Weedy and Invasive Plant Genomics

Weedy and Invasive Plant Genomics

Weedy and Invasive Plant Genomics

Edited by C. NEAL STEWART, JR. Racheff Chair of Excellence in Plant Molecular Genetics Professor, Department of Plant Sciences University of Tennessee

A John Wiley & Sons, Inc., Publication

Edition first published 2009 © 2009 Blackwell Publishing Chapters 4, 8, and 14 remain with the U.S. government. Blackwell Publishing was acquired by John Wiley & Sons in February 2007. Blackwell’s publishing program has been merged with Wiley’s global Scientific, Technical, and Medical business to form Wiley-Blackwell. Editorial Office 2121 State Avenue, Ames, Iowa 50014-8300, USA For details of our global editorial offices, for customer services, and for information about how to apply for permission to reuse the copyright material in this book, please see our website at www.wiley.com/ wiley-blackwell. Authorization to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by Blackwell Publishing, provided that the base fee is paid directly to the Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923. For those organizations that have been granted a photocopy license by CCC, a separate system of payments has been arranged. The fee codes for users of the Transactional Reporting Service are ISBN-13: 978-0-8138-2288-4/2009. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. On the cover: The foreground plants are composed of mature Arabidopsis thaliana, often called “the weed” but is, in reality, not. The background plants are small horseweeds, Conyza canadensis, a weed that has proven adept at evolving resistance to many herbicides. Photo by Reginald Millwood. Library of Congress Cataloging-in-Publication Data Weedy and invasive plant genomics/edited by C. Neal Stewart, Jr.—1st. ed. p. cm. Includes bibliographical references and index. ISBN 978-0-8138-2288-4 (hardback : alk. paper) 1. Weeds—Genetics. 2. Weeds—Germplasm resources. 3. Weeds—Biological control. 4. Invasive plants—Genetics. 5. Invasive plants—Germplasm resources. 6. Invasive plants—Biological control. 7. Genomics. I. Stewart Jr., C. Neal. nSB611.W388 2009 581.6′52—dc22 2009009716 A catalog record for this book is available from the U.S. Library of Congress. Set in 10.5/12pt Times by SNP Best-set Typesetter Ltd., Hong Kong Printed in Singapore 1 2009

Dedication To weed scientists who are not afraid to use genomics to answer agricultural and biological questions, and to genomicists who see the wonders of weedy and invasive plants.

Contents

Contributors Preface Chapter 1

Chapter 2

Chapter 3

Chapter 4

Chapter 5

xi xv Why Should Weed Scientists Care About Genomics? WILLIAM K. VENCILL Genomics To A Weed Scientist Resistance Better Use Of Existing Herbicides An Introduction To Molecular Genetic And Genomic Techniques CHHANDAK BASU AND SAM R. ZWENGER Weeds As A Source Of Genes For Crop Improvement Tools And Approaches For Understanding Weediness At The Molecular Level Arabidopsis Is Not A Weed, And Mostly Not A Good Model For Weed Genomics; There Is No Good Model For Weed Genomics JONATHAN GRESSEL Introduction: Arabidopsis And Weediness Questions About Weeds—Can Arabidopsis Genomics Answer Them? The Misdirected Approach In Using Arabidopsis To Elucidate New Herbicide Targets Arabidopsis Genomics Can Help In Dealing With Transgene Flow— In A Limited Manner Lessons To Be Learned

3 3 4 8 11 11 12 25 25 27 28 29 30

Model Weeds For Genomics Research WUN S. CHAO AND DAVID P. HORVATH What Makes A Good Model Species? Leveraging From Other Models Genomics Tools For Weeds That Are Under Development

33

21st-Century Weed Science: A Call For Amaranthus Genomics PATRICK J. TRANEL AND FEDERICO TRUCCO The Amaranthus Genus Hybridization And Adaptive Evolution Herbicide Resistance Currently Available Genomic Resources Needs And Opportunities

53

34 36 44

53 61 64 71 75 vii

viii Chapter 6

Chapter 7

Chapter 8

Chapter 9

Chapter 10

Chapter 11

CONTENTS

Evolutionary Genomics Of Weedy Rice BRIANA L. GROSS AND KENNETH M. OLSEN Phenotypic Diversity Of Weedy Rice Genomic Diversity Of Weedy Rice The Origin(s) And Evolution Of Weedy Rice The Genetic Basis Of Weediness And Use Of Weedy Rice In Crop Breeding

83

Rhizomatousness: Genes Important For A Weediness Syndrome ANDREW H. PATERSON Developmental Context An Exemplary Case: Johnsongrass Dissecting The Genetic Control Of Rhizomatousness Early Insights Into The Sorghum Rhizo-Transcriptome Future Work And Potential Applications Synthesis

99

Leafy Spurge: An Emerging Model To Study Traits Of Perennial Weeds DAVID P. HORVATH AND JAMES V. ANDERSON Regulation Of Shoot Development And Growth Regulation Of Bud Dormancy Case Study: Leafy Spurge Future Work Herbicide Resistance: Target Site Mutations CHRISTOPHER PRESTON Resistance To Photosystem II-Inhibiting Herbicides Resistance To Acetohydroxyacid Synthase-Inhibiting Herbicides Resistance To Acetyl Coenzyme A Carboxylase-Inhibiting Herbicides Resistance To Glyphosate Resistance To Microtubule Assembly Inhibitors Resistance To Phytoene Desaturase Inhibitors Molecular and Genomic Mechanisms Of Non-Target-Site Herbicide Resistance JUN HU, PATRICK J. TRANEL, C. NEAL STEWART JR., AND JOSHUA S. YUAN Herbicide Application And Resistance Herbicide Classification And Resistance Non-Target Herbicide Resistance Signal Transduction Detoxification and Modification A Herbicide Defense Trait That Is Distinct From Resistance: The Evolutionary Ecology And Genomics Of Herbicide Tolerance REGINA S. BAUCOM Resistance Versus Tolerance In Weed Science Tolerance In Evolutionary Ecology Tolerance Traits And The Genomics Of Tolerance

84 85 89 94

100 101 103 105 107 109 113 113 116 117 122 127 128 131 136 138 140 141 149

149 150 150 150 151 163 163 166 171

CONTENTS

Why Again Should We Focus On Tolerance, Tolerance Traits, And The Genomics Of Tolerance? Chapter 12

Chapter 13

Chapter 14

Index

The Genomics of Plant Invasion: A Case Study In Spotted Knapweed AMANDA K. BROZ AND JORGE M. VIVANCO Why Study Invasive Plant Genomics? Spotted Knapweed Life History Allelopathy And The Novel Weapons Hypothesis Genomics Resources And Approaches For Studying Spotted Knapweed

ix

172 177 177 178 180 185

Molecular Ecology Of Plant Competition DOMINIK D. SCHMIDT, MERIJN R. KANT, AND IAN T. BALDWIN Competition Signals And Their Perception By Plants Molecular Basis Of Competitively Important Traits Transcriptomic Insights Into Competitive Interactions Of Weedy Plants

197

Genomics And Weeds: A Synthesis STEPHEN O. DUKE, SCOTT R. BAERSON, AND JONATHAN GRESSEL From Fundamental Information To Practical Solutions Where Do We Go From Here?

221

198 207 211

222 241 249

Contributors

James V. Anderson

USDA-ARS Plant Science Unit 1605 Albrect Blvd. Fargo, ND 58105 [email protected]

Scott R. Baerson

Natural Products Utilization Research Unit USDA-ARS P.O. Box 8048 University, MS 38677 USA [email protected]

Ian T. Baldwin

Max Planck Institute for Chemical Ecology Hans-Knoell-Str. 8, 07745 Jena, Germany [email protected]

Chhandak Basu

School of Biological Sciences University of Northern Colorado Greeley, Colorado 80639 USA [email protected]

Regina S. Baucom

Department of Genetics University of Georgia Athens, GA 30602 USA [email protected]

Amanda K. Broz

Center for Rhizosphere Biology Department of Horticulture and Landscape Architecture Colorado State University Fort Collins, Colorado, 80523 USA [email protected]

Wun S. Chao

Plant Science Unit USDA-ARS 1605 Albrecht Blvd. Fargo, ND 58105 [email protected] xi

xii

CONTRIBUTORS

Stephen O. Duke

Natural Products Utilization Research Unit USDA-ARS P.O. Box 8048 University, MS 38677 USA [email protected]

Jonathan Gressel

Department of Plant Sciences Weizmann Institute of Science Rehovot, Israel 76100 [email protected]

Briana L. Gross

Department of Biology Box 1137 Washington University St. Louis, MO 63130 USA [email protected]

David P. Horvath

Plant Science Unit USDA-ARS 1605 Albrecht Blvd. Fargo, ND 58105 [email protected]

Jun Hu

Department of Plant Sciences University of Tennessee Knoxville, TN 37996 USA and Institute of Plant Genomics and Biotechnology Department of Plant Pathology and Microbiology Texas A&M University College Station, TX 77843 USA [email protected]

Merijn R. Kant

Max Planck Institute for Chemical Ecology Hans-Knoell-Str. 8, 07745 Jena, Germany [email protected]

Kenneth M. Olsen

Department of Biology Box 1137 Washington University St. Louis, MO 63130 USA [email protected]

Andrew H. Paterson

Plant Genome Mapping Laboratory University of Georgia Athens, GA 30602 USA [email protected]

CONTRIBUTORS

Christopher Preston

School of Agriculture Food and Wine University of Adelaide PMB 1 Glen Osmond SA 5064, Australia [email protected]

Dominik D. Schmidt Max Planck Institute for Chemical Ecology Hans-Knoell-Str. 8, 07745 Jena, Germany [email protected] C. Neal Stewart Jr.

Department of Plant Sciences University of Tennessee Knoxville, TN 37996 USA

Patrick J. Tranel

Department of Crop Sciences University of Illinois Urbana, IL 61801 USA [email protected]

Federico Trucco

Instituto de Agrobiotecnología Rosario S.A. Rosario, Argentina

William K. Vencill

Department of Crop and Soil Sciences University of Georgia Athens, GA 30602 USA [email protected]

Jorge M. Vivanco

Center for Rhizosphere Biology Department of Horticulture and Landscape Architecture Colorado State University Fort Collins, Colorado, 80523 USA [email protected]

Joshua S. Yuan

Institute of Plant Genomics and Biotechnology Department of Plant Pathology and Microbiology Texas A&M University College Station, TX 77843 USA [email protected]

Sam R. Zwenger

School of Biological Sciences University of Northern Colorado Greeley, Colorado 80639 USA [email protected]

xiii

Preface

Pat Tranel and I (and others—many of whom are authors of chapters) have been discussing the genomics of weedy plants for quite a few years now. The Plant and Animal Genome (PAG) conference of 2009 was the fourth instance we’ve held the annual workshop on weedy and invasive plant genomics. The first weedy genomics workshop was held at the 2005 Weed Science Society of America Conference, which was in Hawaii that year. I vividly recall enticing two postdocs in my laboratory with warm beaches on a cold winter ’s day. If they could successfully write and publish a review paper on the genomics of weedy plants, I promised to pay for their travel to present a paper at the conference in Hawaii that February. They did (Trends in Plant Science (2004) 9:391–398) and we all had an aloha good time. While I and many others have been interested in genes and proteins that confer interesting traits to weeds and invasive plants, there has been surprisingly little research at the genome or proteome levels. Molecular biology of weeds has progressed very slowly, which I think is mainly due to three interrelated factors. First, the weed science research community and culture is vastly different from that of plant genomics and evolutionary biology. Simply, most weed scientists don’t know about molecular genetics and most genomicists have yet to discover the world of fast-evolving weeds. I recall a conversation with a weed scientist who asked whether it was offensive to be referred to as a “gene jockey.” I reciprocated by asking him whether “nozzlehead” is a derogatory term. We each had a laugh over this exchange, realizing that the two research worlds are indeed polar opposites. Second, the sources of research funding are also quite different in each of these two areas. Much of the weed science research (done in the United States, anyway) is corporate sponsored and focused on proprietary herbicides. Plant genomicists are typically funded by the National Science Foundation (NSF) or U.S. Department of Agriculture (in the U.S.). No entity has seemed to be very interested in funding weed genomics. Third, and related to the second point, there is little information on the genomes of any weedy and invasive plant species. Being neither models nor crops, their genomes have fallen through the funding cracks. With little basal genomic information, meager funding, and a lack of collaboration between weed scientists and genomicists, the field has not yet blossomed. The situation is about to change as these two cultures have been slowly converging on the field of weed genomics, with the promise of more research to come in the near future. Funding agencies are slowly catching on to the synergies of using genomic tools to address fundamental and practical issues of weediness and invasiveness. At the same time, genomics tools, such as next generation sequencing platforms, are becoming increasingly more accessible and reasonably priced. We are now moving from talking about weed genomics to actually doing weed genomics. This new year finds me with a dataset of more than 400,000 horseweed transcriptome sequences to mine, and, indeed, a few other researchers also have recently obtained transcriptomic sequences from their own favorite weeds. It is truly an exciting time in science, and I think this book captures the scintillation of this emerging area. xv

xvi

PREFACE

My motivation for editing this book is that young scientists will discover the fun of combining weeds and genomics. Having never been burdened with either the label of “weed scientist” or “genomicist,” emerging young scientists interested in weedy plant genomics will be free to define their own field of study. After a sizable amount of genomic information is made available, which should begin to happen in exponential fashion very soon, many important biological questions will become increasingly accessible and fruitful ground for many scientific careers. I firmly believe that as the science matures, increasing funding opportunities that are needed to accelerate knowledge acquisition will emerge, which will in turn be translated to applications for the more effective control of weedy and invasive plants. Neal Stewart

Weedy and Invasive Plant Genomics

1

Why Should Weed Scientists Care About Genomics? William K. Vencill

Genomics To A Weed Scientist

Genomics does not provide any information that cannot be obtained by more traditional genetic approaches. However, traditional approaches analyze one or a few genes at a time. Among other things, genomics seeks to examine the response of the entire genome to a given stimuli —in one of the most pertinent cases in weed science, an herbicide. A better understanding and use of these technologies potentially allows the weed scientist to find new herbicides and herbicide mechanisms-of-action and extend the use of current herbicide mechanisms-ofaction by overcoming weed resistance, developing crop resistance, or making them more efficacious. Weed scientists and those interested in controlling invasive plants face many challenges concerning available control techniques. When examining chemical control of weeds, there are three major issues facing weed scientists: (1) resistance of weeds to existing herbicide mechanisms-of-action, (2) loss of older herbicides, and especially specific herbicide mechanisms-of-action (MOA) through regulatory or economic means, and (3) lack of new herbicides, and especially herbicides with novel mechanisms-of-action. When we examine the past decade in weed science, we see a revolution in weed control through the introduction of herbicide-resistant crops. Currently, in the U.S., between 50% and 75% of the major grain, oilseed, and fiber crops have either an herbicide resistance trait or an insecticide trait, or in some cases, both (Dill et al. 2008). The rapid adoption of herbicideresistant crops has had many positive impacts on weed management, but it has also led to some troubling trends. The widespread reliance on a few herbicides for weed control in the major row crops has led to downward price pressure on other herbicides, which has contributed to industry consolidation. The lower return on investment of newer herbicides has been a contributing factor in fewer herbicide introductions and the lack of new herbicide mechanismsof-action since 1993 (Kraehmer et al. 2007). In some major row crops, such as soybeans and cotton, there has been an overreliance on one herbicide for weed control that has created high selection pressure for resistance development. The conundrum is thus: if widespread resistance occurs to the most commonly used herbicides and we have fewer older herbicides available because of regulatory issues and economic reasons, and there are fewer herbicides and new herbicide mechanisms-of-action in the pipeline, are we far from having a scenario in which we have no herbicides available for certain crops? Furthermore this scenario is building at a time of increasing demand because of population growth, more affluence in the developing world with its modernization of agriculture, and biofuel demand. For a weed scientist, it is obvious that many of the technologies such as screening thousands of organic compounds a year to discover a potential herbicide, which has provided new herbicides and herbicide mechanisms-of-action in the past, might not be viable in the future. Many genomic technologies could provide methods of obtaining the new herbicides and even new classes of herbicides that are the cornerstone of modern weed control. 3

4

WEEDY AND INVASIVE PLANT GENOMICS

Resistance

The first case of herbicide resistance in a weed was documented in the late 1960s, when common groundsel (Senecio vulgaris L.) was found to be resistant to triazine herbicides (Heap 2008). Herbicide resistance in weeds has grown dramatically; there are now 319 cases of herbicide-resistant biotypes in 185 species covering all herbicide mechanisms-of-action (Heap 2008). See Chapters 9 and 10 for more information on herbicide resistance. Herbicide resistance in weeds has had a major impact on herbicide use patterns. As a result, the loss of effective herbicides for weed control has the potential to negatively impact the production of certain crops in some areas where the prospect of having no available herbicides available for weed control is very real. One example is cotton in the southeastern United States, where glyphosate-resistant Palmer amaranth (Amaranthus palmeri) has been confirmed in twenty-nine Georgia counties since 2005 (Stanley Culpepper, personal communication). In addition, acetolactate synthase (ALS) herbicide resistance to Palmer amaranth is present in sixty-one Georgia counties. There is sizable overlap in these same counties and there have been observations of double-resistant Palmer amaranth biotypes to both glyphosate- and ALSinhibiting herbicides (Stanley Culpepper, personal communication). The presence of ALS- and glyphosate-resistant Palmer amaranth will leave cotton growers with few options for control. The current practice of using protoporphyrinogen oxidase (PPO or PROTOX) -inhibiting herbicides comes with the concern that if PPO-resistant Palmer amaranth develops, there would be no available herbicides for controlling Palmer amaranth in cotton. The consolidation process in the agrochemical industry (Copping 2003) has severely reduced overall research and development expenditures. In 2005, there were only eleven companies with significant efforts in crop protection research and development, compared with thirty-five companies in 1985 (Rüegg 2007). Coupled with a loss of herbicide MOA to regulatory action (e.g., organic arsenicals in the U.S.; substituted ureas in Europe), widespread resistance to ALS herbicides (Heap 2008), and large increases in resistance to glyphosate, we are facing a crisis of herbicide availability. To some extent, weed scientists are the victims of their own success. In a survey of growers in Indiana conducted by Johnson and Gibson (2006), 65% of growers reported that they were not concerned about glyphosate resistance problems (now in the future) because new herbicide products would be introduced to replace glyphosate when it was no longer effective because of resistant weeds. The intensive use of a single herbicide such as glyphosate in glyphosate-resistant crops is likely to accelerate the evolution of herbicide resistance. This is especially true if a single herbicide is used in various crops grown in the same rotation, as is currently the case with glyphosate in herbicide-resistant crops in the U.S. (Duke and Powles 2008). In addition, regulatory requirements are increasing worldwide (Rüegg et al. 2007). This has encouraged industry to focus development in “safe herbicide harbors,” or those chemistries that have proven records of positive environmental and toxicological profiles to make the registration process easier, such as ALS or acetyl-CoA carboxylase (ACCase) inhibitors (Rüegg et al. 2007). This compounds the problem when resistance to these chemistries becomes widespread. This can partly explain small variations in chemistries and MOA among recently launched herbicides.

Better Understanding of Resistance

Herbicide resistance can occur via an altered target site (see Chapter 9) or non-target site resistance such as enhanced metabolism, or an exclusion mechanism such as decreased foliar

WHY SHOULD WEED SCIENTISTS CARE ABOUT GENOMICS?

5

uptake or translocation out of treated leaves (see Chapter 10). There are several cases of nontarget herbicide resistance such as glyphosate resistance in horseweed (Conyza canadensis). One particularly intriguing case in which genomics could be effective in characterizing glyphosate resistance is Palmer amaranth. It appears that the EPSPS gene coding for a sensitive enzyme has been duplicated, perhaps over 100 times, leading to very high levels of EPSPS enzyme and resistance (Gaines et al. 2009). In these cases, genomic tools could be powerful in elucidating genes and proteins responsible for resistance and altered translocation, and possibly finding ways to overcome the resistance mechanism to restore utility to the herbicide. Many of the non-target site herbicide resistance cases have been established using enzyme assays and metabolite analysis, but few resistance genes have been cloned and characterized from weeds (Basu et al. 2004). Many important questions regarding the mechanisms of nontarget herbicide resistance have not been answered. For instance, does resistance result from gene transcriptional regulation, an increase in enzyme affinity, altered substrate specificity, or combinations thereof? Does increased enzyme activity involve a site mutation? A functional genomics approach has recently been successfully applied in herbicide resistance studies and led to the identification of several resistance genes (Gachon et al. 2005; Zhen and Singh 2001). In a review, Yuan et al. (2007) proposed an integrated functional genomics approach to identify genes involved with non-target herbicide resistance in weed species. Cytochrome P450, glutathione S-transferase, and ABC transporter gene families have been implicated in non-target herbicide resistance. Genomic technologies might allow the identification of weed taxa with propensity for resistance so growers might be advised to use alternative weed management strategies or agronomic practices (Weller et al. 2001).

New Herbicides And Herbicide Mechanisms-of-action?

In 1960, the number of compounds that had to be screened to yield one single product was 10,000; by 2000 the number had increased to 140,000 (Stenzel 2004). In the 1980s, around 10,000 compounds could be screened to yield a compound showing activity in greenhouse assays. This number increased to 30,000 in the 1990s and reached 100,000 in 1998. Since 1991, when sulcotrione, a 4-hydroxyphenyl-pyruvate-dioxygenase (4-HPPD) -inhibiting herbicide, was introduced in the marketplace, no new herbicide mechanism-of-action has been commercialized (Rüegg 2007). In contrast, between 1970 and 1985, ten new herbicide mechanisms-of-action were introduced in Europe and the U.S. Since the discovery of the auxinic herbicides in the late 1940s, empirical screening has led to the commercialization of nearly 270 active ingredients, representing seventeen mechanismsof-action (Lein et al. 2004). Of these, approximately 50% act on one of three target sites: photosystem II, ALS, and protoporphyrinogen oxidase PPO. Ten of the 270 active ingredients account for 45% of total market value (Lein et al. 2004) and glyphosate accounts for 30% of herbicide sales worldwide and 20% of all pesticide sales. An overreliance on a few herbicides has led to an explosive growth in herbicide resistance worldwide. Agrochemical companies have shifted to a strategy that is driven by in vitro testing rather than whole plant screening of herbicide candidates. Most of the known herbicide MOA involve enzyme inhibition and only a handful disrupt other process such auxin response or cell division. Approximately 20% of the genes in Arabadopsis and rice code for enzymes. Does this mean herbicide targets are restricted to a small subset of plant genes or have previous approaches simply favored their discovery (Lein et al. 2004)? Since the early 1990s, agrochemical

6

WEEDY AND INVASIVE PLANT GENOMICS

companies have shifted from whole plant screening to more target-based approaches. Initially, other enzymes of existing herbicide targets were examined with limited success (Abell 1996). Researchers have examined “key” or “limiting” proteins in essential plant processes, also with limited success. Another approach that is useful in herbicide discovery is to provide evidence that the gene that encodes the target protein is essential to plant growth and development. Abell (1996) suggested that a protein is a suitable target site if inhibition of 60% to 80% of its activity leads to severe growth reduction. The accumulation of large amounts of sequence information from the late 1990s onward from expressed sequence tag (EST) sequencing and full genome sequencing made it possible to use unbiased and genome-wide strategies to identify targets. Unfortunately, the function of more than 30% of genes from completely sequenced species is still unknown or incomplete. Jun et al. (2002) initiated a study in Arabadopsis in which 1,000 antisense lines were created using cDNAs that had been randomly selected. This study indicated that 1% to 2% of Arabidopsis genes (say, a few hundred genes) encode potential herbicide targets. However, the numbers of genes identified were too small to allow any firm conclusions about their distribution in different functional categories. In addition, Arabidopsis is not a weedy species (see Chapter 3) and so perhaps an examination of truly weedy species would reveal some potential targets as well. Lein et al. (2004) created a normalized cDNA library from tobacco, sequenced it, excluded redundant clones, transformed 20,000 randomly selected cDNAs in sense or antisense configuration in tobacco, scored plants for visual phenotypes, and carried out retransformation to confirm the result. As of 2004, about 10,000 genes had been put through the process, resulting in forty-six potential herbicide targets. Genes whose partial inhibition leads to chlorosis, necrosis, and concomitant growth defects have been discovered in this process. They contain known herbicide targets (e.g. glutamine synthetase) and genes for which antisense (lack of expression) has already been reported to mimic herbicide phenotypes (e.g., Rubisco and ferredoxin:NADP oxidoreductase [Stitt 1999 and Palatnik et al. 2003]). About half of these genes identified as encoding herbicide targets are annotated as enzymes. The remaining genes have an extremely imprecise annotation, including a quarter of which with no known function. This finding indicates that current herbicide targets found by traditional approaches only represent a small percentage of potential targets. More recently, some groups have initiated programs to create large numbers of RNAi lines. RNAi produces a partial inhibition of gene expression that generally leads to higher suppression compared with antisense methods. Virusinduced gene silencing methods have the potential to speed up the genetic identification of potential herbicide targets. Within a few years, lists of hundreds of potential herbicide targets might be formulated. There will likely be a premium on the speed and effectiveness with which the next two stages of agrochemical discovery pipeline (role of the protein and development of high-throughput assay) will be developed. Genomics has allowed for the discovery of many genes with unknown functions. Herbicide research could possibly contribute to elucidating the function of these gene products while possibly providing new active ingredients for the marketplace. There are historical parallels in which herbicide research led to much of what we know about photosynthetic function through tracers, inhibitors, and resistant plant species. Bioinformatics tools will allow the common metabolic response (Ott et al. 2003) of the plant to be profiled and compared with known herbicide MOA so that enzymes that are targeted by potential herbicides can be viewed by known pathways and biological processes (Thimm et al. 2004). Genomics should allow high-throughput testing of target-based screening based on genes that are affected by a test compound. Currently, most known herbicides interfere with the synthesis of an essential compound by inhibiting a rate-limiting step in a biosynthetic pathway. The use of genomic technologies

WHY SHOULD WEED SCIENTISTS CARE ABOUT GENOMICS?

7

could allow the discovery of target sites that do not have enzymatic function or any known function at all. These possible target sites could be regulatory proteins or components of signal transduction pathways. Genomics has the potential to increase the efficiency of discovering new herbicide mechanisms-of-action. This can be achieved by two general strategies: (1) reverse genetics in which genes are knocked out in a model organism resulting in a reduced-function or nonfunctional target site leading to a detectable and relevant phenotype, and (2) forward genetics in which a model organism is treated with an active compound with an unknown mechanismof-action and the molecular target is subsequently uncovered (Stenzel 2004; Egner et al. 2005). To identify a target by reverse genetics, it is necessary to generate knock-out mutants that display mimicking effects or lethal phenotypes similar to an herbicide treatment. Such mutants can be generated via multiple genomics methods such as chemical mutagenesis, transposon mutagenesis, antisense down regulation, sense-cosuppression, ribozymes, or RNAi technology (Stenzel 2004). Death of the plant from the disabling of a specific protein might confirm a potential herbicide target site. The ideal target site needs to fulfill at least the criteria of what Stenzel (2004) calls (1) essentiality or proven by genomic knockout in the model organism, (2) druggability or discovery of small molecules binding to the target protein, (3) lethality proven by in vivo activity in a subsequent in vivo screen, and (4) proof by commercial success. Klaus Grossmann (2005) described a physiognomic approach to herbicide discovery in which test compounds are compared to known herbicide mechanisms-of-action on several whole plant levels, including functional gene identification, gene expression, protein profiling, histochemistry, and analysis via metabolite profiling. Gene expression profiling (genomics) and metabolic profiling (metabolomics) can allow fast and reliable detection of known herbicides’ mechanisms-of-action and clear identification and classification of herbicides with an unknown mechanism-of-action (Ott et al. 2003). Artificial neural networks analysis of 1H NMR spectra was used to determine changes in the metabolic profile (or metabolome) of maize caused by herbicide application. Ott et al. (2003) used this method to classify nineteen distinct herbicide mechanisms-of-action in maize. Genomic, metabolomic, and proteomic technology can also be used to analyze potential changes in crop plants from genetic transformation. This could be used to allay consumer fears over genetically modified crops in regard to the nutritional content or allergenicity of a modified crop (Wheelock and Miyagawa 2006). Approaches for mining and exploiting genomic information that rely solely on genetic or molecular techniques typically do not provide sufficient confidence that a potential site of interest can be effectively modulated by chemical intervention. For example, the effect of a genetic knockout of a gene may have more impact than the impact of a chemical inhibitor of the protein. Conversely, genetic redundancy may underestimate the potential effects on an inhibitor that can interact with two or more members of a target encoded by a gene family. The design of new chemistry that interacts with a novel site of interest predicted from genetic evidence requires significant resources and a level of risk that is typically not taken by companies interested in pesticide discovery. There is a great need for shortcuts in this discovery process that can take advantage of genomic information while simultaneously providing insights into chemistry that can effectively interact with new sites-of-action. A hybrid or chemical genetic approach may be the most practical route (Walsh 2007). Chemical genetics can be defined as the use of small molecules to mimic the effect of genetic mutations in a biological system of interest (Stockwell 2000), allowing the production of a specific phenotype in a treated organism or cell that can be investigated in much the same way as a genetic mutant. This approach allows for compounds to be applied and removed at specific times and tissues to rapidly produce their effects, with their effects being readily titratable in

8

WEEDY AND INVASIVE PLANT GENOMICS

a dose response. The use of chemistry to interrupt or modulate key biological processes can produce “phenotypes” with distinct physiological impairment or lethality. In this manner, the principles of forward genetic screening for distinct and desired phenotypes can be readily used to organize a chemical genetic approach for pesticide discovery that takes advantage of both chemical screening and genomic resources. Genomic screening in model organisms can be used to identify phenotypes of interest that might allow the discovery of potential novel target sites. However, no obvious chemical starting points are available now. Validation that a target has the potential to be chemically modulated can be difficult to achieve and might require considerable resources with little chance of return of a commercial product. In addition, the barrier of translating in vitro results to in vivo activity can be difficult to overcome. A chemical genetic approach combines the use of an organized chemical library with phenotype screens and robust target identification to produce novel targets of interest coupled with interacting chemistry (Grossmann 2005). This approach requires more upstream tools than other approaches. There are three components of a chemical genetic process to uncover novel sites of herbicide action: chemical libraries, phenotype screens, and target site identification.

Better Use Of Existing Herbicides Herbicide Safeners

Herbicide safeners are chemicals that reduce herbicide toxicity to crop plants via a physiological mechanism, usually by enhancing herbicide metabolism. They can be used to examine systemwide effects of an herbicide application on a target species. For example, Castro et al. (2005) treated grapevine with flumioxazin and found that thirty-three distinct proteins had altered synthesis patterns compared with untreated plants. These proteins included a diverse range of functions including photosynthesis-related proteins and antioxidant systems, allowing an overview of the systemic effects of the herbicide application. Zhang and Reichers (2004) used a similar approach to examine the influence of the herbicide safener fluxofenim on the chloroacetamide herbicide dimethenamid in wheat. They found that the safener caused eighteen proteins to be induced, including fifteen glutathione-S-transferase (GST) subunits and three proteins with known roles in glycolysis and the Krebs cycle. Herbicide safeners were shown to induce GSTs and glucosyltransferases in maize and Arabidopsis (Edwards et al. 2005). This could be used to differentiate safener use in a wide range of crop species.

Surfactants

Pesticide surfactants are chemicals that improve the emulsifying, dispersing, spreading, and wetting properties of herbicides, improving their foliar uptake. Madhou et al. (2006) examined the role of surfactants on plant gene expression in Arabidopsis. The expression of 169 genes were altered within one hour after plants were treated with 0.2% volume/volume of surfactant NUK1026. Functional category analysis of these genes revealed that the largest categories included metabolism, physiological processes, transport, protein metabolism, response to stimulus, and transcription. Genes coding for cytochrome P450 and GST proteins were unregulated as were enzymes involved in 1-aminocyclopropane-1-carboxylate synthase genes for ethylene production.

WHY SHOULD WEED SCIENTISTS CARE ABOUT GENOMICS?

9

Summary

Herbicide resistance in a growing number of weed species coupled with a lack of new herbicides has brought traditional chemical weed control programs to a crossroads. In the near future there could be several weed species without adequate chemical control in major row crops. However, new genomic technologies could potentially provide weed scientists with more herbicides with novel mechanisms-of-action and a better understanding of herbicide resistance, and provide techniques to improve the efficacy and crop safety of current herbicides. Thus, weed scientists are in a unique position to collaborate with genomicists in discovery research that could lead to better weed management. References Abell L (1996) Biochemical approaches to herbicide discovery: advances in enzyme target identification and inhibitor design. Weed Science 44, 734–742. Basu C, Halfhill MD, Mueller TC, Stewart CN, Jr. (2004) Weed genomics: new tools to understand weed biology. Trends in Plant Science 9, 391–398. Castro A, Carapito C, Zorn N, Magne C, Leize E, Van Dorsselaer A, Clément C (2005) Proteomic analysis of grapevine (Vitis vinifera L.) tissues subjected to herbicide stress. Journal of Experimental Botany 56, 2783–2795. Copping L (2003) The evolution of crop protection companies. Pesticide Outlook 14, 276–279. Dill G, CaJacob C, Padgette C (2008) Glyphosate-resistant crops: adoption, use and future considerations. Pest Management Science 64, 326–331. Duke S, Powles S (2008) Glyphosate: a once-in-a-century herbicide. Pest Management Science 64, 319–325. Edwards R, Del Buono D, Fordham M, Skipsey M, Brazier M, Dixon D, Cummins I (2005) Differential induction of glutathione transferases and glucosyltransferases in wheat, maize, and Arabadopsis thaliana by herbicide safeners. Zeitschrift für Naturforschung 60, 307–316. Egner U, Krätzschmar J, Kreft B, Pohlenz HD, Schneider M (2005) The target discovery process. ChemBioChem 6, 468–479. Gaines T, Preston C, Shaner D, Leach D, Chisholm S, Bulcum B, Ward S, Culpepper AS, Tranel P, Westra P (2009) A novel mechanism of resistance to glyphosate in Palmer amaranth (Amarantus palmeri). Abstracts of WSSA 49, 368. Gachon C, Langlois-Meurinne M, Henry Y, Saindrenan P (2005) Transcriptional co-regulation of secondary metabolism enzymes in Arabidopsis: functional and evolutionary implications. Plant Molecular Biology 58, 229–245. Grossman K (2005) What it takes to get a herbicide’s mode of action. Physionomics, a classical approach in a new complexion. Pest Management Science 61, 423–431. Heap I (2008) The International Survey of Herbicide Resistant Weeds. May 2008. www.weedscience.com. Johnson WG, Gibson KD (2006) Glyphosate-resistant weeds and resistance management strategies: An Indiana grower perspective. Weed Technology 20, 768–772. Jun JH, Kim CS, Cho DS, Kwak JM, Ha CM, Park YS, Cho BH, Patton D, Nam HG (2002) Random antisense cDNA mutagenesis as an efficient functional genomic approach in higher plants. Planta 214, 668–674. Kraehmer H, Schulz A, Laber B (2007) Where are the new herbicides modes of action. FarmTech 2007 Proceedings 88–97. Lein W, Börnke F, Reindl A, Ehrhardt T, Stitt M, Sonnewald U (2004) Target-based discovery of novel herbicides. Current Opinion in Plant Biology 7, 219–225. Madhou P, Raghavan C, Wells A, Stevenson TW (2006) Genome-wide microarray analysis of the effect of a surfactant application in Arabidopsis. Weed Research 46, 275–283. Ott KH, Araníbar N, Singh B, Stockton G (2003) Metabonomics classifies pathways affected by bioactive compounds. Artificial neural network classification of NMR spectra of plant extracts. Phytochemistry 62, 971–985. Palatnik J, Tognetti V, Poli H, Rodríguez R, Blanco N, Gattuso M, Hajirezaei MR, Sonnewald U, Valle EM, Carrillo N (2003) Transgenic tobacco plants expressing antisense ferredoxin-NADP(H) reductase transcripts display increased susceptibility to photo-oxidative damage. Plant Journal 35, 332–341. Powles S, Duke SO (2008) Evolved glyphosate-resistant weeds around the world: lessons to be learnt. Pest Management Science 64, 360–365. Rüegg WT, Quadranti M, Zoschke A (2007) Herbicide research and development: challenges and opportunities. Weed Research 47, 271–275.

10

WEEDY AND INVASIVE PLANT GENOMICS

Stenzel K (2004) From genes to compound discovery: Unique research platform combining innovative screening technologies. Pflanzenschutz-Nachrichten Bayer 57, 34–45. Stitt M (1999) The first will be the last and last will be first: non-regulated enzymes call the tune? In: Plant Carbohydrate Biochemistry. Bryant JA, Burrell MM, Kruger NJ, eds. BIOS Scientific, Oxford, UK, pp. 1–16. Thimm O, Bläsing O, Gibon Y, Nagel A, Meyer A, Krüger P, Selbig J, Müller L, Rhee S, Stitt M (2004) MAPMAN: a user-driven tool to display genomics data sets onto diagrams of metabolic pathways and other biological process. Plant Journal 37, 914–939. Walsh TA (2007) Chemical genetic approaches to uncover new sites of pesticide action. Pages 285–294 in Pesticide Chemistry: Crop Protection, Public Health, Environmental Safety; Ohkawa H, Miyagawa H, Lee PW, eds. WileyVCH, Weinheim, Germany. Weller S, Bressan R, Goldsbrough P, Fredenburg T, Hasegawa P (2001) The effect of genomics on weed management in the 21st century. Weed Science 49, 282–289. Wheelock C, Miyagawa H (2006) The omicization of agrochemical research. Journal of Pesticide Science 31, 240–244. Yuan JS, Tranel PJ, Stewart CN Jr. (2007) Non-target site herbicide resistance: a family business. Trends in Plant Science 12, 6–13. Zhang Q, Reichers D (2004) Proteomic characterization of herbicide safener-induced proteins in the coleoptiles of Triticum tauschii seedlings. Proteomics 4, 2058–2071. Zhen RG, Singh B (2001) From inhibitors to target site genes and beyond—herbicidal inhibitors as powerful tools for functional genomics. Weed Science 49, 266–272.

2

An Introduction To Molecular Genetic And Genomic Techniques Chhandak Basu and Sam R. Zwenger

Introduction

This chapter offers a list of molecular tools that should be useful to gain a better understanding of the genetic basis of weediness. A large majority of agriculturists and botanists alike seek to understand weediness traits on a molecular level, and perhaps, in some cases, even use the information to help control weeds in agriculture. Before one attempts to defeat an enemy, one should learn its characteristic strengths and weaknesses and use them advantageously. Toward that end, we list various molecular markers and tools such as map-based cloning and functional genomics that should help in determining these characteristics.

Weeds As A Source Of Genes For Crop Improvement

In one view of weed biology, weeds are just as domesticated as crop plants (Warwick and Stewart 2005). After all, both types of plants have suites of traits that adapt them to cultivated agricultural fields and have been selected by agricultural practices. However, unlike crop plants, nature selects the weedy species that perform best, whereas mankind is constrained to grow certain crop species. Thus, the deck is stacked for weeds to win. Indeed, weeds can tolerate various types of severe biotic and abiotic stresses and out-compete crop plants. Despite the power of weeds, as of yet, a weed genome has not been sequenced and so we lack a thorough understanding of weediness traits. Although Arabidopsis is often referred to as a weed species, it is not a weed, as explained by Jonathan Gressel (see Chapter 3). From the unique agronomic and physiological properties of various weeds, it can be said that weed genomes are certainly repositories for innumerable economically important plant genes in at least two respects. First, the genes in weeds cost U.S. farmers billions of dollars annually in yield losses. Second, these same genes, if transferred into crop genomes, could make crops much more productive. High throughput sequencing methods, such as development of an expressed sequence tag (EST) library (discussed below) or complementary DNA (cDNA) libraries, or using nextgeneration genome sequencing, can help open the treasure chest of important genes. We expect these genes will confer weediness advantages such as disease resistance, dormancy, pest resistance, salt tolerance, and drought tolerance. Some plants will emerge as better candidates for becoming a model weed species than others (see Chapter 4). For instance, models should preferably be diploid organisms, have relatively small genomes, be economically significant, and exemplify weediness traits such as fast growth, high fecundity, and vigorous seed dispersal (Basu et al. 2004). However, close relatives of crop plants, which sometimes have good weediness traits, are being studied for their potential agronomical benefits. Ellis et al. (2000) nicely reviewed the possibility of using wild barley (Hordeum vulgare subsp. spontaneum C. Koch) as a source of genes for crop improvement. Wild barley as well as Middle Eastern landraces are shown 11

12

WEEDY AND INVASIVE PLANT GENOMICS

to contain useful sources of genes for cultivated crop development (Ellis et al. 2000). The mlo gene in barley confers resistance to powdery mildew disease and wild type mlo gene is the origin of mlo-mediated resistance in cultivated barley (Lyngkjaer et al. 2000). Dormancy genes from weedy rice have been identified and it was subsequently proposed that dormancy genes in weedy rice could be studied to understand the germination mechanisms in cultivated rice (Gu et al. 2003). The dormancy genes from weedy rice could also be used to develop resistance to preharvest sprouting in cultivated rice (Gu et al. 2003). An important study, which describes weediness characteristics and their significance in agricultural settings, was done by Kane and Rieseberg (2008). They created an EST library from sunflower (Helianthus annuus), and subsequently demonstrated that weediness can emerge several times within a population. Perhaps of even greater importance is that they were able to identify putative weediness loci under selective pressure. This approach to understanding weediness traits will surely play an important role in our future understanding of most weedy species. Similarly, Riera-Lizarazu et al. (2005) proposed that genes from jointed goatgrass (Aegilops cylindrica), a close relative of cultivated wheat (Triticum aestivum), might be used to enhance genetic diversity in cultivated wheat populations. Okuno and Ebana (2003) identified seven quantitative trait loci (QTL) controlling allelopathic effects in rice. They proposed the QTL could be used to control weeds in rice fields. Molecular biology techniques have, therefore, been useful to identify genes in weeds in aid of crop improvement, even in the absence of genomic sequencing.

Tools And Approaches For Understanding Weediness At The Molecular Level Map-based Cloning And Molecular Markers

In map-based cloning (also known as positional cloning) short sequences (or markers) are identified near the gene of interest in the chromosome. There are many types of molecular markers that have been used in weed science and these have been reviewed by Slotta (2008). These markers sometimes flank or lie near the gene of interest and can be isolated. Map-based Cloning. If necessary, a technique known as chromosome walking can be performed to isolate the gene of interest. While chromosome walking may be feasible in smaller plant genomes, it may be complicated in larger plant genomes such as wheat, which has 16 billion base pairs. This technique may also be difficult to perform in plant genomes with highly repetitive DNA (Tanksley et al. 1995). Therefore, map-based cloning in polyploid weed species could be a challenging task. In some cases, it may be necessary to start with a mutant and ultimately identify the gene responsible for the altered phenotype (Jander et al. 2002). But as pointed out by Jander et al. (2002), many of the steps of chromosome walking in Arabidopsis thaliana have been either eliminated or made much easier because of the availability of the entire sequence of its genome (Arabidopsis Genome Initiative, 2000), availability of thousands of randomly distributed markers at the Cereon Arabidopsis Polymorphic Collection (http://www.arabidopsis.org/ cereon), and access to new tools such as single feature polymorphisms (Borevitz et al. 2007) etc. to detect DNA polymorphisms. Thus, we envisage map-based cloning as an important intermediate tool for gene discovery until weed genomes are sequenced and annotated. Map-based cloning, a forward genetics approach, begins with a mutant phenotype and later identification of the gene responsible for the mutant phenotype. The advantage of map-based

MOLECULAR GENETIC AND GENOMIC TECHNIQUES

13

cloning is that prior knowledge of the gene-of-interest or genome is not required (Jander et al. 2002). One of the great resources to initiate a map-based cloning project is the website developed by TAIR (The Arabidopsis Information Resource, www.arabidopsis.org) at http://171.66.71.74/portals/mutants/mapping.jsp. This website contains many valuable webbased analysis tools for map-based cloning experiments in Arabidopsis. The steps of map-based cloning of a hypothetical gene from a weed is described below (adapted from Liu et al. 2007): 1.

Identify an amplified fragment length polymorphism (AFLP) marker (discussed later) or another DNA marker closely associated with a dormancy gene (or any gene of interest). 2. Use thermal asymmetric interlaced polymerase chain reaction (PCR) (or TAIL PCR, Liu and Huang 1998) to isolate the gene along with the marker. 3. Use this fragment as a probe. 4. Screen a bacterial artificial chromosome (BAC) library or yeast artificial chromosome (YAC) library and selected clones containing the regions (or genes) of interest from the genome. Random Amplification Of Polymorphic DNA (RAPD). Determining variations in nucleotide sequence has become very popular, not only in plant studies, but also in those of animals, fungi, and other organisms. Randomly amplifying sequence polymorphisms, first reported by Williams et al. (1990), was one of the first methods established to determine these variations. Similar to standard PCR, RAPD analysis can be performed with a thermal cycler, basic gel electrophoresis equipment, and an imaging device. Decamer primers are developed without prior knowledge of sequence information; that is to say, the sequence is random and therefore they amplify multiple loci in genomic DNA. Because variations might exist between species, different banding patterns should result by detecting this with a primer that anneals to a region of DNA, but the results of this method are not as robust as some other markers described in subsections below. Xu et al. (2003) reported use of RAPD profiling to study genetic diversity in alligator weed (Alternanthera philoxeroides [Martius] Grisebach) collected from different parts of southern China. They used 108 RAPD primers in the study and concluded that alligator weed has low genetic diversity. Thus, a species does not need to be genetically diverse to be a cosmopolitan weed. Vellekoop et al. (1996) studied the weed Silene latifolia from sixteen geographical locations across Europe. By using RAPD markers they reported east-west division of the weed population. Many other studies using RAPD have been performed. RAPD markers were used to analyze genetic diversity among five broomrapes (Orobanche spp.) from Israel (Katzir et al. 1996). The authors reported clear genetic differences (diversity) among five Orobanche spp. MüllerSchärer and Fischer (2001) reported significant genetic differences among eighty common groundsel (Senecio vulgaris) plants taken from nine populations in two regions of Switzerland. Rutledge et al. (2000) studied the origin of propanil resistance in barnyard grass (Echinochloa crus-galli) in Arkansas. Using RAPD, markers they concluded the spread of resistant biotypes in Arkansas is from seed dispersal and independent mutation events. Stankiewicz et al. (2001) analyzed twenty-five populations of black nightshade (Solanum nigrum) from France, Poland, and U.K. using RAPD markers. They concluded that the spread of black nightshade in these European countries might be from dispersal of seeds by migratory birds. Restriction Fragment Length Polymorphisms (RFLP). As the name implies, RFLP uses restriction enzymes to digest DNA. Because restriction enzymes are sequence specific,

14

WEEDY AND INVASIVE PLANT GENOMICS

variations in sequences can be differentiated. However, it has been shown that relatively large differences (insertions or deletions) and certain nucleotide lengths are easier to detect than smaller sequence variations. After digestion with a restriction enzyme, the resulting fragments are electrophoresed and analyzed for banding differences using specific DNA probes in a DNA blot. One benefit of using RFLP in weed genomics studies is that it provides a relatively simple detection method for previously unpublished sequence data. However, because PCR is not used, the total starting DNA must be available in a larger quantity, and sequence from orthogolous genes-of-interest must be available. Desplanque et al. (1999) identified five RFLP loci and observed high genetic diversity in wild beet (Beta vulgaris L.) in France. They concluded that inland wild beets are closely related to Mediterranean coastal wild beets, but differ from other coastal types such as Biscay, Brittany, and northern France. RFLP has also been used to study phenotypic responses of chlorosulfuron application in kochia (Kochia scoparia L. Schrad.) biotypes (Guttieri et al. 1992). The authors reported strong correlation between the RFLP and the resistant and susceptible biotypes of kochia. Besides studying weed biology, RFLP have been incorporated in mapping. Wang et al. (1998) reported development of an RFLP map of foxtail millet (Setaria italica (L.) P. Beauv.) consisting of 160 loci. RFLP mapping was also used to study evolution of chlorotoluron resistance gene (Su1) in cultivated wheat (Triticum aestivum) and wild wheat (Triticum dicoccoides) (Krugman 1997). Amplified Fragment Length Polymorphism (AFLP). One technique that has been useful in determining genetic variation and phylogenetic relationships has been amplified fragment length polymorphisms (AFLP) (Vos et al. 1995). Similar to RFLP, restriction digestion is performed on the DNA and subsequently gel electrophoresis is performed. Like RAPD markers, AFLPs do not require any knowledge of genome sequences (Nguyen and Wu 2005). This aspect is especially useful in weed genomics research, where a priori knowledge of weed genomes might not be available. Amplification of polymorphisms based on sequence variation uses PCR to amplify fragments, therefore less DNA is required than for RFLP analysis. AFLP markers offer many benefits to areas beyond weed genomic studies. For example, AFLPs have been applied to forensic science when officials gather evidence for understanding distribution patterns of marijuana (Cannabis sativa) (Coyle et al. 2003). In a more recent report by Datwyler and Weiblen (2006), AFLP was used to differentiate between psychoactive marijuana and nonpsychoactive strains of hemp. Because these two plants often appear to be similar, AFLP markers provide another method of discriminating between them. AFLP analysis has also been used to look at the potential of hybridization in weedy plants. Wassom and Tranel (2005) provide a cluster analysis of hybridization potential based on genetic diversity in agronomically important pigweeds (Amaranthus spp.). In a study by Choi et al. (2003), a weedy soybean was crossed with a cultivated soybean, which provides a better understanding of the polymorphisms within the species. Many studies have used AFLPs to make genetic comparisons between wild-type (weedy) plants and their cultivated counterparts, including chicory (Van Cutsem et al. 2003), potato (Solis et al. 2007), azuki bean (Yoon et al. 2007), and carrots (Magnussen and Hauser 2007). In a review by Burger et al. (2008) describing the domestication of agronomically important plants, recognition is given to AFLP analysis in wheat and barley studies. Microsatellites. Microsatellites are simple sequence repeats (SSR) of two to five (usually two to three) nucleotides long dispersed throughout the genome (Holton 2001). For example,

MOLECULAR GENETIC AND GENOMIC TECHNIQUES

15

an SSR of CAA might be repeated five times and shown as (CAA)n, where n is the number of times the sequence is repeated at that locus. Microsatellites are found in both prokaryotic and eukaryotic genomes in coding and non-coding regions and are usually polymorphic (Zane et al. 2002). Microsatellites have the following advantages over other types of molecular markers (Holton 2001): (1) they show higher levels of polymorphisms, (2) are relatively easy to detect, and (3) are inherited as a codominant marker. The SSR may be a dinucleotide repeat (CA) or trinucleotide repeat (CAA) and can even be tetra and penta nucleotide motifs (Weising et al. 1995). Varshney et al. (2002) reported that trinucleotide repeats are more frequent (54% to 78%) than both dinucleotide repeats (17.1% to 40.4%) and tetranucleotide repeats (3% to 6%). With the sudden surge of ESTs (expressed sequence tags, discussed later) being deposited into various databases (e.g. dbEST, Boguski et al. 1993, http://www.ncbi.nlm.nih.gov/dbEST/), additional microsatellite and SSR identification are expected. With more sequence information we might expect to find basic patterns within the plant kingdom. For example, it has been shown that the (CA)n repeat is comparatively rare in plants and more common in humans, whereas (AT)n appears to be more common in plants (Lagercrantz et al. 1993). After identifying SSRs, designing a microsatellite motif specific primer and performing a standard PCR are necessary for a detailed microsatellite detection protocol (Hamilton et al. 1999). Development of working sets of primers to detect polymorphic microsatellite loci can sometimes be very challenging tasks (Squirrell et al. 2003). Nevertheless, development of allele-specific microsatellite markers will be an invaluable tool to enhance plant science research (Varshney et al. 2004), and thus, should also advance weed genomics research. Many weed researchers have benefited from using microsatellite markers. Green et al. (2001) used nine microsatellite markers and screened 131 samples of grass weed barren brome (Anisantha sterilis). They identified genetic diversity within and among different fields. They concluded Anisantha sterilis exists as many separate inbred genetic lines, which rarely outcross. In another study, Genton et al. (2005) developed five polymorphic microsatellite loci in common ragweed (Ambrosia artemisiifolia) populations from North America and France. They suggest that the microsatellite loci might be useful in studying invasion patterns of ragweed in North America and Europe. Jump et al. (2002) identified nine microsatellite markers in stemless thistle (Cirsium acaule). It was stated that identification of these loci could be useful in identifying several species throughout the genus Cirsium. Kane and Rieseberg (2008) studied 106 microsatellites in weedy populations of sunflower (Helianthus annuus). They reported that the weedy sunflower populations were more genetically similar with nearby wild populations than with one another. From their report it can also be concluded that weedy sunflower populations evolved multiple times within the species and retained their weediness traits despite the gene flow from nearby wild sunflower populations (Kane and Rieseberg 2008). SSR markers were also used to study weedy rice (Oryza sativa f. spontanea) populations in China (Cao et al. 2006). The authors proposed that weedy rice populations might have originated from cultivated rice (Oryza sativa) populations by mutation and intervarietal hybrids. Besides microsatellite identification in the nuclear genome of plants, chloroplast genomes of weed species have also been used to study polymorphisms. Seven microsatellite loci in the chloroplast genome of shepherd’s purse (Capsella bursa-pastoris) revealed a small population size (Ceplitis et al. 2005). Another interesting finding from this study was that over a brief period of evolutionary time, mononucleotide repeats within chloroplast genomes might be gradually lost. Inter Simple Sequence Repeats (ISSR). Inter simple sequence repeats (ISSR), a PCR-based marker system, uses simple sequence repeats or microsatellites in genome (Lagercrantz et al.

16

WEEDY AND INVASIVE PLANT GENOMICS

1993). ISSR was used to study genetic diversity among six biennial wormwood (Artemisia biennis Willd.) populations and one annual wormwood (Artemisia annua L.) population (Mengistu et al. 2004). Biennial wormwood is an important weed in the soybean and dry bean fields in the Dakotas and Minnesota (Mengistu et al. 2004). They concluded that biennial wormwood is different from the annual populations, although the biennial wormwood behaves much like annual wormwood. Tranel and Wassom (2001) used the ISSR technique to study 217 common cockleburs (Xanthium strumarium L.) to generate twenty-seven polymorphic markers. They concluded that genetic variation among the U.S. cocklebur population is distributed along a latitudinal gradient. Slotta et al. (2005) developed polymorphic markers for Canada thistle (Cirsium arvense) populations to understand its genetic make-up. They identified an average of nine polymorphic alleles per microsatellite locus and eleven loci per ISSR locus. This data should be very useful in finding strategies to control the spread of Canada thistle. A combination of ISSR and RAPD markers was also successfully used to determine genetic diversity among wild barley (Hordeum. vulgare subsp. spontaneum) populations in Turkey (Tanyolac 2003) Single Nucleotide Polymorphisms (SNPs). Single nucleotide polymorphisms, or SNPs (pronounced as “snips”), are DNA sequence variations observed between species due to differences in a single nucleotide (either A, G, C, or T). For example, the following two hypothetical weed biotypes have the following DNA sequences: Weed biotype 1 ACGTGCTGCAGCTAGCAATCGC Weed biotype 2 ACGTGCAGCAGCTAGCAATCGC In this case, it can be said that these two species have a SNP, that is, a single base polymorphism. SNPs have some advantages over other mapping methods (Kahl et al. 2005 and Gupta et al. 2001), including low mutation rates, relative ease of detection (even distribution across the genome), and choice of a non-gel based system. The number of SNPs and their positions in the respective genomes may also vary considerably (Kahl et al. 2005). Some of the freely available plant SNP databases accessible via the Internet are found in Table 2.1. It is evident that different types of SNPs might be present in biotypes that have altered expression patterns of specific genes (Figure 2.1). For example, promoter SNPs might be important to characterize since they could vary in their binding of transcription factors to the promoter region (Kahl et al. 2005). Some of the widespread ways of identifying SNPs have been outlined by Gupta et al. (2001). • • • •

Locus specific PCR primers are designed to identify SNPs between individuals. Post-PCR sequencing of amplified product leads to discovery of new SNPs. Expressed sequence tags (discussed later) can be aligned to identify new SNPs. BACs from sequenced genomes can be compared for mismatches to detect SNPs.

Table 2.1. Single nucleotide polymorphism databases for plants. SNP database name

Website

Reference

PlantMarkers dbSNP indica/japonica SNPs SIGnAL

http://markers.btk.fi/ http://www.ncbi.nlm.nih.gov/projects/SNP/ http://www.plantgenome.uga.edu/snp http://signal.salk.edu/cgi-bin/AtSFP

Rudd et al. 2005 Sherry et al. 1999 Feltus et al. 2004 Borevitz et al. 2007

MOLECULAR GENETIC AND GENOMIC TECHNIQUES

17

Figure 2.1. Genomic classification of SNPs. (Based on Kahl et al. 2005.)

Délye et al. (2002) developed SNP markers for blackgrass (Alopecurus myosuroides Huds.) resistant to acetyl CoA-carboxylase–inhibiting herbicides. Chloroplastic ACCase encoding gene from black-grass was cloned and the authors identified two point mutations in the gene (Délye et al. 2002). This information could be extremely useful in understanding the molecular biology of herbicide resistance in weeds. It should be noted that because they indicate a single base pair difference, SNPs are one of the most specific descriptors of sequence variation.

Fluorescence in situ Hybridization (FISH)

The FISH technique has been used to physically map genes and DNA sequences directly on chromosomes (Hass-Jacobus and Jackson 2005). It uses fluorochromes conjugated to an antibody to detect probes (Hass-Jacobus and Jackson 2005). Each probe is visualized as a separate color on chromosomes under a microscope slide. This technique can be applied in many ways. For example, Yang et al. (1999) used a fluorescence hybridization technique to study the genome structure and evolution of wild oat (Avena fatua L.). They used rRNA sequences from Avena fatua and genomic DNA from A. strigosa (AA-genome diploid) and A. clauda (CCgenome diploid) as probes. Based on their results, they concluded that the A. fatua genome organization was similar to that of other Avena species. Fregonez et al. (2004) used the FISH technique to differentiate among three weed genomes, namely Crepis japonica, Galinsoga parviflora, and Chaptalia nutans, which are all in the Asteraceae family. Although these three weed species seemed to consistently exhibit karyotypic differences, the 45S rDNA sites always occurred on the short arm of the chromosome. FISH has been modified in various ways to suit experimental needs. For example, Benabdelmouna et al. (2001) used GISH (genomic in situ hybridization) to differentiate between the A and B genomes in diploid and tetraploid foxtail millet weeds (Setaria sp.). GISH uses the same principle as FISH with the exception that in GISH, genomic DNA is used as the probe instead of specific genes (Hass-Jacobus and Jackson 2005).

18

WEEDY AND INVASIVE PLANT GENOMICS

Comparative Genomics

Comparative genomics involves comparing genomes of two or more species. Bioinformatics are important tools for comparative genomics among several plant species. For example, Chittenden et al. (1994) proposed possible relationships between Sorghum bicolor (a cultivated crop) and Sorghum halepense (Johnsongrass). With 2n = 40, Daggett (1976) proposed Sorghum halepense may be a polyploid descendent of Sorghum bicolor × Sorghum propinquum (Chittenden et al. 1994). RFLP mapping revealed 30% of thirteen S. bicolor restriction fragments were shared with S. halepense (Chittenden et al. 1994). From these observations it can be concluded that S. propinquum may have contributed to the genomic development of S. halepense. The authors also concluded that the RFLP data generated will be an invaluable tool in mapping QTLs (quantitative trait loci) associated with evolution of grain sorghum from its weedy relatives. Comparative genomics is a branch of genomics in which genome sequences of several species are compared. A study using comparative genomics (Lan et al. 2000) reported a comparative map of Brassica oleracea and Arabidopsis thaliana. They identified 57% of the comparative loci between Brassica and Arabidopsis, which was indicative of widespread synteny between Brassica and Arabidopsis. The comparative mapping approach by Lan et al. (2000) might be used to map weediness genes and compare them with a well-studied model plant. For example, Arabidopsis cDNA sequences could be used to identify and clone homologous genes from Brassica (Lan et al. 2000). Arabdopsis BAC/YAC contigs will also be useful in Brassica for map-based cloning (Lan et al. 2000).

Functional Genomics

Functional genomics is a branch of specialization in the field of genomics that studies the functions and interactions among genes. Functional genomics includes transcriptomics (e.g. mRNA profiling or microarray, discussed later), metabolomics, proteomics, etc. Use Of ESTs In Functional Genomics. Creating an EST library has become a widespread basis for the analysis of the transcriptional state of an organism. The process is similar to microarray (discussed later) analysis in that mRNA is isolated from an organism and reverse transcribed into complimentary DNA (cDNA). The cDNA fragments, which vary in length, can be amplified using traditional PCR methods and then inserted into vectors (i.e. plasmids). The vectors can then be inserted into a host (i.e. bacterial cells) and grown so as to clonally propagate the inserts. These inserts can be stored or removed using basic molecular laboratory methods. Sequencing of the cloned inserts is then performed and these represent the EST library. Therefore, an EST library is generated by sequencing of randomly selected cDNA clones from a cDNA library (Sreenivasulu, 2002). However, it is necessary to exclude the redundant clones from the EST collection to generate an EST library with novel clones (genes) (Andreas 2004). More than 3 million sequences from more than 200 plant species have been deposited in a publicly available EST database (http://www.ncbi.nlm.nih.gov/dbEST/) (Rudd 2003). Some of the weed ESTs deposited in this database are: Artemisia annua (sweet wormwood), Brachypodium distachyon (commonly called purple false brome), Glycine soja (wild soybean), Eschscholzia californica (California poppy), Euphorbia esula (leafy spurge), Hordeum vulgare subsp spontaneum (wild barley), Sorghum halepense (Johnson grass), etc. ESTs can be used for gene discovery, gene annotation, and comparative genomics (Rudd 2003). Development

MOLECULAR GENETIC AND GENOMIC TECHNIQUES

19

of an EST library is cheaper than the whole genome sequencing using first generation sequencing and is therefore sometimes known as the “poor man’s genome” (Rudd 2003). Development of an EST library will be very applicable in weed genomics research because no prior knowledge of a genome is required for making an EST library. ESTs can be “scanned” for putative “weediness” genes. They can also be used for weed plant metabolism analysis (e.g. allelopathy, stress response, plant growth, etc.). EST-based molecular markers (e.g. SNPs, SSRs, etc.) can be successfully used in weed biology research as described earlier. Spotted knapweed (Centaurea maculosa), originally from Eurasia, is a serious problem in the northwestern United States (Broz et al. 2007, see Chapter 12). Broz et al. (2007) developed EST libraries from seven populations of this invasive weed. The database could be very useful for weed biologists to understand the genetic basis of invasiveness. ESTs can also be used to develop microarray chips to study genome-wide gene expression in the invasive species Centaurea maculosa or in other knapweed species such as C. diffusa, C. solstitialis, C. virgata, or Acroptilon repens (Broz et al. 2007). The cDNA library or the EST database is surely a starting point to understanding the genetic basis of invasiveness in plants similar to knapweed. Anderson et al. (2007) reported an EST database of leafy spurge (Euphorbia esula L.), an economically important perennial weed in the northern plains of the United States and Canada (Foley 2002, see Chapter 4). They identified 19,015 unigenes with 15.5% of the unigenes unique to leafy spurge. The unigene database can be accessed at http://www.ncbi.nlm.nih.gov/ sites/entrez?db=unigene. Wang et al. (2006) created an EST library of the weed salt cedar (Tamarix androssowii). More recently, Gao et al. (2008) created an EST library of salt cedar specifically to better understand salt stress. Both of these studies have contributed to a greater understanding of the expression of weediness genes in salt cedar. Microarray Analysis. Microarray technology (Schena et al. 1995) is a powerful technology that allows the analysis of the simultaneous expression of thousands of genes. In this technology thousands of oligonucelotides or cDNAs (representing genes) are immobilized on a solid surface such as a glass microscope slide, and fluorescently labeled cDNAs from treated and control samples, respectively, are both hybridized onto the slide. After a series of washes, the slide is scanned with a laser-based fluorescence imager. Gene expression patterns can be inferred based on the fluorescence intensity of the spots. The hybridization patterns can also be used to study heterologous hybridization where labeled cDNAs from one species can be hybridized onto a microarray chip constructed for a different species. This includes making labeled cDNA from a weedy species to hybridize to a microarray chip constructed for Arabidopsis. While comparing known expression profiles from previously studied microarrays, heterologous hybridization (also called cross-species hybridization; Bar-Or et al. 2007) is helpful to understand evolutionary relationships and comparative genomics, or to analyze complex biochemical pathways in weed species. For example, Renn et al. (2004) compared heterologous hybridization across various fish species and found less hybridization with further evolutionary distances. Microarray technology can be successfully applied to understand weed biology physiology, and Lee and Tranel (2008) have provided an in-depth review on this aspect. In an interesting study by Horvath et al. (2006), it was demonstrated that gene expression in corn plants changes when grown in competition with velvetleaf weed (Abutilon theophrasti). In this case, the authors hybridized velvetleaf cDNAs onto a corn (Zea mays) microarray chip. They demonstrated that corn genes involved in carbon utilization, photosynthesis, red light signaling, and cell division were differentially expressed when corn was grown in competition with velvetleaf.

20

WEEDY AND INVASIVE PLANT GENOMICS

Because microarrays can yield copious amounts of data, experiments incorporating crossspecies hybridizations ostensibly have the potential to provide information on poorly understood pathways and evolutionary relationships of weedy plant species. One caveat has been that cross-species arrays can yield ambiguous results, so interpretations can vary, sometimes with unhelpful results. Options remain, however, such as using closely related species to that of the microarray or formulating a hypothesis about weed-related observations while using non-weed plants that have a cDNA microarray chip already established (Lee and Tranel 2008).

Next Generation Sequencing

New platforms for transcriptome or genome sequencing have emerged in recent years: next generation sequencing. These technologies are currently led by three companies (454 sequencing technology by Roche Applied Sciences, Solexa genome analysis system by Illumina, and SOLiD system by Applied Biosystems). Although each is similar in the beginning steps of sample preparation, they differ in respect to the sequencing approaches and chemistry. After DNA is randomly fragmented the resulting molecules are immobilized on a solid support. The support type can either be a macroscopic planar surface, such as a slide, or microscopic beads. Whereas one DNA molecule is placed on a bead, several can be randomly placed on a slide or flow cell. The support carrying the molecules is then subjugated to PCR, yielding polonies (polymerase colonies). These polonies remain attached to the surface. However, if they are on beads, the beads themselves are fixed onto a planar surface. The sequencing, which varies depending upon the method used, is performed next. Although no weed species have been sequenced using any of the next generation sequencing strategies, the potential exists for use of these technologies to sequence weeds in the future. In addition, technologies for next-next generation sequencing are moving quickly. Currently there is a tradeoff between quantity of sequence and the number of contiguous bases in data. For example, a 454 run (plate) will yield perhaps 300 Mb with an average read length of 400 bases, whereas a Solexa run will yield up to 2 Gb, but with only 50 base reads. Therefore, de novo genomic sequencing would likely require 454 to produce data indicating sequence scaffold and use Solexa to infill. There are promising technologies that could give long reads and much higher amounts of data, which will, no doubt, revolutionize genomics and systems biology for weeds and other organisms (Yuan et al. 2008).

Conclusion

Weeds, especially invasive weeds, are a threat to biodiversity (Broz et al. 2007). There is a saying that if we want to defeat our enemy we must know the enemy very well. Weeds are champions in agro-ecosystems with limited resources and cause tremendous damage to crop yields (Basu et al. 2004). Emerging tools of molecular biology (Dyer 1991), genomics (Basu et al. 2004), and bioinformatics (Larrinua and Belmar 2008) could be successfully used in the field of weed science to understand weed physiology, weed growth and development, and mechanisms of herbicide resistance, to name a few. For example, the emerging science of bioinformatics has sought to increase molecular data on the variation in sequences residing in all organisms. Weeds should not be left out of this useful field. Although not covered in this chapter, many additional techniques and databases exist, which will undoubtedly facilitate the bioinformatics/weeds relationship (Larrinua and Belmar 2008).

MOLECULAR GENETIC AND GENOMIC TECHNIQUES

21

To understand the many proteomic aspects of weed species (e.g. herbicide resistance), future research challenges mainly revolve around the lack of sequence data (Zhang and Riechers 2008). Therefore, the future of weed research will undoubtedly use map-based cloning methods and functional and comparative genomic approaches to help dissect weediness traits. Of course it should be remembered that while high-throughput methods used to gain molecular insight into weed genomics should increase the flow of information, it will be a challenging task to channel this information into the real-life agricultural field in order to intelligently combat weed species. Hopefully, after exploring the traits of weeds, it will be understood that sometimes even a worst enemy can become a best friend. References Anderson JV, Horvath DP, Chao WS, Foley ME, Hernandez AG, Thimmapuram J, Lei L, Gong GL, Band M, Kim R, Mikel MA (2007) Characterization of an EST database for the perennial weed leafy spurge: an important resource for weed biology research. Weed Science 55, 193–203. Andreas GR (2004) A simple hybridization-based strategy for the generation of non-redundant EST collections—a case study in barley (Hordeum vulgare L.). Plant Science 167, 629–634. Arabidopsis Genome Initiative (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796–815. Bar-Or C, Czosnek H, Koltai H (2007) Cross-species microarray hybridizations: a developing tool for studying species diversity. Trends in Genetics 23, 200–207. Basu C, Halfhill MD, Mueller TC, Stewart CN Jr. (2004). Weed genomics: new tools to understand weed biology. Trends in Plant Science 9, 391–398. Benabdelmouna A, Shi Y, Abirached-Darmency M, Darmency H (2001) Genomic in situ hybridization (GISH) discriminates between the A and the B genomes in diploid and tetraploid Setaria species. Theoretical and Applied Genetics 44, 685–690. Boguski MS, Lowe TM, Tolstoshev CM (1993) dbEST—database for “expressed sequence tags.” Nature Genetics 4, 332–333. Borevitz JO, Hazen SP, Michael TP, Morris GP, Baxter IR, Hu TT, Chen H, Werner JD, Nordborg M, Salt DE, Kay SA, Chory J, Weigel D, Jones JD, Ecker, JR (2007) Genome-wide patterns of single-feature polymorphism in Arabidopsis thaliana. Proceedings of the National Academy of Sciences of the United States of America 104, 12057–12062. Broz AK, Broeckling CD, He J, Dai X, Zhao PX, Vivanco JM (2007) A first step in understanding an invasive weed through its genes: an EST analysis of invasive Centaurea maculosa. BMC Plant Biology 7, 25. Burger JC, Chapman MA, Burke JM (2008) Molecular insights into the evolution of crop plants. American Journal of Botany 95, 113–122. Cao Q, Lu BR, Xia H, Rong J, Sala F, Spada A, Grassi F (2006) Genetic diversity and origin of weedy rice (Oryza sativa f. spontanea) populations found in North-eastern China revealed by simple sequence repeat (SSR) markers. Annals of Botany 98, 1241–1252. Chittenden LM, Schertz KF, Lin YR, Wing RA, Paterson AH (1994) A detailed RFLP map of Sorghum bicolor × S. propinquum, suitable for high-density mapping, suggests ancestral duplication of Sorghum chromosomes or chromosomal segments. Theoretical and Applied Genetics 87, 925–933. Choi P, Yoshiro M, Ishikawa A, Odashima M, Umezawa T, Fujimura T, Takahata Y, Komatsuda T (2003) Construction of a high-density AFLP and SSR map using recombinant inbred lines of cultivated × weedy soybean. Breeding Science 53, 335–344. Ceplitis A, Su Y, Lascoux M (2005) Bayesian inference of evolutionary history from chloroplast microsatellites in the cosmopolitan weed Capsella bursa-pastoris (Brassicaceae). Molecular Ecology 14, 4221–4233. Coyle HM, Shutler G, Abrams S, Hanniman J, Neylon S, Ladd C, Palmbach T, Lee HC (2003) A simple DNA extraction method for marijuana samples used in amplified fragment length polymorphisms (AFLP) analysis. Journal of Forensic Science 48, 343–347. Daggett H (1976) Sorghum. In: Evolution of Crop Plants. Simmonds NW, ed. Longman Scientific and Technical, Essex, UK pp. 112–117. Datwyler SL, Weiblen GD (2006) Genetic variation in hemp and marijuana (Cannabis sativa L.) according to amplified fragment length polymorphisms. Journal of Forensic Science 51, 371–375. Délye C, Calmès E, Matéjicek A (2002) SNP markers for black-grass (Alopecurus myosuroides Huds.) genotypes resistant to acetyl CoA-carboxylase inhibiting herbicides. Theoretical and Applied Genetics 104, 1114–1120.

22

WEEDY AND INVASIVE PLANT GENOMICS

Desplanque B, Boudry P, Broomberg K, Saumitou-Laprade P, Cuguen J, Van Dijk H (1999) Genetic diversity and gene flow between wild, cultivated and weedy forms of Beta vulgaris L. (Chenopodiaceae), assessed by RFLP and microsatellite markers. Theoretical and Applied Genetics 98, 1194–1201. Dyer WE (1991) Applications of molecular biology in weed science. Weed Science 39, 482–488. Ellis RP, Forster BP, Robinson D, Handley LL, Gordon DC, Russell JR, Powell W (2000) Wild barley: a source of genes for crop improvement in the 21st century? Journal of Experimental Botany 51, 9–17. Feltus FA, Wan J, Schulze1 SR, Estill JC, Jiang N, Paterson AH (2004) An SNP resource for rice genetics and breeding based on subspecies indica and japonica genome alignments. Genome Research 14, 1812–1819. Foley ME (2002) Weeds, seeds, and buds—opportunities and systems for dormancy investigations. Weed Science 50, 267–272. Fregonezi JN, Torezan JMD, Vanzela ALL (2004) A karyotypic study of three southern Brazilian Asteraceae species using fluorescence in situ hybridization with a 45S rDNA probe and C-CMA3 banding. Genetics and Molecular Biology 27, 223–227. Gao C, Wang Y, Liu G, Yang C, Jiang J, Li H (2008) Expression profiling of salinity-alkali stress responses by large-scale expressed sequence tag analysis in Tamarix hispida. Plant Molecular Biology 66, 245–258. Genton BJ, Jonot O, Thévenet D, Fournier E, Blatrix R, Vautrin D, Solignac M, Giraud T (2005) Isolation of five polymorphic microsatellite loci in the invasive weed Ambrosia artemisiifolia (Asteraceae) using an enrichment protocol. Molecular Ecology Notes 5, 381–383. Green JM, Barker JH, Marshall EJ, Froud-Williams RJ, Peters NC, Arnold GM, Dawson K, Karp A (2001) Microsatellite analysis of the inbreeding grass weed Barren Brome (Anisantha sterilis) reveals genetic diversity at the within- and between-farm scales. Molecular Ecology 10, 1035–1045. Gu X, Chen Z, Foley ME (2003) Inheritance of seed dormancy in weedy rice. Crop Science 43, 835–843. Gupta PK, Roy JK, Prasad M (2001) Single nucleotide polymorphisms: a new paradigm for molecular marker technology and DNA polymorphism. Current Science 80, 524–535. Guttieri MJ, Eberlein CV, Mallory-Smith CA, Thill DC, Hoffman DL (1992) DNA sequence variation in domain A of the acetolactate synthase genes of herbicide-resistant and -susceptible weed biotypes. Weed Science 40, 670–677. Hamilton MB, Pincus EL, Di Fiore A, Fleischer RC (1999) Universal linker and ligation procedures for construction of genomic DNA libraries enriched for microsatellites. Biotechniques 27, 500–502, 504–507. Hass-Jacobus B, Jackson SA (2005) Physical mapping of plant chromosomes. In: The Handbook of Plant Genome Mapping. Meksem K, Kahl G eds. Wiley-VCH Verlag GmbH and Co. KGaA, Weinheim pp. 131–150. Holton TA (2001) Plant genotyping by microsatellites. In: Plant Genotyping: The DNA Fingerprinting of Plants. Henry RJ, ed. CABI Publishing, Wallingford, UK pp. 15–25. Horvath DP, Llewellyn D, Clay SA (2006) Heterologous hybridization of cotton microarrays with velvetleaf (Abutilon theophrasti) reveals physiological responses due to corn competition. Weed Science 55, 546–557. Jander G, Norris SR, Rounsley SD, Bush DF, Levin IM (2002) Arabidopsis map-based cloning in the post-genome era. Plant Physiology 129, 440–450. Jump AS, Dawson DA, James CM, Woodward FI, Burke T (2002) Isolation of polymorphic microsatellites in the stemless thistle (Cirsium acaule) and their utility in other Cirsium species. Molecular Ecology Notes 2, 589–592. Kahl G, Mast A, Tooke N, Shen R, van der Boom D (2005) Single nucleotide polymorphisms: Detection techniques and their potential for genotyping and genome mapping. In: The Handbook of Plant Genome Mapping. Meksem K, Kahl G, eds. Wiley-VCH Verlag GmbH and Co. KGaA, Weinheim pp. 75–108. Kane NC, Rieseberg LH (2008) Genetics and evolution of weedy Helianthus annuus populations: adaptation of an agricultural weed. Molecular Ecology 17, 384–394. Katzir N, Portnoy V, Tzuri G, Joel DM, Castejon M (1996) Use of random amplified polymorphic DNA (RAPD) markers in the study of the parasitic weed Orobanche. Theoretical and Applied Genetics 93, 367–372. Krugman T, Levy O, Rubin B, Snape JW, Korol A, Nevo E (1997) Comparative RFLP mapping of chlorotoluron resistance gene (SU1) in cultivated wheat (Triticum aestivum) and wild wheat (Triticum dicoccoides). Theoretical and Applied Genetics 94, 46–51. Lagercrantz U, Ellegren H, Anderson L (1993) The abundance of various polymorphic microsatellite motifs differs between plants and vertebrates. Nucleic Acids Research 21, 1111–1115. Lan TH, Delmonte TA, Reischmann KP, Hyman J, Kowalski SP, McFerson J, Kresovich S, Paterson AH (2000) An ESTenriched comparative map of Brassica oleracea and Arabidopsis thaliana. Genome Research 10, 776–788. Larrinua IM, Belmar SB (2008) Bioinformatics and its relevance to weed science. Weed Science 56, 297–305. Lee RM, Tranel PJ (2008) Utilization of DNA microarrays in weed science research. Weed Science 56, 283–289. Liu Y, Huang N (1998) Efficient amplification of insert end sequences from bacterial artificial chromosome clones by thermal asymmetric interlaced PCR. Plant Molecular Biology Reporter 16, 175–181.

MOLECULAR GENETIC AND GENOMIC TECHNIQUES

23

Liu Z, Faris JD, Edwards MC, Friesen TL (2007) The initiation of map-based cloning of an avirulence gene from Pyrenophora teres. Meeting Abstract. 24th Fungal Genetics Conference, Pacific Grove, California, USA (March 20–25, 2007) pp 159. Lyngkjaer MF, Newton AC, Atzema JL, Baker SJ (2000) The barley mlo-gene: an important powdery mildew resistance source. Agronomie 20, 745–756. Magnussen LS, Hauser TP (2007). Hybrids between cultivated and wild carrots in natural populations in Denmark. Heredity 99, 185–192. Mengistu LW, Christoffers MJ, Kegode GO (2004) Genetic diversity of biennial wormwood. Weed Science 52, 53–60. Müller-Schärer H, Fischer M (2001) Genetic structure of Senecio vulgaris in relation to habitat type and population size. Molecular Ecology 10, 17–28. Nguyen HT, Wu X (2005) Molecular marker systems for genetic mapping. In: The Handbook of Plant Genome Mapping. Meksem K, Kahl G, eds. Wiley-VCH Verlag GmbH and Co. KGaA, Weinheim pp. 23–52. Okuno K, Ebana K (2003) Identification of QTL controlling allelopathic effects in rice: genetic approaches to biological control of weeds. Japan Agricultural Research Quarterly 37, 77–81. Renn SP, Aubin-Horth N, Hofmann HA (2004) Biologically meaningful expression profiling across species using heterologous hybridization to a cDNA microarray. BMC Genomics 5, 42. Riera-Lizarazu O, Watson CJW, Zemetra RS, Mallory-Smith CA, Vales MI (2005) Development of a molecular marker linkage map of jointed goatgrass (Aegilops cylindrica Host). 2005. Agronomy Abstracts. ASA, CSSA, SSSA Annual Meeting, Salt Lake City, Utah, USA (Nov. 6–10, 2005). Rudd S (2003) Expressed sequence tags: alternative or complement to whole genome sequences? Trends in Plant Science 8, 321–329. Rudd S, Schoof H, Mayer K (2005) PlantMarkers—a database of predicted molecular markers from plants. Nucleic Acids Research 33, 628–632. Rutledge J, Talbert RE, Sneller CH (2000) RAPD analysis of genetic variation among propanil-resistant and -susceptible Echinochloa crus-galli populations in Arkansas. Weed Science 48, 669–674. Schena M, Dari S, Davis RW, Brown PO (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270, 467–470. Sherry ST, Ward M, Sirotkin K (1999) dbSNP—Database for single nucleotide polymorphisms and other classes of minor genetic variation. Genome Research 8, 677–679. Slotta T, Foley M, Horvath D (2005) Development of polymorphic markers for Cirsium arvense, Canada thistle, and their amplification in closely related taxa. Molecular Ecology Notes 5, 917–919. Slotta TAB (2008) What we know about weeds: insights from genetic markers. Weed Science 56, 322–326. Solis JS, Ulloa DM, Rodríguez LA (2007) Molecular description and similarity relationships among native germplasm potatoes (Solanum tuberosum ssp. tuberosum L.) using morphological data and AFLP markers. Electronic Journal of Biotechnology DOI: 10.2225/vol10-issue3-fulltext-14. http://www.ejbiotechnology.info/content/vol10/issue3/full/14 Squirrell J, Hollingsworth PM, Woodhead M, Lowe AJ, Gibby M, Powell W (2003) How much effort is required to isolate nuclear microsatellites from plants? Molecular Ecology 12, 1339–1348. Sreenivasulu N, Kavi Kishor PB, Varshney RK, Altschmied L (2002) Mining functional information from cereal genomes —the utility of expressed sequence tags. Current Science 83, 965–973. Stankiewicz M, Gadamski G, Gawronski SW (2001) Genetic variation and phylogenetic relationships of triazine-resistant and triazine-susceptible biotypes of Solanum nigrum analysis using RAPD markers. Weed Research 41, 287–300. Tanksley SD, Ganal MW, Martin GB (1995) Chromosome landing: a paradigm for map-based gene cloning in plants with large genomes. Trends in Genetics 11, 63–68. Tanyolac B (2003) Inter-simple sequence repeat (ISSR) and RAPD variation among wild barley (Hordeum vulgarae subsp. spontaneum) populations from west Turkey. Genetic Resources and Crop Evolution 50, 611–614. Tranel PJ, Wassom JJ (2001) Genetic relationships of common cocklebur accessions from the United States. Weed Science 49, 319–325. Van Cutsem P, du Jardin P, Boutte C, Beauwens T, Jacqumin S, Vekemans X (2003) Distinction between cultivated and wild chicory gene pools using AFLP markers. Theoretical and Applied Genetics 107, 713–718. Varshney RK, Thiel T, Stein N, Graner A (2002) In silico analysis on frequency and distribution of microsatellites in ESTs of some cereal species. Cellular and Molecular Biology Letters 7, 537–546. Varshney RK, Zhang H, Potokina E, Stein N, Langridge P, Graner A (2004) A simple hybridization-based strategy for the generation of non-redundant EST collections—a case study in barley (Hordeum vulgare L.). Plant Science 167, 629–634. Vellekoop P, Buntjer JB, Maas JW, van Brederode J (1996) Can the spread of agriculture in Europe be followed by tracing the spread of the weed Silene latifolia? A RAPD study. Theoretical and Applied Genetics 92, 1085–1090.

24

WEEDY AND INVASIVE PLANT GENOMICS

Vos P, Hogers R, Bleeker M, Reijans M, van de Lee T, Hornes M, Friters A, Pot J, Paleman J, Kuiper M, Zabeau M (1995) AFLP: A new technique for DNA fingerprinting. Nucleic Acid Research 23, 4407–4414. Wang YC, Yang CP, Liu GF, Jiang J, Wu JH (2006) Generation and analysis of expressed sequence tags from a cDNA library of Tamarix androssowii. Plant Science 170, 28–36. Wang ZM, Devos KM, Liu CJ, Wang RQ, Gale MD (1998) Construction of RFLP-based maps of foxtail millet, Setaria italica (L.) P. Beauv. Theoretical and Applied Genetics 96, 31–36. Warwick SI, Stewart CN Jr. (2005) Crops come from wild plants—domestication, transgenes, and linkage together shape ferality, pp. 9–30. In: J. Gressel (ed.) Crop Ferality and Volunteerism. CRC Press, Boca Raton, FL. Wassom JJ, Tranel PJ (2005) Amplified fragment length polymorphism-based genetic relationships among weedy Amaranthus species. Journal of Heredity 96, 410–416. Weising K, Nybom H, Wolff K, Meyer W (1995) DNA Fingerprinting in Plants and Fungi. CRC Press, Boca Raton, FL. Williams JGK, Kubelik AR, Livak KJ, Rafalski JA, Tingey SV (1990) DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Research 18, 6531–6535. Xu CY, Zhang WJ, Fu CZ, Lu BR (2003) Genetic diversity of alligator weed in China by RAPD analysis. Biodiversity and Conservation 12, 637–645. Yang Q, Hanson L, Bennett MD, Leitch IJ (1999) Genome structure and evolution in the allohexaploid weed Avena fatua L. (Poaceae). Genome 42, 512–518. Yoon MS, Lee J, Kim CY, Baek HJ (2007) Genetic relationships among cultivated and wild Vigna angularis (Willd.) Ohwi et Ohashi and relatives from Korea based on AFLP markers. Genetic Resources and Crop Evolution 54, 875–883. Yuan JS, Galbraith DW, Dai SY, Griffin P, Stewart CN Jr. (2008) Plant systems biology comes of age. Trends in Plant Science 13, 165–171. Zane L, Bargelloni L, Patarnello T (2002) Strategies for microsatellite isolation: a review. Molecular Ecology 11, 1–16. Zhang Q, Riechers DE (2008) Proteomics: an emerging technology for weed science research. Weed Science 56, 306–313.

3

Arabidopsis Is Not A Weed, And Mostly Not A Good Model For Weed Genomics; There Is No Good Model For Weed Genomics Jonathan Gressel

Introduction: Arabidopsis And Weediness

The widely used model species for plants, the fully sequenced Arabidopsis thaliana cv Columbia (hereafter called Arabidopsis, unless another strain or species is stated), is widely known as “the weed” by the plant biology community, and has often been called a weed by those who study it. The electronic newsletter of Arabidoptologists was called “Weeds World,” the Weed Science Society of America lists it as a weed (common name: “mouse–eared cress, #ARBTH”), and many renowned scientists have referred to it as a weed in publications in prestigious journals (Bartel and Last 2004; Buell 1998; Ecker 1998; Federspiel 2000; Kunkel 1996; Meyerowitz 1989; Rensink and Buell 2004). Does this consensus make it a weed? A more careful examination suggests that Arabidopsis is at best a wild species that has few weedy characters. The common lab strains are actually partly domesticated, losing some traits that might be classified in the syndrome called “wildness” (Table 3.1). Strains such as the commonly used Columbia germinate upon imbibition, having no secondary dormancy to overcome or light requirements to induce germination. When researchers want to study such phenomena, they must go to more wild strains such as cvi (Carrera et al. 2008). Selfcompatibility is a trait that is selected in most major domesticated crops, as is the case with domesticated A. thaliana but not so in wild A. lyrata, where researchers interested in this trait had to go to find genomic information (Nasrallah et al. 2002). There are claims that A. thaliana Columbia has lost disease resistance genes that are found in wild strains (Stahl et al., 1999). The diminutive Arabidopsis does not abide in any agro-ecosystems in which weeds are rampant; it performs best under low irradiance conditions in greenhouses and growth chambers in pots. It is not competitive with crops or much else (Figure 3.1). It is a cosmopolitan, high latitude species, whose highly-specific niche is where winter deciduous species grow and there is bare soil in early spring, in forests and disturbed (ruderal) ecosystems. Its life cycle allows it to germinate, grow, and set seed in a very short period before other wild species close the canopy that would smother it if it had not finished its annual life cycle. It has limited seed production compared to many weeds, limited plasticity, and a limited ability to adapt because of an inability to compete with other species. Its ability to adapt seems to be limited to genotypic plasticity in recognizing the ideal light regime so that germination and flowering occur at the most appropriate times for a given ecosystem (Banta et al. 2007). There have been lengthy discussions about the great extent of phenotypic plasticity of Arabidopsis (Pigliucci and Kolodynska 2006), but its plasticity was not compared to that of other species, especially weedy species. Part of the phenotypic plasticity of most weeds comes from a symbiotic synergy with mycorrhizae, but alas, Arabidopsis is not infected by these fungi (Cipollini et al. 2008). Weediness is a syndrome, and as with many syndromes there is not a single diagnostic trait. Indeed, a weed typically has a group of weediness traits, with different weeds having a variety of groupings. How can one plant, chosen because it has the smallest known genome in higher 25

26

WEEDY AND INVASIVE PLANT GENOMICS

Table 3.1. Traits of Arabidopsis thaliana cv Columbia (in underlined bold) compared to the properties that make up the weediness, the wild, and the domestication syndromes. Many of the wild traits are found in other A. thaliana strains other than Columbia or Landsbergii, or other Arabidopsis species. Wild traits

Traits of domesticated crops

Weedy traits

Propagules that are not adapted to long-distance dispersal

Retention of the seed/fruit on the plant at maturity

Seed dormancy Discontinuous germination (secondary dormancy)

Loss of germination inhibition Synchrony in germination (loss of secondary dormancy) No requirement for light

Propagules that are adapted to long-distance dispersal and easily distributed Seed dormancy Discontinuous germination (secondary dormancy)

Requires light for germinations Reduced ability to germinate in a wide range of conditions Long-lived seeds (seed bank) Slow growth to flowering, perennial More determinate growth Lower seed output Limited geographic range Seed produced in a narrower range of environmental conditions Propagule (seed) shattering No special adaptations for seed dispersal over both short and long distances Reduced plasticity of growth Reduced competitive ability Not adapted to disturbed habitats Diploid and/or polyploid

Short-lived seeds (no seed bank) Synchrony of flowering and fruit development More determinate growth Smaller numbers of larger fruits or inflorescences Wide geographic range Seed produced in a wide range of environmental conditions Reduction in seed dispersal (shattering)

Broad germination requirements Ability to germinate in a wide range of conditions Long-lived seeds (seed bank) Rapid growth to flowering, annual (Arabidopsis faster than weeds) Continuous seed production for as long as growing conditions permit Very high seed output Wide geographic range Seed produced in a wide range of environmental conditions Propagule (seed) shattering Special adaptations for seed dispersal over both short and long distances

Increase in apical dominance Reduced competitive ability Adaptation to disturbed habitats

Plasticity of growth Strong competitive ability Adaptation to disturbed habitats

Polyploidy frequent

Polyploidy frequent

Traits of weeds, wild plants, and crops modified from Warwick and Stewart (2005) and Gressel (2008) as well as other sources. (Prepared with the help of Dr. Yuval Eshed, Arabidoptologist.)

plants, be a model for weeds with their many-fold larger genomes? An organism that has a longer (e.g. growing-season-long) life cycle can afford the time and energy to synthesize a large amount of DNA per genome, allowing it the flexibility to exist in the variable habitats and climates of major global weeds. A very short per-season life cycle favors small genomes. This was demonstrated long ago by Spiegelman and colleagues (Levisohn and Spiegelman 1969; Spiegelman 1971), who grew Escherichia coli with its phage in a chemostat. As they ramped up the speed of E. coli cell division with rich media, and increased the dilution rate, the bacteriophages were selected for those that had faster division. This was achieved by eliminating genes; the shorter DNA replicated faster. Thus, they selected for phages with smaller and smaller genome sizes until in the end the phages had a replicase gene and little more left. A streamlined genome is sufficient for fast replication in a rich media-enabled monoculture, but the fewer the genes an organism has, the less adaptive and competitive it can be. The Columbia strain of Arabidopsis is now strictly adapted to monocultures in pots in a low-light lab environment, hardly comparable to a weed that must adapt to various habitats.

