PREFACE
Origin of This Book
F
orest pathology in North America first appeared around the beginning of the 20th century as an integration of taxonomic fungal plant pathology and conservation forestry (Peterson and Griffith 1999). Although it has developed in many directions during the interceding century, this field still fundamentally addresses the same two issues. What has changed, of course, are the tools used for research and the conceptual basis used to interpret research results. Molecular biology/genetics has become the domain of taxonomic plant pathology, and spatial ecology has become the domain of conservation forestry. Integration of these two domains is no less the goal of our field today than it was 100 years ago, and its emphasis still makes forest pathology unique. The compilation of papers in this book is based on a symposium presented at the 1999 Montreal American Phytopathological (APS) Society Annual Convention, entitled “From molecules to ecosystems: bridging scales in forest disease concepts’’. Papers presented in the symposium addressed the use of concepts and methods from epidemiology, spatial ecology, molecular biology, and molecular genetics in forest pathology in forests managed for both timber and non-timber resources to examine the vast range of temporal and spatial scales that forestry encompasses. For this book, we asked authors to update and expand on the themes originally presented at the Montreal symposium. As a consequence, the chapters cut across a range of the current most rapidly advancing topics in Forest Pathology. The application of molecular biology to studies of forest pathogens has expanded phenomenally in the last couple decades, as it has in many other commodity areas of plant pathology. In Chapter 1, Hamelin reviews some of the progress in the use of genomics in forest pathology, points to possible applications, and discusses the weaknesses, caveats, and challenges encountered when investigating forest pathogens, as opposed to better characterized agricultural or model pathosystems. Most early molecular biology studies of forest pathogens focused on characterizing diversity among pathogens using various molecular markers. In Chapter 2, Kim and her coauthors review molecular genetic tools used to analyze pathosystem processes—such as pathogen diagnostics, phylogenetics, population genetics, population genomics, and
genetic mapping—and speculate on how these tools would be integrated into forest management in the future. In Chapter 3, Richardson and his colleagues introduce the current theoretical and empirical concepts of population genetic processes in forest pathosystems. The chapter begins with an overview of the processes that influence population structure of plant-pathogen interactions in forest ecosystems, followed by review of studies addressing population genetics of pathosystems and a discussion of potential management implications. A molecular population genetics approach can be a powerful adjunct to field observations and experimentation in defining patterns of variation and elucidating how these different organisms interact, and how each responds to their environment. In Chapter 4, Six reviews how molecular techniques have been used to look at population level variation in bark beetle/fungal systems and discusses directions for future studies. The chapter begins with an introduction to bark beetles, their fungal associates, and association types to provide background and to acquaint readers with the complexities and challenges involved in the study of variation in these associations. In Chapter 5, McDonald and his coauthors discuss the concepts of naturalization of forest pathogens to North American hosts and environments, with reference to white pine blister rust (Cronartium ribicola). This review draws upon the extensive body of theory regarding coevolutionary interactions, induced defenses, and numerical descriptions of epidemics of “victim-exploiter” interactions in wild ecosystems in order to critically analyze the process of naturalization of plant pathogens into wild ecosystems that are highly heterogeneous in space and time for epidemic development, topography, and physical environment. Such approaches provide valuable insights toward understanding changes associated with pathogen introductions. Forest diseases result from the interaction of a virulent pathogens and susceptible hosts interacting in an enabling environment. Previous analyses of virulence and pathogenesis at a molecular level have focused largely on agricultural model systems, mostly because few data are available for forest systems. In Chapter 6, Temple and Hinzt present a detailed overview of the molecular basis of pathogenesis and virulence in forest diseases. -iii-
PREFACE
Melanin synthesis in sapstain causes a cosmetic defect that reduces the value of commercial wood products. The key enzymes of Melanin synthesis are often targeted by fungicides in plant pathogens but are not yet a target for sapstain control at sawmills. At tree harvest sites, legislation and public concern make biocontrol agents more acceptable than antifungal chemicals. Such agents must be inexpensive, reliable and ecologically compatible. Ensuring reliable and responsible use of biocontrol agents requires detailed ecological and genetic information. In Chapter 7, Breuil, Fleet, and Loppnau discuss two such research areas: 1) the impact, nature and distribution of sapstain fungi, and 2) the use of biochemical and molecular biological approaches to clarify the biosynthetic pathways for pigment production in these fungi. Many genetic engineering strategies that have shown great potential for disease resistance in agricultural crops are just beginning to be tested and applied to forest trees. Poplar was the first trees species to be genetically transformed. In fact, because it is easy to propagate clonally, and because many poplar species are susceptible to Agrobacterium, poplar has been called “the tobacco of forest trees”. In Chapter 8, Mentag and Séguin discuss genetic engineering strategies that have been used to create pathogen-resistant poplars. Decades of rather unproductive effort in searching for genetic resistance to Dutch elm disease and chestnut blight had a damping effect on the selection and breeding of disease resistant tree species. However, the past two decades have seen renewed interest in such operational tree improvement programs, particularly in pine species. In Chapter 9, Hunt and Ying attribute this resurging interest to the maturation of long-term provenance and progeny tests and public intolerance to pesticides. Impact is a relative term because diseases affect so many different elements of the forest. Trees killed by diseases reduce timber volume in stands managed for timber production, but dead trees can contribute to stand structure, the abundance and composition of snags and downed logs and wildlife habitat. Forest pathogens can lead to changes in composition of recolonizing vegetation and thus further impact wildlife habitat and diversity. Whether a disease has a negative or positive impact, or no impact at all, depends on what a forest is managed for. With the increasing number of stakeholders concerned about the use and condition of forests, non-timber impacts have become increasingly important. Few guidelines exist for managing diseases for non-timber resources. In Chapter 10, Stubblefield, Lundquist and van der Kamp discuss the influences and effects of forest pathogens on wildlife habitat, and present a framework for examining these interactions. In Chapter 11, Lundquist and Ward expand on some of the insights presented in Chapter 10 by describing a field study that examined the relative effects of diseases and other types of disturbances on small mammal prey of the Mexican spotted owl. The study illustrates that studying the connection between forest diseases and specific wildlife
