently in fMRI, specifically patients diagnosed with prosopagnosia. This project covers results from 9 subjects who underwent functional magnetic resonance imaging (fMRI) testing while performing a facial recognition and memory test. The task was composed of four parts: 1) a localization test 2) a Cambridge Facial Memory Test (CFMT) 3) a specificity test, 4) an adaptation test. Current results show that subjects from the general population not diagnosed with prosopagnosia have increased neural activation in the fusiform face area (FFA), limbic lobe and regions of the temporal lobe during our facial recognition and memory tasks. Further fMRI tests will be done on 41 more general population subjects and members of the 3 diseased families, in order to make more robust conclusions on the relationship between genotype at the specific gene’s location, facial recognition ability, and neural activity in those specific brain regions.

This basic genetic, behavioural, and physiologic inquiry into the brain’s neural activity in the general population and patients with different genotypes in a specific gene associated with prosopagnosia has broad clinical diagnostic applications and is a step forward to better understanding the mechanisms and underpinnings behind this disease.

A Quantitative Assessment of Telomere Length and Mortality in Arsenic-exposed Bangladeshi Populations Sammi Dean The University of Chicago, Department of Public Health Sciences, Pierce Research Group Telomeres serve as a cap, protecting chromosome ends from damage. Shorter telomere length (TL) has been associated with overall mortality, although the literature on this topic is not entirely consistent. In this study, we assessed the association between prospectively-measured TL and subsequent mortality within three arsenic-exposed Bangladeshi cohorts: Health Effects of Arsenic Longitudinal Study (HEALS), HEALS Expansion (HEALS-E) and Bangladesh Vitamin E and Selenium Trial (BEST). We investigated overall mortality, mortality from chronic diseases, respiratory diseases, and circulatory diseases, and mortality from neoplasms (growths indicative of cancer). Telomere length was measured using quantitative PCR and QuantiGene Plex technology. The association between TL and mortality was assessed using odds ratios produced by conditional logistic regressions. In HEALS-E there was no significant association between TL and any of the investigated mortality outcomes. In BEST and HEALS there was a significant association between shorter telomeres and overall mortality (P=0.008 and P=0.048 respectively). This inverse association was also present in both cohorts for mortality from chronic disease (P=0.01 and P=0.02), and mortality from circulatory disease (P=0.008 and P=0.04). In BEST alone, the association was see in mortality from neoplasm (P=0.04). In the pooled analysis, shorter TL was significantly associated with overall mortality (P=0.004), death due to chronic disease (P=0.004) and death due to circulatory disease (P=0.0009). The results suggest an inverse association between telomere length and mortality in this population.

11

Winter 2017

Associations between Gut Microbiota and Neurological Disorders in the American Gut Data

Remy Lee University of Chicago, Department of Human Genetics, Ober Lab

There is growing interest in gut microbial influences on brain functions in humans. Here we investigated associations between gut microbial composition and phenotype status for seven neurological disorders. Data were drawn from the American Gut Project, which collects fecal samples through a crowdsourcing model and detects the abundance of bacterial strains, termed operational taxonomic units (OTUs). The dataset includes abundance counts of each OTU in each sample and a compilation of participant responses to a questionnaire on health and demographics. Statistical analyses showed that for three of the seven disorders, cases contain significantly more diverse microbial compositions than controls. A regression-based test for association between the abundance of individual OTUs and disorder status showed few significant signals across the board, especially when the regression accounted for confounding variables strongly correlated with each disorder. Without the additional covariates, however, 48 of 142 genera were significantly associated with migraines. This finding calls for a focus on confounders that mediate the microbial influence of neurological disorders. One genus Brevibacterium was strongly associated with epilepsy; a different genus in the same order has previously been linked with seizure. The study demonstrates that while difference in microbial composition may influence neurological disorders, individual microbial effects are rarely significant.

Integration of Behavioral Health and Type 2 Diabetes Care at Community Health Centers in the Midwest Melissa Li University of Chicago Department of Medicine Section of General Internal Medicine Patients with Type 2 diabetes have a two to three-times higher risk of developing comorbid depression, resulting in decreased quality of life, increased medical care costs, and increased mortality along with other adverse outcomes.

Many of these patients are cared for in community health centers (CHCs), which have higher rates of diabetes and depression. While previous studies have shown that behavioral healthcare integration through the Collaborative Care Model, which calls for practices such as care coordination and screening, is effective in managing comorbid diabetes and depression, little is known about the current practices of CHCs regarding behavioral healthcare integration. In this study, a cross-sectional survey of 67 MidWest Clinician’s Network affiliated CHC directors was used to describe behavioral health (BH) and diabetes care integration at Midwest CHCs and to examine relationships between CHC-reported stage of integration to their integration practices. The majority of surveyed CHCs were found to have already implemented or completed their BH and diabetes care integration. Also, CHC-reported stage of integration was most closely associated with BH and primary care office co-location and a shared BH and diabetes registry manager. No association was seen between CHC reported stage of integration and tracking diabetes in a BH registry or tracking depression screening in a diabetes registry. These data suggest that there is a need for improvement of existing levels of BH and diabetes care integration in Midwest CHCs, especially in increasing the level of integration of BH and diabetes registries.

12

Scientia

Adapting a Mobile Cessation (mCessation) Intervention for a Pacific Islander Context: The Guam Experience Noel Martin D. Rubio The University of Guam Cancer Research Center Text messaging (SMS) has been proven effective for tobacco cessation. Most of the research and development has taken place in developed countries with comprehensive tobacco control policies, and education and media campaigns. Research is needed to investigate the potential to transfer proven effective SMS interventions to less developed countries with very different tobacco control and cultural contexts. The National Cancer Institute recently launched its SmokefreeTXT program, which helps users set a quit date and provides motivational messages leading up to the date and for the weeks following the quit date. The program is available to Guam residents and can be accessed with domestic SMS charges. We conducted a usability study of the SmokefreeTXT program with Chamorro and other Micronesian young adults to: (1) explore their beliefs about quitting benefits and barriers, and their interest in the textbased program; (2) follow participants’ use of the program over a two-week trial period; and (3) explore their evaluations of the program, usability problems, and reactions to new messages about chewing tobacco with areca nut/betel quid as well as new messages tailored to Chamorro and other Micronesian cultures. Pre and post-trial focus groups with 15 Chamorro and Micronesian smokers were augmented by an online survey with around 100 respondents. Young smokers in Guam were highly receptive to a text-based cessation program. Concerns about addiction and cancer, and the desire to quit increased after the 2-week SMS trial, while current smoking decreased. Survey data collection is ongoing. The potential for text-based cessation is significant within the Pacific Island community and the fine-tuning of message content to contextualize the program within the Pacific Islander culture will enhance program acceptability.

Investigating Social and Infrastructural Parameters Concerning Toxoplasmosis in Panama Aliya Moreira and Abhi Pandey Hospital Santo Tomas

The protozoan Toxoplasmosis gondii is one of the most prevalent parasites in the world. The major source of transmission is through cat oocysts, which can come into contact with humans through a variety of ways. Infection is generally asymptomatic in nature, with the parasite lying dormant in over 2,000,000,000 people worldwide. However, congenital transmission can cause brain and eye damage, fetal malformation, or even death. Panama specifically has one of the highest rates of infection in the world. Therefore, it is important to both enforce screening of pregnant mothers and to determine environmental sources of infection in the country. Thus, there was a dual objective of this study: to map seroprevalence incidence rates and to track screening rates at public health centers and private clinics. We hoped to differentiate areas of high incidence in order to examine potential environmental differences and to determine which health centers were and were not providing mandatory screening. To do so, we spent 5 weeks collecting screening data, IgG/IgM test results, demographic information, and addresses from patient prenatal control sheets at Hospital Santo Tomas. Data was collected from a total of 665 pregnant women. We found that 38% had received screening, with individual townships varying between 20-60% screened. Private clinics had a greater rate of screening than public health centers, and the neighborhoods of Pueblo Nuevo and Curundu had higher rates of screening. Of those screened, indigenous people and those living east of the canal were more likely to test positive for toxoplasmosis.

Certain townships, like Curundu, have drastically greater rates of positivity. Additionally, positive points are clustered around roads and waterways. The results of this study provide useful information for the Panamanian Ministry of Health to enforce universal screening within the country’s public health centers. Also, it identifies areas of future study—like the neighborhood of Curundu—that could provide clues to methods of transmission of the disease within Panama.

13

Winter 2017

Life Sciences Flightless-I and Daam1 are Proteins with Opposing Functions that Interact at the Barbed End of Actin Filaments Diane Alkatout The University of Chicago, Department of Molecular Genetics and Cell Biology, Lab of David Kovar

Actin is a protein that is present in cells as part of the cytoskeleton, which gives cells their structure and shape. Actin monomers can be assembled to form filaments, and these filaments are important for several cellular processes, such as division, endocytosis, and motility. In order for these processes to successfully occur, actin filaments need to be assembled at the correct time and place, and with the proper architecture and dynamics. This is coordinated through the action of different actin binding proteins, some of which bind to monomers and nucleate filaments, and others which can bind to filaments and either cap, sever, or crosslink them. One such protein is Flightless-I, which is named for the mutant phenotype in Drosophila. While this protein has been shown to colocalize with actin in vivo, its interaction with actin has not been well characterized. Biochemical assays have shown that filaments are shorter in the presence of this protein, but it is uncertain whether this is due to severing or capping. I utilized fluorescent spectroscopy as well as total internal reflection fluorescence microscopy (TIRFM) to characterize the Gelsolin-like domain (GLD) of Flightless-I, which is the actin-binding domain. Results show that this protein caps actin filaments by binding to the dynamic barbed end, inhibiting filament elongation. In vivo, antibody imaging shows that Flightless-I localizes to filopodia, which are membrane protrusions that aid in cellular motility. Daam1 is a binding partner of Flightless-I that also localizes to filopodia, and its function is to increase the elongation rate of filaments at the barbed end. Because of the opposing functions of Flightless-I and Daam1, I again used fluorescent spectroscopy and TIRF to determine if there is any interaction between these proteins. TIRF reveals that these proteins can colocalize at the barbed end, and both TIRFM and fluorescent spectroscopy suggest that Daam1-CT increases the dissociation rate of Fli-IGLD, even at high concentrations of Fli-I-GLD.

14

Scientia

Dynamics of Asterias forbesi and A. vulgaris microbial communities in relation to sea star wasting disease (SSWD) of the Northeast coast of the U.S.

Isa Alvarez Marine Biological Laboratory, Bay Paul Center, David Mark Welch Lab

Sea stars play a crucial role in coastal marine ecosystems as keystone predators. Recently, a new disease has caused thousands of stars to wash up on the west and east coast shores of the U.S., puzzling researchers everywhere. Previous works have established a connection between the sea star associated densovirus (SSaDV) and the wasting disease affecting different species on the west coast. In this study, I closely monitor the effects of sea star wasting disease on A. forbesi and A. vulgaris on the east coast of the U.S. in regards to the microbiome of sea stars and the effects of the microbiome on susceptibility to SSWD. First, I examined several sea stars using FISH (fluorescence in situ hybridization) to determine what structures, if any, possess a complex microbiome. Only the aboral surface had a microbiome. I then collected DNA samples from the aboral surface of fresh-caught stars and from the same stars after acclimating to the aquarium environment of the MRC. We will examine infected and uninfected stars under the confocal microscope, amplify and sequence DNA, and isolate viral fraction of infected stars for metagenomic sequencing to determine the effect of handling/swabbing stars in regards to passing the infection. Potential findings in this study could help future researchers studying the disease on the west coast and provide information on the potential differences in the disease due to geography.

Probing the Kinetic Isotope Effect and pH Sensitivity of the Cyanobacterial Circadian Clock through In Vitro Reconstitution Daniel Beaglehole University of Chicago, Rust Lab

Circadian rhythms exist within a range of organisms from bacteria to humans. They are precisely characterized by the ability to maintain stable oscillations with a period of ~24 hours in constant environmental conditions and the ability to adjust to external stimuli by aligning the phase of oscillation with the local environment. In the cyanobacterium Synechococcus elongatus, critical processes such as cell division, global gene expression, photosynthesis, and nitrogen fixation are regulated by a circadian oscillator composed of three core proteins: KaiA, KaiB, and KaiC. The timing information is stored in a series of phosphorylation states of KaiC, whose inherent autophosphatase and autokinase activities are regulated by KaiA and KaiB in a rhythmic manner throughout the internal (subjective) day-night cycle.

Notably, the oscillations of KaiC phosphorylation generated by its autophosphatase and autokinase activities can be reconstituted in vitro by combining the core clock components with adenosine triphosphate (ATP). The ability to reconstitute this clock reaction in vitro with relative ease allows us to directly perturb various conditions of the clock reaction, leading to an understanding of the mechanism underlying the clock’s interesting properties that have been observed and possibly conserved across species, without introducing the complications of the in vivo environment. It has been observed in certain mammalian cells that the period of circadian oscillations is lengthened when grown in the presence of deuterium oxide (heavy water). Furthermore, it remains unclear whether and how circadian clocks remain robust to changes in intracellular pH. In this study, we addressed these questions through probing the sensitivity of the KaiABCoscillatortopHandisotopeeffectsinvitro.Remarkably,wefoundthatrhythmicoscillationsinKaiCphosphorylation were maintained at a variety of pH conditions and deuterium oxide contents, demonstrating the adaptability and robustness of the clock to external perturbations. However, we found that the period of the clock was highly sensitive to both of these perturbations, prompting further investigation into the mechanisms leading to these responses.

15

Winter 2017

TEL-mediated Dysregulation is Highly Sensitive to Changes in its Polymerization

Juana Delao University of Chicago, Ben May Department of Cancer Research, Rebay Lab

In order for a cell to survive and function, it must be able to properly regulate the transcription of genes. Transcription factors are proteins that bind to regulatory regions in the genome and form complexes that control the expression of other genes. Some cancers can be driven by mutations in transcription factors that result in dysregulation of transcription. How cells respond to this dysregulation, apart from oncogenesis, is not well understood. TEL is a polymerizing transcription factor that regulates blood development in mammals. Chromosomal translocations, where the polymerizing domain of TEL is fused to the DNA-binding domain of another transcription factor, have been shown to lead to some forms of leukemia. In essence, these translocations produce novel polymerizing transcription factors, but how these transcription factors operate in a new context to disrupt endogenous transcription is not well understood. In order to study the effects of introducing a strong novel polymerizing transcription factor into a genome, mammalian TEL was overexpressed in the eye of the fruit fly Drosophila melanogaster using the GAL4UAS system. This led to a large disruption in the normally well-organized pattern of the eye, resulting in a rough eye phenotype. In order to determine how TEL is able to cause this roughness, several point mutations of TEL that disrupt its ability to polymerize to varying degrees were made and overexpressed in the eye as well. Preliminary results suggest that TEL’s ability to produce a rough eye is dependent on its ability to polymerize, as monomeric TEL produces a wild-type eye. Additionally, point mutants that only slightly decrease TEL’s ability to polymerize also produce an eye indistinguishable from wild type, further suggesting that TEL is highly sensitive to changes in polymerization. Future work will attempt to elucidate the molecular mechanism for TEL-mediated dysregulation.

Potentiation of D1 Dopamine Receptor Signaling Pathways Using Positive Allosteric Modulators (PAMs) David Gao National Institute of Neurological Disorders and Stroke

Dopamine is a catecholamine neurotransmitter that regulates several important processes including cognition, motivation, and reward. Its dysregulation has been linked to a number of diseases, including schizophrenia, Alzheimer’s Disease, and Parkinson’s Disease. There are five G-protein coupled dopamine receptors, grouped into two families: the D1-like, including D1 and D5, and the D2-like, comprised of D2, D3, and D4. The D1-like family is a target for treating cognitive decline resulting from neuropsychiatric diseases, as activation of these receptors enhances cognition.

However, D1-like receptor orthosteric agonists have a number of liabilities, such as hypotension and a narrow therapeutic window. As an alternative, the Sibley lab searched for D1 receptor (D1R) positive allosteric modulators (PAMs). These compounds bind to an allosteric site separate from the receptor’s orthosteric binding pocket and increase dopamine’s binding affinity and/or signaling efficacy. Our laboratory previously screened the NIH Molecular Libraries Program to search for D1R PAMs and discovered two compounds, MLS1082 and MLS6585, which potentiate dopamine binding and signaling at the D1-like dopamine receptors. My project this summer was to compare our two PAMs to a novel D1R PAM, Compound B. We confirm Compound B is a non-orthosteric ligand via competition binding experiments. Compound B acts as a PAM by potentiating both G-protein (cAMP accumulation) and ß-arrestin (recruitment)-mediated signaling, while not having any intrinsic agonist activity. Compound B did produce a 2-fold increase in dopamine’s potency for both b-arrestin recruitment and 5-fold shift in cAMP accumulation with no effect on Emax. This potentiation was lower than that reported in the literature and compared to MLS1082 and MLS6585.

Finally, we ran Compound B, MLS1082, and MLS6585 in combination with the previous experiments to determine if the PAMs could be functioning through the same or separate sites on the receptor. Our results suggest that Compound B may be competing with MLS1082 and MLS6585 at their binding sites on the D1R dopamine receptor. In conclusion, we have further characterized a new PAM, Compound B, and compared its activity to our own compounds. These D1R PAMs may serve as scaffolds for the further development of drug leads to treat neuropsychiatric disorders.

16

Scientia

CRISPR-Cas9 as a Novel DNA-Damage Tool for Studying Senescence Ellen Iverson University of Chicago, Department of Molecular Genetics and Cell Biology, Kron Laboratory Common experimental means of inducing DNA damage, including radiation or drug treatment, are unable to produce pure DNA damage without having other secondary impacts on cells. This limitation has prevented rigorous investigation into the precise role of DNA damage in various cellular processes, including accelerated senescence. Using CRISPRCas9, I have developed a means of inducing pure, dose-dependent DNA damage in cells that could help overcome this obstacle. Experiments have shown that the CRISPR-Cas9 construct is capable of inducing extensive DNA damage. I have also demonstrated that the level of stable Cas9 protein present in transfected cells and the degree of DNA damage can easily be modulated using a small molecule. Studies have linked senescence to various age-related pathologies, including serious lung conditions like idiopathic pulmonary fibrosis. Improving understanding of the relationship of DNA damage to accelerated senescence and the mechanism behind this relationship through research facilitated by this novel application of the CRISPR-Cas9 technology may help suggest new treatments for age-related diseases.

Assessing Sequence Conservation and Phylogenetic Relationships Amongst X-Family DNA Polymerases Shantal Jayawickreme National Institute of Environmental Health Sciences, Genome Integrity and Structural Biology Laboratory The X-family in humans contains four DNA polymerases: beta, lambda, mu and terminal transferase, or TdT. These polymerases are generally involved in DNA repair. All four polymerases operate through a catalytic subdomain which contains three Asp residues interacting with two metal cations at the catalytic site. The polymerases differ, however, with respect to insertion accuracy and the template dependence. Beta and lambda have a higher insertion accuracy than TdT and mu. Additionally, TdT is able to repair short gaps in DNA without a template, while beta, lambda and mu require the presence of a template. In this project, we wanted to better understand the differences that may correspond to the varying fidelities amongst the polymerases by investigating the sequence conservation and structural differences amongst the four X-family members. In order to do this, we first had to compare the sequences of the polymerases in a variety of species, using the genome browser, Ensemble. From the browser, we collected hundreds of sequences and then performed multiple sequence alignments using the program, Mafft. We then created residue conservation diagrams for each polymerase using Jalview in order to identify functionally important residues of each polymerase. Finally, we performed a maximum-likelihood test with the program, MEGA 7.0, to generate phylogenetic trees. A few different observations were gleaned from this analysis. In performing residue conservation analysis of a specific region (called “Loop 1” in mu and lambda), it was found that beta and lambda have the best residue conservation, indicating that the residues of the region are quite important in polymerase beta and lambda function, compared to TdT and mu. It was also found that between TdT and mu, TdT has better conservation than mu, indicating that the “Loop 1” region may play a more important role in TdT. Secondly, in creating phylogenetic trees for the four polymerases, it was found that TdT may have evolved from polymerase mu as chordates evolved. These results corporately indicate that that TdT may have branched out from a common ancestor of polymerase mu (a metazoan ancestor) in order to fulfill a function required specifically in chordates.