ARABIDOPSIS: THERE IS NO GOOD MODEL FOR WEED GENOMICS

27

Figure 3.1. Arabidopsis thaliana (right) and a weed (Sonchus oleraceus, sow thistle). (Photo courtesy of Sagit Meir.)

Even if one uses the definition widely attributed to the poet/philosopher Ralph Waldo Emerson that “a weed is a plant whose virtues have not yet been discovered” (Lewis 2008), Arabidopsis is surely not a weed because it does have many virtues. By virtue of the genomics and its adaptedness to the lab, we have learned ever so much about plant growth, development, and metabolism. Scientists were able to annotate thousands of plant genes from other species (though not always correctly) based on what is known from the studies of Arabidopsis. Still, after hearing the repetitive mantra that Arabidopsis is a model plant, many are surprised at the huge number of novel putative genes that are found every time a new plant species is sequenced, genes that are not in Arabidopsis. The initial hubris of Arabidoptologists was that studying Arabidopsis was enough to understand all plants including weeds, similar to the (in)famous statement by Nobel laureate Jacques Monod a few generations ago that “Anything that is true of Escherichia coli must be true of elephants, only more so” (Friedmann 2004). Even though an elephant’s body contains orders of magnitude more E. coli cells than elephant cells, we gain little understanding about elephant biology from their E. coli excrement. Questions About Weeds—Can Arabidopsis Genomics Answer Them?

We have many important basic and applied questions about weeds, and genomics can assist in answering them. For example, just over a decade ago Amaranthus tuberculatus = A. rudis (Pratt and Clark 2001) was a relatively rare wild species limited to river bottoms and disturbed lands and described in Gray’s Manual of Botany (Fernald 1950) and Weeds of the North Central States (Buchholtz et al. 1954) as Acnida altissima. It was not considered one of the important weedy amaranths (Holm et al. 1997; Holm et al. 1977). There was only one isolated early, possibly prophetic, report about its weediness in agriculture before 1995 (Feltner et al. 1969).

28

WEEDY AND INVASIVE PLANT GENOMICS

Suddenly it became a major weed, with more than 150 peer-reviewed reports appearing since 1995 with its new names—A. tuberculatus is now preferred by taxonomists (Pratt and Clark 2001). These publications describe its weediness insofar as it left its limited habitat and it displayed an amazing propensity to evolve resistance to herbicides of various chemical groups with quite varying modes of action (Costea et al. 2005; Hager et al. 2002; Hartzler et al. 2004; Horak and Peterson 1995; Owen and Zelaya 2005; Patzoldt et al. 2002; Steckel and Sprague 2004; Uscanga-Mortera et al. 2007; Volenberg et al. 2007). Based upon findings of Gressel and Kleifeld (1994), it appears that single gene mutations rarely cause a syndrome change; it takes more than a propensity to evolve resistance to make a pernicious weed. As if by intelligent design or divine retribution, A. tuberculatus evolved to be a pernicious and herbicide-resistant weed, especially in areas where a goodly part of the human population does not believe in the existence of evolution. What genes mutated or were their regulations altered to bring this species out of the swamp? What/how many/which genes turn a wild species into a pernicious weedy species? We are not likely to get the answer by studying Arabidopsis gene chips.

The Misdirected Approach In Using Arabidopsis To Elucidate New Herbicide Targets

Until recently, new herbicide discovery had been a principal research endeavor of agrochemical companies. No new major targets for herbicide action have been found for decades. Biotechnology companies were founded and major agrochemical company investments were made by using Arabidopsis genomics to find new targets so that novel herbicides could be discovered. Gene expression was either suppressed in Arabidopsis by RNAi or down regulated by anti-sense, or the genes were knocked out by T-DNA insertion mutagenesis to find lethal mutants. These genes were further characterized and putative target enzymes for herbicides were synthesized (Rice et al. 2004). These enzymes were used as targets to be challenged with huge libraries of chemicals that were tested for reactions with the enzymes. Patents were filed claiming the targets on a routine basis. In most cases no chemical was found that inhibited the target enzyme with a low Ki, and even when one was found, treating plants with it did not kill the plants. With the robots now idle, many of the companies have gone bankrupt or are no longer looking for herbicides. What was wrong with such an erstwhile promising approach? First, the greatest need for herbicides today is to control graminaceaous weeds (especially in grain crops) and perennial weeds with storage organs, which like the Phoenix, arise from the ashes of having their aboveground tissues killed by conventional herbicides. Graminaceous weeds are so closely related to graminaceous crops that it was not hard for them to evolve unregulated herbicide catabolic pathways similar to those used by the crops (Gressel 1988; Gressel 2002). Most of the herbicides that did or do control graminaceous weeds do not control Arabidopsis. It was an errant assumption that a compound that kills Arabidopsis will control graminaceous weeds. A herbicide must systemically translocate to dormant or underground buds to kill many perennial or rosette type weeds to prevent their Phoenix-like response to herbicides. How can a species that has no such buds be a useful model of herbicide discovery for such weeds? It was a mistake to assume that there always will be a chemical that binds to an enzyme and inhibits its function. Even when a chemical that at a low rate gives 95% inhibition, it does not follow that the plant will die. At least nine targets have been documented in which the supposedly primary enzyme is inhibited, yet the plant does not die (see Table 2.4 in Gressel 2002). Conversely, some of the best herbicides are lethal when their target is only partially

ARABIDOPSIS: THERE IS NO GOOD MODEL FOR WEED GENOMICS

29

inhibited; the lethality is due to accumulation of toxic intermediates, which would not be discovered in a high throughput enzyme assay. The first goal in classical herbicide development had been to find a lead chemical that killed plants in novel ways, suggesting a new target, and then to synthesize around it to optimize structure, function, and selectivity. No such lead compounds seem to have been found using the Arabidopsis genomics approach. Could Arabidopsis genomics be used for useful target discovery in the future? If industry is not completely soured on Arabidopsis, it might consider taking the hundreds of lead chemicals that are available that kill plants (e.g. Table 2.7 in Gressel 2002) and use the ones that kill Arabidopsis, elucidating their targets using Arabidopsis functional genomics and metabolomics. These lead chemicals are the hundreds of natural products that kill plants, yet whose target site(s) are unknown. These could have resulted in multi-site inhibitors, which would be far less prone to the evolution of resistance. This multisite property might also be synergistic, as has been demonstrated with sorgoleone, a natural allelochemical produced by sorghum roots that is an inhibitor of both photosystem II (Czarnota et al. 2001) and HPPD (4-hydroxyphenylpyruvate-dioxygenase) (Meazza et al. 2002). This multi-site synergy theoretically delayed the evolution of resistance for millions of years. Once potentially good sites are found using this strategy, there could be high throughput assays developed to test newly designed and simplified versions of the natural products to find functional herbicides. With known genes, X-ray crystallography can be performed on the target proteins to elucidate structures, which would assist in optimizing chemicals that bind to the target. Arabidopsis Genomics Can Help In Dealing With Transgene Flow—In A Limited Manner

A real and major issue with transgenic crops, especially herbicide-resistant transgenic crops, is transgene flow to related weeds, where related weeds exist. Transgene flow to the wild is a far less real issue, due to rarity as well as unlikelihood of having wild relatives that derive a selective advantage or disadvantage in meaningful population levels (Gressel and Al-Ahmad 2005). One way to preclude transgenic crop-weed hybrids from establishing is by transgenic mitigation, in which the transgene of choice is covalently and genetically linked with a gene that is good or neutral for the crop but would be deleterious to a related weed, preventing its establishment (Gressel 1999). When a dwarfing gene was stacked in tandem with an herbicide resistance gene and was then introduced into oilseed rape (Brassica napus), it enhanced the yield of oilseed rape when cultivated by itself, but severely reduced its ability to compete with the wild-type oilseed rape (Al-Ahmad et al. 2006). Hybrids of the transgenic plants with B. rapa were also unable to compete with this weedy relative (Al-Ahmad and Gressel 2006). Oilseed rape is related to (but does not hybridize with) Arabidopsis. A series of Arabidopsis genes whose functions correspond with silique (seed capsule) opening have been isolated, e.g. shatterproof, indehiscent 1, etc. (Liljegren et al. 2004; Yanofsky and Liljegren 2006). One such gene ectopically expressed when introduced into a Brassica sp. did inhibit silique opening (Østergaard et al. 2006), and thus might be an excellent mitigating gene for oilseed rape, since it is not fully domesticated. Despite breeding efforts, oilseed rape can still shatter (prematurely drop) a significant proportion of seed, which reduces yield. Thus, such genes might be of use to oilseed rape. Conversely, they would be deleterious to the weedy relatives. Seeds of a crop × weed hybrid would remain on the hybrid plant, until harvested with the crop seed, without replenishing the weed seed bank. If such hybrid seed is crushed for oil and not used for planting, hybrids and their backcross offspring cannot establish to significant levels in the field.

30

WEEDY AND INVASIVE PLANT GENOMICS

Indeed, one such gene has been introduced into a Brassica sp. and prevents shattering (Østergaard et al. 2006). Why is this of limited utility? The same gene will not work outside of Brassicaceae. Crops and weeds in other families do not have siliques. Indeed, the genes controlling shattering differ among grasses; the genes of wheat, rice, and sorghum are all different from each other (Konishi et al. 2005; Li and Gill 2006).

Lessons To Be Learned

The variability of weeds and the various constellations of properties that confer weediness should warn us against the E. coli = elephants, Arabidopsis = all plants paradigm. This should teach us not to call for, as some have, obtaining the gene sequence of one major weed as “representative”, so “let’s now choose one” (although, see Chapter 4). Dicot Amaranthus is not representative of dicot Conyza in growth habit or weed habitat. Monocot weedy Avena is not representative of monocot Cyperus in their habits nor habitats. The sequence of the crop sorghum (Sorghum bicolor) will not explain the weediness of shattercane (also Sorghum bicolor), nor will knowing the sequence of shattercane explain the weediness of perennial Johnsongrass (Sorghum halepense), even though the latter contains the Sorghum bicolor genome along with that of a cousin. To date, there is no unified theory of weediness. Biology is always more diverse than physics. We may learn much about weediness from comparative functional genomics of various weed and crop species. The more we learn, the more we can conceive novel weed control mechanisms for specific groups of weeds. Weed science needs genomics: (1) to understand weediness, (2) as an adjunct to genetics and relatedness studies, (3) for gene flow studies, both to understand the dynamics and to manipulate them, (4) to find new herbicide targets, (5) to elucidate targets of known herbicides, (6) to understand the mechanisms of evolved herbicide resistance, and (7) to assist in the development of new weed control strategies. Arabidopsis has been of little help in such endeavors, and at times counterproductive, all because people considered it to be a model for weeds—a model so inclusive that it was considered superfluous to study the genomics of other weeds. It is hoped that this lesson has finally been learned.

References Al-Ahmad H, Dwyer J, Moloney MM, Gressel J (2006) Mitigation of establishment of Brassica napus transgenes in volunteers using a tandem construct containing a selectively unfit gene. Plant Biotechnology Journal 4, 7–21. Al-Ahmad H, Gressel J (2006) Mitigation using a tandem construct containing a selectively unfit gene precludes establishment of Brassica napus transgenes in hybrids and backcrosses with weedy Brassica rapa. Plant Biotechnology Journal 4, 23–33. Banta JA, Dole J, Cruzan MB, Pigliucci M (2007) Evidence of local adaptation to coarse-grained environmental variation in Arabidopsis thaliana. Evolution 61, 2419–2432. Bartel B, Last RL (2004) Weed power, translating Arabidopsis. Plant Physiology 135, 601. Buchholtz KP, Grigsby BH, Lee OC, Willard CJ, Volk NJ (1954) Weeds of the North Central States. University of Illinois Agricultural Experiment Station, Urbana. Buell CR (1998) Arabidopsis: A weed leading the field of plant-pathogen interactions. Plant Physiology and Biochemistry 36, 177–186. Carrera E, Holman T, Medhurst A, Dietrich D, Footitt S, Theodoulou FL, Holdsworth MJ (2008) Seed after-ripening is a discrete developmental pathway associated with specific gene networks in Arabidopsis. Plant Journal 53, 214–224. Cipollini D, Stevenson R, Cipollini K (2008) Contrasting effects of allelochemicals from two invasive plants on the performance of a non-mycorrhizal plant. International Journal of Plant Sciences 169, 371–375.

ARABIDOPSIS: THERE IS NO GOOD MODEL FOR WEED GENOMICS

31

Costea M, Weaver SE, Tardif FJ (2005) The biology of invasive alien plants in Canada. 3. Amaranthus tuberculatus (Moq.) Sauer var. rudis (Sauer) Costea and Tardif. Canadian Journal of Plant Science 85, 507–522. Czarnota MA, Paul RN, Dayan FE, Nimbal CI, Weston LA (2001) Mode of action, localization of production, chemical nature, and activity of sorgoleone: A potent PSII inhibitor in Sorghum spp. root exudates. Weed Technology 15, 813–825. Ecker JR (1998) Genes blossom from a weed: The Arabidopsis genome initiative. FASEB Journal 12, A1306–A1306. Federspiel N (2000) Deciphering a weed. Genomic sequencing of Arabidopsis. Plant Physiology 124, 1456–1459. Feltner KC, Hurst HR, Anderson LE (1969) Tall waterhemp competition in grain sorghum. Weed Science 17, 214–216. Fernald ML (1950) Gray’s Manual of Botany, Eighth Edition American Book Co., Boston. Friedmann HC (2004) From Butyribacterium to E. coli, an essay on unity in biochemistry. Perspectives in Biology and Medicine 47, 47–66. Gressel J (1988) Multiple resistances to wheat selective herbicides: new challenges to molecular biology. Oxford Surveys of Plant Molecular and Cell Biology 5, 195–203. Gressel J (1999) Tandem constructs: preventing the rise of superweeds. Trends in Biotechnology 17, 361–366. Gressel J (2002) Molecular Biology of Weed Control. Taylor and Francis, London. Gressel J (2008) Genetic Glass Ceilings: Transgenics for Crop Biodiversity. Johns Hopkins University Press, Baltimore. Gressel J, Al-Ahmad H (2005) Molecular containment and mitigation of genes within crops: prevention of gene establishment in volunteer offspring and feral strains. In: Crop Ferality and Volunteerism Gressel J, ed., CRC Press, Boca Raton. pp. 371–388. Gressel J, Kleifeld Y (1994) Can wild species become problem weeds because of herbicide resistance? Brachypodium distachyon: a case study. Crop Protection 13, 563–566. Hager AG, Wax LM, Stoller EW, Bollero GA (2002) Common waterhemp (Amaranthus rudis) interference in soybean. Weed Science 50, 607–610. Hartzler RG, Bruce B, Nordby D (2004) Effect of common waterhemp (Amaranthus rudis) emergence date on growth and fecundity in soybean. Weed Science 52, 242–245. Holm L, Doll J, Holm J, Pancho E, Herberger J (1997) World Weeds: Natural Histories and Distribution. Wiley, New York. Holm LG, Plucknett JD, Pancho LV, Herberger JP (1977) The World’s Worst Weeds, Distribution and Biology, University of Hawaii Press, Honolulu. Horak MJ, Peterson DE (1995) Biotypes of Palmer amaranth (Amaranthus palmeri) and common waterhemp (Amaranthus rudis) are resistant to imazethapyr and thifensulfuron. Weed Technology 9, 192–195. Konishi S, Lin SY, Ebana K, Izawa T, Sasaki T, Yano M (2005) Molecular cloning of a major QTL, QSH-1, controlling seed shattering habit in rice. Plant and Animal Genomes XIII Conference W306 (ABSTRACT). Kunkel BN (1996) A useful weed put to work: Genetic analysis of disease resistance in Arabidopsis thaliana. Trends in Genetics 12, 63–69. Levisohn R, Spiegelman S (1969) Further extracellular Darwinian experiments with replicating RNA molecules: Diverse variants isolated under different selective conditions. Proceedings of the National Academy of Sciences of the United States of America 63, 805–811. Lewis JJ (2008) http://www.wisdomquotes.com/002720.html, like many others, attributes this to Emerson, but the definition is not found in his complete works online at http://www.emersoncentral.com/search.htm accessed March 23, 2008. Li W, Gill BS (2006) Multiple genetic pathways for seed shattering in the grasses. Functional and Integrative Genomics 6, 300–309. Liljegren SJ, Roeder AHK, Kempin SA, Gremski K, Østergaard L, Guimil S, Khammungkhune D, Yanofsky MF (2004) Control of fruit patterning in Arabidopsis by INDEHISCENT. Cell 116, 843–853. Meazza G, Scheffler BE, Tellez MR, Rimando AG, Romagni JG, Duke SO, Nanayakkara D, Khan IA, Abourashed EA, Dayan FE (2002) The inhibitory activity of natural products on plant p-hydroxyphenylpyruvate dioxygenase. Phytochemistry 60, 281–288. Meyerowitz EM (1989) Arabidopsis, a useful weed. Cell 56, 263–269. Nasrallah ME, Liu P, Nasrallah JB (2002) Generation of self-incompatible Arabidopsis thaliana by transfer of two S locus genes from A. lyrata. Science 297, 247–249. Østergaard L, Kempin SA, Bies D, Klee HJ, Yanofsky MF (2006) Pod shatter resistant Brassica fruit produced by ectopic expression of the FRUITFULL gene. Plant Biotechnology Journal 4, 45–51. Owen MDK, Zelaya IA (2005) Herbicide-resistant crops and weed resistance to herbicides. Pest Management Science 61, 301–311. Patzoldt WL, Tranel PJ, Hager AG (2002) Variable herbicide responses among Illinois waterhemp (Amaranthus rudis and A. tuberculatus) populations. Crop Protection 21, 707–712. Pigliucci M, Kolodynska A (2006) Phenotypic integration and response to stress in Arabidopsis thaliana: a path analytical approach. Evolutionary Ecology Research 8, 415–433.

32

WEEDY AND INVASIVE PLANT GENOMICS

Pratt DB, Clark LG (2001) Amaranthus rudis and A. tuberculatus—one species or two? Journal of the Torrey Botanical Society 128, 282–296. Rensink WA, Buell CR (2004) Arabidopsis to rice. Applying knowledge from a weed to enhance our understanding of a crop species. Plant Physiology 135, 622–629. Rice JW, Guo L, Davis K, et al. (2004) Methods for the identification of inhibitors of serine acetyltransferase activity in plants. US Patent 6,770,452. Spiegelman S (1971) An approach to experimental analysis of precellular evolution. Quarterly Review of Biophysics 4, 213–253. Stahl EA, Dwyer G, Mauricio R, Kreitman M, Bergelson J (1999) Dynamics of disease resistance polymorphism at the Rpm1 locus of Arabidopsis. Nature 400, 667–671. Steckel LE, Sprague CL (2004) Common waterhemp (Amaranthus rudis) interference in corn. Weed Science 52, 359–364. Uscanga-Mortera E, Clay SA, Forcella E, Gunsolus J (2007) Common waterhemp growth and fecundity as influenced by emergence date and competing crop. Agronomy Journal 99, 1265–1270. Volenberg DS, Patzoldt WL, Hager AG, Tranel PJ (2007) Responses of contemporary and historical waterhemp (Amaranthus tuberculatus) accessions to glyphosate. Weed Science 55, 327–333. Warwick SI, Stewart CN Jr. (2005) Crops come from wild plants: How domestication, transgenes, and linkage together shape ferality. In: Crop Ferality and Volunteerism Gressel J, ed. CRC Press, Boca Raton. pp. 9–30. Yanofsky MF, Liljegren SJ (2006) Control of fruit dehiscence in plants by Indehiscent 1 genes. US Patent 7,135,621.

4

Model Weeds For Genomics Research Wun S. Chao and David P. Horvath

Introduction

Various tools and resources are needed for in-depth molecular studies on the functional genomics of weeds. Likewise, studies on the physiological processes and genetic mechanisms that impart or impact weediness would also be beneficial to better understand how weedy and invasive plants make their living. However, such studies require extensive resources that are generally not available to most weed scientists. One way to address this issue is to focus resources and efforts on a few key, or “model” weedy species, which might serve to encourage and facilitate collaborations among weed scientists and scientists working on traditional model species such as Arabidopsis thaliana. By selecting one or a few model weeds, weed scientists might be able to emulate the success of the Arabidopsis research community. The advantages of model organisms lie primarily in the ability to leverage vast resources needed to develop the powerful but expensive genomic-based tools, services, and repositories of information, mutants, and gene libraries. Work on A. thaliana is an excellent example of how research communities can rapidly drive production of tools such as whole genome sequences (AGI 2000), collections of insertion mutations (Alonso et al. 2003), expression databases (Grennan 2006), and development of central repositories of information (http:// www.arabidopsis.org/) to expedite numerous and diverse research efforts. Tools and information developed for A. thaliana can be extrapolated economically to studies in non-model species. Furthermore, there are advantages to having numerous scientists working on a single species, such as opportunities for developing synergistic relationships and increased discoveries in one area of research serendipitously impacting other unrelated studies. In modern biology, there are many instances in which such synergy and serendipity have proven worthwhile for other model species. For example, in A. thaliana, studies on drought resistance identified the same class of regulatory genes involved in responses to cold (Stockinger et al. 1997, Shinozaki and Yamaguchi-Shinozaki 2000), thus strengthening the hypothesized connection between these two stresses. Such connections provided additional information on gene regulatory pathways that rapidly enhanced research efforts in both areas of study. Likewise, similar regulatory genes were identified in both sugar responses and responses to the plant hormone abscisic acid (ABA) (Arenas-Huertero et al. 2000). Unlike the cold- and drought-stress studies, in which some connection was previously suspected, the sugar and ABA studies provided a completely novel linkage and opened up a new avenue of study that has impacted not only plant-environmental interaction research, but also research in plant growth and development. There are countless other instances that illustrate how our understanding of plant biology has benefited from the use of model species. Such devotion of resources and potential for impact would not be possible for all weed species of interest. In addition to exceptional scientific advances, many technical issues and lessons have been learned through the development of model species such as A. thaliana. The evolution of webbased information for sharing technologies and scientific societies are best exemplified by The Arabidopsis Information Resource (TAIR) website (http://www.arabidopsis.org/). TAIR is not 33

34

WEEDY AND INVASIVE PLANT GENOMICS

only a repository for A. thaliana genomic information, materials, and resources; it also provides linkages to many other websites that have been developed to analyze the massive amounts of expression and developmental data derived from the many individual studies on given genes, regulatory and developmental signaling pathways, and physiological processes. Already, lessons learned from A. thaliana have been used and modified for emerging model systems such as poplar, rice, and Medicago trunculata. These models have been useful in answering some questions pertinent to weed science. Indeed, several notable studies and reviews illustrate the impact these model systems have on weed science (Foley 2002; Horvath et al. 2003; Horvath et al. 2006; Basu et al. 2004; Gu et al. 2005; Yuan et al. 2007). Despite the fact that most plant species generally share many genes and physiological functions, no species incorporates all possible biological responses and developmental pathways. Thus, no single species can serve as a model for all plants. For example, although A. thaliana has proven very useful for identifying the key genes involved in flower development, there are notable differences between how these floral regulatory genes are expressed, regulated, and function in other species. For example, although floral regulatory genes such as CONSTANS and FLOWERING LOCUS T are present in plants such as Pharbitis (Ipomoea nil) and regulate flowering time in Pharbitis, these genes do not respond to environmental signals in the same way as they do in A. thaliana (Hayama et al. 2007). Likewise, both A. thaliana and rice are annual species with a floral transition marked by bolting, and thus are unlikely to provide much useful information on endodormancy in vegetative buds, nor on the signals regulating other notable plant reproductive and developmental growth patterns in perennial systems. Consequently, there is a need to develop additional model systems that can help fill in the experimental knowledge gaps left by rice and A. thaliana. Some emerging models such as poplar, tomato, and Medicago will fill some research niches; however, there are still many questions pertaining specifically to weed science that will not be served by these new models. It can be argued that weeds are special cases with regard to many traits such as competitive ability and herbicide resistance that crop models would not address. It is expensive and time consuming to develop models. A considerable amount of equipment and expertise was needed to provide genomic sequencing and expression data as well as to prepare web-based platforms for the distribution of the vast amounts of information that were generated for model species such as A. thaliana and rice. With the completion of these initial massive genome sequencing efforts, much of the resources and equipment used for these efforts are now available for new projects. Consequently, now marks a unique opportunity for leveraging these resources to develop new models that can effectively and efficiently tackle previously untenable questions. Because weeds markedly decrease crop yield, a strong argument can be made for using these resources to improve weed science. However, before such efforts can be undertaken, it would be prudent to consider exactly which species should be used as models.

What Makes A Good Model Species?

Before any scientific community commits its resources to a limited number of model species, it is necessary to explore and understand what attributes comprise a good model system. In the case of weed science, the first and most important step is to identify the important characteristics and traits of both weeds and models and then choose candidates in which weeds and models intersect. Many characteristics (see the box entitled Characteristics Of Weeds That Might Be Considered In Selection Of A Model Experimental System) make weeds particularly persistent

MODEL WEEDS FOR GENOMICS RESEARCH

35

Box 4.1. Characteristics of Weeds that Might Be Considered in Selection of a Model Experimental System*

• • • • • • • • • • • • • • • • • • •

Crop mimicry in life cycle, morphology, or physiology Cross-pollination by wind or unspecialized organisms Dormancy in seeds or vegetative propagules Facultatively self-compatibility Germination in many environments Resistance to control measures, e.g. herbicide, mowing Interspecific interference, e.g. competitive, allelopathic, parasitic Longevity of seeds and vegetative propagules Perennials deep rooted; hard to uproot Perennials readily regenerated from plant fragments Polyploid and low nuclear DNA content Propagules adapted for short and long distance dispersal Rapid growth to flowering Seed appendages for dispersal e.g. awns, barbs, pappus, etc. Seed output high under favorable environmental conditions Seed output under a variety of environmental conditions Prolonged seed production throughout growing season Seed shattering Vegetative reproduction, e.g. root/crown buds, rhizomes, tubers

*Chao et al. 2005 and pernicious in relation to human activities. Baker (1974) proposed the idea of “generalpurpose genotypes”—that weeds evolved to include multiple weedy characteristics that impart tolerance to adverse environments and therefore the potential for expansion in their niches and distributions. Other criteria for the selection of model weeds should also be based on wide and pertinent geographic distributions, serious economic impacts, tractable genome size and growth habits, and ease in genetic manipulation through transformation and/or crossing. Consideration should also include broad organizational and political support (Chao et al. 2005). One of the most important lessons learned from the Arabidopsis era is that a large number of researchers focusing their particular questions of interest on the given species is very powerful. One key attribute of Arabidopsis that was inspiring to plant science researchers was that A. thaliana requires little space or equipment to grow and maintain. This meant that researchers with little infrastructure (e.g. greenhouse or field space) could embark on suitably-sized studies using Arabidopsis thaliana. Also, A. thaliana grows quickly and thus, many experiments, even those that required multiple generations, could be completed quickly. Finally, A. thaliana had several enthusiastic champions who promoted the use of this species and made their stocks of seeds and mutant varieties freely available to any researcher who asked. It is important to note that “investigator territorialism” must be minimized for the model approach to be effective. In addition to these practical attributes, A. thaliana has a very small genome, is easily manipulated genetically, and thanks to some very early efforts by a small group of European researchers, there was a good collection of mutant lines that were ripe for study by newly emerging molecular techniques (Meinke et al. 1998). The large numbers of researchers investigating their particular questions of interest in A. thaliana provided the

36

WEEDY AND INVASIVE PLANT GENOMICS

justification to develop tools and resources such as sequence databases for members of the Arabidopsis research community. Finally, Arabidopsis researchers share a sense of community, punctuated by an annual research conference. No weedy species has all of these important characteristics. However, given the major influence of obtaining critical mass of researchers to warrant investment in expensive genomicsbased resources, weeds that are particularly problematic over a wide geographic range or that have already inspired major combined scientific efforts should be thoroughly considered. Toward that end, a search of thirty commonly studied weeds on Google Scholar indicated the five most studied (based on the number of publications with the scientific or common name of the weed in the title and with reference to weediness) are as follows: Amaranthus ssp., witchweed (Striga asiatica), leafy spurge (Euphorbia esula), wild oat (Avena sativa), and nutsedge (Cyperus ssp.). The number is likely an indication not only of scientific interest, but also ease of study. Consequently, any of these taxa would likely have significant scientific support as model species. Other attributes of a good model species include a small diploid genome size to facilitate genomic sequencing efforts and cloning of specific genes, small stature and ease of growth in greenhouse and lab environments, facile genetic manipulation, and well developed protocols for tissue culture and transformation. Although many of these weeds have generated great scientific interest, most of them have significant problems that need to be resolved before they can be considered as excellent model systems. For instance, many weeds cannot be easily transformed genetically. Because testing certain specific hypotheses often requires the ability to turn on and off the expression of various genes, a good transformation system is important. Of these top five weeds, only leafy spurge and amaranthus have been transformed, albeit with very low efficiencies. The genetics of these wild species is sometimes problematic. For instance, leafy spurge is a semi-, self-incompatible auto-allo hexaploid, and amaranthus is a complex group of sexually compatible dioecious and monoecious species. Moreover, most weeds are not very small in physical stature (relative to A. thaliana), and thus require considerable greenhouse space for screening of specific genetic mutants and propagation. It is not absolutely necessary to develop model weeds with little biological information or starting genomics tools. There are several well developed model crops that are highly similar to various weedy species that might be co-opted. The genomics resources and information developed for these models can be directly applied to assist in analyzing certain weeds, thus bypassing the most expensive and difficult requirements for model development. Many of these model crops have been mentioned by Chao et al. (2005). This chapter updates the information on genomics resources of the crops and their value in weed research. It also introduces and updates genomics tools under development for some weedy species. See Chao et al. (2005) for more information.

Leveraging From Other Models Tomato For Nightshades

Tomato (Solanum lycopersicum) is a member of the dicot family Solanaceae, which contains valuable vegetable species such as potato (Solanum tuberosum), tobacco (Nicotiana tabacum), eggplant (Solanum melongena), and pepper (Capsicum annuum), and important weeds such as nightshades (Solanum ssp.), horse nettle (Solanum carolinense), Jerusalem cherry (Solanum pseudocapsicum), and tropical soda apple (Solanum viarum). Despite differences in genome

MODEL WEEDS FOR GENOMICS RESEARCH

37

sizes among Solanaceae members, most species possess the same chromosome number of 12 (2n = 24). The haploid genome size of tomato (Figure 4.1) is 953 Mb, which is among the smallest in the Solanaceae family (Arumuganathan and Earle 1991b). Tomato was considered to be an ideal model of the Solanaceae family because it has a relatively small genome, short generation time, and routine methods for Agrobacterium-mediated transformation. Also, there are ample genetic and genomic resources available including high density genetic maps (Fulton et al. 2002) and more than 200,000 expression sequence tag (EST) accessions (http://plantta.tigr. org/). Commercially available microarrays from various solanaceous species are available, and studies testing the applicability of potato microarrays to study gene expression in other solanaceous species proved successful (Rensink et al. 2005). This suggests that these tools could be used to characterize gene expression in solanaceous weeds. Nightshades (Solanum spp.) (Figure 4.2) are annual to short-lived perennials in the Solanaceae family. Black nightshade (S. nigrum L.) is the best known noxious weed among nightshade species (Ogg et al. 1981; Defelice 2003) and is reported as a weed in more than thirty-seven crops and sixty-one countries around the world (Holm et al. 1991). Nightshades, in general, reproduce by seeds. Berries can contain fifteen to ninety-six seeds and a single plant can produce up to 30,000 seeds in a single season. Seeds remain viable after years in the soil seed bank and can germinate intermittently under favorable conditions (Defelice 2003). Nightshades also are toxic, compete with crops, impede harvest, and reduce crop quality by seed discoloration (Cooper and Johnson 1984; Lampe and McCann 1985). In addition, some nightshades have evolved resistance to photosystem II and ALS inhibitors, and bipyridiliums (Heap 2008). Black nightshade is a hexaploid with a chromosome number of 2n = 72. Eastern black nightshade (S. ptycanthum), however, is common to the U.S. and is a diploid species with a chromosome number of 2n = 24, the same as tomato. It is possible that transformation systems developed for the various solanaceous crops could be used to transform nightshades. Also, given the sequence similarity of many genes from solanaceous crops, identifying and cloning genes of interest within the nightshades should be relatively efficient. Furthermore, most of the markers developed for mapping in tomato should be useful for mapping in the nightshade complex. Therefore, tomato resources might be of great assistance to study weedy characteristics of eastern black nightshade and other nightshade species. Besides being an important weed with several characteristics that might make it a suitable model species, nightshades have several particularly weedy traits that could be studied using resources developed from tomato. For example, some nightshade species are annuals, while other closely related sub-species are perennials (Hobbs et al. 2000). Some closely related nightshades also have different invasiveness characteristics (John Masiunas, personal communication). Direct comparison of related varieties of nightshade using tomato microarrays could identify the mechanisms by which the species has developed perennial growth patterns and potentially identify genes involved in different levels of invasiveness. Likewise, markers developed for tomato could allow mapping of genes and quantitative trait loci (QTL) that influence invasiveness and perennial growth.

Wheat For Goatgrass

Cultivated wheat (Triticum aestivum) (Figure 4.3) is the most widely grown crop in the world. It is an allohexaploid species with three homeologous genomes (2n = 6× = 42; AABBDD). The genome sizes of wheat are 17,000 Mb (Bennett and Smith 1976), and more than 80% of

Figure 4.1. Tomato (Solanum lycopersicum).

Figure 4.2. Black nightshade (Solanum nigrum). (http://aphotoflora.com/DevonandCornwall/Solanum%20nigrum16-0906.jpg)

38

MODEL WEEDS FOR GENOMICS RESEARCH

39

Figure 4.3. Cultivated wheat (Triticum aestivum).

the genome is repetitive noncoding DNA (Smith and Flavell 1975). The large genome size, combined with the high percentage of repetitive DNA, makes comprehensive sequencing of the wheat genome extremely challenging. However, there are rich genomic resources for wheat, as more than 850,000 ESTs are in dbEST (http://wheat.pw.usda.gov/genome/), which provides direct information on the mature transcripts for the coding portion of the genome. In addition, more than 20,000 EST loci have been mapped on the twenty-one chromosomes (http:/ www.ncbi.nlm.nih.gov/dbEST). In addition to these resources, hundreds of aneuploid (one or a few chromosomes above or below the haploid number) stocks have been developed owing to the homoeology existing among the three component genomes, which allows various aneuploidy to be tolerated. Wheat aneuploid stocks include monosomic (loss of one chromosome), nullisomic (loss of a chromosome pair), trisomic (three chromosomes instead of two for a given pair), tetrasomic (four chromosomes instead of two for a given pair), and telosomic lines (half of a pair of chromosomes is missing). These lines are valuable tools for the construction of a molecular map of wheat (http://www.jic.ac.uk/GERMPLAS/prec_ce/). In 2005, the International Wheat Genome Sequencing Consortium (IWGSC) was established to facilitate and coordinate international efforts toward obtaining the complete sequence of the

40

WEEDY AND INVASIVE PLANT GENOMICS

common wheat genome, and the French National Institute for Agricultural Research (INRA) led the project to sequence the largest wheat chromosome (3B) based on a chromosomespecific approach (Leroy et al. 2006). Transformation protocols also exist for wheat (Jones 2005). Many of these resources can be used to study a serious weed, jointed goatgrass (Aegilops cylindrical Host). Jointed goatgrass (Figure 4.4) is a winter annual grass; it spreads exclusively by seed. Jointed goatgrass is found in most major U.S. winter wheat (Triticum aestivum) production regions and has infested more than 5 million acres of winter wheat cropland. Total losses from jointed goatgrass infestation in the western U.S. annually exceed $145 million (Westbrooks 1998). Jointed goatgrass seeds are similar in size and shape to wheat seeds, making them difficult to separate from one another. Jointed goatgrass seeds are dormant after shattering, require afterripening to germinate, and remain viable for three to five years in the soil seed bank (Donald and Ogg 1991). Few, if any, herbicides can selectively control this weed in winter wheat because of species similarity (Seefeldt et al. 1998; Zemetra et al. 1998). These characteristics make crop protection extremely difficult and therefore herbicide-resistant transgenic wheat has been considered (Anderson et al. 2004). However, inter-specific hybridization between these species occurs and would likely result in herbicide-resistant jointed goatgrass (Seefeldt et al. 1998; Zemetra et al. 1998; Hanson et al. 2005). Jointed goatgrass is an allotetraploid and shares the D genome with winter wheat (Donald and Ogg 1991). It is thus a competitive weed that mimics the life cycle of wheat. However, this feature also allows for the utilization of wheat genomics resources for goatgrass genomics research. Goatgrass also has some characteristics that might make it an excellent model weed. For example, goatgrass is a grassy weed that is easily propagated and manipulated in the green-

Figure 4.4. Jointed goatgrass (Aegilops cylindrical). (http://jointedgoatgrass.wsu.edu/jointedgoatgrass/gallery/Photos/ JGG1-close.jpg)

MODEL WEEDS FOR GENOMICS RESEARCH

41

house. Because it is so similar to wheat, it could serve as the model for evolution and selection in crop mimicry. Also, since its seed dormancy is one of the characteristics that contributes to difficulty of control, resources and studies on wheat seed dormancy and pre-harvest sprouting could be directly applicable to goatgrass and vice versa.

Rice For Weedy Rice

Rice (Figure 4.5) is one of the world’s most important cereal crops. It belongs to the Poaceae family which includes cereals, turf, and many serious weeds. Rice contains 389 Mb per haploid genome (IRGSP 2005), which is the smallest among all the cereal crops and only three times larger than the Arabidopsis thaliana genome. Rice can be transformed routinely using Agrobacterium. More than 400,000 rice ESTs and 32,000 full-length cDNA clones are deposited in dbEST (Vij et al. 2006). Microarray chips are commercially available (Meyers et al. 2004; Rensink and Buell 2004). These features make rice an excellent genetic model for cereal crops and grasses. The map-based, whole genome rice (Oryza sativa ssp. japonica cv. Nipponbare) sequence was reported by IRGSP in 2004 (http://rgp.dna.affrc.go. jp/IRGSP/celebrates/celebrates.html), which has had a profound impact on rice and cereal grass research.

Figure 4.5. Rice (Oryza sativa subspecies indica).

42

WEEDY AND INVASIVE PLANT GENOMICS

With completion of the rice genome, about 19,000 simple sequence repeat (SSR) markers have been identified and are publicly available (Vij et al. 2006), which will aid marker-assisted breeding and map-based gene cloning of related weed species. These resources ultimately allow scientists to create more desirable rice genotypes, e.g., disease- and stress-resistant, more nutritious, and varieties that are more resistant to pests. In the perspective of weed research, all these resources can be directly applied to investigate a serious annual weed, weedy rice (Oryza sativa L.). Weedy (red) rice (Figure 4.6) is an important annual weed worldwide, especially in cultivated rice. Weedy rice could be used to investigate most of the characteristics in the above box, Characteristics Of Weeds That Might Be Considered In Selection Of A Model Experimental System. Weedy rice has been used to examine physiological, biochemical, and genetic aspects of seed germination, dormancy, and after-ripening (Leopold et al. 1988; Footitt and Cohn 1995; Gu et al. 2004), crop interference (Diarra and Talbert 1985), allelopathy (Duke et al. 2003), gene flow between crops and weeds (Gealy et al. 2003), rhizomatousness (Hu et al. 2003), and other weedy related traits (Gu et al. 2005). Topics of recent interest include domestication of cultivated rice (Bres-Patry et al. 2001), weedy traits such as shattering and dormancy (Cai and Morishima 2000, Gu et al. 2005), and gene flow from cultivated rice to weedy rice (Chen et al. 2004). One important motivation for using weedy rice as a model system is that map-based cloning of seed dormancy genes is possible because of the ample available genomic resources in cultivated rice, which is the same species as weedy rice. In fact, currently a dormancy gene, qSD7-1, has been cloned from the weedy rice based on the rice genomics information (Gu et al. 2006, 2007), providing

Figure 4.6. Weedy rice (Oryza sativa). (Courtesy of Xingyou Gu.)

MODEL WEEDS FOR GENOMICS RESEARCH

43

a good example of using the resources of a model crop in weed research. In this instance, cloning of qSD7-1, the gene encoding a transcription factor which underlies a QTL previously shown to have pleiotrophic effects on dormancy, red pericarp color, and grain weight, would have been very difficult without the assistance of rice fine maps and whole genome sequence.

Sorghum For Johnsongrass

Sorghum (Sorghum bicolor) is the world’s fifth most important cereal crop, following wheat, rice, maize, and barley. The Sorghum genus also includes one of the world’s most noxious weeds, Johnsongrass (Sorghum halepense) (Figure 4.7). Sorghum grows well on marginal lands and can endure both drought and water-logging (Smith and Frederiksen 2000), both environments in which weedy species are well adapted (See the box entitled Characteristics Of Weeds That Might Be Considered In Selection Of A Model Experimental System). The sorghum genome is denoted by more than 200,000 ESTs, representing a 22,000 unigene set. In addition, many physical and genetic maps and large bacterial artificial chromosome (BAC) libraries are available (Sorghum Genomics Planning Workshop Participants 2005). The genome size of sorghum ranges from 700 Mb to 772 Mb (Arumuganathan and Earle 1991a; Peterson et al. 2002). Large-scale shotgun sequencing of sorghum was started at the end of 2005 and

Figure 4.7. Johnsongrass (Sorghum halepense). (http://plants.usda.gov/java/profile?symbol=SOHA&photoID=soha_022_ avp.tif)

44

WEEDY AND INVASIVE PLANT GENOMICS

completed in 2007 (http://www.phytozome.net/sorghum). These genomic resources are excellent tools to study the weedy characteristics of Johnsongrass. Johnsongrass is a perennial, and a serious weed within the Poaceae family. Holm (1969) listed Johnsongrass as one of the ten worst weeds in the world. It is a noxious or prohibited weed in twenty-two U.S. states (http://invader.dbs.umt.edu/Noxious_Weeds/noxlist.asp last accessed September 25, 2007). Johnsongrass can interfere with the production of crops such as cotton (Gossypium spp.), corn (Zea mays), sorghum, soybean (Glycine max), and sugarcane (Saccharum officinarum) (CABI 2004a). It reproduces through seeds and rhizomes. A single plant can produce more 28,000 seeds per year (Monaghan 1979), which remain viable and germinate intermittently for as long as six years (Leguizamon 1986). Johnsongrass can produce 8 kg fresh weight and 70 m of rhizomes, per plant, in a single growing season (Monaghan 1979), and this vigorous rhizome system can be spread effectively by tillage. Paterson et al. (1995) identified QTLs affecting rhizomatousness and tillering using progenies from the cross between S. bicolor and S. propinquum. To date, Johnsongrass has evolved resistance to several groups of herbicides: ACCase and ALS inhibitors, dinitroanilines, and others (Heap 2008). The ability to overcome herbicidal control poses a great threat to crop production. Sorghum halepense has a chromosome number of either 2n = 20 (diploid) or 2n = 40 (tetraploid). Johnsongrass, the tetraploid species found in the U.S., contains 1,617 Mb per haploid genome (Bennett and Leitch 2003). The species likely originated from naturally occurring crosses between Sorghum bicolor (2n = 20) and Sorghum propinquum (a grassy weed, 2n = 20) (CABI 2004a). Consequently, because of its close relationship to cultivated sorghum and growth habits, Johnsongrass makes a good model weed for studying perennial growth, and vegetative reproduction in grassy weeds. Rudimentary genomics tools are available for this weed. For example, 1,200 ESTs were generated from a rhizome cDNA library of Johnsongrass (http://fungen.org/Index.htm).

Genomics Tools For Weeds That Are Under Development Canada Thistle

Canada thistle is a competitive broadleaf perennial in the Asteraceae family. It is a noxious weed or prohibited weed in forty-three states (http://invader.dbs.umt.edu/Noxious_Weeds/ noxlist.asp), the highest number of states for any weed on the list. It is also a noxious weed in Canada and many other areas of the world (Holm et al. 1977). Canada thistle reproduces from vegetative buds and from seed. Root fragments as small as 8 mm long by 3 mm in diameter are able to develop new plants (Moore 1975). Seed production is prolific and they can be easily spread long distances. An analysis of Canada thistle genotypes in North American populations indicates gene flow has occurred between Alaskan and Maryland populations (Tracey Slotta, personal communication). Seeds can remain viable for up to twenty years in soil seed banks (Madsen 1962). Debris of Canada thistle is reported to have allelopathic effects on surrounding plants (Donald 1994) and ecotypes with resistance to synthetic auxins have been found (Heap 2008). In addition, attempts at biological control have proven problematic, suggesting additional research is required (Louda et al. 1997; Jordon-Thaden and Louda 2003). Thus, this plant has a multitude of characteristics (see the box entitled Characteristics Of Weeds That Might Be Considered In Selection Of A Model Experimental System) that can be studied to improve weed management and increase the communities’ understanding of the physiological and genetic mechanisms underlying various weedy traits.

MODEL WEEDS FOR GENOMICS RESEARCH

45

Canada thistle is a diploid (2n = 34) with a moderate genome size (slightly more than three times the size of rice) of 15–19 Mbp per haploid genome (Bennett and Leitch 2003). This deep-rooted perennial weed is well adapted to a variety of environmental and edaphic conditions in temperate regions infesting agronomic and horticultural crops, rangelands, turf and urban landscapes, riparian areas, and recreational and natural lands. Canada thistle also serves as an alternate host for insects and pathogenic microorganisms that attack various crops (Donald 1994). Because it has a wide geographic distribution and impacts multiple ecosystems, further research on Canada thistle is likely to garner political support. Recently, the U.S. Department of Agriculture-Agricultural Research Service (USDA-ARS) and National Science Foundation (NSF) have supported investigations on the population genetics of Canada thistle and related species, and a whole plant normalized cDNA library has been developed and used to initiate a small collection of ESTs (James V. Anderson, personal communication).

Waterhemp

Waterhemp (Amaranthus spp.) is a troublesome annual broadleaf weed with a wide geographic distribution. Waterhemp is native to the Midwest U.S.; however, populations of waterhemp have spread rapidly in corn and soybean fields since the late 1980s and 1990s (Hartzler 2003). Strong weedy characteristics and development of herbicide-resistant biotypes have contributed to the increased frequency and severity of waterhemp infestations (Hartzler et al. 2004; Nordby and Hartzler 2004). Such rapid proliferation is not common in agriculture and has drawn the attention of weed scientists. Waterhemp has many weedy characteristics that are desirable in a model weed. It is a dioecious species that cross-pollinates, which leads to high genetic variability. Waterhemp can produce 3,000 to 300,000 seeds per plant under unfavorable or favorable conditions, respectively (Hartzler et al. 2004). The high seed production increases the probability of seeds being dispersed over long distances. These seeds display dormancy (Leon and Owen 2003) and can germinate under a wide range of environmental conditions throughout the growing season. Waterhemp seedlings grow very fast; the relative growth rate is 50% to 70% greater than that of other tested annual weedy species (Horak and Loughin 2000; Seibert and Pearce 1993). The high growth rate of waterhemp has been attributed to the fact that it is capable of C4 photosynthesis, which is rare among dicots (Pearcy and Ehleringer 1984). Most importantly, waterhemp has evolved resistance to ALS-inhibiting, PPO, triazine, and glyphosate herbicides (Boerboom and Owen 2006; Heap 2008). A waterhemp population was also found to have developed three-way resistance (ALS, PPO, and triazine) (Boerboom and Owen 2006). These traits enable waterhemp to be a rapid invader in many cropping systems. Traditionally, waterhemp has been divided into two species, common waterhemp (Amaranthus rudis) and tall waterhemp (Amaranthus tuberculatus) (Sauer 1955). However, Pratt and Clark (2001) have questioned whether common and tall waterhemp are in fact one polymorphic species. Common waterhemp infestations are frequent from Nebraska to Texas, while tall waterhemp is prevalent in Indiana and Ohio. Both species are widespread in Iowa, Illinois, and Missouri (Hartzler 2003). Waterhemp has a relatively small to moderate genome size (657 Mbp per haploid genome in tall waterhemp; Jeschke et al. 2003). It produces a large number of small seeds (10,000 to 480,000 per plant) that could facilitate screening for phenotypes in seedling stage (Trucco et al. 2005). Substantial efforts have been made to develop genomics resources including genomic and cDNA sequence data, a BAC library, microsatellite markers, and a recombinant inbred line. See Chapter 5 for more information.

46

WEEDY AND INVASIVE PLANT GENOMICS

Leafy Spurge

Leafy spurge (Euphorbia esula) is an invasive perennial broadleaf weed. It is a member of the genetically diverse Euphorbiaceae family that includes important crops, horticultural, weedy, and endangered species such as Akoka (Chamaesyce spp.) and telephus spurge (Euphorbia telephioides) (Chao et al. 2005). Leafy spurge is a highly competitive and invasive plant found in at least thirty-five states and six Canadian provinces (CABI 2004b). Leafy spurge has many important weedy characteristics including vegetative reproduction, seed and bud dormancy, seed longevity, resistance to control measures, a deep and extensive root system, and early spring growth and flowering. Leafy spurge is being used as a model system to study growth and development of vegetative propagules because of its abundance of underground adventitious buds that form on the crown and roots (commonly referred to as crown and root buds) and its relatively fast growth rate under greenhouse and field conditions. Life cycle time from propagation of meristem cuttings in the greenhouse to development of visible crown and root buds can be as short as two months. An average population of 2,800 leafy spurge plants can be maintained in a 10 × 10 meter greenhouse. Leafy spurge is also easily maintained in garden plots, and portable potting systems have been designed that allow plants grown under field conditions to be transferred to controlled environments without major disturbance to the root system. Currently, specific leafy spurge genotypes with high regeneration rates have been identified and maintained using tissue culture technique (Xu et al. 2008). In addition, an Agrobacterium-mediated transformation system is being developed. Currently, a single transgenic leafy spurge was obtained and confirmed by both southern blot analysis and beta glucuronidase (GUS) -staining (unpublished data). Efforts are underway to improve the efficiency of transformation protocol. Leafy spurge has a relatively complex genetic makeup (Schulz-Schaeffer and Gerhardt 1987, 1989). The chromosome number of leafy spurge varies from 2n = 48 to 2n = 60, and a hexaploid (2n = 60) is most prevalent (CABI 2004b). The hexaploid species of leafy spurge contains 2,069 Mbp per haploid genome. Of all the potential model perennial broadleaf weeds being considered, leafy spurge might lead the way in genomics resources. A leafy spurge EST database was generated from a normalized cDNA library constructed from whole plant leafy spurge tissues subjected to multiple abiotic and biotic stresses, different dormancy states, and growth induction regimes (Anderson et al. 2007). This EST database, being developed by USDA-ARS and the University of Illinois, contains 45,314 high-quality sequences which were assembled into 23,472 unique sequences representing 19,015 unigenes. In addition to the normalized library, three other cDNA libraries are available and have been used to generate additional EST resources. One cDNA library was developed from three-day growth-induced root buds (Anderson and Horvath 2001), and was originally used to isolate ESTs for development of preliminary microand macro-arrays (Anderson et al. 2004). This library was designed to allow two hybrid screening for protein-protein interactions. Two subtracted libraries (forward, preferentially expressed in growing buds and reverse, preferentially expressed in dormancy buds) have been constructed and used to screen for preferentially expressed dormancy genes (Jia et al. 2006). A leafy spurge genomic library is available and serves as an important resource for identifying and characterizing cis-acting elements. Currently, microarrays containing 19,808 leafy spurge and 4,129 cassava DNA probes were developed by USDA-ARS and the University of Illinois (both species are members of the Euphorbiaceae family). These microarrays have been used to examine transcriptome profiles associated with seasonal dormancy transitions, and responses to drought stress and pathogens.

MODEL WEEDS FOR GENOMICS RESEARCH

47

In addition to studies on bud dormancy and vegetative development, leafy spurge could serve as a model for understanding invasiveness. Leafy spurge in its native range is not considered to be an invasive plant. Transcriptome comparisons between ecotypes collected in the native range compared to those from its introduced range could help identify genes and physiological processes involved in invasiveness.

Brachypodium

Brachypodium distachyon, commonly called purple false brome, is an invasive, monocot, annual weed. B. distachyon is a weed of grassland and woodland (http://www.cal-ipc.org/ip/ inventory/weedlist.php) and is found in at least six states in the U.S. (http://plants.usda.gov/ java/profile?symbol=BRDI2). In Israel, B. distachyon has evolved resistance to photosystem II inhibitor herbicides and infests roadsides (http://www.weedscience.org/Case/Case. asp?ResistID=95). B. sylvaticum (false brome), a perennial member of the genus, is a more serious invasive species. B. sylvaticum displaces tree seedlings and takes over forests, meadows, and rangeland in the state of Oregon at an alarming rate (USDA Forest Service 2003). It tolerates a wide range of habitats, reproduces rapidly from seeds, and tends to form large coalesced clumps (USDA Forest Service 2003). Rich genomics resources have been developed from B. distachyon since it is used as a model for grain crops (wheat and barley in particular), forages, turfgrasses, and herbaceous biofuel crops such as switchgrass (http://www.gramene.org/species/brachypodium/brachypodium_ intro.html). Many attributes make B. distachyon an excellent model organism, including a small genome (355 Mb per haploid genome), diploid accessions, a series of polyploid accessions (tetraploid and hexaploid), a small physical stature, inbreeding, simple genetics, a short life cycle (two months), and simple growth requirements (Draper et al. 2001; Garvin 2007, Garvin et al. 2008). In addition, it is amenable to efficient transformation procedures (Christiansen et al. 2005; Vogel et al. 2006a). Genomics resources being developed include a set of inbred diploid Brachypodium lines (Vogel et al. 2006a), more than 20,000 ESTs and 8,500 potential SSR markers (Vogel et al. 2006b; http://brachypodium.pw.usda.gov/), and deep bacterial artificial chromosome (BAC) libraries and high-density BAC colony filters (Huo et al. 2006, 2008). In 2006, the U.S. Department of Energy (DOE) supported a project to sequence its genome and an additional 200,000 ESTs of B. distachyon. This work was expected to be completed by the end of 2007 (Garvin 2007, Garvin et al. 2008). In fact, a draft of 8× genome coverage has been completed (http://www.brachypodium.org/). Sequence information will help scientists understand how genes work in concert to control the growth and development of an entire organism and identify genes of interest in a given region of the genome directly. Furthermore, these genomics resources should be useful for investigating weedy characteristics of Brachypodium species and many other related annual and perennial monocot weeds. Weed scientists should thus exploit these resources pertinent to their research interests.

Conclusions

Pimentel et al. (2005) estimate that weeds cause an overall 12% reduction in crop yields in U.S. agriculture. This represents about $33 billion in lost crop production annually. Therefore, it is necessary to perform basic research to elucidate genes and pathways regulating weedy

48

WEEDY AND INVASIVE PLANT GENOMICS

characteristics. New information can be used to develop alternative strategies to improve the effectiveness of existing control measures or to develop novel control strategies. Selection of good model weeds will facilitate the achievement of these goals. Ideally, a model weed should have multiple weedy characteristics; serious economic impact; wide geographic distribution; political support; and pre-existing tools such as EST databases, established protocols for growth, transformation, etc.; and information on biology and ecology. Because no single weed will serve as a general purpose model for all weedy characteristics, it is crucial to select the most appropriate models to efficiently address the scientific questions posed in weed science. In this review, we introduce a few weedy species that might be studied using genomics tools developed from their closely related crop species. We also present a number of weedy species that have genomics tools under development. Based on the criteria set in this review, many other model weeds could be studied as well. The cost to develop some of the genomic tools such as ESTs and molecular markers is continually decreasing because of new sequencing technologies and greater efficiencies of scale. The application of emerging technologies such as microarrays, real-time PCR, and proteomics are also becoming less expensive. These exciting developments in technology allow more and more weed scientists to incorporate genomic tools and emerging technologies into their research programs. However, obtaining a full understanding of weedy characteristics from any weed requires multifaceted approaches: not only ESTs and microarrays, but also genetic linkage maps for comparative genomics, transformation systems, mutagenized lines, and genome sequencing (Basu et al. 2004). The task cannot be accomplished without focused financial resources and involvement of various research disciplines including ecology, physiology, and bioinformatics, to name a few. Developing model weed systems will ensure that focused funding and research occur in the weed science community. In addition, through careful planning and collaborative efforts, resources can be used effectively.

References AGI (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796–815. Alonso JM, Stepanova AN, Leisse TJ, et al. (2003) Genome-wide insertional mutagenesis of Arabidopsis thaliana. Science 301, 653–657. Anderson JA, Matthiesen L, Hegstad J (2004) Resistance to an imidazolinone herbicide is conferred by a gene on chromosome 6DL in the wheat line cv. 9804. Weed Science 52, 83–90. Anderson JV, Delseny M, Fregene MA, Jorge V, Mba C, Lopez C, Restrepo S, Soto M, Piegu B, Verdier V, Cooke R, Tohme J, Horvath DP (2004) An EST resource for cassava and other species of Euphorbiaceae. Plant Molecular Biology 56, 527–539. Anderson JV, Horvath DP (2001) Random sequencing of cDNAs and identification of mRNAs. Weed Science 49, 590–597. Anderson JV, Horvath DP, Chao WS, Foley ME, Hernandez AG, Thimmapuram J, Liu L, Gong GL, Band M, Kim R, Mikel MA (2007) Characterization of an EST database for the perennial weed leafy spurge: an important resource for weed biology research. Weed Science 55, 193–203. Arenas-Huertero F, Arroyo A, Zhou L, Sheen J, Léon P (2000) Analysis of Arabidopsis glucose insensitive mutants, gin5 and gin6, reveals a central role of the plant hormone ABA in the regulation of plant vegetative development by sugar. Genes and Development 14, 2085–2096. Arumuganathan K, Earle E (1991a) Estimation of nuclear DNA content of plants by flow cytometry. Plant Molecular Biology Reporter 9, 208–218. Arumuganathan K, Earle ED (1991b) Nuclear DNA content of some important plant species. Plant Molecular Biology Reporter 9, 208–218. Baker HG (1974) The evolution of weeds. Annual Review of Ecology and Systematics 5, 1–24.

MODEL WEEDS FOR GENOMICS RESEARCH

49

Basu C, Halfhill MD, Mueller TC, Stewart CN Jr. (2004) Weed genomics: new tools to understand weed biology. Trends in Plant Science 9, 391–398. Bennett MD, Leitch IJ (2003) Plant DNA C-values Database Online Release 2.0, January 2003. www.kew.org/cval Bennett MD, Smith JB (1976) Nuclear DNA amounts in angiosperms. Philosophical Transactions of the Royal Society of London B 274, 227–274. Boerboom C, Owen M (2006) Facts about glyphosate-resistant weeds. In: Glyphosate, Weeds, and Crops. http://www.ces. purdue.edu/extmedia/GWC/GWC-1.pdf. Bres-Patry C, Lorieux M, Clement G, Bangratz M, Ghesquiere A (2001) Heredity and genetic mapping of domestication— related traits in temperate Japonica weedy rice. Theoretical and Applied Genetics 102, 118–126. CABI (2004a) Sorghum halepense. In: Crop Protection Compendium, 2004 edition. Wallingford, UK: CAB International. CABI (2004b) Euphorbia esula (original text by Chao W and Anderson JV). In: Crop Protection Compendium, 2004 edition. Wallingford, UK: CAB International. Cai HW, Morishima H (2000). Genomic regions affecting seed shattering and seed dormancy in rice. Theoretical and Applied Genetics 100, 840–846. Chao WS, Horvath DP, Anderson JV, Foley M (2005) Potential model weeds to study genomics, ecology, and physiology in the 21st century. Weed Science 53, 927–937. Chen LJ, Lee DS, Song ZP, Suh HS, Lu BR (2004) Gene flow from cultivated rice (Oryza sativa) to its weedy and wild relatives. Annals of Botany 93, 67–73. Christiansen P, Andersen CH, Didion T, Folling M, Nielsen KK (2005) A rapid and efficient transformation protocol for the grass Brachypodium distachyon. Plant Cell Reports 23, 751–758. Cooper MR, Johnson AW (1984) Poisonous Plants in Britain and their Effects on Animals and Man. Her Majesty’s Stationery Office, London, England. p 305. Defelice MS (2003) The black nightshades, Solanum nigrum L. et al.—poison, poultice, and pie. Weed Technology 17, 421–427. Diarra ARJ, Talbert RE (1985) Interference of red rice (Oryza sativa) with rice (O. sativa). Weed Science 33, 644–649. Donald WW (1994) The biology of Canada thistle (Cirsium arvense). Reviews in Weed Science 6, 77–101. Donald WW, Ogg AG Jr. (1991) Biology and control of jointed goatgrass (Aegilops cylindrica), a review. Weed Technology 5, 3–17. Draper J, Mur LAJ, Jenkins G, Ghosh-Biswas GC, Bablak P, Hasterok R, Routledge APM (2001) Brachypodium distachyon. A new model system for functional genomics in grasses. Plant Physiology 127, 1539–1555. Duke SO, Baerson SR, Dayan FE, et al. (2003) United States Department of Agriculture-Agricultural Research Service research on natural products for pest management. Pest Management Science 59, 708–717. Foley ME (2002) Weeds, seeds, and buds—opportunities and systems for dormancy investigations. Weed Science 50, 267–272. Footitt S, Cohn MA (1995) Seed dormancy in red rice (Oryza sativa). IX. Embryo fructose-2,6-bisphosphate during dormancy breaking and subsequent germination. Plant Physiology 107, 1365–1370. Fulton TM, Van der Hoeven R, Eannetta NT, Tanksley SD (2002) Identification, analysis, and utilization of conserved ortholog set markers for comparative genomics in higher plants. Plant Cell 4, 1457–1467. Garvin DF (2007) Brachypodium: a new monocot model plant system emerges. Journal of the Science of Food and Agriculture 87, 1177–1179. Garvin DF, Gu YQ, Hasterok R, Hazen SP, Jenkins G, Mockler TC, Mur LAJ, Vogel JP (2008) Development of genetic and genomic research resources for Brachypodium distachyon, a new model system for grass crop research. Crop Science 48, S-69–S-84. Gealy DR, Mitten DH, Rutger JN (2003) Gene flow between red rice (Oryza sativa) and herbicide-resistant rice (O-sativa): implications for weed management. Weed Technology 17, 627–645. Grennan AK (2006) Genevestigator. Facilitating web-based gene-expression analysis. Plant Physiology 141, 1164–1166. Gu XY, Foley ME, Horvath D, Anderson JV, Chen ZX (2007) Positional cloning of qSD7-1, a dormancy locus associated with red pericarp color in weedy rice. The Annual Meeting of the American Society of Plant Biologists, July 7–11. Chicago, IL (ABSTRACT). Gu XY, Kianian SF, Foley ME (2004) Multiple loci and epistases control genetic variation for seed dormancy in weedy rice (Oryza sativa). Genetics 166, 1503–1516. Gu XY, Kianian SF, Foley ME (2005) Seed dormancy imposed by covering tissues interrelates to shattering and seed morphological characteristics in weedy rice. Crop Science 45, 948–955. Gu XY, Kianian SF, Foley ME (2006) Isolation of three dormancy QTLs as Mendelian factors in rice. Heredity 96, 93–99. Gu XY, Kianian SF, Hareland GA, Hoffer BL, Foley ME (2005) Genetic analysis of adaptive syndromes interrelated with seed dormancy in weedy rice (Oryza sativa). Theoretical and Applied Genetics 110, 1108–1118.

50

WEEDY AND INVASIVE PLANT GENOMICS

Hanson BD, Mallory-Smith CA, Price WJ, Shafii B, Shill DC, Zemetra RS (2005) Interspecific hybridization: Potential for movement of herbicide resistance from wheat to jointed goatgrass (Aegilops cylindrica). Weed Technology 19, 674–682. Hayama R, Agashe B, Luley E, King R, Coupland G (2007) A circadian rhythm set by dusk determines the expression of FT homologs and the short-day photoperiodic flowering response in Pharbitis. Plant Cell 19, 2988–3000. Hartzler RG (2003) Waterhemp—The perfect weed? ISU Weed Science Online. WSSA-North American Weed Management Association Invasive Plant Species Workshop, Kansas City, MO. Feb. 12–13. www.weeds.iastate.edu/mgmt/2003/ symposium.shtml (ABSTRACT). Hartzler RG, Battles BA, Nordby D (2004) Effect of common waterhemp (Amaranthus rudis) emergence date on growth and fecundity in soybean. Weed Science 52, 242–245. Heap I (2008) The International Survey of Herbicide Resistant Weeds. www.weedscience.com. Hobbs HA, Eastburn DM, D’Arcy CJ, Kindhart JD, Masiunas JB, Voegtlin DJ, Weinzierl RA, McCoppin NK (2000) Solanaceous weeds as possible sources of Cucumber mosaic virus in southern Illinois for aphid transmission to pepper. Plant Disease 84, 1221–1224. Holm L (1969) Weed problems in developing countries. Weed Science 17, 113–118. Holm GL, Plucknett DL, Pancho JV, Herburger JP (1977) The World’s Worst Weeds: Distribution and Biology. Hawaii Univ. Press, Honolulu, HI. Holm GL, Plucknett DL, Pancho JV, Herburger JP (1991) The World’s Worst Weeds: Distribution and Biology. Malabar, FL: Krieger Publishing. p. 609. Horak MJ, Loughin TM (2000) Growth analysis of four Amaranthus species. Weed Science 48, 347–355. Horvath DP, Schaffer R, Wisman E (2003) Identification of genes induced in emerging tillers of wild oat (Avena fatua) using Arabidopsis microarrays. Weed Science 51, 503–508. Horvath DP, Gulden R, Clay S (2006) Microarray analysis of velvetleaf (Abutilon theophrasti) impact on maize. Weed Science 54, 983–994. Hu FY, Tao DY, Sacks E, Fu BY, Xu P, Li J, Yang Y, McNally K, Khush GS, Paterson AH, Li ZK (2003) Convergent evolution of perenniality in rice and sorghum. Proceedings of the National Academy of Sciences of the United States of America 100, 4050–4054. Huo N, Gu YQ, Lazo GR, Vogel J, Coleman-Derr D, Luo MC, Thilmony R, Garvin D, Anderson O (2006) Construction and characterization of two BAC libraries from Brachypodium distachyon, a new model for grass genomics. Genome 49, 1099–1108. Huo N, Lazo GR, Vogel JP, You FM, Ma Y, Hayden DM, Coleman-Derr D, Hill TA, Dvorak J, Anderson OD, Luo MC, Gu YQ (2008) The nuclear genome of Brachypodium distachyon : analysis of BAC end sequences. Functional Integrated Genomics 8, 135–147. IRGSP (International Rice Genome Sequencing Project) (2005) The map-based sequence of the rice genome. Nature 436, 793–800. Jeschke MR, Tranel PJ, Rayburn AL (2003) DNA content analysis of smooth pigweed (Amaranthus hybridus) and tall waterhemp (A. tuberculatus): implications for hybrid detection. Weed Science 51, 1–3. Jia Y, Anderson JV, Horvath DP, Gu YQ, Lym RG, Chao WS (2006) Subtractive cDNA libraries identify differentiallyexpressed genes in dormant and growing buds of leafy spurge (Euphorbia esula). Plant Molecular Biology 61, 329–344. Jones HD (2005) Wheat transformation: current technology and applications to grain development and composition. Journal of Cereal Science 41, 137–147. Jordon-Thaden IE, Louda SM (2003) Chemistry of Cirsium and Carduus: a role in ecological risk assessment for biological control of weeds? Biochemical Systematics and Ecology 31, 1353–1396. Lampe KF, McCann MA (1985) AMA Handbook of Poisonous and Injurious Plants. American Medical Association. Chicago, Illinois, United States of America. p. 432. Leguizamon ES (1986) Seed survival and patterns of seedling emergence in Sorghum halepense (L.) Pers. Weed Research 26, 397–403. Leon RG, Owen MDK (2003) Regulation of weed seed dormancy through light and temperature interactions. Weed Science 51, 752–758. Leopold AC, Glenister R, Cohn MA (1988) Relationship between water content and afterripening in red rice. Plant Physiology 74, 659–662. Leroy P, Legeai F, Gicquello E, Itoh T, Feuillet C (2006) Structural and comparative genomics in bread wheat, the TriAnnotPipeline project. In: 4th INRA/AFFRCS French-Japanese Seminar on Food Safety, Novel Food, and Sustainable Environment to Promote Animal and Human Health, pp. 31–33. Louda SM, Kendall D, Connor J, Simberloff D (1997) Ecological effects of an insect introduced for the biological control of weeds. Science 277, 1088–1090.

MODEL WEEDS FOR GENOMICS RESEARCH

51

Madsen SB (1962) Germination of buried and dry stored seeds III, 1934–1960. Proceedings of the International Seed Testing Association 27, 920–928. Meinke DW, Cherry JM, Dean CD, Rounsley S, Korneef M (1998) Arabidopsis thaliana: a model plant for genome analysis. Science 282, 662–682. Meyers BC, Galbraith DW, Nelson T, Agrawal V (2004) Methods for transcriptional profiling in plants. Be fruitful and replicate. Plant Physiology 135, 637–652. Monaghan N (1979) The biology of Johnsongrass (Sorghum halepense). Weed Research 19, 261–267. Moore RJ (1975) The biology of Canadian weeds. 13. Cirsium arvense (L.) Scop. Canadian Journal of Plant Science 55, 1033–1048. Nordby DE, Hartzler RG (2004) Influence of corn on common waterhemp (Amaranthus rudis) growth and fecundity. Weed Science 52, 255–259. Ogg AG, Rogers BS, Schilling EE (1981) Characterization of black nightshade (Solanum nigrum) and related species in the United States. Weed Science 29, 27–32. Paterson AH, Schertz KF, Lin YR, Liu SC, Chang YL (1995) The weediness of wild plants: Molecular analysis of genes influencing dispersal and persistence of Johnsongrass, Sorghum halepense (L.) Pers. Proceedings of the National Academy of Sciences of the United States of America 92, 6127–6131. Pearcy RW, Ehleringer J (1984) Comparative ecophysiology of C3 and C4 plants. Plant, Cell and Environment 7, 1–13. Peterson DG, Schulze SR, Sciara EB, Lee SA, Bowers JE, Nagel ANJ, Tibbitts DC, Wessler SR, Paterson AH (2002) Integration of Cot analysis, DNA cloning, and high-throughput sequencing facilitates genome characterization and gene discovery. Genome Research 12, 795–807. Pimentel D, Zuniga R, Morrison D (2005) Update on the environmental and economic costs associated with alien-invasive species in the United States. Ecological Economics 52, 273–288. Pratt DB, Clark LG (2001) Amaranthus rudis and A. tuberculatus-one species or two? Journal of the Torrey Botanical Society 128, 282–296. Rensink WA, Buell CR (2004) Arabidopsis to rice. Applying knowledge from a weed to enhance our understanding of a crop species. Plant Physiology 135, 622–629. Rensink WA, Lee Y, Liu J, Iobst S, Ouyang S, Buell CR (2005) Comparative analyses of six solanaceous transcriptomes reveal a high degree of sequence conservation and species-specific transcripts. BMC Genomics 6, 124. Sauer JD (1955) Revision of the dioecious amaranths. Madroño 13, 5–46. Schulz-Schaeffer J, Gerhardt S (1987) Cytotaxonomic analysis of the Euphorbia spp. (“leafy spurge”) complex. Biologisches Zentralblatt 106, 429–438. Schulz-Schaeffer J, Gerhardt S (1989) Cytotaxonomic analysis of the Euphorbia spp. (“leafy spurge”) complex. II. Comparative study of the chromosome morphology. Biologisches Zentralblatt 108, 69–76. Seefeldt SS, Zemetra R, Young FL, Jones SS (1998) Production of herbicide-resistant jointed goatgrass (Aegilops cylindrica) × wheat (Triticum aestivum) hybrids in the field by natural hybridization. Weed Science 46, 632–634. Seibert AC, Pearce RB (1993) Growth analysis of weed and crop species with reference to seed weight. Weed Science 41, 52–56. Shinozaki K, Yamaguchi-Shinozaki K (2000) Molecular responses to dehydration and low temperature: Differences and cross-talk between two stress signaling pathways. Current Opinion in Plant Biology 3, 217–223. Smith CW, Frederiksen RA (2000) Sorghum: Origin, History, Technology, and Production. New York, NY: John Wiley and Sons, 824. Smith DB, Flavell RB (1975) Characterisation of the wheat genome by renaturation kinetics. Chromosoma 50, 223–242. Sorghum Genomics Planning Workshop Participants (2005) Toward sequencing the sorghum genome. A U.S. National Science Foundation-sponsored workshop report. Plant Physiology 138, 1898–1902. Stockinger EJ, Gilmour SJ, Thomashow MF (1997) Arabidopsis thaliana CBF1 encodes an AP2 domain-containing transcription activator that binds to the C-repeat/DRE, a cis-acting DNA regulatory element that stimulates transcription in response to low temperature and water deficit. Proceedings of the National Academy of Sciences of the United States of America 94, 1035–1040. Trucco F, Jeschke MR, Rayburn AL, Tranel PJ (2005) Promiscuity in weedy amaranths: high frequency of female tall waterhemp (Amaranthus tuberculatus) × smooth pigweed (A. hybridus) hybridization under field conditions. Weed Science 53, 46–54. USDA Forest Service, USDI Bureau of Land Management, Oregon Department of Agriculture, US Army Corps of Engineers, OSU College of Forestry, Institute for Applied Ecology, Starker Forests Inc., The Nature Conservancy, Native Plant Society of Oregon (2003) Invasive Plant Alert False-brome (Brachypodium sylvaticum). Prepared by Kaye T. http://www.appliedeco.org/invasive-species-resources/FBWG/brsybrochure.pdf. Vij S, Gupta V, Kumar D, Vydianathan R, Raghuvanshi S, Khurana P, Khurana JP, Tyagi AK (2006) Decoding the rice genome. BioEssays 28, 421–432.

52

WEEDY AND INVASIVE PLANT GENOMICS

Vogel JP, Garvin DF, Leong OM, Hayden DM (2006a) Agrobacterium-mediated transformation and inbred line development in the model grass Brachypodium distachyon. Plant Cell Tissue Organ Culture 84, 199–211. Vogel JP, Gu YQ, Twigg P, Lazo GR, Laudencia-Chingcuanco D, Hayden DM, Donze TJ, Vivian LA, Stamova B, ColemanDerr D (2006b) EST sequencing and phylogenetic analysis of the model grass Brachypodium distachyon. Theoretical and Applied Genetics 113, 186–195. Westbrooks RL (1998) Invasive Plants, Changing the Landscape of America: Fact Book. Federal Interagency Committee for the Management of Noxious and Exotic Weeds (FICMNEW). Washington, D.C. 4–5, pp. 16–17. Xu B, Dai W, Chao WS (2008) An efficient method for in vitro regeneration of leafy spurge (Euphorbia esula L.). In Vitro Cellular and Developmental Biology-Plant. 44(6), 548–556. Yuan JS, Tranel PJ, Stewart CN Jr. (2007) Non-target-site herbicide resistance: a family business. Trends in Plant Science 12, 6–13. Zemetra RS, Hansen J, Mallory-Smith CA (1998) Potential for gene transfer between wheat (Triticum aestivum) and jointed goatgrass (Aegilops cylindrica). Weed Science 46, 313–317.

5

21st-Century Weed Science: A Call For Amaranthus Genomics Patrick J. Tranel and Federico Trucco

Introduction

Over the last decade, weed science research has shifted from a focus on herbicide discovery and physiology to the study of weed biology and ecology. An overriding research question is, what makes a species a successful agricultural weed? This central question has been tackled from a variety of angles, identifying and characterizing attributes contributing to weed success. Molecular biology and genomic research tools are being used increasingly in these efforts. Most attributes contributing to weed success are specific variants of common plant characters—“rapid” growth rates, “high” fecundity, “discontinuous” germination, and herbicide “tolerance,” just to name a few items on the list. The mechanisms that control these characters are presumed to be conserved throughout plants; therefore, their basic study can be accomplished using model non-weedy organisms. However, identifying and understanding allelic variants by which a character may achieve weedy values cannot escape the selection of a model weed species. Aside from being amicable to biotechnologies, the selection of a model weed species (or species complex) often has been discussed in terms of the wealth of weedy attributes exhibited by a given taxon (Basu et al. 2004; Chao et al. 2005). From a weed management perspective, however, the most significant aspect of weediness is the ability of a species to modify a given attribute over time and in response to selection. The fact that herbicide resistance is of greater concern than herbicide tolerance is an example of this assertion. (Herbicide resistance, as it is used in this chapter, is an evolutionary phenomenon following herbicide selection, and is differentiated from herbicide tolerance, which is innate to the wild type of a species.) The evolution of herbicide resistance often forces dramatic changes in weed management practices, whereas tolerance is responsible at best for manageable and subtle shifts (see Chapter 11). With this in mind, selection of a model weed species should place high priority on its suitability for the study of adaptive evolution. The complex of Amaranthus weeds, often referred to as pigweeds, provides an excellent model for the study of important weed attributes as well as a compelling case of rapid adaptation.

The Amaranthus Genus

The genus Amaranthus (Caryophyllales: Amaranthaceae) is comprised of seventy species distributed worldwide (Mosyakin and Robertson 2003). Although this genus includes crop and ornamental species, it is best known for its many weedy species. The weedy Amaranthus species are abundant, widely distributed, and among the most damaging weeds in the world. The current importance of Amaranthus weeds is evidenced by its leading status in weed science literature (Figure 5.1). Most weedy Amaranthus species typify agronomic weeds in that they are summer annual species capable of competing strongly with crop plants, expressing high plasticity in response 53

Digitaria Helianthus Senna Panicum Euphorbia Oryza Raphanus Alopecurus Conyza Xanthium Bromus Aegilops

Genus

Ambrosia Striga Cirsium Centaurea Ipomoea Sorghum Abutilon Chenopodium Setaria Solanum Echinochloa Cyperus Orobanche Avena Lolium Amaranthus 0

20

40

60

80

100

Occurrences in article titles (Weed Research,Weed Science, and Weed Technology, 2000-2007)

Figure 5.1. Predominance of several weed-containing genera in weed science literature. Genera were used as keyword searches of article titles published in Weed Research, Weed Science, or Weed Technology from 2000 through 2007. Sixty genera expected to have high occurrences were chosen from the Weed Science Society of America’s Composite List of Weeds (www.wssa.net/Weeds/ID/WeedNames/namesearch.php); thirty-two of those chosen (not shown) had less than ten occurrences. Article titles containing only crop species with the genus name were excluded from the tallies.

54

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

55

to environmental changes, and ensuring their future survival by producing large numbers of seeds. For example, the pigweeds produce more seeds on a per plant basis than most other weeds (Zimdahl 1999). A single pigweed plant may produce well over 100,000 seeds and, even when growing in competition with a crop, pigweed species average several thousand seeds per plant (Knezevic and Horak 1998; Massinga et al. 2001; Steckel and Sprague 2004; Steckel et al. 2003; Uscanga-Mortera et al. 2007). The seeds are very small (about 1 mm diameter), making them easily transported by wind and water, and seeds also may be transported very long distances by the activities of humans and birds (DeVlaming and Proctor 1968; Weaver and McWilliams 1980). The pigweeds use the C4 photosynthetic pathway, making them efficient at net photosynthesis and providing them with a competitive advantage, particularly in environments characterized by high light intensity, high temperature, and limited water availability. Many of the weedy Amaranthus species have rapid growth rates, and their competitiveness with crops has been well documented (Klingaman and Oliver 1994; Knezevic et al. 1994; Toler et al. 1996; Cowan et al. 1998; Rowland et al. 1999; Massinga et al. 2001; Hager et al. 2002; Steckel and Sprague 2004). To be widely successful, weeds must be adaptable; here again, the pigweeds excel. In addition to having a worldwide distribution, pigweeds invade a variety of niches, including most cropping systems, pastures and rangelands, waste areas, fence-rows, and right-of-ways. Exemplifying their adaptability, and described in more detail later, pigweeds are notorious for their propensity to evolve resistance to herbicides. Pigweeds use two contrasting strategies of sexual reproduction: most of the species are monoecious (bisexual individuals with unisexual flowers), but some are dioecious (unisexual individuals, that is, separate male and female plants). The monoecious species are primarily self-pollinated owing to the proximal arrangement of male and female flowers (Murray 1940; Weaver and McWilliams 1980), while the dioecious species are obligately allogamous. Dioecism encourages genetic diversity by forcing outcrossing, whereas monoecism allows colonization of new areas by a single plant. Both monoecious and dioecious pigweeds are highly successful invaders. The presence of both strategies within the genus provides an ideal opportunity to investigate the advantages and disadvantages of each reproductive tactic in terms of contributions toward adaptability and overall weediness. The various Amaranthus species can be difficult to distinguish (Horak et al. 1994; Pratt et al. 1999), and there are documented cases of misidentification by scientists (Sauer 1953; Ahrens et al. 1981). Much of the difficulty in taxonomic discrimination of species within the group can be attributed to attempts at recognizing taxa based on pigmentation or growth forms, which are extremely variable within amaranths (Sauer 1967). However, examination of floral parts can result in constant characters from which correlated discontinuities can be used to define well-established taxa. On the basis of these characters three subgenera are presently recognized: Acnida, encompassing dioecious weeds; Amaranthus, including monoecious weeds and crop species; and Albersia, where many of the poorly characterized amaranth taxa are grouped. Some of the most notable weedy Amaranthus species are listed in Table 5.1.

Monoecious Species, The “Old” Weeds

Historically, the most notorious Amaranthus weeds in the United States belonged to the group of monoecious species. Among these, three different clusters can be recognized: (i) the species of subgenus Albersia, A. blitoides and A. albus; (ii) A. spinosus and its allotetraploid A. dubius;

56

WEEDY AND INVASIVE PLANT GENOMICS Table 5.1. Major Amaranthus weeds. Species

Common name

A. albus A. arenicola A. australis A. blitoides A. hybridus A. lividus A. palmeri A. powellii A. quitensis A. retroflexus A. spinosus A. tuberculatus var. rudis A. tuberculatus var. tuberculatus A. viridus

Tumble pigweed Sandhills amaranth Giant amaranth Prostrate pigweed Smooth pigweed Livid amaranth Palmer amaranth Powell amaranth Yuyo colorado Redroot pigweed Spiny amaranth Common waterhemp Tall waterhemp Slender amaranth

and (iii) the interbreeding complex of A. retroflexus-A. powellii-A. hybridus. Two of these species, A. hybridus and A. spinosus, are ranked among the eighteen most serious weeds in the world (Holm et al. 1991), and A. retroflexus is among the most widely distributed (Holm et al. 1997). Unless referenced otherwise, much of the observations provided on the distributions, habitats, and evolutionary histories of monoecious amaranths are derived from Jonathan Sauer ’s (1967) revision of the subject. Amaranthus blitoides and A. albus were classified in old North American floras as A. graecizans, a species thought to have Old World origins (Mosyakin and Robertson 2003). However, both species are presumed to be native to the central United States and are now widely distributed in temperate North America and other warm to temperate regions of the world. Amaranthus spinosus and A. dubius have an earlier distribution based in the tropical lowlands of America, with A. spinosus migrating to the warmest parts of the world as early as 300 years ago. Amaranthus dubius’ spread has been slower, with its presence recorded in only a few areas of the Old World and never outside the tropics. Amaranthus dubius is the only known polyploid among amaranths, presumably an amphidiploid generated by a crossing of A. spinosus with either A. quitensis or A. hybridus. Yet, this species is not considered a significant weed. Amaranthus retroflexus’ earlier range expanded from the central-eastern United States to adjacent Canada and Mexico; whereas A. hybridus expanded from milder and moister eastern North America to the northernmost part of South America. Both species were initially riverbank pioneers and are today ubiquitous agricultural weeds. Introduction in Europe was reported as early as 300 years ago, with A. retroflexus spreading quickly throughout temperate regions of the continent and rest of the world. Amaranthus hybridus’ spread has been slower, with presence in western North America, eastern Asia, Australia, and South Africa only recorded as of the early to mid 1900s. These species have been regarded as weeds of agriculture since their earliest characterizations (Linnaeus 1753). Amaranthus powellii’s initial distribution included canyons, desert washes, and other open habitats west of the Cordilleran system of America, with wide gaps in wetter regions of Central America. The earliest European record of this species is found in German herbarium specimens

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

57

from the late 1800s, and later introductions can be interpreted from samples of southern India and South Africa. Expansion of A. powellii to eastern North America occurred only during the last century. Partially fertile hybrid swarms between these species can be found in the United States, in areas where their distributions overlap, and in Europe, where all three species are recent immigrants. The amaranth grain crop is derived of ancient domestications of these species or their hybrids.

Domesticated Species

One reason for interest in Amaranthus is its unrealized potential as crops. Amaranths can be cultivated for grain (Figure 5.2), forage, horticultural, and ornamental production. The earliest recorded history of amaranths as crops dates to the Aztecs in central Mexico; archeological evidence, however, suggests Amaranthus species were used as food crops for thousands of years (Brenner et al. 2000). Although currently planted on only a few hundred hectares in the United States, some Amaranthus cultivars are registered as grain crops (Baltensperger et al. 1992; Schulz-Schaeffer et al. 1991; Sooby et al. 1999). China is the world’s largest producer of amaranth, growing an estimated 150,000 ha, mainly as forage for swine (Brenner et al. 2000). There has been increased worldwide interest in cultivating Amaranthus species in recent years, because of its nutritional profile and its tolerance to high temperature, drought, salinity, and aluminum toxicity in acid soils (Brenner et al. 2000). Amaranthus hypochondriacus, one of the three grain amaranths, is cultivated as an alternative crop in North America and Asia. Although initially thought to have Asian origins, it is now believed that this distribution is secondary and that the species derives from an A. powellii

Figure 5.2. Field of grain amaranth grown in Argentina. Production of grain amaranth is gaining interest in different areas of the world. In Argentina, the Instituto de Agrobiotecnologia Rosario (INDEAR) is leading efforts on crop genetics, production, and industrialization, aiming at expanding crop acreage in dry and semi-dry regions of this country. (Photo courtesy of J. Albertengo, Asociación Argentina de Productores en Siembra Directa [AAPRESID].)

58

WEEDY AND INVASIVE PLANT GENOMICS

domestication in North America. Hybridization has had a significant role in the evolution of A. hypochondriacus, with several hybrid races cultivated by American Indians. Sauer (1955) identified stable hybrid cultivars derived from crosses, presumably between A. hypochondriacus and local admixtures of A. cruentus—an A. hybridus domesticated form originating in southern Mexico or Guatemala. Amaranthus caudatus, the grain amaranth of South America, is thought to originate from a domestication of A. quitensis in the Andean region. Amaranthus quitensis is a weedy member of the A. hybridus aggregate, with original distribution as a riverbank pioneer of South America, in mountains in the northwest, and at lower elevations in the temperate south. Cultivation of A. quitensis forms with incipient domestication is observed from Ecuador to northern Argentina, mainly for the production of pigments needed for coloring of chicha and other maize dishes. Although some cultivated forms of A. caudatus-A. quitensis are suspected to be the result of interbreeding with A. cruentus, the South American amaranths are not considered to readily hybridize with the North American members of this cluster. The close evolutionary relationship between important weedy and domesticated amaranths provides excellent raw material for the genomic study of evolutionary processes. With detailed genotyping we can examine at the DNA level genomic shifts associated with the acquisition of crop traits, at a detriment to weediness in many instances, and vice versa. We could dissect change resulting from direct selection apart from collateral shifts due to linkage. At the same time, we could address long-standing questions regarding domestication of grain amaranths, fundamentally, whether the three species resulted from independent events or share a common origin, consolidating our understanding of pre-Colombian agriculture and associated cultures.