species can be challenging because cause and effect in forestry is often difficult to prove. As our understanding of the effects of forest pathogens on natural ecosystems has expanded, the view that diseases have positive sustaining effects on forest health has become increasingly prevalent, but guidelines on how much disease is useful still need to be established. In Chapter 12, Rubin and Manion examine the concept of pathological baseline conditions and argue that a predictable level of mortality is required for the maintenance of forest structure at the landscape level. Recent management activities and various land uses have dramatically altered edaphic and environmental conditions from those under which forest tree species and ecosystems have evolved. In Chapter 13, Otrosina describes how the creation of unnatural landscape conditions that differ from pathological baseline conditions can cause previously unknown or insignificant pathogens to become significant damaging agents. Our traditional concepts of forest diseases are inadequate to describe this phenomenon. Variable expressions of diseases have both causes and effects that interact to create changing patterns and processes within forests. These dynamics have ecological and socioeconomic impacts that differ depending on the spatial scale. One of the most daunting challenges is finding a way to integrate impact data and still maintain its spatial integrity. Analyses of spatially explicit data have essentially created a revolution in the ways forest diseases can be assessed and modeled. In Chapter 14, Reich and Lundquist present a framework for assessing impacts of different types of diseases using spatial statistics. Lundquist expands on this theme using dwarf mistle toe as a case study in Chapter 15. Research at the scale management decisions are made is more likely to be relevant to managers and other forest stakeholders, more likely to be integrated into management decision-making, and therefore more likely to be implemented operationally. Ecosystem management, conservation forestry, restoration ecology, sustainable forestry, and other similar programs increasingly define relevant scale as the landscape. In Chapter 16, Lundquist introduces the concept of landscape pathology by describing how concepts of spatial ecology and landscape ecology might be integrated into the more traditional core competency areas within forest pathology. We gratefully acknowledge the editorial assistance of Bob Hamre, and the page layout preparation by Connie Lemos. We also acknowledge the helpful suggestions and comments by the many anonymous reviewers, and especially appreciate the patience and counsel of our senior editor Gareth Hughes. Most of all, we appreciate the keen insights and cutting edge knowledge generously shared by the authors. Peterson, P. D. and C. S. Griffith. 1999. Herman von Schrenk: the beginnings of forest pathology in the U.S. Forest. History Today Fall: 29-34. -iv-
FOREST PATHOLOGY: FROM GENES TO LANDSCAPES
Contents Preface—Origin of This Book
iii
Chapter 1—Forest Pathology in the Era of Genomics
1
Richard C. Hamelin
Chapter 2—Application of Molecular Genetic Tools to Studies of Forest Pathosystems
9
Mee-Sook Kim, Ned B. Klopfenstein, and Richard C. Hamelin
Chapter 3—Assessing Forest-pathogen Interactions at the Population Level
21
Bryce Richardson, Ned B. Klopfenstein, and Tobin L. Peever
Chapter 4—Population Genetics of Bark Beetles and Their Associated Blue-stain Fungi with the Use of Molecular Markers
31
Diana L. Six
Chapter 5—Naturalization of Host-Dependent Microbes After Introduction into Terrestrial Ecosystems
41
Geral I. McDonald, Paul J. Zambino, and Ned B. Klopfenstein
Chapter 6—Molecular Analysis of Fungal Pathogenesis in Forest Pathogens
59
Brad Temple and William E. Hintz
Chapter 7—Sap Stain in Trees, Logs and Lumber: Fungi, Pigment, and Pigment Biosynthetic Pathways
69
C. Breuil, C. Fleet, and P. Loppnau
Chapter 8—Transgenic Approaches to Increase Pathogen Disease Resistance in Forest Trees: A Case Study with Poplar
79
R. Mentag and A. Séguin
Chapter 9—Operational Uses of Disease Resistance in Conifer Tree Improvement Programs
89
R.S. Hunt and C.C. Ying
Chapter 10—Forest Disease Impacts on Wildlife: Beneficial?
95
Cindy H. Stubblefield (Holte), John E. Lundquist, and Bart van der Kamp
Chapter 11—Impacts of Diseases and Other Disturbances on Non-Timber Forest Resources: A Case Study Involving Small Mammals
105
John E. Lundquist and James P. Ward, Jr.
Chapter 12—Characterizing Regional Forest Health and Sustainability— A Case Study Using Diameter Distributions, Baseline Mortality, and Cumulative Liabilities Benjamin D. Rubin and Paul D. Manion
-v-
113
CONTENTS
Chapter 13—Exotic Ecosystems: Where Root Disease Is Not a Beneficial Component of Temperate Conifer Forests
121
William J. Otrosina
Chapter 14—Use of Spatial Statistics in Assessing Forest Diseases
127
Robin M. Reich and John E. Lundquist
Chapter 15—Patterns in Diseased Landscapes: A Case Study of a Lodgepole Pine Forest Infected by Dwarf Mistletoe
145
John E. Lundquist
Chapter 16—Landscape Pathology—Forest Pathology in the Era of Landscape Ecology
155
John E. Lundquist
Conclusion—Forest Pathology in the Era of Integration and Synergy
167
John E. Lundquist and Richard C. Hamelin
Subject Index
169
Scientific Names Index
174
-vi-
CHAPTER 1
Forest Pathology in the Era of Genomics Richard C. Hamelin Natural Resources Canada, Laurentian Forestry Service, 1055 rue du PEPS, Quebec, Quebec, Canada
Introduction
microbiologists, biologists, or mycologists who learned about forestry. Since the 1980s, more and more forest pathologists have also been trained as geneticists, biochemists, or molecular biologists. In the last decade, genomics has affected the way we look at the biological world from medicine to agriculture and forestry. With entire genomes being sequenced for bacteria, fungi, plants, and animals, the view of how cells and multi-cellular organisms work and how they evolve is becoming clearer, while at the same time, paradigms are being shifted. In plant pathology, great advances have been made in the discovery and characterization of genes involved in resistance to biotic and abiotic stress, and of genes involved in pathogenicity and virulence. Genomics research in forest pathology is still mostly in the data-gathering stage. However, more and more investigators are using these tools to test hypotheses and increase our understanding of forest pathosystems. Some of the new knowledge and tools have and will help better manage forests either through direct interventions (for example by harnessing the power of genetic engineering; see chapter 8 this volume), or by integrating knowledge at the landscape level (chapter 16 this volume).