17

Winter 2017

Signaling Centers That Pattern Developing Cerebral Cortex Are Conserved between Mice and Ferrets William D. Jones The University of Chicago, Department of Neurobiology, Elizabeth A. Grove Laboratory

The functionally specialized regions of mammalian cerebral cortex are responsible for higher cognitive functions such as decisionmaking, perception, and memory. These cortical regions are specified by signaling molecules in the developing embryo, where secreted factors from sources called signaling centers direct different cortical regions to adopt distinct fates. Furthermore, many of these signaling molecules are redundant at different time points in development, so both the spatial pattern of signal expression and the temporal changes in expression during development contribute to the particular fate that the cortex will develop. While many cortical signaling molecules have been identified and their roles described in the lissencephalic mouse, their relevance to cortical development in gyrencephalic species, including humans, is unknown. Differences in cortical size and morphology, most prominently the highly folded cortical surface of gyrencephalic species, suggest that signaling centers might contend with different spatial considerations in these two types of cortex. To assess the relevance of area patterning mechanisms in the mouse to species with a gyrencephalic cortex, we used a gyrencephalic model: the ferret Mustela putorius furo. We first described the external features of the ferret embryo and the morphology of the brain at several developmental time-points when cortical signaling centers are active, and we used this data to determine the temporally equivalent neural developmental stages in mouse and human embryos. Based on these equivalent stages, we showed that the ferret neocortex is no larger than the mouse neocortex when signaling centers are active. We next characterized, with in situ hybridization, the expression patterns of several genes implicated in cortical patterning at these developmental stages, including Fibroblast Growth Factors, Bone Morphogenetic Proteins, Wnt proteins, and Emx and Nr2f1 transcription factors. Comparison of the expression patterns in mice and ferrets reveals strong similarities, providing evidence that highly similar signaling mechanisms could govern cortical area patterning in lissencephalic and gyrencephalic species. Overall, we have shown that principal regionalization of the telencephalon occurs before the gyrencephalic cortex substantially increases in size and complexity, allowing signaling centers to pattern both gyrencephalic and lissencephalic brains using a conserved developmental program despite these brains’ morphological disparities.

The Effects of Amyloid Beta (Aβ) on Mitochondrial Function in Hippocampal Neurons Nikita Mehta Whitman Center, Marine Biological Laboratory

Alzheimer's Disease (AD) is a neurological disorder characterized by cognitive and motor deterioration that affects memory, attention, speech, and behavior. One of the factors underlying the progression of AD is the accumulation of amyloid beta (Aβ) protein plaques that may block synaptic transmission. It has been previously suggested that mitochondrial dysfunction may play a role in the onset or progression of AD. We now demonstrate that Aβ contributes to cellular dysfunction by affecting both mitochondrial membrane potential and calcium retention capacity -- two measures of mitochondrial function and efficiency. After isolating and purifying mitochondria through density-gradient centrifugation, we found that Aβ-treated mitochondria exhibited a reduction in calcium uptake ability in comparison to control mitochondria. Using confocal microscopy of the mitochondrial membrane potential indicator, TMRE, we visualized and quantified a significant difference in mitochondrial membrane potential between control and Aβ-treated hippocampal neurons. In addition to drastically reducing the membrane potential, Aβ peptide also affected cell morphology – the neurons and mitochondria were irregular in shape, size, frequency, and distribution. We predict that these findings are the result of an Aβ-induced inner mitochondrial membrane leak, and hypothesize that Aβ’s interaction with B-cell lymphoma extra-large (Bcl-xL), a specialized Bcl-2 family protein, may cause the opening of the mitochondrial permeability transition pore (mPTP), resulting in decreased mitochondrial competency. Bcl-xL’s binding to the beta subunit of the F1F0-ATP synthase reduces the inner membrane ion leak, thereby increasing the flow of ions through the H+ translocator of ATP synthase to make ATP and enhancing the efficiency of energy metabolism in the mitochondria. Since Bcl-xL, among other regulatory proteins, is vital for preventing the neurotoxicity that leads to AD, our findings suggest that Aβ may disrupt regulation of the gating of the inner membrane leak channel within ATP synthase, leading to functional decline of the mitochondria.

18

Scientia

Observed in vitro Lipid Dynamics to Confirm Novel Model for CNS Myelinogenesis Andrew Molina Lee Lab, University of Chicago

The myelin sheath is an insulating, compacted, multilamellar biological membrane that facilitates efficient propagation of action potentials down neuronal axons and is critical for normal physiological function. Demyelinating diseases, such as multiple sclerosis, lead to severe physical and mental impairment, and even death. Recently, University of Chicago neurobiologist Sara Szuchet has proposed a novel mechanism for myelin formation in the central nervous system (CNS) to address fundamental flaws in the canonical model. To investigate this mechanism, we designed in vitro experiments to probe the biophysical interactions of myelin lipids as they (1) self-assemble into tubules and (2) transition into lamellar form. Specifically, using TEM imaging, we have observed the selfassembly of tubules consisting of 100% galactosylceramide (GalCer), a fundamental glycolipid of CNS myelin. We have similarly investigated how two-component systems of GalCer/Cholesterol and GalCer/Plasmalogen transition from stable tubule structures to larger aggregates of lower curvature upon decreasing GalCer concentration. Our results suggest that these structures of major myelin lipids can be stable precursors for myelination. Using fluidcell AFM, we have begun to investigate how these self-assembled tubules interact with supported lipid bilayers in HEPES buffer solution (pH 7.4) with hopes of observing a fusion event between the structures that would be experimental evidence of the proposed Szuchet model for the biological mechanism of myelin formation.

The Effect of CEACAM-7 on Interactions Between the Small Intestinal Epithelium and Patient-Derived Strains of B. fragilis Using Intestinal Epithelial Cell Line Monolayers as an in vitro Model An Nguyen University of Chicago Department of Medicine, Chang Lab

Patients with ulcerative colitis (UC), an inflammatory bowel disease that affects the colon and rectum, may have their colon and rectum removed and replaced with a pouch constructed from the ileum connected to their anus. Nearly half of these patients will suffer from pouchitis, or inflammation of the ileal pouch, the pathogenesis of which is still unclear. Preliminary data from a pilot study following UC and non-UC pouch patients has revealed gene expression consistent with colonic metaplasia in the UC pouch that is not seen in the non-UC pouch or the pre-pouch ileum. These alterations may promote the virulence and attachment of select microbial species in the colonic-like pouch microbiota, leading to pouchitis. Assessing the effects of these colonic metaplasia gene products on interactions between ileal pouch tissues and patient-derived potential pathobionts could elucidate the pathogenesis of pouchitis, as well as UC. We found that CEACAM7, which is normally expressed in the colon but not the ileum, is highly expressed in the ileal UC pouch. CEACAM7 is a part of a gene family involved in cell matrix adhesion, which may be critical in bacterial invasion.

Other preliminary data indicates that eight out of nine pouchitis patients have significant increases in the relative abundance of Bacteroides species. Using monolayers of human colon-derived Caco2 cells that express very low levels of endogenous CEACAM7 as an in vitro model of the ileum, we transiently transfected them with CEACAM7 and then infected them with patient-derived strains of B. fragilis. We developed an attachment assay to determine whether CEACAM7 affects B. fragilis adhesion to ileal epithelial cells. Additionally, we examined cell death and NF-ÎşB pathway activation with Western blots and cell barrier integrity with immunofluorescence tight junction staining. Preliminary results indicate that CEACAM7 has no significant effect on B. fragilis attachment or monolayer viability and integrity following infection. However, CEACAM7 seems to be coexpressed with other members of the CEACAM family, so we may need to transiently transfect the monolayers with CEACAM7 in conjunction with other CEACAMs. The assay developed will be useful in assessing the relationships between other colonic metaplasia genes and potential pathobionts.

19

Winter 2017

TBX5-microRNA Negative Feedback Loop Maintains Atrial Rhythm Kaitlyn M. Shen University of Chicago, Departments of Pediatrics, Pathology, and Human Genetics, Moskowitz Lab We investigated the role of small regulatory RNAs, specifically microRNAs (miRs), in atrial fibrillation (AF), the most common cardiac arrhythmia world-wide. AF is a highly heritable disease, and TBX5, a T-box transcription factor, has been linked to AF by genome wide association studies (GWAS). We generated a Tbx5-dependent AF animal model by removing Tbx5 from the adult mouse using an inducible Cre system. Adult-specific Tbx5 deletion resulted in spontaneous and sustained AF, characterized by electrocardiographic and echocardiographic approaches. We hypothesized that Tbx5 drove a network of regulatory RNAs in the adult atrium. In order to define this network, we performed small RNA sequencing on left atria from Tbx5-deleted and corresponding control mice and found that 108 miRNAs were differentially expressed. Interestingly, the most downregulated small RNA was miR10b, which has been previously shown to directly regulate Tbx5. Chromatin immunoprecipitation (ChIP) of the adult atrium showed TBX5 occupancy at the miR10b promoter, indicating direct Tbx5 regulation of this microRNA. Transfection of a miR10b mimic into the HL-1 mouse atrial cardiomyocyte cell line reduced Tbx5 expression. Conversely, transfection of a miR10b inhibitor led to increased Tbx5 expression. Taken together, these data provide evidence for a Tbx5-mir10b regulatory loop. miR10b overexpression in HL-1 cardiomyocytes led to prolongation of the cellular action potential, phenocopying the result of Tbx5 deletion in mouse atrial myocytes.

These data suggest that a TBX5-microRNA negative feedback loop is necessary for maintaining atrial rhythm.

Transcriptional Co-Activators YAP and TAZ Drive Melanoma Progression and Survival Through Differing and Overlapping Pathways Kelsey Ogomori Deborah Lang Lab, University of Chicago Department of Medicine Metastatic melanoma accounts for the majority of skin-cancer related deaths each year. Yes-associated protein (YAP) and transcriptional co-activator with PDZ-binding motif (TAZ) are transcriptional coactivators that have both been shown to promote survival and metastatic potential in a wide variety of cancers. As YAP and TAZ do not contain DNA-binding domains, they are often found to drive cancer progression through interaction with the DNA-binding TEAD family of transcription factors. YAP and TAZ have similar structures and are often cited as paralogs of each other with identical function but differing levels of expression. Although they have been studied to some degree in other cancers, their function in melanoma is relatively unknown. To first investigate the role of YAP and TAZ in melanoma, we performed siRNA-mediated knockdowns targeting YAP, TAZ, and both YAP and TAZ in multiple melanoma cell lines. We immediately noticed that the YAP knockdown groups consistently exhibited a longer, more dendritic phenotype as compared to the siScrambled control and TAZ knockdown groups. RNA-sequencing analysis was then performed on these three knockdown groups (YAP, TAZ, YAP/TAZ) in Mel537 cells whereby we were able to generate lists of differentially expressed genes for each knockdown group. Global ontology analysis clearly showed that these transcriptional coactivators have different gene expression profiles. To further investigate this, we generated lists of YAP-specific, TAZ-specific, and YAP/TAZ specific targets. We then validated several YAP and TAZ specific targets in two additional melanoma cells lines. One of the targets was actin-related protein 2/3 complex subunit 5 (ARPC5), which may be responsible for the YAP specific dendritic phenotype. Several TEAD sites have been identified in the ARPC5 gene promoter, making ARPC5 a possible direct downstream target of YAP-TEAD. Our results suggest that although YAP and TAZ are traditionally thought to have similar structure and overlapping functions, this might not necessarily be the case in melanoma, as there is the presence of both YAP- and TAZ-specific targets.

20

Scientia

Variation in Male Genital Morphology Within and Between Populations of Poecilia mexicana during Ecological Speciation Alexander Okamoto, Ryan Greenway, and Michael Tobler Kansas State University, Division of Biology, Tobler Lab

Male genital morphology is one of the most rapidly evolving traits in internally fertilizing species. Male live-bearing fishes of the family Poeciliidae are characterized by the gonopodium, a modified anal fin, used to transfer sperm to the females. Poecilid gonopodia are morphologically complex, exhibiting variation in the number and position of ornaments and weapons across species. Fishes in the Poecilia mexicana species complex have independently colonized toxic hydrogen sulfide-rich springs in multiple tributaries of the RĂ­o Grijalva in southern Mexico. Sulfide spring fishes are reproductively isolated from ancestral populations in adjacent non-sulfidic habitats. Several barriers to gene flow have been documented in this system, yet they are not sufficient to explain observed levels of reproductive isolation. Variation in male gonopodia morphology between populations could be serving as an additional reproductive barrier. In this study, we quantified variation in gonopodia between sulfidic and nonsulfidic populations of P. mexicana in order to test for lock-and-key mechanisms. We also investigated variation within populations associated with different male mating strategies. These analyses were replicated across four pairs of sulfidic and non-sulfidic populations to determine whether gonopodia morphology has evolved in convergence. We assessed male body size and shape, as well as the size and shape of isolated gonopodia for each of our focal populations. We found a negative allometric relationship between gonopodium length and body length related to male mating strategy. Our results provide no clear evidence for lock-and-key mechanisms between sulfidic and nonsulfidic populations. However, there is a consistent pattern of within population variation in gonopodia, with larger males exhibiting more ornamentation. Several mechanisms of sexual selection could be driving convergent evolution of male gonopodia morphology in P. mexicana. Future studies on the functional consequences of different gonopodia morphologies and the reproductive success of different male mating strategies are necessary to elucidate the processes generating evolutionary convergence in gonopodia structure.

Measuring the Growth of Shewanella oneidensis MR-1 to Understand Nanoparticle Toxicity Ellen Purdy University of Minnesota, Twin Cities, Chemistry Department, Carlson Lab

Nanotechnology is an emerging field of science with diverse applications in materials science and medicine, among other areas. Although nanoparticles show promise in solving many problems, their environmental effects have not been fully studied. Due to the increasing implementation of nanotechnology in everyday life, the toxicity of nanomaterials is of great concern. My work focused on the toxicity of lithium nickel manganese cobalt oxide (NMC), a battery cathode material, against Shewanella oneidensis MR-1, a dissimilatory metal reducing aquatic bacterium. Previous research shows that NMC is toxic to S. oneidensis. One measure of toxicity is bacterial culture growth. Four methods were used to analyze bacteria growth: protein concentration, colony counting, optical density measurements, and dry pellet weight. Protein concentration and colony counting account for viable bacteria only, while optical density measurements and dry pellet weight account for all bacteria present in the culture. Data was collected across three days of bacterial growth. Results showed significant variability between cultures but an overall trend toward decreasing optical density, protein concentration, and dry pellet weight over the time course. We found that colony counting was an ineffective method of measurement, while reproducible data was obtained with the other three methods. The purpose of this research is to further analyze the impact of NMC nanoparticles on the growth of S. oneidensis. Data gathered over this time course will be used to establish a baseline of normal growth against which to measure toxicity.

Further understanding of the mechanisms of toxicity will lead to development of less toxic nanoparticles in the future.

21

Winter 2017

Characterizing the Acoustic Behavior of Chilean Blue Whales (Balaenoptera musculus) using DTAG Accelerometers Mark Saddler Woods Hole Oceanographic Institution, Department of Biology

Vocal behavior of blue whales (Balaenoptera musculus) in the Gulf of Corcovado, Chile was analyzed using both audio and accelerometer data from digital acoustic recording tags (DTAGs). Over the course of three austral summers (2014, 2015, 2016), seventeen tags were deployed, yielding 124 hours of data. We report the occurrence of Southeast Pacific type 2 (SEP2) songs, which exhibit peak frequencies, durations, and timing consistent with previous recordings made using towed and moored hydrophones. We also describe tonal downswept (D) calls, which have not been previously described for this population. Since being able to accurately assign vocalizations to individual whales is fundamental to studying social behavior and for estimating population densities from call rates, we further examine the feasibility of using high-resolution DTAG accelerometers to identify low-frequency calls produced by tagged blue whales. We describe a method of cross-correlating acoustic signals with simultaneous tri-axial accelerometer readings in order to analyze the phase as well as the amplitude of body vibrations associated with low-frequency calls. Analyzing the phase match of acceleration and audio signals in addition to the amplitude of body vibrations provides a reliable method of determining if an acoustic signal is associated with a detectable body vibration. Our results suggest that vocalizations from nearby individuals are capable of exciting detectable body vibrations in the tagged whale. We propose a solution to identifying which of the accelerometerdetected calls originate from the tagged animal by using cross-correlation of acceleration vectors between calls in order to find signature vibration patterns associated with sounds produced within the tagged whale.

Morphological Variation in Himalayan Birds Implies Biogeographical Rules can be Explained by Biotic Interactions

Matthew Schumm The University of Chicago, Department of Ecology and Evolution, Price Lab

Bergmann’s and Allen’s rules state the following: within and across endotherm species, body size and appendage size increase with higher latitudes and altitudes (and decreasing temperatures). This trend has been historically attributed to selection in colder climates favoring increased energy storage capacity and minimized body heat loss. Using previously published data on Himalayan bird morphology, ecology, and altitudinal distribution, we evaluate the patterns predicted by these rules and the factors and processes from which they would follow in 489 bird species along an altitudinal gradient in the eastern Himalayas (a region supporting the second-highest level of bird diversity in the world) that spans climates from subtropical to boreal. This largely forested altitudinal gradient offers limited habitat variability, a steep and stable decline in mean temperature, and no dramatic lineage turnover.

Therefore it is an ideal study system for a test of the hypotheses underlying these biogeographical rules. Because biotic factors such as interactions between competing species and the abundance and diversity of the resources for which they compete are thought to explain many observed trends in morphological diversity along latitudinal and altitudinal gradients (including trends consistent with the predictions of Bergmann’s and Allen’s rules), we use sister-species analyses based on phylogenetic relationships to measure the nature and intensity of competition for resources between closely related species at different altitudes. We also use principal components analysis (PCA) and hypervolume-based computational techniques to assess how patterns of ecological niche space occupation in these species change with altitude and with the availability of resources such as prey insects. We find that Bergmann’s and Allen’s rules are not valid generalizations for these species, a finding consistent with recent studies of bird diversity along montane gradients at tropical latitudes. We also find that biotic pressures (and whether these pressures favor divergence or allow for convergence between species), rather than changes in temperature or other abiotic factors, explain the altitudinal variation that we do see in (a) body size dispersion and (b) in relative sizes of appendages.

22

Scientia

Efficacy of Telavancin in Mouse Skin Infection Model Ryan Singh University of Chicago, Department of Pediatrics, Daum La Telavancin is a relatively new antibiotic that is very similar to structure to the common antibiotic vancomycin. Both have been shown to prevent the growth of Staphylococcus aureus, including MRSA strains, in vitro, namely by inhibiting cell wall synthesis. The main structural difference between telavancin and vancomycin is the presence of an additional hydrocarbon chain in telavancin that is thought to improve its ability to bind to the cell wall of Grampositive bacteria. Due to its relatively new approval status by the FDA, the exact efficacy of telavancin is still being investigated, especially in vivo. One target for this investigation would be the USA300 strain S. aureus. USA300 is a community-associated MRSA strain commonly found in the United States, often causing infections in skin and soft tissue. The specific strain of USA300 known as 923 (ST8, Clonal Cluster 8, SCCmec type IV) was investigated in an in vivo mouse model to determine if telavancin was effective in treating dermonecrosis caused by the 923 bacteria in skin infections. The mice were all infected with the 923 strain, then were treated in subsequent days via telavancin injection or saline injection (control). Pictures were taken of the mouse lesions every day after the infection in order to measure the size of the dermonecrotic area. Analysis of the pictures indicated that the telavancin-treated mice showed significantly smaller regions of dermonecrosis overall than the untreated mice at a 5% significance level, suggesting that telavancin was an effective treatment in this skin infection model.

In vitro studies of Penelope-like retroelements in the genome of the bdelloid rotifer Adineta vaga

Vishok Srikanth Arkhipova Lab of The Josephine Bay Paul Center for Comparative Molecular Biology and Evolution at the Marine Biological Laboratory Penelope-like elements (PLEs) represent a class of unusual retroelements that, despite their widespread presence in eukaryotic genomes, propagate through a poorly understood transposition mechanism. PLEs encode a reverse transcriptase (RT) domain that is highly divergent from that of other retroelements, and have introns within inserted sequences – a counter-intuitive feature for RNA-propagated mobile elements. Further complicating the structurefunction relations underlying PLE behavior, it has recently been reported that the pseudo-long terminal repeat (pLTR) regions flanking some PLE insertions contain previously undiscovered variants of Hammerhead ribozymes (HHRs) – a type of small self-cleaving RNA that could play a role in processing PLE transposons during replication. In the context of the “RNA world” view that postulates genetic and catalytic functions for RNA prior to the emergence of DNA- and protein-based life forms, these RT and HHR motifs may provide valuable insight into the history of transposons and the diversity of RNA-based mobile elements. The current work aims to improve biochemical understanding of PLE domains and of these elements’ transposition mechanism, focusing in particular on an apparently active family of PLEs identified in the genome of a bdelloid rotifer, Adineta vaga. These PLEs are characterized by a continuous open reading frame (ORF) containing RT and endonuclease (EN) domains, flanked by two HHR motif-containing pLTR regions. In silico analysis suggests that part of the pLTR may interact with the EN domain. Expression of this recombinant ORF has been moderately successful and suggests that its EN domain is active within the expression host.