Dioecious Species, The “New” Weeds

Dioecious amaranths have only recently acquired notoriety as problematic weeds. In the United States, A. tuberculatus and A. palmeri are generally ranked among the worst weeds in states of their respective regions of the Midwest and Southeast, especially with the recent identification of glyphosate-resistant populations of these species (Culpepper et al. 2006; Legleiter and Bradley 2008). Of the nine dioecious species listed in the Flora (Mosyakin and Robertson 2003), four are considered significant agronomic weeds. As with the monoecious species, Sauer (1955, 1957, 1972) has contributed extensively to the reconstruction of evolutionary histories of the dioecious amaranths, and unless otherwise referenced, the discussion that follows is based on his work. Distributions Of The Dioecious Species. Jonathan Sauer identified three groups of species based on their distributions: coastal, southwestern, and interior species. Coastal species have narrow distributions along the eastern and southern coasts of the United States. They are found in the wet sand or mud of coastal marshes, swamps, estuaries, lagoons and bayous. Of the four species in this group, A. cannabinus, the eastern waterhemp, has the best herbarium record, indicating a static distribution since the earliest collections (beginning in the 1850s), ranging from New England to the Carolinas. Amaranthus australis, the southern waterhemp, was first identified in Jamaica in the 1840s, and shows a grossly discontinuous distribution along Gulf States. Some herbarium samples have been identified as hybrids between these two species, or between A. australis and A. tuberculatus, an interior dioecious weed. Yet, neither of these

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

59

two coastal species, nor the less characterized A. floridanus and A. greggii, are important agronomic weeds. Of the three southwestern species, two are problematic weeds, and unlike the non-weedy A. aconthochiton, both show dynamic distributions over the past century. These species were not found in wet banks of permanent streams, but they were present originally in sandy floors of desert washes, canyon bottoms, and intermittent stream beds. Amaranthus palmeri is an old weed among dioecious pigweeds, with the majority of collections examined by Sauer coming from artificial habitats, which was not true then of any other dioecious amaranth. The earliest recorded distribution of this species included parts of California, New Mexico, Arizona, and Texas. By the beginning of the last century, northeastward expansion of the species was recorded, overlapping with the distribution of the interior dioecious taxa. Southeastward expansion into Mexico occurred by the 1930s. Amaranthus watsoni, the other southwestern weed, has shown similar but milder weed tendencies. This species was only detected in the peninsula of Baja California before the 1850s, and movement into the mainland is recorded as early as the mid 1870s. Putative hybrids between A. palmeri and A. watsoni are found among herbarium samples. Interestingly, collections from agricultural fields are heavy on hybrid intermediates. Three interior dioecious species were recognized by Sauer with strongholds in sandy and muddy stream banks, lakeshores, and pond margins, or open sites covered or reworked by water during part of every year. All three species were considered incipient agricultural weeds, with A. arenicola and A. tuberculatus having static ranges for the period covered by the samples included in Sauer ’s revision. Extending through the Missouri, Mississippi, and Ohio river systems, A. tuberculatus overlapped with A. arenicola, of western distribution, in South Dakota and Nebraska. Sauer ’s A. tamariscinus (synonymous with A. rudis) represents a different case. This species was first described in Oklahoma in the 1830s and since has shown continuous northward and eastward accretion, moving into A. tuberculatus’ territory. By the 1850s, the species was reported to have been collected outside of Oklahoma in Missouri, Kansas, and Texas. Spread to Iowa, Nebraska, and Illinois occurred by the 1870s; to the Dakotas and Arkansas by the 1900s; to Wisconsin, Indiana, and Tennessee by the 1920s; to Minnesota by the 1950s; and is presently naturalized in thirty-eight North American states and some Canadian regions (Mosyakin and Robertson 2003). Where both A. tuberculatus and A. tamariscinus co-existed, the record of the former was on average forty years prior to that of the latter. Many of the samples collected in these areas were classified as putative A. tuberculatus × A. tamariscinus hybrids, with a higher ratio of hybrids to non-hybrids in artificial habitats compared to natural settings. In Sauer ’s assessment of dioecious amaranths, A. tuberculatus × A. tamariscinus hybrids are the most abundant hybrid combination. The author also notes that actual hybridization among these species may be underestimated due to the nature of morphological determinations based on character intermediacy, which is often diluted after a few generations of backcrossing with the predominant genotype. More recent work using molecular and morphological markers suggested both species to be one and the same (Pratt and Clark 2001), and a single species, A. tuberculatus, is presently recognized (Mosyakin and Robertson 2003). Costea and Tardif (2003), however, encouraged recognition of the two entities at the variety level: A. tuberculatus var. rudis has more weedy tendencies than A. tuberculatus var. tuberculatus. Discrimination of these two variants at the genome level would provide insights into underlying bases for their different weed tendencies. The challenge in such an approach, however, lies in accurate identification of the “pure” variants, and would require collection of plant material outside of their increasingly overlapping ranges. Throughout the remainder of this

60

WEEDY AND INVASIVE PLANT GENOMICS

chapter, A. tuberculatus will be used to refer to all previously recognized species in this taxon (e.g., A. tamariscinus and A. rudis) and collectively to the two presently recognized varieties. Invasion of A. tuberculatus. In terms of modern weed species shifts, there are few examples more striking than the invasion of A. tuberculatus into agronomic fields throughout the midwestern United States (Figure 5.3). In Illinois, for example, the species went from relatively unknown status to being the primary weed in about two decades (Steckel and Sprague 2004; Steckel 2007). The species continues to expand its distribution, and it is now becoming problematic as far north as Canada (Costea et al. 2005). What caused A. tuberculatus’ rapid rise to prominence as a widespread weed? Costea et al. (2005) discussed changes in weed management that may have favored A. tuberculatus. Increased adoption of reduced tillage and more reliance on postemergence rather than preemergence herbicides created a niche more favorable to the small seeds and extended germination characteristics of this species. Additionally, widespread use of herbicides that inhibit acetolactate synthase (ALS),coupled with A. tuberculatus’ ability to rapidly evolve resistance to these herbicides, could have provided the species with a foothold, setting the stage for subsequent invasion. Alternatively, maybe it was not so much a change in management that fostered A. tuberculatus’ invasion but, rather, that A. tuberculatus changed, somehow endowing itself with more weed characteristics. An interesting hypothesis is that hybridization between A. tuberculatus and previously differentiated species (e.g., A. rudis) and/or some other Amaranthus species historically more problematic as weeds (e.g., A. hybridus) could have led to A. tuberculatus’

North Dakota Minnesota

98

South Dakota

Wisconsin

95

96 Nebraska

94 Kansas

92

Michigan

95

94

Iowa Illinois

91

93

98

Indiana

Ohio

Mid 80s Kentucky Missouri

Figure 5.3. Invasion of A. tuberculatus in the Midwestern United States. This weed, considered a significant problem in the 1980s only in a relatively small area centered in Missouri, became problematic over a much larger area during the 1990s. Numbers indicate the year (19xx) in which the species was first considered by local extension weed scientists to be a significant problem. (Data kindly provided by R. Hartzler, Iowa State University.)

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

61

acquisition of the genetic material that enabled its invasion. This hypothesis, which could be investigated with a genomics approach, has profound implications for weed management. Was A. tuberculatus destined to become a problematic weed, or did we facilitate the problem through our management practices (and therefore could have prevented the problem)? The following section embarks on a more detailed discussion on this subject.

Hybridization And Adaptive Evolution

A species’ ability to adapt to changing environmental conditions is found in the genetic diversity of its populations. Success in weed populations facing changing agricultural ecosystems often correlates with an abundance of genetic polymorphisms within those populations (Jasieniuk and Maxwell 2001). Several studies point to a broad genetic base in weedy Amaranthus species (Chan and Sun 1997; Sun et al. 1999; Pratt and Clark 2001), although this is mostly inferred from phenotypic diversity and high outcrossing rates, from 2% to 20% for monoecious species (Agong and Ayiecho 1991) and obligate outcrossing in dioecious populations. The propensity of pigweeds to evolve herbicide resistance, as discussed later, also alludes to their broad genetic diversity. Spontaneous mutations, errors at DNA replication, meiotic recombination, and transposable element activity are some of the mechanistic avenues by which advantageous alleles may originate. In addition, new genetic variants may arise via interspecific hybridization.

Hybridization Between Amaranthus Species

An important and well-recognized component in the evolutionary history of Amaranthus species is interspecific hybridization. Cytogenetic studies point to the paleo-allotetraploid nature of several amaranths (Pal et al. 1982; Greizerstein and Poggio 1992, 1994, 1995). Genetic studies on the mechanism of sex determination in these species showed that progeny could be obtained from several combinations of interspecific crosses (Murray 1940). In particular, hybrids were readily obtained from crosses within subgenus Acnida, within subgenus Amaranthus, and between the two subgenera. As discussed in a prior section, an interesting observation in the historical survey of weedy amaranths is the preponderance of putative interspecific hybrids in managed ecosystems (i.e. agricultural fields). Hybridization has been proposed as a critical stimulus for invasiveness (Ellstrand and Schierenbeck 2000) and is perhaps aiding in the evolution of adaptations critical for the success of amaranths as weeds. Interspecies gene flow may lead to chromosomal evolution (e.g. chromosomal rearrangements). Lack of homology between genetic complements of interspecific hybrids may cause dysfunctional meiosis and thus lead to chromosomal aberrations. Cytogenetic studies of amaranths and their hybrids allude to this possibility (Greizerstein and Poggio 1992; Pandey 1999). In addition, hybridization may lead to gene introgression. As postulated by Anderson (1949), this refers to the stable transfer of genes from one species to another; it implies the formation of hybrids and the nearly complete reconstitution of the recipient species’ genome through recurrent backcrossing. Although species interbreeding is most often maladaptive, it might represent an important route for the evolution of genotypes favored under the intense selection pressure found in

62

WEEDY AND INVASIVE PLANT GENOMICS

agricultural habitats. A clear example of this possibility is herbicide resistance evolution. A resistant individual resulting from a hybridization event may be lacking in health, vigor, and fertility, but may represent the only viable genotype upon herbicide treatment. Several recent studies have aimed at investigating the role of hybridization in the evolution of pigweeds, specifically addressing the likelihood of herbicide resistance transfer among species. Wetzel et al. (1999a) determined that herbicide resistance genes could be transferred among dioecious species, from A. tuberculatus to A. palmeri. Similarly, transfer of herbicide resistance was shown to occur in the reciprocal direction (from A. palmeri to A. tuberculatus [Franssen et al. 2001]). Several hybrids were obtained between the two species using a dominant gene conferring resistance to ALS-inhibiting herbicides as a marker and conducting controlled crosses in a growth chamber. All of these hybrid plants appeared to be fertile and, furthermore, at least one of these hybrids was able to backcross with the herbicide-sensitive parent (A. palmeri) and produce viable offspring. As expected, a proportion of these backcross progeny carried the dominant herbicide-resistance gene, suggesting that introgressive hybridization between these two species could occur. Yet, the lack of genome-wide data may render the conclusion from this and similar research premature. Using flow cytometry to examine genomic makeup of progeny from crosses between A. tuberculatus and A. palmeri, Trucco et al. (2007) provided a non-introgressive explanation to the data obtained previously. The alternative offered was based on the production of non-sexual progeny in these species and triploidy as a route for heterozygosity. Molecular (Wassom and Tranel 2005), morphological (Franssen et al. 2001), and DNA content data (Rayburn et al. 2005), as well as data from hybridization studies (Trucco et al. 2007), suggest that A. palmeri and A. tuberculatus are not sister taxa, notwithstanding their shared dioecism.

Hybridization Between A. tuberculatus And A. hybridus

Earlier, we mentioned that A. tuberculatus is becoming increasingly problematic as a weed. Concurrently, we presented the hypothesis suggesting that weed adaptations are being transferred to A. tuberculatus by long-standing weedy amaranths and proposed A. hybridus as a candidate donor, a coexisting species in much of the Midwest. Some recent studies have investigated this possibility. Transfer of herbicide resistance from A. hybridus to A. tuberculatus was documented by Patrick Tranel and colleagues (2002). In this case, backcrossing to A. tuberculatus was performed for two generations, and the resistance-conferring allele from A. hybridus was transmitted to, and functional in, some of the BC2 progeny. Prior knowledge in this area indicated that hybrids (F1s) between A. tuberculatus and A. hybridus could be produced (Murray 1940). However, subsequent introgression was thought to be compromised by severe sterility in the F1 (Sauer 1957), and the only viable BC1 progeny were considered to be those derived from unreduced gametes from the hybrid parent (Murray 1940), resulting in triploidy. BC1s would have a full complement of the recurrent species’ genome and only a haploid complement of the non-recurrent parent and, therefore, exhibit sterility from abnormal chromosome pairing. Although there are several mechanisms by which triploids could be stabilized (e.g., agamospermy, genome duplication, etc.), stabilization under any of these mechanisms would be rare and likely to result in a speciation event (Rieseberg 1997)—the hybrid population is now reproductively isolated from the parental species.

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

63

These observations suggested little if any chance for an “Andersonian” outcome to hybridization between A. tuberculatus and A. hybridus. No records are found in herbaria that suggest polyploidy has occurred to a noticeable degree in these species. Although the work performed by Tranel and colleagues (2002) showed that gene movement could occur in a homoploid background—in contrast to previous observations—it did not directly address the fertility and genome structure of introgressants. Were heterozygous BC2s more fertile than heterozygous BC1s, or F1s? Was the genomic constitution of these introgressants recombinant (on average 12.5% A. hybridus and 87.5% A. tuberculatus), or a reconstitution of the F1? What about introgression in the reciprocal direction? To what extent could this experiment be replicated in nature? To address some of these questions field hybridization experiments were conducted and hybrids produced at high frequencies under selected conditions. In the case in which the monoecious parent was used maternally, the maximum hybridization frequency obtained accounted for close to 50% of the expected intraspecific outcrossing potential of the species (Trucco et al. 2005a). In the reciprocal case, more than 200,000 hybrids could be obtained from a single A. tuberculatus plant (Trucco et al. 2005b). These data indicate that there is little, if any, gametic incompatibility between the studied species and that F1 production is unlikely to constitute a significant bottleneck in gene exchange. Although they better approximate an agricultural setting than a greenhouse or growth chamber, field hybridization plots do not accurately replicate ecologically representative conditions. What proportion of Amaranthus communities is expected to be composed of hybrids in fields where the two species coexist? This question will be hard to tackle without the use of modern genomic tools, as even first generation hybrids are not easily distinguishable using morphological parameters from the unisexual parent (Trucco et al. 2006). Significant sterility can be observed in A. tuberculatus × A. hybridus F1s, yet as many as 800 seeds could be recovered from a single plant (Trucco et al. 2006). Cytogenetic profiling of these individuals—progeny from backcrosses to the “pure” species—revealed that most (98%) were homoploid (2n=32), and triploidy (2%) was not necessarily explained by the production of unreduced hybrid gametes (Trucco et al. 2005c). In the same study, fertility restitution was not a strict function of the reconstitution of the parental species’ genomes; in fact, hybrid sterility could be explained by as few as five independently assorting loci. In which case, advantageous alleles unlinked to these loci may be introgressed quickly. The introgression of linked alleles (genes linked to postzygotic reproductive barriers) may depend on the significance of the adaptation being transferred (i.e., selection coefficient) and the likelihood of de novo evolution. These observations support a probabilistically feasible role for introgressive hybridization between A. tuberculatus and A. hybridus. If this is the case, one could be puzzled by the fact that ALS-inhibitor resistance, a widespread herbicide resistance in pigweeds, shows independent evolution in sympatric A. hybridus and A. tuberculatus populations. Evaluating introgression at the ALS locus, (interspecific) heterozygosity at ALS showed a fecundity penalty (Trucco et al. 2005c). This penalty together with the relatively high initial frequencies of ALS-inhibitorresistance-conferring mutations in unselected populations (Preston and Powles 2002) may be held responsible for this otherwise unpredicted result. The use of a whole-genome molecular-mapping approach would have allowed for a more detailed examination of the genetics of postzygotic reproductive isolation between these species. Specifically, such an approach could have revealed the number and relative location of loci involved with hybrid sterility and the magnitude of their effects. Yet, this alternative is only feasible given the current knowledge on BC1 genome structure and would have been

64

WEEDY AND INVASIVE PLANT GENOMICS

meaningless if hybrids produced only non-recombinant gametes. Identifying whether newly evolved herbicide resistance traits (such as nuclear triazine resistance, protoporphyrinogen oxidase inhibitor resistance, or glyphosate resistance) map to a sterility locus would be useful in predicting the likelihood of resistance evolution via introgressive hybridization among pigweeds. The real-life significance of introgressive hybridization between A. tuberculatus and A. hybridus may be realized only when evidence of such an evolutionary event is found in nature. The research presented in this section indicates that searching for such evidence is appropriate, and modern genomics tools should be enlisted in this endeavor.

Herbicide Resistance

One or more herbicide options are commercially available to control essentially any given weed species in any major crop (although, a somewhat different perspective is offered in Chapter 1). Through the process of mutation and selection, however, weeds evolve resistance to herbicides when they are used repeatedly. Thus, the phenomenon of herbicide resistance provides an ongoing challenge to effective weed management. A worldwide catalog of herbicide-resistant weeds (Heap 2008) includes more than 300 resistant biotypes. A biotype in this case is a particular weed species with resistance to a particular herbicide or group of herbicides with the same site of action; thus, for example, two populations of A. tuberculatus with different mechanisms of resistance to ALS-inhibiting herbicides would count as only one “biotype.” Although there are more than 180 different weed species in this resistance catalog, Amaranthus species comprise more than 5% of the total number of resistant biotypes. Thus, Amaranthus can be considered a “leader” among weeds in terms of resistance evolution. In addition, and as described in the following sections, several “firsts” in terms of our understanding of herbicide-resistance mechanisms occurred with pigweeds. Amaranthus species also provide current opportunities to explore herbicideresistance mechanisms that have not yet been fully elucidated, and they may serve as models to examine the evolution of resistance across broad geographic regions.

Resistance To Photosystem II Inhibitors

Resistance to photosystem (PS) II inhibitors (e.g. triazine herbicides) in Senecio vulgaris (common groundsel) is often credited as the first major case of evolved herbicide resistance in a weed (Ryan 1970). Since then, more than sixty weed species have been documented with resistance to PS II inhibitors, and it is the second most common type of herbicide resistance (Heap 2008). Amaranthus species joined the growing list of weeds resistant to PS II inhibitors in the late 1960s and early 1970s, and now Amaranthus contributes a total of nine species to the list (Table 5.2). Early research with triazine-resistant A. hybridus contributed much to our understanding of the fitness penalty associated with this resistance (Ahrens and Stoller 1983; Ort et al. 1983) and to the elucidation of the mechanism of resistance (Steinback et al. 1981; Hirschberg and McIntosh 1983). In fact, the identification of a glycine for serine codon change in the psbA gene of triazine-resistant A. hybridus was the first ever DNA-level characterization of evolved herbicide resistance (Hirschberg and McIntosh 1983). This same mutation in the D1 protein (the product of the psbA gene and the molecular target site of PS II-inhibiting herbicides) has

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

65

Table 5.2. Instances of herbicide resistance that have evolved in Amaranthus populations around the world (data from Heap 2008). Herbicide

Species

Year1

Country(ies)2

Photosystem II inhibitors

A. A. A. A.

hybridus powellii lividus retroflexus

1972 1977 1978 1980

A. blitoides A. albus A. cruentus A. palmeri A. tuberculatus A. blitoides A. palmeri A. retroflexus A. hybridus A. lividus A. tuberculatus A. powellii A. quitensis A. palmeri A. tuberculatus A. tuberculatus

1983 1987 1989 1993 1994 1991 1991 1991 1992 1993 1993 1996 1996 2005 2005 2001

USA, France, Italy, Switzerland, Spain, Israel, South Africa Canada, France, Switzerland, Czech Republic, USA Switzerland, France Canada, France, Germany, USA, Switzerland, Bulgaria, Czech Republic, Spain, China, Poland, Italy, Serbia, Greece Israel, Spain Spain Spain USA USA, Canada Israel USA Israel, USA, Canada, Serbia, Italy USA USA USA, Canada USA, Canada Argentina, Bolivia USA USA USA

A. lividus A. palmeri

1990 1989

Malaysia USA

Acetolactate synthase inhibitors

Glyphosate Protoporphyrinogen oxidase inhibitors Bipyridyliums Dinitroanilines 1

Indicates year when the resistant population was identified/documented. Multiple countries are listed in chronological order of documented resistance.

2

now been identified in several other weed species, and it is by far the most frequent mechanism of resistance to PS II-inhibiting herbicides (Gronwald 1994; Tian and Darmency 2006). Other mutations in the D1 protein that confer resistance to triazines and/or other PS II inhibitors have been identified in a few weed populations (Masabni and Zandstra 1999; Mengistu et al. 2000; Park and Mallory-Smith 2006), one of which is an Amaranthus species (Dumont and Tardiff 2002). Resistance to PS II inhibitors not caused by an altered D1 protein also has been identified in weeds in a few instances (Gronwald et al. 1989; Burnet et al. 1993; Cummins et al. 1999), again including an Amaranthus species (Patzoldt et al. 2003). Non-target-site resistance to PS II inhibitors is generally thought to be caused by enhanced rates of herbicide metabolism. There are two important distinctions between target-site and metabolism-mediated PS II inhibitor resistance that influence the evolution of these resistance mechanisms. First, D1 mutations that reduce herbicide-binding affinity also generally reduce the rate of electron transport in PS II, leading to reduced growth and relative fitness (Holt and Thill 1994). Second, because the psbA gene is a chloroplast gene, target-site resistance to PS II inhibitors is largely, if not exclusively, maternally inherited (Gronwald 1994). Why, then, is target-site resistance to PS II inhibitors the norm and metabolism-based resistance the exception? Interestingly, in a survey of A. tuberculatus for atrazine resistance, fourteen of fifty-nine populations collected throughout Illinois exhibited resistance, and only one of these (based on observed maternal inheritance) contained an altered D1 protein as the

66

WEEDY AND INVASIVE PLANT GENOMICS

resistance mechanism (Patzoldt et al. 2002). Further research confirmed the existence of nontarget-site atrazine resistance in A. tuberculatus (Patzoldt et al. 2003). Thus, in the case of A. tuberculatus, despite the fact that target-site resistance was identified in this species, nearly a decade earlier (Table 5.2), non-target-site atrazine resistance appears to be more common. Further examination of the case of triazine-resistant A. tuberculatus in Illinois provides several insights. First, despite the fact that atrazine continues to be used on the majority of Zea mays (maize) hectares in Illinois (NASS 2008), and that A. tuberculatus with target-site atrazine resistance has been present in the state for several years (Foes et al. 1998), atrazineresistant A. tuberculatus is not considered to be a widespread problem. In contrast, resistance to ALS-inhibiting herbicides has become so common in A. tuberculatus populations throughout Illinois (and other Midwest states) that these herbicides are no longer recommended for control of this species (Patzoldt et al. 2002). The fact that target-site atrazine resistance has not become so widespread in A. tuberculatus might be attributable to the fitness penalty associated with it, and to A. tuberculatus’ dioecious nature: target-site atrazine resistance cannot be disseminated via pollen, unlike the case of nuclear-inherited ALS-inhibitor resistance. This could also account for the higher occurrence of non-target-site relative to target-site atrazine resistance in A. tuberculatus, because the non-target-site mechanism is a nuclear trait. This brings us back to the question, then, of why non-target-site triazine resistance is generally the exception. Non-target-site resistance generally provides a lower level of resistance than does target-site resistance to triazine herbicides, which would favor the evolution of the target-site resistance mechanism. For both Abutilon theophrasti (velvetleaf) and A. tuberculatus with non-target-site atrazine resistance, the level of resistance is not sufficient to overcome full labeled rates of atrazine applied preemergence (Gray et al. 1995; Patzoldt et al. 2003). Because atrazine is often used as a preemergence product, non-target-site resistance may provide limited benefit to the plant. It is also interesting to note that the discovery of non-target-site atrazine resistance in A. tuberculatus was somewhat accidental. The discovery occurred not because of a specific report of a resistant population, but rather from a survey of arbitrarily collected accessions. This suggests that non-target-site resistance might also be more common in other species than it is presently thought to be. The type of mutation that is required for a plant to acquire non-targetsite triazine resistance remains unknown, because this type of resistance has not been elucidated to the DNA level. As described in Chapter 10, exploration of non-target-site resistance mechanisms is ripe for genomics approaches, and these could provide more insights into why non-target-site relative to target-site triazine resistance occurs infrequently, despite the fitness cost associated with the latter.

Resistance To Acetolactate Synthase Inhibitors

More than ninety weed species have evolved resistance to ALS inhibitors, making it the leading group of herbicides for resistance evolution (Heap 2008). Here again, pigweeds are well represented, with eight confirmed resistant species (Table 5.2). Target-site resistance is the most common mechanism of resistance to ALS inhibitors, and reasons why ALS target site mutations occur so frequently are discussed in Chapter 9. Based on a catalog of resistance-conferring ALS mutations (Tranel et al. 2008), there are six conserved amino acids at which point mutations have been identified in weeds, and there is some evidence implicating at least two additional amino acids (Duran et al. 2003; Osuna et al. 2003; Laplante 2006). Thus, any of a multitude of ALS point mutations may render plants

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

67

resistant to ALS inhibitors. Again illustrating the prominence of Amaranthus species in resistance evolution, all six of the confirmed mutation sites (amino acid residue positions) are represented by one or more Amaranthus species, and two of the mutation sites (Asp376 and Ser653) are thus far represented exclusively by pigweeds (Tranel et al. 2008). Because of the importance of ALS-inhibiting herbicides and the frequency at which weed populations have evolved resistance to them, a wealth of data has been generated on the subject and several reviews written (e.g., Guttieri et al. 1996; Duggleby and Pang 2000; Tranel and Wright 2002). One of the interesting unanswered questions is what determines which amino acid mutation is selected in a given population. For example, the first ALS mutation identified in resistant Solanum ptycanthum (eastern black nightshade) was an alanine to threonine substitution (Ala122Thr), and was found in two geographically separate populations from the United States (Milliman et al. 2003). Subsequently, twelve of thirteen resistant populations of the same species from Canada were shown to contain an alanine to valine (Ala205Val) ALS substitution (Ashigh et al. 2008). Genetic analysis indicated that the Ala205Val substitution had independently evolved in at least some of the populations (i.e. occurrence of the same mutation in multiple populations was not solely due to gene flow from an initial selection event). Some of the factors that might account for the evolution of different mutations in different populations include: the herbicide selection pressure (different mutations confer different levels of resistance—in some cases no resistance—to different ALS-inhibiting herbicides), pleiotropic effects of the different mutations (e.g. a fitness penalty associated with a particular mutation would be expected to result in a low initial frequency for that mutation), the nucleotide sequence of the amino acid codon (some amino acid substitutions may be more readily obtained than others, depending on what nucleotide point mutation(s) is needed), and random chance. That all six of the confirmed ALS mutation sites identified thus far in weeds are present in Amaranthus species makes this group suited to explore these questions in more detail. Although resistance-conferring ALS mutations are generally thought to be associated with limited pleiotropic fitness penalties, a notable exception was reported in A. powellii (Tardif et al. 2006). Biotypes with the Trp574Leu ALS mutation were observed to produce significantly less biomass and were less competitive than sensitive biotypes. Additionally, the resistant biotypes produced distorted leaves, a characteristic that previously had not been associated with ALS mutations. The Amaranthus species, with their multitude of ALS mutations, provides an opportunity to explore associated fitness penalties in great detail.

Resistance To Protoporphyrinogen Oxidase Inhibitors

Resistance to herbicides that target the enzyme protoporphyrinogen oxidase (Protox) is rare among weeds, having thus far been documented in only three species (Heap 2008). The first of these was A. tuberculatus (Shoup et al. 2003), once again highlighting the preeminence of Amaranthus species in herbicide-resistance evolution. Why resistance to Protox inhibitors occurs infrequently has been a mystery (Dayan and Duke 1997), but the recent elucidation of the resistance mechanism in A. tuberculatus may offer some insights. The DNA mutation that conferred resistance to Protox-inhibiting herbicides in an A. tuberculatus biotype from Illinois was a codon deletion in the gene encoding the mitochondrial Protox isomer (PPX2) (Patzoldt et al. 2006). The mutation is unusual in that it is a deletion of a complete codon, not the typical substitution of one nucleotide for another. The deletion of a codon (three nucleotides) should not occur stepwise (i.e. one nucleotide after

68

WEEDY AND INVASIVE PLANT GENOMICS

another), because deletion of one or two nucleotides will alter the reading frame and, in all likelihood, result in a non-functional enzyme. It is presumed that the codon deletion occurred via a slippage-like mechanism that commonly occurs within short nucleotide repeats (microsatellites), as the deleted codon was within two overlapping repeats of tri-nucleotides (Gressel and Levy 2006; Patzoldt et al. 2006). Interestingly, similar codon insertion/deletion was observed among different A. tuberculatus PPX2 alleles and also among different ALS alleles (within the chloroplastic-targeting domain) from this species, suggesting that such genetic variation may be particularly common in A. tuberculatus (Patzoldt et al. 2006; Patzoldt and Tranel 2007). Although the codon-deletion was a completely novel herbicide-resistance mutation, further research revealed that the same mutation was present in multiple A. tuberculatus populations resistant to Protox-inhibiting herbicides (Lee et al. 2008). Physical distances between the populations, as well as nucleotide polymorphisms among the resistance alleles, suggest that the same codon deletion has independently evolved in these populations. In fact, no other resistance-conferring mutation within an A. tuberculatus Protox gene has yet been identified, suggesting that the codon deletion, rather than a point mutation, is the most facile evolutionary path to Protox-inhibitor resistance. Some support for this idea is provided by results of attempts to engineer resistance to Protoxinhibiting herbicides in crop plants. In these attempts, single point mutations provided either modest resistance or substantial fitness penalties, and thus an iterative approach was used to obtain double mutations (although in this case the mutations were within the chloroplastic PPX rather than the mitochondrial PPX)(Li and Nicholl 2005). Further research with other A. tuberculatus populations resistant to Protox-inhibiting herbicides and with the two other weed species with resistance to these herbicides should help answer the question of whether an unusual mutation (or multiple mutations) in PPX genes are required to provide effective herbicide resistance. In addition to the fact that the A. tuberculatus PPX mutation was a codon deletion, it was also determined that A. tuberculatus contained a dual-targeted PPX (Patzoldt et al. 2006). As was first determined with Spinacia oleracea (spinach) PPX2 (Watanabe et al. 2001), the A. tuberculatus mitochondrial PPX was encoded by a nuclear gene with a chloroplastic targeting sequence in frame and upstream of the mitochondrial targeting sequence. As demonstrated in S. oleracea, translation from the first translation initiation codon results in the enzyme being targeted to the chloroplast, whereas translation from the second methionine codon results in a mitochondrial-targeted Protox. Since the A. tuberculatus resistance-conferring mutation occurs in the gene for the dual targeted Protox, a herbicide-insensitive enzyme will be present in both organelles. Although the significance of having an herbicide-insensitive enzyme in both organelles to whole-plant resistance is unknown, this could further explain the paucity of resistance to Protox-inhibiting herbicides. Clearly, we have much more to learn about the molecular biology and evolution of resistance to Protox-inhibiting herbicides, and A. tuberculatus has emerged as the model for this research.

Resistance To Glyphosate

With the rapid and widespread adoption of transgenic, glyphosate-resistant crops, coupled with the consensus that glyphosate is an extraordinarily valuable herbicide, the evolution of glyphosate-resistant weeds has received increasingly more attention. At present, thirteen species are confirmed to have evolved glyphosate resistance (Heap 2008). Two of these species

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

69

belong to the Amaranthus genus (Table 5.2). Glyphosate-resistant A. palmeri was first identified in Georgia, United States, in 2004 and had evolved from the selection by glyphosate in glyphosate-resistant Gossypium hirsutum (cotton) production (Culpepper et al. 2006). Subsequently, glyphosate-resistant A. tuberculatus populations were identified in the state of Missouri and had evolved in glyphosate-resistant Glycine max (soybean) production systems (Legleiter and Bradley 2008). Additional glyphosate-resistant populations of both these pigweeds have now been identified in several other states (Heap 2008), and are a growing threat to agronomic crop production systems heavily reliant on glyphosate for weed control. Although the mechanism(s) of glyphosate resistance has not yet been determined in Amaranthus species, mutations at the target site (5-enolypyruvylshikimate-3-phosphate synthase, EPSPS) and altered glyphosate translocation (or cellular distribution) have been identified as likely primary mechanisms in several other glyphosate-resistant weed species (e.g., Baerson et al. 2002; Lorraine-Colwill et al. 2003; Feng et al. 2004; Wakelin and Preston 2006). Elevated levels of EPSPS (via increased transcription) and alterations in plant architecture (specifically, increased branching) also have been reported as contributing glyphosate-resistance mechanisms (Dinelli et al. 2006, 2008). In general, the level of resistance to glyphosate is quite low (typically less than ten-fold, in some cases up to about twenty-fold, relative to sensitive biotypes) when compared to many other herbicide resistances (which may be onehundred-fold or more). Because the resistance is low, it is more difficult to investigate, especially when working with genetically diverse weed populations that may express natural variation in glyphosate response. In fact, genetic variability for glyphosate tolerance has been documented for multiple weed species (Duncan and Weller 1987; Baucom and Mauricio 2008; see Chapter 11), including A. tuberculatus (Patzoldt et al. 2002; Smith and Hallett 2006). Such natural tolerance is likely from the combined contributions of multiple genes, suggesting that recurrent selection could lead to multigenic (quantitative) glyphosate resistance. To directly test this idea, two groups have subjected A. tuberculatus to recurrent selection using reduced rates of glyphosate with either whole plant or seedling-based assays (Zelaya and Owen 2005; Tranel et al. 2006). In both cases, small but significant decreases in glyphosate sensitivity were observed in subsequent generations. Zelaya and Owen used a divergent selection strategy and also succeeded in selecting for increased glyphosate sensitivity. Both groups also observed that, even after repeated cycles of selection, populations exhibited highly variable responses, consistent with the hypothesis that glyphosate resistance was acting as a multigenic trait. These selected populations would be well suited for more in-depth analysis using various genomic approaches, such as molecular mapping, to identify and isolate the genes involved. It also will be interesting to determine whether glyphosate-resistant A. tuberculatus and A. palmeri populations identified in agronomic fields will share resistance mechanisms with those obtained from artificial (greenhouse) selections. Regardless, genomic approaches may greatly facilitate identification of the glyphosate-resistance mechanism(s) in Amaranthus and other weed species, particularly non-target-site mechanisms (Yuan et al. 2007; see Chapter 10).

Multiple Herbicide Resistance

As shown in Table 5.2 and elaborated upon in preceding sections, Amaranthus species have proven adept at evolving resistance to particular herbicides. Nevertheless, because there are numerous herbicides (with different sites-of-action) with activity against Amaranthus species,

70

WEEDY AND INVASIVE PLANT GENOMICS

Figure 5.4. Illustration of multiple herbicide resistance in A. tuberculatus. The three plants on the left were from a sensitive population and were treated with a typical field use rate of (from left to right) a triazine herbicide, an acetolactate-synthaseinhibiting herbicide, or a protoporphyrinogen-oxidase-inhibiting herbicide. The plant on the far right was from an Adams County, Illinois, population and was treated with all three herbicides at the corresponding rates. (Reprinted from Patzoldt et al. (2005) courtesy of the Weed Science Society of America.)

resistance to one herbicide (or group of herbicides with the same site of action) does not, in most cases, leave a farmer without herbicidal options. However, when a single population or, worse yet, a single plant, possesses multiple herbicide resistances, then chemical weed control as a strategy may become threatened (see Chapter 1). Amaranthus species, particular A. tuberculatus, have been “leaders” among weeds in this arena as well. Lolium rigidum (rigid ryegrass) is second to none in terms of evolving resistance to multiple herbicides. For example, more than a decade ago a single population in Australia had evolved resistance to nine classes of herbicides in five chemical families (Burnet et al. 1994). In the United States, however, A tuberculatus has been the leader. An A. tuberculatus biotype from Illinois was reported with multiple resistances to triazine and ALS-inhibiting herbicides in 1998 (Foes et al. 1998). Subsequently, another biotype from the same state was reported with additional resistance to Protox-inhibiting herbicides (Figure 5.4) (Patzoldt et al. 2005). This three-way resistance was significant in that it left glyphosate as the only effective postemergence herbicide option in soybean. Then, a population from Missouri was reported with threeway resistances to ALS and Protox inhibitors and to glyphosate (Legleiter and Bradley 2008). Four-way resistance—to ALS and Protox inhibitors, triazines, and glyphosate—is not far away, and will severely limit herbicide options in both corn and soybean production. Amaranthus tuberculatus’ dioecious nature undoubtedly contributes to its ability to evolve multiple resistances. Gene flow via pollen movement from one resistant biotype to another effectively combines different resistance traits. However, two of the monoecious species, A. powellii and A. retroflexus (which are predominantly self-pollinated), have also evolved multiple resistances (Heap 2008). Why are Amaranthus species notorious for herbicide resistance evolution? Their high fecundities along with their common and widespread distributions mean that untold numbers of

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

71

Figure 5.5. Prolific seed production of Amaranthus species is a key to their success as weeds. The remains of an A. tuberculatus plant that survived repeated applications of glyphosate the previous year in an Illinois soybean field can be seen surrounded by thousands of seedlings. Many of these seedlings likely are its offspring and inherited resistance to glyphosate. (Photo courtesy of A. Hager, University of Illinois.)

individuals are the targets of selection every year (Figure 5.5). But do Amaranthus species also possess innate genetic properties that predispose them to rapid evolution, especially for herbicide-resistance? For example, do they have high transposable element activity, or other factors that contribute to high mutation rates? How much can rapid resistance evolution (as was observed with resistance to ALS-inhibiting herbicides) be attributed to multiple, independent founder events versus few selection events followed by rapid gene flow? Does hybridization among Amaranthus species contribute to its innate ability to evolve resistance (perhaps by making a more unstable genome that is susceptible to higher-than-typical mutation rates) or only to the potential for resistance to move from one species to another? Are insertion/ deletion mutations (as observed for resistance to Protox-inhibitors in A. tuberculatus) particularly common in Amaranthus species and, therefore, an unusual contributing factor to herbicide-resistance evolution? Are Amaranthus species particularly adept at evolving second-site mutations that might mitigate pleiotropic fitness penalties associated with herbicide-resistance mutations? These and many other questions are ripe for genomic-based investigations. Given the success of pigweeds in thwarting herbicidal control via resistance evolution, they provide an obvious model system for studying herbicide resistance in weeds.

Currently Available Genomic Resources

From the previous sections it is clear that there are a multitude of questions related to the weediness and evolution of Amaranthus species that would be greatly facilitated by genomic resources. In fact, Amaranthus may serve as one of the best model systems for genomic investigations of weediness (Basu et al. 2004; Chao et al. 2005; see also Chapter 4).

72

WEEDY AND INVASIVE PLANT GENOMICS Table 5.3. Haploid genome sizes of Amaranthus weeds and two model plant species (Amaranthus data from Rayburn et al. 2005, other data from http://data. kew.org/cvalues). Species

Haploid genome (Mbp)

A. palmeri A. spinosus A. hybridus A. powellii A. albus A. blitoides A. retroflexus A. tuberculatus Arabidopsis thaliana Oryza sativa

460 520 520 550 570 570 600 700 160 490

In addition to their collective weediness stature, Amaranthus species possess several characteristics that make them desirable as a model system. Most of the species are functional diploids with n = 16 or 17 (paleo-allotetraploids), although at least one species, A. dubious, is polyploid with 2n = 64 (Greizerstein and Poggio 1995; http://data.kew.org/cvalues). As shown in Table 5.3, major weedy Amaranthus species have genomes of tractable sizes: about three to four times the genome size of the model plant, Arabidopsis thaliana. Amaranthus species are generally easy to culture and manipulate under greenhouse and other experimental conditions. Seed dormancy present in some biotypes/species typically can be readily overcome with stratification, alternating temperatures, and/or red light treatment (see e.g., Gallagher and Cardina 1998; Leon et al. 2006). As is typical of weeds, Amaranthus species exhibit environmental plasticity and thus are easily induced to flower in response to stress or day length. For example, by growing A. hybridus plants with limited rooting volume, mature seeds can be obtained within a few weeks after planting, allowing several generations to be grown per year in a greenhouse. The monoecious species readily self-pollinate but also will outcross, and F1 hybrids are easily obtained if a selectable (e.g., herbicide resistance) or scoreable (e.g., pigmentation) marker is present in one of the parents. Although selfing is precluded in the dioecious species, clonal propagation of stem cuttings facilitates replicate analyses of a single genotype (Smith and Hallett 2006). Production of large numbers of small seeds makes Amaranthus species amenable to highthroughput screenings such as seedling bioassays and mutant screens. An extensive collection of Amaranthus germplasm is maintained at the North Central Region Plant Introduction Station and available for distribution (http://www.ars-grin.gov/npgs/index.html). Amaranthus species are also well represented in many herbarium collections. For example, the Illinois Natural History Survey contains more than 200 accessions of A. tuberculatus, with several collected in the early to mid-1900s (www.inhs.uiuc.edu/cbd/collections/plants.html). These could be used to investigate the recent evolutionary history of this species. The first genomics-based resources derived specifically from Amaranthus species are just now becoming available. A bacterial artificial chromosome (BAC) library was constructed from A. hypochondriacus (Maughan et al. 2008). This library contains about 37,000 clones and a predicted genome coverage of greater than ten-fold. Thus, any specific A. hypochondriacus DNA sequence has a greater than 99% chance of being represented in this library. The utility of the BAC library was demonstrated by obtaining full-length genomic sequences for the herbicide target-site genes, ALS and PPX2. Because of the genetic similarities among the

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

73

Amaranthus species (Figure 5.6), the BAC library from A. hypochondriacus is of direct utility to molecular-based investigations of the weedy species. Illustrating transferability between A. hypochondriacus and the weedy species, ALS and PPX2 sequence information from weedy species was used in the design of primers to generate probes to isolate the A. hypochondriacus BAC clones. One obvious use of the BAC library is as a tool for the isolation and identification of fulllength gene sequences. As a specific example, if altered expression of EPSPS is a suspected glyphosate-resistance mechanism in a particular biotype, the EPSPS promoter sequence could be obtained by screening the BAC library with a probe based on Amaranthus EPSPS cDNA sequence (GenBank accession AY545657). The resulting sequence from A. hypochondriacus could then be used to design primers to obtain and compare the corresponding promoter sequences from glyphosate-resistant and sensitive biotypes of the Amaranthus weed. The BAC library also will be used to generate a physical map of the Amaranthus genome, which ultimately will enable map-based cloning of Amaranthus genes of interest (Maughan et al. 2008). The second Amaranthus genomics resource to become available is a set of microsatellite markers (Mallory et al. 2008). Nearly 400 unique microsatellite markers were obtained primarily from microsatellite-enriched libraries but also from BAC-end sequence data. About 180 of these proved to be polymorphic across three grain Amaranthus species, and at least threefourths of these also amplified products in each of A. hybridus, A. powellii, and A. retroflexus. Although yet to be tested, many of these markers should be transferable to the dioecious Amaranthus species as well. Microsatellite markers are one of the most robust and informative markers for studies of genetic variability and, as such, will significantly build on previous molecular markers used with Amaranthus species (see e.g., Wetzel et al. 1999b; Wassom and Tranel 2005). They will be of great value to investigate questions related to gene flow, evolution, and hybridization within and among Amaranthus weeds. The microsatellite markers also can be used to construct genetic linkage maps from mapping populations (see below) and for multilocus genome scanning to identify genetic targets of selection. The general premise of this “hitchhiking” mapping strategy is that genetic variation will be lowest at and near the targets of selection (MaynardSmith and Haigh 1974). As an example, this approach was used to identify loci that were putative targets of selection during the evolution of the homoploid hybrid species Helianthus deserticola (Gross et al. 2007). A similar approach could be taken to identify potential loci that contributed to A. tuberculatus’ ability to recently and rapidly invade the midwestern United States. A third resource, not yet published, is genomic sequence information obtained from A. tuberculatus. Random sequencing was performed using massively parallel pyro-sequencing technology (454 Sequencing; Margulies et al. 2005). From a single pilot sequencing run, approximately 160,000 sequencing reads were obtained with an average read length of about 270 nucleotides, yielding a total of about 43 million nucleotides of A. tuberculatus sequence data (Tranel, unpublished data). Although this represents only a fraction of the genome (less than 10%), it nevertheless provides a wealth of data that is currently being mined for information. The sequence dataset includes more than 300 hits to candidate genes for herbicide metabolism (P450s, ABC transporters, glutathione S-transferases, and glycosyltransferases) as well as hits to herbicide target-site genes. The latter includes known genes from A. tuberculatus (including ALS and EPSPS) as well as previously unknown sequences (e.g. glutamine synthetase, the target of glufosinate) from this species. Genetic sequences associated with seed dormancy, floral timing, and sex expression also were identified, and could serve as leads to investigate particular aspects of A. tuberculatus’

PPX2 A. h. MVIQSITHLSPKLALPSPLSVSTKNYPVAVMGNISEREEPTSAKRVAVVGAGVSGLAAAY A. t. ...........N................................................ A. h. KLKSHGLNVTLFEADSRAGGKLKTVKKDGFIWDEGANTMTESEAEVSSLIDDLGLREKQQ A. t. .......S.................................................... A. h. LPISQNKRYIARDGLPVLLPSNPAALLSSNILSAKSKLQIMLEPFLWRKRNATELSDEHV A. t. ...........................T.....................H.......... A. h. QESVGEFFERHFGKEFVDYVIDPFVAGTCGGDPQSLSVHHTFPDVWNVEKRFGSVFAGLI A. t. .............................-.......M.....E...I............ A. h. QSTLLSKKEKGGGENASIKKPRVRGSFSFHGGMQTLVDTMCKQIGEDELKLQCEVLSLSY A. t. ............-................Q.............L................ A. h. NQKGIPSLGNWSVSSMSNNTSEDQSYDAVVVTAPIRNVKEMKIMKFGNPFSLDFIPEVTY A. t. ............................................................ A. h. VPLSVMITAFKKDKVKRPLEGFGVLIPSKEQHNGLKTLGTLFSSMMFPDRAPSDMCLFTT A. t. ............................................................ A. h. FVGGSRNRKLAKASTDELKQIVSSDLQQLLGTEDEPSFVNHLFWSNAFPLYGHNYDSVLR A. t. ...........N............................................C... A. h. AIDKMEKDLPGFFYAGNHKGGLSVGKAMASGCKAAELVISYLDSHLYVKMNEKTA A. t. .............................................I....D....

ALS A. h. MASNSSNPPFFYFTKPYKIPNLQSSIYAIPFSNSLKPTSSSS--IPRRPLQISSSSSQSP A. t. ...LLNHQ--.L....N...........L.............SS.L.............. A. h. KPKPPSATITQSPSSLTDDKPSSFVSRFSPEEPRKGCDVLVEALEREGVTDVFAYPGGAS A. t. ..............................D............................. A. h. MEIHQALTRSNIIRNVLPRHEQGGVFAAEGYARATGRVGVCIATSGPGATNLVSGLADAL A. t. .......................................................F.... A. h. LDSVPLVAITGQVPRRMIGTDAFQETPIVEVTRSITKHNYLVLDVEDIPRIVKEAFFLAN A. t. ............................................................ A. h. SGRPGPVLIDIPKDIQQQLVVPNWEQPIKLGGYLSRLPKPTYSANEEGLLDQIVRLVGES A. t. .........................................F.................. A. h. KRPVLYTGGGCLNSSEELRKFVELTGIPVASTLMGLGAFPCTDDLSLHMLGMHGTVYANY A. t. ...............................................Q............ A. h. AVDKADLLLAFGVRFDDRVTGKLEAFASRAKIVHIDIDSAEIGKNKQPHVSICGDVKVAL A. t. ............................................................ A. h. QGLNKILESRKGKVKLDFSNWREELNEQKKKFPLSFKTFGDAIPPQYAIQVLDELTKGDA A. t. R...N....................................................... A. h. VVSTGVGQHQMWAAQFYKYRNPRQWLTSGGLGAMGFGLPAAIGAAVARPDAVVVDIDGDG A. t. I........................................................... A. h. SFIMNVQELATIRVENLPVKIMLLNNQHLGMVVQWEDRFYKANRAHTYLGNPSNSSEIFP A. t. ..................................L......................... A. h. DMLKFAEACDIPAARVTKVSDLRAAIQTMLDTPGPYLLDVIVPHQEHVLPMIPSGAAFKD A. t. ............................................................ A. h. TITEGDGRRAY A. t. ...........

Figure 5.6. Genetic similarities among Amaranthus species as illustrated by protein similarities of two herbicide target sites. Inferred amino acid sequences from protoporphyrinogen oxidase (PPX2) and acetolactate synthase (ALS) genes from the monoecious, cultivated species A. hypochondriacus (A. h.) and the dioecious, weedy species A. tuberculatus (A. t.) are aligned. Identical amino acids are indicated by “.” and amino acid deletions by “-”. The amino acid change in each protein responsible for herbicide resistance in the A. tuberculatus biotype is underlined. GenBank accessions: PPX2 A. h.: EU024569, A. t.: DQ386116; ALS A. h.: EU024568, A. t.: EF157819.

74

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

75

biology and ecology. Additionally, several hundred microsatellites were identified; these are being used to design microsatellite-based DNA markers that will augment the set of such markers just described. Because plastid and mitochondrial DNA were not excluded prior to sequencing, many such sequences were obtained. In fact, a large portion of the plastid genome was obtained, and this information is being used to develop markers that will enable genetic diversity studies of the plastid genome. Results of such studies, when combined with nuclear genome diversity studies, will enable determination of the relative contributions of seed versus pollen movement to A. tuberculatus gene flow, and may also provide new information on interspecific hybridization. Recombinant inbred lines (RILs) have proven to be an extremely valuable resource for plant genomic efforts (see e.g., Uga et al. 2003; Clerkx et al. 2004; Krakowsky et al. 2004). Through single-seed descent, lines derived from a cross between two genetically diverse parents will become fixed for traits that differ between the two parents. Phenotypic analysis of the lines followed by genetic mapping make possible the identification of loci controlling specific traits. A segregating F2 cross between two grain amaranths was obtained and subjected to preliminary linkage analysis using microsatellite markers (Mallory et al. 2008). However, only about 20% of the tested markers were polymorphic, suggesting that lines from more genetically diverse parents are needed. Currently, RILs derived from a cross between A. hypochondriacus cv. “Plainsman: and A. hybridus are in development” (Tranel, unpublished data). Genetic diversity is expected to be high between the cultivated and weedy species, and the RILs should segregate for a wide variety of traits. For example, cursory visual assessment of F2 and F3 lines has revealed obvious differences in stem color, seed size and color, and amount of branching; detailed evaluation of advanced lines should provide insights into these and other traits (such as seed dormancy, developmental plasticity, and nutrient-use efficiency) that may play roles in the weediness and domestication attributes of amaranths.

Needs And Opportunities

As just described, a couple of Amaranthus genomic resources have recently become available and a couple more are in development. There are several additional resources that should be developed to facilitate Amaranthus research. New technologies have made large-scale de novo sequencing projects of non-model organisms feasible (Mardis 2008; Schuster 2008). Although it is premature to suggest full-genome sequencing of multiple Amaranthus species, these technologies could be used for partial genome sequencing as well as for transcriptome (cDNA) sequencing. Partial genome sequence (using BACs and/or random shotgun sequencing) would enable comparative genomics among Amaranthus species (as well as to other related species) and would provide insights on general genomic questions such as gene densities, quantities and types of repetitive elements, and species relatedness. Such an approach also could be used to test specific hypotheses related to interspecies hybridization and to address disparate genome sizes between species (such as the difference between the dioecious A. tuberculatus and A. palmeri; Table 5.3). Transcriptome sequencing would yield expressed sequence tags (ESTs) that could then be used to build microarrays for gene expression studies. For example, 454 sequencing can now yield 400,000 reads of 250 nucleotides each, and increases in both number of reads and read length are expected as the technology develops (Schuster 2008). Thus, a single sequencing run could provide ESTs for nearly all the genes of a species (providing, of course, that the genes

76

WEEDY AND INVASIVE PLANT GENOMICS

were expressed in the template cDNA). Resultant EST information could be used to produce long oligonucleotide microarrays, which could then be used to investigate a myriad of weed science questions related to herbicide resistance, weed biology, weed ecology, and evolution (Lee and Tranel 2008). Given the similarities of cDNAs among Amaranthus species (Figure 5.6), an array designed for one of them likely would be functional across all of them. Although modern DNA sequencing technologies can circumvent the need for DNA libraries, such libraries are valuable for isolation of cDNA clones and gene regulatory regions, as well as for physical genome mapping. The only Amaranthus libraries we are aware of are the A. hypochondriacus BAC and microsatellite libraries discussed in the preceding section. BAC libraries constructed from other Amaranthus species would allow for comparative mapping as well as targeted genomic sequencing of homologous regions among the Amaranthus genomes. Physically mapped BAC clones integrated with genetic linkage maps would serve as a springboard for candidate gene discovery once a locus is identified via either a hitchhiking mapping strategy or RIL analysis, as discussed above. cDNA libraries will greatly facilitate efforts to identify full-length clones, which will be needed for functional analysis of candidate genes. One of the best ways to test the function of a specific gene is to create transgenic plants that either over- or under-express the gene. Although this can be done using a heterologous approach (e.g., expressing an Amaranthus gene in Arabidopsis thaliana), Amaranthus genes that have evolved to confer a specific weediness trait to the species may function improperly or differently when placed in a different genomic context. Although not directly related to weediness, a clear example is a sex-determination gene from the dioecious A. tuberculatus or A. palmeri; although over-expression of such a gene in the hermaphroditic Arabidopsis thaliana might be an interesting experiment, it may provide very misleading information about the native function of the gene; i.e., it needs to be placed in an appropriate genomic context. Thus, although some Amaranthus species are amenable to Agrobacterium-mediated transformation (Bennici et al. 1992, 1997; Jofre-Garfias et al. 1997), transformation is very inefficient. The development of efficient, high-throughput transformation procedures for these species would facilitate functional evaluation of specific genes as well as development of T-DNA insertion lines that could be screened for mutations. Efficient transformation is clearly needed to establish Amaranthus as a weed genomics model system. It is interesting to note that, as mentioned above, the two Amaranthus species that are currently the most problematic in the U.S. are two dioecious species, A. tuberculatus and A. palmeri. As described in the section on hybridization and adaptive evolution, these two species are not as closely related as previously thought, which leads to the hypothesis that dioecy may have evolved independently in the two species. This hypothesis would be most readily tested with a genomics approach, in which the genes controlling the dioecious condition are identified and compared between the two species. The molecular biology of sex determination in dioecious pigweeds goes beyond academic interest in that its manipulation may provide for a unique weed control strategy. The idea of transgenically manipulating weeds to facilitate their control has been discussed previously by Gressel (2002). The rationale for such an approach stems at least in part from successes obtained with the sterile insect technique, in which irradiated (and therefore sterile) males are released in large numbers to compete with healthy males, thereby reducing the number of offspring produced. To improve upon this technique, transgenic manipulation was suggested as an alternative to irradiation, which can reduce fitness of the males and necessitates removal of females for optimum effect (Heinrich and Scott 2000; Thomas et al. 2000). An extension of this type of approach offers the potential for controlling the dioecious pigweeds.

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

77

One possibility would be to introduce a dominant “maleness” gene into, for example, A. tuberculatus, which would then be planted in field borders. Pollen from the transgenic plants would fertilize wild-type plants and the progeny of these fertilizations would all be male and heterozygous for the maleness transgene (assuming the initially released plants were homozygous for the gene). Thus, half the progeny from their subsequent matings would also receive the maleness gene. By inserting multiple copies of the gene into the initial transgenic line (for instance, by linking it to a transposable element), at least one copy of the maleness gene would persist for several generations. In theory, over time this approach would lead to elimination of females and local extinction. To be sure, there are several potential pitfalls of such an approach. To become reality, the first step is a genomics approach to understand the molecular basis of sex determination to see if a maleness gene exists (or can be created). The use of transgenic manipulation to control Amaranthus species is not likely to be reduced to practice in the very near future, but it illustrates a potential and novel weed control strategy that could be born from a genomics effort. In the more near term, a genomics approach focused on Amaranthus is expected to yield new tools that will greatly aid ongoing efforts to understand the biology, ecology, and weediness of these species. As highlighted in this chapter, Amaranthus species rank among the weediest of the weedy. They offer a myriad of questions ripe for genomic analysis, and may serve as an ideal model to answer the question of “what makes a plant a weed?” References Agong SG, Ayiecho PO (1991) The rate of outcrossing in grain amaranths. Plant Breeding 107, 156–160. Ahrens WH, Wax LM, Stoller EW (1981) Identification of triazine-resistant Amaranthus spp. Weed Science 29, 345–348. Ahrens WH, Stoller EW (1983) Competition, growth rate, and CO2 fixation in triazine-susceptible and -resistant smooth pigweed (Amaranthus hybridus). Weed Science 31, 438–444. Anderson E (1949) Introgressive Hybridization. New York: Wiley. Ashigh J, Rajcan I, Tardif FJ (2008) Genetics of resistance to acetohydroxyacid synthase inhibitors in populations of eastern black nightshade (Solanum ptycanthum) from Ontario. Weed Science 56, 210–215. Baerson SR, Rodriguez DJ, Tran M, Feng Y, Biest NA, Dill GM (2002) Glyphosate-resistant goosegrass. Identification of a mutation in the target enzyme 5-enolypyruvylshikimate-3-phosphate synthase. Plant Physiology 129, 1265–1275. Baltensperger DD, Weber LE, Nelson LA (1992) Registration of ‘Plainsman’ grain amaranth. Crop Science 32, 1510–1511. Basu C, Halfhill MD, Mueller TC, Stewart CN Jr. (2004) Weed genomics: new tools to understand weed biology. Trends in Plant Science 9, 391–398. Baucom RS, Mauricio R (2008) The evolution of novel herbicide tolerance in a noxious weed: the geographic mosaic of selection. Evolutionary Ecology 22, 85–101. Bennici A, Schiff A, Bovelli R (1992) In vitro culture of species and varieties of four Amaranthus L. species. Euphytica 62, 181–186. Bennici A, Grifoni T, Schiff A, Bovelli R (1997) Studies on callus growth and morphogenesis in several species and lines of Amaranthus. Plant Cell, Tissue and Organ Culture 49, 29–33. Brenner DM, Baltensperger DD, Kulakow PA, Lehmann JW, Myers RL, Slabbert MM, Sleugh BB (2000) Genetic resources and breeding of Amaranthus. In: Plant Breeding Reviews, Vol. 19, Janick J, ed. pp. 227–285. New York: Wiley. Burnet MWM, Loveys BR, Holtum JAM, Powles SB (1993) Increased detoxification is a mechanism of simazine resistance in Lolium rigidum. Pesticide Biochemistry and Physiology 46, 207–218. Burnet MWM, Hart Q, Holtum JAM, Powles SB (1994) Resistance to nine herbicide classes in a population of rigid ryegrass (Lolium rigidum). Weed Science 42, 369–377. Chan KF, Sun M (1997) Genetic diversity and relationships detected by isozyme and RAPD analysis of crop and wild species of Amaranthus. Theoretical and Applied Genetics 95, 865–873. Chao WS, Horvath DP, Anderson JV, Foley ME (2005) Potential model weeds to study genomics, ecology, and physiology in the 21st century. Weed Science 53, 929–937. Clerkx EJM, Groot SPC, Vreugdenhil D, Koornneef M, Blankestijn-De Vries H, El-Lithy ME, Vierling E, Ruys GJ (2004) Analysis of natural allelic variation of Arabidopsis seed germination and seed longevity traits between the accessions Landsberg erecta and Shakdara, using a new recombinant inbred line population. Plant Physiology 135, 432–443.

78

WEEDY AND INVASIVE PLANT GENOMICS

Costea M, Tardif FJ (2003) Conspectus and notes on the genus Amaranthus (Amaranthaceae) in Canada. Rhodora 105, 260–281. Costea M, Weaver SE, Tardif FJ (2005) The biology of invasive alien plants in Canada. 3. Amaranthus tuberculatus (Moq.) Sauer var. rudis (Sauer) Costea and Tardif. Canadian Journal of Plant Science 85, 507–522. Cowan P, Weaver SE, Swanton CJ (1998) Interference between pigweed (Amaranthus spp.), barnyardgrass (Echinochloa crus-galli), and soybean (Glycine max). Weed Science 46, 535–539. Culpepper AS, Grey TL, Vencill WK, Kichler JM, Webster TM, Brown SM, York AC, Davis JW, Hanna WW (2006) Glyphosate-resistant Palmer amaranth (Amaranthus palmeri) confirmed in Georgia. Weed Science 54, 620–626. Cummins I, Cole DJ, Edwards R (1999) A role for glutathione transferases functioning as glutathione peroxidases in resistance to multiple herbicides in black-grass. Plant Journal 18, 285–292. Dayan FE, Duke SO (1997) Phytotoxicity of protoporphyrinogen oxidase inhibitors: phenomenology, mode of action and mechanisms of resistance. In: Herbicide Activity: Toxicology, Biochemistry and Molecular Biology, Roe RM, Burton JD, Kuhr RJ, eds. pp. 11–35. Amsterdam: IOS Press. DeVlaming V, Proctor VW (1968) Dispersal of aquatic organisms: Viability of seeds recovered from the droppings of captive killdeer and mallard ducks. American Journal of Botany 55, 20–26. Dinelli G, Marotti I, Bonetti A, Minelli M, Catizone P, Barnes J (2006) Physiological and molecular insight on the mechanisms of resistance to glyphosate in Conyza canadensis (L.) Cronq. biotypes. Pesticide Biochemistry and Physiology 86, 30–41. Dinelli G, Marotti I, Bonetti A, Catizone P, Urbano JM, Barnes J (2008) Physiological and molecular bases of glyphosate resistance in Conyza bonariensis biotypes from Spain. Weed Research 48, 257–265. Duggleby RG, Pang SS (2000) Acetohydroxyacid synthase. Journal of Biochemistry and Molecular Biology 33, 1–36. Dumont M, Tardiff FJ (2002) Resistance to linuron in a Powell amaranth population. Weed Science Society of America Abstracts 42, 49–50. Duncan CN, Weller SC (1987) Heritability of glyphosate susceptibility among biotypes of field bindweed. Journal of Heredity 78, 257–260. Duran M, Plaza G, Osuna MD, De Prado R, Rodríguez-Franco A (2003) A mutation in the C domain of the acetolactate synthase (ALS) gene of Bidens pilosa confers resistance to imazethapyr. Proceedings of The BCPC International Congress: Crop Science and Technology, Volumes 1 and 2, pp. 819–824. Ellstrand NC, Schierenbeck KA (2000) Hybridization as a stimulus for the evolution of invasiveness in plants? Proceedings of the National Academy of Sciences of the United States of America 97, 7043–7050. Feng PCC, Tran M, Chiu T, Sammons RD, Heck GR, CaJacob CA (2004) Investigations into glyphosate-resistant horseweed (Conyza canadensis): retention, uptake, translocation, and metabolism. Weed Science 52, 498–505. Foes MJ, Liu L, Tranel PJ, Wax LM, Stoller EW (1998) A biotype of common waterhemp (Amaranthus rudis) resistant to triazine and ALS herbicides. Weed Science 46, 514–520. Franssen AS, Skinner DZ, Al-Khatib K, Horak MJ, Kulakow PA (2001) Interspecific hybridization and gene flow of ALS resistance in Amaranthus species. Weed Science 49, 598–606. Gallagher RS, Cardina J (1998) Phytochrome-mediated Amaranthus germination I: effect of seed burial and germination temperature. Weed Science 46, 48–52. Gray JA, Stoltenberg DE, Balke NE (1995) Absence of herbicide cross-resistance in two atrazine-resistant velvetleaf (Abutilon theophrasti) biotypes. Weed Science 43, 352–357. Greizerstein EJ, Poggio YL (1992) Estudios citogeneticos de seis higridos intersepecificos de Amaranthus (Amaranthaceae). Darwiniana 31, 159–165. Greizerstein EJ, Poggio YL (1994) Karyological studies in grain amaranths. Cytologia 59, 25–30. Greizerstein EJ, Poggio YL (1995) Meiotic studies of spontaneous hybrids of Amaranthus: genome analysis. Plant Breeding 114, 448–450. Gressel J (2002) Molecular Biology of Weed Control. London: Taylor and Francis. Gressel J, Levy AA (2006) Agriculture: The selector of improbable mutations. Proceedings of the National Academy of Sciences of the United States of America 103, 12215–12216. Gronwald JW, Anderson RN, Yee C (1989) Atrazine resistance in velvetleaf (Abutilon theophrasti) due to enhanced atrazine detoxification. Pesticide Biochemistry and Physiology 34, 149–163. Gronwald JW (1994) Resistance to photosystem II inhibiting herbicides. In: Herbicide Resistance in Plants: Biology and Biochemistry, Powles S, Holtum J, eds. pp. 27–60. Boca Raton: CRC Press. Gross BL, Turner KG, Rieseberg LH (2007) Selective sweeps in the homoploid hybrid species Helianthus deserticola: evolution in concert across populations and across origins. Molecular Ecology 16, 5246–5258. Guttieri MJ, Eberlein CV, Mallory-Smith CA, Thill DC (1996) Molecular genetics of target-site resistance to acetolactate synthase inhibiting herbicides. In: Molecular Genetics and Evolution of Pesticide Resistance, Brown TM, ed. pp. 10–16. Washington, D.C.: American Chemical Society.

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

79

Hager AG, Wax LM, Stoller EW, Bollero GA (2002) Common waterhemp (Amaranthus rudis) interference in soybean. Weed Science 50, 607–610. Heap I (2008) The International Survey of Herbicide Resistant Weeds. www.weedscience.com. Accessed April 18, 2008. Heinrich JC, Scott MJ (2000) A repressible female-specific lethal genetic system for making transgenic insect strains suitable for a sterile-release program. Proceedings of the National Academy of Sciences of the United States of America 97, 8229–8232. Hirschberg J, McIntosh L (1983) Molecular basis of herbicide resistance in Amaranthus hybridus. Science 222, 1346–1349. Holm L, Doll J, Holm E, Pancho J, Herberger J (1997) World Weeds: Natural Histories and Distribution. New York: Wiley. Holm LG, Plucknett DL, Pancho JV, Herberger JP (1991) The World’s Worst Weeds: Distribution and Biology. Malabar, Florida: Krieger. Holt JS, Thill DC (1994) Growth and productivity of resistant plants. In: Herbicide Resistance in Plants: Biology and Biochemistry, Powles S, Holtum J, eds. pp. 299–316. Boca Raton: CRC Press. Horak MJ, Peterson DE, Chessman DJ, Wax LM (1994) Pigweed Identification: A Pictorial Guide to the Common Pigweeds of the Great Plains. Manhattan, KS: Kansas State University Cooperate Extension Service. Jasieniuk M, Maxwell BD (2001) Plant diversity: new insights from molecular biology and genomics technologies. Weed Science 49, 257–265. Jofre-Garfias AE, Villegas-Sepúlveda N, Cabrera-Ponce JL, Adame-Alvarez RM, Herrera-Estrella L, Simpson J(1997) Agrobacterium-mediated transformation of Amaranthus hypochondriacus: light- and tissue-specific expression of a pea chlorophyll a/b-binding protein promoter. Plant Cell Reports 16, 847–852. Klingaman TE, Oliver LR (1994) Palmer amaranth (Amaranthus palmeri) interference in soybeans (Glycine max). Weed Science 42, 523–527. Knezevic SZ, Horak MJ (1998) Influence of emergence time and density on redroot pigweed (Amaranthus retroflexus). Weed Science 46, 665–672. Knezevic SZ, Weise SF, Swanton CJ (1994) Interference of redroot pigweed (Amaranthus retroflexus) in corn (Zea mays). Weed Science 42, 568–573. Krakowsky MD, Long MJ, Sharopova N, Lee M, Woodman-Clikeman WL (2004) QTL mapping of resistance to stalk tunneling by the European corn borer in RILs of maize population B73 × De811. Crop Science 44, 274–282. Laplante J (2006). Characterization of resistance to acetolactate synthase inhibitors in green foxtail: molecular genetics and biochemical basis. Weed Science Society of America Abstracts 46, 73–74. Lee RM, Hager AG, Tranel PJ (2008) Prevalence of a novel resistance mechanism to PPO-inhibiting herbicides in waterhemp (Amaranthus tuberculatus). Weed Science 56, 371–375. Lee RM, Tranel PJ (2008) Utilization of DNA microarrays in weed science research. Weed Science 56, 283–289. Legleiter TR, Bradley KW (2008) Glyphosate and multiple herbicide resistance in waterhemp (Amaranthus rudis) populations from Missouri. Weed Science 56, In press. Leon RG, Bassham DC, Owen MDK (2006). Germination and proteome analyses reveal intraspecific variation in seed dormancy regulation in common waterhemp (Amaranthus tuberculatus). Weed Science 54, 305–315. Li X, Nicholl D (2005) Development of PPO inhibitor-resistant cultures and crops. Pest Management Science 61, 277–285. Linnaeus C (1753) Species plantarum. Impensis L. Salvii. Holmiae. Lorraine-Colwill DF, Powles SB, Hawkes TR, Hollinshead PH, Warner SAJ, Preston C (2003) Investigations into the mechanism of glyphosate resistance in Lolium rigidum. Pesticide Biochemistry and Physiology 74, 62–72. Mallory MA, Hall RV, McNabb AR, Pratt DB, Jellen EN, Maughan PJ (2008) Development and characterization of microsatellite markers for the grain amaranths (Amaranthus spp. L.). Crop Science 48, 1098–1106. Mardis ER (2008) The impact of next-generation sequencing technology on genetics. Trends in Genetics 24, 133–141. Margulies M, Egholm M, Altman WE, et al. (2005) Genome sequencing in microfabricated high-density picolitre reactors. Nature 437, 376–380. Masabni JG, Zandstra BH (1999) A serine-to-threonine mutation in linuron-resistant Portulaca oleracea. Weed Science 47, 393–400. Massinga RA, Currie RS, Horack MJ, Boyer J Jr. (2001) Interference of Palmer amaranth in corn. Weed Science 49, 202–208. Maughan PJ, Sisneros N, Luo M, Kudrna D, Ammiraju JSS, Wing RA (2008) Construction of an Amaranthus hypochondriacus bacterial artificial chromosome library and genomic sequencing of herbicide target genes. Crop Science 48, S85–S94. Maynard-Smith J, Haigh J (1974) Hitch-hiking effect of a favorable gene. Genetical Research 23, 23–35. Mengistu LW, Mueller-Warrant GW, Liston A, Barker RE (2000) psbA mutation (valine(219) to isoleucine) in Poa annua resistant to metribuzin and diuron. Pest Management Science 56, 209–217.

80

WEEDY AND INVASIVE PLANT GENOMICS

Milliman LD, Riechers DE, Wax LM, Simmons FW (2003) Characterization of two biotypes of imidazolinone-resistant eastern black nightshade (Solanum ptycanthum). Weed Science 51, 139–144. Mosyakin SL, Robertson KR (2003) “Amaranthus.” In: Flora of North America. North of Mexico, pp. 410–435. New York: Oxford University Press. Murray MJ (1940) The genetics of sex determination in the family Amaranthaceae. Genetics 25, 409–431. NASS (2008) National Agricultural Statistics Service Agricultural Chemical Use Database. www.pestmanagement.info/ nass/. Accessed April 24, 2008. Ort DR, Ahrens WH, Martin B, Stoller EW (1983) Comparison of photosynthetic performance in triazine-resistant and susceptible biotypes of Amaranthus hybridus. Plant Physiology 72, 925–930. Osuna MD, Casado C, De Prado R, Wagner J, Hurle K (2003) A new mutation site in the acetolactate synthase (ALS) gene in Amaranthus quitensis resistant to imazethapyr. Proceedings of The BCPC International Congress: Crop Science and Technology, Volumes 1 and 2, pp. 819–824. Pal M, Pandey RM, Khoshoo TN (1982) Evolution and improvement of cultivated amaranths. Journal of Heredity 73, 353–356. Pandey RM (1999) Evolution and improvement of cultivated amaranths with reference to genome relationships among A. cruentus, A. powellii and A. retroflexus. Genetic Resources and Crop Evolution 46, 219–224. Park KW, Mallory-Smith CA (2006) psbA mutation (Asn(266) to Thr) in Senecio vulgaris L. confers resistance to several PS II-inhibiting herbicides. Pest Management Science 62, 880–885. Patzoldt WL, Tranel PJ, Hager AG (2002) Variable herbicide responses among Illinois waterhemp (Amaranthus rudis and A. tuberculatus) populations. Crop Protection 21, 707–712. Patzoldt WL, Dixon BS, Tranel PJ (2003) Triazine resistance in Amaranthus tuberculatus (Moq) Sauer that is not site-ofaction mediated. Pest Management Science 59, 1134–1142. Patzoldt WL, Tranel PJ, Hager AG (2005) A waterhemp (Amaranthus tuberculatus) biotype with multiple resistance across three herbicide sites of action. Weed Science 53, 30–36. Patzoldt WL, Hager AG, McCormick JS, Tranel PJ (2006) A codon deletion confers resistance to herbicides inhibiting protoporphyrinogen oxidase. Proceedings of the National Academy of Sciences of the United States of America 103, 12329–12334. Patzoldt WL, Tranel PJ (2007) Multiple mutations confer herbicide resistance in waterhemp (Amaranthus tuberculatus). Weed Science 55, 421–428. Pratt DB, Clark LG (2001) Amaranthus rudis and A. tuberculatus—one species or two? Journal of the Torrey Botanical Society 128, 282–296. Pratt DB, Owen MDK, Clark LG, Gardner A (1999) Identification of the Weedy Pigweeds and Waterhemps of Iowa. Ames, IA: Iowa State University, University Extension. Preston C, Powles SB (2002) Evolution of herbicide resistance in weeds: initial frequency of target site-based resistance to acetolactate synthase-inhibiting herbicides in Lolium rigidum. Heredity 88, 8–13. Rayburn AL, McCloskey R, Tatum TC, Bollero GA, Jeschke MR, Tranel PJ (2005) Genome size analysis of weedy Amaranthus species. Crop Science 45, 2557–2562. Rieseberg LH (1997) Hybrid origins of plant species. Annual Review of Ecology and Systematics 28, 359–389. Rowland MW, Murray DS, Verhalen LM (1999) Full-season Palmer amaranth (Amaranthus palmeri) interference with cotton (Gossypium hirsutum). Weed Science 47, 305–309. Ryan GF (1970) Resistance of common groundsel to simazine and atrazine. Weed Science 18, 614–616. Sauer JD (1953) Herbarium species as records of genetic research. The American Naturalist 187, 155. Sauer JD (1955) Revision of the dioecious amaranths. Madroño 13, 5–46. Sauer JD (1957) Recent migration and evolution of the dioecious amaranths. Evolution 11, 11–31. Sauer JD (1967) The grain amaranths and their relatives: A revised taxonomic and geographic survey. Annals of the Missouri Botanical Garden 54, 102–137. Sauer JD (1972) The dioecious amaranths: a new species name and major range extensions. Madroño 21, 425–434. Schulz-Schaeffer J, Baldridge DE, Bowman HF, Salknecht GF, Larson RA (1991) Registration of ‘Amont’ grain amaranth. Crop Science 31, 482–483. Schuster SC (2008) Next-generation sequencing transforms today’s biology. Nature Methods 5, 16–18. Shoup DE, Al-Khatib K, Peterson DE (2003) Common waterhemp (Amaranthus rudis) resistance to protoporphyrinogen oxidase-inhibiting herbicides. Weed Science 51, 145–150. Smith DA, Hallett SG (2006) Variable response of common waterhemp (Amaranthus rudis) populations and individuals to glyphosate. Weed Technology 20, 466–471. Sooby J, Myers R, Baltensperger D, Brenner D, Wilson R, Block C (1999) Amaranth Production Manual for the Central United States, Misc. Pub. EC98-151-S. Sidney, NE: University of Nebraska Cooperative Extension. Steckel LE, Sprague CL, Hager AG, Simmons FW, Bollero GA (2003) Effects of shading on common waterhemp (Amaranthus rudis) growth and development. Weed Science 51, 898–903.

21ST-CENTURY WEED SCIENCE: A CALL FOR AMARANTHUS GENOMICS

81

Steckel LE, Sprague CL (2004) Common waterhemp (Amaranthus rudis) interference in corn. Weed Science 52, 359–364. Steckel LE (2007) The dioecious Amaranthus spp.: here to stay. Weed Technology 21, 567–570. Steinback KE, McIntosh L, Bogorad L, Arntzen CJ (1981) Identification of the triazine receptor protein as a chloroplast gene product. Proceedings of the National Academy of Sciences of the United States of America 78, 7463–7467. Sun M, Chen H, Leung FC (1999) Low-Cot DNA sequences for fingerprinting analysis of germplasm diversity and relationships in Amaranthus. Theoretical and Applied Genetics 99, 464–472. Tardif FJ, Rajcan I, Costea M (2006) A mutation in the herbicide target site acetohydroxyacid synthase produces morphological and structural alterations and reduces fitness in Amaranthus powellii. New Phytologist 169, 251–264. Thomas DD, Donnelly CA, Wood RJ, Alphey LS (2000) Insect population control using a dominant, repressible, lethal genetic system. Science 287, 2474–2476. Tian X, Darmency H (2006) Rapid bidirectional allele-specific PCR identification for triazine resistance in higher plants. Pest Management Science 62, 531–536. Toler JE, Guice JB, Murdock EC (1996) Interference between johnsongrass (Sorghum halepense), smooth pigweed (Amaranthus hybridus), and soybean (Glycine max). Weed Science 44, 331–338. Tranel PJ, Wright TR (2002) Resistance of weeds to ALS- inhibiting herbicides: what have we learned? Weed Science 50, 700–712. Tranel PJ, Wassom JJ, Jeschke MR, Rayburn AL (2002) Transmission of herbicide resistance from a monoecious to a dioecious weedy Amaranthus species. Theoretical and Applied Genetics 105, 674–679. Tranel PJ, Lee RM, Bell MS, Singh S, Walter JR, Bradley KW (2006) What we know (and don’t know) about glyphosate resistance in waterhemp. Proceedings of the North Central Weed Science Society 61, 100. Tranel PJ, Wright TR, Heap IM (2008) ALS mutations from herbicide-resistant weeds. www.weedscience.org/mutations/ MutDisplay.aspx. Accessed April 25, 2008. Trucco F, Jeschke MR, Rayburn AL, Tranel PJ (2005a) Amaranthus hybridus can be pollinated frequently by A. tuberculatus under field conditions. Heredity 94, 64–70. Trucco F, Jeschke MR, Rayburn AL, Tranel PJ (2005b) Promiscuity in weedy amaranths: high frequency of female tall waterhemp (Amaranthus tuberculatus) × smooth pigweed (A. hybridus) hybridization under field conditions. Weed Science 53, 46–54. Trucco F, Tatum T, Rayburn AL, Tranel PJ (2005c) Fertility, segregation at a herbicide resistance locus, and genome structure in BC1 hybrids between two important weedy Amaranthus species. Molecular Ecology 14, 2717–2728. Trucco F, Tatum R, Robertson KR, Rayburn AL, Tranel PJ (2006) Morphological, reproductive, and cytogenetic characterization of Amaranthus tuberculatus × A. hybridus F1 hybrids. Weed Technology 20, 14–22. Trucco F, Zhang D, Woodyard AJ, Walter JR, Tatum TC, Rayburn AL, Tranel PJ (2007) Nonhybrid progeny from crosses of dioecious amaranths: implications for gene-flow research. Weed Science 55, 119–122. Uga Y, Fukuta Y, Cai HW, Iwata H, Ohsawa R, Morishima H, Fujimura T (2003) Mapping QTLs influencing rice floral morphology using recombinant inbred lines derived from a cross between Oryza sativa L. and Oryza rufipogon Griff. Theoretical and Applied Genetics 107, 218–226. Uscanga-Mortera E, Clay SA, Forcella F, Gunsolus J (2007) Common waterhemp growth and fecundity as influenced by emergence date and competing crop. Agronomy Journal 99, 1265–1270. Wakelin AM, Preston C (2006) A target-site mutation is present in a glyphosate-resistant Lolium rigidum population. Weed Research 46, 432–440. Wassom JJ, Tranel PJ (2005) Amplified fragment length polymorphism-based genetic relationships among weedy Amaranthus species. Journal of Heredity 96, 410–416. Watanabe N, Che FS, Iwano M, Takayama S, Yoshida S, Isogai A (2001) Dual targeting of spinach protoporphyrinogen oxidase II to mitochondria and chloroplasts by alternative use of two in-frame initiation codons. Journal of Biological Chemistry 276, 20474–20481. Weaver SE, McWilliams EL (1980) The biology of Canadian weeds. 44. Amaranthus retroflexus L., A. powellii S. Wats. and A. hybridus L. Canadian Journal of Plant Science 60, 1215–1234. Wetzel DK, Horak MJ, Skinner DZ, Kulakow PA (1999a) Transferal of herbicide resistance traits from Amaranthus palmeri to Amaranthus rudis. Weed Science 47, 538–543. Wetzel DK, Horak MJ, Skinner DZ (1999b) Use of PCR-based molecular markers to identify weedy Amaranthus species. Weed Science 47, 518–523. Yuan JS, Tranel PJ, Stewart CN Jr. (2007) Non-target-site herbicide resistance: a family business. Trends in Plant Science 12, 6–13. Zelaya IA, Owen MDK (2005) Differential response of Amaranthus tuberculatus (Moq ex DC) JD Sauer to glyphosate. Pest Management Science 61, 936–950. Zimdahl RL (1999) Fundamentals of Weed Science, 2nd Ed. New York: Academic Press.

6

Evolutionary Genomics Of Weedy Rice Briana L. Gross and Kenneth M. Olsen

Introduction

Wild relatives of crops have been implicated in the evolution of agricultural weeds in seven of the world’s top thirteen crop species (Ellstrand et al. 1999). These conspecific weeds present a special problem for agricultural productivity. Conspecific weeds often share close phenotypic similarity with crops, making them difficult to detect and eradicate by conventional cultivation practices. In addition, infestation of crop fields by conspecific weeds can facilitate gene flow from the crop into populations of weedy and/or wild relatives, potentially with long-term negative consequences. For example, weed-mediated gene flow from crops into wild relative populations can dilute wild germplasm with cultivar alleles, threatening the existence of true wild populations; in the case of lima bean (Phaseolus lunatus), a recent study revealed that gene flow from the crop into populations of the wild ancestor is approximately three times greater than gene flow in the opposite direction (Martinez-Castillo et al. 2007). Crop-weedwild gene flow may also preclude the effective use of engineered herbicide resistance as a weed control strategy, if the resistance trait can be easily introgressed from the crop into conspecific weed populations (Lu and Snow 2005). For rice (Oryza sativa L.), the crop’s conspecific weed is one of the most pervasive and destructive pests of rice fields worldwide. Weedy rice occurs with the crop wherever it is cultivated, both within and outside of the range of wild Oryza species. Weedy rice is often referred to as red rice because of its dark, reddish-brown pericarp that distinguishes it from most domesticated rice varieties. In the U.S., weedy rice has been recognized in rice fields of the Mississippi flood plain for more than one hundred years (Craigmiles 1978), and it is considered one of most destructive weeds of U.S. rice agriculture (Gealy 2005). Studies from the U.S. suggest that weedy rice has detrimental effects on harvest yield at a density of only two plants/m2 due high competitiveness and contamination of the harvest with dark red seeds (Kwon et al. 1991). Weed control efforts are recommended for weedy rice densities of one to three plants/m2, the same density recommended for Sesbania exaltata, a tree-like weed growing up to 3 meters tall (Smith 1988). At high densities, weedy rice can reduce crop yields by nearly 80% through competition with the crop for resources (Estorninos et al. 2005). Weedy rice contamination of harvests and seed stocks also reduces crop marketability, and the economic costs associated with red rice in the U.S. have been estimated at roughly $45 million annually, even with extensive herbicide management practices (Bridges and Baumann 1992). The prevalence and impact of weedy rice appears to be increasing in many parts of the world, including Europe (Bres-Patry et al. 2001; Messeguer et al. 2004), Latin America (Federici et al. 2001), Asia (Chin et al. 2000; Pyon et al. 2000; Watanabe et al. 2000; Cao et al. 2007), and the U.S. (Gealy et al. 2000). In Asia, for example, rice has been grown in close physical proximity to wild and weedy Oryza populations for thousands of years, yet weedy rice has only recently emerged as a major pest—within the last twenty years in China, Malaysia, and Vietnam (Chin et al. 2000; Watanabe et al. 2000; Cao et al. 2006; 83

84

WEEDY AND INVASIVE PLANT GENOMICS

Cao et al. 2007). A move toward mechanized farming (such as direct-seeding) and away from labor-intensive practices (such as hand transplanting and weeding) is probably responsible for this change (Barrett 1983; Watanabe et al. 2000; Cao et al. 2007). Given that rice is the world’s most important food crop, and that weedy rice pressure is increasing, successful control of weedy rice will potentially have a large impact on crop yield and quality world-wide. Because of its dramatic impacts on rice harvests, researchers are increasingly focused on efforts to characterize the biology of weedy rice. However, the general state of the research is still preliminary, and many basic questions remain unanswered. Here, we review the current state of knowledge about the phenotypic and genomic diversity of weedy rice, its origin and evolution, and the genetic basis of traits contributing to weediness.

Phenotypic Diversity Of Weedy Rice

Cultivated rice, Oryza sativa, is traditionally divided into two subspecies or varieties—indica and japonica—based on morphological, physiological, and agronomic qualities, as well as partial reproductive incompatibility. In general terms, indica is nonsticky after cooking, has longer, thinner grains, and is farmed in lowland tropical areas. Rice in the japonica subspecies is sticky after cooking, has shorter grains, and is farmed in temperate, upland areas. Recent research shows that these two subspecies reflect two independent domestication events from different populations of the wild progenitor, O. rufipogon (Londo et al. 2006). There are five major, genetically distinct subgroups within the two subspecies: indica contains indica and aus, while japonica is divided into temperate japonica, tropical japonica, and aromatic varieties (Garris et al. 2005; Caicedo et al. 2007). Only one type of rice tends to be cultivated in a given geographical region. For example, tropical japonica is cultivated in the southern U.S., where the bulk of rice is grown in North America, while temperate japonica is grown exclusively in California. Oryza rufipogon is itself sometimes divided into two species based on differences in life history characteristics: O. nivara (annual form) and O. rufipogon (perennial form). However, because there is no discernable genetic differentiation between these two wild forms (Ge et al. 2005; Caicedo et al. 2007), here we refer to them collectively as O. rufipogon, or simply wild rice. Oryza sativa and O. rufipogon are part of a complex of about seven diploid Oryza species belonging to the AA genome clade, which is roughly 2 million years old (Ge et al. 1999; Ge et al. 2005). Reproductive isolating barriers (RIBs) exist between all species in the AA clade, although fertile hybrids can be recovered from nearly any cross in this clade. Species of Oryza from this and other clades of Oryza are native to tropical areas around the world (Morishima et al. 1992). On a phenotypic level, researchers frequently cite two common morphologies of weedy rice: some weedy rice resembles wild rice, and some phenotypically approaches cultivated rice. In general, wild-type weed forms are distinguished as being much taller than cultivated rice plants, with dark-colored hulls with awns and being prone to shattering. “Crop mimics” are described as having short stature similar to many crop varieties, straw-colored hulls without awns, and lower levels of shattering than wild types. It has been suggested that these two phenotypes represent divergent adaptive strategies for weedy rice survival in a cultivated environment (Federici et al. 2001), although this hypothesis has not been tested. Alternatively, these phenotypes may simply be the product of different evolutionary origins (discussed below). Although these two main phenotypic forms are useful to describe the general appearance of weedy rice, many researchers have shown that strains can also combine wild and crop char-

EVOLUTIONARY GENOMICS OF WEEDY RICE

A.

B.

85

C.

Figure 6.1. (a) Rice field in Arkansas, infested with weedy rice, which is taller and lighter in color than the crop. (b) Panicle of BH weedy rice from Arkansas showing long awns and dark hull. (c) Collection of weedy rice samples from the U.S. Note the diversity of hull colors (from left to right, light brown, straw, dark brown/black) and awn presence/absence.

acters in some way. For example, extensive morphological characterization of U.S. weedy rice has revealed two major classes of weedy rice (Figure 6.1): black-hulled awned (BH) and strawhulled awnless (SH). While the SH type might be considered crop mimic form since it has a similar hull structure to cultivars, it and the BH type share a tall stature compared to cultivars, as well as greater tillering (production of culms), easy shattering, and strong seed dormancy (Diarra et al. 1985; Noldin et al. 1999). The main differences between these types are the hull/ awn morphology as well as earlier flowering in the SH as compared to BH (Diarra et al. 1985; Kwon et al. 1992; Noldin et al. 1999). There are reports of some weedy strains in the U.S. that do not overtop the cultivar (Noldin et al. 1999), but this phenotype is not often seen in morphological studies, perhaps simply because it is more difficult to find and collect. It is also useful to note that although weedy rice is sometimes described as flowering earlier than the crop, U.S. weedy strains are known to flower earlier, synchronously with, or after the crop (Gealy 2005), so that early flowering is not necessarily characteristic of weedy rice. Interestingly, SH is more common than BH in the U.S. (Diarra et al. 1985), although no explanation, either historical or selective, has been proposed to account for this difference. A diversity of coexisting weedy rice phenotypes have been reported in studies from around the world. An extensive survey of weedy rice in Costa Rica (Arrieta-Espinoza et al. 2005) documented forms ranging from crop mimics (including some that lack red pericarps) to some that closely resemble wild rice. Studies of weedy rice in Korea and Bhutan used an alternative system of classification, identifying weeds as either long-grained or short-grained (characteristic of O. sativa indica and japonica, respectively), which places an emphasis on the resemblance of weedy rice to the two major subspecies of cultivated rice (Cho et al. 1995; Ishikawa et al. 2005). This type of classification has not been widely adopted.

Genomic Diversity Of Weedy Rice Regional Studies

One of the earliest studies of the genetic diversity and structure of weedy rice focused on long- and short-grained Korean weedy rice, using restriction fragment length polymorphisms (RFLPs) and morphological characters (Cho et al. 1995). The common, short-grained weeds were found to be closely related to O. sativa japonica (cultivated in Korea), whereas the rare, long-grained weeds were closely related to O. sativa indica (not cultivated in Korea), although

86

WEEDY AND INVASIVE PLANT GENOMICS

Table 6.1. Studies characterizing the genomic diversity of weedy rice and relationship to samples of domesticated and wild rice. “x” denotes that samples were included, “w” indicates that weedy rice was genetically similar to a sample. Samples and relatedness of weedy rice* Weedy rice sampling

indica

japonica

Korea Bhutan Uruguay Northeastern China Southern USA Southern USA Southern USA and California Worldwide Worldwide

x, w x, w

x, w x, w x, w x, w x x x x, w x, w

x x, w x, w x, w x, w

wild rice

x x, w x, w w* x, w

Citation Cho et al. 1995 Ishikawa et al. 2005 Federici et al. 2001 Cao et al. 2006 Vaughan et al. 2001 Gealy et al. 2002 Londo and Schaal 2007 Suh et al. 1997 Ling-Hwa and Morishima 1997

*Some samples of weedy rice were inferred to be related to wild rice, although wild rice was not included in this analysis.

the latter group also showed some japonica-like phenotypic characteristics and japonicaspecific RFLPs (Table 6.1). While these analyses did not compare weedy rice genetic diversity to that of wild rice, and while the sampled pool of indica crop varieties was fairly narrow, this study was important in providing early information about the population genetics of weedy rice. Explicit comparison of these Korean weedy rice populations to a more diverse sample of cultivated and wild Oryza accessions could provide additional insights into the origin and evolution of these weed populations found at the northern limit of rice cultivation. A more recent study of weedy rice in Bhutan (Ishikawa et al. 2005) relied on simple sequence repeat (SSRs), cpDNA, isozymes, and morphology to compare weed strains with locally cultivated varieties of indica and japonica rice. This study is particularly interesting because Bhutan is one of relatively few regions where indica and japonica rice are cultivated in close proximity. The results indicated that weedy rice tends to genetically resemble the subspecies cultivated in the field where it is found, indicating that the weed is derived from both crop subspecies and does equally well as a weed of indica and japonica rice fields. Some weedy strains (and to a lesser degree cultivars) showed evidence of gene flow that appears to transcend the indica-japonica boundary. However, this conclusion might change if sampling were expanded to include locally occurring wild Oryza populations, because these alleles could potentially have been gained via gene flow with wild species. In a study of genetic diversity in South American weedy rice, Uruguayan weeds and samples of japonica rice, the locally cultivated subspecies, were surveyed using amplified fragment length polymorphic markers (AFLPs) (Federici et al. 2001). No Oryza species are native to Uruguay, and wild samples were not included in the study. Analyses of genetic diversity revealed three groups: two consisted entirely of weedy rice, and one consisted of weedy and cultivated rice. Of the two groups of weedy rice, one was straw-hulled and lacked awns (similar to the SH phenotype in the U.S.), and the other was black-hulled and awned (like the BH phenotype in the U.S.). The genetic differences of these weed strains are congruent with the genetically and phenotypically distinct patterns seen in U.S. weedy rice (see above and below). The weedy rice grouping with cultivars suggests some level of hybridization or shared ancestry with cultivars. A recent SSR-based study of weedy rice in northeastern China used a more extensive sampling strategy, including local and foreign japonica varieties, regional and foreign indica

EVOLUTIONARY GENOMICS OF WEEDY RICE

87

varieties (indica is not grown locally in northern China), and wild rice from China and elsewhere (Cao et al. 2006). The SSR analysis revealed that the weedy rice in this region is more closely related to locally cultivated japonica varieties and other japonicas than to either indica varieties or O. rufipogon (Cao et al. 2006). Another study of northeastern Chinese weedy rice, which included slightly fewer samples, has also revealed low levels of genetic diversity in the weed populations occurring in this region (Yu et al. 2005). The study by Cao and co-workers also revealed that over the sixteen years that weedy rice populations were surveyed, those populations persisting for multiple years showed a progressive decline in genetic diversity. The authors suggested that this might be because of strong selective pressures applied by farmers in weed eradication efforts. It would be interesting to know whether this pattern holds in other areas, because it potentially has implications for weed control strategies. For example, if a steady decline in genetic diversity can be taken to be a general characteristic of managed weed populations, then instances of persistent or increasing genetic diversity are likely to be reflecting repeated introductions of the weed, either via contaminated crop seed or gene flow from neighboring weed populations. In such cases, weed control efforts would be best focused on minimizing re-introductions of weed seeds. Several different groups have investigated the genetic structure of weedy rice in the southern U.S., where rice is cultivated well outside the native range of any wild rice species. The first large-scale survey, using eighteen SSRs and fairly wide sampling, was conducted by Vaughan et al. (2001). Analyses included thirty-four weedy rice strains sampled from across the geographical range of rice cultivation in the southern U.S., as well as indica rice, japonica rice (including U.S. cultivars, which in the southern U.S. are of the tropical japonica variety), and wild Oryza samples. The study showed that SH rice was most similar to indica rice and that BH rice was most similar to O. rufipogon (some BH individuals appeared to be identical to one particular accession of wild rice). There was also evidence of rare hybridization between weedy rice and U.S. cultivars. A subsequent study using a different set of eighteen SSR loci but less extensive sampling (only weedy rice and U.S. cultivars were included) confirmed that SH, BH, and U.S. cultivars are genetically distinct groups, and that there appears to be some rare hybridization between weedy rice and cultivars (Gealy et al. 2002). A recent study of U.S. weedy rice (Londo and Schaal 2007) examined samples from both O. sativa subspecies (including U.S. cultivars), wild rice, and a single weedy rice accession from California; genetic diversity was assayed using sixteen SSRs and one nuclear gene DNA sequence data set. This study largely confirmed previous findings that SH, BH, and U.S. cultivated rice are genetically distinct groups and that there are rare cases of hybridization between weedy rice and U.S. cultivars. In addition, however, the genetic resolution provided by the broad sampling and DNA sequence haplotypes revealed new insights into the potential origin of weed strains. SH weedy rice was found to have a mixture of markers that are characteristic of indica rice and O. rufipogon, indicating that this form of the weed may be derived from hybridization between those two groups. The BH weedy rice strains appear to be closely related to aus varieties of cultivated rice, which are allied with the indica subspecies; aus varieties are cultivated in upland areas of the northern Indian subcontinent. No samples of weedy rice from the southern U.S. showed the close genetic similarity to O. rufipogon seen in the Vaughan et al. (2001) study. The single sample of California weedy rice closely resembles O. rufipogon, with no obvious relation to cultivated rice. Weedy rice is far less common in California than in the southern U.S., because of successful weed control efforts, and so this single sample may represent a recent, accidental introduction of an O. rufipogon genotype from Asia. Finally, there was evidence of hybridization between SH and BH rice, which has not, to our

88

WEEDY AND INVASIVE PLANT GENOMICS

knowledge, been previously proposed. This has implications for the future of weedy rice evolution in the U.S., because it would be possible for adaptive mutations, such as herbicide resistance, to spread across all weedy strains through gene flow, rather than being restricted to one type or another. Despite the apparently wide phenotypic diversity present among U.S. strains of weedy rice, it should be noted that levels of genetic diversity are actually low in the weed in comparison to the major variety groups of cultivated rice. Indeed, reported measures of diversity for U.S. weedy rice as a whole are lower than almost every other group of cultivated or wild rice except possibly the geographically restricted aus variety (Londo and Schaal 2007). It would be particularly interesting to document whether levels of genetic diversity in U.S. weed populations are correlated with the length of time a population has been present, as in the study of Chinese weedy rice discussed above (Cao et al. 2006).