F
orest pathology is a science of extremes. Trees are generally long-lived, with individuals that can live for centuries. Trees are not mobile and, although their progenies can disperse, they cannot escape the local environmental conditions where they establish and grow. Their generation times are measured most often in decades. Forests cover vast areas, sometimes as natural near monocultures, with one or a few species dominating the landscape. Yet, trees can be attacked and weakened or killed by microscopic organisms such as fungi, bacteria, phytoplasmas, and viruses. These microorganisms have short generation times, sometimes with multiple generations per year. Unlike their hosts, most pathogens are highly mobile, with means of dispersal ranging from vegetative mycelial growth or long-lasting resting structures, to motile or airborne spores that can travel hundreds of km. In some dramatic examples, entire forest ecosystems have been affected and reshaped by a single microorganism. The best examples in North America are that of Chestnut blight, which reduced the formerly dominant American chestnut to the status of undergrowth vegetation and White Pine Blister Rust, which is threatening the fragile white bark pine and limber pine ecosystems in the West. When one thinks of these encounters between trees and pathogens at the genomic scale, there seems to be an unfair advantage on the side of the microorganisms: the host genome is ďŹ xed spatially and temporally, but interacts with rapidly evolving genomes that can recombine, mutate, adapt, and produce repeated attacks locally or from distant sources over several years. This contrast in scales is part of what attracts and intrigues many forest pathologists. Investigators who make up the forest pathology community have been looking at the problem from both sides of the looking glass. Some are foresters trained in microbiology or mycology; others are
Model Systems Model systems are chosen in genomics research because of their genetic tractability and their ease of characterization. The possibility to test hypotheses and deepen our understanding of gene function is important and comprises genetic transformation systems and gene knockout. In genomics research, model systems also have relatively small genomes, and are amenable to genetic analysis and mapping. Ideally, model systems enable us to understand processes that are broadly applicable. This section will review some of the model systems that may be relevant to forest pathology.
-1-
CHAPTER ONE
Fungal Models
organization in this basidiomycete could be a great tool to forest pathologists. Genes coding for functions that are believed to be conserved across genera can be searched in these model fungal genomic database and studied and compared with economically important pathogens. For example, genes believed to be involved in virulence are often involved in detoxification of host defense such as secondary metabolites (Osbourn, 1999; Delserone, 1999; Han, 2001). Cytochrome P450 enzymes have been sequenced in the white-rot fungus and have been implicated in virulence in Nectria haematococca (George, 2001). This resource appears to be currently under exploited by forest pathologists. A better understanding of this gene family in these model systems should find some applications in forest pathology. Cytochrome P450 genes were recently cloned and sequenced in several pine rusts and non-synonymous mutations were found (Joly, 2003). It has been reported that isoforms of this gene could result in different substrate utilization, which could translate in differences in virulence (George, 2001).
The filamentous ascomycete Neurospora crassa (Davis and Perkins 2002) has been a model system of choice for several decades. It was chosen as a model system in the 1940s before the advent of molecular techniques. This fungus possessed useful features as a model system such as the ease with which the ascospores could be dissected and ordered from within the ascocarps, and the presence of mutants which could be easily selected and recognized. These tools allowed investigators to develop a clear understanding of genetic processes such as recombination and crossing-over. This model system was used to demonstrate the “one-gene-one enzyme” hypothesis (Beadle and Tatum 1941). (Ironically, genomics and proteomics research are currently questioning the universality of this hypothesis (Santucci et al. 2000)). The yeast was also an obvious choice as a fungal model system: it has a very small genome, grows fast, and possesses budding and filamentous phases, and, as a bonus, it has been of significant economic importance for centuries. The advantages of having chosen yeast as a model system are now becoming clear. The entire genome has been sequenced, and most of the approximately 6000 genes have been identified and mapped. Recent developments in yeast functional genomics are yielding exciting results with impact outside the yeast community. The expression profile of the entire yeast genome can now be studied in a single experiment using DNA micro arrays (Devaux et al. 2001; Ge et al. 2001; Holter et al. 2000). In addition, all genes of the yeast have now been knocked out and it is now possible to assign functions to previously unknown genes. These fungi are saprotrophs, however, and many functions essential to pathogenicity are not expected to be present in these genomes. Fortunately for plant pathologists, model fungal plant pathogens have emerged, such as Magnaporthe grisea, Mycosphaerella graminicola, and Ustilago maydis. Research on those organisms is providing insights into the mechanics of infection and the evolution of pathogenicity (Xu and Xue 2002; Yoder and Turgeon 2001). But these fungi are pathogens of monocots, and the possibility to extrapolate this research to tree pathogens is uncertain. A model fungus that may be of more interest to the forest pathology community is the white-rot Phanerochaete chrysosporium. The genome of this fungus is the first basidiomycete genome to be sequenced (Pennisi 2001). Genome sequencing, carried out by the Department of Energy Joint Genome Institute, is now complete. The interest of sequencing the genome of this fungus lies in its ability to produce lignin-degrading peroxidases (LiPs), a family of extracellular glycosylated heme proteins. These enzymes, along with others capable of degrading toxic waste, are viewed as having potential environmental importance. Understanding the genome