Separately, the pLTR motif associated with this family has been incorporated into a vector for in vitro transcription.

Our experiments are aimed at characterizing the self-cleaving activity of the HHRs in these pLTR regions and their interaction with RT, and examining the recognition and cleavage preferences of the EN domain on the pLTR substrate.

23

Winter 2017

Generalized Category Encoding in Posterior Parietal Cortex Across Two Categorization Tasks Oliver Zhu University of Chicago, Department of Neurobiology, Freedman Lab

Neurons in the lateral intraparietal cortex (LIP) encode task-relevant information about visual stimuli, such as color, shape, category membership, and task rules. In a well-studied delayed match-to-category (DMC) paradigm, monkeys are trained to categorize a 360° range of motion directions into categories based on an arbitrary, learned category boundary. In this task, monkeys indicate a category match by releasing a lever. However, it is unknown if LIP neuronal populations reflect a generalized categorization process or represent category signals exclusively in the DMC task because of its particular task demands. To test for generalized category encoding, we trained monkeys on a novel categorization task in which monkeys report the category identity of the visual stimuli with an eye movement. Tentatively called the one-interval categorization (OIC) task, monkeys here are trained to associate the same motion directions to one of two color targets—both of which appear simultaneously after stimulus presentation as the proxy by which monkeys indicate their choice. Through single neuron electrophysiology, we found that although some neurons show task-dependent encoding differences, the majority of neurons in LIP simultaneously encode category signals across both tasks. Our results provide support for a model of LIP with mixed selectivity for task-relevant variables that ultimate guide successful perceptually driven behavior.

24

Scientia

Physical Sciences Hydrogen Diffusion in Olivine Phenocrysts from Episode 1 of the 1959 Kilauea Iki eruption Hannah Bausch Lamont-Doherty Observatory at Columbia University, Department of Geochemistry

The time it takes for magma to reach the top of a volcano can be linked to why some volcanoes erupt explosively while others erupt effusively. One possible way of measuring magma’s ascent time is with hydrogen diffusion profiles.

Hydrogen (H+, colloquially known as water) concentration profiles from olivine phenocrysts from Episode 1 of the 1959 Kilauea Iki eruption were used in this study to help constrain magma ascent rates of that eruption. While ascent times have previously been quantified using melt inclusion and melt embayment data, this study aims to help constrain those times by getting data directly from the crystal itself. Water concentrations were measured in corerim profiles across five crystals by Fourier Transform Infrared (FTIR) Spectroscopy. The FTIR data was processed by 1-D modeling in the python package pynams (Ferriss 2015). Volatiles diffuse out of magma at lower pressures so it is expected that water concentrations should be lower at the rims of the crystal than at its core because diffusion occurs at the rims of the crystals first. Water diffusion profiles from this study do not follow expectations and instead show that water concentrations are either relatively constant across the crystal or show a higher concentration of water at the rim than at the core. Three of the five profiles taken have water concentrations that are 32%, 35%, and 40% higher at the rim than at the core. Rims of the crystals were identified by the visible glass along the edge. The higher water concentrations at the rims may be present because the crystal partially equilibrated in magma that had a higher concentration of water than where it first formed. Because relatively high concentrations of water are at the rims, it is possible that the magma did not spend enough time traveling up the volcano to lose a significant amount of water prior to quenching.

25

Winter 2017

Precision Photometry on the SDSS Photometric System Rebecca Chen and Lindsay Berkhout Yerkes Observatory, Williams Bay, WI High precision photometry involves a measurement of the light coming from different sources in a field and a calculation of the difference between their instrumental and apparent magnitudes. This is done using a number of comparison stars with known magnitudes and a source of unknown flux. One then corrects the instrumental magnitudes of the comparison stars to their magnitudes as measured within a specific photometric system, thereby adjusting the unknown sources in the field by the same zeropoint and accurately putting them on the correct scale. This allows the measured flux to be expressed on a physical scale (Watts m2 Hz1 ). This zeropoint correction is determined by the difference in the known stars’ instrumental and known magnitudes. This allows us to isolate the varying source and measure its true variability without interference from conditions that change from night to night. This kind of photometry is important for projects dealing with measurements of fluctuating sources, as it can accurately give a measurement of the change in brightness over time. However, photometry is difficult with high precision and there is therefore a need for a method to accurately obtain these magnitudes on telescopes such as the 24” reflector at Yerkes and the 20” reflector at Stone Edge Observatory, which can both be remotely controlled via the Internet and are frequently used for projects relating to magnitude measurement. Our project pieced together a methodology for using the programs Aperture Photometry Tool and TOPCAT to convert instrumental magnitudes gained from these two telescopes to magnitudes in the Sloan Digital Sky Survey photometric system. Once we resolved how to get these magnitudes, the primary focus was on improving precision. By the end, we achieved less than 5% external precision and less than 3% internal precision on a number of images in multiple filters on the Yerkes 24” telescope and are currently working on doing the same with SEO. These numbers are reasonable for this kind of photometry, but can continue to be improved through different methods of analyzing the data or improved photometric conditions.

Improved Estimation of Atmospheric Particulate Matter in Three U.S. Regions Using Remote Sensing with Public Health Applications

Shane Coffield NASA Marshall Space Flight Center, Earth Sciences Branch, Center for Applied Atmospheric Research and Education Satellite measurements of Aerosol Optical Depth (AOD) have been shown to be correlated with ground measurements of fine particulate matter less than 2.5 microns (PM2.5), which in turn has been linked to respiratory and heart diseases. The strength of the correlation between AOD and PM2.5 varies for different AOD products, geographic regions, and seasons. We evaluated several Moderate Resolution Imaging Spectrometer (MODIS) AOD products from different satellites (Aqua vs. Terra), retrieval algorithms (Dark Target vs. Deep Blue), Collections (5.1 vs. 6) and spatial resolutions (10-km vs. 3-km) for eight cities in the Western, Midwestern and Southeastern United States. We developed and validated PM2.5 prediction models using remotely sensed AOD data along with meteorological variables (temperature, relative humidity, precipitation, wind speed, and wind direction) from the North American Land Data Assimilation System Phase 2 (NLDAS-2). Adding these remotely sensed meteorological data significantly improved the predictive power of the PM2.5 models compared to previous research, especially in the Western U.S. Temperature, relative humidity, and wind speed were the most significant meteorological variables, with wind speed being more significant in the cold season and temperature being more significant in the warm season.

Finally, our study re-establishes the connection between PM2.5 and public health concerns including respiratory and cardiovascular diseases (asthma, high blood pressure, coronary heart disease, heart attack, and stroke).

26

Scientia

Machine Learning at the Energy Frontier: A Signal Processing Approach Towards Finding Dark Matter & Natural Supersymmetry David Jin Fermi National Accelerator Laboratory A search for dark matter is conducted with LHC proton-proton collisions at center-of-mass energy of 13 TeV, using a data set corresponding to an integrated luminosity of 12.88 fb−1, recorded by the CMS experiment. Dark matter candidates are produced in association with a top quark via neutral flavor changing current. We select events containing a single top quark decaying, leptonically produced in association with missing transverse energy. Events are required to have one isolated muon and large missing transverse momentum. The background is estimated using Monte Carlo generated events. Boosted decision trees are trained to distinguish signal several orders of magnitude lower than background. An analytical framework is constructed in the search for monotop dark matter processes within CMS data.

Neutrino Beam Optimization Tyler Johnson Fermilab Deep Underground Neutrino Experiment (DUNE) The Long Baseline Neutrino Facility neutrino beam will be the most intense neutrino beam in the world. The beam is created by colliding high energy protons with a graphite target. Pions and kaons resulting from this collision are focused by a series of magnetic focusing horns. The focused particles decay into neutrinos with an energy spectrum peaked between 1 and 5 GeV. This beam of neutrinos is sent through the Earth to a far detector 1300 km away in South Dakota.

The goal of DUNE is to measure the extent that CP-Violation and the matter effect plays in the neutrino oscillation process for neutrinos and antineutrinos. These goals can only be accomplished if the neutrino beam produces a significant neutrino flux at the far detector. The purpose of this study is to optimize the flux of muon neutrinos received at the far detector by redesigning the focusing and targeting mechanisms. A focusing horn design from previous neutrino experiments is optimized with an algorithm to produce a far more effective design. The resultant design is simple and lacks realistic characteristics like material descriptions and support structures, thus a derivative design is created to be more realistic. This realistic design exhibited a massive flux deficit in comparison to the original optimized design.

This study examines the realistic design to determine why this deficit is produced and establishes numerous strategies for overcoming said deficit while not adversely affecting the CP-sensitivity or violating engineering principles.

27

Winter 2017

An Investigation of Thermotropic Blue Phase and Nematic Liquid Crystals Khia Kurtenbach University of Chicago Institute of Molecular Engineering De Pablo Group Liquid crystals are a phase of matter that flow like liquids but are highly ordered over long ranges. They have applications in fast electro-optic devices, LED displays, three-dimensional lasers, and sensors. In the De Pablo lab, we study the behavior of liquid crystals on different scales, from atomistic levels to mesoscales, to investigate new applications.

Recent work explores the system mixtures and the manipulation of blue phase liquid crystals, as well as the type and anchoring of liquid crystals near surfaces. Blue phases of liquid crystals are interesting states of matter in which an array regular 3-dimensional defects display Bragg’s reflections in the range of visible light. Liquid crystals can be thermotropic, where phase is determined by temperature or lyotropic, with their phases determined by concentration.

Blue phases are thermotropic, occurring only within a specific temperature range between the nematic and isotropic phases of liquid crystals. We focused on studying the blue phases in systems of chiral liquid crystals confined in spherical droplets, as opposed to previous studies which have focused on bulk properties. Our work focused on how the confinement of droplets within polymer films affects the temperature ranges at which the blue phases are stable. We have also focused on studying the effects of mechanical stresses on liquid crystal phase. Our recent studies have explored the effect of various ions on the reorganization of nematic liquid crystals molecules at water interface. Nematic liquid crystals have a fluidity similar to that of most liquids, but they can be aligned using electric or magnetic fields or through surface anchoring. Previous work has shown that liquid crystals align perpendicularly to the N,N-dimethyl-n-octadecyl-3-aminopropyltrimethoxysilyl chloride (DMOAP) treated glass surface. In our experiment, we confined 4-Cyano-4'-pentylbiphenyl liquid crystals, commonly called 5CB, within TEM gold grids on DMOAP-treated glass immersed in water. We explored the effects of ion concentration and time of ion addition on molecular orientation. We found that increasing ion concentration can provisionally and transiently force liquid crystals to move to an isotropic orientation; the effect is modulated by liquid crystal layer thickness and the timescale.

Optical Resonance Imaging: Sub-diffractionlimited Imaging on an Ultrafast Timescale Kirk Lancaster University of Chicago, Department of Chemistry, Engel Lab The resolution of classical imaging systems is limited by the Abbe diffraction limit. Modern optical superresolution techniques, such as STED or PALM, present one way to overcome the diffraction limit by manipulating the fluorescence of a sample. However, as they are not ultrafast techniques, these methods cannot provide information about transient species such as excitons or charge carriers. We present optical resonance imaging (ORI), an optical analog of MRI, to achieve sub-diffraction-limited imaging using ultrashort femtosecond laser pulses. ORI follows the three-pulse sequence of a photon echo experiment, but uses pulsefront tilt (PFT) in the third pulse to couple time to space. The character of this tilted pulse allows different parts of the sample to emit at different times. Interferometric heterodyne detection of the emitted signal with a fixed time reference pulse allows us to generate a spatial image from the arrival time of the signal. By generating PFT without temporal distortion (following geometry in ref. 1), we will be able to preserve the ultrafast nature of the pulse. The apparatus we are constructing will achieve 4-um spatial resolution and 20-fs temporal resolution, which will provide opportunities to understand exciton dynamics at semiconductor interfaces. These have been routinely studied using pump-probe spectroscopy, however they have not been imaged directly because of limits on the spatio-temporal resolution of current techniques. Imaging with ultrafast superresolution techniques like ORI will offer a better understanding of energy transfer and lead to better control over how energy moves through materials.

28

Scientia

Interaction of Ultrafast Terahertz Electromagnetic Pulses with Metamaterials Jonathan Michelsen

Terahertz Science and Technology Group, Department of Photonics Engineering, Technical University of Denmark Ultrafast terahertz (THz) technologies (using light with frequencies of around 300 GHz to 10 THz and picosecond pulse duration) have become widely studied due to their prospective application in ultrafast communication devices and spectroscopy of systems with electronic transitions on the order of milli-electronvolts. Terahertz timedomain spectroscopy (THz-TDS) has been used successfully to examine the characteristics of naturally occurring substances and recently of metamaterials—artificial structures with customizable permeability and permittivity.

Metamaterials consisting of miniature antenna arrays with particular geometries have been found to interact with the electric and/or magnetic fields of linearly-polarized THz pulses. Notably, THz metamaterials have recently been demonstrated to enhance the electric field by a factor of 300, promising to enable sufficient field strengths for nonlinear experiments on systems that only interact moderately with THz radiation. We theoretically and experimentally quantified the interaction of ultrafast THz pulses with various types of metamaterial antennas on silicon substrates positioned perpendicularly to the incoming pulse. When oriented co-linearly to the electric field of the THz pulse, the antennas were shown to behave as resonators, with the electromagnetic pulse inducing a current within each antenna. Furthermore, we placed the previous metamaterial structures into parallel-plate waveguides (PPWG) to increase the wave-matter interaction length. The antennas were found to substantially manipulate the electric field, mainly through the propagation of the fundamental transverse electric (TE1) mode, with the individual antennas coupling to each other within the PPWG. Through the analysis of the metamaterials as thin film layers, we determined their sheet conductivity. Hereby, we can extract essential metamaterial parameters through a transfer function, yielding insight into the magnetic, electric, and optical properties of these metamaterial structures. These parameters will be used for the determination of group and phase velocities of the propagating modes, in order to develop a fuller understanding of the coupling of THz pulses to metamaterial structures that can be used to increase electric fields for future applications in high-power, nonlinear THz-TDS.

A Type-theoretic Foundation for Unsupervised Learning Aaron Geelon So University of Chicago Computer Science Department Learning is often defined as the ability of a human or computer to increase the performance of some task, given more experience or data. In this presentation, we will focus on understanding as a form of learning. To understand a complex system, we rely on a large family of observables to give us measurements about the system. A snapshot of the system is the combination of individual measurements at a single state; a collection of these snapshots can be thought of as a dataset. In the language above, understanding is the ability of a human or computer to describe with greater efficiency the state of a system given a larger dataset. In particular, we think of efficiency in terms of the amount of space required to encode the information about a state. Intuitively, if we can describe the same amount of information with fewer terms, then the information must have had similarities within itself. The redundant parts need not be repeated; instead, they can be abstracted out, so that the type of the term connotes additional information. In fact, we use the words "type" and "term" in a type theoretic sense. This gives us a hint as to why we should look to type theory as a mathematical foundation to learning. There are many feats of machine learning, but they are difficult to understand because of the size and bulkiness of numerical solutions. The goal of creating a type theory for learning is to give humans an algebraic way to reason about learning and to give computers a computable way to learn. For concreteness, I will focus on object recognition in this presentation, and from that, give a possible type theory, justify how a dataset is a family of continuous functions, and how this leads to topologies on types and type formation. Future work will be to continue strengthening the type theory. If successful, besides implementing it, we may be able to draw on other fields of mathematics, such as algebraic topology, category theory, dynamical systems, and so on, to give us more insight on learning.

29

Winter 2017

Surface Modification of Ion Conducting Ceramic Particles Samuel Veroneau IBM Research Laboratory, Almaden This project describes the functionalization of ion conducting ceramic particles for use in battery membranes. Lithium based batteries have long been seen as the future of battery power. Solid state and lithium-air batteries promise higher energy density and higher specific capacity than conventional lithium batteries. However, even the most contemporary lithium based batteries face an inherent shortcoming: they don’t make use of a metallic lithium anode.

This is not without reason, as such anodes promote destructive growths within batteries known as dendrites. Thus, to make use of metallic lithium anodes, dendrite growth must be suppressed. One possible approach to this is through the use of ion conducting ceramic membranes. Such membranes have been shown to significantly curtail dendrite growth while still permitting the flow of ions. However, the production of such membranes is difficult and demanding.

To streamline this process, the ceramic particles embedded in the membrane need to be readily dispersed in a polymer solution. To achieve this, it was hypothesized that the surfaces of ion conducting ceramic particles could be functionalized using small molecules. Lithium and sodium conducting ceramic particles were functionalized by molecules with long carbon chains and various head groups, including carboxylic acids, phosphonic acids, and silanes. Using various techniques, mainly atomic force microscopy (AFM) and x-ray photoelectron spectroscopy (XPS), the surfaces of these ceramic particles were analyzed to determine how, and to what extent, the surfaces of the ceramic particles were functionalized. This project provided an initial understanding to the functionalization of ionconducting ceramic particles, with the hope of implementing the results to create more effective and efficient batteries.

Purification of Multiply Kinked Nanowire by Configuration Dara Weiss University of Chicago/James Frank Institute, Department of Chemistry, Tian Lab The morphology of multiply kinked nanowires (MKNW) has been hypothesized to be greatly useful in biophysics, as well as in other fields. Studies have been done using one-dimensional nanowires to create a resistor-capacitor circuit using Human Umbilical Vein Endothelial Cells as leaky capacitor-bounded resistors. The use of multiply kinked nanowires in such studies can greatly increase even effectiveness of the circuit and synthesis of the device.

However, despite advances in MKNW synthesis and application, the efficiency with which progress is made in these fields can still be greatly increased by successfully separating cis and trans nanowires before further experimentation.

Thus far, advances have been made regarding the purification of rod-shaped and spherical nanoparticles; however, there have been limited to no attempts to purify and separate MKNWs by cis-trans configuration. Here, we attempt to separate the multiply kinked nanowires by configuration using optimized sonication, followed by sucrose gradient centrifugation. We determined that sucrose gradient centrifugation would be useful in separating the MKNWs because trans configured nanowires tend to be less dense than those in cis configurations. First, we ran experiments on straight, one-dimensional nanowires to find the sonication conditions that would yield the longest nanowires on average. Then, we applied those settings to the MKNWs. We then ran trials to determine the sucrose gradient that most effectively separates the cis and trans configured nanowires. We expect this new approach to purification of multiply kinked nanowires to dramatically improve the optimization of future studies and experiments.

30

Scientia

Novel Carbonized Gelatin-Laponite System with Previously Unreported Thermal Stability for Applications in Bone Repair Kylie Zane University of Chicago, Institute for Molecular Engineering, de Pablo Group We are developing a biocompatible and osteoinductive synthetic bone graft scaffold for applications in bone graft surgery. The important considerations for a scaffold candidate include: 1) mechanical strength comparable to endogenous bone, 2) a porous microstructure that allows bone cell adhesion as well as biomineralization and 3) electrical conductivity for the differentiation of bone cells. We are testing a candidate system of carbonized gelatin and laponite. Gelatin and laponite are mixed in water in various ratios (G/L = 1, 2, 5) with low total solid composition (5wt%). These mixtures are tested for injectability and undergo rheological measurements. Next, they are freeze dried at -80 C and then carbonized at 800 C. These samples are imaged with scanning electron microscopy (SEM) to examine pore structure and thermogravimetric analysis to investigate thermal stability. Carbonized low solid weight samples (1.2wt%), prepared by vacuum filtration, were submerged in simulated body fluid to test biomineralization, quantified by Raman spectroscopy, and imaged with SEM. Our samples are currently being tested for bone cell regeneration and electrical conductivity. Thus far, we have found that our samples exhibit high mechanical strength, a desirable pore-like structure suitable for bone cell adhesion and differentiation, successful biomineralization, and sufficient electrical conductivity. Interestingly, our carbonized samples possess higher thermal stability than any value yet reported to our knowledge for carbon or carbon-clay systems, retaining over 90wt% at 800 C. We have created a novel system of carbonized gelatin and clay that shows promise as a synthetic bone graft candidate and previously unheard of thermal stability. To increase its electrical properties, we will proceed with experiments testing systems with laponite and graphene oxide, which can be thermally reduced to graphite after mixing.