Worldwide Studies

To date, only two studies have included worldwide samples of weedy rice in a single genetic analysis. The first was conducted using morphological characters and isozymes and included weedy rice from Brazil, China, Japan, Korea, India, Nepal, Thailand, and the U.S.; as well as one japonica rice variety, one indica variety, and two wild rice samples (Ling-Hwa and Morishima 1997). A combined analysis revealed three groups of weedy rice, which were characterized as indica crop mimics, indica wild types (this group included the wild Oryza samples), and japonica wild types (Ling-Hwa and Morishima 1997). Note that in this and the following study, the authors of the papers use the terms crop mimic and wild type to describe genetic groupings, so these terms are not necessarily reflective of phenotypic similarity; this usage differs from how the terms are frequently used in the literature to describe broad morphological syndromes of weedy rice (see above). A more complex analysis used morphology, isozymes, RAPDs, and cpDNA (Suh et al. 1997). This study included samples of weedy rice from Bangladesh, Bhutan, Brazil, China, India, Japan, Korea, Nepal, Thailand, and the U.S., as well as two japonica rice varieties and two indica varieties. A factor analysis based on morphological and isozyme data revealed four groups: one contained the japonica samples and weedy rice, one contained the indica samples and weedy rice, and the other two consisted entirely of weedy rice (Suh et al. 1997). These groups were basically the same as those identified by Ling-Hwa and Morishima (1997), the difference being the delineation of four groups instead of three. The weedy rice in the first two groups were labeled as japonica and indica crop mimics, and the weedy rice in the second two groups were labeled as japonica and indica wild types. However, it should be noted that no wild rice was included in the study for comparison, so characterization of a group as wild type was based on similarity to common wild phenotypes rather than direct phenotypic or genomic comparisons. Despite the fact that this study was one of the earliest investigations of weedy rice genetics, the groups identified in the paper showed many patterns that either confirmed the few preceding studies or were confirmed in later studies. Crop mimics resembling japonica were found in Korea (the short-grained weedy rice) and Bhutan, which is in accordance with some of the patterns found by Cho et al. (1995) and Ishikawa et al. (2005). Some of the indica crop mimics were distributed in temperate climates, including the U.S. weedy rice (matching the studies by Vaughan et al. [2001] and Londo and Schaal [2007]). This group also included weedy rice from Bhutan, again consistent with the two types of weedy rice documented by Ishikawa

89

EVOLUTIONARY GENOMICS OF WEEDY RICE

et al. (2005), and the long-grained weedy rice from Korea (Cho et al.1995). A very small number of weedy rice samples from Korea and China made up the japonica wild type group. While the presence of Chinese weedy rice in this group does not correspond with the findings of Cao et al. (2006), who showed that weedy rice (in northeastern China, at least) was very closely related to local cultivars, it should not be considered contradictory; there is a diversity of wild rice in southern China that could contribute to the formation of weedy rice, and there may well be many different forms of weedy rice in China. The indica wild type weeds included weeds from mostly tropical climates, including Brazil and Thailand. These patterns are yet to be confirmed or refuted by other studies.

The Origin(s) And Evolution Of Weedy Rice

A fundamental question about weedy rice is whether it has a single origin or multiple origins. Perhaps because cultivated rice itself has multiple origins (Garris et al. 2005; Londo et al. 2006; Caicedo et al. 2007), researchers have generally been comfortable with the idea of multiple derivations of the conspecific weed. Indeed, the question itself has rarely been explicitly asked, and the default opinion seems to be that weedy rice is independently derived in each geographic area where it is found, with independently derived strains potentially coexisting in the same location. This expectation of multiple weed origins has been borne out in the U.S. (e.g., Londo and Schaal 2007), and is likely to hold true for other world regions as well. However, making the assumption of multiple, independent origins can be an obstacle to asking useful questions about the origins of this weed. For example, are there any specific weed genotypes that are particularly successful in invading more than one location? Are there any types of cultivated and/or wild rice or agricultural situations that facilitate the evolution of weedy rice, either by crop-weed gene flow or by “de-domestication” into weedy phenotypes (discussed below)? These questions can only be answered with broad surveys of weedy rice genetic diversity representing multiple locations—a sampling scheme beyond the scope of most studies to date. There are a variety of possible scenarios describing the origin of weedy rice, and the most frequently mentioned are as follows (see also Table 6.2):

Table 6.2. Four potential scenarios for the origin of weedy rice, the expected genomic patterns that would result from a scenario, the data required to prove the scenario, and the studies that show support for a scenario. Note that some studies are listed more than once. Origin of weedy rice Wild rice Domesticated rice Wilddomesticate hybrid Intersubspecific hybrid

Expected genomic patterns

Required for proof

Studies in support

Weedy rice resembles wild rice Weedy rice resembles domesticated rice Weedy rice has components of wild and domesticated genomes

Absence of domesticate alleles Absence of wild alleles

Weedy rice resembles domesticated rice, has components of both indica and japonica genomes

Absence of wild alleles, presence of domesticate alleles from both varieties

Vaughan et al. 2001; Londo and Schaal 2007 Suh et al. 1997; Ishikawa et al. 2005; Cao et al. 2006 Cho et al. 1995; Ling-Hwa and Morishima 1997; Suh et al. 1997; Londo and Schaal 2007 Cho et al. 1995; Ling-Hwa and Morishima 1997; Suh et al. 1997; Ishikawa et al. 2005

Presence of domesticate and wild alleles

90

• • • •

WEEDY AND INVASIVE PLANT GENOMICS

Weedy rice is a form of wild rice that adapted to agricultural or other disturbed environments, weedy rice is the result of the de-domestication or reversion of domesticated rice to a feral state, weedy rice evolved as a consequence of gene flow between domesticated rice and wild rice, and weedy rice is the result of hybridization between the two subspecies of cultivated rice.

The first three modes of origin are proposed to be general to weed evolution and were first articulated by de Wet and Harlan (1975), while the fourth is unique to rice. Because weedy rice may have many unique origins, all of these theories might be true for some lineages of weedy rice and none is mutually exclusive. In addition, as discussed below, several of these scenarios would be difficult to distinguish from each other definitively, even based on genetic data. Therefore, understanding the origin of weedy rice may not involve proving one of these scenarios to the exclusion of others, but rather establishing to what extent each scenario is important for the evolution of a given strain of weedy rice. Below, we evaluate each of these scenarios in light of the currently available information. Discussion of weedy rice origins has traditionally emphasized potential differences in weed evolution within vs. outside the native range of Oryza populations (Cho et al. 1995; Bres-Patry et al. 2001). The pervasiveness of weedy rice in regions where no wild Oryza species occur (e.g. North America, Europe) is sometimes cited as evidence that weedy rice arises through de-domestication of crop germplasm. However, it should be noted that there is no reason why weedy rice, like many other weeds, cannot be transported over large distances, especially in the context of agricultural exchange. Thus, weedy rice outside of the range of wild rice may still have its origins in wild Oryza germplasm. In contrast to the weed’s origin, the ongoing evolution of weedy rice may indeed be different within vs. outside the geographical range of wild Oryza populations. Within the range of wild rice, weeds can potentially introgress adaptive traits from both wild and cultivated rice, whereas the only potential for genetic exchange outside of the range of wild rice is with crop varieties. The nature of genetic exchange between wild, weedy, and cultivated rice is still being explored, and may offer interesting insights into the success of weedy rice.

Weedy Rice Is Derived From Wild Oryza Strains That Adapted To Agricultural Or Other Disturbed Environments

This scenario has implications for understanding the origin of cultivated plants and the potential repeatability of this process, because the evolution of crops and weeds both reflect adaptation to the human altered, agricultural environment (de Wet and Harlan 1975). Weedy rice thrives in agricultural environments, and traits characterizing at least some weed strains are traits that are also typical of cultivated plants (e.g. upright, compact growth habit), although of course there are other traits that differ greatly from crops, such as independent propagation. If some weed strains evolved directly from wild Oryza populations, with little or no genetic contribution from the domesticate, then evolution of the weedy form would in some respects be a repetition of the original domestication process acting on the same pool of genetic variation. It has been proposed that weedy forms of this type might have served as vital reservoirs of genetic variation that contributed desired traits to early cultivars (Harlan 1965).

EVOLUTIONARY GENOMICS OF WEEDY RICE

91

The long history of rice cultivation near wild Oryza populations in Asia and the widespread occurrence of crop-wild hybrid swarms (Oka and Chang 1961; Majumder et al. 1997) would argue against the likelihood of weedy rice origins with no genetic contribution from the crop. While some weedy rice strains possess traits more characteristic of a wild species than the crop (e.g. shattering, black hulls, long awns), such strains could easily represent weed forms derived from past hybridization between wild and cultivated rice, followed by adaptation to a weedy life history. In at least one case, weed strains have been shown not to be derived from locally occurring wild Oryza populations. In a study of Costa Rican weedy rice, ArrietaEspinoza et al. (2005) used a thorough morphological analysis to eliminate native diploid and polyploid species of New World wild rice (O. glumaepatula [AA genome], O. grandiglumis [CCDD genome], O. latifolia [CCDD genome]) as likely progenitors of the weed. Unfortunately, genetic surveys of weedy rice have frequently failed to include wild rice, making it difficult to know if wild rice populations are contributing a large amount of genetic variation to weedy strains. Of those studies that have included wild rice, a few of them show a close relationship between weedy rice and its wild progenitor, which strongly supports the importance of wild rice in the formation of weedy rice. Vaughan et al. (2001) found that some weedy rice in the southern U.S. were identical to an accession of O. rufipogon, although this was not confirmed in a later study (Londo and Schaal 2007). One sample of weedy rice from California showed an almost complete identity to wild rice at the genetic level (Londo and Schaal 2007). Clearly, a very complete survey of genetic variation in weedy rice would be necessary to eliminate the possibility of gene flow from cultivated plants into weedy lineages (and this ignores the possibility of allele sharing between cultivated and wild rice, see below), so the more realistic goal is to understand the relative contributions of wild and cultivated rice.

Weedy Rice Is The Result Of The De-Domestication Or Reversion Of Cultivated Rice To A Partially Wild State

This scenario resembles the previous one in that weedy rice is envisioned as being derived from a single taxon in response to selective pressures; however, it differs in the direction of evolution inasmuch that evolution occurs from a domesticated phenotype toward a more wild phenotype. There is some support for this scenario, based on the phenotypic and genotypic similarity of weedy strains to cultivated types. Many morphological surveys of weedy strains show overlap between weeds and cultivars, and the term crop mimic is sometimes used to describe weedy rice with awnless, straw-colored hulls (Federici et al. 2001). Given the close genomic similarity between cultivated rice and wild Oryza species (see e.g., Caicedo et al. 2007), it would be quite difficult to completely eliminate the possibility that crop varieties and wild Oryza populations have both contributed to the origin weed strains. This analysis would also be complicated by the difficulty of distinguishing cultivar from wild alleles, since the former would commonly be a subset of the latter. Thus, any conclusions about this scenario will have more to do with identifying important genetic contributions from wild rice to the origin and evolution of weedy rice. Also, as noted above, the widespread occurrence of weedy rice in regions with no wild Oryza populations cannot be taken as definitive evidence of weed evolution through de-domestication. Perhaps the strongest case for the de-domestication origin of weedy rice comes from an examination of weedy rice in northeastern China, where weedy strains showed close genetic similarity to local cultivars over wild samples (Cao et al. 2006). Several other studies have

92

WEEDY AND INVASIVE PLANT GENOMICS

shown that weedy rice is genetically related to domesticated rice (Suh et al. 1997; Ishikawa et al. 2005), but the lack of wild samples again makes these studies ultimately inconclusive in reference to this scenario. Ultimately, this scenario might best be tested by comparative sequencing of genes controlling phenotypes that differ between the weed and the crop (e.g. pericarp color, awn length). Based on DNA sequence haplotypes, one could potentially determine whether haplotypes conferring weed-associated traits are derived from wild Oryza strains or instead represent crop alleles that have undergone mutations allowing reversion to the phenotype of a wild strain. Although many genes controlling weediness in rice are yet to be identified, and no comparison between weedy and cultivated strains has been made for weediness genes to date, there is already some evidence that makes de-domestication unlikely. Many domestication traits have arisen through loss of function mutations, such as the loss of function of the Rc gene that results in the absence of pericarp pigmentation and the white pericarp that characterizes domesticated rice (Sweeney et al. 2006; Sweeney et al. 2007). This pattern is also true for two genes controlling shattering (Konishi et al. 2006; Li et al. 2006) and the Waxy gene, which controls amylose content in cultivated rice (Olsen and Purugganan 2002; Olsen et al. 2006). Thus, evolving weediness solely through de-domestication would require regaining function in nonfunctional alleles, a process that is evolutionarily unlikely.

Weedy Rice Evolved As A Consequence Of Gene Flow Between Domesticated Rice And Wild Rice

Both pre- and post-zygotic reproductive isolating barriers are present between O. sativa and its progenitor O. rufipogon, including pollen competition and hybrid sterility (Chu and Oka 1970; Majumder et al. 1997; Song et al. 2002). Hybrids between cultivated and wild rice are nonetheless present in areas where the two species overlap (Oka and Chang 1961; Chu and Oka 1970; Oka and Morishima 1971), and levels of gene flow are estimated to range from 1.1% to 2.75% (Song et al. 2003; Chen et al. 2004). In an important experiment, Oka and Morishima (1971) grew crop-wild hybrids in a cultivated environment, which included bulkharvesting and hand-sowing, and found an increase in some cultivated characteristics (such as loss of shattering) after as few as five generations. Although this experiment tended to produce more cultivar-like plants than weeds per se (nearly all weedy rice shatters easily), it does show that hybrids might quickly respond to selective pressures to survive in a cultivated environment. Many phenotypic surveys indicate that strains of weedy rice have both wild and cultivated characters, but it is impossible to determine the origin of these characters based on morphology alone. One genetic study has shown that SH weedy rice in the U.S. may be the result of hybridization between O. sativa indica and O. rufipogon (Londo and Schaal 2007). Hybridization has also been proposed to explain mixtures of genetic markers characteristic of O. sativa indica and japonica in Korean long-grained weedy rice (Cho et al. 1995) and in a surveys of weedy rice from around the world (Ling-Hwa and Morishima 1997; Suh et al. 1997). Determining whether a genetic marker is truly characteristic of cultivated or wild rice is difficult, so the ambiguity associated with determining whether an allele is derived from a wild or domesticated background could potentially make this a type of “default” explanation for many cases. Further characterization of the diversity in O. rufipogon will likely shed some light on this issue; if some samples of O. rufipogon are quite similar to both cultivated and weedy rice, then a wild origin for both of them might be the most likely scenario. There is a

EVOLUTIONARY GENOMICS OF WEEDY RICE

93

rapid increase of availability of DNA sequence data for O. rufipogon, and while the breadth of sampling within the species is still quite narrow, a survey of the NCBI Genbank database indicates that PopSets containing multiple O. rufipogon accessions are now available for more than 300 loci. The origin of weedy rice from crop-wild hybrids has implications for the weed’s continued evolution, as ongoing gene flow may be an important source of new adaptations to the cultivated environment. The potential for this type of genetic exchange is of particular interest in the context of the spread of transgenes or other cultivar alleles to the weedy form. Recent research has focused on this area because of the development of herbicide-resistant rice, marketed specifically as a solution to the weedy rice problem in agriculture (Lu and Snow 2005). Extensive experimentation has shown that the outcrossing rate from cultivated to weedy rice is generally below 1%, frequently as low as 0.1% and sometimes undetectable (Messeguer et al. 2001; Zhang et al. 2003; Chen et al. 2004; Shivrain et al. 2007). While these rates are low, the numbers of weedy rice plants in a field can be quite high (forty plants/m2 in a heavy infestation), so there may still be a large number of crop-weed hybrids each year and selection for differential survival of herbicide-resistant weeds would be very high (Gealy 2005). On the other hand, these studies generally measure the number of hybrids formed based on seeds collected after a single season, which ignores many potential post-zygotic reproductive isolating barriers. These barriers appear to be strong in at least some crosses, as in the case of U.S. crop-weed F1 hybrids, which flower very late compared to both parents (potentially not until after harvest occurs) (Langevin et al. 1990; Oard et al. 2000; Zhang et al. 2003; Rajguru et al. 2005). Thus, the actual rate of hybridization and introgression from U.S. elite crop varieties is largely unknown. The importance of ongoing crop-weed hybridization is not yet clear, although genetic data for U.S. weedy rice detected only a few crop-weed hybrids and no contribution of U.S. cultivars to the general gene pool of weedy rice (Gealy et al. 2002; Londo and Schaal 2007).

Weedy Rice Is The Result Of Hybridization Between The Two Subspecies Of Cultivated Rice

Although this scenario is not invoked with great frequency, it has appeared in the literature (Cho et al. 1995; Ling-Hwa and Morishima 1997; Suh et al. 1997; Bres-Patry et al. 2001; Ishikawa et al. 2005). The evolutionary potential for this scenario is generally based on some degree of non-complementation for some domestication traits that occurs in crosses between the independently domesticated O. sativa japonica and O. sativa indica (Bres-Patry et al. 2001). It is thought that the resulting offspring, segregating for wild characteristics, might persist and form a weedy lineage in crop fields. However, a comparison of F1 hybrids to weedy rice has not been conducted, nor has a thorough genetic model for this scenario been proposed. Weedy rice with genetic markers from both O. sativa japonica and indica have been described as potentially resulting from crosses between the two cultivars in several cases (Cho et al. 1995; Ling-Hwa and Morishima 1997; Suh et al. 1997; Ishikawa et al. 2005), although many authors also point out that this mixture of genetic markers might also result from crosses between cultivars of one subspecies and wild rice strains that resemble the other subspecies genetically (Cho et al. 1995; Ling-Hwa and Morishima 1997; Suh et al. 1997). The origin of a weed from intra-domesticate crossing is an interesting possibility, but it has been suggested based on evidence from a limited number of molecular markers, and is generally based on studies that did not include wild rice.

94

WEEDY AND INVASIVE PLANT GENOMICS

The Genetic Basis Of Weediness And Use Of Weedy Rice In Crop Breeding

The genetic basis of weedy traits is likely to be controlled, at least to some extent, by the same genes that differentiate wild from cultivated forms of rice. For example, a single mutation in the Rc gene results in a white pericarp in most cultivated rice, while the wild allele causes a dark red pericarp in wild rice (Sweeney et al. 2006; Sweeney et al. 2007). Weedy rice, most of which has a dark red pericarp, must have either the wild allele or a back-mutation that restores gene function, barring some compensatory mutation that would restore proanthocyanidin production (Sweeney et al. 2006). Shattering is also a prominent characteristic of weedy rice, and again is likely to be caused either by the presence of a wild allele or by back-mutation to a functional allele. Thus far, two genes controlling shattering in rice have been cloned (Konishi et al. 2006; Li et al. 2006) and a total of at least four have been identified in QTL mapping experiments (Cai and Morishima 2000), so changes at a variety of loci could potentially influence this trait in weedy rice. Alternatively, novel loci controlling weediness can be uncovered through QTL mapping experiments. QTL maps based on crosses between wild and cultivated rice have generally showed clusters of QTLs controlling domestication-related traits (Cai and Morishima 2000; Cai and Morishima 2002). Two QTL mapping experiments in weedy rice, one conducted using a weed × O. sativa japonica cross and the other conducted using weed × indica cross, also showed clustering of weediness QTL (Bres-Patry et al. 2001; Gu et al. 2005), although true colinearity with wild × crop maps has not been evaluated. A series of QTL experiments or even simple crosses between weedy rice and both wild and cultivated rice are needed to identify the genetic basis of weediness, particularly to determine whether the same loci control wild type traits such as shattering in both weeds and wild species. Of course, because different weedy strains may have independent origins, it is possible that weedy traits are controlled by different loci in different strains. Crosses between weedy strains may be informative for understanding this, especially for weedy rice that shows crop-like traits such as a lack of awns. A failure to complement for these traits would indicate that different genetic architecture underlies important adaptations in different weedy lineages. There has been considerable interest in the seed dormancy exhibited by weedy rice, which can allow seeds to persist in fields for years after they are shed. While dormancy is selected against in cultivated rice, some strains have developed a detrimental trait of preharvest sprouting (germination of seeds while they are still on the panicle), so dormancy-enhancing alleles are desirable for breeding purposes. Extensive mapping of dormancy-related QTLs with the goal of isolating such alleles in weedy rice has shown that there are pros and cons to the potential use of weedy rice in crop breeding efforts. While weedy rice has at least six loci controlling dormancy, many of them are closely linked to loci harboring alleles that are undesirable in cultivated rice, such as the functional version of the Rc gene (Gu, Kianian, Hareland et al. 2005; Gu, Kianian, and Foley 2005; Gu et al. 2006). If this is the case for many weedassociated traits, it will take careful planning to use the genetic variation present in weedy rice in crop improvement efforts. Conclusions Current Knowledge

Current research into weedy rice reveals some broad, consistent patterns. Some weedy forms closely resemble wild rice while others approach cultivars (either japonica or indica), based

EVOLUTIONARY GENOMICS OF WEEDY RICE

95

on both phenotypic and genomic patterns of variation. Weedy rice appears to have multiple, independent origins, and it is common for more than one strain of weedy rice to be present in a given region. Despite this apparent diversity of origins on a global scale, the few populationlevel surveys of weedy rice that have been conducted reveal low overall levels of variation. There is a variable amount of evidence supporting all four potential scenarios for the origin of weedy rice from pure wild, pure crop, crop-wild hybrids, or intersubspecific hybrids. Genetic evidence is strongest for the origin of weeds through crop-wild hybridization. The other two modes of evolution cannot be eliminated based on current research, although some components of rice domestication and cultivation make these seem evolutionarily difficult but not impossible. The evolution of weedy rice via hybridization of the two subspecies of domesticates is supported based on very limited species and genetic sampling, and has not been thoroughly evaluated.

Future Directions

Our understanding of the origin and evolution of weedy rice is at a preliminary stage, with major questions still unanswered. Future studies must include thorough sampling of cultivated and wild rice at both the population and genomic levels to discriminate between potential scenarios for the origin of weedy rice as well to elucidate current patterns of evolution. This level of resolution should be possible given the germplasm resources available for rice (accessions maintained at the International Rice Research Institute) and genomic resources (two sequenced O. sativa genomes: Goff et al. 2002; Yu et al. 2002; rapidly expanding genomic sequence data and other genomic resources for O. rufipogon and other wild Oryza species: Ammiraju et al. 2006). Studies with worldwide sampling of weedy rice will be more challenging, but are particularly needed to explore the spread or unique origin of the strains of weedy rice infesting agricultural fields around the world. Finally, conspecific crop weeds such as weedy rice can also provide excellent study systems in which to examine more general processes of adaptation in a controlled environment. Like their conspecific crop relatives, these weeds experience strong selective pressures in the context of cultivation. As has been shown for rice, some responses will be likely to resemble those of cultivated plants, while others will obviously be different. For example, weedy rice has an upright growth habit like cultivated rice, but many strains of weedy rice shatter instead of retaining their seeds. In rice and other important domesticates, the genetic basis of many domestication (and, conversely, wild) traits are well characterized. Thus, it should be possible to identify both the specific selective pressure and the adaptive response in the weed, which allows the development of a frame of research for detailing the genetic basis of adaptation.

References Ammiraju JSS, Luo MZ, Goicoechea JL, Wang WM, Kudrna D, Mueller C, Talag J, Kim H, Sisneros NB, Blackmon B, Fang E, Tomkins JB, Brar D, MacKill D, McCouch S, Kurata N, Lambert G, Galbraith DW, Arumuganathan K, Rao KR, Walling JG, Gill N, Yu Y, SanMiguel P, Soderlund C, Jackson S, Wing RA (2006) The Oryza bacterial artificial chromosome library resource: Construction and analysis of 12 deep-coverage large-insert BAC libraries that represent the 10 genome types of the genus Oryza. Genome Research 16,140–147. Arrieta-Espinoza G, Sánchez E, Vargas S, Lobo J, Quesada T, Espinoza AM (2005) The weedy rice complex in Costa Rica. I. Morphological study of relationships between commercial rice varieties, wild Oryza relatives and weedy types. Genetic Resources and Crop Evolution 52, 575–587.

96

WEEDY AND INVASIVE PLANT GENOMICS

Barrett SCH (1983) Crop mimicry in weeds. Economic Botany 37, 255–282. Bres-Patry C, Lorieux M, Clément G, Bangratz M, Ghesquière A (2001) Heredity and genetic mapping of domesticationrelated traits in a temperate japonica weedy rice. Theoretical and Applied Genetics 102, 118–126. Bridges D, Baumann P (1992) Weeds causing losses in the United States. In: Crop Losses due to Weeds in the United States, Bridges D, ed. Champaign: Weed Science Society of America. Cai HW, Morishima H (2000) Genomic regions affecting seed shattering and seed dormancy in rice. Theoretical and Applied Genetics 100, 840–846. Cai HW, Morishima H (2002) QTL clusters reflect character associations in wild and cultivated rice. Theoretical and Applied Genetics 104, 1217–1228. Caicedo A, Williamson S, Hernandez RD, Boyko A, Fledel-Alon A, York T, Polato N, Olsen K, Nielsen R, McCouch SR, Bustamante CD, Purugganan MD (2007) Genome-wide patterns of nucleotide polymorphism in domesticated rice. PLoS Genetics, 3(9), 1745–1756. Cao Q, Lu BR, Xia HUI, Rong JUN, Sala F, Spada A, Grassi F (2006) Genetic diversity and origin of weedy rice (Oryza sativa f. spontanea) populations found in north-eastern China revealed by simple sequence repeat (SSR) markers. Annals of Botany 98, 1241–1252. Cao QJ, Li B, Song ZP, Cai XX, Lu BR (2007) Impact of weedy rice populations on the growth and yield of direct-seeded and transplanted rice. Weed Biology and Management 7, 97–104. Chen LJ, Lee DS, Song ZP, Suh HS, Lu BR (2004) Gene flow from cultivated rice (Oryza sativa) to its weedy and wild relatives. Annals of Botany 93, 67–73. Chin DV, Hien TV, Thiet LV (2000) Weedy rice in Vietnam. In Wild and Weedy Rice in Rice Ecosystems in Asia—A Review, Baki B, Chin D, Mortimer M, eds. Phillipines: International Rice Research Institute. Cho YC, Chung TY, Suh HS (1995) Genetic characteristics of Korean weedy rice (Oryza sativa L.) by RFLP analysis. Euphytica 86, 103–110. Chu YE, Oka HI (1970) Introgression across isolating barriers in wild and cultivated Oryza species. Evolution 24, 344–355. Craigmiles J (1978) Introduction. In: Rice Research and Control. Texas Agricultural Experiment Station Bulletin, Eastin E, ed. College Station: State of Texas Publications Press. de Wet J, Harlan JR (1975) Weeds and domesticates: Evolution in the man-made habitat. Economic Botany 29, 99–107. Diarra A, Smith RJ Jr., Talbert RE (1985) Growth and morphological characteristics of red rice (Oryza sativa) biotypes. Weed Science 33, 310–314. Ellstrand NC, Prentice HC, Hancock JF (1999) Gene flow and introgression from domesticated plants into their wild relatives. Annual Review of Ecology and Systematics 30, 539–563. Estorninos LE Jr., Gealy DR, Gbur EE, Talbert RE, McClelland MR (2005) Rice and red rice interference. II. Rice response to population densities of three red rice (Oryza sativa) ecotypes. Weed Science 53, 683–689. Federici MT, Vaughan D, Tomooka N, Kaga A, Wang XW, Doi K, Francis M, Zorrilla G, Saldain NE (2001) Analysis of Uruguayan weedy rice genetic diversity using AFLP molecular markers. Electronic Journal of Biotechnology 4, 130–145. Garris AJ, Tai TH, Coburn J, Kresovich S, McCouch S (2005) Genetic structure and diversity in Oryza sativa L. Genetics 169, 1631–1638. Ge S, Guo YL, Zhu QH (2005) Molecular phylogeny and divergence of the rice tribe Oryzeae, with special reference to the origin of the genus Oryza. Paper read at “Rice is Life: Scientific Perspectives for the 21st Century,” November 4–7, 2004, at Tokyo and Tsukuba, Japan. Ge S, Sang T, Lu BR, Hong DY (1999) Phylogeny of rice genomes with emphasis on origins of allotetraploid species. Proceedings of the National Academy of Sciences of the USA 96, 14400–14405. Gealy DR (2005) Gene movement between rice (Oryza sativa) and weedy rice (Oryza sativa)—a U.S. temperate rice perspective. In: Crop Ferality and Volunteerism, Gressel J, ed. Boca Raton: CRC Press. Gealy DR, Saldain NE, Talbert RE (2000) Emergence of red rice (Oryza sativa) ecotypes under dry-seeded rice (Oryza sativa) culture. Weed Technology 14, 406–412. Gealy DR, Tai TH, Sneller CH (2002) Identification of red rice, rice, and hybrid populations using microsatellite markers. Weed Science 50, 333–339. Goff SA et al. (2002) A draft sequence of the rice genome (Oryza sativa L. ssp. japonica). Science 296, 92–100. Gu XY, Kianian SF, Foley ME (2005) Phenotypic selection for dormancy introduced a set of adaptive haplotypes from weedy into cultivated rice. Genetics 171, 695–704. Gu XY, Kianian SF, Hareland GA, Hoffer BL, Foley ME (2005) Genetic analysis of adaptive syndromes interrelated with seed dormancy in weedy rice (Oryza sativa). Theoretical and Applied Genetics 110, 1108–1118. Gu XY, Kianian SF, Foley ME (2006) Isolation of three dormancy QTLs as Mendelian factors in rice. Heredity 96, 93–99.

EVOLUTIONARY GENOMICS OF WEEDY RICE

97

Harlan JR (1965) The possible role of weed races in the evolution of cultivated plants. Euphytica 14, 173–176. Ishikawa R, Toki N, Imai K, Sato Y, Yamagishi H, Shimamoto Y, Ueno K, Morishima H, Sato T (2005) Origin of weedy rice grown in Bhutan and the force of genetic diversity. Genetic Resources and Crop Evolution 52, 395–403. Konishi S, Izawa T, Lin SY, Ebana K, Fukuta Y, Sasaki T, Yano M (2006) An SNP caused loss of seed shattering during rice domestication. Science 312, 1392–1396. Kwon SL, Smith RJ Jr., Talbert RE (1991) Interference of red rice (Oryza sativa) densities in rice (O. sativa). Weed Science 39, 169–174. Kwon SL, Smith RJ Jr., Talbert RE (1992) Comparative growth and development of red rice (Oryza sativa) and rice (O. sativa). Weed Science 40, 57–62. Langevin SA, Clay K, Grace JB (1990) The incidence and effects of hybridization between cultivated rice and its related weed red rice (Oryza sativa L.). Evolution 44, 1000–1008. Li C, Zhou A, Sang T (2006) Rice domestication by reducing shattering. Science 311, 1936–1939. Ling-Hwa T, Morishima H (1997) Genetic characterization of weedy rices and the inference on their origins. Breeding Science 47, 153–160. Londo JP, Chiang YC, Hung KH, Chiang TY, Schaal BA (2006) Phylogeography of Asian wild rice, Oryza rufipogon, reveals multiple independent domestication of cultivated rice Oryza sativa. Proceedings of the National Academy of Sciences of the United States of America 103, 9578–9583. Londo JP, Schaal BA (2007) Origins and population genetics of weedy red rice in the USA. Molecular Ecology 16, 4523–4535. Lu BR, Snow AA (2005) Gene flow from genetically modified rice and its environmental consequences. BioScience 55, 669–678. Majumder N, Ram T, Sharma A (1997) Cytological and morphological variation in hybrid swarms and introgresed population of interspecific hybrids (Oryza rufipogon Griff. × Oryza sativa L.) and its impact on evolution of intermediate types. Euphytica 94, 295–302. Martinez-Castillo J, Zizumbo-Villarreal D, Gepts P, Colunga-GarciaMarin P (2007) Gene flow and genetic structure in the wild-weedy-domesticated complex of Phaseolus lunatus L. in its mesoamerican center of domestication and diversity. Crop Science 47, 58–66. Messeguer J, Fogher C, Guiderdoni E, Marfà V, Català MM, Baldi G, Melé E (2001) Field assessments of gene flow from transgenic to cultivated rice (Oryza sativa L.) using a herbicide resistance gene as tracer marker. Theoretical and Applied Genetics 103, 1151–1159. Messeguer J, Marfà V, Català MM, Guiderdoni E, Melé E (2004) A field study of pollen-mediated gene flow from Mediterranean GM rice to conventional rice and the red rice weed. Molecular Breeding 13, 103–112. Morishima H, Sano Y, Oka HI (1992) Evolutionary studies in cultivated rice and its wild relatives. In: Oxford Surveys in Evolutionary Biology: Volume 8, Futuyma DJ, Antonovics J, eds. Oxford: Oxford University Press, New York. Noldin JA, Chandler JM, McCauley GN (1999) Red rice (Oryza sativa) biology. I. Characterization of red rice ecotypes. Weed Technology 13, 12–18. Oard J, Cohn MA, Linscombe S, Gealy D, Gravois K (2000) Field evaluation of seed production, shattering, and dormancy in hybrid populations of transgenic rice (Oryza sativa) and the weed, red rice (Oryza sativa). Plant Science 157, 13–22. Oka HI, Chang WT (1961) Hybrid swarms between wild and cultivated rice species, Oryza perennis and O. sativa. Evolution 15, 418–430. Oka HI, Morishima H (1971) The dynamics of plant domestication: Cultivation experiments with Oryza perennis and its hybrid with O. sativa. Evolution 25, 356–364. Olsen KM, Purugganan MD (2002) Molecular evidence on the origin and evolution of glutinous rice. Genetics 162, 941–950. Olsen KM, Caicedo AL, Polato N, McClung A, McCouch S, Purugganan MD (2006) Selection under domestication: Evidence for a sweep in the rice Waxy genomic region. Genetics 173, 975–983. Pyon J, Kwon W, Guh J (2000). Distribution, emergence, and control of Korean weedy rice. In: Wild and Weedy Rice in Rice Ecosystems in Asia—A Review, Baki B, Chin D, Mortimer M eds. Phillipines: International Rice Research Institute. Rajguru SN, Burgos NR, Shivrain VK, Stewart JM (2005) Mutations in the red rice ALS gene associated with resistance to imazethapyr. Weed Science 53, 567–577. Shivrain VK, Burgos NR, Anders MM, Rajguru SN, Moore J, Sales MA (2007) Gene flow between Clearfield(™) rice and red rice. Crop Protection 26, 349–356. Smith RJ Jr. (1988) Weed thresholds in southern U.S. rice, Oryza sativa. Weed Technology 2, 232–241.

98

WEEDY AND INVASIVE PLANT GENOMICS

Song Z, Lu B, Zhu Y, Chen J (2002) Pollen competition between cultivated and wild rice species (Oryza sativa and O. rufipogon). New Phytologist 153, 289–296. Song ZP, Lu BR, Zhu YG, Chen JK (2003) Gene flow from cultivated rice to the wild species Oryza rufipogon under experimental field conditions. New Phytologist 157, 657–665. Suh HS, Sato YI, Morishima H (1997) Genetic characterization of weedy rice (Oryza sativa L.) based on morphophysiology, isozymes and RAPD markers. Theoretical and Applied Genetics 94, 316–321. Sweeney MT, Thomson MJ, Cho YG, Park YJ, Williamson SH, Bustamante CD, McCouch SR (2007) Global dissemination of a single mutation conferring white pericarp in rice. PLoS Genetics 3, e133. Sweeney MT, Thomson MJ, Pfeil BE, McCouch S (2006) Caught red-handed: Rc encodes a basic helix-loop-helix protein conditioning red pericarp in rice. Plant Cell 18, 283–294. Vaughan LK, Ottis BV, Prazak-Havey AM, Bormans CA, Sneller C, Chandler J M, Park WD (2001) Is all red rice found in commercial rice really Oryza sativa? Weed Science 49, 468–476. Watanabe H, Vaughan D, Tomooka N (2000) Weedy rice complexes: case studies from Malaysia, Vietnam, and Surinam. In: Wild and Weedy Rice in Rice Ecosystems in Asia—A Review, Baki B, Chin D, Mortimer M, eds. Phillipines: International Rice Research Institute. Yu GQ, Bao Y, Shi CH, Dong CQ, Ge S (2005) Genetic diversity and population differentiation of Liaoning weedy rice detected by RAPD and SSR markers. Biochemical Genetics 43, 261–270. Yu J et al. (2002) A draft sequence of the rice genome (Oryza sativa L. ssp. indica). Science 296, 79–92. Zhang N, Linscombe S, Oard J (2003) Out-crossing frequency and genetic analysis of hybrids between transgenic glufosinate herbicide-resistant rice and the weed, red rice. Euphytica 130, 35–45.

7

Rhizomatousness: Genes Important For A Weediness Syndrome Andrew H. Paterson

Introduction

Rhizomes are underground stems that are fundamentally important in plant competitiveness and invasiveness, playing two contrasting roles in agriculture. As a primary means of dispersal, rhizomes are an important component of weediness of many of our most noxious weeds, including Johnsongrass (Sorghum halepense L. Pers.), Bermuda grass (Cynodon dactylon L. Pers.), purple nutsedge (Cyperus rotundus), quack grass (Agropyron repens), and cogon grass (Imperata cylindrica). Both Johnsongrass and Bermuda grass were introduced to the U.S. as prospective crops, but became major weeds, partly because of their aggressive rhizomes. The threat of other such escapes restricts use of valuable germplasm in improvement of several crops. For example, the rhizomatous grasses Oryza longistaminata (sexually compatible with rice) and Saccharum spontaneum (sexually compatible with sugarcane) harbor many genes of potential value for improving rice and sugarcane, but are illegal to grow in the field in the U.S. because of the threat that they may become weeds. By contrast, rhizomes are also a valuable asset in establishment and persistence of dense, productive stands of forage, and turfgrasses cultivated on more than 60 million acres in the southern U.S. alone (Burton 1989), including Cynodon spp. (Bermuda grass), Paspalum spp. (bahia and dallisgrass), Pennisetum/Cenchrus spp. (buffelgrass), and many others. Such grasses have an estimated value of $3 billion per year in the U.S. as forage, and a national economic impact estimated at $24 billion annually (Barnes and Baylor 1995), via meat, dairy, and fiber (wool) production. These species, together with wild perennial grasses, form a dense subterranean “net” which plays a major role in erosion control. Failure to recognize this role was partly responsible for the Dust Bowl epochs that periodically crippled the economies of various parts of the U.S. Many leading candidate cellulosic biofuels crops are perennial plants, at least partially because of rhizomatousness. The expansion of agriculture to provide plant biomass for production of fuels or chemical feedstocks will require greater use of marginal lands to make production of low per-unit value biomass economical. Perennial crops are essential to bringing marginal lands into sustainable biomass production (Cox et al. 2002; Scheinost 2001; Wagoner 1990), maximizing ecosystem productivity (Field 2001) and minimizing losses of topsoil (Pimentel et al. 1995), water, and nutrients. In a Missouri field experiment monitored for more than 100 years, perennial cover crops were more than fifty times more effective than annual crops in maintaining topsoil (Gantzer et al. 1990). Perennials are also thirty to fifty times more effective than annuals at preventing nitrogen losses (Randall and Mulla 2001). There might be substantial room for improving biomass yield in leading candidate cellulosic biofuels crops, in particular regarding the relative allocations of photosynthate to harvestable biomass versus perennation organs such as rhizomes. Wild herbaceous perennial plants may be “overbuilt” for survival under dramatically varying conditions, and for reproductive fitness integrated over many years. It may very well be possible to divert some photosynthate from 99

100

WEEDY AND INVASIVE PLANT GENOMICS

that excess below-ground capacity to other organs, in proportions targeted by the breeder. This adjustment is feasible because an agriculturally managed landscape is not subject to the extreme stresses that prevail in the wild. In a worst-case scenario, farmers can occasionally re-sow a perennial crop, and probably do so periodically to mitigate the accumulation of pests (microbial pathogens, insects, and nematodes). Better understanding of the genetics of rhizomatousness, and in particular the genetics of carbon allocation to harvestable versus perennation organs, may be of fundamental importance to tailoring existing plants to sustainable biomass production.

Developmental Context

Botanically, rhizomes are modified subterranean stems that are diageotropic (e.g. orient their growth perpendicular to the force of gravity), but spawn geotropic shoots that can become independent ramets (Figures 7.1a–c). Rhizomes and tillers both develop from axillary buds at the lowermost nodes of the erect leafy shoot of the plant. These basal buds exhibit a clear positional gradient, with an acropetally increasing tendency to develop into tillers, both in Sorghum halepense and in Agropyron repens (Gizmawy et al. 1985). Primary rhizomes initiate in Johnsongrass seedlings at about the five-leaf stage, and rhizome development accelerates after the formation of ten leaves (Anderson et al. 1960; McWhorter 1961a).

Figure 7.1. a. Johnsongrass (Sorghum halepense), as an example of a rhizomatous plant. Rhizomes account for about 70% of the dry weight of the mature Johnsongrass plant (McWhorter 1961), and can spawn new genetically identical plants at a considerable distance from the original crown. Rhizomes differentiate from vegetative buds below the soil surface. Prior to differentiation, these buds are virtually identical to those that give rise to tillers. The commitment of a bud to rhizome development, tiller development, or quiescence is genetically determined (Paterson et al. 1995a), and is associated with marked differences in gene expression (Jang et al. 2006). (Drawing by Charlene Chang.)

RHIZOMATOUSNESS: GENES FOR A WEEDINESS SYNDROME

101

Figure 7.1. b. Crowns of S. bicolor (left) and S. propinquum. The vast majority of the S. propinquum crown is composed of rhizomes, most of which were truncated during harvest. (Photo by the author.)

Figure 7.1.

c. Yearling Johnsongrass plant, with rhizomes spanning about 1 meter. (Photo by the author.)

Although rhizomes and tillers are anatomically related, they are physiologically very different. Mature rhizomes exhibit the basic properties of storage organs; they are apparently immune to the autolytic senescence processes that affect above-ground plant parts (Oyer et al. 1959a). In perennial Sorghums, rhizomes are the primary repository of sucrose (L. Tarpley and D. Vietor, Texas A&M, unpublished), the main storage carbohydrate in tropical grasses.

An Exemplary Case: Johnsongrass

The Sorghum genus has become a model for dissecting the molecular control of perenniality (Hu et al. 2003; Jang et al. 2006; Paterson et al. 1995c), a subject that has been inadequately explored in view of the many advantages that might be realized by use of perennial crops for biofuel production (above) and perhaps even for food and feed production (Cox et al. 2002). Sorghum halepense L. (Johnsongrass) is one of the world’s most noxious weeds (Holm et al.

102

WEEDY AND INVASIVE PLANT GENOMICS

1977), and a paradigm for the potential dangers of crop-weed introgression. Sorghum halepense is native to western Asia, but was introduced and has naturalized in tropical and warm temperate climates worldwide (Holm et al. 1977). Cytological, morphological, and molecular genetic data suggest that S. halepense is a naturally formed tetraploid hybrid derivative of S. bicolor, an annual, polytypic African grass species which includes cultivated sorghum, the fifth most important grain crop worldwide, and S. propinquum, a perennial native to moist habitats in southeast Asia (Celarier 1958; Doggett 1976; Paterson et al. 1995b). Sorghum halepense is a major contaminant in sorghum seed production, an alternate host and means of overwintering for pests and pathogens of both monocot and dicot crops, and a highly effective competitor for sunlight and other growthlimiting resources. Reductions in economic yield of 45% or greater in monocots such as sugarcane (Millhollen 1970) and dicots such as soybean (McWhorter and Hartwig 1972) are specifically attributable to Johnsongrass. Johnsongrass may have been intentionally introduced into the U.S. as a prospective forage, and/or unintentionally introduced as a contaminant of seed lots, and was well-established in several southern states by 1830 (McWhorter 1971). The name “Johnsongrass” came to supplant some forty common names and is first documented in an 1874 letter, referring to Colonel William Johnson, an Alabama plantation owner who sowed it on his farm (McWhorter 1971). The first federal appropriation for weed control research targeted Johnsongrass (House Bill #121 1900). Sorghum halepense produces extensive rhizomes (subterranean stems that confer perenniality and also provide for clonal propagation) that can account for 70% of an individual plant’s dry weight (Oyer et al. 1959b) and make it difficult and expensive to eradicate. A single plant can produce about 212 linear feet (65 meters) of rhizomes, with mass of more than 8 kg, in 152 days of growth from seed (McWhorter 1961). The same plant produces about 3 kg of leaves, 0.68 kg of seed, and 0.43 kg of roots. Modern herbicides control Johnsongrass in major crops such as maize, soybean, and cotton, but at an estimated cost of $12 to $20 per acre. No herbicide is available to control Johnsongrass in sorghum. The relationship between these species is so close that all known compounds kill both species. The Johnsongrass problem exemplifies constraints on improvement of many crops through both biotechnology and traditional breeding. Although S. bicolor (2n = 2x = 20) and S. halepense (2n = 4x = 40) differ in ploidy, numerous artificial crossing studies have demonstrated that S. bicolor can serve as the pollen parent of triploid and tetraploid interspecific hybrids (reviewed in Tang and Liang 1988; Warwick and Black 1983). The two species frequently grow in close physical proximity and have overlapping flowering periods. Experimental field studies demonstrate the potential for S. halepense × S. bicolor hybrid formation (Arriola and Ellstrand 1996) and persistence (Arriola and Ellstrand 1997). A high level of gene flow from cultivated sorghum to Johnsongrass occurs in regions such as the U.S. Southern Plains, where the two are closely associated, but is spreading thousands of miles to regions such as New Jersey, where sorghum is not known to have ever been produced (Morrell et al. 2005). To my knowledge, no transgenic S. bicolor is commercially available because of the high risk of transgene escape into Johnsongrass. Transformation of sorghum (Casas et al. 1993) with genes for resistance to insects, diseases, or herbicides could have a major economic impact on areas of Kansas, Nebraska, Oklahoma, and Texas, which often receive too little rainfall to economically produce other crops such as maize. First stably transformed in 1993 (Casas et al. 1993), sorghum has been transformed by a wide variety of methods within the last five years alone (Carvalho et al. 2004; Cheng et al. 2004; Gao et al. 2005a; Gao et al. 2005b; Gray et al.

RHIZOMATOUSNESS: GENES FOR A WEEDINESS SYNDROME

103

2004; Howe et al. 2006; Jeoung et al. 2004; Krishnaven et al. 2004; Mythili et al. 2004; Nguyen et al. 2007; Rathus et al. 2004; Sticklen and Oraby 2005; Wang et al. 2007; Williams et al. 2004; Zhao 2006) with a growing set of potentially valuable genes (Bird and Akhurst 2007; Girijashankar et al. 2005; Krishnaven et al. 2004; Tadesse and Jacobs 2004; Yu et al. 2005).

Dissecting The Genetic Control Of Rhizomatousness

Growth and development of rhizomes is an area of plant biology that remains underexplored, perhaps because most major crop gene pools are based on non-rhizomatous genotypes. All members of the cultivated species, Sorghum bicolor, are non-rhizomatous. However, close relatives are rhizomatous (S. propinquum; see Figure 7.2), as is the ancestral form of a sister species, Saccharum spontaneum. These observations suggest that rhizomatousness is ancestral within the Saccharinae clade (Figure 7.2), and that the loss of rhizomes in Sorghum bicolor has been within the past approximately 1 million years since its divergence from a common ancestor shared with S. propinquum, based on a recent estimate of the level of sequence diversity between these two species, (Feltus et al. 2004). Since all S. bicolor genotypes known, both wild and cultivated, are non-rhizomatous, the trait was presumably lost early in the radiation of S. bicolor. S. bicolor and S. propinquum are each well characterized, and their naturally occurring polyploid hybrid Johnsongrass (S. halepense) is even more invasive than S. propinquum, with nearly worldwide distribution. Indeed, hybridization with cultivated S. bicolor has been proposed as a potential cause of increased aggressiveness in the weed (see e.g., de Wet and Harlan 1975; Holm et al. 1977). S. bicolor and S. propinquum have the same ploidy, and are readily crossed to generate fertile hybrids. Detailed phenotypic and DNA marker analysis of progeny from a cross between these two species has led to much of our knowledge of the genetic control of rhizomatousness (Paterson et al. 1995c). About 370 F2 individuals from a cross between these two species were

Rhizomatous? Sorghum bicolor

no

Sorghum halepense

yes

Sorghum propinquum

yes

Miscanthus sinensis

yes

Miscanthus x. giganteus

yes

Miscanthus sacchariflorus

yes

M. sacchariflorus ‘Robustus’ yes Saccharum spontaneum

yes

Saccharum cultivars

no

Saccharum officinarum

no

Saccharum robustum

no

Figure 7.2. Rhizomatousness in current and prospective Saccharinae crops.

104

WEEDY AND INVASIVE PLANT GENOMICS

assessed in the field near College Station, Texas, allowed to overwinter, and assessed again in the spring. Subsequent analysis of F3 families for a subset of these individuals permitted estimates of heritability for rhizome-related traits. The entire population was genotyped with DNA markers at well-spaced intervals across the entire genome, permitting QTL (quantitative trait loci) mapping by an interval analysis approach (using flanking markers to infer the genotype of the intervening region (Lander 1989). Variation in the number of rhizomes producing aboveground shoots was associated with three QTLs. Variation in regrowth (ratooning) after overwintering was associated with QTLs accounting for additional rhizomatous growth and with QTLs influencing tillering. Vegetative buds that become rhizomes are similar to those which become tillers—one QTL appears to influence the number of such vegetative buds available, and additional independent genes determine whether individual buds differentiate into tillers or rhizomes (respectively). Rhizomatousness and regrowth are each complex traits in which QTL mapping using F2 plants and their F3 progeny was only able to account for 14% to 30% of phenotypic variance— much may remain to be learned. Recently, a recombinant inbred line (RIL) population has been produced by selfing the same F2 progeny used for QTL mapping. To my knowledge this is the first RIL population in any plant that segregates for rhizomatousness, offering the possibility to improve sensitivity to detect QTLs with small effect. Seeds of the RIL population are presently being increased and early mapping of the population is in progress. In principle, positional information about rhizomatousness QTLs provides a foundation for cloning of genes responsible for rhizomatousness, aided by the recent, nearly-complete sequencing of the Sorghum bicolor genome (Paterson et al. 2009). In practice, however, this will be a complicated task, in view of the subterranean nature of the phenotype and relatively low variance explained in genetic models. Association genetic approaches, although generally attractive in sorghum (Hamblin 2005), are of limited value for this trait because there is no phenotypic variation within S. bicolor, and there are woefully inadequate collections available for S. propinquum and S. halepense.

Corresponding Quantitative Trait Loci Affecting Rhizomatousness Of Sorghum And Rice

Based on QTL data, it was suggested nearly a decade ago that the independent domestications of a number of divergent wild grasses that were to become the world’s leading cereal crops had occurred largely as a result of allelic variants in genes (QTLs) that were at corresponding genomic locations. In other words, independent but convergent mutations in a relatively small number of corresponding genes may have played a substantial role in early progress to increase seed size, reduce shattering, and adjust flowering to be conducive to optimal productivity (Paterson et al. 1995a). Comparative studies also show close correspondence in genes associated with rhizomatousness of sorghum and rice, respectively (Hu et al. 2003). Rhizomatousness of rice was investigated using an F2 population derived from a cross between cultivated rice (Oryza sativa) and one of its wild relatives, O. longistaminata. Genetic analyses and molecular mapping based on a complete simple sequence repeat (SSR) map revealed two dominant complementary genes, Rhz2 and Rhz3, controlling rhizomatousness. Comparative mapping indicated that each of these genes, plus many additional rice QTLs with smaller effects on rhizomatousness, closely correspond to major QTLs controlling rhizomatousness in Sorghum propinquum. In total, ten (62.5%) of sixteen rice QTLs correspond to nine (100%) of nine sorghum QTLs influencing rhizomatousness. The likelihood of the

RHIZOMATOUSNESS: GENES FOR A WEEDINESS SYNDROME

105

observed number of matches occurring by chance was 0.000005, indicating that QTLs controlling rhizomatousness in O. longistaminata and S. propinquum fall in largely corresponding genomic locations (Hu et al. 2003). Correspondence of these genes in rice and sorghum, which are estimated to have diverged from a common ancestor about 41 million years ago (Paterson et al. 2004), suggests that the Rhz2 and Rhz3 genes in particular may be key regulators of rhizome development in many Poaceae. The genetically dominant nature of the Rhz2 and Rhz3 loci support the notion that perenniality is ancestral to annual habit, suggested by the sympatric distribution of annual and perennial species within many grass genera and the general association of perenniality with wild species and annual habit with cultigens. Strong epistatic interaction between Rhz2 and Rhz3 and their pleiotropic effects on many rhizome traits suggest that Rhz2 and Rhz3 may be regulatory genes. Both also strongly influenced tiller number, a trait associated strongly with regrowth (ratooning) and persistence of perennial grasses. Rhz3 accounted for greater phenotypic variance, and affected more rhizome-related traits than did Rhz2, implying that Rhz3 may be more “upstream” in the rhizomatousness pathway. The pleiotropic effects of Rhz2 on rhizome branching degree, rhizome internode length, and tiller number, on the other hand, suggest its influence on the growth of axillary buds and/or intercalary meristems.

Early Insights Into The Sorghum Rhizo-Transcriptome

Expressed-sequence tag (EST) surveys of cDNA populations isolated from rhizome tissues have provided early insights into the pathways and processes active in rhizome development. More than 10,000 ESTs from nearly 6,000 different clones were sampled randomly from an S. propinquum rhizome cDNA library as part of a broader sampling of the sorghum transcriptome (Pratt et al. 2005). The comparison of random cDNAs from the rhizo-transcriptome to genes that were expressed at substantially higher levels in rhizomes than other plant parts of either S. propinquum or S. halepense has provided insight into their probable functions, genomic organization, and regulation (Jang et al. 2006). Two rhizome cDNA libraries that were screened for highly expressed genes included one from S. propinquum (described above) and one from S. halepense, in each case produced using mRNA from the terminal 1 cm of rhizomes (rhizome tips). From each library, a total of 18,432 (= 48 × 384) recombinant clones were picked into 384-well plates and archived by cryopreservation as a permanent resource (copies of these plates are available to the community; to order, see www.plantgenome.uga.edu/catalog/). The 36,864 arrayed clones were used to produce nylon macroarrays and probed with mRNA to distinguish two broad classes of genes: (1) those predominantly expressed in the growing point, i.e. terminal 1 cm of rhizomes (rhizome tip, or RT) and (2) those predominantly expressed in mature rhizomes (MR) from internodes at least 10 cm distal to the rhizome tip and near the point of maximum rhizome diameter. To differentiate genes that are overexpressed in early rhizome growth and development from those shared with mature rhizome tissue or other plant parts, we compared expression profiles based on mRNA from RT and MR to one another and to mRNA from pooled aboveground (A) plant parts (See Figure 7.1a). All experiments included both biological replicates (mRNA from independent tissue samples) and technical replicates (application to two different replica macroarrays, with each clone represented by two spots per array). Examined were the 768 clones with the highest RT/MR and RT/A expression ratios (192 per ratio from each of the two cDNA libraries) and 192 with the lowest ratios as a negative control. The total of 960 clones showing extreme expression levels (768 high and 192 low) in

106

WEEDY AND INVASIVE PLANT GENOMICS

rhizomes were sequenced, then the resulting 889 high-quality sequences were assembled into 218 singlets and 392 contigs. The sequences were assigned to Gene Ontology-based functional categories to permit general comparison of the clones in different expression categories to one another and to the random sampling of clones from the S. propinquum library. Genes with no homology to previously known sequences were more abundant in the highly expressed genes than in the random set, suggesting that many previously unknown genes show rhizome-enriched expression. The high RT/A set showed striking enrichment for genes implicated in secondary/hormone metabolism and chromatin/DNA metabolism (although to a lesser degree). Both highly-expressed sets showed some enrichment for genes implicated in cell wall structure/metabolism and abiotic stimuli and development. The genes that had low levels of expression in rhizomes were enriched in those involved in protein synthesis and processing, but completely excluded those related to the cytoskeleton. The sorghum sequence was not yet available, but use of the rice sequence permitted us to begin to explore regulatory motifs that might be associated with regulatory signals in rhizomes, implicating gibberellins in regulation of many rhizome-expressed genes. For those genes that could be unambiguously anchored onto a single location of the rice genome sequence, we identified the transcription initiation site and extracted 1,000 bp of 5′ sequence. These regions were analyzed for cis-acting regulatory elements by using the PLACE database (www.dna. affrc.go.jp/PLACE). Sorghum genes that were relatively highly expressed in rhizome tip tissues were enriched for the pyrimidine box, TATCCA box, and CAREs box, and cis-element motifs related to GA response. The results implicate gibberellins in regulation of many rhizome-specific genes, a finding that is supported by prior evidence from other rhizomatous species (Jacobs and Davis 1983; Zheng et al. 2005). Further, the general over- or underrepresentation of these classes of cis regulators in rhizome-derived genes begins to provide a signature that might be used to identify other genes likely to be (or have once been) expressed in rhizomes, and provide a foundation for building hypotheses about the possible functional significance of these elements in regulating gene expression in rhizomes. Alignments of this relatively small gene set suggested that highly expressed rhizome genes were somewhat enriched in QTL likelihood intervals associated with early studies of rhizomatousness or ratooning. For example, we (Jang et al. 2006) found a total of twenty-nine (6.4%) high RT/MR and/or RT/A genes to be within QTL likelihood intervals for rhizome QTLs, vs. 3.7% of random RT/MR and/or RT/A genes. These data were simply too limited to try to draw conclusions about the possibility that there was significance of specific gene functional groups (i.e. specific GO groups or gene sharing common Pfam domains), although some tantalizing observations are beginning to be supported by new data. For example, MADS box transcription factors are a gene family associated with the development of both tubers and rhizomes in other taxa (Kim et al. 2002; Skipper 2002). Three MADS box transcription factors found in the QTL likelihood intervals are now known to also be expressed in the rhizomes of both ginger (Zingiber) and turmeric (Curcuma), thanks to a recent deposit in Genbank of rhizome ESTs for these species (38,139 and 12,593, respectively), largely from an NSF Plant Genome Research project (ag.arizona.edu/research/ganglab/ArREST.htm).

What Is The Fate Of “Rhizome-specific” Genes In Plant Genotypes That Have Lost The Ability To Produce Rhizomes?

As exemplified by S. bicolor (above), many crops have rhizomatous relatives and rhizomatousness appears to have been eliminated genetically from their gene pools, in some cases relatively

RHIZOMATOUSNESS: GENES FOR A WEEDINESS SYNDROME

107

recently. The evolutionary fate in non-rhizomatous genotypes of genes that formerly contributed to rhizome development is interesting both from basic and practical standpoints. From a basic standpoint, rhizomes are an excellent example of a case of organ loss that can be genetically manipulated, perhaps shedding light on basic principles that may apply to understanding the evolution of morphology of other organisms, for example the fate of tail-specific genes in humans. From an applied standpoint, better understanding of the fates of rhizome-specific genes would shed some light on alternative models for how rhizomatousness was lost. For example, the elimination of rhizomes by the progressive shutdown of many genes may show a very different signature in its impact on variation of gene sequences between rhizomatous and non-rhizomatous genotypes than an abrupt macro-mutation in one or a small number of genes. Several lines of evidence indicated that an abrupt macro-mutation in one or a small number of regulatory genes was responsible for other striking morphological modifications during crop domestication. For instance, selection in the regulatory region of the teosinte branched1 gene appears to have contributed substantially to the transformation of maize from the inflorescence morphology of the wild grass teosinte, with long branches with tassels (Wang et al. 1999; Clark et al. 2006), to the short branches typical of cultivated maize. Recently, Li et al. (2006) reported that reduced shattering of the mature inflorescence associated with rice domestication was caused in part by human selection of an amino acid substitution in the DNA binding domain of the sh4 gene. Rhz2 and Rhz3 (Hu et al. 2003) might be targets for such macro-mutations affecting rhizomatousness. However, little is known about the fate of many genes involved in the molecular pathway after the genotype has lost the ability to produce rhizomes. Recently, we (C. Jang and A.H. Paterson, unpublished data) have fully sequenced the coding and upstream (about 1 kb) regions of twenty-four genes shown to have rhizome-enriched expression (Jang et al. 2006) in both S. bicolor and S. propinquum. We found no propensity for mutation in coding regions (e.g., premature stop codon) in the non-rhizomatousness genotype as compared to the rhizomatousness genotype, and indeed found the genes to generally continue to be under purifying selection in non-rhizomatous S. bicolor. This suggests that many of the rhizome-enriched genes may serve multiple functions during growth and development of plants, even if some of these functions are not in rhizomes. Some differences were found, however, in the regulatory features, suggesting an alternative hypothesis that the genes have been “recruited” from their rhizome-specific expression in S. propinquum to serve other purposes in S. bicolor. This area of investigation will benefit greatly from the new availability of the completed sorghum sequence and the prospect of substantial “re-sequencing” efforts in additional germplasm.

Future Work And Potential Applications

A particularly attractive goal that may be facilitated by genomics is identifying and implementing new approaches to regulate growth and development of rhizomes. Identification of genes, regulatory elements, and biochemical functions that are important to rhizome development but dispensable to other plant processes would provide a foundation to search for plant growth regulators that specifically target rhizomes. Such targeting of growth regulation to rhizomes might provide for control of rhizomatous weeds, even in closely related crops, for example Johnsongrass (S. halepense) genes in sorghum. Success in cloning genes responsible for rhizomatousness would provide a point of entry into developmental pathways which might be regulated by endogenous (genetic) or exogenous

108

WEEDY AND INVASIVE PLANT GENOMICS

(chemical) means. For example, while Johnsongrass can be chemically controlled in crop fields, new infestations quickly occur from roadside or other nearby populations. Non-specific eradication of such populations, along with companion grasses, could cause massive erosion. Down regulation of rhizome production might afford integrated management strategies to selectively impede the spread of Johnsongrass, while leaving companion populations intact to ensure erosion control. Better understanding of rhizome development may also benefit improvement of turf, forage, and biomass grasses. Enhancement of rhizomatousness in grasses currently grown on more than 60 million acres in the southern U.S. alone (see above) might afford substantial gains in productivity, as well as improved erosion control. The prospect of expanded importance of rhizomatous grasses in contributing to biomass/biofeedstock production, and development of perennial plant genotypes optimal for meeting this goal on marginal lands, will only increase the importance of better understanding the genetic and developmental control of perennation organs such as rhizomes.

Invasion Biology And Transgene Escape

As detailed above, prior research has demonstrated a high level of gene flow from cultivated sorghum to Johnsongrass—indeed, nearly all members of Johnsongrass populations adjacent to long-term sorghum production fields contain chromatin introgressed from sorghum. Some introgressions spread over long distances, to regions where sorghum is never known to have been grown (Morrell et al. 2005). Concern about transgene escape into Johnsongrass (itself an invasive species) has confined genetic modification of sorghum to experimental purposes, denying society the many potential environmental benefits of genetically modified crops. A promising approach by which this limitation might be overcome is to develop technologies and/or strategies that reduce the risk of gene escape (Gressel 1999; Kuvshinov et al. 2001), applicable to a wide range of transgenes and mitigating the need to make (extremely difficult) assessments of escape probabilities or ecological impacts of individual transgenes. Several approaches have been suggested for mitigation of crop-weed gene flow. Isolation of transgenic sorghum by physical distance is impractical in view of the close physical proximity of naturalized Johnsongrass and the fact that Johnsongrass is a major contaminant of sorghum seed lots. Approaches such as “terminator” genes (Masood 1998) or specialized constructs that prevent flowering (Kuvshinov et al. 2001) risk strong public opposition, particularly in regions where farmers produce their own seed for the next season. The strategy of linking a transgene to flanking “domestication genes,” i.e. mutations that confer phenotypes which are adaptive in agriculture but generally not in the wild (Gressel 1999) is an attractive possibility for mitigating crop-weed gene flow. Few domestication genes have been cloned (Doebley et al. 2006), and none are from sorghum. However, the supply of domestication genes will increase with progress in functional genomics of botanical models generally, and particularly in sorghum as a result of the genome sequence and its amenability to association genetics approaches (Hamblin et al. 2005). Additivity or dominance is a necessity for use of domestication genes to mitigate transgene escape, since the escaped construct will invariably be heterozygous (Gressel 1999). The domesticated alleles of the Rhz2 and Rhz3 loci are genetically recessive (Hu et al. 2003), as are many of the sorghum rhizomatousness QTLs (Paterson et al. 1995c). However, with a few key genes in hand, one could envision producing RNAi-based constructs that would interfere with expression of key rhizome-specific genes in a dominant manner.

RHIZOMATOUSNESS: GENES FOR A WEEDINESS SYNDROME

109

Polyploid Evolution

The nature of tetraploid S. halepense raises a question as to the relative roles of alleles from its probable diploid progenitors, S. propinquum and S. bicolor, in the growth and development of Johnsongrass rhizomes. Johnsongrass (S. halepense), with nearly worldwide distribution, is arguably more invasive than S. propinquum. Could this be partly due to recruitment of S. bicolor genes into rhizome development? As noted above, hybridization with cultivated S. bicolor has been proposed as a potential cause of increased aggressiveness in the weed (see e.g., de Wet and Harlan 1975; Holm et al. 1977). A host of studies report polyploidy to have severe effects on expression of duplicate genes with the diverse patterns including silencing and up- or down-regulation of one of the duplicates (Adams and Wendel 2005). For example, silencing or unequal expression of one of homoeologs was documented for at least 25% genes duplicated by polyploidy in the organ of Gossypium hirsutum (Adams et al. 2003). Extensive analysis of the subgenomic distribution of QTLs in cotton (Jiang et al. 1998; Rong et al. 2007) shows that diploid progenitors (such as S. bicolor) lacking a phenotype (such as rhizomatousness) can still contribute to enhancement of the phenotype in polyploids. The availability of massively parallel re-sequencing technologies sets the stage for detailed comparisons of the Johnsongrass and S. propinquum transcriptomes to the S. bicolor sequence to gain insight into this possibility. One could envision a host of questions being answered by this approach, including the relative abundance of S. propinquum and S. bicolor transcripts (both generally and in specific biochemical pathways), altered patterns of regulation of these transcripts in Johnsongrass relative to those of their diploid progenitors, patterns of coexpression and interaction among S. propinquum and S. bicolor transcripts, and even the possibility that S. propinquum and S. bicolor alleles have recombined to form novel alleles in Johnsongrass. Synthesis

Better understanding of rhizome biology offers opportunities ranging from fundamental insights into organic evolution to applied crop improvement and environmentally-benign avenues for weed control. While rhizome biology has been underexplored, the completed sequence of S. bicolor opens new doors into study of the genetics and development of this trait in a genus (Sorghum) that is both a botanical model for rhizomatousness and an important subject for application of new knowledge about this trait. References Adams KL, Cronn R, Percifield R, Wendel JF (2003) Genes duplicated by polyploidy show unequal contributions to the transcriptome and organ-specific reciprocal silencing. Proceedings of the National Academy of Sciences of the United States of America 100, 4649–4654. Adams KL, Wendel JF (2005) Novel patterns of gene expression in polyploid plants. Trends in Genetics 21, 539–543. Anderson LE, Appleby AP, Weseloh JW (1960) Characteristics of Johnsongrass rhizomes. Weeds 8, 402–406. Arriola PE, Ellstrand NC (1996) Crop-to-weed gene flow in the genus Sorghum (Poaceae): spontaneous interspecific hybridization between johnsongrass, Sorghum halepense, and crop sorghum, S. bicolor. American Journal of Botany 83, 1153–1160. Arriola PE, Ellstrand NC (1997) Fitness of interspecific hybrids in the genus Sorghum: Persistence of crop genes in wild populations. Ecological Applications 7, 512–518. Barnes RF, Baylor JE (1995) Forages in a changing world. In: Forages: An Introduction to Grassland Agriculture Barnes RF, Miller DA, Nelson CJ, eds. pp. 3–13. Iowa State University Press, Ames, IA.

110

WEEDY AND INVASIVE PLANT GENOMICS

Bird LJ, Akhurst RJ (2007) Effects of host plant species on fitness costs of Bt resistance in Helicoverpa armigera (Lepidoptera : Noctuidae). Biological Control 40, 196–203. Burton GW (1989) Progress and benefits to humanity from breeding warm-season forage grasses. In: Contributions from Breeding Forages and Turf Grasses, Sleper DA, Asay KH, Pedersen JF, eds. pp. 21–29. Crop Science Society of America, Madison, WI. Carvalho CHS, Zehr UB, Gunaratna N, et al. (2004) Agrobacterium-mediated transformation of sorghum: factors that affect transformation efficiency. Genetics and Molecular Biology 27, 259–269. Casas AM, Kononowicz AK, Zehr UB, et al. (1993) Transgenic sorghum plants via microprojectile bombardment. Proceedings of the National Academy of Sciences of the United States of America 90, 11212–11216. Celarier RP (1958) Cytotaxonomic notes on the subsection Halepensia of the genus. Sorghum. Bull. Torr. Bot. Cl. 85, 49–62. Cheng M, Lowe BA, Spencer TM, Ye XD, Armstrong CL (2004) Factors influencing Agrobacterium-mediated transformation of monocotyledonous species. In Vitro Cellular and Developmental Biology-Plant 40, 31–45. Clark RT, Nussbaum-Wagler T, Quijada P, Doebley J (2006) A distant upstream enhancer at the maize domestication gene, tb1, has pleiotropic effects on plant and inflorescent architecture. Nature Genetics 38, 594–597. Cox TS, Bender M, Picone C, et al. (2002) Breeding perennial grain crops. Critical Reviews in Plant Sciences 21, 59–91. de Wet JMJ, Harlan JR (1975) Weeds and domesticates: evolution in the man-made habitat. Economic Botany 29, 99–107. Doebley JF, Gaut BS, Smith BD (2006) The molecular genetics of crop domestication. Cell 127, 1309–1321. Doggett H (1976). Sorghum. In: Evolution of Crop Plants, Simmonds NW, ed. pp. 112–117. Longman, Essex, UK. Feltus FA, Wan J, Schulze SR, Estill JC, Jiang N, Paterson AH (2004) An SNP resource for rice genetics and breeding based on subspecies Indica and Japonica genome alignments. Genome Research 14, 1812–1819. Field CB (2001) Sharing the garden. Science 294, 2490–2491. Gantzer CJ, Anderson SH, Thompson AL, Brown JR (1990) Estimating soil erosion after 100 years of cropping on Sanborn Field. Journal of Soil and Water Conservation 45, 641–644. Gao ZS, Jayaraj J, Muthukrishnan S, Claflin L, Liang GH (2005a) Efficient genetic transformation of Sorghum using a visual screening marker. Genome 48, 321–333. Gao ZS, Xie XJ, Ling Y, Muthukrishnan S, Liang GH (2005b) Agrobacterium tumefaciens-mediated sorghum transformation using a mannose selection system. Plant Biotechnology Journal 3, 591–599. Girijashankar V, Sharma HC, Sharma KK, Swathisree V, Prasad LS, Bhat BV, Royer M, San Segundo B, Narasu ML, Altosaar I, Seetharama N (2005) Development of transgenic sorghum for insect resistance against the spotted stem borer (Chilo partellus). Plant Cell Reports 24, 513–522. Gizmawy I, Kigel J, Koller D, Ofir M (1985) Initiation, orientation, and early development of primary rhizomes in Sorghum halepense (L.) Pers. Annals of Botany 55, 343. Gray SJ, Zhang S, Rathus C, Lemaux PG, Godwin ID (2004) Development of sorghum transformation: Organogenic regeneration and gene transfer methods. In: Sorghum Tissue Culture and Transformation, Seetharma N, Goodwin I, eds. pp. 35–43 Science Publishers, Enfield, NH. Gressel J (1999) Tandem constructs: preventing the rise of superweeds. Trends in Biotechnology 17, 361–366. Hamblin MT, Fernandez MGS, Casa AM, Mitchell SE, Paterson AH, Kresovich S (2005) Equilibrium processes cannot explain high levels of short- and medium-range linkage disequilibrium in the domesticated grass Sorghum bicolor. Genetics 171, 1247–1256. Holm LG, Plucknett DL, Pancho JV, Herberger JP (1977) Sorghum halepense (L.) Pers. In: The World’s Worst Weeds: Distribution and Biology, pp. 54–61. University Press of Hawaii, Honolulu, Hawaii. Howe A, Sato S, Dweikat I, Fromm M, Clemente T (2006) Rapid and reproducible Agrobacterium-mediated transformation of sorghum. Plant Cell Reports 25, 784–791. Hu FY, Tao DY, Sacks E, Fu BY, Li J, Yang Y, McNally K, Khush GS, Paterson AH, Li ZK (2003) Convergent evolution of perenniality in rice and sorghum. Proceedings of the National Academy of Sciences of the United States of America 100, 4050–4054. Jacobs WP, Davis W (1983) Effects of gibberellic-acid on the rhizome and rhizoids of the algal coenocyte, Caulerpa prolifera, in culture. Annals of Botany 52, 39–41. Jang CS, Kamps TL, Skinner DN, Schulze SR, Vencill WK, Paterson AH (2006) Functional classification, genomic organization, putative cis-acting regulatory elements, and relationship to QTLs of Sorghum genes with rhizome-enriched expression. Plant Physiology 142, 1148–1159. Jeoung JM, Krishnaveni S, Jayaraj J (2004) Agrobacterium-mediated transformation of grain sorghum. In: Sorghum Tissue Culture and Transformation, Seetharma N, Goodwin I, eds. pp. 57–64. Science Publishers, Enfield, NH.

RHIZOMATOUSNESS: GENES FOR A WEEDINESS SYNDROME

111

Jiang CX, Wright RJ, El-Zik KM, Paterson AH (1998) Polyploid formation created unique avenues for response to selection in Gossypium (cotton). Proceedings of the National Academy of Sciences of the United States of America 95, 4419–4424. Kim SH, Mizuno K, Fujimura T (2002) Isolation of MADS-box genes from sweet potato (Ipomoea batatas (L.) lam.) expressed specifically in vegetative tissues. Plant and Cell Physiology 43, 314–322. Krishnaven S, Jeoung J, Zhu H, Muthukrishnan S, Liang GH (2004) Biolistic transformation of sorghum and influence of a transgenic chitinase gene. In: Sorghum Tissue Culture and Transformation, Seetharma N, Goodwin I, eds. pp. 65–74. Science Publishers, Enfield, NH. Kuvshinov V, Koivu K, Kanerva A, Pehu E (2001) Molecular control of transgene escape from geentically modified plants. Plant Science 160, 517–522. Lander E, Botstein D (1989) Mapping Mendelian factors underlying quantitative traits using RFLP linkage maps. Genetics 121, 185–199. Li C, Zhou A, Sang T (2006) Rice domestication by reduced shattering. Science, 311, 1936–1939. Masood E (1998) Monsanto to back down over “terminator” gene? Nature 396, 503. McWhorter CG (1961) Morphology and development of Johnsongrass plants from seeds and rhizomes. Weeds 9, 558–562. McWhorter CG (1971) Introduction and spread of Johnsongrass in the United States. Weed Science 19, 496. McWhorter CG, Hartwig EE (1972) Competition of johnsongrass and cocklebur with six soybean varieties. Weed Science 20, 56–59. Millhollen RW (1970) MSMA for johnsongrass control in sugarcane. Weed Science 18, 333. Morrell PL, Williams-Coplin D, Bowers JE, Chandler JM, Paterson AH (2005) Crop-to-weed introgression has impacted allelic composition of johnsongrass populations with and without recent exposure to cultivated sorghum. Molecular Ecology 14, 2143–2154. Mythili PK, Rani TS, Sairam RV (2004) Sorghum tissue culture and transformation research. In: Sorghum Tissue Culture and Transformation, Seetharma N, Goodwin I, eds. pp. 51–56. Science Publishers, Enfield, NH. Nguyen TV, Thu TT, Claeys M, Angenon G (2007) Agrobacterium-mediated transformation of sorghum (Sorghum bicolor (L.) Moench) using an improved in vitro regeneration system. Plant Cell Tissue and Organ Culture 91, 155–164. Oyer E, Gries G, Rogers B (1959a) The seasonal reproduction of Johnson Grass plants. Weeds 7, 13. Oyer EB, Gries GA, Rogers BJ (1959b) The seasonal reproduction of Johnson Grass plants. Weeds 7, 13–19. Paterson AH, Bowers JE, Chapman BA (2004) Ancient polyploidization predating divergence of the cereals, and its consequences for comparative genomics. Proceedings of the National Academy of Sciences of the United States of America 101, 9903–9908. Paterson AH, et al. (about 50 authors) (2009) The Sorghum bicolor genome and the diversification of grasses. Nature 457, 551–556. Paterson AH, Lin YR, Li ZK, Schertz KF, Doebley JF, Pinson SRM, Liu S-C, Stansel JW, Irvine JE (1995a) Convergent domestication of cereal crops by independent mutations at corresponding genetic loci. Science 269, 1714–1718. Paterson AH, Schertz KF, Lin Y-R, Liu SC, Chang YL (1995b) The weediness of wild plants: molecular analysis of genes influencing dispersal and persistence of johnsongrass, Sorghum halapense (L.) Pers. Proceeding of the National Academy of Sciences of the United States of America 92, 6127–6131. Paterson AH, Schertz KF, Lin YR, Liu SC, Chang YL (1995c) The weediness of wild plants—molecular analysis of genes influencing dispersal and persistence of Johnsongrass, Sorghum halepense (L) Pers. Proceedings of the National Academy of Sciences of the United States of America 92, 6127–6131. Pimentel D, Harvey C, Resosudarmo P, Sinclair K, Kurz D, McNair M, Crist S, Shpritz L, Fitton L, Saffouri R, Blair R (1995) Environmental and economic costs of soil erosion and conservation benefits. Science 267, 1117–1123. Pratt LH, Liang C, Shah M, et al. (2005) Sorghum expressed sequence tags identify signature genes for drought, pathogenesis, and skotomorphogenesis from a milestone set of 16,801 unique transcripts. Plant Physiology 139, 869–884. Randall GW, Mulla D (2001) Nitrate nitrogen in surface waters as influenced by climatic conditions and agricultural practices. Journal of Environmental Quality 30, 337–344. Rathus C, Nguyen T, Able JA, Gray SJ, Godwin ID (2004) Optimizing sorghum transformation technology via somatic embryogenesis. In: Sorghum Tissue Culture and Transformation, Seetharma N, Goodwin I, eds. pp. 25–34. Science Publishers, Enfield, NH. Rong J-K, Feltus FA, Waghmare VN, et al. (2007) Meta-analysis of polyploid cotton QTLs shows unequal contributions of subgenomes to a complex network of genes and gene clusters implicated in lint fiber development. Genetics 176, 2577–2588. Scheinost P (2001) Perennial wheat: a sustainable cropping system for the Pacific Northwest. American Journal of Alternative Agriculture 16, 147–151.

112

WEEDY AND INVASIVE PLANT GENOMICS

Skipper M (2002) Genes from the APETALA3 and PISTILLATA lineages are expressed in developing vascular bundles of the tuberous rhizome, flowering stem and flower primordia of Eranthis hyemalis. Annals of Botany 89, 83–88. Sticklen MB, Oraby HF (2005) Shoot apical meristem: A sustainable explant for genetic transformation of cereal crops. In Vitro Cellular and Developmental Biology-Plant 41, 187–200. Tadesse Y, Jacobs M (2004) Nutritional quality improvement of sorghum through genetic transformation. In: Sorghum Tissue Culture and Transformation, Seetharma N, Goodwin I, eds. pp. 81–83. Science Publishers, Enfield, NH. Tang H, Liang GH (1988) The genomic relationship between cultivated sorghum [Sorghum bicolor (L) Moench] and johnsongrass [Sorghum halepense (L) Pers]—a reevaluation. Theoretical and Applied Genetics 76, 277–284. Wagoner P (1990) Perennial grain development—past efforts and potential for the future. Critical Reviews in Plant Sciences 9, 381–408. Wang RL, Stec A, Hey J, Lukens L, Doebley J (1999) The limits of selection during maize domestication. Nature 398, 236–239. Wang WQ, Wang JX, Yang CP, Li Y, Liu L, Xu J (2007) Pollen-mediated transformation of Sorghum bicolor plants. Biotechnology and Applied Biochemistry 48, 79–83. Warwick SI, Black LD (1983) The biology of Canadian weeds. 61. Sorghum halepense (L.) Pers. Canadian Journal of Plant Science 63, 997–1014. Williams SB, Gray SJ, Laidlaw HKC, Godwin ID (2004) Particle inflow gun-mediated transformation of Sorghum bicolor. In: Transgenic Crops of the World: Essential Protocols, Curtis IS, ed. pp.89–102. Kluwer Academic, Dordrecht. Yu CKY, Springob K, Schmidt JR, Nicholson RL, Chu IK, Yip WK, Lo C (2005) A stilbene synthase gene (SbSTS1) is involved in host and nonhost defense responses in Sorghum. Plant Physiology 138, 393–401. Zhao ZY (2006) Sorghum (Sorghum bicolor L.). Agrobacterium Protocols, Second Edition, Vol 1 343, 233–244. Zheng CS, Zheng XS, Ohno H, Hara T, Matsui S (2005) Involvement of ethylene and gibberellin in the development of rhizomes and rhizome like shoots in Oriental cymbidium hybrids. Journal of the Japanese Society for Horticultural Science 74, 306–310.

8

Leafy Spurge: An Emerging Model To Study Traits Of Perennial Weeds David P. Horvath and James V. Anderson

Introduction

Weeds contain inherent genetic traits that give them remarkable plasticity, allowing them to adapt, regenerate, survive, and thrive in a multitude of ecosystems (Chao et al. 2005). Many weeds are capable of vegetative regeneration from tissue which can arise spontaneously from root or stem tissues following tilling or destruction of the aerial portion of the plant. Other plants maintain a ready supply of vegetative propogules in the form of shoot buds on roots or stems that can rapidly produce a new shoot following damage to the plant (Anderson et al. 2001; Klimesˇové and Martínková 2004). In such weeds, any control measure that fails to kill the entire vegetative structure, or worse, just break up the vegetative structure, often results in growth of more shoots than were originally present. Many perennial weeds pose a special problem for weed control because of their ability to readily regenerate from vegetative tissues after prolonged absence of growth. They often support a bud bank which, like a seed bank, can produce vegetative propogules years following the initial attempt to control a given infestation. This chapter will review the various mechanisms by which weeds regenerate new shoots and how growth and dormancy is regulated in buds, and we will examine a case study on bud dormancy analysis using a model perennial weed, leafy spurge (Euphorbia esula). The genes and signals regulating these processes might eventually be exploited not only to develop novel weed control strategies, but to also enhance crop production by capitalizing on the traits that make weeds so competitive.

Regulation Of Shoot Development And Growth Shoot Architecture Overview

All weeds must develop a shoot apical meristem (SAM) to grow and reproduce. Some weeds are capable of forming SAMs de novo from roots, stems, and even leaves following tilling or destruction of the existing SAMs by herbicide treatment (Klimesˇové and Martínková 2004). Other plants produce and maintain fully developed but dormant SAMs in leaf axils (axillary buds) or other portions of the plant (adventitious buds). Because SAM development and growth in seeds and buds is required for weed survival and proliferation, the genes required for SAM initiation and proliferation are potential targets for future development of novel weed control measures. In higher plants, the SAM is composed of several well defined groups of cells, often referred to as the central, peripheral, and rib zones (reviewed by Barton 1998; Lenhard and Laux 1999). These three cell zones communicate with each other to ensure sustainable proliferation of cells required for the growing shoot. In vegetative meristems, lateral organs (leaves) are derived from the peripheral zone. Stem tissues are derived from the rib zone. An undifferentiated pool 113

114

WEEDY AND INVASIVE PLANT GENOMICS

Peripheral zone

CLV1-3 CUC

CUC

AS1

AS1

ANT ANT WUS STM

UFO

Central zone

Rib zone

Figure 8.1. Drawing of a typical shoot meristem with tissue-specific expression and function of key regulatory genes noted.

of cells from the central zone is maintained to replace cells in the peripheral and rib zones that are lost when they are incorporated into differentiating organs. Communication and boundary establishment among these cell zones are facilitated by expression of specific signaling proteins, the most well characterized of which include SHOOTMERISTEMLESS (STM), WUSCHEL (WUS), CLAVATA 1 through 3 (CLV1-3), CUP-SHAPED COTYLEDON 1 through 3 (CUC1-3), AINTEGUMENTA (ANT), ASYMMETRIC LEAVES1 (AS1), and UNUSUAL FLORAL ORGANS (UFO) (Sharma and Fletcher 2002). (Figure 8.1) Mutational analysis of genes encoding these signaling proteins has revealed critical roles in establishing and maintaining the identity of cells in zones where they are expressed. These proteins are also involved in communicating the size and activity of their perspective zones relative to the central zone. In addition, expression of the genes encoding these signaling proteins early in embryo development can serve as functional markers for formation of the distinct regions of the SAM (reviewed by Bowman and Eshed 2000). A review of the literature provides evidence for a simplified model indicating that STM is required for maintaining WUS expression (although WUS appears to be expressed prior to STM in the developing embryo and in differentiating callus [Cary et al. 2002]), and both of these genes act to promote maintenance of the central zone. The size of the central zone is limited by CLV1-3 which negatively regulates WUS (Brand et al. 2000, Schoof et al. 2000) and establishes the boundary between the central zone and the peripheral zone. The boundary between the central and peripheral zones also requires the involvement of CUC1-3 (Vroemen et al. 2003) which acts upstream of STM and WUS (Cary et al. 2002). Although CUC expression potentiates STM expression, STM activity appears to define the limits of CUC expression (Aida et al. 1999). AS1 expression is limited by STM and vice versa, thus, AS1 expression is inhibited in the central zone and STM expression is inhibited in the peripheral zone. AS1 also positively regulates ANT expression which is required for initiation of leaf primordia (Elliott et al. 1996), and for maintaining cell division in the developing organs (Mizukami and Fischer 2000). ANT may also be involved in communicating with the central zone to recruit additional cells to the peripheral zone (Bowman and Eshed 2000). Less is known about the signals involved with control and

LEAFY SPURGE: A MODEL TO STUDY PERENNIAL WEED TRAITS

115

development of the rib zone, but the gene UFO has an expression pattern which suggests it may play some role in this process (Bowman and Eshed 2000). In addition to the key genes described above, numerous other genes influence the initiation and development of the SAM. Members of the YABBY gene family, implicated in development of leaf structure, also impact (directly or indirectly) the maintenance of the central zone (Kumaran et al. 2002). Consequently, there appear to be numerous genes with various characterized functions that can influence SAM formation and growth. Most notable among these are genes such as CYCLIN D-(1-4) and other cell cycle regulatory genes (Gaudin et al. 2000).

Initiation Of Adventitious And Axillary Buds

Although the genes that regulate SAM function and formation are fairly well characterized, little is known about the role, if any, these genes might play in initiation or development of shoots. Experimental evidence so far has focused mainly on the development of axillary buds. Some very elegant experiments indicate that a portion of central zone cells in the SAM of Arabidopsis plants are maintained as undifferentiated cells at the base of the developing leaf primordia (Ward and Leyser 2004). These cells appear to be required to form a core for development of axillary buds. Mutations in genes such as LATERALSUPPRESSOR (LS), a VHIID family protein in tomato, inhibit formation of SAMs in the leaf axils of tomato, as do mutations in an MYB family transcription factor named BLIND. Regardless, little is still known about the exact mechanisms by which this occurs (Schumacher et al. 1999). However, there is some indication that LS might be required for maintaining expression of STM in cells at the base of the leaf until the meristem is reinitiated (Ward and Leyser 2004). Interestingly, LS appears to be more important in Arabidopsis axillary bud formation late in development when axillary buds form farther away from the SAM (Greb et al. 2003). Substantially less is known about the development of adventitious buds. Experimental evidence with poplar has shown that overexpression of ARBORKNOX1, a putative orthologue of STM, resulted in formation of multiple adventitious and axillary buds (Groover et al. 2007). Likewise, ectopic expression of STM in Arabidopsis leaves resulted in meristem-like structures that formed along the leaf periphery (Gallois et al. 2002). The perennial weed leafy spurge (Euphorbia esula L.) develops adventitious buds on its roots and crown. STM expression was observed in leafy spurge root and hypocotyl tissue, including root tissue that was devoid of any visible buds (Varanasi et al. unpublished data). Sequence from the promoter of a putative STM orthologue from leafy spurge was attached to a reporter gene (GUS) and this construct was reinserted back into the leafy spurge genome. In the resulting transgenic plants, GUS expression was observed in root and hypocotyl tissues (Varanasi et al. 2008). Interestingly, the same construct, when transformed into Arabidopsis, also caused the reporter gene to be expressed in Arabidopsis roots and hypocotyl. These results were interpreted to mean that expression of STM is required in root and hypocotyl tissues during or prior to adventitious bud formation in leafy spurge. Many weeds such as dandelions or Canada thistle develop SAMs at the site of injury. Thus, it would be interesting to follow expression of STM or other SAM developmental genes under conditions that lead to de novo shoot development. Additional work determining the impact that mutations to genes such LS or BLIND have on adventitious bud development will greatly increase our understanding of how these vegetative propogules form. More importantly for the weed community, understanding the mechanisms and components that regulate these processes could lead to novel weed control strategies.