Forest Pathosystem Models Poplar has been chosen as the tree model system for genomics research (http://genome.jgi-psf.org/poplar0/poplar0. home.html). It has a relatively small genome compared to other tree species, and can be manipulated easily experimentally. Tissue culture, clonal propagation, and genetic transformation can be conducted routinely in poplar (Jouanin et al. 1993). The genome of poplar has been completely sequenced and worldwide consortia of research teams are poised to exploit these resources (http://genome.jgi.psf. org/). Although poplar is attacked by numerous pathogens, a tree pathogen model system has yet to be identified and characterized. Poplar leaf rust, caused by several Melampsora spp. has some advantages. It displays well defined host-pathogen interactions that can be assessed in reproducible leaf disk assays (Hamelin et al. 1994; Newcombe 1998; Pinon 1992; Pinon et al. 1994). Resistance genes have been identified, characterized, and mapped (Newcombe 1998). Rusts are among the most important groups of pathogens worldwide on a variety of trees and crops and their economic impact is very important. Poplar leaf rusts are present on all continents and therefore, can be of interest to international consortia of genomics research groups. In addition, Melampsora is the genus that generated the proof for the gene-for-gene concept, a landmark in plant pathology. But rusts are not ideal model fungi. They can only be cultured with difficulty, and their genetic manipulation is complicated by the fact that they are biotrophs. In addition, Melampsora spp. are heteroecious and therefore require alternate hosts in order to complete their life-cycle. This complicates genetic analyses. Nevertheless, the advantages -2-
FOREST PATHOLOGY: FROM GENES TO LANDSCAPES
Genome Sequencing
overweight the caveats and M. larici-populina is on the list of genomes to be sequenced in the coming years. Another important poplar disease in North America is Septoria canker. It also has fairly well defined host-pathogen interactions and some resistance has been identified (Ostry and McNabb 1985). It is culturable and is probably amenable to genetic manipulations. However, bioassays for resistance and virulence are sometimes difficult to perform and repeat, and the applicability of leaf assays to field resistance is uncertain. An important advantage of developing Septoria as a model tree pathogen would be the access to large databases within the genomics projects in Mycosphaerella gaminicola (Septoria tritici) (Table 1.1). This would make it possible to conduct whole-genome comparisons between these two related pathogens with different hosts. Another important forest pathosystem that is emerging as a leading-edge system for genome research is Heterobasidion annosum. Large resources are devoted to the generation of EST libraries and genome sequencing of the pathogen. Libraries of interactions with the hosts are being exploited and will likely shed light on this important basidiomycete and its interaction with its host. One important question to be addressed is whether or not we need a model forest pathosystem for genomics research. If model systems are to be useful, they should provide data and information that can be applied to other closely or even distantly related organisms. A potentially productive approach is to look for gene homologs in currently available model systems. For example, 40% of single-gene determinants of human heritable diseases find homologues in yeast (Oliver 2002) and filamentous fungi, and human genomes have even more homologues in common (Zeng et al. 2001). It is quite likely that a large number of genes in forest pathogens will find homologues in N. crassa, S. cerevisae, and P. chrysosporium. But it is perhaps the genes that are different between genomes of model systems and forest pathogens that may be of greater interest, since they may be the determinants of the unique set of genes that enable a forest pathogen to infect a tree.
While the first genomes sequenced were those of organelles, bacteria, and viruses, there are currently genome sequencing projects in all three major domains of life: bacteria, archaea, and eukaryota. There are nine major fungal genome sequencing projects where the goal is to sequence the entire genomes. There are also several other genome projects where portions of the genomes, usually only the coding genes, are targeted. These projects aim to study the structure, organization, and function of the genes making up an organism.
Genomics in Plant Pathology Genomics projects allow us to identify genes in organisms and assign putative functions based on homology with known genes, or based on the presence of functional domains. Although plant pathology got a late start in genomics research, several projects are currently underway (Table 1.1). Genomics projects in plant pathology generally aim to increase our understanding of host-parasite interactions through the discovery of novel genes encoding proteins that control the outcome of infection. Currently most projects involve the sequencing of expressed sequence tags (EST) of the transcribed mRNA. These short tags can give enough information to perform database searches for homology and eventually assign function to the genes. This approach is useful for gene discovery, but will be of limited use for whole-genome comparison and understanding of genome organization, structure, and evolution. A white paper was developed by the American Phytopathological Society APS to target genomes to be sequenced in the future, recognizing the importance of the task as well as the impossibility of sequencing all genomes of plant-associated microbes (www.apsnet.org.onhfeature/ microbe). Among the criteria for inclusion of pathogens for genome sequencing was economic importance, uniqueness of biological or environmental features, as well as genetic tractability.
Table 1.1. Internet genome resources for fungi and fungal plant pathogens. Name of site
Internet address
Broad Institute
http://www-broad.mit.edu
DOE joint genome institute Phytopathogenic Fungi and Oomycete EST Database
http://www.jgi.doe.gov/
Phytophthora functional genomics database
Fungi Neurospora crassa, Aspergillus nidulans, Magnaporthe grisea, Fusarium graminearum white rot fungus Phanerochaete chrysosporium Several, including Magnaporthe grisea, Mycosphaerella graminicola, Fusarium sporotrichioides P. sojae, P. infestans, P. ramorum (planned)
http://cogeme.man.ac.uk/
http://www.pfgd.org/
-3-
Resources Whole genome, predicted genes, blast searches Whole genome, Blast searches, GO, Kegg, Kog ESTs
ESTs, Whole genome
CHAPTER ONE
Given the limited resources available for research in genomics of plant pathogens large collaborative groups have joined efforts. For Phytophthora infestans and P. sojae, a consortium was developed to increase throughput and data sharing (Waugh et al. 2000) (Table 1). Other advanced projects comprise F. sporotrichioides, U. maydis and M. grisea. Among forest pathogens, several genomics sequencing projects are underway for P. ramorum, the causal agent of sudden oak death (SOD) and for Heterobasidion annosum. One exciting new development is the project to completely sequence the genomes of microorganisms associated with poplar. Two micro-organisms have been recently approved for sequencing by the JGI: a mycorhizae, and laccaria bicolor, an ecto mycorthiza. The interactions of these symbiotic microbial genomes with poplar should yield useful insight. The comparison of the array of genes expressed during mutualistic and pathogenic interactions should also be helpful in identifying what genetic determinants are responsible for pathogenicity. Poplar rust has been identified as the next poplar-associated micro-organism to be sequenced.