31

Winter 2017

Scientia Original Abstracts Vagabond Viruses: Variations in Human Endogenous Retroviral Sequences of HCP5 and Susceptibility to Exogenous Viruses Darian Doakes Arkansas School for Mathematics, Sciences, & Arts, Dr. Patrycja Krakowiak

Human endogenous retroviruses (HERVs) are remnants of retroviral sequences in our genome that, unlike most of the genes that code for proteins, have most likely not been subjected to neg-ative natural selection. HCP5 is a HERV located within the genomic region of the major histo-compatibility complex I (MHC I). Because of the location, genetic variants in this region may be associated with susceptibility to seasonal viruses. Surveys containing questions regarding the frequency and severity of participants’ seasonal illness experiences and buccal samples of mes-enchymal cheek cells in saliva from 53 participants were collected. DNA was isolated and am-plified using a polymerase chain reaction (PCR) and visualized using gel electrophoresis. The first 523 base pairs of the HCP5 gene were sequenced from 49 samples and analyzed for genetic variations. Three loci were identified to be single nucleotide polymorphisms (SNPs). At each polymorphism, participants were either homozygous duplicates for the variant alleles or the typ-ical alleles. Susceptibilities of participants to colds were statistically compared using two sam-ple t-tests at each SNP to determine if there was a trend between varying levels of susceptibility and whether or not a participant was heterozygous, homozygous variant, or homozygous typi-cal. The susceptibility was determined by scoring participants’ answers to the survey, which in-cluded questions prompting severity of cold symptoms and the number of colds caught annual-ly. Each answer per question received a value, and the averages of these values were recorded. None of the SNPs were found to be statistically associated with increasing or decreasing suscep-tibility to colds due to analysis that yielded p-values greater than the predetermined significance level (α-value). Sources of error include how susceptibility scores were calculated from survey averages amongst participants and the small sample size. Adjustments to these sources of error could produce statistical significance. Statistically significant SNPs could aid in improving our understanding of the possible positive selection for HERVs as a means of immune preparation for phylogenetically similar viruses. Providing short sequences containing SNPs that correlate more with lessened susceptibility to cells through nanotechnological advancement could have widespread implications for seasonal health care.

Altering Lentiviral Tropism: Design and Implication of a Targeted Drug-Delivery System

Rachel Gleyzer ergen Countiy Academies Stem Cell Laboratory, Bergen County Academies, Dr. Robert Pergolizzi Despite many promising discoveries in gene therapy, efficient transfer of therapeutic genes remains a major roadblock. Viral vectors are commonly employed, but have drawbacks since delivery is limited by wild-type viruses’ natural tropisms, which are often too broad or too narrow for targeting. To address this hindrance, this project explores a potent technique called pseudotyping, which embeds proteins with an affinity to cells of interest into the lipid envelope of lentiviral vectors. First, vectors were pseudotyped with vesicular stomatitis virus glycoprotein (VSVg) which significantly improved their infection ability (p<.05) and broadened their tropism, but did not provide specificity in targeting. Subsequently, cancer cells with unique receptors were selectively targeted by pseudotyping viral vectors with the unique receptors’ complementary proteins. Pancreatic cancer was chosen as a model system to demonstrate this selective targeting due to its unique expression of cholecystokinin 1 and 2 (CCK1 and CCK2) receptors, which bind molecules of gastrin-17 (GAST17) and cholecystokinin-8 (CCK8). To generate targeted lentiviral vectors with GAST17 and CCK8 (non-membrane bound proteins) in their envelopes, a novel procedure was designed which fuses these proteins to the platelet derived growth factor receptor (PDGFR) transmembrane domain hence embedding them into the lentiviral lipid envelope. An original computer program was also designed to optimize pseudotyping protein sequences. Thus, this project establishes a pioneering and clinically relevant concept and procedure for selectively targeting organs and cancers with personalized therapeutic-delivery tools.

32

Scientia

In Depth Going for The Goal: The Effects of Removing Preparatory Information on The Fast and Unconscious Reading of Action Goals in a Computer-Simulated Competitive Interaction Shira Eisenberg Laboratory of Dr. Ken Nakayama Harvard Vision Sciences Laboratory

Humans are able to quickly and accurately determine intent from early motions of other agents in the social environment.

Through real-time analysis of biomechanical patterns, the mind uses preparatory information to read and project action goals during interpersonal exchanges. Research in psychology, neurophysiology, and behavior provides evidence for a ‘reactive advantage’ in motion dynamics and execution time for reaction versus intentional movement in naturalistic competition.

Though it is widely assumed that computational involvement alters the naturalistic behavioral and perceptual dynamics of human interaction, here I show that the removal of key preparatory information severely impacts reaction time in simulated competition over a computational interface and elicits response patterns consistent with those expected for real-life models.

In this experiment, subjects competed against a computerized opponent in an interactive computer game modeled after a naturalistic reaching task. Each subject used both a mouse and a trackpad and raced their opponent to one of two on-screen targets aligned on either side of the display window. Reaction time, measured as the difference between initial finger launch of the opponent and cursor launch of the participant, was significantly faster in full trials than in cut trials, in which preparatory information was systematically removed. Regardless of input method, these differences in timing are closely correlated to those in naturalistic settings, demonstrating that human perceptual patterns in human-to-human interaction are still present in interactions with computers. This new insight can substantially contribute to designing intuitive interfaces for humancomputer interaction and enhancing models for AI systems in the military and sport.

Introduction Body movement is a chief constituent of human social interaction. Observers in the social environment often rely on external cues like eye movements, facial expressions, and hand gestures to predict others’ emotions, thoughts, and intentions. When consciously perceived, this information is thought to determine an agent’s course of action in social contexts. Some actions, like dodging a bullet on a battlefield or returning a casual handshake, seem intuitive, and are executed

with little to no preliminary thought. Actions following these patterns can be categorized as neither reflexes, because they are executed with intent, nor planned actions, because they occur more suddenly, are often more erratic in execution, and lack an immediate coherence of volition.

Rather, behavioral research suggests that these actions belong to the latter of two distinct classes in human behavior: planned action and reaction to environmental stimuli (Haggard, 2008; Passingham, Bengtsson & Lau, 2010). The first of these classes, planned action, is thought to be controlled via the fronto-median cortex (Krieghoff, Waszak, Prinz &

33

Winter 2017 Brass, 2011). The latter class of action, on which this study is focused, is thought to be controlled by the parietal and premotor cortices (Waszak et al., 2004; Jenkins et al., 2000). That being said, it is possible, the two classes of actions share common central preparatory mechanisms (Hughes, Schutz-Bosbach & Waszak, 2011).

Though little is known about what distinguishes stimulus-driven reactions from voluntary actions, research suggests that fast mechanisms for environmentally stimulated reactions might exist as a means of promoting survival (Welchman, Stanley, Schomers, Miall & Bulthoff, 2010). Additionally, psychophysical evidence suggests that when an external agent’s bodily motion serves as the initial stimulus, preliminary action decisions are made using knowledge of the biomechanical limitations that dictate body movement (Naber et al., 2013). For instance, we might consciously and automatically know to return a high five because our visual system stores experiential information about the string of precursory movements required to initiate the motion, including raising the elbow, lifting the shoulder, flexing the wrist, and straightening the fingers. In close-range combat between armed opponents, for instance, the difference between life and death is around 380 milliseconds (Blair, 2011). In that time, one party, a soldier or officer for example, must notice the enemy begin to raise a firearm and react before the enemy fires.

Since World War II, researchers in military and sport psychology have been seeking ways of understanding this ability to make split-second decisions through perceptual anticipation. Studies of professional and novice athletes in basketball (Wu et al., 2013), tennis (Farrow & Abernethy, 2003; Jackson & Mogan, 2007), badminton (Abernethy & Zawi, 2007), baseball (Ranganathan & Carlton, 2007), and, most notably, soccer (Savelsbergh, Van der Kamp, Williams & Ward, 2005; Diaz, Fajen & Phillips, 2012; Timmis, Turner & van Paridon, 2014) have shown that professional athletes have a distinct ability to forecast the intentions and ensuing movements of their opponents by quickly and effectively analyzing the preperformance cues scattered across their opponents’ bodies (Caserta, 2007). Evidence from weight lifting (Runeson & Frykholm, 1983; Sebanz & Shiffrar, 2009), speech (Abry, Cathiard, Robert-Ribes & Schwartz, 1994; Stork & Hennecke, 1996; Schwartz, Basirat, Ménard

34

& Sato, 2012), handwriting (Orliaguet, Kandel & Boë, 1997; Kandel, Orliaguet & Boë, 2000), and sign language (Pennel, Coello & Orliaguet, 1999) reports similar patterns of goal-anticipation based on earlier motions, suggesting that advanced anticipation of movement is present across human activity.

Most of these studies, however, have focused on a specific demographic and have tested perceptual anticipation in instance-specific paradigms. To broaden the scope of implication, Cormiea, Vaziri Pashkam, & Nakayama (2015) examined realtime action anticipation in a simple reaching task and found a relationship between the presence of preparatory information and the speed of participants’ reaction relative to the start of an opponent’s motion. These findings demonstrate that advanced perceptual mechanisms in everyday people depend on preparatory cues as a means to anticipate and react to the intentions of opponents and social counterparts.

Whereas technology and computer interfaces are constituents of daily life, the impact of perceptual anticipation on behavior in a computational paradigm has yet to be determined. Here, I introduce an approach to examining action anticipation in competitive interaction via a computer interface.

My results are consistent with action anticipation patterns for naturalistic paradigms and thereby suggest that computational involvement does not hinder the mind’s ability to process preparatory information scattered throughout the body as a means to predict intended action before the launch of motion.

To re-create a computerized version of a simple reaching task, I programed an interactive game in which ten subjects were shown 200 video trials broken down into four separate blocks of 50 videos.

In each video trial, a computerized ‘Attacker’ would reach for a target on either the right or left of the screen. The subjects assumed the role of ‘Blocker’ and were tasked with matching the Attacker’s speed and course of motion to the specified target. Two of the video blocks contained preparatory information, or early motion cues. Two did not; in these two, the motion paths preceding the launch of the Attacker’s finger into the air were systematically removed.

I employed a balancing algorithm to ensure that each subject was shown equal numbers of rightreaching and left-reaching trials in a randomized order determined using a Mersenne Twister

Scientia (Matsumoto, 1998). To prevent device bias, subjects completed 4 blocks of edited and unedited trials using each a mouse and a track pad. Cursor location was determined using a unilateral cursor-tracking function.

Reaction times, measured as the difference between finger launch of the computerized Attacker and cursor launch of the Blocker, were faster for blocks of unedited trials, which contained all of the Attacker’s early motions. These results are consistent with results expected for real-life social interaction.

Additionally, differences in reaction time between full and cut conditions were shown to be outwardly imperceptible to the subjects, most of whom were unable to detect disparity between manipulated video trials and untouched video trials. The results of this study show that humans predict intended action from consciously imperceptible preparatory cues and that the removal of these cues causes a decline in speed of performance in a competitive computer-simulated interaction. Furthermore, the reaction speeds with preparatory actions were close to naturalistic speeds, demonstrating that humans perceive computer movement in the same way they perceive human movement.

Stimuli and Procedure

Methods and Materials

Experimental I – Designing Experimental Paradigm:

Participants

The main experiment—the interactive computerized reaching task—was programmed using Matlab and the Psychtoolbox extension. Each subject who played the game was shown 200 video trials, broken down into four separate blocks by condition (CutMouse, CutTrackPad, FullMouse, FullTrackPad) of 50 trials each. In each video trial, a computerized ‘Attacker’ (moving image of lab member) would reach for a target on either the left or right of the screen. Subjects assumed the role of ‘Blocker’ and were tasked with matching Attacker’s speed and course of motion to the target for which they believed the Attacker was aiming. The order in which conditions were presented was randomized using a Latin Square method of randomization (Knuth, 2011). A representation of the described experimental layout is provided (figure 1).

Final sample size of 9 adults (5 women, 4 men), ranging in age from 18 to 35. All Participants volunteered via the Harvard Study Pool website (husp.sona-systems. com) and received their choice of compensation of $10 or course credit. All subjects were right-handed.

All subjects gave informed consent and were tested individually, using the same computer workstation and hardware. All subjects performed trial blocks from all 4 device-condition combinations.

Apparatus

A lab member was videotaped reaching for one of two foam targets on a Plexiglas screen. The camera (GoPro Hero3+ was positioned at eye-level to the lab member. A motion sensor was attached to the lab member’s finger and was used to track horizontal movements of the finger towards either target. 200 trial clips were extracted (100 reaching towards the left target; 100 reaching towards the right target).

Methods

Computational I – Sorting and Arranging Trials: A Matlab program was written to convert each video trial to a matrix and identify each frame in a given trial. The algorithm identified the frame in which motion was initiated, the frame in which the target was hit, and the direction the computerized opponent’s finger moved. An additional Matlab program was written to sort the trials into categories based on the direction of motion and to store each video matrix with either a ‘left’ substring or ‘right’ substring in its file name.

Stimulus generation, subject testing, and data analysis were done on a MacBook Pro 13-in. computer using Matlab and Psychtoolbox software.

A wireless Logitech mouse set to standard sensitivity was utilized for trials with a mouse. The Computational II – Programming Experimental Apple trackpad built into the computer was utilized Paradigm: for trials requiring a trackpad.

Trials were broken up into four blocks based

35

Winter 2017

Figure 1: Experimental design. a) Design showing block figure. Program began with a start screen calling for the participant to press any key to start. Upon key press, the screen transitioned to the first frame of the first trial. The participant (Blocker) would place the cursor in the green calibration square aligned with the Attacker’s finger. Only once the cursor had been placed within the calibration square would the trial begin. During Full trials, the Attacker’s full range of motion would be visible to the Blocker. During Cut trials, the screen would remain on the first frame on the trial until the launch of the Attacker’s finger. At the start of Attacker motion, the Blocker would move the cursor from the calibration square to the target the Blocker perceived as the Attacker’s goal. The selected target would fill in solid blue, when hit by cursor. b) Subject’s view during trial.

on testing condition and device: MouseFull, MouseCut, TrackPadFull, TrackPadCut. Each trial block contained 50 trials. At the beginning of the general experimental code for each trial block, a manual input was given for the subject number, the condition (Cut or Full), and the device (Mouse or TrackPad). Within the experimental code, an algorithm was used to analyze the input and determine if a subject with the given subject number already existed in the database. If so, (i.e. if the subject had already completed >=1 of the 4 trial blocks) the code would create a new csv file for the data for the newest trial block and store it within the pre-existing directory for that subject. If the given subject did not already have a directory (i.e. if this was the first trial block for which the subject was being tested), the code would create a directory for the subject in which all four csv files (for each condition) would be stored. Each csv file contained columns: trialName (name of the video matrix), lor (this would describe the movement) Frame (frame of the trial video at a particular instance of data recording), trialTime (time elapsed since the first instance of the given trial), exptTime (time elapsed since the start of the first trial in the current trial block), mouseX (cursor coordinate of mouse on the horizontal axis), mouseY (cursor coordinate for mouse on the vertical axis), launchB (Boolean to determine whether or not Blocked had launched for the calibration square), targetHit (Boolean to determine whether or not either of the two target squares had been hit by the cursor), launchTimeB (time of Blocker launch—calculated

36

as the first point in time at which cursor velocity with respect to x exceeded a threshold of 500 px/ms for three consecutive readings), mouseDir (direction of motion of mouse), accuracy (Boolean value to determine whether the Blocker had selected the correct target), launchFr (frame of Blocker launch), hitFr (frame Blocker hit one of two targets), waitTime (randomized number of seconds waited before start of each video trial), LaunchTimeA (time of Attacker Launch), RT/n (Reaction time of Blocker, measured as the difference between the launch of the Blocker and the Launch of the Attacker). Throughout the experiment, data was recorded to the csv file in the appropriate column and reset to the next row for each video trial. This was advantageous in data analysis because it allowed data to be called and extracted based on line number and column.

During data analysis, columns were extracted from csv files and converted into .mat files that could be processed and analyzed on Matlab. A Mersenne Twister algorithm was used to randomize the order of trials. A balancing algorithm was used to ensure each subject was shown equal numbers of rightreaching and left-reaching trials. Cursor location was determined using the GetMouse function in the PsychtoolBox extension for Matlab.

Computational III – Isolating and Removing Preparatory Information in Cut Video Trials: Within the code of the general experiment for each trial block, an algorithm was used to manipulate video matrices for trials in the cut condition. For Cut trials, the algorithm was written to systematically remove frames containing preparatory information—the frames between the first frame of the trial and the frame of Attacker launch—and replace them with a replica of the first frame of the trial. Cut trials and Full

Figure 2: Removal of Preparatory Information in Cut Condition Trials For each trial in the Cut Condition, frames between the first frame of the trial and the start of motion were removed and replaced with replicas of the first frame.

trials were therefore the same length of time, and the same number of frames. A visual representation of the information that was removed for these trials is provided (figures 2 and 3).

Scientia Equations Utilized: (1) (2) Figure 3: Schematic of the two experimental conditions.

Trials in the Full Condition contained the Attacker’s movements prior to launch towards a target. In Cut Condition, all frames before the start of the Attacker launch were systematically removed and replaced with a replica of the first frame of the trial. Diagram shows both conditions, broken into frames, from the first frame of a given trial to the frame directly after attacker launch.

Experimental II – Cut Trials vs. Full Trials: To prevent for device bias, subjects were tested on two devices, a mouse and a trackpad, in Full and Cut conditions. In the Full condition, subjects played against unedited trial clips of the Attacker.

In the Cut condition, frames containing preparatory information, or predictive movements between the first frame and the frame of Attacker launch, were removed.

Mathematical I: Reaction Time (RT) was measured and recorded to the appropriate subject’s csv for each trial in each of the four conditions. RT was calculated as the difference in time between the instance of Attacker Launch and the instance of Blocker Launch in milliseconds. Frame rate was 60 fps. See Equations Utilized (1). Attacker Launch Time (Ta) was calculated as the time at which the computerized Attacker’s finger lifted from its initial position.

Blocker Launch Time (Tb) was calculated as the first point in time at which the horizontal velocity (Vx), measured as the approximate derivative by finite differences of consecutive elements in the vector of recorded x-coordinates for the cursor, divided by the approximate derivative by finite differences of consecutive recorded trial times relative to the start of each independent trial, exceeded a threshold of 500 px/ms for a minimum of 3 consecutive readings in a single and accurate direction. See Equations Utilized (2). Trial time was zeroed to account for trials in which participants changed direction partially through motion. Trial Time was zeroed to Attacker Launch Time. See Equations Utilized (3).

mintr < n < max tr

(3)

RT = Reaction Time Tb = Blocker launch time TA = Attacker launch time x = vector of x cursor position readings t = vector of trial time recordings tr = time recording Tzeroed = trial time zeroed to attacker launch F0 = first frame of the trial Fn = final frame of the trial

Results

Accuracy and average Reaction Time (RT) were calculated across accurate trials for each subject and condition. Nine subjects participated and completed 4 blocks of 50 trials each (200 trials per subject), resulting in N = 9. Accuracy was high across conditions (98.66667% ± 0.74811). The difference between average RT in the Full Mouse condition (M= 0.2060, SEM=0.003958) and average RT in the Cut Mouse condition (M= 0.2716, SEM=0.002771) was drastically significant (t(87.72) = 13.59, p<0.0001), as was the difference between RTs in the Full Trackpad condition (M=0.2382, SEM=0.003315) and RTs in the

Figure 4. Average Reaction Time (RT) and Accuracy: a) Reaction time was recorded and plotted in milliseconds for each subject for accurate trials in each condition. Error bars represent standard deviations of the mean. b) Average Accuracy: Percent accuracy of trials in each condition is displayed. Error bars represent deviations of the mean.

N = 9 subjects who performed 50 trials each per condition (200 trials per subject).

Cut Trackpad condition (M=0.3105, SEM=0.002553) across subjects (t(92.00)=17.27, p<0.0001). Although accuracy was highest in the Cut Mouse and Full Trackpad conditions (99.11111% ± 2.6667), there was no significant difference between the mean accuracy in these conditions and the mean accuracy in the other two conditions (p= 0.5772) (figure 4a and 4b).

For visualization and comparison purposes, a graph

37

Winter 2017

Figure 5: Plots of average RT for each subject: Average reaction time (RT) in all 4 conditions was plotted on a separate graph for each subject. RT are only shown for accurate trials.

of average reaction time for accurate trials for each subject in each condition is provided (figure 5).

A simple linear regression was done to determine if participants became faster over the course of time in each of the trial blocks. In all four conditions, Betavalues in the first and second halves of the blocks were not found to be significantly different from zero.

Neither the Full Mouse condition F(1.000,48.00) =

Figure 6. Comparison of average RT per trial (1-50) across subjects for Full Mouse and Cut Mouse conditions: RTs were averaged across subjects for each accurate trial of a given number in CutMouse and FullMouse conditions. A linear regression analysis was conducted for CutMouse and FullMouse conditions. Average RT for given trial number in MouseFull is plotted in red. Average RT for a given trial number in MouseCut is plotted in blue. Error bands (shaded regions) represent standard deviations throughout. Reactions were consistently faster in MouseFull than in MouseCut.