116

WEEDY AND INVASIVE PLANT GENOMICS

Regulation Of Bud Dormancy Signals Regulating Shoot Growth

Once buds are formed, additional signaling pathways and genes regulate their further development and growth. During the active growing season, axillary and adventitious buds of intact plants are inhibited from differentiating into new shoots by mechanisms associated with apical dominance, and buds are considered to be in a state referred to as paradormancy. Auxin produced in the growing shoot apices prevents outgrowth of more distal buds (Cline 1991; Chao et al. 2007b). The genes, hormones, and mechanisms involved in paradormancy have been well characterized in Arabidopsis, pea, and petunia (Snowden et al. 2005; Beveridge 2006; Johnson et al. 2006; Ongaro and Leyser et al. 2008). Essentially, the auxin signal is transported basipetally from the growing apices primarily through the action of the polar auxin transport protein PIN1. Auxin transport is perceived by an unknown mechanism and two CAROTENOID CLEAVAGE DIOXYGENASE proteins (designated as MAX3 and 4 in Arabidopsis, DAD1 and DAD3 in petunia, and RMS1 and RMS5 in pea) are activated or induced. These CAROTENOID CLEAVAGE DIOXYGENASE proteins produce a novel carotenoid-based hormone. This hormone is likely further modified by a particular P450 (MAX1 or RMS3/4 in Arabidopsis and pea respectively) and is transported and/or perceived in the bud by a specific member of an F-box protein degradation complex (MAX2). Studies on hormone action involving gibberellic acid and auxin, which also act through an F-box, indicate that the perception of carotenoids likely results in specific degradation of growth inhibitors that then allows bud outgrowth. The hormones ABA and cytokinin also play a role in regulating bud outgrowth (Cline et al. 1991). Cytokinin appears to be made in the stem internode and moves into the buds upon paradormancy release (Shimizu-Sato and Mori 2001). High auxin levels inhibit cytokinin accumulation in the transpiration stream (Eklof et al. 1997; Nordstrom et al. 2004). ABA induces ICK1, a known inhibitor of cell division (Wang et al. 1998), and also accumulates in the internode adjacent to paradormant buds (Shimizu-Sato and Mori 2001).

Seasonal Bud Dormancy

Perennials growing in temperate climates face a special problem in the fall. Freezing temperatures and/or short day lengths cause the senescence and death of growing shoots, thus removing the inhibiting basipetal auxin signal. However, intermittent growth-conducive conditions can still occur in the late fall, after shoot senescence and death. Such warm spells could cause induction of bud growth, and the growing buds would soon be killed by subsequent and inevitable return of freezing temperatures. Not only would such untimely growth result in reduction of vegetative prologues, but nutrient reserves needed for over-wintering and shoot establishment from buds the following spring would be squandered. As a result, many temperate perennials have evolved mechanisms to prevent bud outgrowth during the fall when bud growth is no longer inhibited by paradormancy-maintaining mechanisms. This transitional dormant state prevents bud out-growth during intermittent growth-conducive conditions typically experienced in the fall. The transitional dormant state preventing bud growth in the fall is often referred to as endodormancy because physiological signaling processes within the bud itself inhibit it from differentiating into new aerial shoots. Environmental signals that bring about endodormancy, primarily shortening day lengths and cooling temperatures, also regulate flowering. Thus, it has been hypothesized that there might

LEAFY SPURGE: A MODEL TO STUDY PERENNIAL WEED TRAITS

117

be overlap between the mechanisms regulating flowering (vernalization) and those regulating endodormancy (Chouard 1960; Horvath et al. 2003). This was proven to be the case when a key floral regulatory gene FLOWERING TIME LOCUS T (FT) was over-expressed in poplar and caused not only early flowering, but also prevented growth cessation and bud set (Böhlenius et al. 2006; Hsu et al. 2006). Other indications of overlap in floral and endodormancy regulation come from studies of the evergrowing (evg) mutation in peach. This mutation prevents peach buds from entering endodormancy in the fall. When the locus containing the evg mutation was cloned and sequenced, it was found to contain a series of MADS-box transcription factor genes related to a small family of MADS-box transcription factors including SHORT VEGETATIVE PHASE (SVP) and AGAMOUS-LIKE 24 (AGL24) (Bielenberg et al. 2004). SVP and AGL24 have been previously characterized in Arabidopsis and shown to alter flowering time. Interestingly, SVP acts by binding to the promoter of FT and inhibiting its expression (Lee et al. 2007). Microarray analysis of dormancy transitions in leafy spurge, potato, raspberry, and poplar have all indicated similar DORMANCY ASSOCIATED MADS-box (DAM) genes being up-regulated during endodormancy induction or down-regulated upon endodormancy release (Horvath et al. 2008). Combined, the experimental evidence and observations indicate that many of the genes, signaling pathways, and physiological processes regulating bud dormancy and growth are well conserved among diverse perennial species. Consequently, what is learned from these model species is quite likely to be applicable to weeds. Indeed, leafy spurge, a model species used for bud growth and development studies, is a weed. Many tools have been developed to specifically study growth and development of this weed, and much can be learned concerning the evolution of developmental processes that have evolved in this weed to facilitate its perennial life cycle.

Case Study: Leafy Spurge Weedy Characteristics

Because of its ease of growth, manipulation, scientific support, and existing tools, leafy spurge has been proposed as a model for the study of perennial broadleaf weeds (Chao et al. 2005; see also Chapter 4). Leafy spurge is an invasive perennial weed that causes significant economic losses to range, recreational, and right-of-way lands in North American plains and prairies (Bangsund et al. 1999). Vegetative reproduction from an abundance of underground crown and root buds (adventitious buds located on the underground portion of the stem or along the lateral roots) makes this weed particularly difficult to control. Although seasonal development of crown and root buds occurs from early to mid-summer, these buds are inhibited from initiating new shoot growth by mechanisms regulating well-defined phases of seasonal dormancy (para-, endo-, and ecodormancy) until the following spring (Anderson et al. 2005). In fact, dormancy-imposed inhibition of new shoot growth from underground adventitious buds has long been considered a key characteristic leading to the persistence of perennial weeds such as leafy spurge (Coupland et al. 1955). More specifically, dormancy facilitates the distribution of shoot emergence over time, and is therefore a critical factor that allows weeds such as leafy spurge to escape control by chemical, cultural, mechanical, and biological control measures (Anderson et al. 2001; CAB 2004). In leafy spurge, two separate signals produced by the aerial portion of the plant are capable of maintaining crown and root buds in a paradormant phase during the active growing season

118

WEEDY AND INVASIVE PLANT GENOMICS

(Horvath 1999). These signals involve auxin, likely acting through previously described mechanisms, and sugar, acting through ABA/GA signaling that appear to regulate the G1 to S transition of the cell cycle (Horvath et al. 2003; Chao et al. 2006). In the fall, the crown buds of leafy spurge are known to enter a state of endodormancy (Anderson et al. 2005), which is believed to bridge the gap between the transitions from paradormancy to ecodormancy. Establishment of endodormancy is important for preventing initiation of new shoot growth from crown and root buds during the post-senescence period of fall, when environmental conditions can still promote growth. Without the endodormant phase, new shoot growth from crown or root buds would be killed by fall frost and winter freezes, as previously discussed. However, after an extended period of cold temperatures, crown buds are released from endodormancy and become flowering competent (Anderson et al. 2005). At this time, winter conditions effectively block new shoot growth from the buds by making the transition to the ecodormant phase until spring conditions allow for renewed growth.

Genomics

At least three genomics initiatives have been undertaken to characterize genetic pathways/ networks unique to perennial weeds, using leafy spurge as a model (Anderson and Horvath 2001; Anderson et al. 2007; Jia et al. 2006; Horvath et al. 2005b, 2006, 2008). The most important aspect of these genomics initiatives was the development of leafy spurge EST databases which have served as a source of clones and markers for researchers studying specific physiological responses in leafy spurge and related species (Anderson and Horvath 2001; Anderson and Davis 2004; Anderson et al. 2005, 2007; Chao et al. 2007a, 2007c; Horvath et al. 2002, 2005b). Further development of both low-density (Horvath et al. 2005b, 2006) and high-density (Anderson 2008) cDNA arrays from unigene sets identified within leafy spurge EST databases represent the first known microarrays for a perennial weed species. Experiments conducted using low-density microarrays have illuminated the importance of signals regulating expression of genes associated with transitions in well-defined phases of dormancy (Horvath et al. 2005b, 2006) and have also confirmed the need for high-density arrays. Thus, the development and use of high-density leafy spurge microarrays, containing a 19,015 leafy spurge unigene set (Anderson et al. 2007) as well as an additional set of 4,129 unigenes from cassava (Lokko et al. 2007), a member of the Euphorbiaceae family, has enhanced our understanding of genes and genetic networks associated with traits that make weedy perennials such as leafy spurge so invasive and difficult to control. For example, analysis of genomic sequences corresponding to coordinately regulated genes identified from microarray analysis provides opportunities for detecting shared transcription factor binding sites (Tatematsu et al. 2005). These transcription factor binding sites can be used to identify the transcription factors that bind to them using techniques such as 1-hybrid screening methods (Li and Herskowitz 1993). Transcription factors that regulate numerous genes during transitions in well-defined phases of dormancy likely play important roles in regulating perennial growth patterns.

Transcriptomics

Monitoring the transcriptome (expressed genes; mRNAs) provides a snapshot of the genes and/or pathways responsive to signaling events associated with transitional phases of growth

LEAFY SPURGE: A MODEL TO STUDY PERENNIAL WEED TRAITS

119

and development. Classical methods for monitoring gene expression generally involve northern blot hybridization, semi-quantitative RT-PCR, or real-time PCR in which one to several genes of interest can be monitored at a time. With the availability of a leafy spurge EST database, these classical methods have provided substantial insights into mechanisms involved in well-defined phases of dormancy transition in vegetative buds of leafy spurge. Studies using these classic methods have provided important information on cold hardening, paradormancy to growth transitions, meristem development, sugar metabolism, photomorphogenesis, and transitions between para-, endo-, and eco-dormancy (Anderson and Horvath 2001; Anderson et al. 2005; Chao et al. 2007a; Horvath and Olson 1998; Horvath et al. 2002, 2005a; Varanasi et al. 2008). Although these studies have helped to identify marker genes for tracking transitional states of dormancy and cold hardiness required for perennial growth, randomly selecting the transcripts to monitor is biased and can become a laborious and time-consuming effort. Thus, using microarray analysis to screen for differentially expressed transcripts is more efficient and less biased since bioinformatic analysis, based on statistical analysis, steers researchers to differentially expressed transcripts of significance. Further, cluster analyses of processed array data give visual cues to common patterns or coordinate regulation of expression among transcripts. Initial studies performed using low-density leafy spurge microarrays indicated that loss of paradormancy was associated with the down regulation of genes involved in flavonoid biosynthesis (Horvath et al. 2005b). The transition from paradormancy to endodormancy was paralleled by a down regulation of numerous gibberellic acid-responsive genes, and known cold-regulated genes were up regulated (Horvath et al. 2006). Interestingly, genes involved in photomorphogenesis were also induced during endodormancy and ethylene signaling responses were observed during ecodormancy (Horvath et al. 2006). Further screening of leafy spurge crown bud samples (collected over a five-year period) using high-density microarrays identified 999 genes that showed a significant level of differential expression at some point during seasonal development (Horvath et al. 2008). Studies using high-density arrays also implicated differential expression of genes involved in flavonoid biosynthesis, GA responsiveness, photomorphogenesis, and cold-regulation, similar to that observed using the low-density arrays. Further analysis of the 999 differentially expressed genes based on information linked to designated gene ontology (GO), (MIPs), eukaryotic orthologous genes (KOGs), and MAPMAN implicated the importance of other pathways such as cell cycle, chromatin remodeling, glycolysis, protein and lipid metabolism, redox potential, transport, jasmonic acid responsiveness, and stress (Horvath et al. 2008).

Interactomics

A map of protein:protein interactions (the interactome) can provide clues to regulatory and metabolic processes that likely affect cellular functions (Geisler-Lee et al. 2007). Potential protein interactions and functions associated with individual KOGs were obtained using the string program (http://string.embl.de) and incorporated as an additional method for identifying/ visualizing central nodes and edges of interest related to dormancy transitions in leafy spurge. Interactions of interest identified during the progressive seasonal transitions from para-, endo-, ecodormancy included genes involved in cell cycle and mitosis, transcription, chromosome structure, carbon/protein metabolism, membrane modification, signal transduction, and redox/oxidative stress (Figure 8.2). However, when comparing the transition from para- to

Para > Endo transition-specific S-adenosylhomocysteine hydrolase

KOG1370

Cystathionine beta-synthase

KOG1252

Aspartate aminotransferase

Triosephosphate isomerase Glyceraldehyde 3-phosphate KOG1643 KOG0657 dehydrogenase

KOG1411

Squalene synthetase Cis-prenyltransferase KOG1459 KOG1602

HSP104

Glutamate decarboxylase

KOG1051 KOG2670

KOG1383

Enolase DnaJ

KOG2763

KOG1176

Adenylate kinase

KOG0712

Aldehyde Acyl-CoA synthetase dehydrogenase

Acyl-CoA thioesterase

Up Slight Up Both up and down Slight down Down

Cystathionine beta-lyases

KOG0053

KOG1744

Histone H2B

KOG3078

KOG2450

DnaJ

KOG1756

KOG0714

Reductases

KOG0022

Cytochrome P450 Thioredoxin

Cytochrome P450

Protein kinase PCTAIRE

KOG1329

Phospholipase D1

Glutathione peroxidase

DNA-binding protein jumonji/RBP2/SMCY KOG1246

KOG0419

KOG4197

PPR repeat

Ubiquitin-protein ligaseKOG0118 KOG4155

KOG1282

RRM domain

KOG1339

Serine carboxypeptidases

WD40 repeat Aspartyl protease

GATA-4/5/6 T.F. KOG1601

KOG0513 KOG4232

Ca2+-independent phospholipase A2

Delta 6-fatty aciddesaturase

KOG0724

DnaJ superfamily

KOG0055

KOG0014

MADS box T.F. Beta-glucosidase

KOG0594

KOG1471

SEC14

KOG1651

KOG0626

RRC1 (chromosome condensation)

KOG1429

UDP-G.A. decarboxylase

Reductase

KOG0156

KOG1426

KOG1208

KOG0541

KOG0907

KOG0143

KOG0504

Dehydrogenases

KOG0157

Iron/ascorbate family oxidoreductases

Histone 2A

Ankyrin repeat

KOG0725

KOG0024

Sorbitol dehydrogenase

ABC superfamily

KOG0223

Aquaporin

Endo > Eco transition Up

Iron/ascorbate family oxidoreductases KOG0143

Slight Up Both up and down

Cytochrome P450

Acyl-CoA synthetase

Slight down

KOG0157

KOG1176

Down Cytochrome P450 PPR repeat

KOG2456

KOG0156

Retinoblastoma-related

Aldehyde dehydrogenase

KOG1010

KOG4197

Aldo/keto reductase KOG1502

14-3-3

KOG1577

Flavonol reductase

KOG0048

Transcription factor, Myb superfamily

KOG0841

Predicted transporter KOG0254 Leucine rich repeat proteins

KOG1192

KOG1947

UDP-glucosyl transferase

Serine/threonine protein kinase

Cyclin KOG1674

KOG0698

KOG0653

KOG1187 KOG2987

KOG1339

Serine/threonine protein phosphatase

KOG0504

Ankyrin repeat

Cyclin B kinase-activating proteins KOG0379

Fatty acid desaturase

Kelch repeat-containing proteins

KOG4302

Aspartyl protease

Microtubule-associatedanaphase spindle elongation KOG0118

RRM domain

KOG0055

ABC superfamily KOG0327 KOG0054

ABC superfamily

Translation initiation factor 4F KOG2605

cysteine protease

Figure 8.2. Protein:protein interactome maps related to the transition of leafy spurge crown buds through seasonal dormancy transitions (top: paradormancy to endodormancy; bottom: endodormancy to ecodormancy). KOG numbers represent genes that showed a significant (P value < 0.005) level of differential expression during at least one transitional phase. Only those KOGs showing interactions under high-stringency are presented. Shading of the nodes indicate expression trend during the designated transition.

120

LEAFY SPURGE: A MODEL TO STUDY PERENNIAL WEED TRAITS

121

endodormancy vs. the transition from endo- to ecodormancy, some differences were observed for specific KOGs representing the central nodes and edges of these functional groups. In particular, differences in specific transcription factors, cell cycle–related proteins, chromosome structural proteins, and several growth regulators were observed when comparing the two different seasonal transitions in dormancy. Specific Genes Of Interest

Because FT has been shown to be involved in dormancy induction (Böhlenius et al. 2006) and is negatively regulated by SVP in Arabidopsis (Lee et al. 2007), it was of particular interest that expression of DAM genes were inversely correlated with expression of FT in crown buds of leafy spurge during the seasonal transitions from para- to endo- and ecodormancy (Horvath et al. 2008). This observation is even more intriguing since DAM genes from raspberry, poplar, apricot, and peach have also been implicated in bud dormancy regulation (Bielenberg et al. 2004; Horvath et al. 2008; Yamane et al. 2008). In poplar, some chromatin remodeling genes are significantly up regulated prior to endodormancy induction, and chromatin remodeling has been suggested to play a role in dormancy transitions in cambial meristems of poplar (Ruttink et al. 2007; Druart et al. 2007). Because similarities appear to exist between the mechanisms regulating both flowering and dormancy, it is tempting to speculate that the same mechanisms regulating chromatin remodeling processes and genes regulating vernalization and flowering competence in Arabidopsis also regulate the re-initiation of growth competence in vegetative buds following extended cold treatments. Further evidence for this hypothesis comes from the fact that in both leafy spurge and poplar, an SWI2/SNF2-like gene similar to At5g66750 was significantly down regulated (steadily through dormancy transitions in leafy spurge and after four weeks of SD—a treatment known to induce endodormancy in poplar). SWI2/SNF2 ATPases are central subunits of a much larger chromatin remodeling complex that enhance expression of specific genes (Pazin and Kadonaga 1997). Studies conducted with the high-density microarrays indicated that a transcript with similarity to the tomato Blind gene, an R2R3 Myb class transcription factor involved in lateral meristem formation (Schmitz et al. 2002) was up regulated during the transitions through endo- and eco-dormancy. Because lateral or axillary meristems eventually result in axillary buds (AguilarMartinez et al. 2007), this gene may play a significant role in either vegetative bud formation or initiation of branching points within buds. In either case, alteration of this gene could result in a mechanism for modified branch architecture during vegetative reproduction. As indicated by Figure 8.2, transitions in well-defined phases of dormancy highlight an involvement of several core components of nucleosomes (Histones H2A and H2B). Covalent modifications of histones play crucial roles in transcriptional activation, DNA repair, and chromatin condensation (Eckardt 2007). Data observed in Figure 8.2 indicate that these histones likely undergo protein:protein interactions with a chromosome condensation component (KOG1426), which may also affect interactions with key cell cycle proteins (KOG0594). It is intriguing that a ubiquitin protein ligase (KOG0419) sits adjacent to the previously mentioned KOGs in this interactome map, since ubiquination of histones by gene products such as HISTONE MONOUBIQUITINATION1 (HUB1) influence the expression of downstream genes associated with cell cycle regulation, including genes associated with seed dormancy (Eckardt 2007). Many cell cycle genes/markers show a general pattern of down regulation during endoand ecodormancy in leafy spurge crown buds while others show a constitutive pattern of expression (Anderson et al. 2003; Horvath et al. 2005, 2008). Retinoblastoma (Rb) is

122

WEEDY AND INVASIVE PLANT GENOMICS

also involved in cell cycle regulation and acts both through sequestration of several E2F transcription factor family members—positive regulators of cell division—and through chromatin modification of key cell cycle regulatory genes (Shen 2002). Studies conducted specifically on a putative leafy spurge Rb orthologue demonstrated that it was not differentially expressed through dormancy transitions (Anderson et al. 2003), which was further confirmed by the same Rb orthologue present on low- and high-density microarrays (Horvath et al. 2006, 2008). However, the high-density leafy spurge microarrays did contain a second Rb-related gene which is similar but not identical to the previously mentioned putative Rb orthologue. The second leafy spurge Rb-related gene showed significant up regulation during the transition from endo- to ecodormancy (Horvath et al. 2008). Thus, it will be intriguing to determine how this particular Rb-related family member functions within the cell cycle pathway. A MAX2/ORE9-like transcript was up-regulated through endo- and ecodormancy, which would be consistent with the inhibition of auxin transport (Chao et al. 2007). These data also correlate well with previous studies showing that a dormancy-associated, auxin-repressed gene is up-regulated in crown buds of leafy spurge during the endo- and ecodormancy transition (Anderson et al. 2005) and the fact that reduced auxin transport appears to play a role in the endodormant phase in the cambial layer of poplar (Schrader et al. 2004). Loss of paradormancy (Chao et al. 2006) or induction of endo- and ecodormancy (Anderson et al. 2005) in crown buds of leafy spurge results in a rapid breakdown of stored starch. Studies have indicated that several members of the leafy spurge beta-amylase family are rapidly up regulated in crown buds during the loss of paradormancy and induction of endo- and ecodormancy (Chao et al. 2007). Silencing of these beta-amylase genes to prevent the conversion of starch to simple sugars could alter known sugar signaling pathways (Anderson et al. 2005) and may make crown and root buds of leafy spurge less tolerant to overwintering temperatures.

Future Work

It is now well established that the transition from endodormancy to ecodormancy, a period associated with a vernalization requirement, in crown buds of leafy spurge is tightly correlated with the induction of flowering (Anderson et al. 2005; Horvath et al. 2008) and likely involves a mechanism associated with chromatin remodeling (Horvath et al. 2003, 2008). However, almost all of the previous global expression data collected from leafy spurge crown and root buds represent samples collected from outdoor-grown field plants or plants from greenhouse stocks that were not flower competent. Thus, multiple environmental factors or signals affecting the plants that were grown under natural field conditions could be masking differentially expressed genes involved in regulating dormancy transitions and initiation of flower competency. The development of protocols for inducing flowering and endodormancy in leafy spurge plants grown under controlled environmental conditions have recently been established (Foley and Anderson, personal communication). Microarray analysis of crown buds collected from plants grown under controlled environmental conditions (light, temperature, etc.) will help narrow the candidates of differentially expressed transcripts responsive to specific treatments, and focus on coordinately regulated genes specifically involved in well-defined phases of bud dormancy. Additionally, these controlled-environment experiments have begun to yield data useful to characterize the signals regulating flowering in leafy spurge. Control of the flowering

LEAFY SPURGE: A MODEL TO STUDY PERENNIAL WEED TRAITS

123

cycle is more critical to a perennial growth habit than bud dormancy, as evidenced by perennial growth in tropical plants with vegetative buds that never go dormant. Thus, transcriptomic analysis of flowering regulation in leafy spurge could provide much-needed information on the regulation of sexual propagation in perennials. The fact that cDNA microarrays can be used to follow transcriptome changes in related species (Horvath et al. 2003; Rensink et al. 2005) allows the possibility of comparing global changes in gene expression between closely related species with perennial and annual growth habits. For example, there are different ecotypes of the same nightshade species that have perennial or annual life cycles (Hobbs et al. 2000). Potato or tomato microarrays could be used to identify changes in gene expression associated with the different life cycles of these two ecotypes. Likewise, there are annual spurges that do not produce adventitious buds and spurges that produce adventitious buds only on specific below-ground organs and locations (Klimešové and Martínková 2004). The new high density microarray could be used to study differences in both perennial growth and bud development. Microarrays are available for corn and sorghum. Perennial weeds such as Johnsongrass are closely related to these species. Thus, comparisons between these annual and perennial species are now possible. The wealth of available microarrays for numerous crop species provides many opportunities for studying additional aspects of perennial growth in related perennial weeds.

References Aida M, Ishida T, Tasaka M (1999) Shoot apical meristem and cotyledon formation during Arabidopsis embryogenesis: Interaction among the CUP-SHAPED COTYLEDON and SHOOT MERISTEMLESS genes. Development 126, 1563–1570. Aguilar-Martinez JA, Poza-Carrion C, Cubas P (2007) Arabidopsis BRANCHED1 acts as an integrator of branching signals within axillary buds. Plant Cell 19, 458–472. Anderson JV (2008) Emerging technologies: an opportunity for weed biology research. Weed Science 56, 281–282. Anderson JV, Chao WS, Horvath DP (2001) A current review on the regulation of dormancy in vegetative buds. Weed Science 49, 581–589. Anderson JV, Davis DG (2004) Abiotic stress alters transcript profiles and activity of glutathione S-transferase, glutathione peroxidase, and glutathione reductase in Euphorbia esula. Physiologia Plantarum 120, 421–433. Anderson JV, Gesch RW, Horvath DP (2003) Effects of photoperiod on whole-plant carbohydrate partitioning, crown bud development, and gene expression in leafy spurge (Euphorbia esula L.). ASPB Annual Meeting. July 25-30, 2003 in Honolulu, HI. PlantBiology 2003, Final Program and Abstract Supplement, p. 95. http://abstracts.aspb.org/pb2003/ public/P37/0596.html Anderson JV, Gesch RW, Jia Y, Chao WS, Horvath DP (2005) Seasonal shifts in dormancy status, carbohydrate metabolism, and related gene expression in crown buds of leafy spurge. Plant, Cell and Environment 28, 1567–1578. Anderson JV, Horvath DP (2001) Random sequencing of cDNAs and identification of mRNAs. Weed Science 49, 590–59. Anderson JV, Horvath DP, Chao WS, Foley ME, Hernandez A, Thimmapuram J, Liu L, Gong GL, Band M, Kim R, Mikel MA (2007) Characterization of an EST database for the perennial weed leafy spurge: an important resource for weed biology research. Weed Science 55, 193–203. Bangsund DA, Leistritz FL, Leitch JA (1999) Assessing economic impacts of biological control of weeds: The case of leafy spurge in the northern Great Plains of the United States. Journal of Environmental Management 56, 35–43. Barton MK (1998) Cell type specification and self renewal in the vegetative shoot apical meristem. Current Opinion in Plant Biology 1, 37–42. Bielenberg DG, Wang Y, Fan S, Reighard GL, Scorza R, Abbott AG (2004) A deletion affecting several gene candidates is present in the evergrowing peach mutant. Journal of Heredity 95, 436–444. Beveridge CA (2006) Axillary bud outgrowth: sending a message. Current Opinion in Plant Biology 9, 35–40. Böhlenius H, Huang T, Charbonnel-Campaa L, Brunner AM, Jansson S, Strauss SH, Nilsson O (2006) The conserved CO/FT regulatory module controls timing of flowering and seasonal growth cessation in trees. Science 312, 1040–1043.

124

WEEDY AND INVASIVE PLANT GENOMICS

Bowman J, Eshed Y (2000) Formation and maintenance of the shoot apical meristem. Trends in Plant Sciences 5, 110–115. Brand U, Fletcher JC, Hobe M, Meyerowitz EM, Simon R (2000) Dependence of stem cell fate in Arabidopsis on a feedback loop regulated by CLV3 activity. Science 289, 617–619. CAB International (2004) Euphorbia esula (original text by W. Chao and J. V. Anderson). In: Crop Protection Compendium, 2004 edition. Wallingford, UK: CAB International. (CD-ROM). Cary AJ, Che P, Howell SH (2002) Developmental events and shoot apical meristem gene expression patterns during shoot development in Arabidopsis thaliana. Plant Journal 32, 867–677. Chao WS, Anderson JV, Horvath DP (2007a) Changes in well-defined phases of bud dormancy may involve shifts in carbohydrate metabolism. ASPB Annual Meeting, July 7–11, 2007, Chicago, IL. http://2007.botanyconference.org/ engine/search/index.php?func=detailandaid=317 Chao WS, Foley ME, Horvath DP, Anderson JV (2007b) Signals regulating dormancy in vegetative buds. International Journal of Plant Developmental Biology 1, 49–56. Chao WS, Serpe MD, Jia Y, Shelver WL, Anderson JV, Umeda M (2007c) Potential roles for autophosphorylation, kinase activity, and abundance of a CDK-activating kinase (Ee;CDKF;1) during growth in leafy spurge. Plant Molecular Biology 63, 365–379. Chao WS, Serpe MD, Anderson JV, Gesch RW, Horvath DP (2006) Sugars, hormones, and environment affect the dormancy status in underground adventitious buds of leafy spurge (Euphorbia esula L.). Weed Science 54, 59–68. Chao WS, Horvath DP, Anderson JV, Foley MP (2005) Potential model weeds to study genomics, ecology, and physiology in the 21st century. Weed Science 53, 929–937. Chouard P (1960) Vernalization and its relations to dormancy. Annual Review of Plant Physiology 11, 191–238. Cline MG (1991) Apical dominance. Botanical Review 57, 318–358. Coupland RT, Selleck GW, Alex JF (1955) Distribution of vegetative buds on the underground parts of leafy spurge (Euphorbia esula L.). Canadian Journal of Agricultural Science 35, 161–167. Druart N, Johansson A, Baba K, Schrader J, Sjödin A, Bhalerao RR, Resman L, Trygg J, Moritz T, Bhalerao RP (2007) Environmental and hormonal regulation of the activity–dormancy cycle in the cambial meristem involves stagespecific modulation of transcriptional and metabolic networks. Plant Journal 50, 557–573. Eckardt N (2007) Two tales of chromatin remodeling converge on HUB1. Plant Cell 19, 391–393. Eklof S, Astot C, Blackwell J, Moritz T, Olsson O, Sandberg G (1997) Auxin-cytokinin interactions in wild-type and transgenic tobacco. Plant Cell Physiology 38, 225–235. Elliott RC, Betzner AS, Huttner E, Oakes MP, Tucker WQ, Gerentes D, Perez P, Smyth DR (1996) AINTEGUMENTA, an APETALA2-like gene of Arabidopsis with pleiotropic roles in ovule development and floral organ growth. Plant Cell 8, 155–168. Gallois JL, Woodward C, Reddy VG, Sablowski R (2002) Combined SHOOT MERISTEMLESS and WUSCHEL trigger ectopic organogenesis in Arabidopsis. Development 129, 3207–3217. Gaudin V, Lunness PA, Fobert PR, Towers M, Riou-Khamlichi C, Murray JAH, Coen E, Doonan JH (2000) The expression of D-cyclin genes defines distinct developmental zones in snapdragon apical meristems and is locally regulated by the Cyctoidea gene. Plant Physiology 122, 1137–1148. Geisler-Lee J, O’Toole NO, Ammar R, Provart NJ, Millar AH, Geisler M (2007) A predicted interactome for Arabidopsis. Plant Physiology 145, 317–329. Gesch RW, Palmquist D, Anderson JV (2007) Seasonal photosynthesis and partitioning of non-structural carbohydrates in Euphorbia esula. Weed Science 55, 346–351. Greb T, Clarenz O, Schafer E, Muller D, Herrero R, Schmitz G, Theres K (2003) Molecular analysis of the LATERAL SUPPRESSOR gene in Arabidopsis reveals a conserved control mechanism for axillary meristem formation. Genes and Development 17, 1175–1187. Groover AT, Mansfield SD, Di Fazio SP, Dupper G, Fontana JR, Millar R, Wang Y (2006) The Populus homeobox gene ARBORKNOX1 reveals overlapping mechanisms regulating the shoot apical meristem and the vascular cambium. Plant Molecular Bilology 61, 917–932. Hobbs HA, Eastburn DM, D’Arcy CJ, Kindhart JD, Masiunas JB, Voegtlin DJ, Weinzierl RA, McCoppin NK (2000) Solanaceous weeds as possible sources of Cucumber mosaic virus in southern Illinois for aphid transmission to pepper. Plant Disease 84, 1221–1224. Horvath DP (1999) Role of mature leaves in inhibition of root bud growth in Euphorbia esula L. Weed Science 47, 544–550. Horvath DP, Olson PA (1998) Cloning and characterization of cold-regulated glycine-rich RNA-binding protein genes from leafy spurge (Euphorbia esula L.) and comparison to heterologous genomic clones. Plant Molecular Biology 38, 531–538.

LEAFY SPURGE: A MODEL TO STUDY PERENNIAL WEED TRAITS

125

Horvath DP, Schaffer R, West M, Wisman E (2003a) Arabidopsis microarrays identify conserved and differentially expressed genes involved in shoot growth and development from distantly related plant species. Plant Journal 34, 125–134. Horvath DP, Anderson JV, Chao WS, Foley ME (2003b) Knowing when to grow: signals regulating bud dormancy. Trends in Plant Science. 8, 534–540. Horvath DP, Anderson JV, Jia Y, Chao WS (2005a) Cloning, expression, and regulation of CYCLIN D3-2 from leafy spurge (Euphorbia esula). Weed Science 53, 431–437. Horvath DP, Anderson JV, Soto-Suarez M, Chao WS (2006) Transcriptome analysis of leafy spurge (Euphorbia esula) crown buds during shifts in well-defined phases of dormancy. Weed Science 54, 821–827. Horvath DP, Chao WS, Anderson JV (2002) Molecular analysis of signals controlling dormancy and growth in underground adventitious buds of leafy spurge. Plant Physiology 128, 1439–1446. Horvath DP, Choa WS, Suttle JC, Thimmapuram J, Anderson JU (2008) Transcriptome analysis identified novel responses and potential regulatory genes involved in seasonal dormancy transitions of leafy spurge (Euphorbia esula L.) BMC Genomics 9, 536. Horvath DP, Soto M, Jia Y, Chao WS, Anderson JV (2005b) Transcriptome analysis of paradormancy release in root buds of leafy spurge (Euphorbia esula). Weed Science 53, 795–801. Hsu CY, Liu Y, Luthe DS, Yuceer C (2006) Poplar FT2 shortens the juvenile phase and promotes seasonal flowering. Plant Cell. 18, 1846–1861. Jia Y, Anderson JV, Horvath DP, Gu YQ, Chao WS (2006) Identification of differentially-expressed genes in dormant and growing buds of leafy spurge (Euphorbia esula L.). Plant Molecular Biology 61, 329–344. Johnson X, Brcich T, Dun E, Goussot M, Haurogné K, Beveridge CA, Rameau C (2006) Branching genes are conserved across species: genes controlling a novel signal in pea are co-regulated by other long-distance signals. Plant Physiology 142, 1014–1026. Klimeˇsové J, Martínková J (2004) Intermediate growth forms as a model for the study of plant clonality functioning: an example with root sprouters. Evolutionary Ecology 18, 669–681. Kumaran MK, Bowman JL, Sundaresan V (2002) YABBY polarity genes mediate the repression of KNOX homeobox genes in Arabidopsis. Plant Cell 14, 2761–2770. Lee JH, Yoo SJ, Park SH, Hwang I, Lee JS, Ahn JH (2007) Role of SVP in the control of flowering time by ambient temperature in Arabidopsis. Genes and Development 21, 397–402. Lenhard M, Laux T (1999) Shoot meristem formation and maintenance. Current Opinion in Plant Biology 2, 44–50. Li JJ, Herskowitz I (1993) Isolation of ORC6, a component of the yeast origin recognition complex by a one-hybrid system. Science 262, 1870–1874. Lokko Y, Anderson JV, Rudd S, Raji A, Horvath DP, Mikel MA, Kim R, Liu L, Hernandez A, Dixon AGO, Ingenlbrecht IL (2007) Characterization of an 18,166 EST dataset for cassava (Manihot esculenta Crantz) enriched for droughtresponsive genes. Plant Cell Reports 26, 1605–1618. Mizukami Y, Fischer RL (2000) Plant organ size control: AIN-TEGUMENTA regulates growth and cell numbers during organogenesis. Proceedings of the National Academy of Sciences of the United States of America 97, 942–947. Nordstrom A, Tarkowski P, Tarkowska D, Norbaek R, Astot C, Dolezal K, Sandberg G (2004) Auxin regulation of cytokinin biosynthesis in Arabidopsis thaliana: a factor of potential importance for auxincytokinin-regulated development. Proceedings of the National Academy of Sciences of the United States of America 101, 8039–8044. Ongaro V, Leyser O (2008) Hormonal control of shoot branching. Journal of Experimental Botany 59, 67–74. Pazin MJ, Kadonaga JT (1997) SW12/SNF2 and related proteins: ATP-driven motors that disrupt protein-DNA interactions? Cell 88, 737–740. Rensink WA, Lee Y, Liu J, Iobst S, Ouyang S, Buell CR (2005) Comparative analyses of six solanaceous transcriptomes reveal a high degree of sequence conservation and species-specific transcripts. BMC Genomics 6, 124 doi:10.1186/1471-2164-6-124. Ruttink T, Arend M, Morreel K, Storme V, Rombauts S, Fromm J, Bhalero RP, Boerjan W, Rohde A (2007) A molecular timetable for apical bud formation and dormancy induction in poplar, Plant Cell 19, 2370–2390. Schmitz G, Tillmann E, Carriero F, Fiore C, Cellini F, Theres K (2002) The tomato Blind gene encodes a MYB transcription factor that controls the formation of lateral meristems. Proceedings of the National Academy of Sciences of the United States of America 99, 1064–1069. Schoof H, Lenhard M, Haecker A, Mayer KF, Jurgens G, Laux T (2000) The stem cell population of Arabidopsis shoot meristems is maintained by a regulatory loop between the CLAVATA and WUSCHEL genes. Cell 100, 635–644. Schrader J, Moyle R, Bhalerao R, Hertzberg M, Lundeberg J, Nilsson P, Bhalerao RP (2004) Cambial meristem dormancy in trees involves extensive remodeling of the transcriptome. Plant Journal 40, 173–187.

126

WEEDY AND INVASIVE PLANT GENOMICS

Schumacher K, Schmitt T, Rossberg M, Schmitz G, Theres K (1999) The Lateral suppressor (Ls) gene of tomato encodes a new member of the VHIID protein family. Proceedings of the National Academy of Sciences of the United States of America 96, 290–295. Sharma VK, Fletcher JC (2002) Maintenance of shoot and floral meristem cell proliferation and fate. Plant Physiology 129, 31–39. Shen WH (2002) The plant E2F-Rb pathway and epigenetic control. Trends in Plant Science 7, 505–511. Shimizu-Sato S, Mori H (2001) Control of outgrowth and dormancy in axillary buds. Plant Physiology 127, 1405–1413. Snowden KC, Simkin AJ, Janssen BJ, Templeton KR, Loucas HM, Simons JL, Karunairetnam S, Gleave AP, Clark DG, Klee HJ (2005) The decreased apical dominance1/Petunia hybrida CAROTENOID CLEAVAGE DIOXYGENASE 8 gene affects branch production and plays a role in leaf senescence, root growth and flower development. Plant Cell 17, 746–759. Tatematsu K, Ward S, Leyser O, Kamiya Y, Nambara E (2005) Identification of cis-elements that regulate gene expression during initiation of axillary bud outgrowth in Arabidopsis. Plant Physiology 138, 757–766. Varanasi V, Boda-Slotta T, Horvath D (unpublished data) Cloning and characterization of a critical meristem developmental gene (EeSTM) from leafy spurge (Euphorbia esula). Weed Science 56, 490–495. Vroemen CW, Mordhorst AP, Albrecht C, Kwaaitaal MACJ, de Vries SC (2003) The CUP-SHAPED COTYLEDON3 gene is required for boundary and shoot meristem formation in Arabidopsis. Plant Cell 15, 1563–1577. Wang H, Qi Q, Schorr P, Cutler AJ, Crosby WL, Fowke LC (1998) ICK1, a cyclin-dependent protein kinase inhibitor from Arabidopsis thaliana interacts with both Cdc2a and CycD3, and its expression is induced by abscisic acid. Plant Journal 15, 501–510. Ward SP, Leyser O (2004) Shoot branching. Current Opinion in Plant Biology 7, 73–78. Yamane H, Kashwa Y, Ooka T, Tao R, Yonemori K (2008) Suppressive hybridization and differential screening reveal endodormancy-associated expression of an SUP/AGL24-type MADS-box gene in lateral vegetative buds of Japanese apricot. Journal of the American Society of Horticultural Science 33, 708–711.

9

Herbicide Resistance: Target Site Mutations Christopher Preston

Introduction

With a few exceptions, all herbicides cause their primary effect by binding or interacting with a single protein. All other herbicide symptoms flow from this initial interaction. Today the tools are available to understand the molecular nature of herbicide target sites prior to herbicide commercialization. However, it was not always so, and many herbicides were commercialized before their target sites were known and fully characterized. There are still herbicides in which the target site is only poorly understood, if known at all. A consequence of most herbicide action being the result of inhibition of a single enzyme is that single amino acid substitutions within the target enzyme may reduce or destroy herbicide binding to the target. Therefore, it should come as no surprise that target site modifications are commonly reported in weeds that become resistant to herbicides. While target site–based resistance to herbicides is commonly reported in weeds, it is not the only mechanism of herbicide resistance to evolve in weed species. Non-target site mechanisms of herbicide resistance are discussed in the next chapter. However, because target site resistance is, for the most part, easy to identify and usually occurs as a single nucleotide change in the gene encoding the target protein, many examples of target site resistance to herbicides are well understood. This chapter will survey target site–based herbicide resistance in weeds in which the amino acid substitutions within the target enzyme that endow resistance have been identified. It is apparent that for most target enzymes, more than one amino acid substitution providing resistance to herbicides is possible. For some target enzymes, a single amino acid substitution is very frequently found in herbicide-resistant weeds; in others, a variety of amino acid substitutions are regularly observed. For most target sites, amino acid substitutions endowing resistance to herbicides and that have not yet been observed in weeds have been identified in chemical mutants in the laboratory or have been created by site-directed mutagenesis of plant or microbial genes. When herbicide target sites are well known, it is possible to predict which target site amino acid substitutions will result in resistance. However, other factors influence the specific amino acid substitutions that are selected in herbicide-resistant weeds in fields. The major factors influencing the selection of specific target site modifications are: the relative fitness of the resistance allele in the absence of selection, the fitness of the resistant allele under selection, and chance. Fitness of resistance alleles in the absence of herbicide selection dictates their relative frequency in unselected weed populations (Jasieniuk et al. 1996). Target site modifications that have higher fitness in the absence of herbicide selection occur more commonly and hence have greater probability of being selected by herbicide use. Fitness under selection ensures those alleles that provide the highest level of resistance to the herbicide will be preferentially selected by herbicide use (Preston 2002). Lastly, chance plays a significant role in the specific mutants selected in the field (Preston 2002). Resistance alleles are unevenly distributed across weed populations. This means that occasionally, an amino acid substitution with lower fitness in the absence of selection and lower fitness under selection may be selected, 127

128

WEEDY AND INVASIVE PLANT GENOMICS

because it was the only resistance allele present in the population. For many target sites, there are numerous herbicides that inhibit a single target enzyme. Different resistance alleles could result in different fitness under selection by different herbicides. Therefore, the specific herbicide that is used as the selecting agent can also influence the amino acid modification selected.

Resistance To Photosystem II-Inhibiting Herbicides

Photosystem II (PS II) catalyzes the key light reaction of photosynthesis in all green plants. As such, it is a natural target site for herbicides and many herbicides inhibit this reaction (Gronwald 1994). The earliest example of herbicide resistance in weeds that was extensively examined was resistance to atrazine and other triazine herbicides that inhibit PS II. Triazine resistance was first reported in Senecio vulgaris from Washington state in 1970 (Ryan 1970). Within the next two decades, triazine resistance appeared in many weed species in both North America and Europe (Holt and LeBaron 1990). Resistance to triazine herbicides was subsequently discovered to result in the failure of the herbicides to effectively inhibit electron transport in PS II, and a single amino acid change within the D1 protein of PS II was determined to be responsible (Hirschberg and MacIntosh 1983). This change of Ser264Gly was observed in numerous triazine resistant weed species (Table 9.1). This particular amino acid substitution results in resistance to the triazine and triazinone herbicides, low-to-no resistance to the urea herbicides, and no cross-resistance to the nitrile herbicides (Fuerst et al. 1986). However, resistance within herbicide chemistries varies in plants containing this mutation. For example, resistance to the urea herbicides ranges from supersensitivity to moderate resistance, depending on the specific herbicide tested (Fuerst et al. 1986). The amino acid substitution Ser264Gly occurs frequently in weed species in which selection has occurred with the triazine herbicides. A large number of other amino acid substitutions within the D1 protein of PS II were selected in algae in the laboratory (Table 9.2), but none of these mutations appeared in field-selected weed species until 1999. To date, four other amino acid substitutions in the D1 protein of PS II have been discovered in herbicide-resistant weeds (Table 9.1). All of these amino acid substitutions have been selected by non-triazine herbicide use.

Table 9.1. Amino acid substitutions in the D1 protein endowing resistance to PS II-inhibiting herbicides selected in weeds. Relative resistance of PS II activity to herbicidesa

Amino acid substitution

Triazine

Val219Ile Ala251Val Ser264Gly Ser264Thr

−b ++++ +++

++ +

+++ ++

Asn266Thr

−b

−b

+b

Urea

Uracil

++

Triazinone

Nitrile

References

+ ++b +++

−

+b

++b

Mengistu et al. (2005) Mechant et al. (2008) Fuerst et al. (1986) Masabni and Zandstra (1999) Park and MallorySmith (2006)

−

a Relative resistance compared to susceptible enzyme: − = <2-fold; + = 2- to 10-fold; ++ = 11- to 100-fold; +++ = 101- to 1,000-fold; ++++ = >1,000-fold; a blank signifies no data provided for that herbicide chemistry. Where variable responses to different herbicides within an herbicide chemistry occur, the highest level of resistance is indicated. b Determined by whole plant response to herbicides.

Table 9.2. Amino acid substitutions in the D1 protein endowing resistance to PS II-inhibiting herbicides from cyanobacteria, green algae, or plant cell lines created by chemical or site-directed mutagenesis. Relative resistance of PS II activity to herbicidesa

Amino acid substitution

Triazine

Urea

Asp170Glu Gly178Ser Phe211Ser Leu218Phe Val Ser Thr Val219Ile

− + + ++ − − + +

+ ++ − +++ + ++ +++ ++

Tyr237Phe Arg238Val Thr245Ala Ser Ile248Phe Thr Ala250His Asn Ala251Cys Gly Val Ile Leu Tyr254Cys Phe255Tyr Gly256Asp Arg257Val Ala263Pro Ser264Gly Ala Thr Lys Ile Pro Asn266Thr Asn267Tyr Ser268Pro Arg269Gly Leu271Val Met Ala Leu275Phe

− + + + −

+ + + + + +

+ − − − ++ − + +++ ++ ++ ++ ++++ ++++ ++ +++ ++++ +++ ++++ − +++ ++ + + − − −

− + − +

Phe295Ser

+

−

+ − + + − + ++ ++ + +++ +++ ++ +++ − − − −

Uracil

+

Triazinone

Nitrile

References

− −

− +

Wilski et al. (2006) Wilski et al. (2006) Gingrich et al. (1988) Narusaka et al. (1998) Wilski et al. (2006)

− +++ +++

− − ++

− − −

+ + − +

++

+++ ++ ++ − +

++ +++ ++

+ ++ ++++ ++ +++ −

++++ ++++ ++++ +++

++

− − − +

− − − ++

− + + + ++

+ − − − − − − + +

+ + + −

Erickson et al. (1989) Johanningmeier et al. (2006) Kless et al. (1994) Wilski et al. (2006) Wilski et al. (2006) Narusaka et al. (1998) Perewoska et al. (1994) Johanningmeier et al. (2000) Johanningmeier et al. (2000) Johanningmeier et al. (1987) Johanningmeier et al. (2006) Wildner et al. (1990) Förster et al. (1997) Narusaka et al. (1998) Ohad and Hirschberg (1992) Erickson et al. (1989) Kless et al. (1994) Dalla Chiesa et al. (1997) Ohad and Hirschberg (1990) Sigematsu et al. (1989) Johanningmeier et al. (2006) Dalla Chiesa et al. (1997) Wilski et al. (2006) Ajlani et al. (1989) Narusaka et al. (1998) Alfonso et al. (1996) Xiong et al. (1997) Ohad and Hirschberg (1992)

Erickson et al. (1989) Wildner et al. (1989) Narusaka et al. (1998)

a Relative resistance compared to susceptible enzyme: − = <2-fold; + = 2- to 10-fold; ++ = 11- to 100-fold; +++ = 101- to 1,000-fold; ++++ = >1,000-fold; a blank signifies no data provided for that herbicide chemistry. Where variable responses to different herbicides within an herbicide chemistry occur, the highest level of resistance is indicated.

129

130

WEEDY AND INVASIVE PLANT GENOMICS

The change Ser264Thr was selected by linuron use in Portulaca oleracea (Masabni and Zandstra 1999). This amino acid substitution provides resistance to triazine and substituted urea herbicides. A change of Val219Ile was selected in Poa annua by diuron use (Mengistu et al. 2000). A further example of Val219Ile substitution occurs in Kochia scoparia resistant to urea herbicides (Mengistu et al. 2005). This amino acid substitution provides resistance to diuron and metribuzin, but not the nitrile herbicides. A change of Asn266Thr in S. vulgaris was selected by use of the nitrile herbicide bromoxynil and provides resistance to nitriles, uracils, and triazinones, but not to triazines or substituted ureas (Park and Mallory-Smith 2006). A change of Ala251Val in a population of Chenopodium album from Sweden selected with the triazinone herbicide metamitron provides resistance to metamitron, but not atrazine (Mechant et al. 2008).

Impact Of Amino Acid Substitutions In Photosystem II

The reaction center of the cyanobacterium Rhodopsuedomonas viridis has been crystallized and the data produced provide insights into the likely structure of the plant PS II reaction center (Michel and Deisenhofer 1988). A crystal structure of the R. viridis reaction center containing bound herbicides is available and suggests roles for important amino acids in herbicide binding (Lancaster and Michel 1999; Sinning 1992). Triazine herbicides hydrogen bond to Ser264 and have hydrophobic interactions with Phe255. Substituted urea herbicides bind deeper in the binding pocket having hydrophobic interactions with Phe255 and His215. The change Ser264Gly removes a hydrogen bond crucial to the binding of triazine herbicides, which explains the high levels of resistance to the triazine herbicides of plants containing this amino acid substitution. Substitution Ser264Thr introduces a more bulky amino acid into the binding pocket and distorts the position of the hydroxyl group. This alteration could reduce the ability of triazine and other herbicides to enter the binding pocket or the ability of triazine herbicides to form a hydrogen bond with Thr264 and/or hydrophobic interaction with Phe255. Val219 is adjacent to His215 on Helix 14. Ile, while a conservative change at this position, is larger than Val and may interfere with hydrophobic interactions of the substituted urea herbicides with His215. Asn266 is located adjacent to Ser264 in the binding pocket. The exact impact of the Thr substitution is unclear; however, it would remove a positive charge within the binding pocket and also provide greater stearic interference for binding of herbicides. Ala251 is also located within the herbicide binding pocket and is predicted to contribute to herbicide binding (Johanningmeier et al. 2000; Xiong et al. 2007). Val has similar chemistry, but is larger than Ala and the Ala251Val substitution would most likely provide stearic interference for binding of some, but not all herbicides. In PS II, the herbicide binding site is also the substrate binding site and there is competition between herbicides and the natural substrate plastoquinone (Ohad and Hirschberg 1992). Therefore, changes to the D1 protein that result in resistance to herbicides could also alter plastoquinone binding. Several changes within the D1 protein are known to influence PS II activity. The substitution Ser264Gly destabilizes plastoquinone binding, reduces the rate of electron transfer in PS II, and reduces photosynthetic efficiency (Jursinic and Pearcy 1988). Photosynthetic performance of plants carrying this resistance allele is not markedly affected under low light conditions, but is adversely affected at high light (Hart and Stemler 1990a, 1990b). In particular, this change makes the photosynthetic apparatus more susceptible to photoinhibition (Hart and Stemler 1990b). The Ser264Thr substitution has less of an impact on photosynthetic efficiency (Smeda et al. 1993). The Val219Ile and Asn266Thr substitutions

HERBICIDE RESISTANCE: TARGET SITE MUTATIONS

131

have been characterized as causing no reduction in photosynthetic electron transport (Erickson et al. 1989; Ajlani et al. 1989). Substitutions at Ala251 can have a significant impact on photosynthetic activity, because this amino acid is important for assembly and stability of PS II (Lardans et al. 1997). The PS II herbicide binding site has been intensively studied in mutants of cyanobacteria, green algae, or plant cell cultures produced through chemical or site-directed mutagenesis (Oettmeier 1999; Johanningmeier et al. 2006). These studies have determined that amino acid substitutions occurring at more than twenty different sites in the D1 protein provide resistance to one or more PS II-inhibiting herbicides (Table 9.2). Most of the amino acid substitutions resulting in an herbicide-resistant PS II occur around the herbicide binding site covering the D-E loop and helices D and E of the D1 protein (Oettmeier 1999). A small number of the sites are predicted to occur on the luminal side of the membrane and must act on herbicide binding through long-range interactions (Xiong et al. 1997). A large number of the amino acid substitutions produced by mutagenesis result in low levels of resistance to herbicides and are therefore less likely to be selected in the field (Table 9.2). Other amino acid substitutions, such as Leu218Thr, provide novel patterns of resistance (Wilski et al. 2006) and may be selected in weed populations in the future. A number of the amino acid substitutions produced by sitedirected mutagenesis, such as Tyr254Cys (Narusaka et al. 1998), drastically reduce photosynthetic performance and are therefore unlikely to compete with better performing mutations in the field. The most widely observed substitution within triazine-resistant weed populations is Ser264Gly, despite the significant fitness penalty imposed by this amino acid substitution (Gronwald 1994). This amino acid substitution provides the highest level of triazine resistance of all the substitutions so far observed in weeds (Table 9.1) and therefore is preferentially selected under intense triazine herbicide use. The fact that triazine herbicides have been widely and persistently employed worldwide, particularly in maize cropping (Holt and LeBaron 1990), explains the high prevalence of the Ser264Gly substitution in weed populations resistant to the PS II-inhibiting herbicides. The other amino acid substitutions observed in PS II herbicideresistant weeds have occurred in cropping systems reliant on other PS II-inhibiting herbicides (Masabni and Zandstra 1999; Mengistu et al. 2000; Park and Mallory-Smith 2006; Mechant et al. 2008). These herbicides select for amino acid substitutions that provide greater levels of resistance to the selecting herbicides than the Ser264Gly substitution.

Resistance To Acetohydroxyacid Synthase-Inhibiting Herbicides

Acetohydroxyacid synthase (AHAS), also known as acetolactate synthase (ALS), is a key enzyme in branched-chain amino acid biosynthesis. This enzyme catalyses two reactions: the condensation of two molecules of pyruvate to create 2-aceotolactate or the condensation of pyruvate with 2-ketobutyrate to form 2-aceto-2-hydroxybutyrate (Singh 1999). The first AHAS inhibiting herbicides, the sulfonylureas (SUs), were introduced in the early 1980s. Subsequently, several other herbicide chemistries that inhibit AHAS have been commercialized, including imidazolinones (IMIs), triazolopyrimidine sulfonanilides (TPs), and pyrimidinylthiobenzoates (PTBs) (Saari et al. 1994). Weed resistance to sulfonylurea herbicides appeared very quickly after the introduction of these herbicides (Heap and Knight 1986; Mallory-Smith et al. 1990). Currently there are ninety-five species with populations resistant to AHAS inhibitors (Heap 2008) and an altered target site is responsible for resistance in many weed species (Tranel and Wright 2002). To date, six different sites within AHAS where amino acid substitutions endow

132

WEEDY AND INVASIVE PLANT GENOMICS

Table 9.3. Amino acid substitution of AHAS endowing resistance to herbicides selected in weeds. Relative resistance of AHAS to herbicidesa Amino acid substitution

SU

IMI

TP

References

Ala122Thr

+

+++

−

Pro197Ala Pro197Arg Pro197Gln Pro197His Pro197Ile Pro197Leu Pro197Ser Pro197Thr Ala205Val Asp376Glu Trp574Leu

++++

−

+

+++ ++ +++ +++ +++ ++ ++++c ++++

++ + +++ ++ ++ ++ ++c ++++

+ ++ ++ − ++ − +++c ++++

Ser653Asn Ser653Thr

+ +

++ ++

− −

Bernasconi et al. (1995) Trucco et al. (2005) Boutsalis et al. (1999) Guttieri et al. (1995) Guttieri et al. (1995) Eberlein et al. (1997) Boutsalis et al. (1999) Sibony et al. (2001) Yu et al. (2003) Preston et al. (2006) Ashigh and Tardif (2007) Whaley et al. (2007) Bernasconi et al. (1995) Foes et al. (1999) Patzoldt and Tranel (2007) Patzoldt and Tranel (2007)

b b

a Relative resistance compared to susceptible enzyme: − = <2-fold; + = 2- to 10-fold; ++ = 11- to 100-fold; +++ = 101- to 1,000-fold; ++++ = >1,000-fold; a blank signifies no data provided for that herbicide chemistry. Where variable responses to different herbicides within an herbicide chemistry occur, the highest level of resistance is indicated. b Described as resistant to chlorsulfuron, but no data provided. c Determined by whole plant response to herbicides.

resistance to AHAS-inhibiting herbicides have been discovered in weed species and there are unique patterns of resistance associated with each site (Table 9.3). Amino acid substitutions at Pro197 (numbered according to the Arabidopsis thaliana sequence) have occurred in many weed populations resistant to the AHAS-inhibiting herbicides (Tranel and Wright 2002). These populations are highly resistant to the sulfonylurea herbicides, but have lower resistance to other AHAS inhibitors and little resistance to the IMI herbicides. At least eight amino acid substitutions are documented at Pro197 in weeds and give similar, but not identical, patterns of resistance (Table 9.3). The large number of amino acid modifications so far found suggests other amino acids could be substituted at Pro197 and result in a resistant enzyme. This finding suggests other amino acid substitutions, not yet observed at this site, might yet be found in resistant weed species. Amino acid substitutions at Trp574 have appeared less frequently, but provide high levels of resistance to all classes of AHAS inhibitors (Bernasconi et al. 1995; Foes et al. 1999). Trp574Leu is the only amino acid substitution observed at this site so far in weeds. Sitedirected mutagenesis studies on the Xanthium strumarium AHAS suggested other amino acid substitutions at Trp574 result in an inactive enzyme (Bernasconi et al. 1995). However, later site-directed mutagenesis studies in tobacco and A. thaliana suggest that Ser and Phe substitutions at Trp574 produce a resistance enzyme, albeit one with much reduced catalytic activity (Chong et al. 1999; Chang and Duggleby 1998). Amino acid substitutions at the other four sites within AHAS are rarer in weed species and most have so far only been documented a few times. The Ala205Val substitution provides higher resistance to SU herbicides than IMI herbicides and none to the TP herbicides (Ashigh

HERBICIDE RESISTANCE: TARGET SITE MUTATIONS

133

and Tardif 2007). Ala122Thr, Ser653Thr, and Ser653Asn substitutions provide high resistance to the IMI and PTB herbicides only (Bernasconi et al. 1995; Trucco et al. 2006; Patzoldt and Tranel 2007). Asp376Glu, like Trp574Leu, provides resistance to all four classes of AHAS inhibitors (Whaley et al. 2007).

Impact Of Amino Acid Substitutions In Acetohydroxyacid Synthase

Enzymatic studies show that AHAS-inhibiting herbicides do not compete with substrate binding to AHAS (Shaner et al. 1984) and it has been suggested that herbicide binding occurs at a vestigial quinone binding site (Schloss et al. 1988). Recently AHAS from yeast and A. thaliana have been crystallized with the AHAS-inhibiting herbicides bound (Pang et al. 2003; McCourt et al. 2005; McCourt et al. 2006). There are some slight differences between the structures of the yeast and A. thaliana AHAS, but the SU herbicides bind similarly to both enzymes (McCourt et al. 2006). The major structural difference of note is that residues 652 to 660 are located closer to the herbicide binding site in the A. thaliana enzyme. The crystal structure shows the SU herbicides bind at the opening of a channel leading to the active site in AHAS (Pang et al. 2003). Ala122, Pro197, Ala205, Asp376, Trp574, and Ser653, the sites where amino acid substitutions endowing herbicide resistance have been found in weeds, are all located around chlorimuron-ethyl when the herbicide is docked to the binding site (McCourt et al., 2006). The SU herbicides make multiple hydrophobic interactions with AHAS as well as hydrogen bonds (McCourt et al. 2005). There are slight differences in binding between the different SU herbicides, the most notable being an additional hydrogen bond formed with chlorimuron-ethyl (McCourt et al. 2005). One of the most notable interactions of the SU herbicides with AHAS is hydrophobic stacking between the heterocycle ring of the herbicides and Trp574. Other hydrophobic interactions occur with Pro197, Ala205, and Asp376 (McCourt et al. 2005; McCourt et al. 2006). Substitutions at amino acids that provide hydrophobic interactions with the herbicide could destabilize binding of the herbicide to AHAS. The mutation Trp574Leu would result in loss of a significant hydrophobic interaction with the SU herbicides through loss of the indolyl ring of Trp and would be expected to greatly reduce binding of the herbicides (Pang et al. 2003). In contrast, the substitution Ala205Val would not reduce hydrophobicity. However, because Val is larger than Ala, this change may interfere with the phenyl ring of the SU herbicides binding correctly to AHAS. Similarly, the substitution Asp376Glu will not change hydrophobicity because both Asp and Glu are acidic amino acids. However, Glu is larger and, once again, might protrude into the region where the phenyl ring of the SU herbicides binds. The situation with Pro197 is more complex. The crystal structure indicates hydrophobic interactions between Pro197 and the phenyl ring of SU herbicides (Pang et al. 2003; McCourt et al. 2006). Substitutions of acidic, basic, and uncharged amino acids for Pro197 all provide resistance to SU herbicides. These substitutions must affect the size or shape of the binding pocket for the herbicides rather than just influencing hydrophobic interactions with the herbicides. Ala122 is located within the region where the sulfonylurea herbicides bind, but points away from the bound herbicide. Likewise, Ser653 sits away from where SU herbicides bind (McCourt et al. 2006). Amino acid substitutions at these two sites result in resistance to IMI herbicides, but little resistance to sulfonylurea herbicides. The A. thaliana AHAS has been crystallized with the IMI herbicide imazaquin bound to the protein (McCourt et al. 2006). Binding of imazaquin partially overlaps with the SU herbicides, with the dihydroimidazolinone ring of imazaquin overlapping in the binding site with the heterocycle ring of SU herbicides. Imazaquin

134

WEEDY AND INVASIVE PLANT GENOMICS

forms interactions with twelve amino acids including Trp574, Ala122, Asp376, and Ser653. Trp574 forms several hydrophobic interactions with the dihydroimidazolinone ring of imazaquin (McCourt et al. 2006). Substitution of Trp574Leu would result in the loss of hydrophobic interactions with the IMI herbicides. Ala122 makes hydrophobic interactions with the dihydroxyimidazolinone ring of imazaquin (McCourt et al. 2006). Thr has a much bulkier side chain than Ala and the Ala122Thr mutation would likely displace IMI herbicides from the binding site. Likewise, Thr and Asn have larger side chains than Ser and replacement of Ser653 by these residues would likely obstruct binding of the quinoline ring of imazaquin (McCourt et al. 2006). The impact of amino acid substitutions on enzyme activity of AHAS has been examined by site-directed mutagenesis in yeast, tobacco, and A. thaliana (Table 9.4). Catalytic properties have also been investigated in some naturally occurring mutations in weed species. Concentrating on amino acid substitutions that have occurred naturally in weed species, the Pro197Ser substitution resulted in a slight decrease in Km for pyruvate in yeast (Duggleby et al. 2003). However, studies with resistant weeds have produced variable results. Extractable AHAS activity in plants containing Pro197 substitutions varied from no difference to large reductions and even increases over the wild type, and could be influenced by the specific amino acid change at Pro197 (Eberlein et al. 1997; Boutsalis et al. 1999; Yu et al. 2003; Preston et al. 2006). The apparent Km for pyruvate in plants containing these substitutions also varied from no change to a two-fold reduction and again may be different for different substitutions (Eberlein et al. 1997; Boutsalis et al. 1999; Preston et al. 2006). Several substitutions at Pro197 resulted in large reductions in feedback inhibition of the enzyme by branch-chain amino acids (Rathinasabapathi et al. 1990; Eberlein et al. 1997; Preston et al. 2006). The Trp574Leu substitution resulted in increased activity of AHAS in both yeast and A. thaliana (Chang and Duggleby 1998; Duggleby et al. 2003). This phenomenon has also occurred with weed species (Boutsalis et al. 1999). In contrast to Pro197 substitutions, the Trp574Leu substitution increased Km for pyruvate substantially in yeast and also in A. thaliana (Chang and Duggleby 1998; Hattori et al. 1995). However, this increase in Km was not seen with the AHAS of S. orientale containing the Trp574Leu substitution (Boutsalis et al. 1999). Unlike the Pro197 substitutions, the Trp574Leu substitution does not result in altered feedback inhibition from branched-chain amino acids (Hattori et al. 1995). The Ala122Val substitution in both yeast and A. thaliana results in an AHAS with much lower catalytic activity, but little change in Km for pyruvate (Chang and Duggleby 1998; Duggleby et al. 2003). The Ala122Thr substitution in tobacco also resulted in much reduced catalytic activity of AHAS with little change to Km for pyruvate (Chong and Choi 2000). The Ala205Val substitution in yeast greatly reduces AHAS activity and increases Km pyruvate by two-fold (Duggleby et al. 2003). This latter amino acid substitution in AHAS of the weed Solanum ptychanthum resulted in a 56% reduction in extractable activity (Ashigh and Tardif 2007). The Ala205Val substitution in S. ptychanthum also reduced feedback inhibition of AHAS, particularly by valine (Ashigh and Tardif 2007). The Asp376Glu amino acid substitution in tobacco did not change AHAS activity, but reduced the Km for pyruvate three-fold (Le et al. 2005a). Amino acid substitutions at Ser653 in A. thaliana and tobacco slightly increased or decreased AHAS activity, depending on the amino acid substitution (Chang and Duggleby 1998; Lee et al. 1988; Chong and Choi 2000). These amino acid substitutions resulted in a slight decrease to no change in Km for pyruvate. It has been suggested that the much reduced feedback inhibition by branched-chain amino acids in AHAS containing amino acid substitutions at Pro197 may provide the fitness penalty that keeps resistance alleles carrying these amino acid substitutions rare in unselected popula-

Table 9.4. Amino acid substitutions in AHAS endowing resistance to herbicides determined by site-directed mutagenesis of yeast, tobacco, or A. thaliana AHAS. Relative resistance of AHAS to herbicidesa

Amino acid substitution

SU

IMI

Gly121Ser Ala122Val Thr Met124Glu Ile Arg142Lys His143Lys Pro197Ser Arg199Ala Glu Ala205Val Phe206Ala His Trp Tyr Lys256Phe Gln Thr Met351Cys Val His352Met Gln Phe Arg373Phe Lys Asp375Glu Ala Asp376Glu Ala Asn Arg377Lys Cys412Ser Trp504Phe Met513Cys Met570Cys Val571Ile Gln Ala Trp574Leu Ser Phe Phe578Asp Glu Ile Lys Arg Trp Leu Cys608Ser Ser653Asn Thr Phe

++++ ++++ +++ ++ − + ++ ++++ − − +++ ++ ++ ++ ++ +++ +++ +++ ++ ++ +++ +++ + + + + ++ +++ +++ +++ + − + − ++ − ++ ++ ++++ ++++ ++++ ++ ++ + ++ ++ − +++ + ++ + +

+++ + +++ +++ + + ++ + + ++ +++ + ++ ++ + ++++ ++++ +++ +++ ++ +++ +++ ++ + + + ++ + − − + − − ++ +++ + + + +++ +++ ++++ ++ + ++ ++ ++ − ++ − +++ +++ +++

TP

+ − +

− ++ − ++ +++ ++ ++ − ++ − + − − + ++ +

References Duggleby et al. (2003) Chang and Duggleby (1998) Chong and Choi (2000) Ott et al. (1996) Le et al. (2005b) Le et al. (2004) Duggleby et al. (2003) Ott et al. (1996) Duggleby et al. (2003) Jung et al. (2004)

Yoon et al. (2002) Duggleby et al. (2003) Le et al. (2003) Duggleby et al. (2003) Oh et al. (2001)

Le et al. (2005b) Le et al. (2005a) Le et al. (2005a) Duggleby et al. (2003)

− + + − ++ + ++

Le et al. (2005b) Shin et al. (2000) Chong et al. (1999) Le et al. (2003) Le et al. (2003) Jung et al. (2004) Duggleby et al. (2003)

+++

Chang and Duggleby (1998) Hattori et al. (1995) Chong et al. (1999) Jung et al. (2004) Duggleby et al. (2003)

++++ ++ ++ + + + + + +

Shin et al. (2000) Chang and Duggleby (1998) Lee et al. (1988) Chong and Choi (2000)

a Relative resistance compared to susceptible enzyme: − = <2-fold; + = 2- to 10-fold; ++ = 11- to 100-fold; +++ = 101- to 1,000-fold; ++++ = >1,000-fold; a blank signifies no data provided for that herbicide chemistry. Where variable responses to different herbicides within an herbicide chemistry occur, the highest level of resistance is indicated.

135

136

WEEDY AND INVASIVE PLANT GENOMICS

tions (Preston et al. 2006). The reduction in Km for pyruvate that occurs with the Trp576Leu amino acid substitution may be sufficient for the wild type to be favored in the absence of selection by herbicides. Likewise, the large reductions in AHAS activity endowed by the Ala122Thr and Ala205Val amino acid substitutions may be sufficient to keep these mutations rare in unselected populations. To date, the negative impacts of the Asp376Glu and Ser653 amino acid substitutions on AHAS are not clear. The extensive worldwide use of SU herbicides is likely the reason for the regular appearance of amino acid substitutions at Pro197 within resistant AHAS, despite the Trp574Leu and Asp364Glu substitutions also resulting in high levels of SU resistance (Saari et al. 1994; Tranel and Wright 2002; Whaley et al. 2007). While there is no obvious fitness penalty associated with the Asp364Glu substitution, it must be less common in populations than the amino acid substitutions at Pro197; otherwise, one would expect it to be selected more often. The other amino acid substitutions found in AHAS-inhibiting herbicide-resistant weeds provide lower levels of resistance to SU herbicides and would be expected to have lower fitness under selection. They are only likely to appear when the Pro197, Trp574, and Asp364 amino acid substitutions are absent. In contrast to the results of SU selection, selection with IMI herbicides results predominantly in amino acid substitutions at Trp574, Ser653, and Ala122 (Tranel and Wright 2002). These amino acid substitutions give high levels of resistance to IMI herbicides. It would also be possible for some IMI herbicides to select for some amino acid substitutions at Pro197, but the much lower levels of IMI resistance generated by amino acid substitutions at Pro197 means they are less likely to be selected. In addition to the amino acid substitutions described above, a further nineteen sites within AHAS where amino acid substitutions result in an herbicide-resistant AHAS have been produced by site-directed mutagenesis (Table 9.4). Naturally occurring resistance has yet not been identified at any of these sites. However, many of these additional sites provide low levels of resistance to herbicides or AHAS enzymes with large reductions in activity or large increases in Km for pyruvate. This suggests there is much less likelihood for selecting resistance at these sites than at the sites that have already been observed in weeds.

Resistance To Acetyl Coenzyme A Carboxylase-Inhibiting Herbicides

Acetyl coenzyme A carboxylase (ACCase) is the first committed step in lipid biosynthesis in plants. ACCase is the target of two herbicide chemical groups: the aryloxyphenoxypropanonate (APP) and the cyclohexanedione (CHD) herbicides (Devine 1997). The ACCase-inhibiting herbicides are selectively used to control grass species in legume and oilseed crops. They were introduced in the 1970s (Duke and Kenyon 1988) and resistance in weeds appeared a few years later (Heap and Knight 1982). Many examples of ACCase-inhibitor resistance reported are the result of target site mutations (Preston and Mallory-Smith 2001). The patterns of cross-resistance within ACCase-inhibitor resistant weed populations are variable with common patterns involving resistance to APP herbicides only, resistance to CHD herbicides only, resistance to APP herbicides and some CHD herbicides, and resistance to all ACCase-inhibiting herbicides (Devine 1997; Preston and Mallory-Smith 2001). These varying patterns of cross-resistance indicated that multiple mutations within ACCase endowing resistance would be discovered. Enzymatic studies suggested that ACCase inhibitors interfered with CO2 transfer within the enzyme (Burton et al. 1991) and hence, the binding site was likely to be at the carboxytransferase domain. When amino acid substitutions within herbicide-

HERBICIDE RESISTANCE: TARGET SITE MUTATIONS

137

resistant ACCase were first sequenced in Lolium rigidum, they were indeed discovered in the carboxytransferase domain (Zagnitko et al. 2001). To date, six sites within this domain have been identified in which amino acid substitutions occur in resistant populations (Table 9.5). The patterns of cross-resistance of these amino acid substitutions are complex, and our understanding is complicated by enzymatic studies on populations containing heterozygous as well as homozygous individuals. Within the carboxyltransferase domain of ACCase, the Ile1781Leu substitution (numbered according to the Alopecurus myosuroides sequence) provides resistance to most APP herbicides and some CHD herbicides (Zagnitko et al. 2001; Tal and Rubin 2004; White et al. 2005). This substitution has been observed in resistant populations of several weed species including L. rigidum (Zagnitko et al. 2001; Délye et al. 2002; Tal and Rubin 2004; Tan et al. 2007), L. multiflorum (White et al. 2005), A. fatua (Christoffers et al. 2002), A. sterilis (Liu et al. 2007), and A. myosuroides (Délye et al. 2002; Brown et al. 2002). Trp2027Cys, Ile2041Asn, and Gly2096Ala have been observed in resistant populations of A. sterilis, A. fatua, A. myosuroides, and L. rigidum (Délye et al. 2003; Délye et al. 2005; Liu et al. 2007; Malone, Boutsalis, and Preston, unpublished data). These amino acid substitutions provide resistance to APP herbicides only. Asp2078Gly provides resistance to both APP and CHD herbicides and has been observed in A. myosuroides, A. sterilis, and L. rigidum (Délye et al. 2005; Liu et al. 2007; Yu et al. 2007a). Cys2088Arg, which has been observed in resistant L. rigidum populations, likewise provides resistance to both APP and CHD herbicides (Yu et al. 2007). The amino acid substitution Trp1999Cys has been found in a resistant A. sterilis population and provides resistance to the APP herbicide fenoxaprop, but not to the APP herbicide haloxyfop or the CHD herbicides (Liu et al. 2007). The crystal structure of the yeast ACCase carboxyltransferase domain has been determined with the APP herbicides haloxyfop or diclofop bound (Zhang et al. 2004). The yeast ACCase is largely resistant to the ACCase-inhibiting herbicides, so interactions of herbicides with the yeast enzyme might not be the same as with the susceptible enzyme from grasses. In the yeast enzyme, the amino acids equivalent to Ile1781, Ile2041, and Trp1999 all surround the binding pocket, although it is not clear that any of these amino acids interact directly with the herbicide. Therefore, the effect of changes at any of these sites is most likely to change the conformation of the binding site. Gly2096, Trp2027, Asp2078, and Cys2088 are a small distance from the APP herbicide binding site (Zhang et al. 2004). At this stage it is not clear why the amino acid substitutions within ACCase result in the variable patterns of resistance to the APP and CHD herbicides selected with ACCase-inhibiting herbicides in the field. The APP and CHD herbicides must bind differentially within the binding pocket. The crystal structure of the ACCase carboxyltransferase domain from yeast indicates a further fourteen amino acid residues surrounding the binding site of haloxyfop (Zhang et al. 2004), suggesting that other amino acid substitutions within ACCase that provide resistance to herbicides will be found. In yeast ACCase, the propionate carboxylate of haloxyfop and diclofop is hydrogen bonded to the main chain at residues equivalent to Ala1707 and Gly1810. Amino acids equivalent to Tyr1814 and Phe2030 provide hydrophobic interactions to the pyridyl ring of haloxyfop (Zhang et al. 2004). To date, no amino acid substitutions at any of these residues have been observed in weeds selected for resistance to ACCase-inhibiting herbicides in the field. Surveys of amino acid substitutions found in ACCase-resistant field-selected populations of A. myosuroides in France found Trp2027Cys and Ile1781Leu to be the most common amino acid substitutions, although three other amino acid substitutions were also identified (Menchari et al. 2006). The Ile1781Leu substitution has been found regularly in resistant populations of other grass weed species (White et al. 2005; Tan et al. 2007). Most of the known amino acid

138

WEEDY AND INVASIVE PLANT GENOMICS

Table 9.5. Amino acid substitution in ACCase endowing resistance to herbicides selected in weeds. Relative resistance of ACCase to herbicidesa Amino acid substitution

APP

CHD

References

Ile1781Leu

++

+++

Trp1999Cys Trp2027Cys Ile2041Asn Asp2078Gly Cys2088Arg Gly2096Ala

+++b ++ +++ +++ ++ ++

−b + + ++ ++ +

Zagnitko et al. (2001) Tal et al. (2004) White et al. (2005) Liu et al. (2007) Délye et al. (2005) Délye et al. (2003) Délye et al. (2005) Yu et al. (2007a) Délye et al. (2005)

a Relative resistance compared to susceptible enzyme: − = <2-fold; + = 2- to 10-fold; ++ = 11- to 100-fold; +++ = 101- to 1,000-fold; ++++ = >1,000-fold; a blank signifies no data provided for that herbicide chemistry. Where variable responses to different herbicides within an herbicide chemistry occur, the highest level of resistance is indicated. b Determined by growth of yeast replacement lines containing a site-directed mutation within the wheat ACCase.

substitutions in herbicide-resistant ACCase have been selected overwhelmingly with APP herbicides and so confer resistance to these herbicides (Table 9.5). As more mutations within ACCase are identified, other patterns of resistance can be expected, especially where CHD herbicides have been the major selecting agent. At present it is not clear why Ile1781Leu and Trp2027Cys are the most common amino acid substitutions selected with ACCase-inhibiting herbicides. This may be because of differences in fitness in the absence of selection rather than fitness differences under selection, because Ile1781Leu does not provide the highest level of resistance to fenoxaprop (Tal and Rubin 2004). Menchari et al. (2008) reported no significant effect of genotype on plant growth or seed production in A. myosuriodes carrying the Ile1781Leu or Ile2041Asn substitutions, but individuals homozygous for the Asp2078Gly substitution had reduced plant growth and seed production. Detailed studies of enzyme function of the mutant enzymes are required to determine the full impacts of these amino acid substitutions.

Resistance To Glyphosate

Glyphosate is the only commercial herbicide to inhibit 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS), a key enzyme in the shikimate pathway (Herrmann 1995). Glyphosate inhibits EPSPS by competing with phophoenol pyruvate (PEP) once shikimate-3-phosphate is already bound (Schönbrunn et al. 2001). Glyphosate only slowly unbinds from EPSPS, effectively resulting in dead-end complex following glyphosate binding. Glyphosate-resistant weeds were first detected in 1996 (Powles et al. 1998), and there are now thirteen weed species with glyphosate-resistant populations (Heap 2008). To date, most cases of glyphosate resistance are not associated with target site modifications (Preston and Wakelin 2008). Where target site modifications of EPSPS have been identified in weed species, they occur at Pro106 (numbered according to the petunia sequence). So far, substitutions of Ser, Thr, and Ala have been observed in the grass weeds Eleusine indica, Lolium rigidum, and L. multiflorum (Baerson et al. 2002; Ng et al. 2003; Wakelin and Preston 2006; Yu et al. 2007; Perez-Jones et al. 2007; Simarmata and Penner 2008; Jasieniuk et al. 2008; Dolman and Preston, unpub-

HERBICIDE RESISTANCE: TARGET SITE MUTATIONS

139

lished data). At the whole plant level, these target site modifications result in three- to fivefold resistance. Jasieniuk et al. (2008) reported L. multiflorum populations containing the Pro106Ala substitution had higher resistance to glyphosate than those containing the Pro106Ser substitution. Enzymatic studies of site-directed mutants suggest minor differences between the various substitutions at Pro106 in the level of resistance to glyphosate with the Pro106Ala substitution resulting in a more resistant enzyme (Stalker et al. 1985; Baerson et al. 2002; Healy-Fried et al. 2007). The E. coli EPSPS with amino acid modifications at Pro101 (equivalent to Pro106 in the petunia sequence) has been crystallized. Pro101 is not involved in binding of glyphosate, but substitutions at this residue shift Gly96 and Thr97 (equivalent to Gly101 and Thr102 in the petunia enzyme) so they protrude into the glyphosate binding site (Healy-Fried et al. 2007). The crystallization of EPSPS from E. coli has identified the glyphosate binding site in the enzyme (Schönbrunn et al. 2001). There are numerous ionic and hydrogen bonds between glyphosate and amino acids, including Lys22, Gly96, Arg124, Gln171, Arg344, Arg386, and Lys411. These are equivalent to Lys23, Gly101, Arg131, Gln180, Arg362, Arg404, and Lys429 in the petunia enzyme. Site-directed mutagenesis of Gly96 in E. coli has shown that mutations at this residue can also impart glyphosate resistance to EPSPS (Padgette et al. 1991; Eschenberg et al. 2002). Site directed mutations at Lys22 and Lys411 in E. coli completely destroyed EPSPS activity (Shuttleworth et al. 1999). Thus far, site-directed mutations at other residues that bind glyphosate have not yet been reported. Additional site-directed mutagenesis has identified Thr42 Met, Thr97 Ile, and Ala183 Thr as changes that will produce a glyphosateresistant EPSPS (He et al. 2003; Pline-Srnic 2005; Kahrizi et al. 2007). These amino acids are equivalent to Thr102 and Ala192 in the petunia enzyme; however, Thr42 is replaced by Asp43 in petunia and other plant EPSP synthases. It is likely that amino acid substitutions at sites other than Pro106 within EPSPS will be found in glyphosate-resistant weed populations in the future. Enzymatic studies of EPSPS with amino acid modifications at Pro106 show that these reduce the affinity for the substrates shikimate-3-phosphate and phosphoenolpyruvate (PEP), as well as reduce binding of glyphosate (Stalker et al. 1985; Baerson et al. 2002; Healy-Fried et al. 2007). The different substitutions at Pro106 result in enzymes with different properties. Substitution of small side chain amino acids for Pro106, such as Ser or Gly, reduces the binding affinity for the substrates by up to two-fold while reducing affinity for glyphosate (Healy-Fried et al. 2007). The substitution Pro106Leu in the rice EPSPS more dramatically affects binding of substrates, affecting affinity of PEP by more than two-fold (Zhou et al. 2006). As a consequence, the mutant EPSPS have catalytic efficiencies between two and ten times lower than the wild type enzyme (Healy-Fried et al. 2007). This mutation is expected to have an impact on the fitness of individuals carrying the mutant allele. The Gly96Ala substitution created by site-directed mutagenesis in E. coli produced an EPSPS with more than 500-fold resistance to glyphosate. However, this substitution results in a thirty-fold reduction in affinity for PEP (Padgette et al. 1991; Eschenberg et al. 2002). The Thr42Met substitution in EPSPS is more than twenty-fold resistant to glyphosate, but like the Gly96Ala substitution, this comes at a cost of a ten-fold reduction in affinity for PEP (He et al. 2003). The much reduced affinity for PEP associated with these amino acid substitutions suggests they are much less likely to be selected in natural populations than substitutions at Pro106. The large fitness penalty expected for plants containing the Gly101Ala substitution will likely counteract the much greater resistance to glyphosate. Padgette et al. (1991) constructed the double substitution Gly96Ala and Pro101Ser in E. coli. The resulting enzyme was highly resistant to glyphosate, but had reduced affinity for PEP.

140

WEEDY AND INVASIVE PLANT GENOMICS

Another double substitution in the maize EPSPS, Thr102Ile and Pro106Ser, produced an enzyme with 100-fold resistance to glyphosate without loss of affinity for PEP (CaJacob et al. 2007). This enzyme was used in some early Roundup Ready maize lines. It is less likely that double mutants in EPSPS will occur spontaneously in weeds; however, such variants could arise in a step-wise fashion where the Pro106Ser substitution would be selected first and an additional substitution selected following further intensive use of glyphosate.

Resistance To Microtubule Assembly Inhibitors

Microtubules are structural components of plant cells consisting of polymers of α/β-tubulin heterodimers. Microtubules are crucial for cell division, organelle movement, and formation of the cell wall. Microtubules are highly dynamic; they are rapidly assembled and disassembled (Wasteneys 2002). The most widely used herbicides that inhibit microtubule assembly are the dinitroanilines, which were introduced in the 1960s (Probst et al. 1975). Resistance to dinitroaniline herbicides first appeared in Eleusine indica in the 1980s in the U.S. (Mudge et al. 1984) and resistance has since occurred in an additional six grass and two broadleaf weed species (Heap 2008). Amino acid substitutions within α-tubulin genes conferring resistance to dinitroaniline herbicides have only been found in E. indica (Anthony et al. 1998; Yamamoto et al. 1998) and Setaria viridis (Délye et al. 2004) to date. Resistance to trifluralin in E. indica was determined to be the result of one of two amino acid substitutions in α-tubulin. A Thr239Ile substitution (numbered according to the Eleusine indica sequence) provided high levels of resistance to dinitroaniline herbicides including trifluralin and oryzalin (Anthony et al. 1998). A Met268Thr substitution provided intermediate levels of resistance to these herbicides (Yamamoto et al. 1998). The Thr239Ile substitution has also been observed in trifluralinresistant S. viridis (Délye et al. 2004). In S. viridis, a Leu136Phe substitution within α-tubulin was also observed in plants resistant to dinitroaniline herbicides (Délye et al. 2004). The crystal structure is available for the mammalian α/β-tubulin dimer (Nogales et al. 1998). While mammalian microtubules are resistant to dinitroaniline herbicides, the sequences of α-tubulin from plants and mammals show significant homology. Molecular modeling of plant α/β-tubulin dimer indicates a likely dinitroaniline binding site in the area of dimer-to-dimer contact (Blume et al. 2003). This suggests that the molecular mode of action for dinitroaniline herbicides is by disrupting contact between α/β-tubulin dimers (Morrissette et al. 2004). The amino acid residues Thr239 and Leu136 are located adjacent to the putative binding site of the dinitroaniline herbicides in susceptible grass species (Délye et al. 2004). Both the Thr239Ile and Leu136Phe substitutions produce larger side chains that might interfere with dinitroaniline herbicide binding. Met268 is located further away from the dinitroaniline herbicide binding site, but a substitution of Thr at this site could change the position of Asn253 within the binding site (Délye et al. 2004). The impacts of these amino acid substitutions on microtubule function in plants are not clear. The microtubules of some protozoan species are also susceptible to disruption by dinitroaniline herbicides (Morrissette et al. 2004). Mutation of α-tubulin in Toxoplasma gondii to produce oryzalin resistance resulted in parasites with higher levels of replication defects (Ma et al. 2007). This may result from increased stability or altered flexibility of the mutant microtubules. Additional amino acid substitutions within α-tubulin providing resistance to dinitroaniline herbicides have been observed in T. gondii and Chlamydomonas reinhardtii. These include

HERBICIDE RESISTANCE: TARGET SITE MUTATIONS

141

substitutions Tyr24His in C. reinhardtii (James et al. 1993) and at fifteen additional sites in T. gondii (Morrissette et al. 2004; Ma et al. 2007). Many of these substitutions result in very modest levels of resistance to oryzalin and might not be selected in weeds by herbicide use. Some substitutions also produce significant negative impacts on microtubule function in T. gondii (Ma et al. 2007). Several amino acid substitutions in T. gondii resulted in high levels of resistance to oryzalin and might be expected to be observed in weeds in the future. These include Val4Leu, Arg243Ser, and Val252Leu (Morrissette et al. 2004). To date, there are no examples of amino acid substitutions in β-tubulin resulting in resistance to herbicides occurring in weed species. Yamamoto and Baird (1999) found no missense mutations in the β-tubulin genes of trifluralin-resistant E. indica. However, substitutions at Lys350 to Met or Glu conferred resistance to dinitroanilines and other microtubule-disrupting compounds in C. reinhardtii (Schibler and Huang 1991). Both of these substitutions enhanced the stability of microtubules and may be disfavored in the field.

Resistance To Phytoene Desaturase Inhibitors

Carotenoids are essential protective pigments synthesized by all photosynthetic organisms. Phytoene desaturase (PDS) is a key enzyme in the biosynthetic pathway leading to the synthesis of carotenoids, and it catalyzes the conversion of phytoene to -carotene (Pecker et al. 1992). Herbicides that inhibit phytoene desaturase were first introduced in the 1970s (Sandmann and Böger 1989). However, the evolution of weed populations resistant to these herbicides did not occur until 1998, when diflufenican-resistant Raphanus raphanistrum appeared in Australia following five applications of diflufenican over ten years (Walsh et al. 2004). Subsequently, Hydrilla verticillata populations resistant to fluridone appeared in lakes in Florida (Michel et al. 2004; Puri et al. 2006). The mechanism of resistance to PDS-inhibiting enzymes has only been determined in H. verticillata, where a series of amino acid substitutions at Arg304 (numbering according to the H. verticillata sequence) within PDS have been observed (Michel et al. 2004; Puri et al. 2007). Collections of H. verticillata from Florida lakes found substitutions of Ser, Cys, or His for Arg304 in PDS in the resistant populations (Michel et al. 2004). These substitutions arise somatically in H. verticillata and so individual lakes usually contain only a single resistant allele. These amino acid substitutions resulted in two- to five-fold resistance of PDS to fluridone, with Arg304His substitution producing the greatest resistance (Michel et al. 2004). None of the amino acid substitutions at Arg304 reduced PDS specific activity. A series of other amino acid substitutions within PDS were found in H. verticillata in some lakes along with mutations at Arg304; however, site-directed mutagenesis of these alleles to convert residue 304 back to Arg restored susceptibility of PDS to fluridone, suggesting these additional amino acid substitutions were unimportant (Michel et al. 2004). Amino acid substitutions in PDS endowing resistance to herbicides have also been observed in the cyanobacterium Synechococcus and Synechocystis (Chamovitz et al. 1991; Chamovitz et al. 1993; Martínez-Férez et al. 1994), where substitutions of Cys, Ser, and Pro at Arg195 (equivalent to Arg304 in H. verticillata) have been found. In Synechocystis, these amino acid substitutions result in high resistance to norflurazon but only low resistance to other PDS inhibitors including flurtamone and fluridone (Martínez-Férez et al. 1994). Amino acid substitutions endowing norflurazon resistance have been observed at three other sites in PDS in Synechococcus: Leu320, Val403, and Leu436, equivalent to Leu425, Val509, and Leu542 in

142

WEEDY AND INVASIVE PLANT GENOMICS

the H. verticillata sequence (Chamovitz et al. 1991; Chamovitz et al. 1993). These amino acid substitutions occur within regions of PDS that are conserved between cyanobacteria and higher plants. Therefore, it is likely that additional amino acid substitutions within PDS providing resistance to herbicides will be found in weeds. In Synechococcus, each of these amino acid substitutions resulted in reduced activity of PDS and lower conversion of phytoene to colored carotenoids, with the Arg195Pro substitution having the greatest impact (Chamovitz et al. 1993). Given that substitutions at Arg304 in H. verticillata had minimal impact of PDS activity (Michel et al. 2004), it is likely that plant PDS operates somewhat differently to that of cyanobacteria.