A consensus seems to be that sequencing a genome serves as a basis for further discoveries. One of the great challenges of genomics research will be to identify functions for those genes that do not have homologies to annotated genes. Some computational approaches are central to these projects and may help in assigning functions to genes. But function assignment will more often involve mutational and expression analyses (Gold et al. 2001; Sweigard and Ebbole 2001). Ultimately, the goal is to link a DNA sequence to a phenotype and a function (Kamoun et al. 2002). Gene knockout libraries and microarrays will be essential tools in this endeavor.
Genomic Diversity A powerful new approach to identify genes of interest that may be involved in fitness and adaptation and evaluate their function is to scan genomes in search of genetic diversity and to look for the signatures of selection. Such approaches rely upon the discovery of large numbers of SNP within species followed by large-scale SNP-genotyping at population levels using high-throughput assays (Kwok 2000, 2001). By looking at allelic variation within species, it is becoming increasingly possible to determine gene functions by association studies (Whitt, & Buckler 2003) or by comparing allelic variation in gene expression (Yan et al. 2004). Such approaches used extensively in human genetics, are promising in studies of fungal genomics due to the relatively small size of the genomes and the relative ease of phenotypic assays. By studying the pattern of genetic diversity at the genome level and by conducting comparisons in related species occupying different niches, it may be possible to dissect the evolutionary steps causing the differentiation, adaptation and specialization (Luikart et al 2003) and identify species-specific genes (Charlebois et al. 2003). For example, the dimorphic Candida albicans which is able to colonize a large variety of environments, has a larger number of specific genes, i.e. genes not found in other yeast such as S. cerevisae or S. pombe (Herrero et al. 2003). Genes that are involved in the host-pathogen interactions or adaptive processes such as substrate utilization can be hypothesized to evolve at a fast rate. This was observed in Arabidospis, where R-genes were highly polymorphic (Stahl 1999). By measuring the ratios of non-synonymous/ synonymous mutations, it is possible to detect and differentiate positive and purifying selection (Smith 1994; Smith et. al. 1995). Chitinases produced by the plant, for example, were highly polymorphic and nonsynonymous substitution rates often exceeded synonymous rates in other dicots, indicating a succession of adaptively driven amino acid replacements (Bishop 2000). In pine rust pathogens, a subset of genes involved in defense had three times higher ratios of non-synonymous/synonymous mutations than ribosomal proteins; this included a P450 that has been implicated in virulence in other pathosystems (Joly et al. 2003; Han et al. 2001; George et al. 2001).
The Post-sequencing Era Genome projects have already met their own limitations. The first limitation was encountered during the early days of the human genome project when it became obvious that sequencing one genome was not enough. By sequencing several genomes, single nucleotide polymorphisms (SNP) were discovered fairly regularly throughout the genome (every 100 to 300 bp). The SNPs are the basis for the variability that is responsible for the differences in phenotypes that we observe. SNP genotyping is proving to be a very powerful tool that allows the discovery of association between mutations and phenotypes. Although it may not be necessary to entirely sequence several individual genomes for all species, it may be useful to resequence large portions in multiple individuals to discover those SNPs. Another limitation of genome sequencing is that it is not always easy to predict gene functions from the raw sequence data. Even after the whole human genome sequencing has been completed, the exact number of genes is still debated (Claverie 2001). The proteomics, genomics, and bioinformatics communities are not always in agreement on these numbers and on the ways to predict them. Increasingly, it appears that proteomics investigators are finding more genes than were predicted by bioinformatics. In one of the best documented pathosystems, the rice blast, half of the EST derived from a cDNA library 48 h post-inoculation did not find a match in databases and therefore were presumably novel or previously non-described genes (Rauyaree et al. 2001). Clearly, the cloning and sequencing portion of genomics projects is relatively straightforward. Finding the function of the genes will be a great challenge in the coming decade. -4-
FOREST PATHOLOGY: FROM GENES TO LANDSCAPES
Population Genomics
Recombination
This so-called mini-sequencing makes it possible to perform population genomics studies by targeting SNP in genes of interest, or in some cases entire pathways. Surveying the occurrence of thousands of mutations in hundreds of genes with population parameters, including geographic origin, host specialization, and intensity of host selection should yield a wealth of data if applied to forest pathology. Genome-wide population studies have been conducted so far on systems where entire genomes and large SNP databases are available. In the human genome, associations between mutations and genetic disorders are being found in large genome-wide population studies (Pletcher and Stumpf 2002). Novel approaches based on multilocus analysis of selection allow looking at selection on a large number of genes (Navarro & Barton 2002; Barton and Navarro 2003). Increasingly the comparison of genomes is becoming possible; this can lead to the search of genes or regulatory elements that are unique to a group of organisms (Kellis et al. 2003). This can be particularly powerful when several genomes of related organisms with different ecological niches or specialization are compared (Naruya 2002). This possibility to scan entire genomes is revolutionizing population and evolutionary studies. In the yeast, this has lead to the discovery that genome variability was biased toward the ends of chromosomes and was more likely to be found in genes with roles in fermentation or in transport (Winzeler et al. 2003). In Candida albicans, whole-genome scans of populations evolving under different selection pressure have revealed that fungicide resistance evolved independently in different populations but represented a common program of adaptation (Cowen et al. 2002). Population genomics studies allow the detection of signature of selection. In Drosophila, genes involved in the immune response had patterns of variation that differed significantly from neutrality; the patterns of differentiation revealed differences in evolutionary forces (Clark and Wang 1997). In pine rusts, the higher ratios of non-synonymous to synonymous mutations in several genes, including a P450, a putative pathogenicity gene than in other genes could indicate a relaxation of purifying selection, or positive selection (Joly et al. 2003). Co-evolution of hosts and pathogens can be studied by surveying entire pathways of genes believed to be involved in host-pathogen recognition, defense reactions, or pathogenicity (Gold et al. 2001). Genome-wide scans can be conducted to study specialization of entire genomes and move from phylogeny based on single or a few genes to comparison of entire genomes. Phylogenomics can also look at the evolution of families of genes across taxa, and yield important information on adaptation and speciation processes (Lawrence et al. 2002; Sicheritz-Ponten and Andersson 2001).