1.314, p = 0.2573, R2 = 0.02665) nor the Cut Mouse condition (F(1.000,48.00), p = 0.0949, R2= 0.5702) exhibited a meaningful regression equation. RTs were fast in the Full Mouse condition from the first trial of the block (~197 ms) and continued to be fast (~205 ms) throughout all fifty trials with no evidence of learning. Likewise, RTs for Full Mouse trials were consistently faster than RTs for Cut Mouse trials throughout (figures 6 and 7).

Every trial for each subject in each condition was analyzed independently and plotted as one of 50 subplots on a grid designated by condition and subject number. This came to 1,800 plots

38

Figure 7: Comparison of average RT per trial (1-50) across subjects for Full TrackPad and Cut TrackPad conditions: RTs were averaged across subjects for each accurate trial of a given number in the Full TrackPad and CutTrackPad conditions. A linear regression analysis was conducted for FullTrackPad and CutTrackPad conditions.

Average RT for a given trial number in the Full condition on trackpad is plotted in red. Average RT for a given trial number in the Cut condition on trackpad is plotted in blue. Error bands (shaded regions) represent standard deviations throughout.

overall. Trial Time, measured as the difference in milliseconds between the time of the sample and the time at which the first frame of a trial was shown, was zeroed as per Equations Utilized (3). In addition to accounting for trials in which participants experienced a change of mind in direction, zeroing the Trial Time to Attacker Launch allowed for a more accurate and direct comparison of RT amongst trials of the same and different conditions in that it identified trials in which brief and accidental cursor motions might have registered as Blocker Launch points. Trials of this sort were analyzed and actual Blocker Launch Times were determined. Trial Time was plotted against X coordinate data of the cursor for each individual trial. A sample grid of a subjectâ&#x20AC;&#x2122;s trials in the Full TrackPad Condition is provided (figure 8).

There was no significant correlation between the accuracy of a trial and the direction of a trial, nor was there significant correlation between the accuracy of a trial and the device on which the trial was run for either the Full conditions (t(15.35)=1.103, p = 0.2869) or the Cut conditions (t(15.58)=0.1907, p = 0.8512). Below is shown the overlapping paths of cursor motion, across subjects, zeroed to Attacker Launch and separated into Mouse and TrackPad trials, respectively (figure 10). Reaction occurs consistently earlier in the Full (red) trials, than in the Cut (blue) trials, by ~70-100ms (p<0.001).

Discussion This experiment sought to determine whether

Scientia

Figure 8: Plots for trials 1-50 for Full TrackPad Condition for Subject 7: Green line represents Blocker Launch. Black line represents Attacker Launch. Y-axis represents x position of cursor. X-axis represents the trial time zeroed to the time of Attacker Launch. Similar grids were generated for every subject in each condition, totaling to 200 plots per subject and 1,800 plots overall. All 1800 plots are included in the supplemental figures section.

Figure 9: Plots of 3 Variations of Sample Trials: a) plot of trial in which Blocker moved directly toward target. Dashed lines represent respective Launch times of Attacker (Ta) and Blocker (Tb). RT was calculated as the difference between the two. b) plot of trial in which Blocker exhibited a change of mind regarding which target to approach.

In order to include trials such as this in analysis, Launch time of Blocker (Tb) was determined to be the point in time at which the Blocker’s horizontal distance from the Calibration Square in the direction of the accurate target reached its absolute minimum for the trial. c) plot of trial in which Blocker exhibited premature cursor movements. In order to include trials such as this in analysis, Launch Time of Blocker (Tb) was calculated as the first point in time at which the velocity of cursor motion in the direction of the accurate target exceeded a threshold of 500 px/ms. A line representing the Attacker’s course of motion is included in each graph to allow for frame of reference. For uniformity and to facilitate more direct comparison between circumstances expressed in each figure, recorded X coordinate values have been converted to a vector in pixels of distance traveled from the calibration square toward the accurate target at any given point. This has been plotted against Trial Time Zeroed.

humans depend on preparatory cues to predict intent in reactions on the human-computer interface.

Figure 10: Plots of overlapping paths of cursor motion on X position vs.

Time, zeroed to Attacker Launch: dark blue indicates an accurate Cut trial; Cerulean indicates an inaccurate Cut trial; Red indicates an accurate Full trial; Magenta indicates an inaccurate Full trial. Plots are separated by device with one plot for Mouse trials and one plot for TrackPad Trials.

To test this, subjects assumed the role of the Blocker in a computer-simulated version of a natural reaching paradigm task tested in a prior study (Cormiea et al., 2015). Over the course of 200 trials, each subject raced a computerized Attacker to one of two targets in a variation of a prototypical choice reaction time test. Subjects’ reactions were quick in both conditions but showed a marked advantage in the Full conditions, suggesting subjects relied on preparatory cues before the start of the Attacker’s movement towards a target to predict intent.

The reaction times of participants are consistent with those in a previous study, in which the removal of preparatory cues in a naturalistic paradigm resulted in a ~100 ms decrease in reaction speed (Cormiea et al., 2015). Trends in motion dynamics and reaction speed were similar as well. Taken together, this suggests that perceptual dependence on motion cues occurs not only in human-to-human interactions, but also in interactions between human and machine.

Accuracy was near perfect across conditions (98.66667% ± 0.74811), higher than the accuracy reported for the naturalistic task that this experiment sought to mimic computationally (92.7% ± 6.29).

Likewise, reaction times were significantly faster in Full conditions than in Cut and were similar across devices. Reaction times were surprisingly faster than those reported in comparable two-choice reaction tasks. In such tasks, reaction times should follow Hick’s Law, which states reaction time would be proportionate to Log(N), in which N is the number of possible choices in the task (Hick, 1952). For this experiment, Log(2) would yield a reaction time of approximately 301 ms. While this is comparable to

39

Winter 2017 participants’ average reaction time in Cut conditions (~300 ms), it is approximately 100 ms slower than average reaction time observed in Full conditions (~200 ms). In addition, the results exhibit no sign of motion advantage in any particular direction, which would be expected based on literature values.

Recent studies have reported results that suggest that the direction of cursor motion has significant effects on kinematic properties such as average velocity, movement time, movement unit, and peak velocity, with leftward motion showing more favorable results than rightward motion (Meng et al., 2012). In this experiment, however, no such results were found. In fact, no significant correlation was found between the direction in which the Attacker reached and the accuracy of the trial. Likewise, there was no significant correlation between the accuracy of a trial and the device on which the trial was performed for either the Full or Cut conditions.

This could be a result of the fact that this experiment focused on a competitive and dynamic interactive paradigm with a simulated opponent and thus drew upon perceptual patterns of action anticipation previously reported only in physical and naturalistic settings.

A notable finding of this study was that reaction times remained fast throughout the experiment.

Subjects did not display signs of learning throughout trial blocks. This contrasts the results of a similar study on the human-computer interface, in which significant improvements in subject reaction throughout the course of trial blocks were noted (Diaz et al., 2012). The difference in results is likely because the aforementioned study lacked the characteristics of a naturalistic interaction. It used point lighters as stimuli rather than a computerized opponent with a human appearance and displayed only the preparatory information, requiring subjects to identify goals based on this information. As noted in a different study, in naturalistic paradigm interactions, humans do not report predictions explicitly. The process of action anticipation in a naturalistic paradigm is implicit and automatic (Cormiea et al., 2015). Rather, the results of this experiment were more consistent with the results of experiments in the naturalistic paradigm. That is to say, although this experiment examined interactions between a human subject and a computerized attacker, the human’s reaction times mimicked those expected for an interaction with another human

40

being. One possible explanation for this is the competitive nature of the reaching task (Georgiou et al., 2007). The demands of competing against an opponent with the image of a human might impose the same social urgency of a competitive interaction with a human opponent.

In addition, the results reported in this study are in line with previous scientific literature in sports and military, fields in which split-second reactions to environmental stimuli dictate success and, in some cases, survival. Professional athletes, for instance, have been shown to possess a distinct ability to forecast the intentions and ensuing movements of their opponents by quickly and accurately analyzing pre-motor cues scattered across their opponents’ bodies (Caserta, 2007). Evidence of similar perceptual mechanisms of early response anticipation has been found in basketball (Wu et al., 2013), tennis (Farrow & Abernethy, 2003; Jackson & Mogan, 2007), badminton (Abernethy & Zawi, 2007), baseball (Ranganathan & Carlton, 2007), soccer (Savelsbergh, Van der Kamp, Williams & Ward, 2005; Diaz, Fajen & Phillips, 2012; Timmis, Turner & van Paridon, 2014), weight lifting (Runeson & Frykholm, 1983; Sebanz & Shiffrar, 2009), speech (Abry, Cathiard, Robert-Ribes & Schwartz, 1994; Stork & Hennecke, 1996; Schwartz, Basirat, Ménard & Sato, 2012), handwriting (Orliaguet, Kandel & Boë, 1997; Kandel, Orliaguet & Boë, 2000), and sign language (Pennel, Coello & Orliaguet, 1999). This indicates that advanced evaluation of precursory movement dictates reactive response and advantage across human activity and engagement. The findings of this study extend these results and demonstrate that perceptual reading and response to distributed bodily motion cues in the moments prior to the physical initiation of action extends beyond the naturalistic paradigm and into the computational.

These results indicate that humans rely on implicit and frequently imperceptible, action cues of others to quickly and accurately anticipate intent, not only in human-to-human interactions, but also across the human-machine interface. They are consistent with anatomical models of movement and demonstrate that humans rely on motion cues prior to others' actions to predict their goals, further validating the putative link between biomechanics of human action and sensory motor perception.

This knowledge, if combined with machine learning techniques, could revolutionize perspectives

Scientia and theories on technological design for human computer interaction.

The findings of this experiment substantially challenge the accepted paradigm of scientific perceptions regarding the human-machine interface.

In order to design and manufacture technologies capable of improving the quality of human life through intimate assistance and interaction, we must first understand the mechanisms of human interaction, as they exist in both naturalistic paradigms and computational interactions. The results of this study provide potential evidence for a number of theories regarding the nature of human reaction and behavioral responses to environmental stimuli. Further research into this could have significant implications in the future of designing technology for intuitive use as well as in the potential creation of artificially intelligent machines. Furthermore, the study of the link between biomechanics and perception could aid in understanding the anticipatory aspects of competition in real-time.

References [1]

Abernethy, B., & Zawi, K. (2007). Pickup of Essential Kinematics Underpins Expert Perception of Movement Patterns. Journal of Motor Behavior, 39(5), 353-367. doi:10.3200/jmbr.39.5.353-368

[2] Abry, C., Cathiard, M., Robert-Ribes, J., & Schwartz, J. (1994). The coherence of speech in audiovisual integration. Current Psychology Of Cognition, 13(1), 52-59.

[3] Blair, J. P., Pollock, J., Montague, D., Nichols, T., Curnutt, J., & Burns, D.

(2011). Reasonableness and reaction time. Police Quarterly, 14, 323-343.

doi:10.1177/1098611111423737 [4] Caserta, R. (2007). Multidimensional Perceptual-Cognitive Skills Training Reduces

Spatio-Temporal Errors And Improves Dynamic Performance Among Seniors

(Ph. D). University of Florida.

[5] Cormiea, S., Vaziri Pashkam, M., & Nakayama, K. (2015). Unconscious perception of an opponent's goal. Presentation, Banyan Breezeway Session: Perception and Action: Complex interactions, VSS Conference at Harvard University.

[6] Diaz, G., Fajen, B., & Phillips, F. (2012). Anticipation from biological motion: The goalkeeper problem. Journal Of Experimental Psychology: Human Perception And Performance, 38(4), 848-864. doi:10.1037/a0026962 [7] Dillen H., Phillips J.G., Meehan J.W. Kinematic analysis of cursor trajectories controlled with a touchpad. Int. J. Hum.-Comput. Int. 2005;19:223–239 [8] Georgiou, I., Becchio, C., Glover, S., & Castiello, U. (2007). Different action patterns for cooperative and competitive behaviour. Cognition, 102(3), 415-433.

[9] Farrow, D., & Abernethy, B. (2003). Do expertise and the degree of perception - action coupling affect natural anticipatory performance?.

Perception, 32(9), 1127-1139. doi:10.1068/p3323

[10] Haggard, P. (2008). Human volition: towards a neuroscience of will.

Nature Reviews Neuroscience,9(12), 934-946. doi:10.1038/nrn2497 [11] Hick, W. E. 1952. On the rate of gain of information. Quarterly Journal of Experimental Psychology, 4:11-26.

[12] Hughes, G., Schutz-Bosbach, S., & Waszak, F. (2011). One Action System or Two? Evidence for Common Central Preparatory Mechanisms in Voluntary and Stimulus-Driven Actions.

Journal Of Neuroscience, 31(46), 16692-16699. doi:10.1523/ jneurosci.2256-11.2011 [13] Jackson, R., & Mogan, P. (2007). Advance Visual Information, Awareness, and Anticipation Skill. Journal Of Motor Behavior, 39(5), 341-351. doi:10.3200/jmbr.39.5.341-352 [14] Jenkins, I. (2000). Self-initiated versus externally triggered movements: II. The effect of movement predictability on regional cerebral blood flow. Brain, 123(6), 1216-1228. doi:10.1093/ brain/123.6.1216 [15] Johansson, G. (1973). Visual perception of biological motion and a model for its analysis. Perception & Psychophysics, 14(2), 201-211. doi:10.3758/bf03212378 [16] Kandel, S., Orliaguet, J., & Boë, L. (2000). Detecting anticipatory events in handwriting movements. Perception, 29(8), 953-964. doi:10.1068/p2864 [17] Knuth, D. (2011). The art of computer programming. Upper Saddle River, NJ: Addison-Wesley.

[18] Krieghoff, V., Waszak, F., Prinz, W., & Brass, M. (2011). Neural and behavioral correlates of intentional actions. Neuropsychologia, 49(5), 767-776. doi:10.1016/j.neuropsychologia.2011.01.025 [19] Kuhl, P., Conboy, B., Padden, D., Nelson, T., & Pruitt, J. (2005). Early Speech Perception and Later Language Development: Implications for the "Critical Period". Language Learning And Development, 1(3-4), 237-264. doi:10.1080/15475441.2005.967194819 [20] LIBET, B., GLEASON, C., WRIGHT, E., & PEARL, D. (1983). TIME OF CONSCIOUS INTENTION TO ACT IN RELATION TO ONSET OF CEREBRAL ACTIVITY (READINESS-POTENTIAL). Brain, 106(3), 623642. doi:10.1093/brain/106.3.623 [21] Matsumoto, M., & Nishimura, T. (1998). Mersenne twister: A 623-dimensionally equidistributed uniform pseudo-random number generator. ACM Trans. Model. Comput. Simul. TOMACS ACM Transactions on Modeling and Computer Simulation, 3-30.

[22] Meng, L., Chen, H., Lu, C., Chen, M., & Chu, C. (2012). The Effect of Direction on Cursor Moving Kinematics. Sensors, 1919-1929.

[23] Naber, M., Vaziri Pashkam, M., & Nakayama, K. (2013). Unintended imitation affects success in a competitive game. Proceedings of the National Academy of Sciences of the United States of America, 110(50), 20046–20050. http://doi.org/10.1073/pnas.1305996110 [24] Orliaguet, J., Kandel, S., & Boë, L. (1997). Visual perception of motor anticipation in cursive handwriting: Influence of spatial and movement information on the prediction of forthcoming letters. Perception, 26(7), 905-912. doi:10.1068/p260905 [25] Passingham, R., Bengtsson, S., & Lau, H. (2010). Medial frontal cortex: from self-generated action to reflection on one's own performance. Trends In Cognitive Sciences, 14(1), 16-21. doi:10.1016/j. tics.2009.11.001 [26] Pennel, I., Coello, Y., & Orliaguet, J. (1999). Visual Perception of Motor Anticipation in Sign Language. Studies In Perception And Action V, 144-148.

[27] Phillips, J., & Triggs, T. (2001). Characteristics of cursor trajectories controlled by the computer mouse. Ergonomics, 527-536.

[28] Phillips J.G., Meehan J.W., Triggs T.J. Effects of cursor orientation and required precision on positioning movements on computer screens.

Int. J. Hum. Comput. Interact. 2003;15:379–389.

[29] Ranganathan, R., & Carlton, L. (2007). Perception-Action Coupling and Anticipatory Performance in Baseball Batting. Journal Of Motor Behavior, 39(5), 369-380. doi:10.3200/jmbr.39.5.369-380

41

Winter 2017 [30] Runeson, S., & Frykholm, G. (1983). Kinematic specification of dynamics as an informational basis for person-and-action perception: Expectation, gender recognition, and deceptive intention.

Journal Of Experimental Psychology: General, 112(4), 585-615. doi:10.1037//0096-3445.112.4.585 [31] Savelsbergh, G., Van der Kamp, J., Williams, A., & Ward, P. (2005).

Anticipation and visual search behaviour in expert soccer goalkeepers.

Ergonomics, 48(11-14), 1686-1697. doi:10.1080/00140130500101346 [32] Schwartz, J., Basirat, A., MĂŠnard, L., & Sato, M. (2012). The Perception-for-Action-Control Theory (PACT): A perceptuo-motor theory of speech perception. Journal Of Neurolinguistics, 25(5), 336354. doi:10.1016/j.jneuroling.2009.12.004 [33] Sebanz, N., & Shiffrar, M. (2009). Detecting deception in a bluffing body: The role of expertise. Psychonomic Bulletin & Review, 16(1), 170175. doi:10.3758/pbr.16.1.170 [34] Stork, D., & Hennecke, M. (1996). Speechreading by humans and machines. 173-191 Berlin: Springer. doi: 10.1007/978-3-662-13015-5.

[35] Timmis, M., Turner, K., & van Paridon, K. (2014). Visual Search Strategies of Soccer Players Executing a Power vs. Placement Penalty Kick. Plos ONE, 9(12), e115179. doi:10.1371/journal.pone.0115179 [36] Waszak, F., Wascher, E., Keller, P., Koch, I., Aschersleben, G., Rosenbaum, D., & Prinz, W. (2004).

[37] Intention-based and stimulus-based mechanisms in action selection.

Exp Brain Res, 162(3), 346-356. doi:10.1007/s00221-004-2183-820 [38] Welchman, A., Stanley, J., Schomers, M., Miall, R., & Bulthoff, H. (2010). The quick and the dead: when reaction beats intention.

Proceedings Of The Royal Society B: Biological Sciences, 277 (1688), 1667-1674. doi:10.1098/rspb.2009.2123 [39] Wu, Y., Zeng, Y., Zhang, L., Wang, S., Wang, D., & Tan, X. et al.

(2013). The role of visual perception in action anticipation in basketball athletes. Neuroscience, 237, 29-41. doi:10.1016/j. neuroscience.2013.01.048

About the Author Shira is a first-year student at the University of Chicago majoring in Computer Science and minoring in Computational Neuroscience. She hopes to earn a Ph.D in Computer Science or a related field in the future.

42

Scientia

The Effect of Electromagnetic Fields on the Arion distinctus in the Presence of Various Wavelengths of Light.

Sofia Garrick Archer School For Girls

This research has been academically published by the European Journal of Biotechnology and Bioscience.

For more information on this project and official publication details, please contact Sofia Garrick through email (sgarrick@uchicago.edu).

Electromagnetic sensing ability that is found in many different vertebrates and invertebrates is largely mediated through the opioid receptors and internal opioid peptides. This sensing ability has been shown to be light dependent, and is therefore likely dependent on a photoreceptor. Most of this type of research in invertebrates has been conducted using Cepaea nemoralis (the land snail), but this sensing ability has yet to be explored in shell-less gastropods or slugs. We tested the effect of two different wavelengths of light (red light ­650nm and blue light­450nm) on the electromagnetic sensing abilities in the Arion distinctus (the garden slug). This experiment was performed by injecting the slugs with an enkephalinase inhibitor, followed by a 15min exposure to a ±60μT magnetic field in the presence of either red light, blue light, or white light. This experiment demonstrated for the first time that an electromagnetic sensing ability is present in the Arion distinctus and is light sensitive. Blue light was the most effective light setting for slugs to sense an electromagnetic field.

Introduction Although humans are the most advanced mammals, they lack a sixth sense that is found in many different vertebrates and invertebrates.

Unlike humans, other animals such as sharks, birds, rats, salamanders, flies, fish and snails possess a sixth sense of electromagnetic sensing (1). Electromagnetic sensing is used for various purposes including migration, hunting, defense, and possibly pain management. For example, homing pigeons use the earth’s electromagnetic field (EMF) as their ‘compass’ to determine their location or estimate how far they are from a specific goal (2).