Conclusions

Target site modifications endowing herbicide resistance in weeds are easily identified because they are typically single nucleotide modifications resulting in the replacement of a single amino acid in the target enzyme. The examples of weed resistance from target site modification given here demonstrate that for most herbicide target sites, multiple amino acid modifications that result in resistance can be selected in weeds. However, it is possible to create some generalizations regarding selection of target site resistance in weeds. Those amino acid substitutions that provide the highest level of resistance, and hence have the greatest fitness under selection, will overwhelmingly be selected by herbicide use. However, the stochasticity in available variants and selection results in the inability to predict the exact outcome of selection in any one population. Equally, while site-directed mutagenesis is a general guide to the amino acid modifications that might be selected in the field, it has proved to be a poor predictor of the amino acid modifications that actually occur in the field. In addition to the examples given above, there are a number of other enzymes in plants targeted by commercial herbicides. For the majority of these it is likely that target site-based resistance will eventually evolve in weeds. For some target sites, such as protoporphyrinogen oxidase (PPO) and p-hydroxypyruvate dioxygenase (HPPD), mutants are known in microorganisms (Randolph-Anderson et al. 1998; Matringe et al. 2005). In addition, a target site modification of PPO has been identified in Amaranthus tuberculatus where deletion of a whole codon, encoding Gly178, from the gene endowed resistance to lactofen (Patzoldt et al. 2006). This demonstrates that alternatives to single base pair changes in the gene encoding the target site are also possible. The vast laboratory of commercial agriculture where farmers treat large populations of weeds persistently with herbicides is sure to provide additional surprises. Multiple amino acid substitutions, amino acid deletions, and even additions may all occur in weed populations, provided these changes result in an herbicide-resistant enzyme. However, there will be situations in which target site modifications will not appear in weed populations, either because they would result in an inactive enzyme, such as with Photosystem I (Preston 1994), or because they produce insufficient resistance to compete with other mechanisms of resistance under selection (Preston 2002). The advent of protein crystallization and structural studies on plant target sites with bound herbicides (McCourt et al. 2006) will undoubtedly provide greater insights into the dynamics of herbicide binding and which amino acids are important for herbicide binding. However, as discussed, amino acid substitutions at key residues that provide hydrogen bonds or hydrophilic interactions are not necessarily found in weed populations. Changes to these amino acids may result in failure of the enzyme substrate to bind and therefore will be selected against in weed

HERBICIDE RESISTANCE: TARGET SITE MUTATIONS

143

populations. Instead, in many cases identified in herbicide-resistant weeds, amino acid substitutions that produce much more subtle changes to the target enzyme will be selected. One important issue in which our understanding is severely lacking is why amino acids that result in resistance to herbicides are rare in unselected weed populations. In a small number of examples the reason is clear: fitness drag in the absence of herbicide. For example, the Ser264Gly substitution in the D1 protein of PS II reduces the efficiency of electron transport (Jursinic and Pearcy 1988) and increases damage to the photosynthetic apparatus by high light (Hart and Stemler 1990b). There is no doubt that future detailed enzymatic studies with mutant enzymes will help resolve this point.

References Ajlani G, Meyer I, Vernotte C, Astier C (1989) Mutation in phenol-type herbicide resistance maps within the psbA gene in Synechocystis 6714. FEBS Letters 246, 207–210. Alfonso M, Pueyo JJ, Gaddour K, Etienne AL, Kirilovsky D, Picorel R (1996) Induced new mutation of D1 serine-268 in soybean photosynthetic cell cultures produced atrazine resistance, increased stability of S2QB− and S3QB− states, and increased sensitivity to light stress. Plant Physiology 112, 1499–1508. Anthony RG, Waldin TR, Ray JA, Bright SWJ, Hussey PJ (1998) Herbicide resistance caused by spontaneous mutation of the cytoskeletal protein tubulin. Nature 393, 260–263. Ashigh J, Tardif FJ (2007) An Ala205Val substitution in acetohydroxyacid synthase of eastern black nightshade (Solanum ptychanthum) reduces sensitivity to herbicides and feedback inhibition. Weed Science 55, 558–565. Baerson SR, Rodriguez DJ, Tran M, Feng Y, Biest NA, Dill GM (2002) Glyphosate-resistant goosegrass. Identification of a mutation in the target enzyme 5-enolpyruvylshikimate-3-phosphate synthase. Plant Physiology 129, 1265–1275. Bernasconi P, Woodworth AR, Rosen BA, Subramanian MV, Siehl DL (1995) A naturally-occurring point mutation confers broad range tolerance to herbicides that target acetolactate synthase. Journal of Biological Chemistry 270, 17381–17385. Blume YB, Nyporko AY, Yemets AI, Baird WV (2003) Structural modeling of the interaction of plant a-tubulin with dinitroaniline and phosphoroamidate herbicides. Cell Biology International 27, 171–174. Boutsalis P, Karotam J, Powles SB (1999) Molecular basis of resistance to acetolactate synthase-inhibiting herbicides in Sisymbrium orientale and Brassica tournefortii. Pesticide Science 55, 507–516. Brown AC, Moss SR, Wilson ZA, Field LM (2002) An isoleucine to leucine substitution in the ACCase of Alopecurus myosuroides (black-grass) is associated with resistance to the herbicide sethoxydim. Pesticide Biochemistry and Physiology 72, 160–168. Burton JD, Gronwald JW, Keith RA, Somers DA, Gengenbach BG, Wyse DL (1991) Kinetics of inhibition of acetylcoenzyme A carboxylase by sethoxydim and haloxyfop. Pesticide Biochemistry and Physiology 39, 100–109. CaJacob CA, Feng PCC, Reiser SE, Padgette SR (2007) Genetically modified herbicide resistant crops. In: Modern Crop Protection Compounds, Krämer W, Schirmer U., eds. Weinheim, Wiley-VCH. pp. 283–302. Chamovitz D, Pecker I, Hirschberg J (1991) The molecular basis of resistance to the herbicide norflurazon. Plant Molecular Biology 16, 967–974. Chamovitz D, Sandmann G, Hirschberg J (1993) Molecular and biochemical characterization of herbicide-resistant mutants of cyanobacteria reveals that phytoene desaturase is a rate-limiting step in carotenoid biosynthesis. Journal of Biological Chemistry 268, 17348–17353. Chang AK, Duggleby RG (1998) Herbicide-resistant forms of Arabidopsis thaliana acetohydroxyacid synthase: characterization of the catalytic properties and sensitivity to inhibitors of four defined mutants. Biochemical Journal 333, 765–777. Chong CK, Choi JD (2000) Amino acid residues conferring herbicide tolerance in tobacco acetolactate synthase. Biochemical and Biophysical Research Communications 279, 462–467. Chong CK, Shin HJ, Chang SI, Choi JD (1999) Role of tryptophanyl residues in tobacco acetolactate synthase. Biochemical and Biophysical Research Communications 259, 136–140. Christoffers MJ, Berg ML, Messersmith CG (2002) An isoleucine to leucine mutation in acetyl-CoA carboxylase confers herbicide resistance in wild oat. Genome 45, 1049–1056. Dalla Chiesa M, Friso G, Deák Z, Vass I, Barber J, Nixon PJ (1997) Reduced turnover of the D1 polypeptide and photoactivation of electron transfer in novel herbicide resistant mutants of Synechocystis sp. PCC 6803. European Journal of Biochemistry 248, 731–740.

144

WEEDY AND INVASIVE PLANT GENOMICS

Délye C, Matéjicek A, Gasquez J (2002) PCR-based detection of resistance to acetyl-CoA carboxylase-inhibiting herbicides in black-grass (Alopecurus myosuroides Huds) and ryegrass (Lolium rigidum Gaud). Pest Management Science 58, 474–478. Délye C, Menchari Y, Michel S, Darmency H (2004) Molecular bases for sensitivity to tubulin-binding herbicides in green foxtail. Plant Physiology 136, 3920–3932. Délye C, Zhang XQ, Chalopin C, Michel S, Powles SB (2003) An isoleucine residue within the carboxyl-transferase domain of multidomain acetyl-coenzyme A carboxylase is a major determinant of sensitivity to aryloxyphenoxypropionate but not to cyclohexanedione inhibitors. Plant Physiology 132, 1716–1723. Délye C, Zhang XQ, Michel S, Matéjicek A, Powles SB (2005) Molecular bases for sensitivity to acetyl-coenzyme A carboxylase inhibitors in black-grass. Plant Physiology 137, 794–806. Devine MD (1997) Mechanisms of resistance to acetyl-coenzyme A carboxylase inhibitors: a review. Pesticide Science 51, 259–264. Duggleby RG, Pang SS, Yu H, Guddat LW (2003) Systematic characterization of mutations in yeast acetohydroxyacid synthase. Interpretation of herbicide-resistance data. European Journal of Biochemistry 270, 2895–2904. Duke SO, Kenyon WH (1988) Polycyclic alkanoic acids. In: Herbicides: Chemistry, Degradation, and Mode of Action, Volume 3, Kearney PC, Kaufman DD, eds. New York, Marcel Dekker. pp. 71–116. Eberlein CV, Guttieri MJ, Mallory-Smith CA, Thill DC, Baerg RJ (1997) Altered acetolactate synthase activity in ALSinhibitor resistant prickly lettuce (Lactuca serriola). Weed Science 45, 212–217. Erickson JM, Pfister K, Rahire M, Togasaki RK, Mets L, Rochaix JD (1989) Molecular and biophysical analysis of herbicide-resistant mutants of Chlamydomonas reinhardtii: structure-function relationship of the Photosystem II D1 polypeptide. Plant Cell 1, 361–171. Eschenberg S, Healy ML, Priestman MA, Lushington GH, Schönbrunn E (2002) How the mutation glycine96 to alanine confers glyphosate insensitivity to 5-enolpyruvyl shikimate-3-phosphate synthase from Escherichia coli. Planta 216, 129–135. Foes MJ, Lui L, Vigue G, Stoller EW, Wax LM, Tranel PJ (1999) A kochia (Kochia scoparia) biotype resistant to triazine and ALS-inhibiting herbicides. Weed Science 47, 20–27. Förster B, Heifetz PB, Lardans A, Boynton JE, Gillham NW (1997) Herbicide resistance and growth of D1 Ala 251 mutants of Chlamydomonas. Zeitschrift für Natturforschung 52c, 654–664. Fuerst EP, Arntzen CJ, Pfister K, Penner D (1986) Herbicide cross-resistance in triazine-resistant biotypes of four weed species. Weed Science 34, 344–353. Gingrich JC, Buzby JS, Stirewalt VL, Bryant DA (1988) Genetic analysis of two new mutations in herbicide resistance in the cyanobacterium Synechococcus sp. PCC 7002. Photosynthesis Research 16, 83–99. Gronwald JW (1994) Resistance to Photosystem II inhibiting herbicides. In: Herbicide Resistance in Plants: Biology and Biochemistry. Powles SB, Holtum JAM, eds. Boca Raton, Fl, Lewis Publishers. pp. 27–60. Guttieri MJ, Eberlein CV, Thill DC (1995). Diverse mutations in the acetolactate synthase gene confer chlorsulfuron resistance in Kochia (Kochia scoparia) biotypes. Weed Science 43, 175–178. Hart JJ, Stemler A (1990a) Similar photosynthetic performance in low light-grown isonuclear triazine-resistant and -susceptible Brassica napus L. Plant Physiology 94, 1295–1300. Hart JJ, Stemler A (1990b) High light-induced reduction and low light-enhanced recovery of photon yield in triazineresistant Brassica napus L. Plant Physiology 94, 1301–1307. Hattori J, Brown D, Mourad G, Labbe H, Ouellet T, Sunohara G, Rutledge R, King J, Miki B (1995) An acetohydroxy acid synthase mutant reveals a single site involved in multiple herbicide resistance. Molecular and General Genetics 246, 419–425. He M, Nie YF, Xu P (2003) A T42M substitution in bacterial 5-enolpyruvylshikimate-3-phosphate synthase (EPSPS) generates enzymes with increased resistance to glyphosate. Bioscience, Biotechnology and Biochemistry 67, 1405–1409. Healy-Fried ML, Funke T, Priestman MA, Han H, Schönbrunn E (2007) Structural basis of glyphosate tolerance resulting from mutations of Pro101 in Escherichia coli 5–enolpyruvylshikimate-3-phosphate synthase. Journal of Biological Chemistry 282, 32949–32955. Heap I (2008) The International Survey of Herbicide Resistant Weeds. www.weedscience.com. Accessed January 1, 2008. Heap I, Knight R (1986) The occurrence of herbicide cross-resistance in a population of annual ryegrass, Lolium rigidum, resistant to diclofop-methyl. Australian Journal of Agricultural Research 37, 149–156. Heap I, Knight R (1982) A population of ryegrass tolerant to the herbicide diclofop-methyl. Journal of the Australian Institute of Agricultural Science 48, 156–157. Herrmann KM (1995) The shikimate pathway: early steps in the biosynthesis of aromatic compounds. The Plant Cell 7, 907–919. Hirschberg J, MacIntosh L (1983) Molecular basis of atrazine resistance in Amaranthus hybridus. Science 222, 1346–1349.

HERBICIDE RESISTANCE: TARGET SITE MUTATIONS

145

Holt JS, LeBaron HM (1990) Significance and distribution of herbicide resistance. Weed Technology 4, 141–149. James SW, Silflow CD, Stroom P, Lefebvre PA (1993) A mutation in the a1-tubulin gene of Chlamydomonas reinhardtii confers resistance to anti-microtubule herbicides. Journal of Cell Science 106, 209–218. Jasieniuk M, Ahmad R, Sherwood AM, Firestone JL, Perez-Jones A, Lanini WT, Mallory-Smith C, Stednick Z (2008) Glyphosate-resistant Italian ryegrass (Lolium multiflorum) in California: distribution, response to glyphosate, and molecular evidence for an altered target enzyme. Weed Science 56, 496–502. Jasieniuk M, Brûlé-Babel AL, Morrison IN (1996) The evolution and genetics of herbicide resistance in weeds. Weed Science 44, 176–193. Johanningmeier U, Bertalan I, Hilbig L, Schulze J, Wilski S, Zeidler E, Oettmeier W (2006) Engineering the D1 subunit of Photosystem II: application to biosensor technology. In: Biotechnological Applications of Photosynthetic Proteins: Biochips, Biosensors and Biodevices. Giardi MT, Piletska EL, eds. pp. 46–56. New York, Springer Science and Business Media. Johanningmeier U, Bodner U, Wildner GF (1987) A new mutation in the gene coding for the herbicide-binding protein in Chlamydomonas. FEBS Letters 211, 221–224. Johanningmeier U, Sopp G, Brauner M, Altenfeld U, Orawski G, Oettmeier W (2000) Herbicide resistance and supersensitivity in Ala250 or Ala251 mutants of the D1 protein in Chlamydomonas reinhardtii. Pesticide Biochemistry and Physiology 66, 9–19. Jung SM, Le DT, Yoon SS, Yoon MY, Kim YT, Choi JD (2004) Amino acid residues conferring herbicide resistance in tobacco acetohydroxy acid synthase. Biochemical Journal 383, 53–61. Jursinic PA, Pearcy RW (1988) Determination of the rate limiting step for photosynthesis in a nearly isonuclear rapeseed (Brassica napus L.) biotype resistant to atrazine. Plant Physiology 88, 1195–1200. Kahrizi D, Salmanian AH, Afshari A, Moieni A, Mousavi A (2007) Simultaneous substitution of Gly96 to Ala and Ala183 to Thr in 5-enolpyruvylshikimate-3-phosphate synthase gene of E. coli (k12) and transformation of rapeseed (Brassica napus L.) in order to make tolerance to glyphosate. Plant and Cell Reports 26, 95–104. Kless H, Oren-Shamir M, Malkin S, McIntosh L, Edelman M (1994) The D-E region of the D1 protein is involved in multiple quinone and herbicide interactions in photosystem II. Biochemistry 33, 10501–10507. Lancaster CDR, Michel H (1999) Refined crystal structures of reaction centres from Rhodopseudomonas viridis in complexes with the herbicide atrazine and two chiral atrazine derivatives also lead to a new model of the bound carotenoid. Journal of Molecular Biology 286, 883–898. Lardans A, Gillham NW, Boynton JE (1997) Site-directed mutations at residue 251 of the Photosystem II D1 protein of Chlamydomonas that result in a nonphotosynthetic phenotype and impair D1 synthesis and accumulation. Journal of Biological Chemistry 272, 210–216. Le DT, Yoon MY, Kim YT, Choi JD (2003) Roles of conserved methionine residues in tobacco acetolactate synthase. Biochemical and Biophysical Research Communications 306, 1075–1082. Le DT, Yoon MY, Kim YT, Choi JD (2004) Homology modeling of the structure of tobacco acetohydroxy acid synthase and examination of the active site by site-directed mutagenesis. Biochemical and Biophysical Research Communications 317, 930–938. Le DT, Yoon MY, Kim YT, Choi JD (2005a) Two consecutive aspartic acid residues conferring herbicide resistance in tobacco acetohydroxy acid synthase. Biochimica et Biophysica Acta 1749, 103–112. Le DT, Yoon MY, Kim YT, Choi JD (2005b) Roles of three well-conserved arginine residues in mediating the catalytic activity of tobacco acetohydroxy acid synthase. Journal of Biochemistry 138, 35–40. Lee KY, Townsend J, Tepperman J, Black M, Chui CF, Mazur B, Dunsmuir P, Bedbrook J (1988) The molecular basis of sulfonylurea resistance in tobacco. EMBO Journal 7, 1241–1248. Liu W, Harrison DK, Chalupska D, Gornicki P, O’Donnell CC, Adkins SW, Haselkorn R, Williams RR (2007) Single-site mutations in the carboxyltransferase domain of plastid acetyl-CoA carboxylase confer resistance to grass-specific herbicides. Proceedings of the National Academy of Sciences of the United States of America 104, 3627–3632. Ma C, Li C, Ganesan L, Oak J, Tsai S, Sept D, Morrissette NS (2007) Mutations in a-tubulin confer dinitroaniline resistance at a cost to microtubule function. Molecular Biology of the Cell 18, 4711–4720. Mallory-Smith CA, Thill DC, Dial MJ (1990) Identification of sulfonylurea herbicide-resistant prickly lettuce (Lactuca serriola). Weed Technology 4, 163–168. Martínez-Férez I, Vioque A, Sandmann G (1994) Mutagenesis of an amino acid responsible in phytoene desaturase from Synechocystis for binding of the bleaching herbicide norflurazon. Pesticide Biochemistry and Physiology 48, 185–190. Masabni JG, Zandstra BH (1999) A serine-to-threonine mutation on linuron-resistant Portulaca oleracea. Weed Science 47, 393–400. Matringe M, Sailland A, Pelissier B, Rolland A, Zink O (2005) p-Hydroxyphenylpyruvate dioxygenase inhibitor-resistant plants. Pest Management Science 61, 269–276.

146

WEEDY AND INVASIVE PLANT GENOMICS

McCourt JA, Pang SS, Guddat LW, Duggleby RG (2005) Elucidating the specificity of binding of sulfonylurea herbicides to acetohydroxyacid synthase. Biochemistry 44, 2330–2338. McCourt JA, Pang SS, King-Scott J, Guddat LW, Duggleby RG (2006) Herbicide-binding sites revealed in the structure of plant acetohydroxyacid synthase. Proceedings of the National Academy of Sciences of the United States of America 103, 569–573. Mechant E, De Marez T, Hermann O, Olsson R, Bulke R (2008) Target site resistance to metamitron in Chenopodium album L. Journal of Plant Diseases and Protection, Special Issue XXI, 37–40 Menchari Y, Camilleri C, Michel S, Brunel D, Dessaint F, Le Corre V, Délye C (2006) Weed response to herbicides: regional-scale distribution of herbicide resistance alleles in the grass weed Alopecurus myosuroides. New Phytologist 171, 861–874. Menchari Y, Chauvel B, Darmency H, Délye C (2008) Fitness costs associated with three mutant acetyl-coenzyme A carboxylase alleles endowing herbicide resistance in black-grass Alopecurus myosuroides. Journal of Applied Ecology 45, 939–947. Mengistu LW, Christoffers MJ, Lym RG (2005) A psbA mutation in Kochia scoparia (L) Schrad from railway rights-of-way with resistance to diuron, tebuthiruron and metribuzin. Pest Management Science 61, 1035–1042. Mengistu LW, Mueller-Warrant GW, Liston A, Barker RE (2000) psbA mutation (valine219 to isoleucine) in Poa annua resistant to metribuzin and diuron. Pest Management Science 56, 209–217. Michel A, Arias RS, Scheffler BE, Duke SO, Netherland M, Dayan FE (2004) Somatic mutation-mediated evolution of herbicide resistance in the nonindigenous invasive plant hydrilla (Hydrilla verticilliata). Molecular Ecology 13, 3229–3237. Michel H, Deisenhofer J (1988) Relevance of the photosynthetic reaction center from purple bacteria to the structure of photosystem II. Biochemistry 27, 1–7. Morrissette NS, Mitra A, Sept D, Sibley LD (2004) Dinitroanilines bind a-tubulin to disrupt microtubules. Molecular Biology of the Cell 15, 1960–1968. Mudge LC, Gossett BJ, Murphy TJ (1984) Resistance of goosegrass (Eleusine indica) to dinitroaniline herbicides. Weed Science 32, 591–594. Narusaka Y, Narusaka M, Kobayashi H, Satoh K (1998) The herbicide-resistant species of the cyanobacterial D1 protein obtained by thorough and random in vitro mutagenesis. Plant and Cell Physiology 39, 620–626. Ng CH, Wickneswari R, Salmijah S, Teng YT, Ismail BS (2003) Gene polymorphisms in glyphosate-resistant and -susceptible biotypes of Eleusine indica from Malaysia. Weed Research 43, 108–115. Nogales E, Wolf SG, Downing KH (1998) Structure of the a/b tubulin dimer by electron crystallography. Nature 391, 199–203. Oettmeier W (1999) Herbicide resistance and supersensitivity in photosystem II. Cellular and Molecular Life Sciences 55, 1255–1277. Oh KJ, Park EJ, Yoon MY, Han TR, Choi JD (2001) Roles of histidine residues in tobacco acetolactate synthase. Biochemical and Biophysical Research Communications 282, 1237–1243. Ohad N, Hirschberg J (1990) A similar structure of the herbicide binding site in photosystem II of plants and cyanobacteria is demonstrated by site specific mutagenesis of the psbA gene. Photosynthesis Research 23, 73–79. Ohad N, Hirschberg J (1992) Mutations in the D1 subunit of Photosystem II distinguish between quinone and herbicide binding sites. The Plant Cell 4, 273–282. Ott KH, Kwagh JG, Stockton GW, Sidorov V, Kakefuda G (1996) Rational molecular design and genetic engineering of herbicide resistant crops by structure modeling and site-directed mutagenesis of acetohydroxyacid synthase. Journal of Molecular Biology 263, 359–368. Padgette SR, Re DB, Gasser CS, Eicholtz DA, Frazier RB, Hironaka CM, Levine EB, Shah DP, Fraley RT, Kishore GM (1991) Site-directed mutagenesis of a conserved region of the 5-enolpyruvylshikimate-3-phosphate synthase active site. Journal of Biological Chemistry 266, 22364–22369. Pang SS, Guddat LW, Duggleby RG (2003) Molecular basis of sulfonylurea herbicide inhibition of acetohydroxyacid synthase. Journal of Biological Chemistry 278, 7639–7644. Park KW, Mallory-Smith CA (2006) psbA mutation (Asn266 to Thr) in Senecio vulgaris L. confers resistance to several PS II-inhibiting herbicides. Pest Management Science 62, 880–885. Patzoldt WL, Tranel PJ (2007) Multiple ALS mutations confer herbicide resistance in waterhemp (Amaranthus tuberculatus). Weed Science 55, 421–428. Patzoldt WL, Hager AG, McCormick JS, Tranel PJ (2006) A codon deletion confers resistance to herbicides inhibiting protoporphyrinogen oxidase. Proceedings of the National Academy of Sciences of the United States of America 103, 12329–12334. Pecker I, Chamovitz D, Linden H, Sandmann G, Hirschberg J (1992) A single polypeptide catalyzing the conversion of phytoene to z-carotene is transcriptionally regulated during tomato fruit ripening. Proceedings of the National Academy of Sciences, USA 89, 4962–4966.

HERBICIDE RESISTANCE: TARGET SITE MUTATIONS

147

Perewoska I, Etienne AL, Kirilovsky D (1994) S1 destabilization and higher sensitivity to light in metribuzin-resistant mutants. Plant Physiology 104, 235–245. Perez-Jones A, Park KW, Polge N, Colquhoun J, Mallory-Smith CA (2007) Investigating the mechanisms of glyphosate resistance in Lolium multiflorum. Planta 226, 395–404. Pline-Srnic W (2005) Technical performance of some commercial glyphosate-resistant crops. Pest Management Science 61, 225–234. Powles SB, Lorraine-Colwill DF, Dellow JJ, Preston C (1998) Evolved resistance to glyphosate in rigid ryegrass (Lolium rigidum) in Australia. Weed Science 46, 604–607. Preston C (1994) Resistance to Photosystem I disrupting herbicides. In: Herbicide Resistance in Plants: Biology and Biochemistry. Powles SB, Holtum JAM, eds. pp. 61–82. Boca Raton, Fl, Lewis Publishers. Preston C (2002) Common mechanisms endowing herbicide resistance in weeds. In 13th Australian Weeds Conference, Papers and Proceedings, Spafford-Jacob H, Dodd J, Moore JH, eds. pp., 666–674. Perth, Plant Protection Society of WA, Inc. Preston C, Mallory-Smith CA (2001) Biochemical mechanisms, inheritance, and molecular genetics of herbicide resistance in weeds. In: Herbicide Resistance and World Grains, Shaner D, Powles SB, eds., pp. 23–60. Boca Raton, Fl, CRC Press. Preston C, Stone LM, Rieger MA, Baker J (2006) Multiple effects of a naturally occurring proline to threonine substitution within acetolactate synthase in two herbicide-resistant populations of Lactuca serriola. Pesticide Biochemistry and Physiology 84, 227–235. Preston C, Wakelin AM (2008) Resistance to glyphosate from altered herbicide translocation patterns. Pest Management Science 64, 372–376. Probst GW, Golab T, Wright WL (1975) Dinitroanilines. In: Herbicides: Chemistry, Degradation and Mode of Action. Kearney PC, Kaufman DD, eds. pp. 453–500. New York, Marcel Dekker, Inc. Puri A, MacDonald GE, Altpeter F, Haller WT (2007) Mutations in phytoene desaturase gene in fluridone-resistant hydrilla (Hydrilla verticillata) biotypes in Florida. Weed Science 55, 412–420. Puri A, MacDonald GE, Haller WT, Singh M (2006) Phytoene and b-carotene response of fluridone-susceptible and -resistant hydrilla (Hydrilla verticillata) biotypes to fluridone. Weed Science 54, 995–999. Randolph-Anderson BL, Sato R, Johnson AM, Harris EH, Hauser CR, Oeda K, Ishige S, Gillham NW, Boynton JE (1998) Isolation and characterization of a mutant protoporphyrinogen oxidase gene from Chlamydomonas reinhardtii conferring resistance to porphryic herbicides. Plant Molecular Biology 38, 839–859. Rathinasabapathi B, Williams D, King J (1990) Altered feedback sensitivity to valine, leucine, and isoleucine of acetolactate synthase from herbicide-resistant variants of Datura innoxia. Plant Science 67, 1–6. Ryan GF (1970) Resistance of common groundsel to simazine and atrazine. Weed Science 18, 614–616. Saari LL, Cotterman JC, Thill DC (1994) Resistance to acetolactate synthase inhibiting herbicides. In: Herbicide Resistance in Plants: Biology and Biochemistry. Powles SB, Holtum JAM, eds. pp. 83–139. Boca Raton, Fl, Lewis Publishers. Sandmann G, Böger P (1989) Inhibition of carotenoid biosynthesis. In: Target Sites of Herbicide Action. Böger P, Sandmann G, eds. pp. 25–44. Boca Raton, Fl, CRC Press. Schibler MJ, Huang B (1991) The colR4 and colR15 b-tubulin mutations in Chlamydomonas reinhardtii confer altered sensitivities to microtubule inhibitors and herbicides by enhancing microtubule stability. Journal of Cell Biology 113, 605–614. Schloss JV, Ciskanik LM, Van Dyk DE (1988) Origin of the herbicide binding site of acetolactate synthase. Nature 331, 360–362. Schönbrunn E, Eschenburg S, Shuttleworth WA, Scloss JV, Amrhein N, Evans JNS, Kabsch W (2001) Interaction of the herbicide glyphosate with its target enzyme 5-enolpyruvylshikimate-3-phosphate synthase in atomic detail. Proceedings of the National Academy of Sciences of the United States of America 98, 1376–1380. Shaner DL, Anderson PC, Stidham MA (1984) Imidazolinones: potent inhibitors of acetohydroxyacid synthase. Plant Physiology 76, 545–546. Shin HJ, Chong CK, Chang SI, Choi JD (2000) Structural and functional role of cysteinyl residues in tobacco acetolactate synthase. Biochemical and Biophysical Research Communications 271, 801–806. Shuttleworth WA, Pohl ME, Helms GL, Jakeman DL, Evans JNS (1999) Site-directed mutagenesis of the putative active site residues of 5-enolpyruvylshikimate-3-phosphate synthase. Biochemistry 38, 296–302. Sibony M, Michel A, Haas HU, Rubin B, Hurle K (2001) Sulfometuron-resistant Amaranthus retroflexus: crossresistance and molecular basis for resistance to acetolactate synthase-inhibiting herbicides. Weed Research 41, 509–522. Sigematsu Y, Sato F, Yamada Y (1989) The mechanism of herbicide resistance in tobacco cells with a new mutation in the QB protein. Plant Physiology 89, 986–992. Simarmata M, Penner D (2008) The basis for glyphosate resistance in rigid ryegrass (Lolium rigidum) from California. Weed Science 56, 181–188.

148

WEEDY AND INVASIVE PLANT GENOMICS

Singh BK (1999) Biosynthesis of valine, leucine and isoleucine. In: Plant Amino Acids: Biochemistry and Biotechnology, Singh BJ, ed. pp. 227–247. New York, Marcel Dekker. Sinning I (1992) Herbicide binding in the bacterial photosynthetic reaction centre. Trends in Biochemical Sciences 17, 150–154. Smeda RJ, Hasegawa PM, Goldsbrough PB, Singh NK, Weller SC (1993) A serine-to-threonine substitution in the triazine herbicide-binding protein in potato cells results in atrazine resistance without impairing productivity. Plant Physiology 103, 911–917. Stalker DM, Hiatt WR, Comai L (1985) A single amino acid substitution in the enzyme 5-enolpyruvylshikimate-3-phosphate synthase confers resistance to the herbicide glyphosate. Journal of Biological Chemistry 260, 4724–4728. Tal A, Rubin B (2004) Molecular characterization and inheritance of resistance to ACCase-inhibiting herbicides in Lolium rigidum. Pest Management Science 60, 1013–1018. Tan MK, Preston C, Wang GX (2007) Molecular basis of multiple resistance to ACCase-inhibiting and ALS-inhibiting herbicides in Lolium rigidum. Weed Research 47, 534–541. Tranel PJ, Wright TR (2002) Resistance of weeds to ALS-inhibiting herbicides: what have we learned? Weed Science 50, 700–712. Trucco F, Hager AG, Tranel PJ (2006) Acetolactate synthase mutation conferring imidazolinone-specific herbicide resistance in Amaranthus hybridus. Journal of Plant Physiology 163, 475–479. Wakelin AM, Preston C (2006) A target-site mutation is present in a glyphosate-resistant Lolium rigidum population. Weed Research 46, 432–440. Walsh MJ, Powles SB, Beard BR, Parkin BT, Porter SA (2004) Multiple-herbicide resistance across four modes of action in wild radish (Raphanus raphanistrum). Weed Science 52, 8–13. Wasteneys GO (2002) Microtubule organization in the green kingdom: chaos or self-order? Journal of Cell Science 115, 1345–1354. Whaley CM, Wilson HP, Westwood JH (2007) A new mutation in plant ALS confers resistance to five classes of ALSinhibiting herbicides. Weed Science 55, 83–90. White GM, Moss SR, Karp A (2005) Differences in the molecular basis of resistance to the cyclohexanedione herbicide sethoxydim in Lolium multiflorum. Weed Research 45, 440–448. Wildner GF, Heisterkamp U, Bodner U, Johanningmeier U (1989) An amino acid substitution in the QB-protein causes herbicide resistance without impairing electron transport from QA to QB. Zeitschrift für Natturforschung 44c, 431–434. Wildner GF, Heisterkamp U, Trebst A (1990) Herbicide cross-resistance and mutations of the psbA gene in Chlamydomonas reinhardtii. Zeitschrift für Natturforschung 45c, 1142–1150. Wilski S, Johanningmeier U, Hertel S, Oettmeier W (2006) Herbicide binding in various mutants of the Photosystem II D1 protein of Chlamydomonas reinhardtii. Pesticide Biochemistry and Physiology 84, 157–164. Xiong J, Hutchison RS, Sayre RT, Govindjee (1997) Modification of the Photosystem II acceptor side function in a D1 mutant (arginine-269-glycine) of Chlamydomonas reinhardtii. Biochimica et Biophysica Acta 1322, 60–76. Xiong J, Subramaniam S, Govindjee (2007) Modeling of the D1/D2 proteins and cofactors of the photosystem II reaction center: Implications for herbicide and bicarbonate binding. Protein Science 5, 2054–2073. Yamamoto E, Baird WV (1999) Molecular characterization of four b-tubulin genes from dinitroaniline susceptible and resistant biotypes of Eleusine indica. Plant Molecular Biology 39, 45–61. Yamamoto E, Zeng L, Baird WV (1998) a-Tubulin missense mutations correlate with antimicrotubule drug resistance in Eleusine indica. The Plant Cell 10, 297–308. Yoon TY, Chung SM, Chang SI, Yoon MY, Hahn TR, Choi JD (2002) Roles of lysine 219 and 255 residues in tobacco acetolactate synthase. Biochemical and Biophysical Research Communications 293, 433–439. Yu Q, Collavo A, Zheng MQ, Owen M, Sattin M, Powles SB (2007a) Diversity of acetyl-coenzyme A carboxylase mutations in resistant Lolium populations: evaluation using clethodim. Plant Physiology 145, 547–558. Yu Q, Cairns A, Powles SB (2007b) Glyphosate, paraquat and ACCase multiple resistance evolved in a Lolium rigidum biotype. Planta 225, 499–513. Yu Q, Zhang XQ, Hashem A, Walsh MJ, Powles SB (2003) ALS gene proline (197) mutations confer ALS herbicide resistance in eight separated wild radish (Raphanus raphanistrum) populations. Weed Science 51, 831–838. Zagnitko O, Jelenska J, Tevzadze G, Haselkorn R, Gornicki P (2001) An isoleucine/leucine residue in the carboxytransferase domain of acetyl-CoA carboxylase is critical for interaction with aryloxyphenoxypropionate and cyclohexanedione herbicides. Proceedings of the National Academy of Sciences, USA 98, 6617–6622. Zhang J, Tweel B, Tong L (2004) Molecular basis for the inhibition of the carboxytransferase domain of acetly-coenzyme-A carboxylase by haloxyfop and diclofop. Proceedings of the National Academy of Sciences, USA 101, 5910–5915. Zhou M, Xu H, Wei X, Ye Z, Wei L, Gong W, Wang Y, Zhu Z (2006) Identification of a glyphosate-resistant mutant of rice 5-enolpyruvylshikimate 3-phosphate synthase using a directed evolution strategy. Plant Physiology 140, 184–195.

10

Molecular And Genomic Mechanisms Of Non-Target-Site Herbicide Resistance Jun Hu, Patrick J. Tranel, C. Neal Stewart Jr., and Joshua S. Yuan

Herbicide Application And Resistance

There has been a dramatic increase of reliance on herbicides to control weeds in row crops in the past thirty years. In particular, the development of herbicide-resistant crops enabled the widespread application of just a few herbicides. Economically, there is no doubt that herbicides and herbicide-resistant crops have drastically improved agricultural efficiency and yields. However, the broad application and/or sometimes the abuse of the herbicides also created problems. The major problem is evolution of weeds with resistance to herbicides—farmers experience new weed problems that cannot be controlled by once-effective herbicides. Herbicide resistance refers to the capacity for a plant to grow and reproduce under the normally lethal dose of herbicide (Yuan et al. 2007). More than 320 biotypes of a total of 185 species (111 dicots and seventy-four monocots) of weeds have evolved resistance to one or more of all major groups of herbicides (www.weedscience.org; accessed November 2008), of which glyphosate might be of most concern. Between 80% and 90% of the U.S. soybean crop is treated with a glyphosate-based product each year. And in many cases, those acres are treated more than once with glyphosate (http://farmindustrynews.com/mag/farming_saving_glyphosate/). Recently, glyphosate-resistant Johnsongrass has been confirmed in Arkansas and Mississippi (http://ipcm.wisc.edu/ WCMNews/tabid/53/EntryID/477/Default.aspx), which brings the total number of glyphosateresistant weeds in the U.S. to nine species. Among those weeds, some are already resistant to acetolactic synthase (ALS)-inhibiting herbicides. Thus, biotypes with multiple resistances are appearing. As the number of glyphosate-resistance weeds continues to increase, cases of multiple resistance also are expected to increase, particularly for out-crossed weed species. These weeds with multiple resistances will not be simply controlled by tank-mixing herbicides, even under their highest usage dose. This scenario, if not disastrous to farmers, would certainly complicate weed control and might render transgenic herbicide-resistant crops less valuable. This is certain: increased herbicide applications will inevitably result in more rapid evolution of weeds with herbicide resistance. The increased incidence of resistance begs for a much better understanding of resistance mechanisms. Management of herbicide-resistant weeds has become an increasing concern for agriculture, making it increasingly urgent for science to reveal the basic mechanisms of resistance, and to develop strategies to mitigate this problem. The mechanisms of herbicide resistance can be classified into two categories: target-site and non-target-site. Target-site resistance is caused by mutations in genes encoding herbicidetargeted proteins; i.e. altering the binding sites of herbicides. Non-target-site resistance, in which the herbicide-targeted protein does not have a significant change at either protein sequence or expression levels, is far more complicated and less understood at both biological and genomic levels. Hypothetically, non-target-site resistance could result from the possible modification (detoxification) and compartmentalization of herbicides and their derivatives (Yuan et al. 2007). Understanding the molecular mechanisms of non-target-site herbicide resistance is important to 149

150

WEEDY AND INVASIVE PLANT GENOMICS

design novel and more effective herbicide molecules for proper weed management. Furthermore, since the non-target gene(s) of interest are likely to be obscure, large-scale genomic technologies are likely to be the only ones capable of discovering these genes and elucidating their functions. Herbicide Classification And Resistance

The mechanisms of herbicide resistance are relevant to the different herbicide chemistries. Most herbicides inhibit a certain enzyme function within the plant cell. For example, chlorsulfuron, diclofop-methyl, oxyfluorfen, and isoxaben target AHAS (Manabe et al. 2007), ACCase (Matthews et al. 1990; De Prado et al. 2005), PPO (Lee et al. 2000), and CESA6 (cellulose synthase) (Desprez et al. 2002), respectively. These herbicides compete with the normal substrate of the functional enzymes and block the corresponding biochemical pathway, which eventually results in the death of the plant. Because these herbicides specifically inhibit a targeted enzyme, target-site resistance might preferentially evolve in these systems in contrast to non-target-site resistance because of natural and widespread variation among sequence in target enzymes. For other herbicides, such as bipyridyliums, which do not bind a specific enzyme, and chloroacetamides, which may have multiple target sites, non-target-site resistance might be expected to occur at a higher frequency. Non-target-site resistance might also be expected to occur more frequently for herbicides such as glyphosate, for which the herbicide-binding site is in a critical—and thus evolutionarily conserved—domain of the enzyme. Non-Target Herbicide Resistance

Because target-site resistance often involves a specific mutation of the corresponding gene, the molecular mechanisms are relatively straightforward to elucidate experimentally. For instance, genes that code for target sites can be sequenced and predictive active site variation is indicative of resistance-conferring mutations. If the target-site gene is not significantly different between the resistant and sensitive biotypes, then the resistance is the product of a non-target-site mutation or expression difference, and genomics approaches may be needed to elucidate the culprit. We herein focus on non-target-site herbicide resistance, which could involve any of a multitude of genes in the detoxification pathways. The sessile lifestyle of plants leads to the evolution of complicated systems responding to the constantly changing environments. Generally speaking, the detoxification process for the herbicide is similar to that for other xenobiotics and consists of three parts: toxic compound recognition and signal transduction, toxic compound modification and degradation, and transportation and compartmentalization of the toxic compounds and their derivatives. Compound recognition and signal transduction for herbicides are not well studied. Signal transduction components are often involved in herbicide-induced responses. The detoxification process generally involves oxidation, glycosylation, acetylation, and S-glutathionylation modification of herbicide molecules. After such modifications, the herbicide derivates or metabolites can be transported and compartmentalized into the vacuolar, or be secreted out of the plant through trichome-like cells in the roots (Dayan et al., 2007). Signal Transduction

Herbicides are designed to block certain functional enzymes or disturb specific biochemical reactions. When these insults occur, endogenous substrates cannot be converted into corre-

MECHANISMS OF NON-TARGET-SITE HERBICIDE RESISTANCE

151

sponding products and they are accumulated within cells. These substrates could be identified by certain mechanisms, finally resulting in stress tolerance regulation. The changes in cellular homeostasis or distortion of membrane function resulting from the herbicide application could also lead to the detoxification process. One example of signaling events is the rapid increase of reactive oxygen species (ROS) when plants are sprayed with an electrophilic herbicide, such as atrazine (Murgia et al. 2004; Ramel et al. 2007). Although there have been no studies on herbicide signal networks, the invasion of herbicide molecules into the cell likely leads to perturbing signal transduction networks. The limited genomic information of weed species has obviated using systems biology approaches to study non-target herbicide resistance mechanisms. This situation could potentially rapidly change. Recent experiments with heterologous microarrays, SAGE-like transcriptional profiling, and high-throughput sequencing has led to the identification of candidate signal transduction pathway genes involved in non-target glyphosate resistance in Conyza canadensis (Yuan and Stewart, unpublished data). The evolution of next-generation sequencing techniques will allow knowledge to be extended to other herbicides and species, facilitating the identification of key signal transduction pathway genes for their roles in non-target herbicide resistance. In the absence of completely elucidated detoxification networks, we can only piece together probable non-target resistance pathways. Sound data exist for non-target mechanisms for certain herbicides and species, but the pathway model we propose is still not validated (Yuan et al. 2007). It follows that upstream signaling events control downstream detoxification processes, which often involve the oxidation, conjugation, and compartmentalization of the herbicide molecules and their derivatives. These components are discussed and pertinent studies are reviewed below.

Detoxification And Modification Oxidation: Cytochrome P450 Monooxygenases

In plant cells, oxidation usually involves enzymes such as oxidases (cytochrome P450 monooxygenase), peroxidases, and other chemical molecules including hydrogen peroxide and free radicals produced during photosynthetic ROS synthesis. Even though hydrogen peroxide and free radicals could directly react with herbicides to oxidize certain herbicides under some conditions, it appears as if most herbicide oxidization is catalyzed by the P450 monooxygenases and other enzymes. One non-P450 enzymatic example is the detoxification/oxidation of glyphosate. The gox gene encodes glyphosate oxidoreductase (GOX) that degrades glyphosate to glyoxlate (Tan et al. 2006). Most data from herbicide oxidation points to P450s as the chief responsible enzyme family. In plants, P450 acts a critical enzyme for the biosynthesis of hormones, sterols (Bishop et al. 2006; Qi et al. 2006; Bishop 2007), and oxygenated fatty acids. The most common reaction catalyzed by cytochrome P450 is the monooxygenase reaction, in which P450 monooxygenase functions as an oxidative media through its electron-transport system (heme center) to oxidize the hydrogen atom of the substrate to a hydroxyl group. In other words, P450 monooxygenases insert one atom of oxygen into hydrophobic compounds to make them more reactive and hydrophilic so that the compound can be degraded in subsequent metabolic reactions. As a result of the reaction, P450s can catalyze oxidation steps resulting in hydroxylations, sulfoxidation, epoxidations, dealkylations, isomerizations, decarboxylations, and deaminations on various substrates.

152

WEEDY AND INVASIVE PLANT GENOMICS

Cytochrome P450 is a very large superfamily of hemoproteins. The plant P450 gene family consists of several hundred members. In Arabidopsis, 246 P450 genes and twenty-six pseudogenes in forty-four sub-families have been identified. Comparative genomic analysis of P450 genes reveals that the gene family is also diverse among different species (Nelson et al. 2004). Gene duplications and divergence, especially multiple local tandem duplications, indicate a dynamic evolution of P450 genes, which might allow for rapid evolution of substrate specificity as well as regulation of expression. Such evolution is important for plants to adapt to the sessile lifestyle since the rapid evolution of P450 genes enabled plants to detoxify the various xenobiotics in ever-changing environments. The role of P450s in non-target herbicide resistance has been well-established through the correlation of P450 enzyme activity with incidence of herbicide resistance in weeds. Kemp et al. (1990) first identified P450 involvement in blackgrass (Alopecurus myosuroides) resistance to chlorotoluron through exogenous application of a P450 enzyme inhibitor and by analyzing the herbicide metabolites that accumulated following treatment. Similar approaches have been employed to identify other cases in which P450s mediated non-target-site herbicide resistance. More recently, research has suggested that P450s are involved in multiple herbicide resistance, which leads to the significant difficulty in managing weed populations with non-target resistance mechanisms (Cocker et al. 2001; Letouze and Gasquez 2003; Yun et al. 2005). Moreover, recent research also indicated the coordination of P450 enzymes with conjugation detoxifying enzymes such as GSTs and glycosyltransferases (Menendez and DePrado 1996; Cocker et al. 2001; Letouze and Gasquez 2003). Further evidence for P450 involvement in non-target herbicide resistance comes from safener application data. Safeners are chemicals added to herbicides to protect crops from herbicide damage by mostly unknown non-target mechanisms. Safener application can induce the expression of P450 genes along with other detoxifying genes including GSTs and ABC transporters (Edwards et al. 2005), implying that safeners activate non-target detoxification pathways. In addition to correlative studies, more direct evidence has been produced that demonstrates P450 enzymes are involved in non-target herbicide resistance. P450 genes have been cloned and protein-herbicide interactions have been characterized in non-weedy species. One of the first P450 genes identified for herbicide resistance was cloned from Jerusalem artichoke (Robineau et al. 1998; Didierjean et al. 2002). The cytochrome P450 CYP76B1 is strongly induced by several xenobiotics and herbicides, and it catalyzes double N-dealkylation reactions metabolizing them (Robineau et al. 1998). In addition, several plant P450 gene products have been shown to detoxify a range of herbicides in crops and model plant species (Morant et al. 2003; Inui and Ohkawa, 2005). Even though no P450 genes responsible for herbicide resistance have yet been identified from weeds, heterologous expression of P450 genes from other species has been used to engineer herbicide-resistant crops. For example, the overexpression of a yeast CYP51A1 in tobacco sidestepped the endogenous sterol biosynthesis pathway and conferred resistance to triazole herbicides (Grausem et al. 1995). A number of transgenic plants with P450-based herbicide resistance have been produced since these early experiments (Yuan et al. 2007). Mammalian P450 genes have been extensively characterized and have become recent targets for overexpression studies in plants, which have endowed herbicide resistance in crops (Inui and Ohkawa 2005). Transgenic experiments have also helped to elucidate the mechanism behind P450 enzyme-induced herbicide resistance in weeds, since P450s in both transgenic plants and naturally-occurring in weeds have resulted in resistance to multiple herbicides (Yuan et al. 2007). Several P450 genes might be involved in conferring resistance to multiple herbicides.

MECHANISMS OF NON-TARGET-SITE HERBICIDE RESISTANCE

153

Overexpressing a single P450 gene in transgenic plants can confer resistance to up to thirteen different herbicides (Hirose et al. 2005).

Conjugation: Glutathione S-Transferases (GSTs) And S-Glutathionylation

Glutathione S-transferase (GST) is a second well-established non-target herbicide resistance gene family. Plant GSTs are multifunctional enzymes that catalyze the conjugation of glutathione (γ-glutamyl-cysteinyl-glycine) to various substrates (R-X) to form a polar S-glutothionylated product (R-SG) (Yuan et al. 2007). Because the R-X substrates are often hydrophobic and electrophilic, GSTs are considered to be important detoxification agents, since the conjugation reaction often leads to R-SG products that can be sequestered into the vacuole by transporters such as ABC transporters (Dixon et al. 2002; Reade et al. 2004). As with P450s, the diversity and dynamic evolution of the GST gene family allows them to detoxify a wide range of chemicals and to be involved in the synthesis of diverse secondary metabolites. GSTs were first implicated in herbicide resistance in the 1970s, when Jensen et al. reported a relationship between GSH conjugate and atrazine resistance in several grass species (Jensen et al. 1977). Further evidence for the involvement of GSTs in non-target-site herbicide resistance derives from GST activity assays in herbicide-resistant weeds. GST activity is normally studied using a model substrate, 1-chloro-2,4-dinitrobenzene (CDNB), whereby the conjugation of GST with artificial substrates is detectable by the shifted light absorbance. Correlations between the herbicide resistance in a weed and the increased GST activity were first established in velvetleaf (Abutilon theophrasti), in which increased glutathione conjugation of atrazine was observed in the resistant biotype (Anderson and Gronwald 1991). GSTs can also confer multi-herbicide resistance, approaching and sometimes surpassing the diversity conferred by P450s (Hall et al. 1997; Cummins et al. 1999; Hatton et al. 1999; Cocker et al. 2001; Letouze and Gasquez 2003). In some cases, increased GST activity is accompanied by increased GST gene expression (Cummins et al. 1997, 1999), whereas in other cases, herbicide resistance has resulted from the increase in GST enzyme activity alone. As with P450s, further evidence for GST-mediated non-target herbicide resistance comes from safener application data, in which induced GST gene expression has been found under safener treatment (Hatzios and Burgos 2004; Smith et al. 2004; Zhang and Riechers 2004; DeRidder and Goldsbrough 2006). Indeed, glutathione peroxidase can also be induced by safeners or herbicides. The safeners could induce several GST expressions in several plant species including Arabidopsis thaliana (DeRidder and Goldsbrough 2006; Nutricati et al. 2006), wheat (Yin et al. 2008), maize (Scarponi et al. 2006; Yin et al. 2008), and rice (Cho et al. 2007). In addition to the studies in weeds, functional characterization of GST genes in crops also confirms the role of GSTs in herbicide metabolism. In 1979, Guddewar and Dauterman purified and characterized a GST enzyme for its activity in herbicide detoxification (Guddewar and Dauterman 1979). Subsequently, many GST enzymes have been purified and characterized for their activity on a variety of herbicides in soybean, wheat, maize, and other crops (Cummins et al. 1997; Dixon et al. 1997; Andrews et al. 2005). Recently, comparative analysis of rice genomic sequences with maize and wheat GSTsled to the molecular cloning of a rice GST gene whose product had activity toward chloroacetamide herbicides (Cho and Kong 2005). A transgenic approach has been useful to study the overexpression of GST genes to confer herbicide resistance (Milligan et al. 2001; Karavangeli et al. 2005; Skipsey et al. 2005). In one case, increased resistance resulted from coordinated

154

WEEDY AND INVASIVE PLANT GENOMICS

overexpression of both GST and thiol synthase (e.g. hGSH synthase) genes, because GST activity requires available thiol (Skipsey et al. 2005). The studies from different perspectives all indicated that GSTs play an important role in herbicide detoxification and resistance. Herbicide applications induce abiotic stress responses. The application of the herbicide atrazine triggered a significant accumulation of H2O2 in maize lines (Nemat Alla and Hassan 2006), which further led to the up regulation of GST genes. Reports also suggested that the phi and the tau class GST enzymes demonstrate herbicide specificity and could play an important role in the detoxification of herbicides (Cho and Kong 2007). In recent years, the novel LC-MS- and GC-MS-based methods make it possible to track the detoxification process by analyzing the trace breakdown products of herbicides (Farkas et al. 2007). In plant detoxification processes, phytochelatin synthase (PCS) (Blum et al. 2007) and gamma-glutamyl transpeptidase (GGT) (Ohkama-Ohtsu et al. 2007) could catabolize glutathione conjugates (GS-conjugates) and result in GS-conjugate degradation. A recent study also shows that GSherbicide conjugates could be transported into roots and might be exuded to the rhizosphere (Schroder et al. 2007). A recent analysis of metabolites has also aided the characterization of the GS-herbicide conjugation process (Brazier-Hicks et al. 2008).

Glycosyltransferases And Glycosylation

Besides GSTs, glycosyltransferases are another family of enzymes shown to be involved in conjugation-based herbicide detoxification. Glycosyltransferases comprise a large gene family in which proteins conjugate a sugar molecule to a wide range of lipophilic small molecule acceptors including plant hormones, secondary metabolites, and xenobiotics (Bowles et al. 2005). The conjugation reactions allow glycosyltransferases to diversify the secondary metabolites via sugar attachment, maintain cell homeostasis by quickly and precisely controlling plant hormone concentration, and detoxify xenobiotics and herbicides through adding sugars onto molecules. Glycosyltransferases exist as a gene superfamily with very diverse members, are found in all kingdoms, and can be classified into seventy-eight subfamilies. For example, the Arabidopsis genome contains more than 100 genes encoding putative UDP-glycosyltransferases, which transfer glucose from UDP-glucose to low-molecular-mass acceptors in the cytosol of plant cells (Keegstra and Raikhel 2001; Ross et al. 2001; Bowles 2002). As for GSTs and P450s, diversity is an important consideration for glucosyltransferasemediated non-target herbicide resistance, because it allows the enzymes to use a wide range of sugar acceptors including herbicides (Bowles et al. 2005; Bowles et al. 2006). Glycosylation could occur on the O (OH- and COOH-), N, S, and C atoms as catalyzed by glycosyltransferases using nucleotide-activated sugars as the donor substrates (Jones and Vogt 2001). Based on the various target atoms, glycosyltransferases could be divided into N-glycosyltransferase (NGT), O-glycosyltransferase (OGT), and others. The glycosylation reactions often convert reactive, bioactive, and toxic aglycones into stable and non-reactive storage forms, thereby limiting their interaction with other cellular components. Glycosyltransferases have been shown to detoxify a variety of chemicals including xenobiotics and pollutants (Brazier et al. 2002, 2003; Loutre et al. 2003; Messner et al. 2003; Poppenberger et al. 2003). Evidence for glycosyltransferases’ role in non-target-site herbicide resistance first came from the induced glycosyltransferase activity in multiple-herbicideresistant blackgrass (Brazier et al. 2002). Further evidence of glycosyltransferase conferring herbicide resistance comes from characterizing glycosyltransferases with activity toward her-

MECHANISMS OF NON-TARGET-SITE HERBICIDE RESISTANCE

155

bicides in crop and model species. In 1992, Leah et al. first isolated two soybean glycosyltransferases that glycosylate the primary major bentazon metabolite, 6-hydroxybentazone (Leah et al. 1992). More glycosyltransferases have been cloned and characterized for their activity toward herbicides such as 2,4,5-trichlorophenol (Brazier et al. 2003; Loutre et al. 2003; Brazier-Hicks and Edwards 2005). As with other non-target resistance genes, these can also often be induced by safener application, which might indicate a role in herbicide detoxification (Hatzios and Burgos 2004; Edwards et al. 2005).

Acetyltransferases And Acetylation

Acetylation is a process of introducing an acetyl group into a compound. Acetylation is catalyzed by the acetyltransferase gene family and can be classified as N-acetylation and O-acetylation. The O-acetylation replaces the active hydrogen atom of a hydroxyl group with an acetyl group to yield an ester. The enzyme involved is often referred as O-acetyltransferase. N-acetylation often transfers the acetyl group from acetyl-coenzyme A to an amine group. The enzyme catalyzing the reaction is often referred to as N-acetyltransferase. Herbicide molecules could be modified by both types of acetyltransferases and be detoxified to less harmful compounds for further degradation or compartmentation. Acetyltransferase was used as a selection marker against the toxicity of phosphinothricin (PPT) for transgenic tobacco. PPT is an analogue of glutamate and acts as a competitive inhibitor of glutamine synthetase (Block et al. 1987; Wohlleben et al. 1988; Botterman et al. 1991). The most widespread application of acetyltransferase results from the application of PPT as a selection marker against bialaphos (bar) for plant genetic engineering (D’Halluin et al. 1992; Padegimas et al. 1994; Lutz et al. 2001; Choi et al. 2003). Chlorinated aromatic xenobiotics can also be substrates of acetyltransferase (N-malonyltransferase [N-MAT]) (Sandermann et al. 1991). For example, glyphosate N-acetyltransferase (GAT) from the soil microbe Bacillus licheniformis is useful as a glyphosate resistance gene (Stokstad 2004). Thus, it is not unreasonable to expect that weeds might have evolved acetylation mechanisms to detoxify glyphosate (Siehl et al. 2007).

Transportation And Compartmentalization: ABC Transporters

In contrast to the above gene families and mechanisms involved in herbicide biochemical modification through metabolism, ABC transporters might confer resistance through sequestration of herbicides and their metabolites. ABC transporters are targeted to membranes and have one or two ATP binding cassettes for active transport using ATP hydrolysis. In higher plants, ABC transporters have been characterized for a wide range of functions including excretion of toxic compounds, sequestration of secondary metabolites, and translocation of fatty acids and phospholipids, as well as transporting of chlorophyll catabolites, auxins, and heavy metals to maintain cell homeostasis (Schulz and Kolukisaglu 2006). ABC transporters are also one of the most diverse gene families in plants, which far surpass non-plant species both in number and in diversity. Diversity, again, is an important consideration for ABC transporter function in herbicide resistance. ABC transporters can be targeted to any part of the endomembrane system, but perhaps the tonoplast might be the most important target for herbicide resistance; sequestration in the vacuole could render an herbicide harmless. Even though herbicide metabolites have long

156

WEEDY AND INVASIVE PLANT GENOMICS

been identified in plant vacuoles, limited research has linked ABC transporters with non-target herbicide resistance in weeds. Nonetheless, ABC transporter activity toward herbicide metabolites has been well established in model species and crop plants. Plant ABC transporters were shown to be able to transport glutathione-conjugated chemicals (Martinoia et al. 1993). A similar experiment showed the glucoside conjugate of the herbicide derivate (5-hydroxyphenyl) primisulfuron could be sequestered in barley mesophyll vacuoles via ABC transporters (Gaillard et al. 1994). AtMRP1 was cloned and characterized as the first ABC transporter gene shown to be able to transport the GS-conjugated herbicide metolachlor (Lu et al. 1997). In addition, several ABC transporters from Arabidopsis thaliana and other species have been shown to transport different herbicides and herbicide metabolites (Liu et al. 2001; Klein et al. 2006; Schulz and Kolukisaglu 2006). Interestingly, reduced translocation has been indicated as a main mechanism for glyphosate resistance, and this might involve ABC transporters (Feng et al. 2004; Koger and Reddy 2005). ABC transporters can also be up regulated upon safener application, thereby providing a possible mechanism for the coordinated overexpression of glutathione-conjugates, ABC transporters, and GST enzymes (DeRidder and Goldsbrough 2006). A GS-conjugate-transporting ABC transporter, AtOPT6, was also up regulated in response to the herbicide primisulfuron (Cagnac et al. 2004). Recent experiments also indicated that glyphosate up regulates several Conyza canadensis ABC transporters in resistant biotypes (Yuan and Stewart, unpublished data). More direct evidence of ABC transporter-mediated herbicide resistance comes from genetic engineering experiments. Overexpression of AtPgp1, a multi-drug resistant family member, and its pea homolog, psNTP9, have been shown to confer multi-herbicide resistance in Arabidopsis (Windsor et al. 2003). In addition to herbicide resistance, resistance to the xenobiotic kanamycin has also been shown in transgenic plants overexpressing an ABC transporter gene (Mentewab and Stewart 2005).

Conclusion

Overall, the molecular and genomic mechanisms of non-target herbicide resistance most likely involve the signal transduction, detoxification, and compartmentalization processes, or, at least, a subset of these. It is clear that transgenic overexpression of a single gene in each of the above five gene families is sufficient to confer herbicide resistance for a number of species and chemicals, which indicates that non-target-site herbicide resistance can be monogenic (Windsor et al. 2003; Inui and Ohkawa 2005). However, increasing evidence points to the coordinative regulation of detoxifying genes conferring herbicide resistance in weeds as shown by safener application data (Gaillard et al. 1994; Persans et al. 2001; Hatzios and Burgos 2004; Edwards et al. 2005). Tandem mechanisms have been demonstrated for a few cases. It has been shown that glycosylation is required for transport of chlorsulfuron-derived 5-hydroxychlorsulfurn into vacuoles via ABC transporters, and glutathionation is required for transport of chlorimuronethyl metabolites into vacuoles via ABC transporters (Bartholomew et al. 2002). The coordinative function of different components of the detoxification pathway might be important to render strong resistance, because the effects could be additive. It is therefore important to understand the mechanisms of non-target-site resistance from a systems biology perspective (Basu et al. 2004; Yuan et al. 2008). Many important questions regarding the mechanisms of non-target herbicide resistance remain unanswered. Microarray and genome analysis have been proposed to study non-target herbicide resistance (Yuan et al.

MECHANISMS OF NON-TARGET-SITE HERBICIDE RESISTANCE

157

2007). However, most weedy species do not have sequence information available. The recent development of next-generation sequencing technology such as Roche GS FLX (454) and the Illumina Genome Analyzer provides a unique opportunity to study the mechanisms of nontarget herbicide resistance at the systems biology level (Yuan et al. 2007). With the availability of more sequence information, we will be able to accelerate our understanding of the genes and pathways crucial in non-target-site resistance.

References Anderson MP, Gronwald JW (1991) Atrazine resistance in a velvetleaf (Abutilon theophrasti) biotype due to enhanced glutathione-S-transferase activity. Plant Physiology 96, 104–109. Andrews CJ, Cummins I, Skipsey M, Grundy NM, Jepson I, Townson J, Edwards R (2005) Purification and characterisation of a family of glutathione transferases with roles in herbicide detoxification in soybean (Glycine max L.): selective enhancement by herbicides and herbicide safeners. Pesticide Biochemistry and Physiology 82, 205–219. Bartholomew DM, Van Dyk DE, Lau SM, O’Keefe DP, Rea PA, Viitanen PV (2002) Alternate energy-dependent pathways for the vacuolar uptake of glucose and glutathione conjugates. Plant Physiology 130, 1562–1572. Basu C, Halfhill MD, Mueller TC, Stewart CN Jr. (2004) Weed genomics: new tools to understand weed biology. Trends in Plant Science 9, 391–398. Bishop G, Nomura T, Yokota T, Montoya T, Castle J, Harrison K, Kushiro T, Kamiya Y, Yamaguchi S, Bancos S, Szatmari AM, Szekeres M (2006) Dwarfism and cytochrome P450-mediated C-6 oxidation of plant steroid hormones. Biochemical Society Transactions 34, 1199–1201. Bishop GJ (2007) Refining the plant steroid hormone biosynthesis pathway. Trends in Plant Science 12, 377–380. Block MD, Botterman J, Vandewiele M, Dockx J, Thoen C, Gossele V, Movva NR, Thompson C, Montagu MV, Leemans J (1987) Engineering herbicide resistance in plants by expression of a detoxifying enzyme. EMBO Journal 6, 2513–2518. Blum R, Beck A, Korte A, Stengel A, Letzel T, Lendzian K, Grill E (2007) Function of phytochelatin synthase in catabolism of glutathione-conjugates. Plant Journal 49, 740–749. Botterman J, Gossele V, Thoen C, Lauwereys M (1991) Characterization of phosphinothricin acetyltransferase and C-terminal enzymatically active fusion proteins. Gene 102, 33–37. Bowles D (2002) A multigene family of glycosyltransferases in a model plant, Arabidopsis thaliana. Biochemical Society Transactions 30, 301–306. Bowles D, Isayenkova J, Lim EK, Poppenberger B (2005) Glycosyltransferases: managers of small molecules. Current Opinion in Plant Biology 8, 254–263. Bowles D, Lim EK, Poppenberger B, Vaistij FE (2006) Glycosyltransferases of lipophilic small molecules. Annual Reviews in Plant Biology 57, 567–597. Brazier M, Cole DJ, Edwards R (2002) O-glucosyltransferase activities toward phenolic natural products and xenobiotics in wheat and herbicide-resistant and herbicide-susceptible black-grass (Alopecurus myosuroides). Phytochemistry 59, 149–156. Brazier M, Cole DJ, Edwards R (2003) Partial purification and characterisation of a 2,4,5-trichlorophenol detoxifying O-glucosyltransferase from wheat. Phytochemistry 64, 419–424. Brazier-Hicks M, Edwards R (2005) Functional importance of the family 1 glucosyltransferase UGT72B1 in the metabolism of xenobiotics in Arabidopsis thaliana. Plant Journal 42, 556–566. Brazier-Hicks M, Evans KM, Cunningham OD, Hodgson DR, Steel PG, Edwards R (2008) Catabolism of glutathione conjugates in Arabidopsis thaliana. Role in metabolic reactivation of the herbicide safener fenclorim. Journal of Biological Chemistry 283, 21102–21112. Cagnac O, Bourbouloux A, Chakrabarty D, Zhang MY, Delrot S (2004) AtOPT6 transports glutathione derivatives and is induced by primisulfuron. Plant Physiology 135, 1378–1387. Cho HY, Kong KH (2005) Molecular cloning, expression, and characterization of a phi-type glutathione S-transferase from Oryza sativa. Pesticide Biochemistry and Physiology 83, 29–36. Cho HY, Kong KH (2007) Study on the biochemical characterization of herbicide detoxification enzyme, glutathione S-transferase. Biofactors 30, 281–287.

158

WEEDY AND INVASIVE PLANT GENOMICS

Cho HY, Lee HJ, Kong KH (2007) A phi class glutathione S-transferase from Oryza sativa (OsGSTF5): molecular cloning, expression and biochemical characteristics. Journal of Biochemistry and Molecular Biology 40, 511–516. Choi YE, Jeong JH, In JK, Yang DC (2003) Production of herbicide-resistant transgenic Panax ginseng through the introduction of the phosphinothricin acetyl transferase gene and successful soil transfer. Plant Cell Reports 21, 563–568. Cocker KM, Northcroft DS, Coleman JOD, Moss SR (2001) Resistance to ACCase-inhibiting herbicides and isoproturon in UK populations of Lolium multiflorum: mechanisms of resistance and implications for control. Pest Management Science 57, 587–597. Cummins I, Cole DJ, Edwards R (1997) Purification of multiple glutathione transferases involved in herbicide detoxification from wheat (Triticum aestivum L.) treated with the safener fenchlorazole-ethyl. Pesticide Biochemistry and Physiology 59, 35–49. Cummins I, Cole DJ, Edwards R (1999) A role for glutathione transferases functioning as glutathione peroxidases in resistance to multiple herbicides in black-grass. Plant Journal 18, 285–292. Cummins I, Moss S, Cole DJ, Edwards R (1997) Glutathione transferases in herbicide-resistant and herbicide-susceptible black-grass (Alopecurus myosuroides). Pesticide Science 51, 244–250. D’Halluin K, De Block M, Denecke J, Janssens J, Leemans J, Reynaerts A, Botterman J (1992) The bar gene as selectable and screenable marker in plant engineering. Methods in Enzymology 216, 415–426. Dayan FE, Watson SB, Nanayakkara NP (2007) Biosynthesis of lipid resorcinols and benzoquinones in isolated secretory plant root hairs. Journal of Experimental Botany 58, 3263–3272. De Prado JL, Osuna MD, Heredia A, De Prado R (2005) Lolium rigidum, a pool of resistance mechanisms to ACCase inhibitor herbicides. Journal of Agricultural and Food Chemistry 53, 2185–2191. DeRidder BP, Goldsbrough PB (2006) Organ-specific expression of glutathione S-transferases and the efficacy of herbicide safeners in Arabidopsis. Plant Physiology 140, 167–175. Desprez T, Vernhettes S, Fagard M, Refregier G, Desnos T, Aletti E, Py N, Pelletier S, Hofte H (2002) Resistance against herbicide isoxaben and cellulose deficiency caused by distinct mutations in same cellulose synthase isoform CESA6. Plant Physiology 128, 482–490. Didierjean L, Gondet L, Perkins R, Lau S-MC, Schaller H, O’Keefe DP, Werck-Reichhart D (2002) Engineering herbicide metabolism in tobacco and Arabidopsis with CYP76B1, a cytochrome enzyme from Jerusalem artichoke. Plant Physiology 130, 179–189. Dixon D, Cole DJ, Edwards R (1997) Characterisation of multiple glutathione transferases containing the GST I subunit with activities toward herbicide substrates in maize (Zea mays). Pesticide Science 50, 72–82. Dixon DP, Lapthorn A, Edwards R (2002) Plant glutathione transferases. Genome Biology 3, reviews 3004.1–3005.10. Edwards R, Del Buono D, Fordham M, Skipsey M, Brazier M, Dixon DP, Cummins I (2005) Differential induction of glutathione transferases and glucosyltransferases in wheat, maize and Arabidopsis thaliana by herbicide safeners. Zeitschrift Fur Naturforschung C 60, 307–316. Farkas M, Berry JO, Aga DS (2007) Determination of enzyme kinetics and glutathione conjugates of chlortetracycline and chloroacetanilides using liquid chromatography-mass spectrometry. Analyst 132, 664–671. Feng PCC, Tran M, Chiu T, Sammons RD, Heck GR, CaJacob CA (2004) Investigations into glyphosate-resistant horseweed (Conyza canadensis): retention, uptake, translocation, and metabolism. Weed Science 52, 498–505. Gaillard C, Dufaud A, Tommasini R, Kreuz K, Amrhein N, Martinoia E (1994) A herbicide antidote (Safener) induces the activity of both the herbicide detoxifying enzyme and of a vacuolar transporter for the detoxified herbicide. FEBS Letters 352, 219–221. Grausem B, Chaubet N, Gigot C, Loper JC, Benveniste P (1995) Functional expression of Saccharomyces cerevisiae CYP51A1 encoding lanosterol-14-demethylase in tobacco results in bypass of endogenous sterol biosynthetic pathway and resistance to an obtusifoliol-14-demethylase herbicide inhibitor. Plant Journal 7, 761–770. Guddewar MB, Dauterman WC (1979) Purification and properties of a glutathione-S-transferase from corn which conjugates s-triazine herbicides. Phytochemistry 18, 735–740. Hall LM, Moss SR, Powles SB (1997) Mechanisms of resistance to aryloxyphenoxypropionate herbicides in two resistant biotypes of Alopecurus myosuroides (blackgrass): Herbicide metabolism as a cross-resistance mechanism. Pesticide Biochemistry and Physiology 57, 87–98. Hatton PJ, Cummins I, Cole DJ, Edwards R (1999) Glutathione transferases involved in herbicide detoxification in the leaves of Setaria faberi (giant foxtail). Physiologia Plantarum 105, 9–16. Hatzios KK, Burgos N (2004) Metabolism-based herbicide resistance: regulation by safeners. Weed Science 52, 454–467. Hirose S, Kawahigashi H, Ozawa K, Shiota N, Inui H, Ohkawa H, Ohkawa Y (2005) Transgenic rice containing human CYP2B6 detoxifies various classes of herbicides. Journal of Agricultural and Food Chemistry 53, 3461–3467.

MECHANISMS OF NON-TARGET-SITE HERBICIDE RESISTANCE

159

Inui H, Ohkawa H (2005) Herbicide resistance in transgenic plants with mammalian monooxygenase genes. Pest Management Science 61, 286–291. Jensen KIN, Stephenson GR, Hunt LA (1977) Detoxification of atrazine in 3 Gramineae subfamilies. Weed Science 25, 212–220. Jones P, Vogt T (2001) Glycosyltransferases in secondary plant metabolism: tranquilizers and stimulant controllers. Planta 213, 164–174. Karavangeli M, Labrou NE, Clonis YD, Tsaftaris A (2005) Development of transgenic tobacco plants overexpressing maize glutathione S-transferase I for chloroacetanilide herbicides phytoremediation. Biomolecular Engineering 22, 121–128. Keegstra K, Raikhel N (2001) Plant glycosyltransferases. Current Opinion in Plant Biology 4, 219–224. Kemp MS, Moss SR, Thomas THE (1990) Herbicide resistance in Alopecurus myosuroides. ACS Symposium Series 421, pp. 376–393. Klein M, Burla B, Martinoia E (2006) The multidrug resistance-associated protein (MRP/ABCC) subfamily of ATP-binding cassette transporters in plants. FEBS Letters 580, 1112–1122. Koger CH, Reddy KN (2005) Role of absorption and translocation in the mechanism of glyphosate resistance in horseweed (Conyza canadensis). Weed Science 53, 84–89. Leah JM, Worrall TL, Cobb AH (1992) Isolation and characterization of 2 glucosyltransferases from glycine-max associated with bentazone metabolism. Pesticide Science 34, 81–87. Lee HJ, Lee SB, Chung JS, Han SU, Han O, Guh JO, Jeon JS, An G, Back K (2000) Transgenic rice plants expressing a Bacillus subtilis protoporphyrinogen oxidase gene are resistant to diphenyl ether herbicide oxyfluorfen. Plant and Cell Physiology 41, 743–749. Letouze A, Gasquez J (2003) Enhanced activity of several herbicide-degrading enzymes: a suggested mechanism responsible for multiple resistance in blackgrass (Alopecurus myosuroides Huds.). Agronomie 23, 601–608. Liu G, Sanchez-Fernandez R, Li ZS, Rea PA (2001) Enhanced multispecificity of arabidopsis vacuolar multidrug resistanceassociated protein-type ATP-binding cassette transporter, AtMRP2. Journal of Biological Chemistry 276, 8648–8656. Loutre C, Dixon DP, Brazier M, Slater M, Cole DJ, Edwards R (2003) Isolation of a glucosyltransferase from Arabidopsis thaliana active in the metabolism of the persistent pollutant 3,4-dichloroaniline. Plant Journal 34, 485–493. Lu YP, Li ZS, Rea PA (1997) AtMRP1 gene of Arabidopsis encodes a glutathione S-conjugate pump: Isolation and functional definition of a plant ATP-binding cassette transporter gene. Proceedings of the National Academy of Sciences of the United States of America 94, 8243–8248. Lutz KA, Knapp JE, Maliga P (2001) Expression of bar in the plastid genome confers herbicide resistance. Plant Physiology 125, 1585–1590. Manabe Y, Tinker N, Colville A, Miki B (2007) CSR1, the sole target of imidazolinone herbicide in Arabidopsis thaliana. Plant and Cell Physiology 48, 1340–1358. Martinoia E, Grill E, Tommasini R, Kreuz K, Amrhein N (1993) ATP-dependent glutathione S-conjugate export pump in the vacuolar membrane of plants. Nature 364, 247–249. Matthews JM, Holtum JA, Liljegren DR, Furness B, Powles SB (1990) Cross-resistance to herbicides in annual ryegrass (Lolium rigidum): I. properties of the herbicide target enzymes acetyl-coenzyme a carboxylase and acetolactate synthase. Plant Physiology 94, 1180–1186. Menendez J, DePrado R (1996) Diclofop-methyl cross-resistance in a chlorotoluron-resistant biotype of Alopecurus myosuroides. Pesticide Biochemistry and Physiology 56, 123–133. Mentewab A, Stewart CN Jr. (2005) Overexpression of an Arabidopsis thaliana ABC transporter confers kanamycin resistance to transgenic plants. Nature Biotechnology 23, 1177–1180. Messner B, Thulke O, Schaffner AR (2003) Arabidopsis glucosyltransferases with activities toward both endogenous and xenobiotic substrates. Planta 217, 138–146. Milligan AS, Daly A, Parry MAJ, Lazzeri PA, Jepson I (2001) The expression of a maize glutathione S-transferase gene in transgenic wheat confers herbicide tolerance, both in planta and in vitro. Molecular Breeding 7, 301–315. Morant M, Bak S, Moller BL, Werck-Reichhart D (2003) Plant cytochromes: tools for pharmacology, plant protection and phytoremediation. Current Opinion in Biotechnology 14, 151–162. Murgia I, Tarantino D, Vannini C, Bracale M, Carravieri S, Soave C (2004) Arabidopsis thaliana plants overexpressing thylakoidal ascorbate peroxidase show increased resistance to araquat-induced photooxidative stress and to nitric oxide-induced cell death. Plant Journal 38, 940–953. Nelson DR, Schuler MA, Paquette SM, Werck-Reichhart D, Bak S (2004) Comparative genomics of rice and Arabidopsis. Analysis of 727 cytochrome genes and pseudogenes from a monocot and a dicot. Plant Physiology 135, 756–772.

160

WEEDY AND INVASIVE PLANT GENOMICS

Nemat Alla MM, Hassan NM (2006) Changes of antioxidants levels in two maize lines following atrazine treatments. Plant Physiology and Biochemistry 44, 202–210. Nutricati E, Miceli A, Blando F, De Bellis L (2006) Characterization of two Arabidopsis thaliana glutathione S-transferases. Plant Cell Reports 25, 997–1005. Ohkama-Ohtsu N, Zhao P, Xiang C, Oliver DJ (2007) Glutathione conjugates in the vacuole are degraded by gammaglutamyl transpeptidase GGT3 in Arabidopsis. Plant Journal 49, 878–888. Padegimas L, Shul’ga OA, Skriabin KG (1994) Creation of transgenic plants Nicotiana tabacum and Solanum tuberosum, resistant to the herbicide phosphinothricin. Molekuliarnaia Biologiia (Moskva) 28, 437–443. Persans MW, Wang J, Schuler MA (2001) Characterization of maize cytochrome monooxygenases induced in response to safeners and bacterial pathogens. Plant Physiology 125, 1126–1138. Poppenberger B, Berthiller F, Lucyshyn D, Sieberer T, Schuhmacher R, Krska R, Kuchler K, Glossl J, Luschnig C, Adam G (2003) Detoxification of the Fusarium mycotoxin deoxynivalenol by a UDP-glucosyltransferase from Arabidopsis thaliana. Journal of Biological Chemistry 278, 47905–47914. Qi X, Bakht S, Qin B, Leggett M, Hemmings A, Mellon F, Eagles J, Werck-Reichhart D, Schaller H, Lesot A, Melton R, Osbourn A (2006) A different function for a member of an ancient and highly conserved cytochrome P450 family: from essential sterols to plant defense. Proceedings of the National Academy of Sciences of the United States of America 103, 18848–18853. Ramel F, Sulmon C, Cabello-Hurtado F, Taconnat L, Martin-Magniette ML, Renou JP, El Amrani A, Couee I, Gouesbet G (2007) Genome-wide interacting effects of sucrose and herbicide-mediated stress in Arabidopsis thaliana: novel insights into atrazine toxicity and sucrose-induced tolerance. BMC Genomics 8, 450. Reade JPH, Milner LJ, Cobb AH (2004) A role for glutathione S-transferases in resistance to herbicides in grasses. Weed Science 52, 468–474. Robineau T, Batard Y, Nedelkina S, Cabello-Hurtado F, LeRet M, Sorokine O, Didierjean L, Werck-Reichhart D (1998) The chemically inducible plant cytochrome CYP76B1 actively metabolizes phenylureas and other xenobiotics. Plant Physiology 118, 1049–1056. Ross J, Li Y, Lim E, Bowles DJ (2001) Higher plant glycosyltransferases. Genome Biology 2, reviews 3004.1–3004.6. Sandermann H Jr., Schmitt R, Eckey H, Bauknecht T (1991) Plant biochemistry of xenobiotics: isolation and properties of soybean O- and N-glucosyl and O- and N-malonyltransferases for chlorinated phenols and anilines. Archives of Biochemistry and Biophysics 287, 341–350. Scarponi L, Quagliarini E, Del Buono D (2006) Induction of wheat and maize glutathione S-transferase by some herbicide safeners and their effect on enzyme activity against butachlor and terbuthylazine. Pest Management Science 62, 927–932. Schroder P, Scheer CE, Diekmann F, Stampfl A (2007) How plants cope with foreign compounds. Translocation of xenobiotic glutathione conjugates in roots of barley (Hordeum vulgare). Environmental Science and Pollution Research International 14, 114–122. Schulz B, Kolukisaglu HU (2006) Genomics of plant ABC transporters: The alphabet of photosynthetic life forms or just holes in membranes? FEBS Letters 580, 1010–1016. Siehl DL, Castle LA, Gorton R, Keenan RJ (2007) The molecular basis of glyphosate resistance by an optimized microbial acetyltransferase. Journal of Biological Chemistry 282, 11446–11455. Skipsey M, Cummins I, Andrews CJ, Jepson I, Edwards R (2005) Manipulation of plant tolerance to herbicides through co-ordinated metabolic engineering of a detoxifying glutathione transferase and thiol cosubstrate. Plant Biotechnology Journal 3, 409–420. Smith AP, DeRidder BP, Guo WJ, Seeley EH, Regnier FE, Goldsbrough PB (2004) Proteomic analysis of Arabidopsis glutathione S-transferases from benoxacor- and copper-treated seedlings. Journal of Biological Chemistry 279, 26098–26104. Stokstad E (2004) Biotechnology. A new tack on herbicide resistance. Science 304, 1089. Tan S, Evans R, Singh B (2006) Herbicidal inhibitors of amino acid biosynthesis and herbicide-tolerant crops. Amino Acids 30, 195–204. Windsor B, Roux SJ, Lloyd A (2003) Multiherbicide tolerance conferred by AtPgp1 and apyrase overexpression in Arabidopsis thaliana. Nature Biotechnology 21, 428–433. Wohlleben W, Arnold W, Broer I, Hillemann D, Strauch E, Puhler A (1988) Nucleotide sequence of the phosphinothricin N-acetyltransferase gene from Streptomyces viridochromogenes Tu494 and its expression in Nicotiana tabacum. Gene 70, 25–37. Yin XL, Jiang L, Song NH, Yang H (2008) Toxic reactivity of wheat (Triticum aestivum) plants to herbicide isoproturon. Journal of Agricultural and Food Chemistry 56, 4825–4831. Yuan JS, Galbraith DW, Dai SY, Griffin P, Stewart CN Jr. (2008) Plant systems biology comes of age. Trends in Plant Science 13, 165–171.

MECHANISMS OF NON-TARGET-SITE HERBICIDE RESISTANCE

161

Yuan JS, Tranel PJ, Stewart CN Jr. (2007) Non-target-site herbicide resistance: a family business. Trends in Plant Science 12, 6–13. Yun MS, Yogo Y, Miura R, Yamasue Y, Fischer AJ (2005) Cytochrome P-450 monooxygenase activity in herbicideresistant and -susceptible late watergrass (Echinochloa phyllopogon). Pesticide Biochemistry and Physiology 83, 107–114. Zhang O, Riechers DE (2004) Proteomic characterization of herbicide safener-induced proteins in the coleoptile of Triticum tauschii seedlings. Proteomics 4, 2058–2071.

11

A Herbicide Defense Trait That Is Distinct From Resistance: The Evolutionary Ecology And Genomics Of Herbicide Tolerance Regina S. Baucom

Introduction

Weed resistance to herbicides presents one of the greatest current economic challenges to agriculture. The worldwide cost of herbicide resistance has been estimated to be as high as $8 billion per year (D. Pimentel, personal communication). This number is a result of herbicide expenditures by agriculturists as well as the yield loss of crops that must compete with weed infestations for water and nutrients. More than 315 biotypes of weed are known to be resistant to herbicides (Heap 2008), with some biotypes of weed species exhibiting resistance to multiple herbicide classes (Neve et al. 2004). The problems that herbicide-resistant weeds pose to agriculture has prompted the cataloguing of herbicide resistance cases, which are published and easily accessible to the public (www.weedscience.org). This necessary and important focus on cases of herbicide resistance has had an unintended consequence: there has been little consideration for other types of plant defense to herbicides, namely, herbicide tolerance. The lack of distinction between these two alternative, perhaps redundant, plant defense traits is partly to blame. In addition, there are specific criteria for diagnosing herbicide resistance, but no such criteria for establishing that a population is tolerant (as defined herein) to herbicides. This is an unfortunate omission, because a complete understanding of how weed populations can adapt to herbicide application requires that all defense traits be examined. This chapter discusses tolerance as an important weediness characteristic and how its investigation would aid studies of resistance to herbicides. It first expands upon the concepts of herbicide resistance vs. tolerance in weed science, with special emphasis on cases of resistance or tolerance to the herbicide glyphosate. Next it examines the theoretical and empirical contributions of the field of evolutionary ecology on the concept of tolerance, and provides a basic understanding of how to measure tolerance as well as some of the problems and pitfalls encountered in its estimation. Finally, this chapter discusses how weed studies can benefit from this established evolutionary framework, and it discusses the potential for investigating the genomics of herbicide tolerance and tolerance-related traits. Resistance Versus Tolerance In Weed Science Historic And Current Definition Of Herbicide Resistance

The first case of herbicide resistance is credited to a biotype of Senecio vulgaris able to survive the field application of simizine (Radosevich and Appleby 1973). Use of the term “resistance” to describe this phenomenon was prompted by the discovery of a novel site-of-action mutation in a chloroplast gene (Holt 1992). The term was later expanded to include field-selected weed biotypes that were resistant because of a variety of mechanisms, such as differential absorption 163

164

WEEDY AND INVASIVE PLANT GENOMICS

or translocation, with the important addendum that to be considered resistant, the weed should be able to withstand the field dose of herbicide (Gressel 1985). Currently, a weed species is considered resistant if its response to herbicide application can fulfill both a scientific definition and a field definition of resistance. The scientific definition of resistance states that it is “a genetically inherited statistical difference in herbicide response between two weed populations of the same weed species,” (Heap 2005), whereas the field definition requires that the resistant population survive the recommended rate of herbicide under normal field conditions (Heap 2005). These definitions are used in tandem to diagnose a case of resistance; it is thought that use of only the scientific definition would lead to cases of weed resistance that were of little practical importance. For example, weed species that exhibit a low level of resistance would likely be considered resistant under the scientific definition but not the field definition. Thus, the stringency of the current criteria ensures that only weeds that pose a significant threat to the agriculturist are included. While the current practical definition of resistance is well established, the scientific definition of resistance is neither uniform nor consistently applied (Radosevich et al. 1997). A commonality among the various definitions is that resistance is achieved when a once susceptible population has become resistant following an increase of resistant individuals in the population (Holt and LeBaron 1990). By this definition, herbicide resistance is an evolved state rather than due to natural variation in a weed species for herbicide response. Unfortunately, use of this definition, while also based in practicality, has the potential to exclude cases of herbicide resistance that may be in the initial states of resistance evolution.

Historic And Current Definition Of Herbicide Tolerance

Tolerance is a term that has been used loosely in weed science (Radosevich et al. 1997). One use of the term has been in reference to a weed’s ability to withstand herbicide application through mechanisms other than an alteration at the herbicide’s site of action (Holt 1992), such as differences in herbicide uptake and translocation or differences in plant metabolism and herbicide detoxification (Warwick 1991). Tolerance has also often been used to describe the natural variability of a weed species to respond to an herbicide, generally prior to selection by the herbicide; i.e. sometimes regarded in terms of physiology rather than genetics (LeBaron and Gressel 1982). However, the response to herbicide that the plants exhibit is not well characterized in terms of a specific trait or traits, and the phrase “natural variation” is often applied to describe the differences in response to herbicides, either within or between species. One of the consequences of a vague definition of tolerance is that it is often confused with resistance. Both traits generally refer to the vegetative response of a weed, or weed population, after damage from herbicide application; their distinction has rested on either their respective mechanism-of-action or whether or not the weed population has experienced selection from herbicide application. The delineation between the two traits has been blurred by considering weed species as resistant if they are able to withstand herbicide application through mechanisms previously ascribed to tolerance, such as reduced herbicide translocation (Holt and LeBaron 1990; Holt 1992). Further confusing this distinction is the suggestion that tolerance and resistance represent a continuum of response to herbicide application, such that tolerance might be defined as the lowest level of resistance (Radosevich et al. 1997). It is entirely possible that the natural variation of response to herbicides that is present in a weed population prior to selection by an herbicide can be selected upon, eventually leading the weed population

ECOLOGY AND GENOMICS OF HERBICIDE TOLERANCE

165

to a state that would be considered resistant. Unfortunately, very little is currently known about the initial stages of resistance evolution in weed populations (Maxwell and Mortimer 1994), and the lack of clarity between the terms “resistance” and “tolerance” perhaps further confuses an understanding of this process. Therefore, some researchers have chosen not to use the term tolerance, because it is likely more reflective of low-level and non-selected resistance (LeBaron and Gressel 1982). The lack of a clear definition for herbicide tolerance is interesting, because the notion and study of tolerance originated in the agricultural literature with the acknowledgment that crops could produce mature grain after exhibiting symptoms of disease (Cobb 1894). Reginald Painter later described herbivore-tolerant plants as those that are able to grow and reproduce or repair injury in spite of hosting a population of herbivores equal to the size that would damage a susceptible host (Painter 1951). Thus, in the crop science literature, the term “tolerance” has been used to describe the fitness response of plants given attack by either herbivores or pathogens. Unlike resistance, tolerance does not prevent the plant from being damaged, but allows it to compensate for damage (Mauricio 2000). Tolerance defined in this way refers to a specific plant trait and as such has garnered much attention in both crop sciences and in the field of evolutionary ecology.

Unified Definition Of Tolerance

Defining tolerance in the same manner as it is considered in both crop sciences and evolutionary ecology would greatly aid the distinction between herbicide resistance and tolerance. In these fields, tolerance refers to maintenance of fitness following damage, and is often considered in a quantitative genetics framework. In this light, tolerance is a trait that can be directly selected upon by the plant breeder for an increase in the mean value (Painter 1951) or indirectly selected for by the application of herbicide on a weed population (Baucom and Mauricio 2004), much the same as any trait that exhibits underlying genetic variation. One of the benefits of studying tolerance as a fitness response after herbicide application is that this approach complements herbicide resistance studies. It has recently been acknowledged that an emphasis on the process of resistance evolution in weed science is needed, rather than focusing primarily on the description of new cases of resistance (Neve 2007). An important first step toward this end would be to more clearly delineate between tolerance and resistance; defining tolerance as the fitness of plants following damage by herbicide would fulfill this need. Next, studying tolerance in the framework previously laid out by evolutionary ecologists would shed light on the evolutionary trajectory of tolerance in weed populations. The researcher could address the following pertinent questions: •

• • •

What is the relative contribution of tolerance vs. resistance to herbicide in a weed population, and do populations of some weed species persist in the face of herbicide application because of a high level of tolerance? Are tolerance and resistance correlated traits such that selection on one indirectly selects for the other? Are there fitness costs of tolerance, and can potential costs slow the rate of tolerance evolution? What role do weed shifts, or the process whereby a new herbicide regime selects for different weed communities than previously established (Radosevich et al. 1997), play in the process of tolerance and resistance evolution?