Genome-wide scans facilitate the sensitive detection of recombination events. These techniques have revealed a surprisingly wide spectrum of genetic diversity in bacterial populations (Joyce et al. 2002). Knowledge of recombination is essential for locating pathogenicity loci by association studies or population genetic approaches (Awadalla 2003). Cryptic sexual reproduction has been detected using multiple gene genealogies and looking for incongruence among trees, a signature of recombination (Burt et al. 1996). Recombination complicates the use of phylogenetic approaches to estimate evolutionary parameters such as selection pressures. This is affecting our understanding of population structures and of how these affect host pathogen relationships. Population genomics analysis is allowing us to query genomes and test hypotheses about genomic diversity and recombination. Are there advantages to high genomic diversity? Are there disadvantages? The paradigm of more diversity meaning more adaptability could be properly tested using these approaches. Certainly, we find successful pathogens with very narrow genetic basis, but we also find other successful pathogens with broad genetic basis. The analysis and comparison of sister species with different genomic diversity could be exploited to test these hypotheses.
Genomes and the Environment Host and pathogen genomes are interacting with one another in complex ways that are intricately linked with their environment and their ecosystems. The power of genomics studies is that the host and pathogen genomes can be scanned simultaneously and the “pathosystems genome” can be studied at the population level. Are particular host-pathogen genomic combinations more likely to occur than is expected by chance alone? Are other combinations selected against? One of the main advantages that forest pathologists have over molecular biologists working with model systems or plant pathologists working on agricultural pathosystems is the great diversity found in forest pathosystems. This diversity is found at the host level, particularly in natural forest situations, but also in plantations of conifers where clonality is not the rule. Diversity is also found at the landscape level. Some pathogens can occur across vast sections of landscapes. For example, white pine blister rust occurs from the Atlantic Coast to the Pacific Coast, at low elevation as well as high elevation, and can infect several species of 5-needle pines. Relevant questions include whether or not these various environments and hosts are shaping the rust genome. It is possible to use population genomics to detect incipient speciation in this fungus. Populations of C. ribicola are differentiated along an east-west line, with apparent -5-
CHAPTER ONE
In the meantime, forest pathologists should fully exploit the benefit of having their own model system and generating the critical mass that will allow the dissection of host-pathogen interactions at the molecular level. Human genomics has the Drosophila and worm model systems, plant genomics has the Arabidopsis, and microbiologists have E. coli, S. serevisae, and C. albicans. Poplar appears to be the model tree system of choice as its genome is already sequenced and several poplar-associated microorganisms will soon be sequenced. This should deepen our understanding of the interactions between trees and pathogens and allow us to decipher the genomic signature that differentiates symbiotic from parasitic interactions.
panmixis within these epidemiological units (Et touil et al. 1999; Hamelin et al. 2000). Looking for the signature of selection at the genome level should be informative to test hypotheses concerning the effect of different hosts and environments on shaping the genome of this rust. Evolution and population genomics studies are also benefiting from the work with model systems. The entire genome of a yeast, Candida albicans, can be monitored when the fungus is exposed to various doses of an anti-microbial agent. It appears that resistance to the anti-microbial agent always occurs in experimental populations, but the pathways to this resistance can differ among populations (Cowen et al. 2001; Cowen et al. 2000; Cowen 2001). Is a similar situation also true for host resistance to tree pathogen?