This electromagnetic sensing ability is mediated through opioid receptors, which also use endogenous opioids to control the feeling of pain (3). In order to relieve pain, endogenous opioids are released by the nervous system when the body experiences external pain sensations. However, the electromagnetic field negates the effects of an endogenous opioid, a natural painkiller, and makes the subject feel even more pain by distorting the opioid receptors and rendering them unable to accept opioids (3,4). Thus, in order to test whether an electromagnetic field is affecting an animal, a pain reaction is measured. If the animal has been exposed to an electromagnetic field, the reaction to

43

Winter 2017 a pain stimulus will be more extreme.

Previous studies demonstrate a connection between exposure to an EMF and reaction to a pain stimulus. These studies showed that 15 minute exposures to extremely low frequency ((ELF) magnetic fields increase the amount of pain experienced in snails. In these same studies, the injection of an enkephalinase inhibitor, which prevents the breakdown of the opioid encephalin, decreased the pain experienced by the snails (5). Enkephalinase inhibitors (e.g. racecadotril) increase the amount of enkephalins, enhancing the analgesic effect and acting similar to a painkiller (6). Since an ELF electromagnetic field makes the opioid receptors less able to receive opioids, such as enkephalin, the snails experienced more pain after being placed in the electromagnetic field (5).

The opposing interaction between the EMF and the enkephalinase inhibitor on pain demonstrates that the EMF sensing ability of the snail is opiate mediated.

Researchers have also tested how different conditions during exposure to an electromagnetic field affect the snails’ sensitivity to the EMF induced pain. Snails have been shown to react differently to the electromagnetic fields when they are exposed in lighted conditions versus dark conditions. When snails were placed in a magnetic field with light exposure, the snails reacted faster to the pain stimulus because the electromagnetic field affected more opioid receptors making them less able to receive the enkephalins, which would normally inhibit pain (7). The magnetic field’s orientation, in relation to the placement of the snail’s body, affects the degree of pain noted (8). Much of the previous research on the sensing ability of electromagnetic fields and their relation to pain has been conducted on snails, but none appears to have been done with slugs. Although slugs are in the same phylum as snails (mollusca), they have evolved into different species (9). Because of their close taxonomic relationship to snails, slugs are predicted to behave in a similar manner when placed in an electromagnetic field.

One goal of this research is to compare the ELF sensing abilities in closely related species to further understand if evolution has affected this sensing ability.

It is important to test various wavelengths of light given the wide range of animals that have an EMF sensing ability, as different animals in

44

different environments are exposed to different wavelengths of light. In addition, the light pollution from cities along the coast could be spreading into the ocean and affecting the EMF sensing abilities of marine animals, as well as cities around the country interfering with the EMF sensing abilities of land animals. The red wavelength chosen for this experiment was chosen because in previous research (Garrick, S. 2015, unpublished preliminary data) red was shown to be the least effective wavelength for enhancing pain in the slugs. The blue wavelength was chosen because sharks and other ocean creatures experience blue wavelengths more frequently than other wavelengths as the water filters out other colors. This experiment tests the differing effects of red light (730nm), blue light (455nm), and broad spectrum white light on the slug’s response to pain under each light condition.

This research is significant because very little is known about the relationship of electromagnetic sensing in invertebrates and how it is affected by light (10). In addition, all the research on invertebrates has been done using a different species than slugs.

This information could give us more insight into a sixth sense that many animals possess and how this natural interaction of light and electromagnetic fields can be affected by human intervention, either deliberate or accidental.

Methods

Helmholtz Coil Construction The electromagnetic field was created using a Helmholtz Coil, built and designed based on the previous research by Kirschvink (11). A Helmholtz Coil consists of two to four solenoid magnets of the same size and placed on the same axis, creating a uniform electromagnetic field. The Merritt et. al coil was chosen because it had the largest area of constant magnetic field strength in the center of its four coils, according to research done by Kirschvink who previously evaluated various coils and their corresponding field areas. The coils were built using a laser cutter to shape the acrylic support structure and then each was wrapped in magnet wire once the structures were finalized. The coils generate a constant ±60μT electromagnetic field, which is comparable to the Earth’s natural magnetic field.

The Helmholtz coil had modifications in order to

Scientia accommodate both snails and slugs. Two trays were designed for holding the slugs and snails: one that was able to keep the snails contained while also accommodating for the size of their shells and one with a smaller cavity for the slugs (Figure 1).

garden snails (Niles Biological, Sacramento, CA).

Slugs varied in weight from 0.08 grams to 0.8 grams.

As previous research has been published using snails as subjects, slugs served as the experimental group and snails as the control. The slugs and snails all were kept in similar conditions: small terrarium, dirt covered base, plenty of vegetables to eat, a wet paper towel covering the bottom, and an automatic mister fogging up the terrariums to keep the environments damp and cool.

Core Assays

Figure 1. Power source and Helmholtz Coil generating ±60μT electromagnetic field.

Light Box Construction Two light boxes were constructed for experimental trials. Holes were cut into the top and sides of the box where 10 blue or red LEDs were evenly placed. The LEDs were soldered together on top of the box and connected to a power source to create a uniform intensity of light (LED World 2.5V 3W red high power LED 620-630nm 20mm star base 700mA; 460-470nm 3.5V 3W, blue). The Helmholtz coil was then placed on the platform with the box fitting over it. Experiments with slugs or snails not exposed to single wavelength light as a control were peformed in standard fluorescent lighting.

Slug and Snail Core Conditions The research subjects are brown slugs and

For the core assays, a group of twelve snails and fifteen slugs were placed in white light with four different conditions: 1) slugs without racecadotril solution and with sham EMF, 2) slugs and snails injected with racecadotril (R; 0.1μg racecadotril/2μl DMSO) solution (Sigma Aldrich, dissolved with DMSO) followed by sham EMF; 3) slugs and snails not injected with racecadotril solution and placed in a ±60μT EMF; and 4) slugs and snails injected with racecadotril solution (0.1μg racecadotril/2μl DMSO) and placed in a ±60μT EMF. Snails and slugs were injected with this solution because the impact of the EMF is supposed to negate the results of the solution, which allows the effects of the EMF to be amplified. After each trial the slugs and snails were placed on a hotplate set at 40+/- oC and their nociceptive response was measured. The pain response was measured based on the amount of time (measured by a stopwatch) the snails or slugs could endure on the hotplate before lifting their “foot” reflecting pain. They were then immediately removed from the hot plate and placed back in their terrarium (8). The outcome measure was time to lift their foot once placed on the 40+/- oC hot plate.

Wavelength Groups There were two different experimental trial groups of slugs: observed and tested at red 650nm and blue - 455nm. Each slug was injected with a Racecadotril solution and then placed in an electromagnetic field set at ±60μT for 15 minutes exposed to their assigned wavelength color. Once they were removed from the electromagnetic field, they were placed on a hotplate set at 40+/- oC and their nociceptive response was measured.

45

Winter 2017 Statistics: All of the statistical testing was done using the “R” statistics application, through the “R Studio” (RStudio, Boston, Massachusetts) interface. For each set of groups, individual data vectors were created.

These vectors were then used with the boxplot() function to create boxplots for the purpose of identifying the presence of any outliers. Each vector was then run through the qqnorm() function to produce a normal probability plot to visually check that the data appeared to be normally distributed.

As an additional check, each data vector was also run through the shapiro.test() function to perform a Shapiro-Wilk test of normality. The data vectors for each group were then combined into a data frame.

The function LeveneTest() was then used to perform Levene’s test for homogeneity on the groups in the set.

These tests were all performed for the purpose of ensuring that the data met the underlying assumptions for an analysis of variance test. In the case of both sets of slug data, the group variances were found to be non-homogeneous. As a result, an ANOVA test would not be appropriate. Instead, Welch’s test is used to compare the group means.

This was done in R using the function oneway.test().

Finally, pairwise post-hoc testing was completed for the groups in each set by using the oneway.test() function in conjunction with the apply() function to produce a table of p-values for each comparison.

Figure 2. The mean amount of time (sec) for each slug group to respond to the pain stimulus under different conditions +/- 95% Confidence. R = Racecadotril preinjection. EMF = Electromagnetic Field. Compared to no EMF and no R (Group 1), the enkephalin inhibitor increased amount of time to pain response. EMF increased the response to pain, ie. shorten the time to pain response (Group 3 vs 4, and Group 1 vs 3). EMF without R had the shortest time to pain response.

The control trials with the snails using the same groups as the slugs, show very similar results (Figure 3). From this collection, the group with the shortest response to the pain stimulus was the group with only EMF (5.63 +/- 0.77s) in comparison to no EMF and no racecadotril (11.25 +/- 1.24s). The second shortest was with racecadotril and EMF, with neither EMF or racecadotril following closely behind (11.25 +/- 1.24s). Finally, the group with only racecadotril (17.13 +/- 1.7s) had the longest delay before the subjects showed signs of pain.

Results Slugs injected with racecadotril (R) and not exposed to EMF took longer to react (13.08 +/1.18s) to a pain stimulus than slugs injected with racecadotril and exposed to EMF (6.8 +/- 0.43s) (Figure 2). The slugs exposed only to EMF, without injections of racecadotril, had the shortest reaction time (5.2 +/- 0.73s) compared to the slugs tested for their natural reaction to the pain stimulus (10.67 +/1/19s). Each of the control trials were conducted in white/fluorescent light (Figure 2).

Welch’s test on the 4 groups are: One-way analysis of means, not assuming equal variances (F=41.7333, Df 3, Denominator df = 23.6672) p<0.05. Pair-wise post-hoc testing revealed all 4 groups were distinct from each of the other 3 (p< 0.05).

46

Figure 3. The amount of time (sec) it took for each snail group to respond to the pain stimulus under different conditions +/- 95% Confidence. R = Racecadotril preinjection. EMF = Electromagnetic Field. Compared to no EMF and no R (Group 1), the enkephalin inhibitor, increased the amount of time to pain response(Group 4).

EMF increased the response to pain, ie. shortened the time to pain response (Group 2 vs 4, and Group 1 vs 3). EMF without R had the shortest time to pain response (Group 3).

Welch’s test on the 4 groups are: One-way analysis of means, not assuming equal variances (F=75.548, Df = 3, Denominator df = 15.022) p<0.05.

Pair-wise post-hoc testing revealed all groups are

Scientia seen as being different from one another, since the p-values are all less than 0.05.

When the slugs were placed in the electromagnetic field in the presence of blue light, their reaction to the heat was faster than the control (4.24 +/- 0.54 vs 6.8 +/- 0.43) (Figure 4). When the slugs were placed in the electromagnetic field in the presence of red light, their nociceptive response was delayed significantly (11.12 +/- 1.02s vs 6.8 +/- 0.43s).

Figure 4 shows the time recorded (+/- 95% confidence) for each trial’s slugs’ nociceptive responses after being injected with racecadetril and then placed in the electromagnetic field exposed to different wavelengths of light. In addition, the control most similar to the experimental trial is added as white+R+EMF.

The results of Welch’s test are based on a oneanalysis of means between the three groups that does not assume equal variances (F=84.672, Df=2, Denominator df=29.072, p< 0.05). Pair-wise posthoc testing was performed to reveal distinct groups (p<0.05).

Discussion These results demonstrate for the first time that the species, Arion distinctus, has an EMF sensing ability. In previous studies (e.g. 3,5,7, 8), the electromagnetic sensing ability has been evaluated in a species of snails that were closely related to the slugs evaluated in this study. Since slugs and snails have a similar evolutionary ancestry, these findings suggest that this sensing ability could also be found in the evolutionarily more recent species of garden slugs.

Comparisons of the four slug control groups, all observed under white light, demonstrate that each group is significantly different from the others.

These results demonstrate strong support for an

EMF sensing ability in slugs that is largely opiate mediated. This work confirms and repeats the prior research conducted in snails that concluded opiate mediated EMF sensing ability (3,5,7, 8).

This study also demonstrates that blue light appears to heighten the slugs’ response to the electromagnetic field (i.e. faster time to demonstrate pain) compared to any other group. The slugs tested in the red light, on the other hand, had a delayed reaction to the hot plate when compared to tests done under control white lighting. The electromagnetic field in white light was associated with the slugs reacting faster to the pain stimulus compared to slower response in red light. These findings extend prior research suggesting the slugs’ sensing of the electromagnetic field involves both light and opiate-mediated mechanisms (7).

The slug core assay group using racecadotril and exposed to EMF is the most comparable to the experimental trials exposed to different wavelengths of light, as all of those trials contained slugs injected with racecadotril and also subjected to the electromagnetic field. Each of the experimental trials included racecadotril because the EMF is theorized to nullify the pain-suppressing effects of the racecadotril. For the control group with racecadotril and with the EMF in white light, the average time for the slugs to respond to pain was 6.8 seconds (+/- 0.43 s), whereas for the red light plus EMF it was 11.12 seconds (+/- 1.02s) (Figure 4). The red light could have negated the effects of electromagnetic exposure because the response time to the pain stimulus is close to the response time for the control trial of slugs just injected with racecadotril (11.2 +/- 1.02s vs 13.08 +/- 1.18s) (Figure 2). Since the blue light demonstrates the slugs’ shortest reaction to the hot plate of any group at 4.24 seconds (+/- 0.54s), the blue light was found to be the most effective environment for the slugs in order to sense an electromagnetic field.

A limitation in this study is the comparison of an earlier performed control group (white light) with later experimental groups (red and blue light).

However, the standard errors were low and the groups were highly significantly different from each other. The most important factor that could have contributed to potential bias in this study would be the measurement of each slug’s or snail’s response to the hot plate test following experimental procedures did not have a blinded rater. This could be minimized

47

Winter 2017 in future studies using a researcher blinded to the solutions injected or the EMF exposures previously used in each experimental subject.

Several other variables may have affected the results in this study. In previous studies testing the electromagnetic sensing abilities of snails, the Enkephalinase inhibitor used was SCH 34826. In this study racecadotril was used, as the Enkephalinase inhibitor SCH 34826 is no longer manufactured.

These substances are both Enkephalinase inhibitors and, in these studies, accomplish the same biological effects. However, since racecadotril is insoluble in water and saline solution it was instead dissolved in DMSO. The DMSO could have had some effect on the slugsâ&#x20AC;&#x2122; response to the electromagnetic field and/or the hot plate.This variable most likely did not have a large effect, as the core trials with snails reacted similarly to previous studies (3,5,7,8).

Another factor that may have affected variation in response to the pain stimulus is the difference in the weight of each slug. Slugs varied in weight and some slugs appeared to have higher pain tolerances than others, which would vary responses to. The heaviest slug weighed 0.8 grams while the lightest slug weighted 0.08 grams. This discrepancy in weight could have impacted the amount of racecadotril absorbed by each slug. This potentially increased the variability of the results and could have affected the data collected for certain wavelengths depending on which randomly selected slugs were used for each trial. Future studies could more carefully track the weight of each slug per trial and how this may have affected their response to pain.

Future research addressing the effects of different wavelengths of light on the ability to sense an electromagnetic field could consider more wavelengths of light and dark (7) in order to fully understand the photoreceptive qualities of this sensing ability. Furthermore, based on the results of this study, it will be useful to investigate how blue light relates to the electromagnetic abilities of marine animals since many marine animals (such as sharks) have electromagnetic sensing abilities and blue light is the most penetrant in water. The slugs responded better to blue light compared to red or white light. Finally, in order to apply this research to supplement environmental change, it would be beneficial to see how light pollution, as a whole and of different wavelengths of light, affect animals electromagnetic sensing abilities in different

48

regions of the world. The effects of artificial light for the benefit of human use could also have unknown and unintended effects.

Acknowledgements This research was supported in full by Archer School for Girls Honors Research program. Special thanks to Mr. Mike Carter for assistance with designing, constructing and testing the Helmholtz Coil used in this project.

References [1]

Johnsen, S., & Lohmann, K. (2005). The Physics and Neurobiology of Magnetoreception. Nature Reviews/ Neuroscience, Vol 6, 703-712.

[2]

Papi, F., Luschi, P. & Limonta, P. (1992). Orientation-Disturbing Magnetic Treatment Affects the Pigeon Opioid System. J. exp. Biol. 166. 169-179.

[3]

Prato, F. S., Thomas, A., & Cook, C. (2005). Extremely Low Frequency Magnetic Fields (ELFMF) and Pain Therapy.

Advances in Electromagnetic Fields in Living Systems, Vol. 4, 155-187.

[4]

Kavaliers, M., & Ossenkopp, K. P. (1991) Opioid systems and magnetic field effects in the land snail, Cepaea nemoralis.

Biol. Bull. 180, 301-309.

[5]

Prato, F. S., Carson, J., Ossenkopp, K., & Kavaliers, M. (1995).

Possible mechanisms by which extremely low frequency magnetic fields affect opioid function. The FASEB Journal: Vol. 9, 807-814.

[6]

Sakisda, L., Galea, L., & Kavaliers, M. (1992). Antinociceptive Effects of the Enkephalinase Inhibitor, SCH 34826, in the Snail, Cepaea nemoralis. Peptides: Vol. 14, pp. 763-765.

[7]

Prato, F.S., Kavaliers, M., & Carson, J.J.L.(1996). Behavioural responses to magnetic fields by land snails are dependent on both magnetic field direction and light. The Royal Society: Biological Sciences. Proc. R. Soc. Lond. B. Vol. 263 No. 1376 1437-1442.

[8]

Prato, F. S., Kavaliers, M. Carson, J. (1996). Behavioural Evidence That Magnetic Field Effects in the Land Snail, Cepaea nemoralis, Might Not Depend on Magnetite or Induced Electric Currents. Bioelectromagnetics 17: 123-130.

[9]

Gordon, David G. Field Guide to the Slug. Seattle: Sasquatch Books, 1994.

[10]

Deutschlander, M., Phillips, J., & Borland, S. (1999). The Case for Light-Dependent Magnetic Orientation in Animals. The Journal of Experimental Biology 202, 891-908.

[11]

Kirschvink, J. (1992). Uniform Magnetic Fields and DoubleWrapped Coil Systems: Improved Techniques for the Design of Bioelectromagnetic Experiments. Bioelectromagnetics 13: 401-411.

About the Author

Sofia Garrick is planning on double majoring in Biology with a specialization in Ecology and Evolution with Environmental Studies on the economics/policy track. She plans to go to graduate school after college to study either marine animal behavior/ecology, the human impact on our oceans, and animal conservation policy. She loves marine biological research, especially that which is related to elasmobranchs.

Scientia

Statistical Microlensing and Gravitationally Lensed Black Holes Patrick Halkett, Janie Hoormann, and Henric Krawczynski

The focus of my research at Washington University this summer was the gravitational lensing of black holes, and working on code that provides a statistical approach to microlensing due to individual stars. I edited and developed a code initially created by Henric Krawczynski to create a random field of stars and simulate hundreds of millions of light rays passing through. Originally, the code was designed to provide a magnification map that illustrates how light is warped by the presence of individual stars in different ways.

I manipulated the code in such a way that I could observe very “zoomed-in” sections of the original map and then made histograms of each individual pixel and its magnification. While I wasn’t able to fully reach a statistical generalization about microlensing, my research revealed much about the nature of microlensing and its effect on the magnification of light rays. Gravitational lensing and microlensing especially are incredibly fascinating and up-and-coming topics in the astronomy and astrophysical sciences, and will hopefully lead to numerous discoveries in the upcoming decades.

Introduction If one considers space-time to be a bed sheet, next think of massive objects, like stars, as bowling balls on this sheet. Mass causes space-time to curve in effectively the same way that a bowling ball would cause a depression in the sheet. Because light travels in a straight path through space-time, the rays of light will follow the depression of the sheet.

Thus their trajectory will be curved as they travel in a “straight line” through curved space-time. Since space-time is curved parabolically based on distance (the gravitational force equation is dependent on the radius squared), the light rays that are closer to the mass get curved (or lensed) more strongly than those further away from the point mass. When many light rays gets curved in a similar pattern, they actually become magnified, and can help us see faint light sources that would otherwise be unobservable.

This curvature of light, called lensing, can occur on a variety of different scales; for example, the light of an entire galaxy can be lensed by another galaxy millions of light years away. However, even a single star, can cause fluctuations in the strength of the magnification if it happens to cross the image forming rays. This effect is known as microlensing and is a type of gravitational lensing that occurs on a non-galactic scale. That can relay important information about the sources from which the light is emanating.