166

WEEDY AND INVASIVE PLANT GENOMICS

Case Studies

There are indications in the literature that herbicide tolerance, as the fitness response following herbicide damage, is present and variable in weed populations. Ipomoea purpurea, the common or tall morning glory, exhibits genetic variation within southeastern U.S. populations for tolerance to glyphosate applied at a rate of 1.12 kg ai ha−1, while Ipomoea hederacea, ivy-leaf morning glory, exhibits variation among populations at the same herbicide rate (Baucom and Mauricio 2008a). This means that two different species of morning glory that are treated with the recommended dose of glyphosate recover and produce seed after experiencing damage by the herbicide. Furthermore, at least for I. purpurea, the ability to tolerate glyphosate exhibits genetic variation and thus can increase in the population if glyphosate is consistently applied. Similarly, Abutilon theophrasti, velvetleaf, has been reported as suffering a 90% reduction in biomass after the application of 0.84 kg ai ha−1 of glyphosate in comparison to untreated plants, yet the plant still produces viable seed after damage by the herbicide (Hartzler and Battles 2001). Other species, such as field bindweed (Convolvulus arvensis), Asiatic dayflower (Commelia communlus), birdsfoot trefoil (Lotus corniculatus), tropical spiderwort (Commelina benghalensis), and common lambsquarters (Chenopodium album), are currently considered “naturally resistant” to glyphosate (Cerdeira and Duke 2006), and should be investigated for being tolerant to glyphosate. It remains to been seen how prevalent tolerance, and genetic variation for tolerance, are in weed populations and across weed species, since the practice of putting weed fitness after herbicide application into this context is relatively unexplored. Fortunately, the framework for studying the evolutionary dynamics of tolerance has already been well established by researchers in evolutionary ecology.

Tolerance In Evolutionary Ecology Theoretical And Empirical Contributions

The study of weed resistance to herbicides presents a classic example of human-mediated evolution. It is a phenomenon similar to the adaptation of insects to pesticides, plants to heavy metals, or bacteria to antibiotics. An excellent review of the basic evolutionary process as applied to herbicide resistance in weeds can be found in Maxwell and Mortimer (1994). These authors point out that the precursors for resistance evolution in a weed population are the same as the requirements for the evolution of any trait—the presence of heritable genetic variation and natural selection. These same requirements apply to tolerance. Work from evolutionary ecologists has expanded on the evolutionary forces affecting tolerance both in terms of the theory behind its prevalence in natural systems and empirical verifications of these forces (Simms and Triplett 1994; Simms 2000; Stowe et al. 2000). These researchers have also discussed the various ways in which tolerance can be measured as well as problems inherent in its estimation. While these studies have focused on tolerance to herbivory and pathogens, the concepts explored in this work can be applied to the phenomenon of herbicide tolerance. Plant Tolerance To Herbivory. Much of the theory underlying the dynamics of tolerance has been prompted by attempts to explain variable levels of defense in nature (Simms and Rausher 1987) and from the acknowledgement that tolerance could be an alternative defensive strategy to resistance (Simms and Triplett 1994; Fineblum and Rausher 1995). Researchers studying

ECOLOGY AND GENOMICS OF HERBICIDE TOLERANCE

167

the evolution of defense against herbivory suggested that the level of defense within a population results from a compromise between the benefits of reduced herbivory and the cost of diverting resources to defense from other fitness-enhancing functions (Simms and Triplett 1994). This concept was formalized by the graphical model of Simms and Rausher (1987), which showed that an intermediate, optimal level of defense should be maintained by stabilizing selection as allocation to defense increases, the benefit asymptotes, and cost corresponds to a linear or concave-upward function of the allocation to defense. This work, as a series of papers, (Simms and Rausher 1987; Rausher and Simms 1989; Simms and Rausher 1989) provided a methodology to test for the costs and benefits of defense. This method involves growing individuals of known genetic relationships, such as family lines, in both an environment with and without the herbivores. The level of defense is then measured on individuals in the presence of herbivores, as is the fitness of each individual from both environments. A negative relationship between the average level of defense and fitness for each family in the environment without herbivores is indicative of a fitness cost of defense, whereas a positive relationship between fitness and defense in the presence of herbivores indicates that there is a benefit of being well defended. The logic behind the methodology is as follows: in the presence of herbivores, an individual’s fitness is the result of the benefit of being able to defend oneself from attack, which is mediated by the costs of allocating resources from other fitness-enhancing functions. In the absence of herbivory, the benefit of being able to defend oneself is not present such that only the cost is expressed (Simms and Triplett 1994). This method has been applied to the study of tolerance across multiple biological systems. In general, most investigated plant species exhibit significant fitness costs of tolerance (Mauricio 1997; Tiffin and Rausher 1999; Fornoni and Núñez-Farfán 2000; Pilson 2000; Stinchcombe 2002; Weinig et al. 2003a; Fornoni et al. 2004). These reports also often find evidence of a benefit of tolerance, albeit in some cases only at high levels of herbivory (Mauricio 1997; Tiffin and Rausher 1999; Pilson 2000; Fornoni et al. 2004). This means that tolerance to herbivores is costly to the plant to maintain and is under negative selection in the absence of herbivores. However, in the presence of herbivory, there is often a net benefit of being tolerant, such that individuals that allocate more resource to tolerance should be selected for over time and the mean level of tolerance in the population will increase, given that the abundance of herbivores remains stable. Plant Tolerance To Abiotic Stresses. This type of cost/benefit analysis has also been applied to the evolution of tolerance to other types of stresses—namely, tolerance to the herbicide glyphosate in I. purpurea (Baucom and Mauricio 2004) and to frost damage in the wild radish Raphanus raphanistrum (Agrawal et al. 2004). Fitness costs were present in both studies, yet only tolerance to glyphosate in I. purpurea appeared to exhibit a fitness benefit, because it was under positive directional selection in the presence of glyphosate. In the case of frost tolerance, the authors suggested that the costs of tolerance outweighed the benefits in that study year (Agrawal et al. 2004). Negative Correlation Between Tolerance And Resistance. Another type of trade-off potentially responsible for the level of defense in natural populations is that of a negative correlation between tolerance and resistance. The idea behind this type of trade-off is that both traits can function as alternative, redundant defense strategies (Simms and Triplett 1994). This hypothesis suggests that natural selection would not act to increase resistance if fitness is not reduced in the population because of a high level of tolerance. Thus, highly tolerant genotypes would not experience selection on resistance, and vice versa (Fineblum 1991). It would be expected,

168

WEEDY AND INVASIVE PLANT GENOMICS

then, that a negative genetic correlation should exist between tolerance and resistance if they are redundant in terms of their effect on fitness, given that both involve fitness costs (Fineblum 1991). The empirical support for this phenomenon is rather low, however, with only a few studies finding trade-offs between resistance and tolerance to herbivory (Fineblum and Rausher 1995; Stowe 1998; Pilson 2000; Fornoni et al. 2003), and other studies finding no evidence of a trade-off between the two traits (Mauricio et al. 1997; Tiffin and Rausher 1999; Stinchcombe and Rausher 2002; Valverde et al. 2003; Leimu and Koricheva 2006). This has led researchers to investigate the conditions under which a mixed pattern of defense allocation, or the presence of both resistance and tolerance, might or might not evolve in plant populations (reviewed in Núñez-Farfán et al. 2007). Presently there is only one study that considers the potential for a negative genetic correlation between tolerance and resistance to an herbicide. However, work from the morning gloryglyphosate system suggests that resistance and tolerance are negatively genetically correlated in I. purpurea, with tolerance defined as the fitness response of plants after glyphosate application, and resistance estimated as the proportion of the plant exhibiting vegetative damage (Baucom and Mauricio 2008b). This finding means that if either trait is selected for by the application of the herbicide, its evolution will be constrained by the negative correlation between the two traits. This finding also suggests that the fitness response of a weed following herbicide application and the vegetative response represent alternative plant defense strategies. Obviously, the definition of each type of plant response is important, because the definition will influence how the trait is measured.

How To Estimate Tolerance

Tolerance is most often estimated as a norm of reaction in response to variable levels of damage, or as the difference in fitness of related individuals planted in two different environments. The norm of reaction is a graphical depiction of how fitness, or any trait, changes in response to a changing environment. The most important aspect of measuring tolerance is that the fitness of groups of related individuals, such as replicates of a genotype (i.e. clones), a family, or population, must be assessed in more than one environment (Simms 2000). This is because if tolerance were considered in only one environment, it would be impossible to determine what aspect of the shared environment caused any potential difference in fitness among different genotypes. However, by experimentally varying an environmental factor, such as salt concentration, and assessing plant fitness from replicates of a genotype planted along a gradient of salt concentrations, one can estimate the level of tolerance that the genotype exhibits in response to the environmental factor—in this case, salt. Thus, to estimate tolerance as a norm of reaction, replicates of individuals from a group of related individuals are planted in multiple environments that vary for the level of the environmental stressor, and fitness is assessed for each individual belonging to that group (Simms 2000; Stowe et al. 2000). Operationally, tolerance, as a norm of reaction, is estimated as the slope of the line in a regression of fitness on damage. In Figure 11.1, genotype A exhibits incomplete, or low level, tolerance, whereas genotype B would be considered perfectly tolerant, because its fitness response does not change across environments. Figure 11.1 thus exhibits two important aspects about assessing tolerance in a study population—tolerance is estimated as the slope of a line from a regression of fitness on damage, and, genetic variation for tolerance is present in the study population given that the genotypes are responding differently to increased damage, i.e. their slopes vary. Genotype C, although expressing higher fitness across all environments in

Relative fitness

ECOLOGY AND GENOMICS OF HERBICIDE TOLERANCE

169

B D C A 0

25

50

75

100

Damage environment Figure 11.1. A fitness reaction norm of four idealized plant genotypes exposed to different levels of damage, with 0 being no damage to 100 being completely damaged. The level of tolerance is indicated as the slope of the reaction norm, with the height of the reaction norm indicating general vigor. Genotype A displays a low level of tolerance, B would be considered perfectly tolerant to damage, and genotype C exhibits the same level of tolerance as genotype A, but is more vigorous. Genotype D overcompensates at low levels of damage, but its fitness is low at high levels of damage, suggesting that a quadratic term would best explain its tolerance level.

comparison to genotype A, exhibits the same level of tolerance as genotype A, given that the slopes are similar. Thus, differences in the height of the respective reaction norms are from differences in overall plant vigor rather than tolerance (Stowe et al. 2000). The second method of estimating tolerance likely has more application to the study of herbicide tolerance. Similar to the above method, the fitness of replicates of related individuals is assessed, but this time using only two environments—one with and one without the selective agent. In this case, the level of tolerance is estimated as the difference in the average fitness of related individuals in the two different environments. For any given genotype, family, or population, the level of tolerance is estimated as fitness in the presence of the herbicide minus fitness in the absence of herbicide, or Wd − Wu. The relative level of tolerance among the groups of related individuals can still be visualized as a reaction norm, but only the two environments are plotted on the x-axis (as in Figure 11.2). Comparison of the level of tolerance among groups is similar to the previous method, with negative values indicating imperfect tolerance, values near zero indicating complete tolerance, and positive values indicating that the plants overcompensate for damage (Tiffin and Rausher 1999; Tiffin and Inouye 2000). Genetic Variation For Tolerance. Whereas visualizing tolerance using a reaction norm approach indicates the relative differences among families or populations in the level of tolerance, the presence of genetic variation for tolerance is determined using an analysis of variance. The relative fitness of each individual is used as the dependent variable with the treatment environment and the family line of the individual is the independent variable. A significant family term indicates that the families vary for fitness, whereas a significant interaction between family and treatment indicates that significant genetic variation exists for tolerance, since the families vary in their fitness response to the different environments (Simms and Triplett 1994; Mauricio et al. 1997; Tiffin and Rausher 1999; Stinchcombe and Rausher 2002; Baucom and Mauricio 2004).

170

WEEDY AND INVASIVE PLANT GENOMICS 0.4 0.3

Relative fitness

0.2 0.1 0.0 –0.1 –0.2 –0.3 –0.4

0

1 Treatment environment

Figure 11.2. The relationship between relative fitness and glyphosate treatment for thirty-two maternal lines of the common morning glory, Ipomoea purpurea. On the x-axis, 0 = glyphosate absent and 1 = glyphosate present at a rate of 1.12 kg ai ha−1. The residuals of fitness were used after the effects of block were removed. Slopes of the lines represent each maternal line’s level of tolerance.

The Problems In Estimating Tolerance. Measuring tolerance has its own brand of difficulties. In the above scenarios, tolerance is estimated as a linear term in the regression of fitness on damage, or by quantifying fitness in two extremes of the environment, i.e. the presence or absence of the selective agent. In reality, the linear term in the regression of fitness on damage might not be the best estimate of tolerance for all families, because some families might overcompensate given slight damage, but show a decrease in fitness at higher levels of damage (Mauricio 2000; Pilson 2000; Simms, 2000). This would produce a significant quadratic term in the regression, indicating that a polynomial regression best explains the relationship of fitness and damage for a family rather than the linear term (Figure 11.1, D). Estimating the area under the non-linear function is a method that has been suggested to take this into account (Pilson 2000), but has yet to receive much empirical treatment. Further, the non-linear response cannot be determined by estimating tolerance using only two environments. This is likely an acceptable caveat when assessing tolerance to an herbicide because, in the practical sense, most weed populations are either treated or not treated with herbicides. However, this could become an important issue when considering the evolution of tolerance in weed populations on the edges of crops, or those that do not receive the full application of herbicide spray. It has recently been shown that the application of low levels of herbicides can lead to an elevated rate of resistance evolution to herbicides in Lolium rigidium (Neve and Powles 2005); thus, studying the evolution of tolerance to reduced rates of herbicide application, simulating herbicide drift, warrants attention. Another difficulty inherent in the measurement of tolerance lies in how the researcher quantifies fitness. The fitness of an individual is comprised of its total reproductive output as well as the relative viability and the survival of its progeny, and the choice of fitness measure can be somewhat subjective. These choices can have experimental consequences. Whereas most studies use seed production as an estimate of total plant fitness, the level of tolerance can vary between female vs. male plant fitness parameters (reviewed in Strauss and Agrawal 1999). Because of some of these problems, researchers have attempted to study the traits that are

ECOLOGY AND GENOMICS OF HERBICIDE TOLERANCE

171

potentially responsible for tolerance, such as re-growth characters (Juenger and Bergelson 2000). The study of such traits provides a complementary framework for assessing plant tolerance, and is perhaps the best route for studying the genomics of tolerance. Tolerance Traits And The Genomics Of Tolerance

The operational definition of tolerance described above is extremely useful because it integrates all possible tolerance mechanisms into one measure and allows the researcher to investigate the evolutionary and ecological forces affecting tolerance (Tiffin 2000). However, relying solely on the operational definition limits an understanding of the various traits or mechanisms responsible for tolerance, because these traits (and genes) could potentially vary among weed species, or even among weed populations. Therefore, investigating the types of tolerance traits underlying the operational measure of tolerance, as well as the genomic architecture of tolerance and tolerance traits, is the next step in understanding tolerance in both natural and human-mediated ecosystems. Tolerance Traits

Tolerance traits are defined as those that increase fitness after plant damage (Juenger and Bergelson 2000). Traits such as the release of lateral dormant buds and increases in photosynthetic rate and regrowth, branch production, branch height, and phenology (Juenger et al. 2000; Tiffin 2000; Weinig et al. 2003a) have all been investigated as putative tolerance traits. Branch production and phenological changes following damage were correlated to tolerance in Ipomopsis aggregata (Juenger and Bergelson 2000). Delayed senescence and the production of additional and taller inflorescence branches were shown to be correlated with tolerance in Arabidopsis thaliana (Weinig et al. 2003a). These studies were again concerned with the evolution of tolerance to herbivory in natural systems. Unfortunately, most studies investigating traits related to herbicide tolerance are likely reporting on the phenomenon I would term low-level, non-selected resistance. For example, “tolerance” in birdsfoot trefoil, as measured by the percent change in fresh weight of plants treated with 0.5 kg ai ha−1 of glyphosate, was not found to be correlated to plant size and the number of crown buds producing regrowth, which were potential tolerance traits (Boerboom et al. 1990). It is likely that the biomass change examined in this study is a trait more closely linked to resistance, and the potential regrowth measures they investigated would likely have been positively correlated to tolerance as measured by fitness, had fitness been assessed. Other mechanisms of herbicide tolerance that have been suggested are the differential absorption of herbicide, differences in epicuticular wax, or herbicide translocation differences (Owen and Zelaya 2005); however, again, these traits are thought to be mechanisms responsible for tolerance as it is defined in the weed science literature, rather than as an aspect of plant fitness following herbicide application. It remains to be seen how each of these types of traits, either the regrowth and plant architecture traits that have received treatment in the herbivory literature, or the traits hypothesized to be responsible for herbicide defense, are correlated to the operational measure of tolerance. Given that tolerance is estimated as the fitness response of plants growing in more than one environment, one might wonder what environment putative tolerance traits should be measured in—the control environment, or the environment with damage present? It is possible that both environments could provide different pieces of the tolerance puzzle. Increased photosynthetic

172

WEEDY AND INVASIVE PLANT GENOMICS

capacity following damage could be correlated to tolerance; however, it would be informative to know if highly tolerant plants also exhibited a higher photosynthetic rate in the control environment compared to less tolerant plants. Investigating putative tolerance traits in both environments can provide information about how their adaptive value changes with damage, and thus, how each trait minimizes the effect of damage on fitness. Once the relative contributions of each type of trait to tolerance are understood, the genomics of such traits, as well as tolerance itself, can be considered.

A Quantitative Trait Loci Approach

As yet, only one study has attempted to investigate the genomics of tolerance and tolerance traits. In a field study assessing the quantitative trait loci (QTL) architecture of herbivore defense traits, it was concluded that tolerance was conferred by many different loci, each having a small effect (Weinig et al. 2003b), because there were no significant QTLs for tolerance, but significant among-family heterogeneity for tolerance. Given that tolerance is an estimate based in fitness, it is not too surprising that tolerance is due to many loci, because fitness is a very complex trait and likely controlled by many genes (Mauricio 2001). Before this conclusion can be drawn, however, more studies investigating tolerance QTL are needed. Studies from the crop science literature have mapped tolerance traits, and generally find that these traits can be explained by anywhere from one to nine QTL (Alam and Cohen 1998; Agrama et al. 2002; Soundararajan et al. 2004).

Genomics Of Tolerance

An understanding of the genomics of weediness traits, including herbicide tolerance and even resistance, is in its infancy (Basu et al. 2004). In fact, little is known about the basic genetics underlying herbicide tolerance, in terms of how it is inherited (nuclear vs. plastid genome) and how many genes underlie the trait. This is important to understand because most cases of herbicide resistance conform to the “resistance paradigm,” in that resistance is generally inherited as a complete or incompletely dominant gene (Cousens and Mortimer 1995; Neve 2007). If tolerance is a quantitative trait, its evolutionary trajectory will be different than that of herbicide resistance, and management efforts to control tolerance evolution will need to be different than those aimed at herbicide resistance. At this point, a QTL approach to finding genomic regions correlated to herbicide tolerance and tolerance traits, following in the same vein as the Weinig et al. (2003) study, is perhaps the best approach for understanding the genetic architecture of tolerance to herbicide.

Why Again Should We Focus On Tolerance, Tolerance Traits, And The Genomics Of Tolerance?

Tolerance to herbicide represents another weed defensive strategy that can promote the survival of weed populations in agricultural systems, thus reducing the effectiveness of herbicides. It is possible that weed populations have evolved separate types of defensive adaptations, such that one population might be resistant to a particular herbicide whereas another population of the same weed species could be highly tolerant to its application. Integrating the study of a

ECOLOGY AND GENOMICS OF HERBICIDE TOLERANCE

173

weed’s fitness response to an herbicide with studies of the vegetative response will provide more information not only on the relative contribution of different types of defense, but on the whole process by which a weed can become problematic to the agriculturist. Many questions remain to be answered about herbicide tolerance. For example, how can we be sure that tolerance is not just the retention of fitness in plants given a low-level of resistance? To answer this question, one would need to either show a negative genetic correlation between resistance and tolerance or experimentally produce genetic lines that exhibited the same level of herbicide resistance and then determine if variation for tolerance exists among the lines. Also, it would be informative to understand if tolerance is due to a mutation at the herbicide’s site of action or if the ability to tolerate herbicide damage is due to mechanisms that allow the plant to mitigate the effects of general damage, or damage from herbivores, pathogens, or other abiotic sources, such as drought (Yuan et al. 2007; see also Chapter 10). It is also imperative to understand the relative contribution of tolerance in the evolution of herbicide resistance. Does tolerance provide a temporal bridge that allows weed populations to survive at low frequency until a resistance gene appears, either through random mutation or gene flow? Are tolerance mechanisms separate from resistance mechanisms? Will the same herbicide rotation schemes designed to delay the evolution of herbicide resistance also mitigate tolerance evolution? Do different mechanisms underlie tolerance in different weed species, or different populations of the same weed species? Many exciting research opportunities exist in the study of herbicide tolerance evolution, and linking the tolerance phenotype to its genetic basis will further our understanding of how to best mitigate its effects. Acknowledgements

I thank R. Mauricio and V. Koelling for offering advice and help with the text, J.R. Stinchcombe for discussions about tolerance traits, and P. Tranel and S. Powles for providing critical feedback that greatly improved my thoughts on the topic. I also thank N. Stewart for the opportunity to be included in this important and interesting volume of work. References Agrama HA, Widle GE, Reese JC, Campbell LR, Tuinstra MR (2002) Genetic mapping of QTLs associated with greenbug resistance and tolerance in Sorghum bicolor. Theoretical and Applied Genetics 104, 1373–1378. Agrawal AA, Conner JK, Stinchcombe JR (2004) Evolution of plant resistance and tolerance to frost damage. Ecology Letters 7, 1199–1208. Alam SN, Cohen MB (1998) Detection and analysis of QTLs for resistance to the brown planthopper, Nilaparvata lugens, in a doubled-haploid rice population. Theoretical and Applied Genetics 97, 1370–1379. Basu C, Halfhill MD, Mueller TC, Stewart CN Jr. (2004) Weed genomics: new tools to understand weed biology. Trends in Plant Science 9, 1360–1385. Baucom RS, Mauricio R (2004) Fitness costs and benefits of novel herbicide tolerance in a noxious weed. Proceedings of the National Academy of Sciences of the United States of America 101, 13386–13390. Baucom RS, Mauricio R (2008a) The evolution of novel herbicide tolerance in a noxious weed: the geographic mosaic of selection. Evolutionary Ecology 22, 85–101. Baucom RS, Mauricio R (2008b) Constraints on the evolution of tolerance to herbicide in the common morning glory: Resistance and tolerance are mutually exclusive. Evolution 68, 2842–2854. Boerboom CM, Wyse DL, Somers DA (1990) Mechanism of glyphosate tolerance in birdsfoot-trefoil (Lotus corniculatus). Weed Science 38, 463–467. Cerdeira AL, Duke SO (2006) The current status and environmental impacts of glyphosate resistant crops: A review. Journal of Environmental Quality 35, 1633–1658. Cobb N (1894) Contributions to an economic knowledge of Australian rusts (Uredineae). Agriculture Gazette N.S.W 5, 239–250.

174

WEEDY AND INVASIVE PLANT GENOMICS

Cousens R, Mortimer AM (1995) Dynamics of Weed Populations. Cambridge University Press, Cambridge. Fineblum WL (1991) Genetic Constraints on the Evolution of Resistance to Host Plant Enemies. Dissertation, Duke University. Fineblum WL, Rausher MD (1995) Tradeoff between resistance and tolerance to herbivore damage in a morning glory. Nature 377, 517–520. Fornoni J, Núñez-Farfán J (2000) Evolutionary ecology of Datura stramonium: Genetic variation and costs for tolerance to defoliation. Evolution 54, 789–797. Fornoni J, Núñez-Farfán J, Valverde PL, Rausher MD (2004) Evolution of mixed strategies of plant defense allocation against natural enemies. Evolution 58, 1685–1695. Fornoni J, Valverde PL, Núñez-Farfán J (2003) Quantitative genetics of plant tolerance and resistance against natural enemies of two natural populations of Datura stramonium. Evolutionary Ecology Research 5, 1049–1065. Gressel J (1985) Herbicide tolerance and resistance: alteration of site of activity. In: Weed Physiology Duke SO, ed. p. 189. CRC Press, Inc., Boca Raton, FL. Hartzler RG, Battles BA (2001) Reduced fitness of velvetleaf (Albutilon theophrasti) surviving glyphosate. Weed Technology 15, 492–496. Heap I (2005) Criteria for Confirmation of Herbicide Resistant Weeds, pp. 1–4. http://www.weedscience.org/resgroups/ Detect%20Resistance.pdf. Heap I (2008) The International Survey of Herbicide Resistant Weeds (www.weedscience.com). Holt JS (1992) History of identification of herbicide resistant weeds. Weed Technology 6, 615–620. Holt JS, LeBaron HM (1990) Significance and distribution of herbicide resistance. Weed Technology 4, 141–149. Juenger T, Bergelson J (2000) The evolution of compensation to herbivory in scarlet gilia, Ipomopsis aggregata: Herbivoreimposed natural selection and the quantitative genetics of tolerance. Evolution 54, 764–777. Juenger T, Lennartsson T, Tuomi J (2000) The evolution of tolerance to damage in Gentianella campestris: natural selection and the quantitative genetics of tolerance. Evolutionary Ecology 14, 393–419. LeBaron HM, Gressel J (1982) Herbicide Resistance in Plants. John Wiley and Sons, New York. Leimu R, Koricheva J (2006) A meta-analysis of genetic correlations between plant resistances to multiple enemies. American Naturalist 168, E15–E37. Mauricio R (1997) Experimental manipulation of putative selective agents provides evidence of the role of natural enemies in the evolution of plant defense. Evolution 51, 1435–1444. Mauricio R (2000) Natural selection and the joint evolution of tolerance and resistance as plant defenses. Evolutionary Ecology 14, 491–507. Mauricio R (2001) Mapping quantitative trait loci in plants: Uses and caveats for evolutionary biology. Nature Reviews Genetics 2, 370–381. Mauricio R, Rausher MD, Burdick DS (1997) Variation in the defense strategies of plants: are resistance and tolerance mutually exclusive? Ecology 78, 1301–1311. Maxwell BD, Mortimer AM (1994) Selection for herbicide resistance. In: Herbicide Resistance in Plants: Biology and Biochemistry Powles SB, Holtrum JAM, eds. pp. 1–26. Lewis, Boca Raton, FL. Neve P (2007) Challenges for herbicide resistance evolution and management: 50 years after Harper. Weed Research 47, 365–369. Neve P, Powles S (2005) Recurrent selection with reduced herbicide rates results in the rapid evolution of herbicide resistance in Lolium rigidum. Theoretical and Applied Genetics 110, 1154–1166. Neve PB, Saldler J, Powles SB (2004) Multiple herbicide resistance in a glyphosate resistant Lolium rigidum biotype. Weed Science 52, 231–239. Núñez-Farfán J, Fornoni J, Valverde PL (2007) The evolution of resistance and tolerance to herbivores. Annual Review of Ecology and Systematics 38, 541–566. Owen MDK, Zelaya IA (2005) Herbicide-resistant crops and weed resistance to herbicides. Pest Management Science 61, 301–311. Painter RH, (1951) Insect Resistance in Crop Plants. Wiley and Sons, New York. Pilson D (2000) The evolution of plant response to herbivory: simultaneously considering resistance and tolerance in Brassica rapa. Evolutionary Ecology 14, 457–489. Radosevich SR, Appleby AP (1973) Relative susceptibility of two common groundsel (Senecio vulgaris) biotypes to 6 s-triazines. Agronomy Journal 65, 553–555. Radosevich SR, Holt J, Ghersa C (1997) Genetics and evolution of weeds. In: Weed Ecology, pp. 69–102. John Wiley and Sons, Inc., New York. Rausher MD, Simms EL (1989) The evolution of resistance to herbivory in Ipomoea purpurea: I. Attempts to detect selection. Evolution 43, 563–572. Simms EL (2000) Defining tolerance as a norm of reaction. Evolutionary Ecology 14, 563–570. Simms EL, Rausher MD (1987) Costs and benefits of plant defense to herbivory. American Naturalist 130, 570–581.

ECOLOGY AND GENOMICS OF HERBICIDE TOLERANCE

175

Simms EL, Rausher MD (1989) The evolution of resistance to herbivory in Ipomoea purpurea. 2. Natural selection by insects and costs of resistance. Evolution 43, 573–585. Simms EL, Triplett J (1994) Costs and benefits of plant response to disease: resistance and tolerance. Evolution 48, 1973–1985. Soundararajan RP, Kadirvel P, Gunathilagaraj K, Maheswaran M (2004) Mapping of quantitative trait loci associated with resistance to brown planthopper in rice by means of a doubled haploid population. Crop Science 44, 2214–2220. Stinchcombe JR (2002) Environmental dependency in the expression of costs of tolerance to deer herbivory. Evolution 56, 1063–1067. Stinchcombe JR, Rausher MD (2002) The evolution of tolerance to deer herbivory: modifications caused by the abundance of insect herbivores. Proceedings of the Royal Society of London Series B-Biological Sciences 269, 1241–1246. Stowe KA (1998) Experimental evolution of resistance in Brassica rapa: Correlated response of tolerance in lines selected for glucosinolate content. Evolution 52, 703–712. Stowe KA, Marquis RJ, Hochwender CG, Simms EL (2000) The evolutionary ecology of tolerance to consumer damage. Annual Review of Ecology and Systematics 31, 565–595. Strauss SY, Agrawal AA (1999) The ecology and evolution of plant tolerance to herbivory. Trends in Ecology and Evolution 14, 179–185. Tiffin P (2000) Mechanisms of tolerance to herbivore damage: what do we know? Evolutionary Ecology 14, 523–536. Tiffin P, Inouye BD (2000) Measuring tolerance to herbivory: Accuracy and precision of estimates made using natural versus imposed damage. Evolution 54, 1024–1029. Tiffin P, Rausher MD (1999) Genetic constraints and selection acting on tolerance to herbivory in the common morning glory Ipomoea purpurea. American Naturalist 154, 700–716. Valverde PL, Fornoni J, Núñez-Farfán J (2003) Evolutionary ecology of Datura stramonium: equal plant fitness benefits of growth and resistance against herbivory. Journal of Evolutionary Biology 16, 127–137. Warwick SI (1991) Herbicide resistance in weedy plants: Physiology and population biology. Annual Review of Ecology and Systematics 22, 95–114. Weinig C, Stinchcombe JR, Schmitt J (2003a) Evolutionary genetics of resistance and tolerance to natural herbivory in Arabidopsis thaliana. Evolution 57, 1270–1280. Weinig C, Stinchcombe JR, Schmitt J (2003b) QTL architecture of resistance and tolerance traits in Arabidopsis thaliana in natural environments. Molecular Ecology 12, 1153–1163. Yuan JS, Tranel PJ, Stewart CN Jr. (2007) Non-target site herbicide resistance: a family business. Trends in Plant Science 12, 6–13.

12

The Genomics Of Plant Invasion: A Case Study In Spotted Knapweed Amanda K. Broz and Jorge M. Vivanco

Why Study Invasive Plant Genomics?

It is estimated that more than 25,000 non-native plant species have been introduced into the United States (Pimentel et al. 2000). Although some of these species represent valuable food crops, such as corn, rice, and wheat, other introduced plants have become some of the nation’s worst weed problems, threatening the viability of food production and the community structure of native ecosystems. Weeds are responsible for agricultural losses of nearly $20 billion per year in the United States alone (Basu et al. 2004), and it is estimated that the economic cost of all invasive plant species exceeds $34.5 billion per year (Pimentel et al. 2000). Although the monetary cost of these weeds is astounding, their negative impact on the biodiversity of native ecosystems may be of greater concern. It is estimated that introduced plant species invade more than 700,000 hectares per year of United States wildlife habitat (Pimentel et al. 2000). In addition, hundreds of native species are being threatened by exotic invasive plants. For instance, nearly half of Hawaii’s 1750 native plant species are endangered, and more than 200 species endemic to Hawaii are thought to have become extinct due to displacement by invasive species (Pimentel et al. 2000). Although the economic and biological issues associated with invasive weeds are recognized internationally, the majority of research efforts have focused on ecological consequences of plant invasion, leaving the genomic basis of these consequences relatively unexplored. This is due at least in part to the paucity of molecular tools available for invasive plants and other weedy species (Basu et al. 2004). In addition, most genomics efforts have focused to a greater extent on crop plants, as opposed to their weedy relatives. However, as genomics resources become more affordable and available to scientists, an understanding of the genetic basis of invasiveness becomes increasingly possible. It has been widely observed that many weedy species possess certain characteristics that enhance their ability to succeed in the environment. These traits include high output of seeds, long distance seed dispersal, competitive ability, discontinuous dormancy, and rapid growth, to name a few (Basu et al. 2004). Although these traits are often associated with weeds, many of them are actually desirable for breeding in crop plants. Thus, plant breeders and geneticists have begun to use genomics resources to understand agronomically important phenotypes, including increased yield, pathogen resistance, and competitive ability of crops. In addition, the sequencing of both the Arabidopsis and rice genomes have led to the identification and characterization of many biochemical pathways involved in growth, fecundity, defense, and other important plant phenotypes. Because many agronomic weeds are close relatives of crop plants, these studies are likely to provide clues concerning weed phenotypes and their underlying gene networks. Because genes conferring important traits are often conserved throughout plant lineages, it is possible that these studies of crops and their close relatives could provide insight into the biology of other weeds. However, the overlap of gene sequence and gene function between plant species is often not reliable, and thus it will be important to develop genomics resources for multiple plant families (Shimamoto and Kyozuka 2002). 177

178

WEEDY AND INVASIVE PLANT GENOMICS

Invasive species present an interesting case study in both weed research and evolutionary biology. Although members of a given species will, in theory, possess the same capacity to exhibit fundamental weedy traits, in their native range these plants are generally seen as benign, whereas in the invasive or introduced range they are extremely problematic pests. Many of the plants currently listed as invasive species were originally introduced as ornamentals or food and fiber crops (Pimentel et al. 2000); therefore, these plants likely underwent years of human selection in the native range that may have conferred an advantageous genotype to founding populations in the invasive range. Many other invasive plant species were not purposefully introduced to North America, however, and do not appear to have undergone human selection for hardiness, competitive ability, or other traits. There is some evidence to suggest that environmental adaptation and evolution play an important role in the success of these invasive species, and ecological hypotheses of plant invasion have been developed based on this evidence (Lee 2002; Blossey and Notzold 1995; Callaway and Aschehoug 2000). Molecular marker studies have revealed differences in population structure and diversity between the native and introduced ranges for many invasive plant species (Bossdorf et al. 2005; Lee 2002). However, there has been little work examining how the genetic diversity of these populations influences the gene expression and protein accumulation that are ultimately responsible for the phenotypic characteristics that allow for invasive success. These questions could begin to be addressed by creating genomics resources for invasive species, and coupling them with molecular markers, mapping studies, and characterization of important phenotypic traits. Genomics investigations may then be able to provide links between plant gene expression and ecological hypotheses of plant invasion into new environments. Multiple research challenges face investigators interested in the genomics of plant invasion. A major limitation in this field is a lack of sequence information for the organisms of interest; however, resources are currently being developed for a few target plant species (Anderson et al. 2007; Broz et al. 2007a). Recently an expressed sequence tag (EST) library was developed from seven populations of the invasive plant spotted knapweed (Centaurea maculosa), an extremely problematic weed in western North America (Broz et al. 2007). Seventy-seven percent of the 4,423 unique sequences in this library were able to be annotated based on homology searches to known sequences, and they represented a wide variety of gene ontology (GO) categories. Sequence information and annotation of this library is publicly available through the PLAN database (http://bioinfo.noble.org/plan/, project 30060; He et al. 2007). Additional EST sequences from invasive spotted knapweed and yellow starthistle (Centaurea solstitialis) have been deposited in GenBank (National Center for Biotechnology Information) as part of a sequencing effort by the Compositae Genome Project at the University of California, Davis (http://compgenomics.ucdavis.edu/). Spotted knapweed is an interesting case study, in that it occurs in both diploid and tetraploid forms in its native Eurasian environment, but only the tetraploid form is invasive in North America (Ochsmann 2001; Treier et al. 2009). In addition, there is evidence to suggest that allelopathy plays a role in the invasive success of this weed. This is one of the best researched examples of an allelopathic invasive plant for which genomics resources are beginning to be developed; this chapter presents an in-depth discussion of the weed and the opportunities and challenges in reconciling genomics approaches with ecological hypotheses of plant invasion. Spotted Knapweed Life History

Spotted knapweed (Centaurea maculosa Lam.) is a particularly aggressive invasive weed in the northwestern United States, infesting more than 4.7 million acres in Montana alone (Mauer

GENOMICS OF PLANT INVASION: SPOTTED KNAPWEED CASE STUDY

179

et al. 2001). A Eurasian native, spotted knapweed is proposed to have been accidentally introduced on both coasts of North America in the late 1800s as a contaminant of alfalfa seed. The weed has since expanded its range to all but three states in the continental U.S. (plants.usda. gov), and is prominent in western and central Canadian provinces. Spotted knapweed inhabits a variety of environments in the native and invaded ranges, and is common in disturbed areas, pastures, prairies, rocky slopes, and rangeland. In Eurasia it is not considered very weedy or problematic; however, in many areas of western North America spotted knapweed has invaded native ecosystems, displacing native species and forming near monocultures. In addition, the weed increases water runoff, leading to erosion (Lacey et al. 1989), and reduces forage for livestock and wildlife (Thompson 1996). In the native range, taxonomists have identified at least two forms of the weed, although the nomenclature is often confusing and inconsistent. The diploid form (Centaurea stoebe L. spp. stoebe, synonym: C. maculosa L. spp maculosa) and the tetraploid form (Centaurea stoebe L. spp. micranthos (Gugler) Hayek, C. biebersteinii, commonly known as C. maculosa Lam), occur in separate and mixed stands in the native range (Treier et al. unpublished data). Recently a mixed stand of diploids and tetraploids was identified in Canada (Treier et al., unpublished data), but otherwise only the tetraploid form has been found in North America (Ochsmann 2001; Treier et al. 2009). Karyotyping and other molecular techniques have been used to identify the two forms because they are indistinguishable based on morphological characters (Oschmann 2001). The diploid form of the weed contains eighteen chromosomes (Powell et al. 1974) with a 2C DNA content near 3.6 pg, based on measures from closely related Centaurea species of the same chromosome number (Grime et al. 1985). This translates to an estimated genome size of 1,800 Mbp, more than ten times larger than the genome of the model species Arabidopsis thaliana (125 Mbp), which contains more than 25,000 genes (www.arabidopsis.org). Both diploid and tetraploid forms of spotted knapweed are tap-rooted, short-lived perennials of the aster family, but the diploid is monocarpic and the tetraploid polycarpic (Oschmann 2001). Both are capable of remaining in rosette form for many years before bolting (Mauer et al. 2001; Freville et al. 1998). In addition to flowering multiple years, the tetraploid is capable of producing multiple flowering stems with up to fifteen capitula each (Mauer et al. 2001), whereas the diploid produces only one stem (Ochsmann 2001). Flowers bloom from late summer to fall, producing up to thirty-five seeds per capitula (Mauer et al. 2001). Seeds germinate in both the spring and early fall and are capable of remaining viable in the soil for up to ten years. Gravity is the major plant mechanism for seed dispersal; however, human and animal dispersal also play an important role in seed transport. Both forms of the weed are insect pollinated and predominately out-crossing. Genetic diversity studies of plants in the native and invaded range suggest that there have been multiple introductions of spotted knapweed (Hufbauer and Sforza 2008). A large amount of allelic richness was found in both native and invasive populations using chloroplast haplotype sequence data, and there was some evidence to indicate possible introgression of chloroplasts between taxa. Spotted knapweed was found to share multiple haplotypes with another Centaurea species (diffuse knapweed, C. diffusa) that also occurs as a diploid and tetraploid in the native range, suggesting hybridization between the two species may have occurred. In addition, studies of Centaurea species in southern France suggest that the rare, endemic C. corymbosa and the more widespread C. maculosa spp albida are likely derived from spotted knapweed (C. maculosa spp maculosa), and have undergone ecological specialization (Freville et al. 1998). These studies provide an interesting context for studying spotted knapweed in both the native and invasive ranges, in that the species contains a range of genetic diversity from which new taxa are able to evolve.

180

WEEDY AND INVASIVE PLANT GENOMICS

In the native range, the tetraploid form of the weed occupies a more extensive area than the diploid and is able to survive at greater latitudinal clines (Ochsmann 2001). In the invaded range, the tetraploid weed occupies niches climatically distinct from those where it exists in the native range, which may be a result of both ecological and evolutionary change (Broennimann et al. 2007). In addition, the tetraploid can withstand dense vegetation (Ochsmann 2001) and drier environments (Broennimann et al. 2007; Treier et al. 2009), which is presumed to be the main reason why only the tetraploid form is invasive in North America, while neither form is considered invasive or even predominant in its native Eurasian habitat. Accurate taxonomical assessment and karyotyping of plants from both the native and invasive range is often lacking in ecological studies of invasive weeds, which can be problematic for drawing appropriate conclusions (Bossdorf et al. 2005). Plants of different ploidy levels often occupy distinct environments or exhibit differences in morphology or life history (Bossdorf et al. 2005), as has been noted to some extent in spotted knapweed (Broennimann et al. 2007). Although correct identification of a plant and its chromosome number should be considered in all types of studies, it is absolutely necessary to understand these plant characteristics when using genomics techniques. This provides an additional challenge for researchers interested in investigating plant populations from both the native and invasive range.

Allelopathy And The Novel Weapons Hypothesis

Ecologists have long been interested in identifying key factors involved in the invasive success of exotic plant species, but the mechanism of invasion is still not well understood. Multiple non-exclusive hypotheses have been developed to explain plant invasion of new environments, all of which are partially supported by empirical evidence. Some of these hypotheses can be logically extended to make predictions about genetic inheritance, gene expression, and protein accumulation, and could be further researched using genomics resources. In particular, hypotheses that propose changes in resource allocation or rapid evolution of plant species upon introduction could be investigated using molecular techniques in addition to more traditional physiological and ecological techniques. The release of secondary metabolites by plants may also facilitate invasion and could be best studied using metabolomic and genomic resources. Recently, Callaway and Aschehoug (2000) generated the novel weapons hypothesis (NWH) to explain plant invasion into new environments. This hypothesis is based primarily on results from investigations of allelopathy (discussed below) in invasive plants, particularly Centaurea species. This hypothesis suggests that plants come to the new environment equipped with chemical or biochemical weapons that have a greater negative effect against plants in the invaded range than similar species in the native range. These weapons give the invader a competitive advantage in the new environment, because they are proposed to exert strong phytotoxic effects against other plants. The NWH has also been expanded to encompass possibilities of rapid evolution, called the allelopathic advantage against resident species hypothesis (AARS). AARS suggests that if plant invaders gain a competitive advantage through the use of novel weapons in the invaded range they will evolve to have greater concentrations of these weapons than populations in the native range. Thus, invasive populations should be better chemically defended and thus better competitors against other plant species than their native counterparts.

GENOMICS OF PLANT INVASION: SPOTTED KNAPWEED CASE STUDY

181

Allelopathy: Definition

Allelopathy is generally defined as any direct or indirect effect of a plant compound on another plant or microbe; however, current literature typically refers to allelopathy as a chemical inhibition of one plant by another (Willis 2004). Plants use a variety of mechanisms to release compounds into the surrounding environment that may mediate interactions between plants and other organisms (Singh et al. 2003). Decomposing litter from plant tissues, volatile release, and root exudation can all contribute novel chemical compounds to the environment that may play important roles in plant competition, succession, diversity, dominance, and ultimately the community composition (Steeghs et al. 2004; Bertin et al. 2003). There is considerable evidence to suggest that direct communication between plants is mediated by plant-derived compounds or allelochemicals (Bertin et al. 2003; Chamberlain et al. 2001; Palmer et al. 2004; Singh et al. 1999; Weir et al. 2004; Baerson et al. 2008), and that non-resource-driven interactions can play important roles in structuring plant communities (Hierro and Callaway 2003).

Challenges In The Study Of Allelopathy

Although allelopathy has long been suspected to be important in both agronomic and native ecosystems (Weston and Duke 2003), understanding the importance and relative influence of allelopathy in direct plant-plant interactions and plant community structure still remains a challenge in both crop plants and weeds. New, more sensitive techniques are available for monitoring litter decomposition, exudation in the rhizosphere, and volatilization of plant compounds; however, generating the necessary proof for these interactions continues to be technically difficult. Many problems in the study of allelopathy stem from the fact that laboratory studies do not adequately reflect the natural environment, because they are typically conducted in Petri dishes or in liquid media. In these experiments, it is often difficult to know if the proposed allelochemical is having direct effects on another plant, or if it is merely changing the pH or osmotic potential of the growing media (Wardle et al. 1998). Additionally, many studies use extremely large amounts of compounds that do not have the observed phytotoxic effects at lower, ecologically relevant amounts. In greenhouse experiments plant material is often crushed or extracted and then mixed into soil to determine the potential for allelopathy. It is unlikely that this technique mimics the natural environment, because a variety of factors modify leaf litter composition and integration into field soil. Activated carbon has traditionally been used to ameliorate the effects of allelopathic root exudates in greenhouse experiments or the field. However, activated carbon nonspecifically adheres to multiple compounds in soil, altering soil chemistry, and thus it cannot provide direct evidence of allelopathy (Wardle et al. 1998). Drawing conclusions from studies conducted in the field is also difficult, because multiple chemical, biological, and environmental factors are at work. For instance, an allelochemical may complex with nutrients in the soil, modifying the soil chemistry and reducing the growth of other plants through indirect interactions. In the same way, an allelochemical may have a large effect on the soil microbial community, such that the change in microbial composition (as opposed to a direct effect of the allelochemical) is responsible for observed negative effects on surrounding plants. Because of these challenges, the idea of allelopathy has continually met with much skepticism and debate. Widely accepted techniques to study this phenomenon are still lacking;

182

WEEDY AND INVASIVE PLANT GENOMICS

therefore, understanding the direct and indirect influence of allelochemicals remains problematic. However, genomics resources and plant transformation techniques may be of considerable assistance in this area, as discussed later in this chapter.

Evidence For Allelopathy In Spotted Knapweed

Allelopathic interactions have long been proposed to be responsible for the aggressive behavior of spotted knapweed (Suchy and Herout 1962; Fletcher and Renney 1963; Locken and Kelsey 1987) in western North America. Extracts from leaves, stems, seeds, and roots of both spotted and diffuse knapweed were found to have a negative impact on germination of barley and lettuce seeds (Fletcher and Renney 1963). The sesquiterpene lactone cnicin was later identified as the major chemical constituent of knapweed leaves and in bioassays inhibited growth of a variety of plant species, including spotted knapweed, at concentrations over 1 mg/mL (Locken and Kelsey 1987). Previous analyses of cnicin found it had strong antimicrobial activity (Cavallito and Bailey 1949), and it was postulated that both of these properties could influence the invasive success of knapweed species. However, analysis of soil samples in knapweedinfested areas recovered only trace amounts of cnicin, suggesting that this compound was not biologically available at phytotoxic concentrations, and thus was unlikely to have ecologically relevant allelopathic effects (Locken and Kelsey 1987). Although bioassays had suggested that other chemical compounds in knapweeds are phytotoxic (Fletcher and Renney 1963), identification of these potential allelochemicals has not been pursued. However, a renewed interest in the potential allelopathic effects of knapweed species was initiated by experiments performed by Ragan Callaway at the University of Montana, Missoula. Callaway and colleagues observed both diffuse and spotted knapweed coexisting with multiple plant genera in their native range of Eurasia. However, in the invaded range the knapweeds effectively excluded many of these same plant genera from communities where it had gained a foothold, presumably through some sort of competitive advantage (Callaway and Ridenour 2004). Because the plant community composition was similar between the native and invaded range, it seemed unlikely that knapweeds were exploiting a novel niche in North America. In addition, Callaway and colleagues were well aware that multiple biological control agents had been introduced to North America in an attempt to curb the knapweed problem, suggesting that release from co-evolved enemies might not be the major factor in their invasive success. In an attempt to understand their observation, the researchers grew related Eurasian and North American bunchgrass species in competition with diffuse knapweed and included activated carbon, which adsorbs organic compounds, in half of the treatments. In sand alone, diffuse knapweed reduced the growth of North American plant species nearly 70% more than their Eurasian congeners; however, the negative effect of diffuse knapweed on North American species was partially ameliorated by the addition of activated carbon (Callaway and Aschehoug 2000). This suggested that allelopathy, through root exudation, might be involved in the interactions between the plants, and led Callaway and Aschehoug (2000) to propose the novel weapons hypothesis. Further greenhouse experiments using spotted knapweed showed similar results to those obtained from its close relative, diffuse knapweed. Spotted knapweed was grown in competition with Idaho fescue (Festuca idahoensis), a North American native bunchgrass, with and without activated carbon (Ridenour and Callaway 2001). Biomass and root elongation of Idaho fescue were significantly inhibited when grown with a spotted knapweed competitor versus a

GENOMICS OF PLANT INVASION: SPOTTED KNAPWEED CASE STUDY

183

conspecific competitor, and the addition of activated carbon reduced this inhibitory effect (Ridenour and Callaway 2001). However, spotted knapweed outperformed Idaho fescue with and without carbon, suggesting that resource competition and allelopathy may both influence success of the invasive. To more directly assess the possibility that root exudates of spotted knapweed were responsible for the observed inhibition in growth of native grasses, Bais et al. (2002) grew spotted knapweed plants in sterile cultures of liquid media under laboratory conditions and applied the media to a variety of target plants. All plants tested except spotted knapweed showed significant reductions in germination, root elongation, and shoot differentiation after fourteen days, providing additional evidence for the presence of an allelochemical in the root exudates (Bais et al. 2002). The phytotoxic activity was confined to one organic fraction which was found to contain a racemic mixture of the flavonoid (±)-catechin (Bais et al. 2002). Interestingly, the presence of (+)-catechin is relatively widespread in the plant kingdom, and catechins are well known for their antioxidant activity and potential human health benefits (Ju et al. 2007). The presence of (−)-catechin has also been documented in plants (Nahrstedt et al. 1987), but it appears to be much less common than its enantiomer. Using bioassays, it was determined that (−)-catechin is a phytotoxin, whereas (+)-catechin has some antimicrobial activity and is a weaker phytotoxin (Bais et al. 2002). Bais et al. (2002) found phytotoxic effects of (–)-catechin using concentrations as low as 100 ug/mL on plants tested in liquid culture. In addition, catechin was reported to be present in spotted knapweed fields at a concentration of 300 ug g-1 (Bais et al. 2002), suggesting the putative allelochemical was present at ecologically relevant amounts. However, Blair et al. (2005, 2006) found soil catechin amounts that ranged from zero to less than 0.11 ppm (0.11 ug g-1) in spotted knapweed infested soils, and reported that soil moisture had a large effect on catechin stability. In an attempt to clarify the presence of catechin in soils, Perry et al. (2007) conducted a broad study sampling spotted knapweed−infested soils on a monthly basis for more than one year, excluding the winter months. Catechin was found very rarely among a range of sites and dates sampled; however, a spike of catechin (650 ug g-1) was found in one site at one particular time point. These results suggest that catechin may only be present in ecologically relevant amounts at certain times of the year or in certain field sites, which would add to the complexity of studying allelopathy under natural conditions. The inconsistencies in these studies could be the result of multiple factors, including seasonal or environmental variation in knapweed root exudation, differences in soil physical and chemical properties, variation in soil microbial communities, or variation in the seasonal presence of plant competitors. Differences in sampling techniques, soil storage and extraction, or other methodological issues could also help to explain the large variation in results. In addition, phenolic compounds such as catechin tend to oxidize rapidly (Appel 1993), making the quantification of soil catechin difficult and potentially unreliable. New techniques using resins to bind root exudates (Morse et al. 2000) are currently being examined as a better methodology for examining soil catechin concentrations in the field and in greenhouse experiments. Traditionally these experiments have been performed by extracting soils with organic solvents (Bais et al. 2002; Perry et al. 2007; Blair et al. 2005, 2006). It is also possible that genetic variation or differences in gene regulation between knapweed populations could influence the exudation of certain allelochemicals, including catechin. It is well known that both genetics and environmental conditions dictate the production of secondary metabolites in plants (Walker et al. 2003), and thus catechin concentrations could vary depending on the age of the plant and other outside factors. In fact, Locken and Kelsey (1987)

184

WEEDY AND INVASIVE PLANT GENOMICS

found that cnicin concentrations in spotted knapweed leaves varied seasonally, by plant part and over years. In addition, herbivory and other biotic stresses have been shown to induce production of defense-related secondary metabolites (Walker et al. 2003). Obviously, there are multiple possible explanations for the observed discrepancies in soil catechin concentrations. Understanding the regulation of allelochemical production in the plant, the stability of the allelochemical in the environment, and the potential effects of the allelochemical on other plants are still great challenges to researchers studying allelopathy. It is important to note here that although catechin is suspected to be responsible for the allelochemical activity of spotted knapweed, it is entirely possible that the plant produces other phytotoxic compounds. The in vitro culturing experiments in which catechin was first purified were conducted under highly controlled laboratory conditions. Liquid culturing of plants lends itself to studying root exudation, but has proven to be a poor representation of field conditions. Thus, the results of laboratory studies often do not directly reflect what happens in the outside environment. Ideally, new laboratory methodologies should be developed that reflect more natural field-like conditions and still allow for relatively simple chemical analyses. Potentially important conditions that are often overlooked in laboratory studies include the presence of plant competitors and soil microbes, varying nutrient conditions, and high incidence of UV radiation, among other factors.

Spotted Knapweed Allelopathy And Soil Microbial Communities

As discussed above, the proposed allelochemical catechin has only been found at ecologically relevant amounts in spotted knapweed infested soils at certain sites and at certain points in time. This could be due to multiple factors, including environmental characteristics such as resident soil microbial communities which may be able to utilize, sequester, or transform allelochemicals. Although there have been no studies that directly examine potential allelopathic effects of spotted knapweed on soil microbial communities, there is evidence to suggest that catechin has some antimicrobial properties (Bais et al. 2002) and that microbial elicitors can increase catechin exudation by spotted knapweed (Bais et al. 2002). Thus, there may be a link between spotted knapweed allelochemical exudation and soil microbial community composition that has an indirect impact on the native plant community (Bais et al. 2002). The link between spotted knapweed and North American soil microflora has been explored in common garden (Callaway et al. 2004c) and greenhouse studies (Marler et al. 1999). In both studies soil fungicide applications had no direct effect on either spotted knapweed or the native grass Idaho fescue when they were grown separately. However, fungicide application reduced spotted knapweed’s competitive advantage over the native grass when the two species were grown together. These results suggest that an indirect interaction between spotted knapweed, Idaho fescue, and the local soil fungi may cause a benefit for spotted knapweed (Marler et al. 1999; Callaway et al. 2004c). Interestingly, other reports provide evidence that spotted knapweed reduces both the amount and diversity of the soil fungal community in areas it has infested (Mummey and Rillig 2006; Broz et al. 2007b), and can cause a shift in the types of fungi associated with neighboring grasses (Mummey et al. 2005; Broz et al. 2007b). These reductions in fungal community composition appear to be more extreme in high density patches of spotted knapweed than in lower density patches (Broz et al. 2007b). Broz et al. (2007b) found that spotted knapweed favored phylotypes that were rare and often undetectable in the native soil microbial community. Other studies have found root exudates to be a major deter-

GENOMICS OF PLANT INVASION: SPOTTED KNAPWEED CASE STUDY

185

minant of soil fungal community composition in greenhouse studies (Broeckling et al. 2008), and thus, knapweed root exudation may be an important factor in the alteration of native soil microflora. Because spotted knapweed alters the native soil microbial community in areas it invades (Mummey et al. 2005; Mummey and Rillig 2006; Broz et al. 2007b), it is interesting to understand how this alteration affects subsequent generations of plants. Meiman et al. (2006) found that spotted knapweed emergence increased in knapweed-infested soils compared to soils taken from native plant communities; however, no significant difference was found in total plant biomass. Alternatively, spotted knapweed exhibited greater biomass in Montana soils preconditioned by spotted knapweed than in soils pre-conditioned by Idaho fescue (Callaway et al. 2004b). In addition, soil microbes from spotted knapweed’s native range were found to inhibit its growth to a greater extent than microbes from the invaded range, suggesting that the weed may have escaped native soil pathogens or has encountered beneficial microbes in the invaded range (Callaway et al. 2004b). The aforementioned studies provide evidence that spotted knapweed is altering North American soil microbial communities, and that this alteration may be to its benefit. However, the reason for this alteration remains unknown and cannot be directly associated with potential allelochemicals released by the weed. Future work could be aimed at identifying specific microbial species that are affected by spotted knapweed invasion, and examining how these species respond to potential allelochemicals. However, the interaction between knapweed, native species, and soil microbes appears to be extremely complex, and a multi-faceted approach will be needed to understand the relevance, if any, of allelochemicals in this system.

Genomics Resources And Approaches For Studying Spotted Knapweed

Although the remainder of this chapter focuses on genomics approaches to understanding invasion ecology of spotted knapweed, it might be more appropriately termed “transcriptomics” approaches. Currently there is no ongoing effort aimed at sequencing and assembling the genome of spotted knapweed, which makes studies of the genome itself extremely difficult. Because invasive spotted knapweed occurs predominately as a tetraploid and has a large genome, it is unlikely that efforts in this area will be initiated in the near future. In addition, the plant is predominately outcrossing, which provides great difficulties in generating recombinant inbred lines (RILs), which can be of great use in molecular mapping studies and classical genetic work (Broman 2005). Combining information from physical genome maps with gene expression profiling can reveal a wealth of information not only about the organism of interest and its response to various environmental conditions, but also about basic transcriptional control inside the cell. However, because full genome sequences and maps are not currently available for spotted knapweed, the following sections will focus on genomics resources in the form of EST libraries. Many other weedy plants provide genomics challenges similar to that of spotted knapweed, and are rarely thought of as ideal model organisms or crops, which are currently the target of most genome sequencing and mapping efforts. However, EST libraries are rapidly being developed for many of these non-model organisms and thus it is important to understand the opportunities and limitations that this type of resource provides. Currently the NCBI GenBank database contains nearly 45,000 ESTs derived from spotted knapweed. Since EST sequencing projects often contain large numbers of redundant transcripts, the number of unique gene sequences represented in the database appears to be closer to 28,200. A small-scale EST sequencing and annotation project by Broz et al. (2007a) found

186

WEEDY AND INVASIVE PLANT GENOMICS

only 18% redundancy in their library of nearly 5,000 sequences. However, this library was normalized to enrich for low abundance transcripts and reduce the redundant transcripts. Assembly of spotted knapweed sequences from the Compositae Genome Project at UC, Davis, revealed approximately 40% to 50% redundancy. Based on homology searches to genes in the NCBI database, Broz et al. (2007a) annotated nearly 3,400 of the unique sequences contained in the smaller EST library. These gene annotations are easily searchable and publicly available at http://bioinfo.noble.org/plan/, project 30060. Gene ontology assignments were also given to a large proportion of the annotated sequences in this library, which represented a wide variety of functional categories (Broz et al. 2007a). Many of the transcripts sequenced in the Compositae Genome Project were able to be annotated based on homology searches, but the information is not currently in a user-friendly format, making it difficult to navigate (http://compgenomics.ucdavis.edu, assembly information link). These two sequencing efforts represent a first step in examining spotted knapweed at a molecular level. Both sequence information and annotation can be extremely useful in determining candidate genes that may be involved in important physiological and ecological processes. In addition, the Compositae Genome Project has created EST libraries for a variety of plants in the aster family, including multiple varieties of sunflower (Helianthus) and lettuce (Lactuca). This sequence information has already led to some insights in the adaptation and hybridization of sunflowers (Rieseberg et al. 2007; Kane and Rieseberg 2007), and can potentially be used in comparative genomics studies to investigate species divergence between Centaurea species or between species within the aster family.

Allelopathic Potential of Catechin

Considering the generally controversial status of allelopathy research and the inconsistencies in field-level allelochemical concentrations from spotted knapweed−infested sites, a more direct approach is desirable for determining the relative influence of allelopathy in the success of this invasive weed. With the advent of a spotted knapweed EST library resource (Broz et al. 2007) and additional sequence information from the Compositae Genome Project, the potential for more directly understanding allelopathy in this invasive plant is greatly increased. However, many challenges remain in elucidating biosynthesis pathways and understanding how these are affected by both the environment and genetic makeup of the plant. Public databases contain sequence information on the putative genes involved in multiple metabolite biosynthetic pathways, and this information can potentially be used to identify candidate genes of similar function in spotted knapweed. However, many plant secondary metabolites are unique and many plant biosynthetic pathways remain uncharacterized, providing a greater challenge for researchers interested in unique secondary metabolic products. Catechin, the putative allelochemical identified in spotted knapweed root exudates, presents both opportunities and challenges in this area. This “novel weapon” occurs as a racemic mixture in root exudates, but the (−)-catechin form has greater phytotoxic activity than the (+)-catechin form (Bais et al. 2002). The synthesis pathway of the (+)-catechin form has been somewhat well-characterized (Tanner et al. 2003; Marles et al. 2003), because this compound occurs in many other plant species, and exhibits some properties beneficial to human health (Ju et al. 2007). However, the enzyme that produces (−)-catechin remains unknown. Multiple isoforms and conjugates of catechin occur in both plants and bacteria (Tanner et al. 2003). Leucocyanidin appears to be the major chemical precursor in the catechin bio-

GENOMICS OF PLANT INVASION: SPOTTED KNAPWEED CASE STUDY

187

synthetic pathway (Tanner et al. 2003; Marles et al. 2003). This molecule has a specific stereochemistry but is relatively unstable, and can be oxidized to the more stable, nonstereospecific cyanidin. The enzyme leucoanthocyanidin reductase (LAR) converts leucocyanidin to (+)-catechin (2R,3S trans), one form of catechin found in spotted knapweed root exudates. LAR is closely related to other isoflavone reductase enzymes, and is part of the reductase-epimerase-dehydrogenase (RED) protein family (Tanner et al. 2003). In an alternative pathway, the enzyme anthocyanidin reductase (ANR, also known as BANYLUS) appears to be responsible for the conversion of cyanidin to (+)-epicatechin (2R, 3R, cis), but the enzyme is very stereospecific and does not produce any form of (−)-catechin. Because cyanidin lacks any specific stereochemistry, it is presumed to be the most likely precursor of both the (−)-catechin (2S, 3R trans) and (+)-catechin (2R, 3S) found in spotted knapweed root exudates (C. Broeckling, personal communication). Catechins are the initiating monomers of proanthocyanidins (PAs), which are also known as condensed tannins. In Arabidopsis, mutants conferring the transparent testa (TT) phenotype have been used to identify genes involved in the PA synthesis pathway (Abrahams et al. 2002). Many genes were identified, including putative transporters, transcription factors, and pathway regulators (Abrahams et al. 2002; Tanner et al. 2003; Tian et al. 2007). Currently, seventeen genes involved in PA biosynthesis have been cloned and at least partially characterized (Routaboul et al. 2006). An Arabidopsis ANR has been the subject of further study and its gene sequence appears most closely related to dihydroflavone reductase (DFR), an upstream enzyme in the flavonoid synthesis pathway (Abrahams et al. 2002). Interestingly, Arabidopsis does not contain a known homolog to LAR. Further identification and characterization of PArelated genes in Arabidopsis may help provide clues to this biosynthetic pathway, but clearly research of multiple plant families will be essential in understanding both catechin and PA biosynthesis. A search of the Centaurea EST library revealed multiple genes that are likely involved in the general phenylpropanoid synthesis pathway including phenylalanine ammonia lyase, cinnamate 4-hydroxylase, 4-coumaryl-CoA ligase, flavonoid 3’-hydroxylase (F3’H), chalcone synthase, and chalcone isomerase (Broz et al. 2007). However, homologs of LAR, ANR, and DFR were unable to be identified. This is disappointing, but not entirely surprising as very few EST sequences in GenBank have been annotated as putative forms of these enzymes. As more sequence and biochemical information becomes available it may be possible to isolate specific domains within these sequences that are essential for their activity, and thus homology searches could focus on specific areas of a sequence as opposed to the entire sequence. However, even with clues from sequence information, putative biosynthesis genes must be cloned, characterized, and shown to have in planta activity to be of the most use in biological and ecological studies. Although putative genes directly responsible for catechin synthesis were not identified in homology searches, it may be possible to characterize genes further up in the flavonoid synthesis pathway that are indirectly responsible for catechin production. Monitoring the expression of these genes over time and in different genetic backgrounds of spotted knapweed plants could provide useful information related to developmental or seasonal catechin production or variability of production between different knapweed populations. If these upstream genes are shown to be good indicators of catechin production, they could be monitored in field situations as well as in the greenhouse. By designing appropriate field studies, it might be possible to resolve some of the discrepancies in reports of soil catechin concentrations, integrating plant genomics, and environmental explanations. Gene expression studies could also be used to test the AARS hypothesis, because

188

WEEDY AND INVASIVE PLANT GENOMICS

it posits that plant populations from the invasive range will evolve to produce more of their novel weapon (in this case catechin) than those populations from the native distribution. However, mere monitoring of gene expression will not be enough to provide conclusive evidence that catechin is the major allelochemical released by spotted knapweed, and it will say little about the evolution of this novel weapon or how it can influence plant-plant competition. It is desirable to have both plants that produce catechin and those that do not to unequivocally demonstrate the relative influence of allelopathy in spotted knapweed invasions, (Inderjit et al. 2006). This could be achieved by screening an extremely large amount of spotted knapweed plants from either mutant populations or different genetic backgrounds, or through biotechnology. Biotechnological techniques may be preferable because they are often the most direct way to create plants with a certain phenotype. Plant transformation technologies, such as Agrobacterium-mediated transformation, have been developed for many plant species, including sunflower and other members of Asteraceae (Lewi et al. 2006); thus these techniques can likely be applied to the study of spotted knapweed. Transgenic plants could be created to knock down or knock out expression of key genes involved in the synthesis or transport of catechin. Plant competition experiments could then be designed to determine the relative influence of these allelochemicals in the competitive competency of spotted knapweed. Ideally, these experiments could be performed in the field or in different environmental conditions, because soil chemistry, soil microbes, and other environmental factors may influence allelopathic effects (Wardle et al. 1998). It is difficult to know how other plant processes might be affected by knocking out genes related to catechin production. Ideally, the biosynthetic enzyme(s) directly involved in both (+)- and (−)-catechin production could be knocked out, resulting in little or no protein accumulation, no catechin exudation, and minimal additional effects. However, with the information currently available, only genes further up in the biosynthetic pathway could be directly targeted. Changes in these genes could potentially lead to other unexpected phenotypes, so plant transgenic lines would need to be studied carefully before solid conclusions concerning allelopathy could be drawn.

Allelopathic Potential Of Cnicin

A strategy similar to that outlined for catechin could be laid out to determine the significance of another potential allelochemical cnicin, a sesquiterpene lactone that accumulates to high levels in spotted knapweed tissues. Multiple Centaurea ESTs were annotated as components of the sesquiterpene lactone biosynthesis pathway (Broz et al. 2007a). One particular gene of interest showed extremely high similarity to a p450 enzyme involved in the synthesis of artemisinic acid, a sesquiterpene lactone isolated from Artemisia annua. This gene may be involved in the synthesis of cnicin or other sesquiterpene lactones in spotted knapweed. These types of chemicals are of great interest, not only because of their potential allelopathic effects, but also because of their potential medicinal uses. For instance, artemisinin is valued as an antimalarial drug (Towie 2006), and other sesquiterpene lactones have been shown to reduce inflammation and have positive effects when used in both cancer chemotherapy and cancer prevention (Zhang et al. 2005). In fact, cnicin has been shown to possess broadspectrum anti-fungal activity (Panagouleas et al. 2003) and may also have value as a pesticide (Meepagala et al. 2006). Cnicin is also thought to act as a deterrent to generalist herbivores of spotted knapweed, but as an oviposition stimulant for specialist herbivores (Landau et al. 1994), and is thus likely to play a role in individual plant fitness.

GENOMICS OF PLANT INVASION: SPOTTED KNAPWEED CASE STUDY

189

Transgenic plants could be created to knock out the expression of genes related to sesquiterpene lactone biosynthesis to understand the relevance of such compounds in respect to both allelopathy and plant defense against herbivores. However, in this case, it may be of greatest interest to create transgenic plants that are over-expressers of genes related to cnicin biosynthesis. If results continue to support the use of cnicin as an effective antifungal agent or pesticide, it may be desirable to generate a large amount of this chemical in planta because it cannot currently be cheaply and easily synthesized in the lab. Further, identification of the biosynthesis pathway for cnicin or other important secondary metabolites could be reconstructed in other, more laboratory-friendly organisms. For instance, the entire synthesis pathway of artemisic acid, the precursor of artemisinin, has recently been characterized and reconstructed in a yeast expression system (Ro et al. 2006). A combination of genomics, proteomics, and metabolomics resources will aid not only in an ecological understanding of these secondary metabolites, but may also hold promise for the directed synthesis of a variety of valueadded plant compounds.

Allelopathic Potential of Other Secondary Metabolites

Although catechin and cnicin have been the major chemicals of study in spotted knapweed, Centuarea species are known to synthesize a variety of secondary metabolites. Some of these metabolites may be involved in allelopathic interactions, defense response, or plant competition, and help facilitate the success of invasive knapweeds. These metabolites include polyacetylenes, flavonoids and their glycosides, anthocyanins, phenolics, lignans, coumarins, terpenoids, and steroidal compounds (Benderson 2003). Sequences from the spotted knapweed library were annotated as being involved in many of these secondary metabolite pathways (Broz et al. 2007a). Investigation into these secondary metabolites could prove useful in understanding their in planta function and ecological relevance. For instance, polyacetylenes are characteristic of many plants in the aster family, including the knapweeds. A variety of polyacetylenes have been identified that have allelopathic, antimicrobial, antifungal, or insecticidal activities; however, the ecological relevance of these compounds remains unknown for the most part (Minto and Blacklock 2008). Interestingly, a phytotoxic polyacetylene, thiophene, has been identified in Russian knapweed (Centaurea repens, recently renamed Acroptilon repens) root exudates (Stevens 1986). Thiophene was found at ecologically relevant concentrations in soils surrounding the plant, and is hypothesized to play a role in the invasive success of this weed (Stevens 1986). Sequences encoding putative acetylenases and fatty acid desaturases were identified in the spotted knapweed EST library (Broz et al. 2007a), and may prove useful in the study of polyacetylene biosynthesis, accumulation, and biological significance in knapweeds. A diverse group of flavonoid compounds are also found in Centaurea species, including catechins. Although these compounds may function as allelochemcials, they are generally implicated in plant response to both biotic and abiotic stress in a wide variety of species. As discussed previously, multiple genes potentially involved in the flavonoid synthesis pathway were identified in the spotted knapweed EST library. Characterizing these genes and their functions could not only lead to insight regarding the allelopathic potential of specific spotted knapweed flavonoids, but could also be used to investigate the roles of these compounds in plant defense against herbivores. The identification of genes directly involved in the biosynthetic pathways of putative allelochemicals is of great importance for understanding when and where these metabolites are

190

WEEDY AND INVASIVE PLANT GENOMICS

made. However, gene transcripts, their respective protein products, and the metabolites themselves are subject to further regulation and reorganization inside the cell. Thus, it may be beneficial to investigate genes involved in the transport and sequestration of a compound of interest. Eight percent of all annotations in the spotted knapweed EST library were designated as transporters, some of which are potentially involved in the transport of xenobiotics and other small molecules (Broz et al. 2007a). Identifying the physiological location, substrate specificity, expression, and activity of these transporters could be just as biologically important as characterizing genes involved in metabolite biosynthesis pathways. This is particularly relevant in the case of spotted knapweed and other plants which are thought to actively exude allelopathic compounds from their roots into their local environment through transporters (Badri et al. 2008).

Spotted Knapweed And The EICA Hypothesis

Because the relative influence of allelopathy in the invasive success of spotted knapweed remains undetermined, it is important to examine other ecological hypotheses of plant invasion in regard to this species. The enemy release hypothesis, originally developed by Darwin (1859) and expanded by Elton (1958), suggests that plants escape their co-evolved pathogens and herbivores upon introduction to a new environment, which allows them to increase in numbers as their population growth continues, unchecked by native enemies. Blossey and Notzwold (1995) expanded on enemy release, developing the evolution of increased competitive ability hypothesis (EICA), which suggests that after escaping their enemies, invaders would rapidly evolve to put fewer resources into defense and more resources into growth and reproduction. Thus, invasive populations should be more poorly defended against herbivores and pathogens than their native counterparts. In addition, invasive populations are predicted to be faster growing and larger, and have a higher reproductive capacity than populations from the native distribution. Only anecdotal evidence exists to suggest that invading spotted knapweed populations allocate more resources toward growth and fecundity and fewer resources toward defense, compared with native populations. One study of native diploids and invasive tetraploids found greater compensatory rooting intensity of the latter (Müller 1989), and recent work by Ridenour et al. (2008) suggests that invasive plants have greater biomass and thicker leaves than natives. Clearly, more physiological studies are needed in this area; however, genomics tools will likely be able to complement these studies. The EICA hypothesis predicts that invaders will evolve reduced defenses; by extension, it suggests that expression of defense-related transcripts would be reduced in invasive versus native populations of plants. Multiple sequences from the Centaurea EST library were annotated as genes potentially involved in plant defense response (Broz et al. 2007a). These represent chitinases, glucanases, PR proteins, transcription factors, and others. In addition, multiple genes involved in secondary metabolism were identified, and production of these chemicals is often involved in constitutive and induced defense responses in plants. These genes could be good candidates for investigating the EICA hypothesis in spotted knapweed. A quantitative PCR-based approach was initiated to investigate the feasibility of using spotted knapweed sequence information to better understand the EICA hypothesis (Broz et al. 2009). Measurements of gene expression were coupled with measures of plant performance and life cycle traits for the three spotted knapweed geo-cytotypes (native diploid, native tetraploid, and invasive tetraploid) grown together in a greenhouse environment. A preliminary analysis of gene expression for three unique PALs, one chitinase, and one glucanase revealed

GENOMICS OF PLANT INVASION: SPOTTED KNAPWEED CASE STUDY

191

lower expression of these defense-related transcripts in spotted knapweed populations from the invaded range than tetraploid populations from the native range (Broz et al. 2009), in accordance with the EICA hypothesis. Interestingly, measures of plant vegetative growth characteristics examined were not significantly different between geo-cytotypes; however, analysis of lifestyle characteristics suggests that the invasive tetraploid has the greatest fecundity of the three geo-cytotypes, although further analysis is needed to confirm this (Broz et al. 2009; Trier et al. 2009). Differences in gene expression were also noted between diploid and tetraploid populations, highlighting the importance of using appropriate plant types when examining a particular species in both the native and invasive range (Broz et al. 2009). This analysis demonstrates the utility of coupling measurements of gene expression with more traditional physiological measures of plant performance, and provides a starting point from which further studies can be developed. Invasive weeds, such as spotted knapweed, are potentially problematic for genomics studies because populations in both the native and invasive range contain a great deal of genetic diversity (Hufbauer and Sforza 2008). In this sense, spotted knapweed is much more difficult to work with than a typical crop species that has undergone a long history of human selection and contains only a fraction of the genetic diversity found in its wild relatives. Crop plants are generally more reliable to work with using genomics techniques because plants of a particular cultivar share a similar genetic background and provide reproducible phenotypes. Using these techniques to study invasive weeds may require more extensive sampling protocols and pooling to ensure that the genetic diversity and related differences in gene expression are accurately assessed. If a founder population for an invasive plant has been identified it may be appropriate to study that population in reference to the invasive populations. However, for plants such as spotted knapweed, no specific founder population has been identified (Hufbauer and Sforza 2008).

Other Applications Of Spotted Knapweed Genomics Resources

The EST libraries developed for spotted knapweed represent a first step toward investigating gene-specific expression in this species. This sequence-based information will allow the development of other resources such as microarrays that will permit a global view of gene expression in plants from different populations that are grown under multiple relevant environmental conditions. This will not only aid in our understanding of differences between populations from the native and invasive range, but will also allow researchers to investigate differences induced by chromosome doubling. The developmental and environmental regulation of chemical-based plant defenses such as catechin and cnicin could be monitored to investigate the NWH and the AARS hypotheses. In addition, the response of native and invasive spotted knapweed populations to multiple pathogens and herbivores could be examined at a molecular level, allowing tests of the EICA hypothesis and revealing important information about plant defense. This could be especially interesting in regard to studies of both specialist and generalist herbivores. Molecular characteristics related to plant defense could be extremely important to developing management strategies for the eradication of invasive weeds. Currently, a popular management strategy based on ideas of enemy release and EICA involves the introduction of biological control agents to the invaded range. Pathogens or specialist herbivores that do damage to the weed of interest in the native range are identified and brought to the invaded range with the idea that they will help keep the weed of interest under control. These introductions have met with success for some but not all plant invaders. The reasons for limited biocontrol success

192

WEEDY AND INVASIVE PLANT GENOMICS

might be, to some extent, related to inherent plant defense responses or potentially the rapid evolution of increased basal defenses in the invaded range. These defense characteristics could be investigated in invasive plant populations to more effectively determine the efficacy of proposed biocontrol agents. It is likely that the resources created for spotted knapweed can be used in investigations of other closely related invasive knapweeds that are problematic in North America, including diffuse knapweed (C. diffusa), yellow starthistle (C. solstitialis), squarrose knapweed (C. virgata), and Russian knapweed (Acroptilon repens). Potential allelochemicals have been identified in other invasive knapweeds (Stevens 1986; Quintana et al. 2008) so these plants may represent good candidates to help determine the relative influence of allelopathy in plant invasion. Comparative studies of these weeds also could be extremely interesting in that they exhibit somewhat different lifestyle characteristics. For instance, diffuse knapweed occurs predominately as a diploid in the invaded range, although both diploid and tetraploid forms have been identified in the native range (Marrs et al. 2007). It is thought that a hybrid between spotted and diffuse knapweed is also invasive in North America (Hufbauer and Sforza 2008), and the effects of this hybridization could be investigated at the level of gene expression. Understanding the genetic basis and molecular mechanisms for increased invasive capacities in plants is critical to determining how these exotic invasions occur. Genomics resources can complement traditional physiological and ecological studies aimed at understanding these plant characteristics and will be valuable tools for a variety of researchers. Conclusions

Invasive weeds represent a persistent threat to native ecosystem biodiversity and have large economic and social impacts. The creation of genomics resources for these weeds will aid in understanding the genetic basis for their invasive success. Understanding these mechanisms will provide insight into plant evolutionary biology, and may provide information that allows the development of more appropriate and effective management strategies. In plants that are proposed to be allelopathic, genomics resources might finally provide unequivocal evidence of the relative influence of allelopathy in the invasive success of these plants. Genomics resources will not replace more traditional physiological and ecological studies, but will supplement these techniques and advance our knowledge of invasion biology. References Abrahams S, Tanner GJ, Larkin PJ, Ashton AR (2002) Identification and biochemical characterization of mutants in the proanthocyanidin pathway in Arabidopsis. Plant Physiology 130, 561–576. Anderson JV, Horvath DP, Chao WS, Foley ME, Hernandez AG, Thimmapuram J, Lei L, Gong GL, Band M, Kim R, Mikel MA (2007) Characterization of an EST database for the perennial weed leafy spurge: an important resource for weed biology research. Weed Science 55,193–203. Appel HM (1993) Phenolics in ecological interactions: The importance of oxidation. Journal of Chemical Ecology 19, 1521–1552. Badri DV, Loyola-Vargas VM, Broeckling CD, De-la-Peña C, Jasinski M, Santelia D, Martinoia E, Sumner LW, Banta LM, Stermitz F, Vivanco JM (2008) Altered profile of secondary metabolites in the root exudates of Arabidopsis ATP-binding cassette transporter mutants. Plant Physiology 146, 762–71. Baerson SR, Dayan FE, Rimando AM, Nanayakkara, Liu C-J, Schroder J, Fishbein M, Pan Z, Kagan IA, Pratt LH, Cordonnier-Pratt M-M, Duke SO (2008) A functional genomics investigation of allelochemical biosynthesis in Sorghum bicolor root hairs. Journal of Biological Chemistry 283(6), 3231–3247.

GENOMICS OF PLANT INVASION: SPOTTED KNAPWEED CASE STUDY

193

Bais HP, Walker TS, Stermitz FR, Hufbauer RA, Vivanco JM (2002) Enantiomeric-dependant phytotoxic and antimicrobial activity of (±)-catechin. A rhizosecreted racemic mixture from spotted knapweed. Plant Physiology 128, 1173–1179. Basu C, Halfhill MD, Mueller TC, Stewart CN Jr. (2004) Weed genomics: new tools to understand weed biology. Trends in Plant Science 9, 391–398. Benderson W: Natural substances in the compositae: The Bohlmann Files at the BGBM (Freie Universitat Berlin, Botanischer Garten und Botanisches Museum) Berlin-Dahem, 2003. http://bohlmann.bgbm.org/bohlmann/. Bertin C, Yang X, Weston L (2003) The role of root exudates and allelochemicals in the rhizosphere. Plant and Soil 256, 67–83. Blair A, Hanson B, Brunk G, Marrs R, Westra P, Nissen S, Hufbauer, RA (2005) New techniques and findings in the study of a candidate allelochemical implicated in invasion success. Ecology Letters 8, 1039–1047. Blair A, Nissen SJ, Brunk GR, Hufbauer RA (2006) A lack of evidence for an ecological role of the putative allelochemical (±)-catechin in spotted knapweed invasion success. Journal of Chemical Ecology 32, 2327–2331. Blossey B, Notzwold R (1995) Evolution of increased competitive ability in invasive nonindigenous plants—a hypothesis. Journal of Ecology 83, 887–889. Bossdorf O, Auge H, Lafuma L, Roger WE, Siemann E, Prati D (2005) Phenotypic and genetic differentiation between native and introduced plant populations. Oecologia 144, 1–11. Broeckling CB, Broz AK, Bergelson J, Manter DK, Vivanco JM (2008) Root exudates regulate soil fungal community composition and diversity. Applied and Environmental Microbiology 74, 738–744. Broennimann O, Treier UA, Müller-Schärer H, Thuiller W, Peterson AT, Guisan A (2007) Evidence of climatic niche shift during biological invasion. Ecology Letters 10, 701–709. Broman KW (2005) The genomes of recombinant inbred lines. Genetics 169, 1133–1146. Broz AK, Broeckling CD, He J, Dai X, Zhao PX, Vivanco JM (2007a) A first step in understanding an invasive weed through its genes: an EST analysis of invasive Centaurea maculosa. BMC Plant Biology, 7, 25. Broz AK, Manter DK, Vivanco JM (2007b) Soil fungal abundance and biodiversity: another victim of the invasive plant Centaurea maculosa. ISME Journal 1, 763–765. Broz AK, Manter DK, Bowman G, Müller-Schärer H, Vivanco JM (2009) Plant origin and ploidy influence gene expression and life cycle in an invasive weed. BMC Plant Biology 9, 33. Callaway RM, Aschehoug E (2000) Invasive plants versus their new and old neighbors: a mechanism for exotic invasion. Science 290, 521–523. Callaway RM, Ridenour WM (2004) Novel weapons: invasive success and the evolution of increased competitive ability. Frontiers in Ecology and the Environment 2, 436–443. Callaway RM, Thelen GC, Rodriguez A, Holben WE (2004b) Soil biota and exotic plant invasion. Nature 427, 731–733. Callaway RM, Thelen GC, Barth S, Ramsey PW, Gannon JE (2004c) Soil fungi alter interactions between the invader Centaurea maculosa and North American natives. Ecology 85, 1062–1071. Cavallito CJ, Bailey JH (1949) An antibacterial principle from Centaurea maculosa. Journal of Bacteriology 57, 207–212. Chamberlain K, Guerrieri E, Pennacchio F, Pettersson J, Pickett JA, Poppy GM, Powell W, Wadhams LJ, Woodcock CM (2001) Can aphid-induced plant signals be transmitted aerially and through the rhizosphere? Biochemical Systematics and Ecology 29, 1063–1074. Darwin C (1859) On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life. John Murray, London. Elton C (1958) The Ecology of Invasions by Animals and Plants. Metheun, London. Fletcher RA, Renney AJ (1963) A growth inhibitor found in Centaurea spp. Canadian Journal of Plant Science 43, 475–481. Freville H, Colas B, Ronfort J, Riba M, Olivieri I (1998) Predicting endemism from population structure of a widespread species: case study in Centaurea maculosa Lam (Asteraceae). Conservation Biology 12, 1269–1278. Grime JP, Shacklock JML, Band SR (1985) Nuclear DNA contents, shoot phenology and species co-existence in a limestone grassland community. New Phytologist 100, 435–445. He J, Dai X, Zhao PX (2007) PLAN: A Web platform for automating high-throughput BLAST searches and for managing and mining results. BMC Bioinformatics 8, 53. Hiero JL, Callaway RM (2003) Allelopathy and exotic plant invasion. Plant and Soil 256, 29–39. Hufbauer RA, Sforza R (2008) Multiple introductions of two invasive Centaurea taxa inferred from cpDNA haplotypes. Diversity and Distributions, 14, 252–261. Inderjit, Callaway RM, Vivanco JM (2006) Can plant biochemistry contribute to understanding of invasion ecology? Trends in Plant Science 11, 574–580.

194

WEEDY AND INVASIVE PLANT GENOMICS

Ju J, Gang L, Lambert JD, Yang CS (2007) Inhibition of carcinogenesis by tea constituents. Seminars in Cancer Biology 17, 395–402. Kane NC, Rieseberg LH (2007) Selective sweeps reveal candidate genes for adaptation to drought and salt tolerance in common sunflower, Helianthus annuus. Genetics 175, 1803–1812. Kelsey RG, Locken LJ (1987) Phytotoxic properties of cnicin, a sesquiterpene lactone from Centaurea maculosa (spotted knapweed). Journal of Chemical Ecology 13, 19–33. Lacey JR, Marlow CB, Lane JR (1989) Influence of spotted knapweed (Centaurea maculosa) on surface runoff and sediment yield, Weed Technology 3, 627–631. Landau I, Müller-Schärer H, Ward PI (1994) Influence of cnicin, a sequiterpene lactone of Centaurea maculosa (Asteraceae), on specialist and generalist herbivores. Journal of Chemical Ecology 20, 929–942. Lee CE (2002) Evolutionary genetics of invasive species. Trends in Ecology and Evolution 17, 386–391. Lewi DM, Hopp HE, Esaudon AS (2006) Sunflower (Helianthus annuus L.). Methods in Molecular Biology 343, 291–297. Locken LJ, Kelsey RG (1987) Cnicin concentrations in Centaurea maculosa, spotted knapweed. Biochemical Systems Ecology 15, 313–320. Marler MJ, Zabinski CA, Callaway RM (1999) Mycorrhizae indirectly enhance competitive effects of an invasive forb on a native bunchgrass. Ecology 80, 1180–1186. Marles MAS, Ray H, Gruber MY (2003) New perspectives on proanthocyanidin biochemistry and molecular regulation. Phytochemistry 64, 367–383. Marrs RA, Sforza R, Hufbauer RA (2007) When invasion increases population genetic structure: a study with Centaurea diffusa. Biological Invasions 1387–3547. online doi: 10.1007/s10530–007–9153–6. Mauer T, Russo MJ, Evans M (2001) Element Stewardship Abstract for Centaurea maculosa. Arlington, Virginia: The Nature Conservancy. Meepagala KM, Osbrink W, Sturtz G, Lax A (2006) Plant-derived natural products exhibiting activity against Formosan subterranean termites (Coptotermes formosanus). Pest Management Science 62, 565–570. Meiman PJ, Redente EF, Paschke MW (2006) The role of the native soil community in the invasion ecology of spotted (Centaurea maculosa auct. Non Lam.) and diffuse (Centaurea diffusa Lam.) knapweed. Applied Soil Ecology 32, 77–88. Minto RE, Blacklock BJ (2008) Biosynthesis and function of polyacetylenes and allied natural products. Progress in Lipid Research 47, 233–306. Morse CC, Yevdokimov IV, DeLuca TH (2000) In situ extraction of rhizosphere organic compounds from contrasting plant communities. Communications in Soil Science and Plant Analysis 31, 725–742. Müller H (1989) Growth pattern of diploid and tetraploid spotted knapweed, Centaurea maculosa Lam. (Compositae) and effects of the root-mining moth Agapeta zoegana (L.) (Lep.: Cochylidae). Weed Research 29, 103–111. Mummey DL, Rillig MC (2006) The invasive plant species Centaurea maculosa alters arbuscular mycorrhizal fungal communities in the field. Plant and Soil 288, 81–90. Mummey DL, Rillig MC, Holben WE (2005) Neighboring plant influences on arbuscular mycorrhizal fungal community composition as assessed by T-RFLP analysis. Plant and Soil 271, 83–90. Nahrstedt A, Proksch P, Conn EE (1987) Dhurrin, (-) catechin, flavanol glycosides and flavones from Chamaebatia foliolosa. Phytochemistry 26, 1546–1547. Ochsmann, J (2001) On the taxonomy of spotted knapweed (Centaurea stoebe L.) In: Proceedings from The First International Knapweed Symposium of the Twenty First Century. Pp. 33–41. Palmer AG, Gao R, Maresh J, Erbil WK, Lynn DG (2004) Chemical biology of multi-host/pathogen interactions: Chemical perception and metabolic complementation. Annual Reviews in Phytopathology 42, 439–464. Panagouleas C, Skaltsa H, Lazari D, Skaltsounis AL, Sokovic M (2003) Antifungal activity of secondary metabolites of Centaurea raphanina ssp mixta, growing wild in Greece. Pharm Biology 41, 266–270. Perry LG, Thelen GC, Ridenour WM, Callaway RM, Paschke MW, Vivanco JM (2007) Concentrations of the allelochemical (±)-catechin IN Centaurea maculosa soils. Journal of Chemical Ecology 33, 2337–2344. Pimentel D, Lach L, Zuniga R, Morrison D (2000) Environmental and economic costs associated with non-indigenous species in the United States. BioScience 50, 53–65. Powell AM, Kyhos DW, Raven PH (1974) Chromosome numbers in Compositae. X. American Journal of Botany 61, 909–913. Quintana N, Weir TL, Du J, Broeckling CD, Rieder JP, Stermitz FR, Paschke MW, Vivanco JM (2008) Phytotoxic polyacetylenes from roots of Russian knapweed (Acroptilon repens (L) DC.). Phytochemistry 69, 2572–2578. Ridenhour WM, Callaway RM (2001) The relative importance of allelopathy in interference: The effects of an invasive weed on a native bunchgrass. Oecologia 126, 444–450. Rieseberg LH, Kim SC, Randell RA, Whitney KD, Gross BL, Lexer C, Clay K (2007) Hybridization and the colonization of novel habitats by annual sunflowers. Genetica 129, 149.