References
Metagenomes
Awadalla, P. 2003. The evolutionary genomics of pathogen recombination. Nat. Rev. Genet. 4 (1):50-60. Beadle, G.W. , and E.L. Tatum. 1941. Genetic control of biochemical reaction in Neurospora. Proc. Natl Acad. Sci. USA 27:49-506. Bowling, Scott A., Ailan Guo, Hui Cao, A. Susan Gordon, Daniel F. Klessig, and Xinnian Dong. 1994. A mutation in arabidopsis that leads to constitutive expression of systemic acquired resistance. Plant Cell 6 (12):1845-1857. Burt, A., D. A. Carter, G. L. Koenig, T. J. White, and J. W. Taylor. 1996. Molecular markers reveal cryptic sex in the human pathogen Coccidioides immitis. Proc. Natl. Acad. Sci. USA 93 (2):770-3. Clark, A. G., and L. Wang. 1997. Molecular population genetics of Drosophila immune system genes. Genetics 147 (2):713-724. Claverie, J. M. 2001. Gene number. What if there are only 30,000 human genes? Science 291 (5507):1255-7. Cowen, L. E., L. M. Kohn, and J. B. Anderson. 2001. Divergence in fitness and evolution of drug resistance in experimental populations of Candida albicans. J. Bacteriol. 183 (10):2971-8. Cowen, L. E., D. Sanglard, D. Calabrese, C. Sirjusingh, J. B. Anderson, and L. M. Kohn. 2000. Evolution of drug resistance in experimental populations of Candida albicans. J Bacteriol 182 (6):1515-22. Cowen, LE. 2001. Predicting the emergence of resistance to antifungal drugs. FEMS MICROBIOLOGY LETTERS 204 (1):1 - 7. Davis, R. H., and D. D. Perkins. 2002. Timeline: Neurospora: a model of model microbes. Nat. Rev. Genet. 3 (5):397-403. Delaney, T. P., L Friedrich, and J. A. Ryals. 1995. Arabidopsis signal transduction mutant defective in chemically and biologically induced disease resistance. Proc. Nat. Acad. Sci. USA 92 (14):6602-6606. Delserone, L M, K McCluskey, D E Matthews, and H D Vanetten. 1999. Pisatin demethylation by fungal pathogens and nonpathogens of pea: Association with pisatin tolerance and virulence. Physiol. and Mol. Plant Path. 55 (6):317-326. Devaux, Frederic, Philippe Marc, and Claude Jacq. 2001. Transcriptomes, transcription activators and microarrays. FEBS Letters 498 (2-3):140-144. Et touil, K., L. Bernier, J. Beaulieu, J. A. Berube, A. Hopkin, and R. C. Hamelin. 1999. Genetic structure of Cronartium ribicola populations in eastern Canada. Phytopathology 89 (10): 915-919. Fleissner, A., C. Sopalla, and K. M. Weltring. 2002. An ATPbinding cassette multidrug-resistance transporter is necessary for tolerance of Gibberella pulicaris to phytoalexins and
It may be necessary now to think of genomes, not in isolation within the boundaries of nuclei and organelles belonging to single species, but as metagenomes. Culture-based methods are revealing only a fraction of the microbial communities present in soils. DNA-based methods are making it possible to describe soil and rhizosphere microbial communities using culture-independent methods. This allows us to investigate the linkage of organisms to ecosystem function (Kent and Triplett 2002). By using molecular phylogenomics, DNA micro arrays, functional genomics, and in situ activity measurements, we can increase our understanding of the structure and function of ecosystems, and the interactions that occur within them (Torsvik and Ovreas 2002). This concept can be applied to forest pathology. The host and pathogen genomes can be seen as interacting genomes or an “interactomes” that will carry its own molecular signature and will be modulated by selective and environmental factors. This exciting new prospect of being able to scan large tracks of the genome, including mutations with known phenotypic effects, at the population level is opening new doors. In the coming decade, it should become possible to better understand the relationship between genomes and their environment.
Conclusion The last quantum leap that allowed the current advances in molecular biology is the advent of the polymerase chain reaction. Among the next quantum leaps expected to revolutionize the way we look at genomics are: the possibility to cheaply genotype hundreds of thousands of SNPs in thousands of individuals; the possibility to cheaply sequence entire genomes in a few days. This should truly open the possibility to compare and match entire genomes of individuals under different adaptive and selection scenarios and highlight their differences and similarities. -6-
FOREST PATHOLOGY: FROM GENES TO LANDSCAPES
virulence on potato tubers. Mol. Plant Microbe Interact. 15 (2):102-8. Ge, Hui, Zhihua Liu, George M Church, and Marc Vidal. 2001. Correlation between transcriptome and interactome mapping data from Saccharomyces cerevisiae. Nature Genetics 29 (4):482-486. George, H. L., and H. D. VanEtten. 2001. Characterization of Pisatin-Inducible Cytochrome P450s in Fungal Pathogens of Pea That Detoxify the Pea Phytoalexin Pisatin1. Fungal Genetics and Biology 33:37-48. Glazebrook, Jane, and Frederick M Ausubel. 1994. Isolation of phytoalexin-deficient mutants of Arabidopsis thaliana and characterization of their interactions with bacterial pathogens. Proc. Nat. Acad. Sci. USA 91 (19):8955-8959. Gold, S. E., M. D. Garcia-Pedrajas, and A. D. Martinez-Espinoza. 2001. New (and used) approaches to the study of fungal pathogenicity. Annu. Rev. Phytopathol 39:337-65. Hamelin, R C, R S Hunt, B W Geils, G D Jensen, V Jacobi, and N Lecours. 2000. Barrier to gene flow between eastern and western populations of Cronartium ribicola in North America. Phytopathology 90 (10):1073-1078. Hamelin, R. C., R. S. Ferris, and L. Shain. 1994. Prediction of poplar leaf rust epidemics from a leaf-disk assay. Can. J. For. Res. 24:2085-2088. Han, YN., XG. Liu, U. Benny, HC. Kistler, and HD. VanEtten. 