Gravitational Lensing (GL), in particular, is extremely helpful in studying one of the universe’s most elusive phenomena, black holes. Returning to our metaphor of the sheet, a black hole is a tear in the fabric itself. Surrounding black holes are accretion disks, which are disks of matter formed when the black hole draws in matter from nearby objects such as stars. The matter is moving so

49

Winter 2017 rapidly and is so hot that it emits radiation of several different wavelengths. The light emitted by these accretion disks around black holes can be used as a pseudo-radiating body (since black holes do not actually emit anything at all). Another component of black holes is the “corona” which has multiple different models. For now, the “lamppost corona” model that treats the corona as a point hovering a certain distance above the black hole that emits radiation, like a candle or lamppost, will be used.

From here, the emission of the corona can either traverse the universe in a straight line to the observer, or can reflect off the accretion disk and be observed after scattering. An observer would witness a time difference between corona light that has traveled directly and scattered light. The notable component of the reflected corona emission that takes place at lower energy levels (6.4 keV) is known as the Fe-Kα line. This component comes from the, “reprocessing of photons in the inner accretion disk and receives its characteristic broadened profile due to gravitational redshift, Doppler effects, and relativistic beaming. This profile can then be fit to determine the BH’s spin,” (Hoorman, Beheshtipour & Krawczynski, 2016). The Fe-Kα line and its evolution also help explain why the magnification of lower energy light-curves differs from that of the higher energy light-curves. According to Chartas et al., 2012, “[the magnification difference] can be explained by the difference between a compact, softer-spectrum corona that is producing a more extended, harder spectrum reflection component off the disk,” (Chartas et al., 2012). The time delay between the direct and reflected emissions, “can be fit with numerical models to deduce system parameters such as the inclination and lamppost height [and an extended corona geometry]” (Hoorman, Beheshtipour & Krawczynski, 2016).

Gravitational lensing systems are studied through both observation and simulation. A simple GL system consists of a source, a lens, and an observer, each within its own plane. Light travels from the source through the lens, gets altered by the matter distribution in the lens plane, and then reaches the observer. The light can be dimmed or magnified by any number of celestial bodies, but for the research done on this project, Professor Henric Krawczynski randomly generated a star field to serve as the lens plane. In addition to generating 100 stars of varying mass, Professor Krawczynski’s code, called

50

the Inverse Ray-Shooting (IRS) code, also runs many simulated light rays through this field of stars. The result is a 2D histogram, called a magnification map, (shown) that illuminates the different magnifications throughout the lens plane. This process, known as lens mapping, is a tool utilized by GL researchers to study the geometry, physics, and statistics of a GL system. The magnification map depicts changes in the magnification of the simulated light rays and some of these changes are sharp and termed “caustics”. As explained in Gravitational Lenses, “the number of images changes by two if, and only if, the source crosses a caustic. Depending on the direction of crossing, two images with opposite parity merge into one and then disappear, or vice versa. Shortly before their fusion (or after their creation) the images are very bright, since they are in the vicinity of the corresponding critical curve in the lens plane,” (Schneider, Ehlers & Falco, 1992).

Thus, these critical curves help physicists better understand magnification mapping and how the images of a source are involved. As put in Chartas et al., 2015, “stars in the foreground lens galaxy produce complex microlensing magnification maps that pass over the quasar. The characteristic size of a magnification caustic produced by a single star is referred to as the Einstein radius. The strength and size of the magnification depends on the relative size of the emitting region and the Einstein radius,” (Chartas et al., 2015.)

This summer I worked closely with Professor Krawczynski to add additional functions to his IRS code and to do a statistical analysis of these magnification maps that I have briefly outlined above. My focus was on the density of the caustics in the source plane, and how the caustic is affected by changing the parameters in the lens equation, detailed below.

Methods and Results The IRS code that was developed by Professor Krawczynski is based on the Ray-trace (or lens) equation that is used to determine a source position in a GL system. This equation is explained in Kayser et al.’s Astrophysical applications of gravitational micro-lensing as, “[describing] the mapping from the deflector plane onto the source plane for rays traced to the observer.” This equation is normalized to be: ζ = sign(σ) S(z) + z and leaves us

Scientia Both Clear: Histogram 7

Histogram 10

Histogram 22

Histogram 39

Histogram 48

Both Somewhat: Histogram 38

Histogram 1

Histogram 25

Histogram 17

No Caustic: Histogram 11

Histogram 6

with the parameters σ and γ which are the normalized surface density of stars and the normalized shear, respectively. When discussing a gravitationally lensed imaged, shear is a mathematical property that stretches the image of the source around the lens. These parameters are generally low value numbers, but can be either positive or negative and they determine what pattern the lensing will follow. Kayser et al. write in their Astronomy and Astrophysics journal article that, “When comparing models with and without shear, but with the same σ, we find that the shear term causes a much higher frequency of such high peaks for small σ” (36). He continues, “for larger values of σ some stars may be sufficiently isolated that some single star critical curves will appear” (38).

As previously explained, Professor Krawczynski’s code initially generated a random star field, light ray simulations, and created several histograms. Furthermore, it also generated 50 random points in the 2D representation of the plane, which is where I began developing my own

51

Winter 2017 code. My first addition to Krawczynski’s code was to use these 50 random points as midpoints for zoomins on small random regions of the larger map to generate 50 more histograms. These small regions depict how each individual light ray is magnified (or not) by the matter in the plane, and I have included several of the ones that are generated by new IRS code.

In addition to the colored 2D histograms that represent random close-ups of the magnification map, I have also included a side-by-side comparison to the 2D histograms’ one dimensional counterparts.

These 1D histograms are created by my second addition to the IRS code. The code looks over the number of simulated rays in each section of the colored 2D histograms, the magnification of each ray, and then creates more histograms that show the counts of differently magnified rays. As indicated by the scale on the histograms, the black and blue rays have little to no magnification, red and orange rays are somewhat magnified, and yellow and white rays are strongly magnified. The reason these maps are colored is to demonstrate how caustics will appear in a magnification map. I have organized the magnification maps as follows: Both Clear histograms are ones with strong or obvious caustics. Both Somewhat histograms are harder to find caustics in, and No Caustic histograms do not contain a caustic.

These histograms illustrate different points. The first point is that the easiest way to tell if a histogram has a caustic is by looking at the relative scales of the X and Y axes. The histograms with caustics, on average, will have a smaller maximum Y value (the X axis is static). If you look in the histograms that have clear caustics, you can see by their respective magnification histograms that some have max Y values around the 60s-70s, but can be as high as 120 and yet as low as 45 (Histogram 22). When you look at Histograms 6 and 11 however, you can see that the max Y values are around 180. My reasoning for this difference in Y values is that caustics will tend to have larger values of magnification nearby and on the caustic, meaning that the simulated rays will tend to be more spread out over the range, rather than concentrated at the lower end of the X axis.

You can also tell if this region has a caustic based on the spread of values over the X axis. As you can see from looking at histograms 6 and 11, the spread of the histogram is incredibly small, with all the data remaining in the first bin. But when you

52

look at histograms like 7, 10, and 48, you can see that there are values ranging all the way from 0 to 300. Regions without caustics will not have a spread like this because very rarely will there exist a magnified ray that is not part of a caustic (which contains many more).

Lastly, the shape of the histogram is a good indicator of the region and if it contains a caustic.

Histograms that are bimodal or demonstrate skew to the right are, in general, strong characteristics of regions that have caustics in them. While finding clear examples of bimodal distribution is somewhat rare, there is an almost perfect bimodal distribution in histogram 22. A solid example of this can be seen in histogram 39. The histograms for this region are the most obvious and clear-cut in terms of depicting caustics. There is a blatant two-sided caustic in the colored histogram, as two white edges surrounded by dark blue/black and the profile histogram contains two almost symmetric peaks can be seen.

After creating the one and two dimensional histograms of the magnification maps, my next goal was to obtain an equation for each caustic.

One possible method that I attempted to use for identifying the location of the caustics in the magnification map was to use ROOT to filter out the most highly magnified pixels in each histogram.

I used the top 10%, but this number could be changed as desired). Then, I created a fit function to have a best fit line that “connects the dots” between the brightest pixels. This would allow you to see the general shape of the caustic, as well as give you a starting point to finding an equation.

One possible drawback to this method, however, is that if there were multiple caustics in a given region, this strategy may not be able to properly identify all of them.

Unfortunately, my novice abilities in ROOT kept me from being able to carry out all the research and programming that I wanted to undertake. I was unable to finish the code that would plot and overlay the best fit function on top of my 2D histograms. However, of the fifty regions of the magnification map that I analyzed, between 23 and 25 of them contained at least one caustic.

Discussion This project could certainly be developed going

Scientia forward in a variety of different ways. My approach to the continuation of this research would be to first finish the best fit function of the code. From there, one could automate the decision-making process for determining if the region contains a caustic by taking the average value of the magnification in each small region and if enough of the pixels are several times the average value of magnification, the code would then print out which histograms contain at least one caustic. Once each causticcontaining histogram has a best fit function/line that traces the pattern of the caustic, one could write a code that would loop over each simulated light ray in the region, find the distance from each ray to the best fit line and make a magnification profile (1D histogram) plotting the distance from each photon to the caustic. The caustic location is so important because, “it is easy to determine the dependence of the number of images on the position of the source, just by using the [locations of the caustics]” (35). Kayser et al. write that, “If the profile of an emission line varies over the source, different line profiles should be expected in different images. Of particular interest are the sudden changes in luminosity, which occur when a compact source (e. g. quasar continuum source) crosses a critical curve” (36). Once the magnification profile has been completed, you could then move on to changing the values of parameters σ and γ in order to better understand how these shape the magnification maps.

Alternatively, you could use the values for convergence and shear of gravitationally lensed quasar RXJ 1131-1231 provided in Dai et al. (2010) to conduct a statistical analysis of a real quasar.

To carry this out, one would just need to replace the parameters in the code with different values for the surface mass density of stars (σ) and shear (γ). The code would then make another 50 one and two dimensional histograms based on the new magnification map, and one could do a statistical analysis on the new histograms.

assistance, they have been invaluable throughout the entirety of my research both as collaborators and as mentors.

References [1]

Chartas, G., C. S. Kochanek, X. Dai, D. Moore, A. M. Mosquera, and J. A. Blackburne. "Revealing the Structure of an Accretion Disk through Energy Dependent X-Ray Microlensing." The Astrophysical Journal 757 (2012): ADS Database.

[2] Chartas, G., C. Rhea, C. Kochanek, X. Dai, C. Morgan, J. Blackburne, B. Chen, A. Mosquera, and C. MacLeod. "Gravitational Lensing Size Scales for Quasars." Astronomische Nachrichten 337 (2015): ADS Database.

[3] Dai, X., C. S. Kochanek, G. Chartas, S. Kozlowski, C. W. Morgan, G. Garmire, and E. Agol. "The Sizes of the X-Ray and Optical Emission Regions of RXJ 1131-1231." The Astrophysical Journal (2010): 278-85.

[4] Hoormann, Janie K., Banafsheh Beheshtipour, and Henric Krawczynski.

"Testing General Relativity's No-Hair Theorem with X-Ray Observations of Black Holes." Physical Review D 93.4 (2016): ADS Database.

[5] Kayser, R., S. Refsdal, and R. Stabell. "Astrophysical Applications of Gravitational Micro-lensing." Astronomy and Astrophysics 166 (1986): 36-52.

[6] Schneider, P., J. Ehlers, and E. E. Falco. Gravitational Lenses. Berlin: Springer-Verlag, 1992.

[7] Wambsganss, Joachim. "Gravitational Lensing as a Powerful Astrophysical Tool: Multiple Quasars, Giant Arcs, and Extrasolar Planets." Annalen Der Physik 518 (2006): 43-59. ADS Database.

About the Author Patrick Halkett is a third-year physics major at the University of Chicago. He is also pursuing a specialization in astrophysics and a minor in statistics.

He hopes to work at a place like NASA or Boeing in the future. He has had a fascination with astronomy and astrophysics for several years, and wants to help advance astrophysical research in any way possible.

Acknowledgements I would like to thank the Washington University McDonnell Center for the Space Sciences for financial support. Additionally, I would like to express my gratitude to Professor Henric Krawczynski and Janie Hoormann for their perpetual guidance and

53

Winter 2017

ERBB2 Germline Mutations Predispose to Cancer and Schizophrenia

Hyesan S Lee, Riyue Bao, Anita Ng, Mark M Sasaki, Alyna Y Katti, Lei Huang, Andrew D Skol, Jaume Mora, and Kenan Onel Though the mechanisms and symptoms of many complex diseases are known, the genetic factors that predispose individuals to such diseases are relatively unexplored. To aid in addressing this issue, we performed whole exome sequencing paired with functional studies on a family of Spanish descent, from which both cancer and schizophrenia have been tracked over three generations. Samples of the individuals were obtained from the Hospital Sant Joan de DĂŠu, Barcelona, Spain. For each of these samples, 98% of the exome was covered at 5x and 85% at 20x, resulting in an average coverage depth of 55x across the entirety of the exome. We were able to isolate candidate germline variants associated with cancer and schizophrenia, and their functional consequences were investigated in HEK293T cells. This allowed us to identify a single rare germline variant (I654V) that is linked to a common polymorphism (I655V) in ERBB2, which is a gene associated with both cancer and schizophrenia, inherited by all affected family members. The functional studies showed that the linked mutations stabilized the ERBB2 intermembrane protein, allowing activation of ERBB2 signaling, suggesting that the pathway may be associated with both diseases. The following provides genetic and functional data supporting a model of genetic pleiotropy, a phenomenon in which a single gene is responsible for several phenotypes.

Introduction individuals with schizophrenia may be less prone to the development of cancer (Barak et al., 2005; The majority of cancers are unpredictable Cohen et al., 2002; Lichtermann et al.,2001). and isolated, but the continued study of familial This occurs despite the fact that individuals with cancer syndromes has suggested that candidate schizophrenia have a higher incidence of smoking genes may contribute to the development of and an unhealthy diet. While these epidemiological specific types of cancer. For instance, inherited Rb1 data suggest an inverse relationship between mutations are thought to result in the occurrence schizophrenia and cancer, our results show a direct of intergenerational retinoblastomas (cancer of the correlation through the identified risk allele. Its eye) (Ali et al. 2010), while other mutations can give heritable nature, in combination with the lower rise to syndromes that cause members to develop incidence of cancer, suggests that the diseases one of several different types. Though many of are mediated by a common genetic underpinning. these syndromes have been identified, they have Schizophrenia is a heritable disease that affects not been officially categorized, suggesting that up to 1% of the population, and it is commonly further investigation may lead to the discovery of characterized by hallucination, delusion, and new susceptibility genes and syndromes. abnormal social behavior. Those with a first-degree

Previous studies have suggested that relative suffering from schizophrenia have a 10%

54

Scientia greater risk for developing the disease, while those with a monozygotic twin that has schizophrenia have 40% increased risk.

A functional genomics analysis of the Spanish family was performed using whole exome sequencing on germline DNA obtained from multiple members with cancer, schizophrenia, or both. This allowed us to identify a rare variant in ERBB2 (I654V, rs1801201), a gene that is known to be associated with both cancer and schizophrenia. This variant is located near another common polymorphism (I655V, rs1136201) with which it is linked. Subsequent functional studies were performed demonstrating that the double mutant (I654V;I655V) stabilized the intermembrane protein and activated the ERBB2 signaling pathway, suggesting that this is mechanism is associated with the two diseases.

Methods

Genomic Analysis A pedigree of the subject family was synthesized and is presented in Figure 1. The DNA necessary for whole exome sequencing (WES) was obtained from whole blood and at least 1µg of DNA was used to generate sequence reads. This process was performed at the University of Chicago Functional Genomics Facility. These results were then aligned to the human reference genome, and the variants were identified using the Illumina HiSeq2000 instrument. The alignment was achieved using three aligners: BWA, Bowtie2, and Novoalign. The population minor allele frequencies were obtained from three different databases, and each variant was marked for pathogenicity using prediction programs. To prioritize candidate germline variants, several requirements were designated. Candidates that fulfilled the designated conditions and were associated with already known cancer-associated genes were prioritized. These candidate genes were then confirmed using PCR reactions followed

by Sanger sequencing.

Functional Analysis To begin the functional analysis of the variants, we obtained cDNA of ERBB2 contained within pcDNA3.1-ERBB2-WT. GENEART Site-Directed Mutagenesis System (from Invitrogen) was used to

mutate A1963F in order to generate the desired I655V mutant; A1960G/A1963G was used to produce the I654V;I655V double mutant. These products were digested with DpnI and transformed into chemically competent bacterial cells. Both mutants were verified by Sanger Sequencing, and a plasmid containing MCherry was used to determine transformation efficiency. HEK293T cells from the American Type Culture Collection were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1x penicillin, and 1x streptomycin. We transfected these populations using Lipofectamine 3000 at 80% confluency. Those used in phospho-Akt analysis were transfected under serum-reduced conditions (0.1% FBS), while those used in phospho-p38 analysis were transfected at 10% FBS. Cell lysate was subsequently harvested 24-48 hours posttransfection for protein degradation analysis.

The cells were lysed using RIPA lysis buffer and were subjected to sonication. Protein concentrations were determined using the Pierce BCA Protein Assay Kit (Thermo Scientific) and 10mg of protein were analyzed using SDS/PAGE gels transferred onto PVDF membranes. These were exposed to Rabbit polyclonal anti-RAN (1:1000), rabbit polyclonal antiAkt (1:3000), rabbit polyclonal anti-p38 (1:3000), rabbit monoclonal anti-phospho-p38 (Thr180/182) (1:2000), and rabbit monoclonal anti-phospho-Akt 9Ser473) (1:4000) paired with anti-rabbit HRPconjugated secondary (1:2000). These were used to determine relative concentrations of Akt to phospho-Akt, and p-38 to phospho-p38 (and RAN served as the positive control). Rabbit anti-Neu2 antibody (1:10000) was used to detect ERBB2 protein levels.

Results

Identifying cancer-predisposing germline mutations In order to isolate candidate mutations, we hypothesized that siblings I-1 and I-3 shared at least one high penetrance rare cancer-predisposing variant which were inherited by II-2 and III-1 (labeling with respect to Figure 1), who were diagnosed with Hodgkin’s lymphoma (HL) and neuroblastoma (NBL), respectively. Since the father of III-1 is genetically unrelated to the family, variants that were shared with this individual were removed. After exclusion,

55

Winter 2017 ERBB2 I654V and I655V variants

Figure 1.

68 variants were shared among the cancer-affected individuals, which were categorized according to mutation types. By doing so, we isolated 48 candidate variants, from which we ultimately identified a deleterious variant in the AHCY gene and ERBB2 gene (also known as the HER2 gene), the latter of which is commonly mutated in breast cancer cases and is linked to familial breast and lung cancers (Yamamoto et al., 2014). The ERBB2 variant resulted in an isoleucine to valine substitution at amino acid 654, a mutation that is known to be associated with familial breast cancer (Frank et al.

2005).

Identifying schizophrenia-predisposing mutations Following a similar line of logic, we hypothesized that individuals I-2 and II-3 shared at least one schizophrenia-predisposing high-penetrance mutation and that I-3 was an obligate carrier.

The rare or novel variants were identified and categorized. Among those candidates identified (77 genes were predicted to be deleterious from this process), 5 were in the schizophrenia-associated gene list compiled by the NHGRI schizophrenia GWAS database. We also found four genes thought to carry variants associated with schizophrenia outside of the database (CBS, CSF2RB, ERBB2, and PLG). Of these, ERBB2 is known to form a heterodimer with ERBB4, which modulates NRG1 signaling to increase expression of DISC1, a gene that is known to contribute to neurodevelopment and whose disruption potentially contributes to schizophrenia. Therefore, ERBB2 is the only gene in which high-penetrance variants were identified and implicated in cancer and schizophrenia etiology.

56

The I654V amino acid change is known to be in complete linkage disequilibrium with a more common polymorphism causing a similar isoleucine to valine amino acid change at position 655 (I655V), which has been studied in breast cancer with inconclusive results. Therefore, all individuals with the rare rs1801201 allele (I654V) also carry the more common rs1136201 minor allele causing the I655V amino acid change, but those with the minor allele carry the wild-type rs1801201 allele. Thus, as expected, the co-occurrence of the two mutations were found in the affected family members.

Functional Studies The ERBB2 protein functions by forming a homodimer or a heterodimer with other ERBB family members to activate Akt and p38 signaling.