2001. Genes determining pathogenicity to pea are clustered on a supernumerary chromosome in the fungal plant pathogen Nectria haematococca. Plant J. 25 (3):305 - 314. Joly, D.L., L. Bernier, S.F. Covert, and R. C. Hamelin. 2003. Molecular analysis of Cytochrome P-450 genes from the pine stem rusts. Phytopathology 93:S41. Holter, Neal S, Madhusmita Mitra, Amos Maritan, Marek Cieplak, Jayanth R Banavar, and Nina V Fedoroff. 2000. Fundamental patterns underlying gene expression profiles: Simplicity from complexity. Proc. Nat. Acad. Sci. USA 97 (15):8409-8414. Jakoby, Marc, Bernd Weisshaar, Laser Wolfgang Droege, Carbajosa-Jesus Vicente, Jens Tiedemann, Thomas Kroj, and Francois Parcy. 2002. bZIP transcription factors in Arabidopsis. Trends in Plant Science 7 (3):106-111. Joly, D.L. , L. Bernier, S.F. Covert, and R. C. Hamelin. 2003. Molecular analysis of Cytochrome P-450 genes from the pine stem rusts. Phytopathology 93:S41. Jouanin, L., A.C.M. Brasileiro, J. C. Leplé, G. Pilate, and D. Cornu. 1993. Genetic transformation: a short review of methods and their applications, results and perspectives for forest trees. Ann. Sci. For. 50:325-336. Joyce, E. A., K. Chan, N. R. Salama, and S. Falkow. 2002. Redefining bacterial populations: a post-genomic reformation. Nat. Rev. Genet. 3 (6):462-73. Kamoun, S, S Dong, W Hamada, E Huitema, D Kinney, W R Morgan, A Styer, A Testa, and T A Torto. 2002. From sequence to phenotype: Functional genomics of Phytophthora. Can. J Plant Path. 24 (1):6-9. Kent, A. D., and E. W. Triplett. 2002. Microbial communities and their interactions in soil and rhizosphere ecosystems. Annu. Rev. Microbiol 56:211-36. Kunkel, Barbara N. 1996. A useful weed put to work: Genetic analysis of disease resistance in Arabidopsis thaliana. Trends in Genetics 12 (2):63-69. Kwok, P. Y. 2001. Methods for genotyping single nucleotide polymorphisms. Annu Rev Genomics Hum. Genet. 2:235-58. Lawrence, CJ., RL. Malmberg, MG. Muszynski, and RK. Dawe. 2002. Maximum likelihood methods reveal conservation of function among closely related kinesin families. J. Mol. Evolution. 54 (1):42 - 53. Liu, X.G., U. Benny, H.C. Kistler, and H.D. VanEtten. 2001. Genes determining pathogenicity to pea are clustered on a
supernumerary chromosome in the fungal plant pathogen Nectria haematococca. Plant J. 25 (3):305 - 314. Lorenz, M. 2002. Genomic approaches to fungal pathogenicity. Curr. Opin. Microbiol. 5 (4):372. Newcombe, G. 1998. Association of Mmd1, a major gene for resistance to Melampsora medusae f. sp. deltoidae, with quantitative traits in poplar rust. Phytopathology 88 (2):114-121. Oliver, Stephen G. 2002. Functional genomics: Lessons from yeast. Philosoph. Trans. Royal Soc. of London B Biol. Sci. 357 (1417):17-23. Osbourn, Anne E. 1999. Antimicrobial Phytoprotectants and Fungal Pathogens: A Commentary. Fungal Genetics and Biology 26:163-168. Ostry, M. E., and H. S. McNabb Jr. 1985. Susceptibility of Populus species and hybrids to disease in the north central USA. Plant Disease 69 (9):755-757. Pennisi, E. 2001. Genomics. New genomes shed light on complex cells. Science 292 (5520):1280-1. Pinon, J. 1992. Frequency and evolution of Melampsora-laricipopulina Klebahn races in northwestern France. Annal. des Sci. Forest. Paris 49 (1):1-15. Pinon, J, G. Newcombe, and G. A. Chastagner. 1994. Identification of races of Melampsora larici-populina, the Eurasian poplar leaf rust fungus, on Populus species in California and Washington. Plant Disease 78 (1):101. Pletcher, S. D., and M. P. Stumpf. 2002. Population genomics: ageing by association. Curr Biol 12 (9):R328-30. Rauyaree, Payungsak, Woobong Choi, Eric Fang, Barbara Blackmon, and Ralph A Dean. 2001. Genes expressed during early stages of rice infection with the rice blast fungus Magnaporthe grisea. Mol. Plant Path. 2 (6):347-354. Santucci, Annalisa, Lorenza Trabalzini, Lucia Bovalini, Elisa Ferro, Paolo Neri, and Paola Martelli. 2000. Differences between predicted and observed sequences in Saccharomyces cerevisiae. Electrophoresis 21 (17):3717-3723. Sicheritz-Ponten, T., and S.G.E. Andersson. 2001. A phylogenomic approach to microbial evolution. Nucleic Acids Res. 29 (2):545 - 552. Silva, A C R da , J A Ferro, F C Reinach, C S Farah, L R Furlan, and others. 2002. Comparison of the genomes of two Xanthomonas pathogens with differing host specificities. Nature 417 (6887):459-463. Sweigard, J. A., and D. J. Ebbole. 2001. Functional analysis of pathogenicity genes in a genomics world. Curr Opin Microbiol 4 (4):387-92. Thomma, B. P., B. P. Cammue, and K. Thevissen. 2002. Plant defensins. Planta 216 (2):193-202. Torsvik, V., and L. Ovreas. 2002. Microbial diversity and function in soil: from genes to ecosystems. Curr. Opin. Microbiol. 5 (3):240-5. Waugh, Mark, Peter Hraber, Jennifer Weller, Yihe Wu, Guonghong Chen, Jeff Inman, Don Kiphart, and Bruno Sobral. 2000. The Phytophthora Genome Initiative database: Informatics and analysis for distributed pathogenomic research. Nucleic Acids Res. 28 (1):87-90. Xu, Jin Rong, and Chaoyang Xue. 2002. Time for a blast: Genomics of Magnaporthe grisea. Mol. Plant Path. 3 (3):173-176. Yoder, OC., and BG. Turgeon. 2001. Fungal genomics and pathogenicity. Current Opinion Plant Biology 4 (4):315 - 321. Yu, Xueshu, Abul K. M. Ekramoddoullah, Doug W. Taylor, and Nina Piggott. 2002. Cloning and characterization of a cDNA of cro rI from the white pine blister rust fungus Cronartium ribicola. Fungal Genetics and Biology 35 (1):53-66. Zeng, Qiandong, Arturo J. Morales, and Guillaume Cottarel. 2001. Fungi and humans: Closer than you think. Trends in Genetics 17 (12):682-684. Zhang, Shuqun, and Daniel F. Klessig. 2001. MAPK cascades in plant defense signaling. Trends in Plant Science 6 (11):520-527. -7-