The transmembrane domain of the protein plays a critical role in stabilizing these dimers. So, we hypothesized that the I654V mutation, located in a glycine zipper motif and â&#x20AC;&#x160;previously found to increase stabilization of the dimers (Yamamoto et. Al, 2014), contributed to increased activation of these pathways. We tested this hypothesis by comparing wild-type ERBB2 (WT) to that of the I655V single mutant (SM) and I654V;I655V double mutant (DM) in HEK 293T cells. This cell line was chosen because it lacks endogenous ERBB2 and ERBB4. Protein synthesis was blocked 24 hours after transfection using CHX (100 ug/ml) and ERBB2 stability was determined using western blots (Figure 2). We found

Figure 2.

that WT ERBB2 degraded 12-18 hours after CHX administration, whereas SM degradation occurred 18-24 hours post-administration; DM degradation was even slower, with higher levels of protein 24 hours post-administration of CHX. To determine downstream effects of increased stability, we investigated phospho-Akt and phospho-p38 levels 24 hours following transfection. We found that Akt

Scientia signaling was upregulated similarly in SM and DM whereas p38 signaling was markedly greater in DM than in SM samples (Figure 3).

family, a better understanding of the interactions between the presented ERBB2 mutants and ERBB4 may provide a means to decipher the oncogenic potential of the I654V and I655V variants.

References [1]

Aberg, K., Adkins, D.E., Bukszar, J., Webb, B.T., Caroff, S.N., Miller, D.D., Sebat, J., Stroup, S., Fanous, A.H., Vladimirov, V.I., et al. (2010). Genomewide association study of movementrelated adverse antipsychotic effects. Biol Psychiatry 67, 279-282.

[2]

Agim, Z.S., Esendal, M., Briollais, L., Uyan, O., Meschian, M., Martinez, L.A.M., Ding, Y.M., Basak, A.N., and Ozcelik, H. (2013). Discovery, Validation and Characterization of Erbb4 and Nrg1 Haplotypes Using Data from Three Genome-Wide Association Studies of Schizophrenia. Plos One 8.

[3]

Ali, M.J., Parsam, V.L., Honavar, S.G., Kannabiran, C., Vemuganti, G.K., and Reddy, V.A. (2010). RB1 gene mutations in retinoblastoma and its clinical correlation. Saudi J Ophthalmol 24, 119-123.

[4]

Ameyaw, M.M., Tayeb, M., Thornton, N., Folayan, G., Tariq, M., Mobarek, A., Evans, D.A.P., Ofori-Adjei, D., and McLeod, H.L. (2002). Ethnic variation in the HER-2 codon 655 genetic polymorphism previously associated with breast cancer. J Hum Genet 47, 172-175.

[5]

Arpel, A., Sawma, P., Spenle, C., Fritz, J., Meyer, L., Garnier, N., Velazquez-Quesada, I., Hussenet, T., Aci-Seche, S., Baumlin, N., et al. (2014). Transmembrane Domain Targeting Peptide Antagonizing ErbB2/Neu Inhibits Breast Tumor Growth and Metastasis. Cell Rep 8, 1714-1721.

[6]

Barak, Y., Achiron, A., Mandel, M., Mirecki, I., and Aizenberg, D. (2005). Reduced cancer incidence among patients with schizophrenia. Cancer 104, 2817-2821.

[7]

Baselga, J., and Swain, S.M. (2009). Novel anticancer targets: revisiting ERBB2 and discovering ERBB3. Nat Rev Cancer 9, 463-475.

[8]

Beauclair, S., Formento, P., Fischel, J.L., Lescaut, W., Largillier, R., Chamorey, E., Hofman, P., Ferrero, J.M., Pages, G., and Milano, G. (2007). Role of the HER2 [lle655Val] genetic polymorphism in tumorogenesis and in the risk oftrastuzumab-related cardiotoxicity. Ann Oncol 18, 1335-1341.

[9]

Bergen, S.E., O'Dushlaine, C.T., Ripke, S., Lee, P.H., Ruderfer, D.M., Akterin, S., Moran, J.L., Chambert, K.D., Handsaker, R.E., Backlund, L., et al. (2012). Genome-wide association study in a Swedish population yields support for greater CNV and MHC involvement in schizophrenia compared with bipolar disorder. Mol Psychiatry 17, 880-886.

[10]

Brennan, P.J., Kumagai, T., Berezov, A., Murali, R., and Greene, M.I. (2002). HER2/Neu: mechanisms of dimerization/ oligomerization (vol 19, pg 6093, 2000). Oncogene 21, 328-328.

[11]

Brix, D.M., Clemmensen, K.K., and Kallunki, T. (2014). When Good Turns Bad: Regulation of Invasion and Metastasis by ErbB2 Receptor Tyrosine Kinase. Cells 3, 53-78.

[12]

Catts, V.S., Lai, Y.L., Weickert, C.S., Weickert, T.W., and Catts, S.V. (2016). A quantitative review of the postmortem evidence for decreased cortical N-methyl-d-aspartate receptor expression levels in schizophrenia: How can we link molecular abnormalities to mismatch negativity deficits? Biol Psychol 116, 57-67.

[13]

Chen, P., Huang, K., Zhou, G.Q., Zeng, Z., Wang, T., Li, B.J., Wang, Y., He, L., Feng, G.Y., and Shi, Y.Y. (2011). Common SNPs in CSF2RB are associated with major depression and

Figure 3.

Discussion This study allowed us to investigate the relationship between cancer and schizophrenia.

Since schizophrenia developed in the subsequent generation in the offspring of an individual affected with cancer, we hypothesized that variants were present in this individual that increased susceptibility to both cancer and schizophrenia. Following identification of rare deleterious variants, we concluded that ERBB2 was the best candidate since the gene has been associated with both diseases.

The protein is a member of the epidermal growth factor receptor (EGFR) family of receptor tyrosine kinases. The binding between these growth factors ultimately regulates cell growth, proliferation, and differentiation through a signal transduction cascade that is incompletely defined. As expected, ERBB2 has been observed in a variety of cancers including breast, lung, ovarian, and colorectal cancers (Muzny et al., 2012; Slamon et al. 1989; Tateishi et al., 1991).

The specific variants isolated in our study were previously associated with increased risk for breast cancer.

The members of the EGFR family can form both heterodimers and homodimers with the exception of ERBB2, which prefers heterodimerization with ERBB4 in the presence of NRG1 (GrausPorta et al., 1997). The upregulation of both factors was found in postmortem brains of those with schizophrenia.

In addition, a polymorphism in DISC1, a downstream target of ERBB2, has also been associated with increased rick for schizophrenia, a discovery which presents a potential mechanism by which the mutated ERBB2 may increase susceptibility to schizophrenia.

Due to the occurrence of both diseases within the

57

Winter 2017 schizophrenia in the Chinese Han population. World J Biol Psychia 12, 233-238.

[14]

[15]

[16]

Cohen, M.E., Dembling, B., and Schorling, J.B. (2002). The association between schizophrenia and cancer: a populationbased mortality study. Schizophr Res 57, 139-146.

Cross-Disorder Group of the Psychiatric Genomics, C. (2013).

Identification of risk loci with shared effects on five major psychiatric disorders: a genome-wide analysis. Lancet 381, 1371-1379.

Dahabreh, I.J., and Murray, S. (2011). Lack of replication for the association between HER2 I655V polymorphism and breast cancer risk: A systematic review and meta-analysis.

Cancer Epidemiol 35, 503-509.

[29]

Lisman, J.E., Coyle, J.T., Green, R.W., Javitt, D.C., Benes, F.M., Heckers, S., and Grace, A.A. (2008). Circuit-based framework for understanding neurotransmitter and risk gene interactions in schizophrenia. Trends Neurosci 31, 234-242.

[30]

Maier, W., Lichtermann, D., Minges, J., and Heun, R. (1994).

Personality-Disorders among the Relatives of SchizophreniaPatients. Schizophrenia Bull 20, 481-493.

[31]

Mavaddat, N., Dunning, A.M., Ponder, B.A.J., Easton, D.F., and Pharoah, P.D. (2009). Common Genetic Variation in Candidate Genes and Susceptibility to Subtypes of Breast Cancer. Cancer Epidem Biomar 18, 255-259.

[32]

Mei, L., and Xiong, W.C. (2008). Neuregulin 1 in neural development, synaptic plasticity and schizophrenia. Nat Rev Neurosci 9, 437-452.

[33]

Mill, C.P., Zordan, M.D., Rothenberg, S.M., Settleman, J., Leary, J.F., and Riese, D.J., 2nd (2011). ErbB2 Is Necessary for ErbB4 Ligands to Stimulate Oncogenic Activities in Models of Human Breast Cancer. Genes Cancer 2, 792-804.

[34]

Mishra, R.K., Marcotte, E.R., Chugh, A., Barlas, C., Whan, D., and Johnson, R.L. (1997). Modulation of dopamine receptor agonist-induced rotational behavior in 6-OHDA-lesioned rats by a peptidomimetic analogue of Pro-Leu-Gly-NH2 (PLG).

Peptides 18, 1209-1215.

[35]

Muzny, D.M., Bainbridge, M.N., Chang, K., Dinh, H.H., Drummond, J.A., Fowler, G., Kovar, C.L., Lewis, L.R., Morgan, M.B., Newsham, I.F., et al. (2012). Comprehensive molecular characterization of human colon and rectal cancer. Nature 487, 330-337.

[36]

Papadopoulou, E., Tripsianis, G., Anagnostopoulos, K., Tentes, I., Kakolyris, S., Galazios, G., Sivridis, E., Simopoulos, K., and Kortsaris, A. (2008). The influence of serum HER-2 levels and HER-2 codon 655 polymorphism on breast cancer outcome. Neoplasma 55, 113-121.

[17]

Einarsdottir, K., Rosenberg, L.U., Humphreys, K., Bonnard, C., Palmgren, J., Li, Y., Li, Y., Chia, K.S., Liu, E.T., Hall, P., et al. (2006). Comprehensive analysis of the ATM, CHEK2 and ERBB2 genes in relation to breast tumour characteristics and survival: a population-based case-control and follow-up study. Breast Cancer Res 8.

[18]

Frank, B., Hemminki, K., Wirtenberger, M., Bermejo, J.L., Bugert, P., Klaes, R., Schmutzler, R.K., Wappenschmidt, B., Bartram, C.R., and Burwinkel, B. (2005). The rare ERBB2 variant Ile654Val is associated with an increased familial breast cancer risk. Carcinogenesis 26, 643-647.

[19]

Gilbertson, R.J., Clifford, S.C., MacMeekin, W., Meekin, W., Wright, C., Perry, R.H., Kelly, P., Pearson, A.D., and Lunec, J. (1998). Expression of the ErbB-neuregulin signaling network during human cerebellar development: implications for the biology of medulloblastoma. Cancer Res 58, 3932-3941.

[20]

Golimbet, V., Korovaitseva, G., Abramova, L., and Kaleda, V. (2009). The 844ins68 polymorphism of the cystathionine betasynthase gene is associated with schizophrenia. Psychiat Res 170, 168-171.

[21]

GrausPorta, D., Beerli, R.R., Daly, J.M., and Hynes, N.E. (1997).

ErbB-2, the preferred heterodimerization partner of all ErbB receptors, is a mediator of lateral signaling. Embo J 16, 1647-1655.

[37]

Pitcher, G.M., Kalia, L.V., Ng, D., Goodfellow, N.M., Yee, K.T., Lambe, E.K., and Salter, M.W. (2011). Schizophrenia susceptibility pathway neuregulin 1-ErbB4 suppresses Src upregulation of NMDA receptors. Nat Med 17, 470-478.

[22]

Hahn, C.G. (2011). A Src link in schizophrenia. Nat Med 17, 425-427.

[38]

[23]

Iwakura, Y., and Nawa, H. (2013). ErbB1-4-dependent EGF/ neuregulin signals and their cross talk in the central nervous system: pathological implications in schizophrenia and Parkinson's disease. Front Cell Neurosci 7.

Saetre, P., Agartz, I., De Franciscis, A., Lundmark, P., Djurovic, S., Kahler, A., Andreassen, O.A., Jakobsen, K.D., Rasmussen, H.B., Werge, T., et al. (2008). Association between a disruptedin-schizophrenia 1 (DISC1) single nucleotide polymorphism and schizophrenia in a combined Scandinavian case-control sample. Schizophr Res 106, 237-241.

[24]

Klapper, L.N., Kirschbaum, M.H., Sela, M., and Yarden, Y. (2000). Biochemical and clinical implications of the ErbB/HER signaling network of growth factor receptors. Adv Cancer Res 77, 25-79.

[39]

Sasaki, M.M., Skol, A.D., Bao, R.Y., Rhodes, L.V., Chambers, R., Vokes, E.E., Cohen, E.E.W., and Onel, K. (2015). Integrated Genomic Analysis Suggests MLL3 Is a Novel Candidate Susceptibility Gene for Familial Nasopharyngeal Carcinoma.

Cancer Epidem Biomar 24, 1222-1228.

[25]

Klingberg, S., Herrlich, J., Wiedemann, G., Wolwer, W., Meisner, C., Engel, C., Jakobi-Malterre, U.E., Buchkremer, G., and Wittorf, A. (2012). Adverse Effects of Cognitive Behavioral Therapy and Cognitive Remediation in Schizophrenia Results of the Treatment of Negative Symptoms Study. J Nerv Ment Dis 200, 569-576.

[40]

Schizophrenia Working Group of the Psychiatric Genomics, C. (2014). Biological insights from 108 schizophreniaassociated genetic loci. Nature 511, 421-427.

[41]

Seshadri, S., Kamiya, A., Yokota, Y., Prikulis, I., Kano, S., Hayashi-Takagi, A., Stanco, A., Eom, T.Y., Rao, S., Ishizuka, K., et al. (2010). Disrupted-in-Schizophrenia-1 expression is regulated by beta-site amyloid precursor protein cleaving enzyme-1-neuregulin cascade. Proc Natl Acad Sci U S A 107, 5622-5627.

[42]

Silberberg, G., Darvasi, A., Pinkas-Kramarski, R., and Navon, R. (2006). The involvement of ErbB4 with schizophrenia: Association and expression studies. Am J Med Genet B 141b, 142-148.

[43]

Slamon, D.J., Godolphin, W., Jones, L.A., Holt, J.A., Wong, S.G., Keith, D.E., Levin, W.J., Stuart, S.G., Udove, J., Ullrich, A., et al. (1989). Studies of the Her-2/Neu Proto-Oncogene in Human-Breast and Ovarian-Cancer. Science 244, 707-712.

[44]

Slamon, D.J., Leyland-Jones, B., Shak, S., Fuchs, H., Paton, V.,

[26]

Law, A.J., Kleinman, J.E., Weinberger, D.R., and Weickert, C.S. (2007). Disease-associated intronic variants in the ErbB4 gene are related to altered ErbB4 splice-variant expression in the brain in schizophrenia. Hum Mol Genet 16, 129-141.

[27]

Lichtermann, D., Ekelund, J., Pukkala, E., Tanskanen, A., and Lonnqvist, J. (2001). Incidence of cancer among persons with schizophrenia and their relatives. Arch Gen Psychiat 58, 573-578.

[28]

Lindor, N.M., McMaster, M.L., Lindor, C.J., Greene, M.H., National Cancer Institute, D.o.C.P.C.O., and Prevention Trials Research, G. (2008). Concise handbook of familial cancer susceptibility syndromes - second edition. J Natl Cancer Inst Monogr, 1-93.

58

Scientia Bajamonde, A., Fleming, T., Eiermann, W., Wolter, J., Pegram, M., et al. (2001). Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. New Engl J Med 344, 783-792.

[45]

Tateishi, M., Ishida, T., Mitsudomi, T., Kaneko, S., and Sugimachi, K. (1991). Prognostic Value of C-Erbb-2 Protein Expression in Human Lung Adenocarcinoma and SquamousCell Carcinoma. Eur J Cancer 27, 1372-1375.

[46]

Varley, J.M. (2003). Germline TP53 mutations and Li-Fraumeni syndrome. Hum Mutat 21, 313-320.

[47]

Weiner, D.B., Kokai, Y., Wada, T., Cohen, J.A., Williams, W.V., and Greene, M.I. (1989). Linkage of Tyrosine KinaseActivity with Transforming Ability of the P185neu Oncoprotein.

Oncogene 4, 1175-1183.

[48]

Welter, D., MacArthur, J., Morales, J., Burdett, T., Hall, P., Junkins, H., Klemm, A., Flicek, P., Manolio, T., Hindorff, L., et al. (2014). The NHGRI GWAS Catalog, a curated resource of SNP-trait associations. Nucleic Acids Res 42, D1001-1006.

[49]

Wen, H.X., Kim, Y.C., Snyder, C., Xiao, F.X., Fleissner, E.A., Becirovic, D., Luo, J.T., Downs, B., Sherman, S., Cowan, K.H., et al. (2014). Family-specific, novel, deleterious germline variants provide a rich resource to identify genetic predispositions for BRCAx familial breast cancer. Bmc Cancer 14.

[50]

Yamamoto, H., Higasa, K., Sakaguchi, M., Shien, K., Soh, J., Ichimura, K., Furukawa, M., Hashida, S., Tsukuda, K., Takigawa, N., et al. (2014). Novel Germline Mutation in the Transmembrane Domain of HER2 in Familial Lung Adenocarcinomas. Jnci-J Natl Cancer I 106.

59

Winter 2017

About The Triple Helix

at The University of Chicago

The Triple Helix, Inc. (TTH) is the world’s largest completely studentrun organization dedicated to evaluating the true impact of historical and modern advances in science. Of TTH’s more than 25 chapters worldwide, the University of Chicago chapter is one of the largest and most active. We, TTH at The University of Chicago, are extremely proud of our chapter’s accomplishments, campus presence, and strong connection to other TTH chapters around the world.

We continue to work closely with an ever-increasing number of faculty members, and have notably acquired the generous support of the founding Pritzker Director of Chicago’s Institute for Molecular Engineering, Matthew Tirrell, and his department. We have expanded our local organization so that now, we can confidently say that there is a place here for each and every one of our fellow college students. We have consciously and dramatically increased the size of our production, marketing, and events teams, and have watched our group of talented writers and editors grow at unprecedented levels. In fact, we have further expanded the intellectual diversity of our chapter, with TTH members having declared for more than 30 of the University’s different major and minor programs. Finally, we are absolutely thrilled to present the newest issue of our journal of original undergraduate research, Scientia.

Over the years, TTH UChicago members have found themselves in research positions around campus, taking advantage of the hundreds of opportunities we are fortunate to have here. We found ourselves wishing, however, that there was an outlet where we could reach out to our peers on campus, to share our projects and project ideas – and to hear or read about their work as well. So we decided to create that outlet ourselves. This was the impetus for our journal, Scientia, which is the culmination of many of our members’ hard work. Since its inception, we have continued to strive for excellence in the diverse world of research by publishing the highest quality of undergraduate research. We hope you enjoy reading our latest issue!

Stephen Yu President of The Triple Helix, Inc.

60

Autumn 2016

Acknowledgements The Triple Helix at the University of Chicago would like to thank the following individuals for their generous and continued support: Dr. Matthew Tirrell

Founding Pritzker Director of the Institute for Molecular Engineering

David Clark

Assistant Vice President for Student Life and Associate Dean of the College

Arthur Lundberg

Senior Assistant Director, Administration and Financial Advising

Tempris Daniels

Student Involvement Advisor

We also thank the following departments and groups: The Institute for Molecular Engineering The Biological Sciences Division The Physical Sciences Division The Social Sciences Division University of Chicago Annual Allocations Student Government Finance Committee (SGFC) Chicago Area Undergraduate Research Symposium UChicago Undergraduate Research Symposium

Finally, we would like to acknowledge all our Faculty Review Board members and the mentors of our abstract authors for their time and effort.

Research Submission Undergraduates who have completed substantial work on a topic are highly encouraged to submit their manuscripts. We welcome both full-length research articles and abstracts.

Please email submissions to uchicago.print@thetriplehelix.org. Please include a short description of the motivation behind the work, relevance of the results, and where and when you completed your research.

If you would like to learn more about Scientia and The Triple Helix, visit http://thetriplehelix.uchicago.edu or contact us at uchicago@thetriplehelix.org.

61

Meet the Staff Scientia

Production Scientia Director Helena Zhang SISR Director Ariel Goldszmidt

Editors in Chief Jeremy Chang Clara Sava-Segal Managing Editors Nilanjana Ray Abhinav Ranjan Quang Tran Associate Editors Derrick Tang Hyesan Lee Zainab Aziz Kathyrn Hicks Margaret Lazarovits Writers: Inquiries Samara Singh (Dr. Greg Norman) In Depth Shira Eisnberg Sofia Garrick Patrick Halkett Hyesan Lee

Marketing Director Nilanjana Ray

Web Webmaster Tima Karginov

Events Director Peter Ryffel Coordinators Jonathan Chuang Franklin Rodriguez

Science in Society Review Editor in Chief Annie Albright Managing Editor Aya Nimer

E-Publishing Editor in Chief Aliya Moreira Managing Editor Bella Pan

Executive President Stephen Yu Vice President Salman Arif

Contact Us Email: uchicago.president@thetriplehelix.org Website: http://thetriplehelix.uchicago.edu