GENOMICS OF PLANT INVASION: SPOTTED KNAPWEED CASE STUDY

195

Ro DK, Paradise EM, Ouellet M, Fisher K, Newman KL, Ndungu JM, Ho KA, Eachus RA, Ham TS, Kirby J, Chang MCY, Withers ST, Shiba Y, Sarpong R, Keasling JD (2006) Production of the antimalarial drug precursor artemisinic acid in engineered yeast. Nature 440, 940–943. Routaboul JM, Kerhoas L, Debeaufon I, Pourcel L, Caboche M, Einhorn J, Lepiniec L (2006) Flavonoid diversity and biosynthesis in seed of Arabidopsis thaliana. Planta 224, 96–107. Shimamoto K, Kyozuka J (2002) Rice as a model for comparative genomics of plants. Annual Review of Plant Biology 53, 399–419. Singh HP, Batish DR, Kohli RK (1999) Autotoxicity: Concept, organisms, and ecological significance. Critical Reviews in Plant Science 18, 757–772. Singh HP, Batish DR, Kohli, RK (2003) Allelopathic interactions and allelochemicals: New possibilities for sustainable weed management. Critical Reviews in Plant Sciences 22, 239–311. Steeghs M, Bais HP, de Gouw J, Goldan P, Kuster W, Northway M, Fall R, Vivanco JM (2004) Proton-transfer-reaction mass spectrometry (PTR–MS) as a new tool for real time analysis of root-secreted volatile organic compounds (VOCs) in Arabidopsis thaliana. Plant Physiology 135, 47–58. Stevens KL (1986) Allelopathic polyacetylenes from Centaurea repens (Russian knapweed). Journal of Chemical Ecology 12, 1205–1211. Suchy M, Herout V (1962) On terpenes. CXXXVI. Identity of the bitter principle from Centaurea stoebe (L.) with cnicin. Collection of Czechoslovak Chemical Communications [0010–0765] vol:27. Tanner GJ, Francki KT, Abrahams S, Watson JM, Larkin PJ, Ashton AR (2003) Proanthocyanidin biosynthesis in plants; purification of legume leucoanthocyanidin reductase and molecular cloning of its cDNA. Journal of Biological Chemistry 278, 31647–31656. Thompson MJ (1996) Winter foraging responses of elk to spotted knapweed removal. Northwest Science 70, 10–19. Tian L, Pang Y, Dixon RA (2007) Biosynthesis and genetic engineering of proanthocyanidins and (iso)flavonoids. Phytochemistry Reviews DOI 10.1007/s11101–007–9076-y. Towie N (2006) Malaria breakthrough raises spectre of drug resistance. Nature 440, 852. Treier UA, Broennimann O, Normand S, Guisan A, Schaffner U, Steinger T, Muller-Scharer H (2009) Shift in cytotype frequency and niche space in the invasive plant Centaurea maculosa. Ecology, 90, 1366–1377. Walker TS, Bais HP, Halligen KM, Stermitz FR, Vivanco JM (2003) Metabolic profiling of non-polar compounds in root exudates of Arabidopsis thaliana in vitro; study of dynamic interface for the comprehensive characterization of rhizopsheric interactions. Journal of Agricultural and Food Chemistry 51, 2548–2554. Wardle DA, Nilsson MC, Gallet C, Zackrisson O (1998) An ecosystem-level perspective of allelopathy. Biological Reviews 73, 305–319. Weir TL, Park S, Vivanco JM (2004) Biochemical and physiological mechanisms mediated by allelochemicals. Current Opinion in Plant Biology 7, 472–479. Weston LA, Duke SO (2003) Weed and crop allelopathy. Critical Reviews in Plant Sciences 22, 367–389. Willis RJ (2004) Justus Ludewig von Uslar, and the First Book on Allelopathy. Dordrecht, The Netherlands: Springer Publications. Zhang S, Won YK, Ong CN, Shen HM (2005) Anti-cancer potential of sesqiterpene lactones: bioactivity and molecular mechanisms. Current Medicinal Chemistry, Anticancer Agents 5, 239–249.

13

Molecular Ecology Of Plant Competition Dominik D. Schmidt, Merijn R. Kant, and Ian T. Baldwin

Introduction

The number of plant species that have moved across their biogeographic barriers because of human activities has increased dramatically over the last two centuries (Vitousek et al. 1997; Mack et al. 2000). This epoch of homogenized biodiversity was baptized “the homogecene” by behavioral ecologist Gordon Orians, and Hal Mooney coined the term “the New Pangaea,” referring to the modern world as “one big man-made continent” (Dukes and Mooney 1999; Rosenzweig 2001). In the New Pangaea, some of the non-native species have become extremely abundant in their newly occupied habitats, often causing problems to native species and thereby causing major environmental and economic problems (Pimentel 2000). Although the need to prevent and control such biological invasions has led to an explosion of scientific studies over the last two decades, satisfactory explanations of why some species are invasive in novel ecosystems and others are not remain elusive. Recent advances in molecular biology, genetics, and analytical chemistry have made it feasible to ask if an invader ’s success can be related to universal traits and to study the extent to which this success is something more than coincidental. In general, successful invaders have certain distinguishing characteristics: they tend to escape local herbivores and pathogens, have efficient dispersal strategies, compete efficiently for scarce resources, and share a relatively high degree of phenotypic plasticity. Baker (1965) reasoned that a weed’s success must result from a “general purpose genotype” that allows for a wide degree of phenotypic plasticity. However, not all weeds are cosmopolitan. Therefore, Richards et al. (2006), among others, suggested three types of successful invaders: (1) “jack of all trades,” (2) “master of some,” or (3) the combined “jack-and-master” type sometimes referred to as “master of all.” A jack of all trades represents a generalist species that performs reasonably well under diverse conditions. Such a generalist will evolve into a master of some, a specialist, when confined to a particular condition for a period of time. This gives natural selection sufficient opportunity to optimize particular jack traits, turning them into master traits at the expense of other traits. So-called jack and master traits are thus the theoretical borders of the fitness landscape of a successful weed. It has been suggested that some of the most aggressive weed species have accumulated multiple master traits, which in turn give rise to the jack-and-master phenotype, a hyper-generalist. However, for identifying marker traits of successful competitors, it is irrelevant how well those traits are optimized. Therefore, we focus on traits that play a role in competition in general: the traits that enable a plant to monitor the status of its environment, respond to that status (e.g. to competition, by establishing strategies of defense, offense or tolerance), and do all these things better than others. Although considerable effort has been invested in describing the phenotypic traits required for invasiveness, little is known about their genetic basis (Weinig et al. 2007). However, as a result of the technological revolution in functional genomics, this situation is rapidly changing. Importantly, it has brought molecular genetics within the reach of ecologists, since gene197

198

WEEDY AND INVASIVE PLANT GENOMICS

expression profiling, also called transcript profiling, has become technically feasible even in less well-equipped laboratories. In addition, the frequent use of transcript profiling has lowered its cost and increased the possibility of outsourcing analyses (Kant and Baldwin 2007). Today, most laboratories are capable of identifying genes of interest and monitoring either their individual expression via quantitative PCR or RNA gel-blots or their simultaneous expression using microarrays and high-throughput quantitative transcriptome analysis. These advances have allowed the “ask-the-plant” experimental approach to be implemented on a genome-wide, and thus unbiased, scale. Moreover, our increased grip on how such candidate genes can be manipulated, via forward and reverse genetics, allowed for assessing the function of alleged plant competition genes efficiently. In this chapter we focus on the molecular mechanisms underlying plant competition, especially those identified through transcript analysis, and discuss how modern molecular tools have illuminated our understanding of the genetic basis of the dynamic and variable ecological phenomenon of invasiveness.

Competition Signals And Their Perception By Plants

The word competition implies that the object competed for is limited—that there are limited resources that constrain the fitness of individual plants and shape plant communities. Plants perceive competition via environmental cues either as a direct consequence of resource depletion or as an indirect consequence of the activity of a neighboring plant (Figure 13.1). Such resources are often related to a direct definition of space, i.e. growing space, or an indirect definition, e.g. the pool of available nutrients. One of the first consequences of nutrient depletion is that a plant starts to prioritize resource allocation more stringently. This means energy and resources are (re)allocated to those organs and processes that are needed (1) to maintain primary metabolism at a basal level or (2) to repair damage and/or restore a malfunctioning process, and in some instances (3) to address the root of the problem to prevent it from happening again. How a plant prioritizes depends strongly on its developmental stage and the local environment. In general, responses to competition are characterized by how resources are allocated to those plant parts that contribute most to the offensive or defensive responses. In many cases the result is a directional growth response enabling a plant to adjust to or outcompete a competitor. These changes in growth priorities are thought to be regulated by phytohormones (see the box entitled Phytohormones). For these growth responses to be initiated, direct and indirect signals of competition must be perceived and processed, and the responses metabolically and physiologically integrated (Figure 14.1).

Indirect Signals: Resource Availability

Light. Sunlight supplies both energy and information. The spectral composition of light is altered by the objects on which it shines or is reflected, and such alterations provide information that plants use to assess the type and structure of the surrounding vegetation. Light is composed of different wavelengths and colors and in distinct ratios. Plant chlorophylls absorb blue and red light, which leads to altered red/far-red (R/FR) ratio and less blue light irradiance under plant canopies. In response to altered red and blue light, plants increase the growth of stems and petioles, and this switch in growth priorities allows plants to overtop their neighbors.

MOLECULAR ECOLOGY OF PLANT COMPETITION

199

Figure 13.1. Plants face a multitude of exogenous signals which they receive and integrate with numerous endogenous signals (bottom panel). Competition is perceived by many direct and indirect cues (top panel; see text for detailed explanation).

This so-called shade avoidance response requires photosensory systems that enable plants to decipher the light signals from their direct environment. The R/FR wavelengths are perceived by the intracellular family of photoreversible phytochrome photoreceptors, while perception of blue and ultraviolet (UV)-A wavelengths requires cryptochromes and phototropins. Three major phytochromes exist in angiosperms: phyA, phyB, and phyC; two types

200

WEEDY AND INVASIVE PLANT GENOMICS

of cryptochromes and two phototropins have been characterized (Quail 2002; Franklin et al. 2005). Among the phytochromes, phyB is a key molecule that enables plants to respond to shading. Mutants deficient in phyB exhibit constitutive shade avoidance responses, such as elongated stems and petioles, even in the absence of neighboring plants (Ballaré 1999). Phytochromes are proteins that lie inactive in the cytosol of a plant’s cell but are activated by R/FR, after which they mediate transcriptional changes. A well-described example involves the interaction between phytochromes and the phytochrome-interacting factor PIF3. PIF3 is a transcriptional activator of genes that encode for repressor proteins of photomorphogenesis. Upon activation by red light, the phyB protein is transported from the cytosol into the nucleus, where it binds and phosphorylates PIF3. Photoconversion of phyB back to its inactive form causes dissociation from phosphorylated PIF3, after which the phytochrome may be transported back into the cytosol. Importantly, however, phosphorylated PIF3 is unstable in light and is readily degraded via ubiquitination (Quail 2002; Jiao et al. 2007), thereby releasing the repressed photomorphogenesis response. Transcriptomic studies have revealed that large changes are elicited by changes in light intensity or specific wavelength. Several common transcription factors as well as common promoter motifs in the plant DNA, to which those factors bind, have been identified in relation to early light-responsive genes (Casal and Yanovsky 2005; Jiao et al. 2007). Many of these early genes are believed to orchestrate lightassociated developmental and morphological changes such as etiolation and de-etiolation, shade avoidance, and phototropism. Water. Water is a resource that provides turgor pressure to cells and allows CO2 to be taken up through the plant’s stomata, but also provides information. In natural habitats, water availability is frequently limiting and competition for soil water can profoundly influence the composition of plant communities. Bunce et al. (1977) showed that the competitive advantage of a more drought-adapted plant species is in part related to its ability to “waste” water by maintaining a high transpiration rate that depletes the soil water for less drought-adapted competitors. Although Darwin (1880) recognized that plant roots penetrate the soil in search of water, relatively little is known about the mechanism of hydrotropic sensing. How the hydro-stimulus is perceived is still unknown, but it is clear that the root cap harbors the hydrosensors necessary to detect moisture gradients and that the resulting hydrotropic growth response depends on a complex interplay of in planta signals, namely calcium, auxin, and ABA, to mediate this process (Eapen et al. 2005). Ion fluxes (Ca2+ and H+ resulting in pH changes), as in gravitropic sensing, are among the initial responses of plants to their roots being exposed to a water gradient, and these fluxes redirect the auxin flow and polarity, redirecting growth (Box 13.1). Additionally, ABA and ethylene appear to positively regulate hydrotropic growth by influencing auxin distribution as well (Eapen et al. 2005; Esmon et al. 2005). The close coordination of hydrotropic and gravitropic responses makes them experimentally difficult to disentangle. For example, the no hydrotropic response (nhr1) mutant is also altered in its gravitropic responses and has dwarfed growth (Eapen et al. 2003). Interestingly, another hydrotropic mutant, miz1 (an acronym for mizu kussei, meaning water tropism in Japanese), grows like wild-type plants except that it does not respond to changes in water availability. The occurrence of the MIZ1 gene (whose function is so far unknown) appears to be restricted to terrestrial plants which, unlike their aquatic counterparts, frequently have to deal with limited water (Kobayashi et al. 2007).

MOLECULAR ECOLOGY OF PLANT COMPETITION

Box 13.1.

201

Phytohormones

Complex responses are often regulated by hormones; these are transported substances that mediate specific processes in ways other than directly acting on enzymatic activity, often via receptors (Chow and McCourt 2006). Therefore, downstream of hormonal action we typically find secondary signaling, i.e. via G-proteins (Jones and Assmann 2004), phospholipids (Xue et al. 2007), or Ca2+ (Hetherington and Brownlee 2004). Such signaling often acts through kinase/phosphatase cascades to (in)activate gene expression. Alternatively, hormones can directly facilitate protein degradation (Dreher and Callis 2007), i.e. that of transcriptional repressors. Here, we summarize the general functions of the well-studied phytohormones. Some of their more specific functions as these relate to weediness are addressed in the text of this chapter. Abscisic acid (ABA) functions in many plant-developmental processes, including abscission, from which its name is derived, bud dormancy, and seed maturation, as well as in plant responses to environmental stress and pathogens. In plants, indole-3-acetic acid (IAA) is the most abundant auxin. Auxins stimulate the expression of specific genes (also called ARFs) via the SCF-type ubiquitin-protein ligasemediated degradation of the Aux/IAA transcriptional repressors. Auxin enables the plant to maintain the polarity of growth, i.e. the direction of growth. Auxin distribution, for example, is essential for apical dominance: auxin produced by the growing tip diffuses downward and inhibits lateral bud growth. Auxin translocation occurs via the phloem and from cell to cell; the latter is directional and tightly regulated through auxin efflux carriers. In general, auxins regulate cell growth by affecting cell division and cellular expansion (Woodward and Bartel 2005). Brassinosteroids are a group of steroidal plant hormones, and more than seventy brassinosteroids have been isolated from plants. Brassinosteroids are synthesized from campesterol, and the biosynthetic and signal transduction genes that they elicit are expressed in a wide range of plant organs (Fujioka and Yokota 2003). Brassinosteroids operate largely but not exclusively at short distances (Symons et al. 2008) and are often associated with processes in which auxin is also involved (Hardtke et al. 2007). They promote cell expansion and cell elongation, vascular differentiation, pollen elongation and pollen tube formation, and possibly chilling and drought stress. Cytokinins (i.e. kinetin, zeatin, and 6-benzylaminopurine) are involved in promoting cell division, cell growth, and differentiation, as well as shoot and root morphogenesis, chloroplast maturation, cell enlargement, auxiliary bud release, and senescence. The ratio of auxin to cytokinin is crucial during cell division and the differentiation of plant tissues and auxin is known to regulate the biosynthesis of cytokinin (Nordstrom et al. 2004). Adenine-type cytokinins occur throughout the plant but are synthesized predominantly in the roots, i.e. in the actively dividing cells of the cambium. Cytokinin is involved in both local and longdistance signaling and shares the same transport systems used by the plant for moving purines and nucleosides (Choi and Hwang 2007). Ethylene (ET) is a volatile plant hormone; its genes were discovered via alterations in the response of dark-grown mutant seedlings to ethylene (the “triple response”) (Guzman and Ecker 1990). Active in minute quantities, ethylene is essential for fruit ripening and senescence. The ethylene receptor is located in the endoplasmic reticulum (ER), and ethylene biosynthesis as well as signal processing is relatively well documented (Hall et al. 2007). It has been shown that Skp1, Cullin, F-box (SCF) -dependent ubiquitination of

202

WEEDY AND INVASIVE PLANT GENOMICS

ethylene insensitive 3 (EIN3) is critical for proper ethylene signaling and plant growth (Gagne et al. 2004). Gibberellins, first discovered as a metabolite of the plant pathogen Gibberella fujikuroi, regulate growth and influence various developmental processes, including stem elongation, germination, dormancy, flowering, sex expression, enzyme induction, and leaf and fruit senescence (Ueguchi-Tanaka et al. 2007; Weiss and Ori 2007). All known gibberellins are diterpenoid acids (van Schie et al. 2007) that are synthesized in the plastids and then modified in the ER and cytosol until they attain their biologically active form. All gibberellins are derived from the ent-gibberellane skeleton but are synthesized via ent-kaurene (Otsuka et al. 2004). Jasmonic acid (JA), or jasmonate, was originally thought to be a hormone essential only for plant defense, but it plays a role in primary metabolic processes, too, even though these functions have been studied in far less detail. Because JA is essential for the production of proteinase inhibitors, some of which interfere with gut digestive proteases, it is considered the regulator of “anti herbivore defense; however, involvement in pathogenesis (i.e. necrotrophic pathogens) has been widely acknowledged as well. Its biosynthesis via the octadecanoid pathway is well documented (Wasternack 2007), and although some of its conjugates are inactive—hydroxides, sulfonates, and glucosides—others have bioactivity similar to that of JA. JA likely activates downstream gene expression via its isoleucine (Ile) conjugate; this conjugate binds to the SFC-coronatine insensitive (COI) complex to initiate the degradation of the so-called jasmonate ZIM-domain proteins (JAZ) transcriptional repressors. Whether the activity of other conjugates depends on conversion into JA-Ile or which form or forms of JA are transported is still unknown (summarized by Farmer 2007). Salicylic acid, or salicylate, is the classical phytohormone essential for defense against pathogens, although it has a role in primary processes as well (Loake and Grant 2007). The bulk of salicylic acid (SA) is synthesized through the shikimate pathway, but the phenylpropanoid pathway downstream of the enzyme phenylalanine ammonia lyase (PAL) may also be used. Salicylate is inactivated, for example, through glucosylation and transported in phloem in the form of methyl salicylate (MeSA) (Park et al. 2007). The mechanism by which SA accumulation leads to gene expression is unknown. SA and JA are known to antagonize one another, likely via a protein called NPR1, and the switch from a JA response to an SA response is often mediated by ET (von Dahl and Baldwin 2007). However, the sequential interdependence and the kinetics of this ménage à trois differ markedly among plant species and are also likely affected by ABA acting on JA biosynthesis (Adie et al. 2007). While ET is often regarded as the only volatile phytohormone, it needs to be mentioned that there are volatile hormone derivatives or precursors produced by plants: methyl jasmonate (MeJA), MeSA, and the gibberellin-precursor, ent-kaurene. MeJA, although a common plant compound, is not commonly released by plants as a volatile into the air (e.g. Karban 2007). MeSA, although commonly released, has hardly been explored as a volatile hormone (Shulaev et al. 1997); however, its role in attracting insects (de Boer and Dicke 2004; Ament et al. 2004) and its endogenous role as a signal molecule have been described in detail (Park et al. 2007). Finally the gibberellin-precursor ent-kaurene—although also not a common plant volatile—emitted by transgenic plants over-expressing an ent-kaurene synthase gene was shown to rescue mutants deficient in gibberellin-biosynthesis upstream of ent-kaurene (Otsuka et al. 2004); and thus, has the potential to act like a volatile hormone.

MOLECULAR ECOLOGY OF PLANT COMPETITION

203

The phytohormone ABA is well known for its role in drought stress and its impact on shoot growth (Sauter et al. 2001). Three ABA receptors have been described; all mediate diverse processes (Hirayama and Shinozaki 2007). Stomatal closure during water stress is the best documented. When roots experience water stress, they signal leaves to produce ABA precursors. These are translocated to the roots, where ABA is formed, and translocated to leaves via the vascular system. Back in the leaf, ABA increases Ca2+ flux into the cytosol of the stomatal guard cells, which is, in turn, mediated by the Ca2+-dependent phosphatase ABI1 (ABAinsensitive 1). The activity of this phosphatase regulates potassium or sodium uptake into guard cells and causes them to lose turgidity, closing the stomata (Schroeder et al. 2001). Thus, ABA signaling is a critical player in determining the water-use efficiency of plants in water-depleted environments. Nutrients. Nutrient availability is highly variable. It fluctuates through time, and nutrient distribution is often very heterogeneous in natural soils. How do plants find nutrient patches? The ability of plants to efficiently assess where nutrients are is crucial for their competitive ability (Hodge 2004). When nutrients become limited, growth is reduced and followed by alterations in physiology and morphology that increase whole-plant uptake rates. Different species respond very differently to nutrient limitations, often in ways that are associated with the particular environments in which they evolved (Grime 2001). While short-term responses appear to be general, long-term responses tend to be specific for the limited nutrient. The tip of the root is where nutrient limitations are primarily perceived. Changes in nutrient composition lead to immediate changes in the membrane potential of root cells, which, in turn, are associated with rapid changes in gene expression (Schachtman and Shin 2007). Nutrients such as nitrate, sulfate, phosphate, or iron act as signals, triggering molecular responses that change roots’ architecture through cell division and differentiation. Important developmental processes, such as root-hair formation, primary root growth, and lateral root formation, are particularly sensitive to changes in the internal and external concentration of nutrients (Schachtman and Shin 2007; Svistoonoff et al. 2007). Finally, because root architecture can be modified by plant growth regulators, such as auxins, cytokinins, and ethylene, it is safe to assume that nutritional control of root development is also mediated via phytohormones to a large extent (López-Bucio et al. 2003). Nitrate signaling exemplifies the complexity of nutrient-induced molecular changes. The MADS box gene, ANR1 (Arabidopsis Nitrate Inducible 1), functions as a key transcriptional regulator of root responses during nitrate starvation. Plants with decreased ANR1 transcript levels are unable to forage effectively for patchily distributed nitrate by means of lateral root proliferation (Zhang and Forde 1998; Forde 2002). In normal plants, the protein ANR1 is controlled by upstream nitrate-sensing machinery that requires the acquisition of nitrate by a nitrate transporter (NRT1.1; Remans et al. 2006). The responses downstream of ANR1, mediated by auxin, translate the availability of nitrates into lateral root growth (Forde 2002). This relatively simple signaling cascade is, however, highly entangled with other parts of plant metabolism, as shown by experiments in which nitrate was added to nitrate-deprived Arabidopsis roots, resulting in the differential expression of nearly 1,200 genes (Wang et al. 2003). Moreover, the acquisition of different nutrients does not happen independently. The interactions between the processes responsible for the uptake of different nutrients and between the different metabolic pathways that process different nutrients are well known and visible on the molecular level. For example, transcript levels of nitrate transporters are down regulated in potassium-deprived plants. Moreover, potassium and phosphorus deprivation elicit the same regulatory genes, such as specific MAP kinases and transcription factors (Schachtman and Shin 2007).

204

WEEDY AND INVASIVE PLANT GENOMICS

Direct Signals

Volatile Compounds. Plants emit large amounts of volatile organic compounds (VOCs) from various tissues, including flowers and leaves. It has been estimated that up to 36% of the assimilated carbon is released back to the atmosphere (Kesselmeier et al. 2002). Why do plants apparently dispose of so much of their fixed carbon aerially? Some VOCs are essential for heat regulation or to repel herbivores but others may be mere waste. VOCs can be small and highly volatile molecules (e.g. ethylene, methanol, isoprene) or heavier compounds, such as mono- and sesquiterpenes, aromatics such as methyl salicylate (MeSA), and lipoxygenasederived C6-volatiles such as hexenal, often collectively referred to as green-leaf volatiles (GLVs). The emission of VOCs is influenced mostly by abiotic factors, such as nutrient availability, temperature, or the spectral composition of light, i.e., especially photosynthetically active radiation. Although most plants release VOCs constitutively, the composition of the VOC blend often changes markedly in response to different biotic stresses, such as herbivore and pathogen attack. These induced changes in the VOC bouquet play important roles in establishing mutualistic interactions between plants and animals (Dicke et al. 2003; Baldwin et al. 2006). They provide information to organisms that interact with plants, and neighboring plants may use this information to adjust their growth to increase their competitive capacity. The observation that a plant’s metabolic status seems encrypted in the profile of VOCs it releases led, more than two decades ago, to the hypothesis that those volatiles might facilitate plant-plant information exchange, i.e. plants eavesdropping on one another. Baldwin and Schultz (1983) discovered that exposing undamaged, individually potted sugar maples and poplars to the VOCs of mechanically damaged conspecifics increased the trees’ levels of tannins and phenolics. Similar results were reported by Rhoades (1983), who found that fieldgrown willows growing next to herbivore-attacked willows were less palatable to larvae than were unattacked willows growing next to unattacked willows. While in the second study, the exchange of below-ground signals could not be excluded, in the first study, the observed effects could only have been mediated by above-ground airborne VOCs. A number of researchers have tried to discern which VOCs elicited such effects and many laboratory and field studies have found additional evidence for VOC-mediated defense activation among plants; however, most studies use synthetic VOCs and fumigate plants with unnaturally high quantities (Dicke et al. 2003; Baldwin et al. 2006). The phenomenon of VOC-mediated defense activation among plants raises intriguing questions about how plants discriminate between signals and noise. The activation of defense responses requires energy, which, by definition, is undesired in the absence of herbivory. Solutions to this problem have been proposed. For example, it would make sense if plants did not activate their complete defense arsenal after the first signs of trouble but were to “prime” their defense metabolisms so as to be in a temporal state of so-called enhanced alertness (Conrath et al. 2006). Priming is thought to entail low-cost metabolic changes. These are difficult to detect experimentally but they enable a plant to launch defense responses more rapidly and more strongly when attacked. By allowing themselves to be primed instead of induced through eavesdropping on a neighbor, plants might very well prevent resources from being wasted on unnecessary defenses and thereby realize an overall fitness benefit (Kessler et al. 2006; Ton et al. 2007). This leaves open the question of why plants are apparently not under strong selection to remain silent. Clearly the advantages of releasing volatiles are larger than the disadvantage of alerting one’s neighbors. Some GLVs were shown to elicit root elongation, depending on GABA signaling in Arabidopsis (Mirabella et al. 2008), and evidence that GLVs mediate plant-plant interactions

MOLECULAR ECOLOGY OF PLANT COMPETITION

205

was found in barley experiments. Unidentified VOCs released by emitter barley plants altered biomass allocation between roots and shoots without influencing total biomass production of receiver barley plants (Ninkovic 2003). In response to the GLVs released from their neighbors, plants seem to redistribute biomass among different organs for optimal growth (Grime 2001). A recent study using native tobacco plants genetically engineered to be “mute” in different aspects of their volatile vocabulary revealed that the transcriptional responses of neighboring eavesdropping plants was not elicited by the presence of GLVs in the volatile bouquet but, rather, by their absence (Paschold et al. 2006). This study underscores the need to keep an open mind about the nature of the information encrypted in the VOC bouquet. Although the idea that VOCs can be used in different forms of plant-plant signaling is now generally accepted, it remains unclear how VOCs are perceived by plants. The volatiles, ethylene, MeSA and MeJA, may simply adhere to the leaf and diffuse from the surface into the epidermal cells where, with or without additional modification, they may play direct signaling roles. However, most plant volatiles seem unlikely to be easily converted into phytoactive substances. Moreover, VOC receptors have not yet been identified, with the notable exception of the ethylene (ET) receptors. Ambient ET concentrations can be elevated in dense stands in which individual plants exhibit a neighbor-induced shade avoidance response that is characterized by stem elongation and leaf hyponasty and mediated by the light signals described earlier. Pierik et al. (2003) demonstrated that ET perception plays a central role in this competitionmediated shade avoidance response. ET-insensitive cultivated tobacco plants carrying a mutated ET receptor (etr1–1) exhibited decreased shade avoidance and were overgrown by wild-type plants with intact ET sensing. Ethylene likely influences blue light perception; such perception is important for increasing vertical growth and phototropism (Briggs and Christie 2002; Binder 2007). Rhizosphere Compounds: Exudates and Quorum Signals. Plant roots exude multiple compounds into the surrounding soil (i.e. the rhizosphere, an arbitrarily defined soil-air-water zone in which interactions take place). Such exudates are composed of low-molecular-weight molecules (e.g. amino acids, organic acids, monosaccharides, and diverse plant secondary metabolites such as lignin precursors and derivatives) to large molecules such as polysaccharides or proteins. Importantly, compounds exuded into the rhizosphere accumulate more readily near the plant surface compared to volatile compounds released from the cuticula or stomata, before finally diffusing into air. As with volatiles, exuded non-volatile metabolites can function as signals that mediate interactions among plants, microbes, and herbivores (Bais et al. 2006). (The well-studied negative plant-plant interactions referred to as allelopathy are covered by Broz and Vivanco in Chapter 12 of this book.) In contrast, little is known about the chemical signals that mediate root-root interactions that are beneficial for receiver plants. Root-exuded secondary metabolites of herbivore-infested plants have been shown to decrease the attractiveness of neighboring plants to above-ground herbivores, such as aphids, while increasing plants’ attractiveness to the herbivores’ natural enemies (Chamberlain et al. 2001; Dicke and Dijkman 2001). Moreover, it was suggested that root exudates might help plants to discriminate kin from non-kin. Dudley and File (2007) found increased allocation of biomass to the roots of Great Lakes sea rocket (Cakile edentula) when it was grown together with non-kin conspecifics, as opposed to when it was grown with its siblings. Similarly, strawberry (Fragaria vesca) root growth was stimulated by contact with roots of Glechoma hederacea, suggesting the existence of mechanisms allowing for self–non-self recognition (Semchenko et al. 2007). In both cases however, the nature of the signals responsible for these responses are unknown. In contrast, the chemical communication between soil microorganisms and plants is better understood.

206

WEEDY AND INVASIVE PLANT GENOMICS

Plants respond to products associated with pathogenic microorganisms; pathogen-associated molecular patterns (PAMPs) can be proteins from bacterial flagella or effector molecules that pathogens inject into plant cells to facilitate infection (Jones and Dangl 2006). The responses elicited in plants after the perception of PAMPs are typically defensive but may also be offensive, aimed at disturbing bacterial group processes. Many bacteria coordinate activities such as conjugation, biofilm formation, and infection of host tissues through specific substances, mostly N-acyl homoserine lactones (AHLs). AHLs are autoinducers. Molecules that are constitutively produced, released, and taken up again by bacteria and the promoter of both the AHL synthase and receptor gene are responsive to the compound itself. Hence, the production and accumulation of AHLs increase when bacterial densities increase. In a bacterial population in soil that has accumulated an AHL concentration above a threshold level, individual gene expression changes dramatically and initiates coordinated processes, as happens, for example, in host colonization. This communication process, called quorum sensing, depends on population density—responses to be coordinated and activated only when sufficient bacteria are present for the mechanism to be effective (Bauer and Mathesius 2004). Although colonizing bacteria can be symbiotic, many are pathogenic and some plants may have evolved mechanisms to monitor and manipulate bacterial communication to anticipate bacterial infection. Applying synthetic AHLs to tomato or Medicago truncatula, for example, has been found to result in the regulation of many genes and proteins associated with the plant stress metabolism (Bauer and Mathesius 2004). In addition, some plants appeared to secrete unidentified AHL mimics that disturb bacterial quorum sensing that might benefit the plant (Gao et al. 2003). It is known, for example, that plant-secreted g-aminobutyric acid disrupts Agrobacterium’s quorum-sensing ability and decreases its infection success (Chevrot et al. 2006; Shelp et al. 2006). Although many questions remain unanswered, the bacterial quorumsensing system is clearly prone to manipulation. Plants have evolved multiple ways of interacting with microbes and it is likely that the more a plant is able to interact with its microbial community, the better able it may be to compete with other plants that have remained deaf to the local microbial dialect.

Integration Of Multiple Signals: Information Management Within Plants

Plants perceive external and internal stimuli with different tissues or organs. Communication and behavioral integration of interconnected plant parts are essential for an efficient overall response for several reasons (de Kroon et al. 2005). For example, plants need to integrate signals from above-ground sources (such as reflected light or VOCs) and below-ground sources (i.e. exudates, nutrients, water, root contact) that come in via their different parts (i.e. shoots and roots; Figure 13.1) to successfully respond to competitors. Inter-organ communication is largely achieved through the vascular system, especially through the phloem, which transports not only photosynthates (sugars) but also a wide variety of other small or large signal molecules, such as RNAs and proteins (Lough and Lucas 2006). The vasculature transmits signals that orchestrate differentially timed responses within a plant’s primary and secondary metabolism, such as nutrient acquisition (daily), flowering (at distinct developmental stages), or coping with drought stress (at more or less random times). For example, the induction of flowering is a systemic process which requires one or more signals from leaves to be translocated to the shoot apex. Since the 1930s, it has been known that flowering can be elicited under non-inductive conditions via a graft-transmissible signal.

MOLECULAR ECOLOGY OF PLANT COMPETITION

207

Chailakhyan (1937) proposed a universal floral stimulus called florigen, but such a compound has yet to be isolated despite several claims. The alternative multifactorial control hypothesis suggested that known hormones, such as gibberellic acid, and other compounds interact to induce flowering after photoperiodic changes (Corbesier and Coupland 2006). In 2007, genetic studies using Arabidopsis and rice mutants identified the Flowering Locus T (FT) protein as being responsible for activating the signal that moves from leaves via the phloem to the shoot apex where it, in turn, activates floral meristem identity genes (Corbesier et al. 2007; Tamaki et al. 2007). Nutrient acquisition in competitive environments depends on information about need versus availability and how this information is integrated in the various parts of the plant. Shoots exercise regulatory control over the nitrogen acquisition activity of roots by a feedback mechanism involving root-shoot (xylem-phloem) communication, which allows plants to respond to heterogeneity in nitrogen availability by increasing the levels of nitrate uptake and lateral root proliferation when roots approach richer soil patches. Split-root experiments demonstrate that the regulation of this process involves both local and systemic signaling from distal plant parts. In these studies, different parts of the root system were exposed to either high or low nitrate levels. Consistent with the overall plant response, nitrate starvation in one part of the root system was compensated for by increased uptake from the other root zone (Forde 2002). Apparently, inorganic signals related to nitrate and the plant’s N-status are integrated with organic signals such as the phytohormones cytokinin and auxin (Box 13.1), coordinating the optimization of root developmental and growth responses for N acquisition (Forde 2002; Sakakibara et al. 2006). The stress hormone ABA plays a central role in coordinating these systemic allocation responses; interestingly, it seems to have distinct roles at different stages in lateral root development and thus might allow for plasticity in the face of changing conditions not only above ground but also in the soil (De Smet et al. 2006). The hormone, which is primarily synthesized in the roots but also in the leaves, is also present in the soil as a result of root exudation and the activity of ABA-producing soil organisms (Jiang and Hartung 2007). In planta, ABA can move rapidly through both the xylem and the phloem and be distributed largely as a function of pH (Jiang and Hartung 2007). A crucial step in controlling ABA levels in Arabidopsis is the hydrolysis of the inactive glucose ester of ABA (ABA-GE). The hydrolysis is catalyzed by the b-glucosidase, AtBG1, and mutants devoid of AtBG1 activity have lower free ABA levels, defective stomatal movements, and stress-sensitive phenotypes (Lee et al. 2006). Additionally, ABA-GE is ideally suited for its role as a long-distance xylem signal: although ABA-GE does not permeate membranes, ABA is freely diffusible, which contributes to the stability of ABA-GE xylem concentrations (Jiang and Hartung 2007).

Molecular Basis Of Competitively Important Traits

The ability to compete can be enhanced by different means. Some of these means are offensive, such as vigorous growth and allelopathy, some are evasive or defensive, such as a high seed output with early mass emergence, and some result in stress tolerance. Although most of these traits are clearly polygenic, there is evidence that some genes, often those encoding regulatory proteins, act as master switches in particular processes. For example, kinase signaling pathways (i.e. cascades of proteins activating other proteins through phosphorylation) are key regulators in numerous developmental and stress responses. The Arabidopsis kinases KIN10 and KIN11 play a central role in the transcriptional reprogram-

208

WEEDY AND INVASIVE PLANT GENOMICS

ming that occurs in response to stimuli, such as darkness, carbohydrate deprivation, and other stresses (Baena-González et al. 2007). Mitogen-activated kinases (MAPKs) are evolutionarily conserved and regulate many signaling processes in growth and stress responses (Jonak et al. 2002; Hamel et al. 2006), often by activating transcriptional regulators. In addition to central regulators, other common features give structure to signaling networks, such as proteasomedependent degradation of repressors (Huq 2006), protein inactivation through dephosphorylation, dimerization, and other types of modifications that are not necessarily transcriptionally regulated. Such mechanisms are ubiquitous in the regulation of phytohormone- and lightsignaling during plant competition. The identification of the hubs in the networks (likely genes or proteins) will be important because those either fundamentally influence the identity of a trait or mediate interactions among discrete traits. Plasticity in setting priorities in growth and development, efficient use of resources, and high stress tolerance are all traits shared by aggressive competitors such as weeds. So can we identify the traits that give rise to so-called jacks and masters? To address these questions, several well-characterized but non-invasive model plants can provide a template against which invasive weed models can be compared.

Regulation Of Above-Ground Architecture

Plants occupy space that contains nutrients as well as competitors, and architectural flexibility allows plants to deal with both of these. Rapid growth allows a plant to monopolize space and the resources in that space. Preemptively occupying space with lateral spreading shoots likely gives absolute and relative fitness benefits; such fitness benefits are typically accrued by weeds because the occupied space cannot be used by others (Grime 2001). Shoot architecture is determined by the apical meristem, which exerts apical dominance over lateral shoots, and thereby branching. Apical dominance is tightly controlled by the interplay between auxin and cytokinin. Auxin is produced in the shoot apical meristem and transported down to the shoot, where it inhibits bud growth by down regulating cytokinin levels. Cytokinins, in turn, are known to counteract the growth inhibition in dormant auxiliary buds. The inhibitory effect of auxin on cytokinin signaling depends on at least three genes, MAX, RMS and DAD, named after the phenotype of plants deficient in their functioning (i.e. MORE AXILLARY BRANCHING, RAMOSUS—meaning “branched”—and DECREASED APICAL DOMINANCE). In addition to auxin and cytokinin, hormonal control of shoot branching involves a third carotenoid-derived hormone, which is transported acropetally in the plant (Ongaro and Leyser 2008). Shoot-branching mutants, such as the Arabidopsis max mutants, show decreased apical dominance and increased shoot branching. The mechanism appears to rely on increased auxin transport capacity in the stems, through the up regulation of auxin transporters. The auxin transport machinery is not saturated by apex-produced auxin, but lateral shoots are also able to export auxin and hence outgrow the suppressing effect of the apical meristem (Ongaro and Leyser 2008). Hormonal control of shoot branching is conserved among monocots and dicots (Doust 2007), and although experimental evidence is lacking, it is likely that MAX/ DAD/RMS genes directly affect competitive interactions. Other phytohormones, such as jasmonates, are known to influence the rate of branch growth (Zavala and Baldwin 2006) and are likely to influence the impact of the auxin-cytokinin duet either by directly influencing signaling or by redeploying the resources used in jasmonate-mediated resistance traits for shoot growth.

MOLECULAR ECOLOGY OF PLANT COMPETITION

209

Rooting Success

Below-ground space is occupied by roots. Plant root systems are highly plastic and variable, but the basic root system morphology and its plasticity are controlled by inherent genetic factors mainly related to hormone signaling. Root architecture is largely determined by the degree of branching. We have mentioned how resources can direct root growth, and clearly root development is regulated by the complex interplay of different signaling pathways (Osmont et al. 2007). The root cap is instrumental in enabling a root to “plow” through the soil, a process which can impose severe physical stress on the growing root. Exudation from the root cap and increased sloughing of border cells decrease friction (Bengough et al. 2006). Border cells originate from root cap meristems and are continuously separated from the root tip by pectolyases. Different plant species produce different amounts of border cells, ranging from none to a few hundred in the Brassicaceae and Solanaceae to thousands in the Pinaceae and Malvaceae (Hawes et al. 2003). Regulation of border cell production is influenced by auxin and ethylene, and a suite of cell-wall-degrading enzymes, such as pectin methylesterases, are involved in the separation process (Driouich et al. 2007). These not only facilitate growth, but together with exudation are also a means of modifying the rhizosphere and are implicated in plant-microbe interactions. Border-cell-produced chemicals can dramatically alter a plant’s interactions with microorganisms, for example, by repelling pathogenic bacteria or controlling symbionts (Hawes et al. 2000). The genetic dissection of root growth and morphology will yield important information about why some plants are more successful than others. In the wild tobacco Nicotiana attenuata, a gene that contributes to the plant’s ability to manipulate the rhizosphere has been characterized in detail. Rapid alkalinization factor (RALF) is a polypeptide that is highly expressed in roots. Plants silenced in the expression of RALF have wild-type above-ground growth, but their roots grow longer and develop abnormal root hairs. As expected for plants without functional root hairs, RALF-silenced plants are outcompeted by wild-type plants in the basic soils of the plant’s native habitat. Intriguingly, root hair development and root growth in general are partially restored in acidic soils. Wu et al. (2007) demonstrate that RALF is required to regulate the extracellular pH important for root hair development.

Determinants Of Photosynthetic Efficiency

Some weeds are superior competitors as a result of their efficient nutrient assimilation or photosynthesis (Black et al. 1969). Photosynthesis depends on the key enzyme of the CalvinBenson Cycle RuBPCase (ribulose-1,5-bisphosphate carboxylase/oxygenase), which fixes CO2 through the carboxylation of ribulose 1,5-bisphosphate (C5), as a result of which two glycerate3-phosphate (C3) molecules are formed. This type of fixation is referred to as C3 carbon fixation in contrast to C4 photosynthesis. During the latter process phosphoenolpyruvate is carboxylated, for which CO2 is used, to produce malate (C4) during the day in the chloroplasts of the mesophyll. This product is broken down to pyruvate and CO2, after which the latter enters the Calvin-Benson Cycle in the chloroplasts of the bundle sheath; C4 photosynthesis is commonly found in plants growing under high temperatures. This enzyme’s oxygenase activity under high O2 and low CO2 partial pressures and ribulose1,5-bisphosphate is erroneously oxidized to phosphoglycolate in a process known as photorespiration. The C4 pathway has likely evolved to solve the principal liability of RuBPCase: its oxygenase activity under high O2 and low CO2 partial pressures, whereby ribulose-1,5-

210

WEEDY AND INVASIVE PLANT GENOMICS

bisphosphate is erroneously oxidized to phosphoglycolate in a process known as photorespiration. RuBPCase’s oxygenase activity increases under high O2 concentrations in leaf cellular spaces when stomata are closed, as commonly occurs when plants are heat stressed. C4 plants are more efficient than C3 plants under heat stress conditions because they are able to sequester RuBPCase from atmospheric oxygen in the bundle sheath cells and deliver CO2 to these cells via a special transport system, which significantly reduces photorespiration (Sage 2004). Many weeds are either C4 plants, e.g. from the families Poaceae or Amaranthaceae, or C3–C4 intermediate plants, such as the (sub)tropical weed Mollugo verticillata (Baker 1974; Kennedy and Laetsch 1974), probably because C4 plants have a clear competitive advantage over C3 plants when facing conditions of drought, high temperatures, and nitrogen or carbon dioxide limitation. Moreover, the discovery that some typical C3 plants—for example, tobacco—share some of the anatomical and biochemical characteristics of C4 plants, such as photosynthetic cells around the vasculature (Hibberd and Quick 2002), and the discovery that all C3 plants possess the necessary enzymes for C4 photosynthesis (Sage 2004), suggest that facultative C4 photosynthesis could be one of the mechanisms that allows such plants to increase their photosynthetic efficiency under stress conditions. The fact that the C4 photosynthetic advantage occurs during periods of heat stress suggests that weeds might be competitively superior, especially during periods of environmental stress, and thus require such stress to be successful. RuBPCase activase (RCA), which controls the activity of RuBPCase by functioning as a catalytic chaperone, may help to optimize photosynthetic processes under normal and stress conditions, such as elevated temperature (Portis 2003). The RuBPCase function in C3 and C4 plants with decreased or absent RCA expression is severely compromised and plants perform poorly under ambient CO2 concentrations or under moderate high temperature stresses (Somerville et al. 1982; He et al. 1997; von Caemmerer et al. 2005). Clearly, plants with the physiological capability to cope with particular environmental stresses will realize a fitness benefit when faced with the stress conditions because their competitors suffer more, and provided their stress-resistant physiology does not also impose equal fitness costs under other conditions.

Stress Tolerance

The success of weeds is often associated with a relative high tolerance for abiotic and biotic stresses. For example, the common dandelion (Taraxacum officinale) has impressive regenerative abilities. After mechanical or herbivore-inflicted defoliation, it is able to regrow from any part of the rootstock (Baker 1974). The methods that plants use to tolerate stress, such as storing resources in below-ground tissues, shade avoidance responses, nutrient starvation responses, and photosynthetic adjustments, are well described at the physiological level but remain largely uncharacterized at the molecular and genetic levels. The traits that enable plants to tolerate stress often result from complex genetic interactions, which is confirmed by the fact that hundreds of genes constitute the loci that correlate with the quantitative traits (Tiffin 2000). However, certain tolerance responses have recently been shown to depend strongly on single genes, such as the SNF1 (sucrose non-fermenting-1)-related kinase, SnRK1, of N. attenuata that mediates sugar allocation during herbivory stress. Collectively, SnRK1s are kinases that function as cellular fuel gauges and play central roles in energy metabolism, development, and stress responses, e.g. during carbohydrate starvation (Lu et al. 2007; Polge and Thomas 2007).

MOLECULAR ECOLOGY OF PLANT COMPETITION

211

When the larvae of the lepidopteran herbivore Manduca sexta attack leaves of N. attenuata plants, the gene GAL83, which encodes the β-subunit of the SnRK1 protein, is strongly down regulated. Moreover, leaf herbivory is accompanied by increased carbon allocation to roots, which diverts resources to a less vulnerable location within the plant (Hermsmeier et al. 2001; Schwachtje et al. 2006). Accordingly, transgenic N. attenuata plants in which GAL83 was silenced mimicked the phenotype of herbivore-attacked plants in that they allocated more resources to the root in the absence of leaf herbivory. Hence, Schwachtje et al. (2006) could demonstrate the existence of a complex stress-tolerance trait by manipulating a single gene. SnRK1-mediated regulation of carbon and amino acid metabolism is conserved among fungi, plants, and animals (Halford et al. 2004), and it may be that SnRK1 has been under selection to regulate the many physiological responses that allow plants to become weeds. Other genes are also associated with competitive ability by increasing stress tolerance. The prosystemin gene encodes systemins, which are polypeptide signal molecules found in solanaceous plants, such as nightshade, and have been well described as peptide hormone mediators of defense gene expression in tomato (Solanum lycopersicum; McGurl et al. 1992; Ryan et al. 2002). However, in black nightshade (S. nigrum), the systemin ortholog is not involved in regulating defense gene expression but, rather, mediates a plastic growth response (reminiscent of SnRK1 in N. attenuata). Immediately after herbivory or mechanical wounding, black nightshade’s prosystemin transcripts are down regulated and greater root mass results. Applying synthetic black nightshade systemin to wounded plants suffices in experiments to compensate for the protein’s natural down regulation and inhibits the increase in root mass. This result is consistent with results from wild tomato (S. pimpinellifolium), in which the growth of lateral roots and root hairs was shown to be inhibited in response to systemin treatment (Holton et al. 2007). Moreover, the reproductive performance of transgenic S. nigrum plants with constitutively low systemin levels was better than that of wild-type plants during competition (Schmidt and Baldwin 2008). Notably, the systemin-mediated root-growth response was found to depend on ethylene and brassinosteroid signaling (Box 13.1); the interface between systemin and BR is provided by the receptor BRI1/SR160 (BRASSINOSTEROID INSENSITIVE1/systemin receptor 160), which is a dual ligand receptor for both systemin and the BR brassinolid (Montoya et al. 2002; Scheer and Ryan 2002). It is intriguing to wonder if regulators such as hormones, signaling molecules, or their receptors, as the hubs in the signaling network, facilitate the rapid evolution of new stress-responsive pathways, since changing single hubs might fully reprogram the upstream regulation of the large metabolic and physiological processes involved in stress tolerance.

Transcriptomic Insights Into Competitive Interactions Of Weedy Plants

Transcriptional profiling can be an extremely useful tool for mapping complex and highly dynamic physiological responses to environmental stresses such as competition. However, interpreting transcriptome data requires skill and a certain amount of luck. In order to reduce complexity of the data, many studies have examined specific parts of processes, comparing wild-type plants to identical plants harboring a specific single mutation. For example, lab studies on phytochrome signaling, a central component of the shade avoidance response, compared the transcriptional profiles of R/FR-exposed phytochrome mutants and wild-type plants using microarrays; complex transcriptional patterns were found (Franklin and Whitelam 2005).

212

WEEDY AND INVASIVE PLANT GENOMICS

Although experimentally manipulating R/FR ratios mimics one aspect of a nearby competitor, whether lab studies can ever sufficiently mimic the complexity of a plant’s natural environment is questionable. Moreover, reconstructing complex processes by summing the results obtained from a range of plants with single mutations in different parts of those processes prompts the question: How realistic is it to assume that the sum adds up to complete process? We do not have sufficient comparative data to answer this question since only a few studies have looked at transcriptional responses of plants to competition and rarely in field settings. Results from Solanum nigrum (see Box 13.2) field and glasshouse experiments led us to conclude that twenty-eight of 568 examined genes were differentially expressed between

Box 13.2.

Solanum nigrum: A Model Ecological Expression System

Black nightshade (Figure 13.3) has all the characteristics of an ideal experimental weed system. Black nightshade (Solanum nigrum s. l.) occurs throughout the world (Edmonds and Chweya 1997). Taxonomically it belongs to the section Solanum of the Morelloid clade that consists of about seventy-five species worldwide (Bohs 2005). S. nigrum s. str. mainly occurs in Europe but was introduced into the U.S., Canada, and Australia in the nineteenth century (Edmonds and Chweya 1997; Defelice 2003). Black nightshade is mostly known as a noxious weed, although in particular varieties the leaves or fruits can be used as vegetable (Edmonds 1997). Darlington (1859) described S. nigrum as “a homely, worthless, and even deleterious weed, which ought to be carefully expelled from the vicinity of all dwellings.” Black nightshade naturally grows on open, disturbed, and nutrient-rich soils, such as riverbanks, and has entered many rural habitats, such as fields, gardens, and wasteland. In Canada and the U.S., many studies have focused on two particular black nightshades, the introduced S. nigrum and the native S. ptycanthum (eastern black nightshade). Both species occur largely in rural places and irrigated agricultural fields. S. ptycanthum has spread dramatically in the U.S. and Canada due to its ability to rapidly evolve resistance to selective herbicides (acetolactate inhibitors) (Ogg and Rogers 1989; Holm et al. 1991; Volenberg et al. 2000; Milliman et al. 2003). Black nightshade is also able to colonize new habitats, suggesting it has a general purpose genotype (Hermanutz and Weaver 1996). For example, the germination behavior of black nightshade is very plastic and differs among populations. In rural populations it emerges earlier and at a more limited temperature range than in agrestal populations (Hermanutz and Weaver 1991). Black nightshades are aggressive competitors of tomato and broccoli. They diminish water soil content drastically and reduce crop yields by 35% to 80% (Ogg and Rogers 1989; McGiffen et al. 1992a; McGiffen et al. 1992b). Additionally, black nightshade is a known reservoir of viral diseases and a host for many pathogens and herbivores which may negatively affect other solanaceous crops (Defelice 2003). Given its plasticity and rich natural history, S. nigrum is an ideal model for the study of adaptive phenotypic responses. In order to identify the genetic basis of ecologically important responses, we developed molecular tools (transformation, microarrays, cloning) for S. nigrum (Schmidt et al. 2004). Moreover, ample resources for closely related solanaceous species, such as tomato and potato (SOL Genomics Network, http://www.sgn.cornell. edu/; NSF Potato Genome Project, http://www.potatogenome.org), facilitate gene discovery and characterization.

MOLECULAR ECOLOGY OF PLANT COMPETITION

213

Figure 13.2. Transcriptional responses of Solanum nigrum to competition and methyl jasmonate (MeJA) elicitation in field mesocosms analyzed with a 568-gene microarray (Schmidt and Baldwin 2006). (a) Venn diagram of the numbers of overlapping and non-overlapping significantly up- or down-regulated genes elicited by either the combination of competition and MeJA-elicitation or competition only. Pie charts summarize the gene categories of the regulated genes for each section of the Venn diagram. (b) The pie chart indicates the overall distribution of gene categories on the 568-gene microarray. (c,d) Degree of over-representation of the up- (c) and down-regulated (d) genes in the functional categories. The distribution of the differentially expressed genes over the functional categories is presented relative to the distribution of all genes on the microarray (set at 100% for each functional category). Photosynthesis and primary metabolism genes are commonly down regulated by both treatments. Genes related to signaling (jasmonate signaling in MeJA-treated plants, and NtETR1 [Nicotiana tabacum ethylene response factor] in competing unelicited plants), defense-related genes, and genes of unknown function distinguish unelicited and MeJA-induced plants. (Reproduced from Schmidt and Baldwin [2006] by permission of Blackwell Publishing.)

competing and noncompeting plants. In particular, genes related to primary metabolism appeared to be suppressed during competition as well as, notably, a gene coding for an ethylene receptor (Figure 13.2) (Schmidt and Baldwin 2006). Similarly, using larger scale microarrays, Horvath et al. (2006) examined the transcriptional responses of competition between the weed velvetleaf (Abutilon theophrasti) and maize (Zea mays). Under competition with velvetleaf, 253 maize genes were differentially regulated. Many of the down regulated maize genes indicated the competing weed negatively affected the crop’s primary metabolism since transcripts of genes involved in photosynthesis and carbohydrate and nitrogen metabolism, many of which are associated with development and growth, were down regulated. These transcriptional changes coincided with a significant decrease in nitrogen content and a 27% reduction of seed production in competing maize plants. Moreover, a large proportion of genes relating to signal transduction, such as transcription factors involved in phytohormone signaling, were suppressed during competition with velvetleaf. Subsequently, Horvath and Llewellyn (2007) performed the reciprocal analysis of velvetleaf gene expression during its competition with maize. In contrast to maize genes, velvetleaf genes that were related to photosynthesis, carbon metabolism, growth, and development were up regulated during competi-

214

WEEDY AND INVASIVE PLANT GENOMICS

tion. Furthermore, transcript abundance of genes involved in ethylene and gibberellin signaling was increased. Unfortunately, the transcriptional profiling of velvetleaf and maize was conducted at different stages of the interaction, which made it difficult to compare the responses between the two interacting species. Taken together, the results of gene expression studies during competition suggest that the differential expression of genes involved in hormonal signaling, i.e. ethylene and gibberellins, as well as phytochromes, potentially regulate the interaction. However, these experiments also make very clear that to identify key processes that mediate competition, we may have to resort to the high-throughput profiling of large-scale transcriptional, metabolic, morphological, and ecological changes in an unbiased manner within the window of natural permutations to establish correlations between those parameters and a plant’s competitive strength. Such an approach will require a holistic attitude in molecular biologists and an appreciation for reductionism in ecological research.

Conclusions And Outlook

How feasible will it be to identify “weed-genes”—those that allow a plant to behave like a weed? Clearly, the ability of a plant species to successfully compete and become a weed can sometimes be attributed to a single locus, pathway, or gene family, which suggests the existence of universal weed traits across species (Basu et al. 2004). However, such universal weed genes are often masked by the interaction between its product, the organism’s local environment, and the feedback from the environment. Such interactions alter the expression of these genes and make them difficult to identify. For example, a trait enabling a plant to grow tall is constrained by seasonal and environmental factors such as nutrient richness. Depending on those factors, a plant might alternate between root and shoot growth. Moreover, such growth might occur at particular developmental stages, and the fitness consequences of tallness will depend on the height and growth rates of the surrounding vegetation. This conditionality suggests that identifying master traits across species will require comparative data from different species living under diverse circumstances and, above all, a lot of luck. Therefore, instead of searching for common traits across species, it may be more feasible to search for traits of single species that play an important role across circumstances; in other words, to search for the jack-of-all-trades traits, the generalist traits which allow a single species to successfully colonize a wide range of habitats (Richards et al. 2006). Although intuitively the advantages of being a generalist are clear, it is hard to come up with plausible scenarios for how the jack-of-all-trades traits are maintained by natural selection, since most suggest simultaneous selection of many different traits. In contrast, the evolution of master traits seems quite straightforward, since plants exposed to a uniform selection regime will specialize with respect to a single trait or few relatively discrete traits. As an example, consider Solanum nigrum’s superior ability (see Box 13.2 and Figure 13.3) to compete in open high-nitrogen environments. Therefore, the existence of supposed jack-ofall-trades weeds suggest three selection scenarios: (1) jack weeds simply encounter a high degree of recurring environmental variability, thereby keeping their diverse collection of traits under sufficient selection pressure, (2) the jack weed phenotype is not constituted by a broad range of traits; rather, jack weeds are equipped with only a small number of broad-range traits—single genes that give rise to an extraordinary degree of plasticity, and (3) jack weeds harbor a high degree of intraspecific variation. This allows for local adaptations, rendering the species, but not the individuals, jacks-of-all-trades. These three scenarios are not mutually

MOLECULAR ECOLOGY OF PLANT COMPETITION

215

Figure 13.3. Black nightshade (Solanum nigrum; flowering [left] and fruit-bearing [right]) in its native habitat, an agricultural field near Jena, Germany. (Photograph on left reproduced by permission of Markus Hartl.)

exclusive. It may well be that weeds that compete in a wide range of habitats have a phenotype that is the product of single genes or traits with a broad range of function (e.g. thorns against herbivores) and within-genome diversity of complementary or additive traits (e.g. facultative apomixes in clonal species), as well as genetic variability within the species (e.g. as for some cases of herbicide resistance). Identifying these traits will require unbiased analysis of large-scale patterns. Such fingerprints will provide detailed descriptions of the functional characteristics of those weed traits that cannot be pinned down to single genes or comprehensive processes, providing an alternative for studying such mechanistically complex traits. Although there are many suitable techniques for profiling, only a few have been applied to field experiments (Kammenga et al. 2007), and this omission might lead to misconceptions and data artifacts (Kant and Baldwin 2007). To find answers to the questions posed in this chapter, these technologies need to be integrated with ecological research because the field is the only place where critical tests can be conducted.

References Adie BAT, Perez-Perez J, Perez-Perez MM, Manuel M, Godoy M, Sanchez-Serrano JJ, Schmelz EA, Solano R (2007) ABA is an essential signal for plant resistance to pathogens affecting JA biosynthesis and the activation of defenses in Arabidopsis. Plant Cell 19, 1665–1681. Ament K, Kant MR, Sabelis MW, Haring MA, Schuurink RC (2004) Jasmonic acid is a key regulator of spider mite-induced volatile terpenoid and methyl salicylate emission in tomato. Plant Physiology 135, 2025–2037. Baena-González E, Rolland F, Thevelein JM, Sheen J (2007) A central integrator of transcription networks in plant stress and energy signalling. Nature 448, 938–942. Bais HP, Weir TL, Perry LG, Gilroy S, Vivanco JM (2006) The role of root exudates in rhizosphere interactions with plants and other organisms. Annual Review of Plant Biology 57, 233–266.

216

WEEDY AND INVASIVE PLANT GENOMICS

Baker HG (1965) Characteristics and modes of origins of weeds. In: Baker HG, Stebbins GL, eds. The Genetics of Colonizing Species. Academic, New York, pp 147–172. Baker HG (1974) The evolution of weeds. Annual Review of Ecology and Systematics 5, 1–24. Baldwin IT, Halitschke R, Paschold A, von Dahl CC, Preston CA (2006) Volatile signaling in plant-plant interactions: “Talking trees” in the genomics era. Science 311, 812–815. Baldwin IT, Schultz JC (1983) Rapid changes in tree leaf chemistry induced by damage: evidence for communication between plants. Science 221, 277–279. Ballaré CL (1999) Keeping up with the neighbors: phytochrome sensing and other signaling mechanisms. Trends in Plant Science 4, 97–102. Basu C, Halfhill MD, Mueller TC, Stewart CN Jr. (2004) Weed genomics: new tools to understand weed biology. Trends in Plant Science 9, 391–398. Bauer WD, Mathesius U (2004) Plant responses to bacterial quorum sensing signals. Current Opinion in Plant Biology 7, 429–433. Bengough AG, Bransby MF, Hans J, McKenna SJ, Roberts TJ, Valentine TA (2006) Root responses to soil physical conditions; growth dynamics from field to cell. Journal of Experimental Botany 57, 437–447. Binder BM (2007) Rapid kinetic analysis of ethylene growth responses in seedlings: New insights into ethylene signal transduction. Journal of Plant Growth Regulation 26, 131–142. Black CC, Chen TM, Brown R (1969) Biochemical basis for plant competition. Weed Science 17, 338–344. Bohs L (2005) Major clades in Solanum based on ndhF sequence data. In: Keating RC, Hollowell VC, Croat TB, eds. A Festschrift for William G. D’Arcy: The legacy of a taxonomist. Monographs in Systematic Botany 104, Missouri Botanical Garden, 27–49. Briggs WR, Christie JM (2002) Phototropins 1 and 2: versatile plant blue-light receptors. Trends in Plant Science 7, 204–210. Bunce JA, Miller LN, Chabot BF (1977) Competitive exploitation of soil water by five eastern North American tree species. Botanical Gazette 138, 168–173. Casal JJ, Yanovsky MJ (2005) Regulation of gene expression by light. International Journal of Developmental Biology 49, 501–511. Chailakhyan MK (1937) Concerning the nature of plant development processes. Doklady Akademii Nauk SSSR 16, 227–230. Chamberlain K, Guerrieri E, Pennacchio F, Pettersson J, Pickett JA, Poppy GM, Powell W, Wadhams LJ, Woodcock CM (2001) Can aphid-induced plant signals be transmitted aerially and through the rhizosphere? Biochemical Systematics and Ecology 29, 1063–1074. Chevrot R, Rosen R, Haudecoeur E, Cirou A, Shelp BJ, Ron E, Faure D (2006) GABA controls the level of quorum-sensing signal in Agrobacterium tumefaciens. Proceedings of the National Academy of Sciences of the United States of America 103, 7460–7464. Choi J, Hwang I (2007) Cytokinin: perception, signal transduction, and role in plant growth and development. Journal of Plant Biology 50, 98–108. Chow B, McCourt P (2006) Plant hormone receptors: perception is everything. Genes and Development 20, 1998–2008. Conrath U, Beckers GJ, Flors V, García-Agustín P, Jakab G, Mauch F, Newman MA, Pieterse CM, Poinssot B, Pozo MJ, Pugin A, Schaffrath U, Ton J, Wendehenne D, Zimmerli L, Mauch-Mani B (2006) Priming: Getting ready for battle. Molecular Plant-Microbe Interactions 19, 1062–1071. Corbesier L, Coupland G (2006) The quest for florigen: a review of recent progress. Journal of Experimental Botany 57, 3395–3403. Corbesier L, Vincent C, Jang S, Fornara F, Fan Q, Searle I, Giakountis A, Farrona S, Gissot L, Turnbull C, Coupland G (2007) FT protein movement contributes to long-distance signaling in floral induction of Arabidopsis. Science 316, 1030–1033. Darlington W (1859) American Weeds and Useful Plants. A.O. Moore, New York, 460 p. Darwin C (1880) The Power of Movements in Plants. John Murray, London, 592 p. de Boer JG, Dicke M (2004) The role of methyl salicylate in prey searching behavior of the predatory mite Phytoseiulus persimilis. Journal of Chemical Ecology 30, 255–271. de Kroon H, Huber H, Stuefer JF, van Groenendael JM (2005) A modular concept of phenotypic plasticity in plants. New Phytologist 166, 73–82. de Smet I, Zhang H, Inzé D, Beeckman T (2006) A novel role for abscisic acid emerges from underground. Trends in Plant Science 11, 434–439. Defelice M (2003) The black nightshades, Solanum nigrum L. et al.—Poison, poultice, and pie. Weed Technology 17, 421–427. Dicke M, Agrawal AA, Bruin J (2003) Plants talk, but are they deaf? Trends in Plant Science 8, 403–405.

MOLECULAR ECOLOGY OF PLANT COMPETITION

217

Dicke M, Dijkman H (2001) Within-plant circulation of systemic elicitor of induced defence and release from roots of elicitor that affects neighbouring plants. Biochemical Systematics and Ecology 29, 1075–1087. Doust AN (2007) Grass architecture: genetic and environmental control of branching. Current Opinion in Plant Biology 10, 21–25. Dreher K and Callis J (2007) Ubiquitin, hormones and biotic stress in plants. Annals of Botany 99, 787–822. Driouich A, Durand C, Vicré-Gibouin M (2007) Formation and separation of root border cells. Trends in Plant Science 12, 14–19. Dudley SA, File AL (2007) Kin recognition in an annual plant. Biology Letters 3, 435–438. Dukes JS and Mooney HA (1999) Does global change increase the success of biological invaders? Trends in Ecology & Evolution 14, 135–139. Eapen D, Barroso ML, Campos ME, Ponce G, Corkidi G, Dubrovsky JG, Cassab GI (2003) A no hydrotropic response root mutant that responds positively to gravitropism in Arabidopsis. Plant Physiology 131, 536–546. Eapen D, Barroso ML, Ponce G, Campos ME, Cassab GI (2005) Hydrotropism: root growth responses to water. Trends in Plant Science 10, 44–50. Edmonds JM, Chweya JA (1997) Black Nightshades, Solanum nigrum L. and Related Species, Gatersleben, Rome, 113 p. Esmon CA, Pedmale UV, Liscum E (2005) Plant tropisms: providing the power of movement to a sessile organism. International Journal of Developmental Biology 49, 665–674. Farmer EE (2007) Plant biology: jasmonate perception machines. Nature 448, 659–660. Forde BG (2002) Local and long-range signaling pathways regulating plant responses to nitrate. Annual Review of Plant Biology 53, 203–224. Franklin KA, Larner VS, Whitelam GC (2005) The signal transducing photoreceptors of plants. International Journal of Developmental Biology 49, 653–664. Franklin KA, Whitelam GC (2005) Phytochromes and shade-avoidance responses in plants. Annals of Botany 96, 169–175. Fujioka S and Yokota T (2003) Biosynthesis and metabolism of brassinosteroids. Annual Review of Plant Biology 54, 137–164. Gagne JM, Smalle J, Gingerich DJ, Walker JM, Yoo S-D, Yanagisawa S, Vierstra RD (2004) Arabidopsis EIN3-binding F-box 1 and 2 form ubiquitin-protein ligases that repress ethylene action and promote growth by directing EIN3 degradation. Proceedings of the National Academy of Sciences of the United States of America 101, 6803–6808. Gao MS, Teplitski M, Robinson JB, Bauer WD (2003) Production of substances by Medicago truncatula that affect bacterial quorum sensing. Molecular Plant-Microbe Interactions 16, 827–834. Grime JP (2001) Plant Strategies, Vegetation Processes, and Ecosystem Properties, 2nd ed. John Wiley, Chichester, 417 p. Guzman P, Ecker JR (1990) Exploiting the triple response of Arabidopsis to identify ethylene-related mutants. Plant Cell 2, 513–523. Halford NG, Hey S, Jhurreea D, Laurie S, McKibbin RS, Zhang Y, Paul MJ (2004) Highly conserved protein kinases involved in the regulation of carbon and amino acid metabolism. Journal of Experimental Botany 55, 35–42. Hall BP, Shakeel SN, Schaller GE (2007) Ethylene receptors: Ethylene perception and signal transduction. Journal of Plant Growth Regulation 26, 118–130. Hamel LP, Nicole MC, Sritubtim S, Morency MJ, et al. (2006) Ancient signals: comparative genomics of plant MAPK and MAPKK gene families. Trends in Plant Science 11, 192–198. Hardtke CS, Dorcey E, Osmont KS, Sibout R (2007) Phytohormone collaboration: zooming in on auxin-brassinosteroid interactions. Trends in Cell Biology 17, 485–492. Hawes MC, Bengough G, Cassab G, Ponce G (2003) Root caps and rhizosphere. Journal of Plant Growth Regulation 21, 352–367. Hawes MC, Gunawardena U, Miyasaka S, Zhao X (2000) The role of root border cells in plant defense. Trends in Plant Science 5, 128–133. He Z, von Caemmerer S, Hudson GS, Price GD, Badger MR, Andrews TJ (1997) Ribulose-1,5-bisphosphate carboxylase/ oxygenase activase deficiency delays senescence of ribulose-1,5-bisphosphate carboxylase/oxygenase but progressively impairs its catalysis during tobacco leaf development. Plant Physiology 115, 1569–1580. Hermanutz L, Weaver S (1991) Variability in temperature-dependent germination in eastern black nightshade (Solanum ptycanthum). Canadian Journal of Botany 69, 1463–1470. Hermanutz L, Weaver S (1996) Agroecotypes or phenotypic plasticity? Comparison of agrestal and ruderal populations of the weed Solanum ptycanthum. Oecologia 105, 271–280. Hermsmeier D, Schittko U, Baldwin IT (2001) Molecular interactions between the specialist herbivore Manduca sexta (Lepidoptera, Sphingidae) and its natural host Nicotiana attenuata. I. Large-scale changes in the accumulation of growth- and defense-related plant mRNAs. Plant Physiology 125, 683–700.

218

WEEDY AND INVASIVE PLANT GENOMICS

Hetherington AM, Brownlee C (2004) The generation of Ca2+ signals in plants. Annual Review of Plant Biology 55, 401–427. Hibberd JM, Quick WP (2002) Characteristics of C4 photosynthesis in stems and petioles of C3 flowering plants. Nature 415, 451–454. Hirayama T, Shinozaki K (2007) Perception and transduction of abscisic acid signals: keys to the function of the versatile plant hormone ABA. Trends in Plant Science 12, 343–351. Hodge A (2004) The plastic plant: root responses to heterogeneous supplies of nutrients. New Phytologist 162, 9–24. Holm LG, Plucknett DL, Pancho JV, Herberger JP (1991) The World’s Worst Weeds: Distribution and Biology. Krieger Publishing, Malabar, 609 p. Holton N, Caño-Delgado A, Harrison K, Montoya T, Chory J, Bishop GJ (2007) Tomato BRASSINOSTEROID INSENSITIVE1 is required for systemin-induced root elongation in Solanum pimpinellifolium but is not essential for wound signaling. Plant Cell 19, 1709–1717. Horvath D, Llewellyn DCPI (2007) Heterologous hybridization of cotton microarrays with velvetleaf (Abutilon theophrasti) reveals physiological responses due to corn competition. Weed Science 55, 546–557. Horvath DP, Gulden R, Clay SA (2006) Microarray analysis of late-season velvetleaf (Abutilon theophrasti) effect on corn. Weed Science 54, 983–994. Huq E (2006) Degradation of negative regulators: a common theme in hormone and light signaling networks? Trends in Plant Science 11, 4–7. Jiang F, Hartung W (2007) Long-distance signaling of abscisic acid (ABA): the factors regulating the intensity of the ABA signal. Journal of Experimental Botany. Jiao Y, Lau OS, Deng XW (2007) Light-regulated transcriptional networks in higher plants. Nature Reviews Genetics 8, 217–230. Jonak C, Ökrész L, Bögre L, Hirt H (2002) Complexity, cross talk and integration of plant MAP kinase signaling. Current Opinion in Plant Biology 5, 415–424. Jones AM, Assmann SM (2004) Plants: the latest model system for G-protein research. EMBO Reports 5, 572–578. Jones JD, Dangl JL (2006) The plant immune system. Nature 444, 323–329. Kammenga JE, Herman MA, Ouborg NJ, Johnson L, Breitling R (2007) Microarray challenges in ecology. Trends in Ecology and Evolution 22, 273–279. Kant M, Baldwin IT (2007) The ecogenetics and ecogenomics of plant-herbivore interactions: rapid progress on a slippery road. Current Opinion in Genetics and Development 17, 519–524. Karban R (2007) Experimental clipping of sagebrush inhibits seed germination of neighbors. Ecology Letters 10: 791–797. Kennedy RA, Laetsch WM (1974) Plant species intermediate for C3, C4 photosynthesis. Science 184, 1087–1089. Kesselmeier J, Ciccioli P, Kuhn U, Stefani P, Biesenthal T, Rottenberger S, Wolf A, Vitullo M, Valentini R, Nobre A, Kabat P, Andreae M (2002) Volatile organic compound emissions in relation to plant carbon fixation and the terrestrial carbon budget. Global Biogeochemical Cycles 16, 1126. Kessler A, Halitschke R, Diezel C, Baldwin IT (2006) Priming of plant defense responses in nature by airborne signaling between Artemisia tridentata and Nicotiana attenuata. Oecologia 148, 280–292. Kobayashi A, Takahashi A, Kakimoto Y, Miyazawa Y, Fujii N, Higashitani A, Takahashi H (2007) A gene essential for hydrotropism in roots. Proceedings of the National Academy of Sciences of the United States of America 104, 4724–4729. Lee KH, Piao HL, Kim HY, Choi SM, Jiang F, Hartung W, Hwang I, Kwak JM, Lee IJ, Hwang I (2006) Activation of glucosidase via stress-induced polymerization rapidly increases active pools of abscisic acid. Cell 126, 1109–1120. Loake G, Grant M (2007) Salicylic acid in plant defense—the players and protagonists. Current Opinion in Plant Biology 10, 466–472. López-Bucio J, Cruz-Ramírez A, Herrera-Estrella L (2003) The role of nutrient availability in regulating root architecture. Current Opinion in Plant Biology 6, 280–287. Lough TJ, Lucas WJ (2006) Integrative plant biology: role of phloem long-distance macromolecular trafficking. Annual Review of Plant Biology 57, 203–232. Lu CA, Lin CC, Lee KW, Chen JL, Huang LF, Ho SL, Liu HJ, Hsing YI, Yu SM (2007) The SnRK1A protein kinase plays a key role in sugar signaling during germination and seedling growth of rice. Plant Cell 19, 2484–2499. Mack RN, Simberloff D, Lonsdale WM, Evans H, Clout M, Bazzaz FA (2000) Biotic invasions: causes, epidemiology, global consequences, and control. Ecological Applications 10, 689–710. McGiffen M, Masiunas J, Hesketh J (1992a) Competition for light between tomato and nightshade (Solanum nigrum or S. ptycanthum). Weed Science 40, 220–226. McGiffen M, Masiunas J, Huck M (1992b) Tomato and nightshade (Solanum nigrum L. and S. ptycanthum Dun.) effects on soil water content. Journal of the American Society for Horticultural Science 117, 730–735.

MOLECULAR ECOLOGY OF PLANT COMPETITION

219

McGurl B, Pearce G, Orozco-Cardenas M, Ryan CA (1992) Structure, expression, and antisense inhibition of the systemin precursor gene. Science 255, 1570–1573. Milliman L, Riechers D, Wax L, Simmons F (2003) Characterization of two biotypes of imidazolinone-resistant eastern black nightshade (Solanum ptycanthum). Weed Science 51, 139–144. Mirabella R, Rauwerda H, Struys EA, Jakobs C, Triantaphylidès C, Haring MA, Schuurink RC (2008) The Arabidopsis her1 mutant implicates GABA in E-2-hexenal responsiveness. Plant Journal 53, 197–213. Montoya T, Nomura T, Farrar K, Kaneta T, Yokota T, Bishop GJ (2002) Cloning the tomato curl3 gene highlights the putative dual role of the leucine-rich repeat receptor kinase BRI1/SR160 in plant steroid hormone and peptide hormone signaling. Plant Cell 14, 3163–3176. Ninkovic V (2003) Volatile communication between barley plants affects biomass allocation. Journal of Experimental Botany 54, 1931–1939. Nordstrom A, Tarkowski P, Tarkowska D, Norbaek R, Astot C, Dolezal K, Sandberg G (2004) Auxin regulation of cytokinin biosynthesis in Arabidopsis thaliana: a factor of potential importance for auxin-cytokinin-regulated development. Proceedings of the National Academy of Sciences of the United States of America 101, 8039–8044. Ogg AG, Rogers B (1989) Taxonomy, distribution, biology, and control of black nightshade (Solanum nigrum) and related species in the US and Canada. Reviews of Weed Science 4, 25–58. Ongaro V, Leyser O (2008) Hormonal control of shoot branching. Journal of Experimental Botany 59, 67–74. Osmont KS, Sibout R, Hardtke CS (2007) Hidden branches: developments in root system architecture. Annual Review of Plant Biology 58, 93–113. Otsuka M, Kenmoku H, Ogawa M, Okada K, Mitsuhashi W, Sassa T, Kamiya Y, Toyomasu T, Yamaguchi S (2004) Emission of ent-kaurene, a diterpenoid hydrocarbon precursor for gibberellins, into the headspace from plants. Plant Cell Physiology 45, 1129–1138. Park S-W, Kaimoyo E, Kumar D, Mosher S, Klessig DF (2007) Methyl salicylate is a critical mobile signal for plant systemic acquired resistance. Science 318, 113–116. Paschold A, Halitschke R, Baldwin IT (2006) Using ‘mute’ plants to translate volatile signals. Plant Journal 45, 275–291. Pierik R, Visser EJW, de Kroon H, Voesenek L (2003) Ethylene is required in tobacco to successfully compete with proximate neighbours. Plant Cell and Environment 26, 1229–1234. Pimentel D (2000) Environmental and economic costs of nonindigenous species in the United States. Bioscience 50, 53–65. Polge C, Thomas M (2007) SNF1/AMPK/SnRK1 kinases, global regulators at the heart of energy control? Trends in Plant Science 12, 20–28. Portis AR (2003) Rubisco activase—Rubisco’s catalytic chaperone. Photosynthesis Research 75, 11–27. Quail PH (2002) Phytochrome photosensory signalling networks. Nature Reviews Molecular Cell Biology 3, 85–93. Remans T, Nacry P, Pervent M, Filleur S, Diatloff E, Mounier E, Tillard P, Forde BG, Gojon A (2006) The Arabidopsis NRT1.1 transporter participates in the signaling pathway triggering root colonization of nitrate-rich patches. Proceedings of the National Academy of Sciences of the United States of America 103, 19206–19211. Rhoades DF (1983) Responses of alder and willow to attack by tent caterpillars and webworms: evidence for pheromonal sensitivity of willows. In: Hedin PA, ed. Plant Resistance to Insects. American Chemical Society, Washington, pp 55–68. Richards CL, Bossdorf O, Muth NZ, Gurevitch J, Pigliucci M (2006) Jack of all trades, master of some? On the role of phenotypic plasticity in plant invasions. Ecology Letters 9, 981–993. Rosenzweig ML (2001) The four questions: what does the introduction of exotic species do to diversity? Evolutionary Ecology Research 3, 361–367. Ryan CA, Pearce G, Scheer J, Moura DS (2002) Polypeptide hormones. Plant Cell 14, S251–264. Sage RF (2004) The evolution of C4 photosynthesis. New Phytologist 161, 341–370. Sakakibara H, Takei K, Hirose N (2006) Interactions between nitrogen and cytokinin in the regulation of metabolism and development. Trends in Plant Science 11, 440–448. Sauter A, Davies WJ, Hartung W (2001) The long-distance abscisic acid signal in the droughted plant: the fate of the hormone on its way from root to shoot. Journal of Experimental Botany 52, 1991–1997. Schachtman DP, Shin R (2007) Nutrient sensing and signaling: NPKS. Annual Review of Plant Biology 58, 47–69. Scheer JM, Ryan CA (2002) The systemin receptor SR160 from Lycopersicon peruvianum is a member of the LRR receptor kinase family. Proceedings of the National Academy of Sciences of the United States of America 99, 9585–9590. Schmidt DD, Baldwin IT (2006) Transcriptional responses of Solanum nigrum to methyl jasmonate and competition: a glasshouse and field study. Functional Ecology 20, 500–508. Schmidt DD, Kessler A, Kessler D, Schmidt S, Lim M, Gase K, Baldwin IT (2004) Solanum nigrum: A model ecological expression system. Molecular Ecology 13, 981–995.

220

WEEDY AND INVASIVE PLANT GENOMICS

Schmidt S, Baldwin IT (2008) Down-regulation of systemin after herbivory is associated with increased root allocation and competitive ability in Solanum nigrum. Oecologia 159, 473–482. Schroeder JI, Kwak JM, Allen GJ (2001) Guard cell abscisic acid signalling and engineering drought hardiness in plants. Nature 410, 327–330. Schwachtje J, Minchin PE, Jahnke S, van Dongen JT, Schittko U, Baldwin IT (2006) SNF1-related kinases allow plants to tolerate herbivory by allocating carbon to roots. Proceedings of the National Academy of Sciences of the United States of America 103, 12935–12940. Semchenko M, John EA, Hutchings MJ (2007) Effects of physical connection and genetic identity of neighbouring ramets on root-placement patterns in two clonal species. New Phytologist 176, 644–654. Shelp BJ, Bown AW, Faure D (2006) Extracellular gamma-aminobutyrate mediates communication between plants and other organisms. Plant Physiology 142, 1350–1352. Shulaev V, Silverman P, Raskin I (1997) Airborne signalling by methyl salicylate in plant pathogen resistance. Nature 385, 718–721. Somerville CR, Portis AR, Ogren WL (1982) A mutant of Arabidopsis thaliana which lacks activation of RuBP carboxylase in vivo. Plant Physiology 70, 381–387. Svistoonoff S, Creff A, Reymond M, Sigoillot-Claude C, Ricaud L, Blanchet A, Nussaume L, Desnos T (2007) Root tip contact with low-phosphate media reprograms plant root architecture. Nature Genetics 39, 792–796. Symons GM, Ross JJ, Jager CE, Reid JB (2008) Brassinosteroid transport. Journal of Experimental Botany 59, 17–24. Tamaki S, Matsuo S, Wong HL, Yokoi S, Shimamoto K (2007) Hd3a protein is a mobile flowering signal in rice. Science 316, 1033–1036 Tiffin P (2000) Mechanisms of tolerance to herbivore damage: What do we know? Evolutionary Ecology 14, 523–536. Ton J, D’Alessandro M, Jourdie V, Jakab G, Karlen D, Held M, Mauch-Mani B, Turlings TC (2007) Priming by airborne signals boosts direct and indirect resistance in maize. Plant Journal 49, 16–26. Ueguchi-Tanaka M, Nakajima M, Motoyuki A, Matsuoka M (2007) Gibberellin receptor and its role in gibberellin signaling in plants. Annual Review of Plant Biology 58, 183–198. van Schie CCN, Ament K, Schmidt A, Lange T, Haring MA, Schuurink RC (2007) Geranyl diphosphate synthase is required for biosynthesis of gibberellins. Plant Journal 52, 752–762. Vitousek PM, Mooney HA, Lubchenco J, Melillo JM (1997) Human domination of earth’s ecosystems. Science 277, 494–499. Volenberg D, Stoltenberg D, Boerboom C (2000) Solanum ptycanthum resistance to acetolactate synthase inhibitors. Weed Science 48, 399–401. von Caemmerer S, Hendrickson L, Quinn V, Vella N, Millgate AG, Furbank RT (2005) Reductions of Rubisco activase by antisense RNA in the C4 plant Flaveria bidentis reduces Rubisco carbamylation and leaf photosynthesis. Plant Physiology 137, 747–755. von Dahl CC, Baldwin IT (2007) Deciphering the role of ethylene in plant-herbivore interactions. Journal of Plant Growth Regulation. 26, 201–209. Wang R, Okamoto M, Xing X, Crawford NM (2003) Microarray analysis of the nitrate response in Arabidopsis roots and shoots reveals over 1,000 rapidly responding genes and new linkages to glucose, trehalose-6-phosphate, iron, and sulfate metabolism. Plant Physiology 132, 556–567. Wasternack C (2007) Jasmonates: an update on biosynthesis, signal transduction and action in plant stress response, growth and development. Annals of Botany 100, 681–697. Weinig C, Brock MT, Dechaine JA, Welch SM (2007) Resolving the genetic basis of invasiveness and predicting invasions. Genetica 129, 205–216. Weiss D, Ori N (2007) Mechanisms of cross talk between gibberellin and other hormones. Plant Physiology 144, 1240–1246. Woodward AW, Bartel B (2005) Auxin: regulation, action, and interaction. Annals of Botany 95, 707–735. Wu J, Kurten EL, Monshausen G, Hummel GM, Gilroy S, Baldwin IT (2007) NaRALF, a peptide signal essential for the regulation of root hair tip apoplastic pH in Nicotiana attenuata, is required for root hair development and plant growth in native soils. Plant Journal 52, 877–890. Xue H, Chen X, Li G (2007) Involvement of phospholipid signaling in plant growth and hormone effects. Current Opinion in Plant Biology 10, 483–489. Zavala JA, Baldwin IT (2006) Jasmonic acid signalling and herbivore resistance traits constrain regrowth after herbivore attack in Nicotiana attenuata. Plant Cell and Environment 29, 1751–1760. Zhang H, Forde BG (1998) An Arabidopsis MADS box gene that controls nutrient-induced changes in root architecture. Science 279, 407–409.

14

Genomics And Weeds: A Synthesis Stephen O. Duke, Scott R. Baerson, and Jonathan Gressel

Introduction

Most of this book deals with genomics of weeds (including invasive plants) as it relates to weed biology. Through the design and development of transgenic, herbicide-resistant crops, genomics and molecular biology have had more impact on weed management worldwide than any other technology since the advent of synthetic herbicides (Duke and Powles 2008). Although acquisition of genomic information with regard to weeds and weediness is expanding rapidly, the progress has thus far not generated significant practical effects in dealing with weeds. In this chapter we discuss genomics with regard to future solutions to weed problems. The genomics of the syndrome called weediness has been hardly studied or understood. Why should it be important? Humans (and other animals) typically pick on cohorts that are “different” in sadistic ways, often with dire effects for those that are different. For the benefit of agriculture, we must be harsh with weeds, especially those closely related to crops. The best way to deal with these weeds is to understand where they are different so we can exploit those differences to facilitate weed management. Examples are given in later sections on how those genomic differences might be used for weed control. For example, we must know what controls seed shatter in shattercane, wild sunflowers, and other weeds that are closely related to crops (for which there are no conventional controls) or what genomically confers underground rhizomes with their storage capacity and dormancy mechanisms. Agriculture has developed a “chemical dependency” when it comes to weed control. This dependency has become almost universal in the developed world because synthetic herbicides have been so efficient and cost effective. Despite sometimes undesirable side effects, the addicted farmers consider the chemicals wonderful, as long as the “high” lasts. Farmers can end up on a bad trip when herbicide resistance evolves, which often is a function of a dependency on a single chemical or family of chemicals with a single molecular target site. This is especially the case when only a single herbicide is cost effective, as has been the case with glyphosate and glyphosate-resistant crops (Duke and Powles 2008). Although evolution of the first cases of glyphosate resistance had a long lag period (twenty-two years), glyphosate-resistant weeds are now being reported at an alarming rate, especially in glyphosate-resistant crops. For example, glyphosate-resistant Sorghum halepense evolved in Argentina in glyphosate-resistant soybeans and has spread throughout that country in only a few years (Vila-Aiub et al. 2007, 2008; Valverde et al. 2007). There are no systemic herbicides other than glyphosate that will kill both the shoots and the underground rhizomes of this weed. In some cases there is no adequate herbicide. There are no good chemical control options for many grass weeds such as the various species of Echinochloa, Avena, Alopecurus, Lolium, and Setaria, which have evolved resistance to most of the selective herbicides that can control them in major grain crops such as wheat, rice, and maize. Worse yet, some weeds are conspecific and congeneric and freely hybridize with crops, e.g. Avena spp. with oats, wild beets with sugar beets, shattercane with sorghum, wild sunflowers with sunflowers, and the worst of all, 221

222

WEEDY AND INVASIVE PLANT GENOMICS

weedy, feral, or red rice with rice. The chemical industry’s only effective answer in some cases has been to apply a herbicide safener to the crop so that the weedy relative can be controlled. So far, such solutions are available only for the sorghum pair, and in the rest of the cases, there are no selective chemical solutions. Urban allergenic weeds are a major problem, especially in poorer, tropical countries with disturbed or wild areas in cities; there are no selectively effective herbicides. To exacerbate the problems with chemistry, no herbicides with major new modes of action have been commercialized in decades—yes, decades. In this chapter, we discuss how genomics can be used to solve some of these problems, especially in areas that have not been covered in previous chapters. We try to highlight limitations and knowledge gaps.

From Fundamental Information To Practical Solutions Herbicide And Herbicide Target Site Discovery

Herbicides with new target sites are needed to fight evolved herbicide resistance, provide weed management tools with more useful selectivity (new markets), and replace older herbicides lost because of undesirable toxicology and/or environmental effects, as well as unfavorable markets. Despite these needs, no significant new product with a new molecular target site has been introduced for more than twenty years. There are fewer than twenty target sites exploited by commercial herbicides. Indeed, the huge technical (so far) (Dill et al. 2008; Duke and Powles 2008) and economic success (Gianessi 2008) of the use of glyphosate with glyphosateresistant crops, and the concomitant reduced use of many other herbicides (Gianessi 2008), has probably prevented the introduction of new products and caused or hastened the demise of some old products. Glyphosate-resistant crops significantly devalued the herbicide market (Nelson and Bullock 2003), resulting in abandonment of research on new herbicides. As pesticide companies reach diminishing returns with brute force screening of huge numbers of compounds for herbicidal activity, they have tried to develop less labor-intensive strategies for discovery. Initially, they tried what were termed “biorational” approaches that were based on understanding inhibitor interactions with known molecular target sites (Gressel 2002). This approach relied on in vitro assays that could be miniaturized and automated. The biorational approach has improved the activity of certain herbicide classes, but has apparently not led to the discovery and development of commercial herbicides with new modes of action. This was followed by the combinatorial chemistry stratagem, which involved screening even larger numbers of compounds generated by automated synthesis. At least this approach was generally coupled with a whole plant bioassay, making it more probable that a new target site would be discovered. This tactic has apparently not been very productive either, although the requirements for launching a new product may be more rigorous due the advent of glyphosateresistant crops. Genomics has ushered in a new approach for herbicide discovery. If the success of this strategy is judged by the introduction of new products with new molecular target sites, the concept of using genomics for discovering new target sites for herbicides using Arabidopsis has apparently been a failure, too (Gressel 2002; see also Chapter 3). However, considering the new paradigm created by transgenic, herbicide-resistant crops, potential new products resulting from such approaches, although effective, might have been judged as uncompetitive with existing products, particularly glyphosate.

GENOMICS AND WEEDS: A SYNTHESIS

223

The development and application of high throughput, systematic, genome-wide approaches to ascertain gene function (functional genomics) using bioinformatics for the analysis of gene function was widely thought to have the potential to lead to the discovery of many new herbicide molecular target sites (Bouchez and Hofte 1998; Cole and Rodgers 2000; Cole and Rodgers 2000; Cole et al. 2000; Hay 1998; McLaren, 2000). Herbicide target site discovery genomics can be more specifically defined as the systematic production of directed mutations or knockouts of specific genes, followed by analysis of phenotype to see if the genetic alteration is lethal. Identification of such target sites would be followed by discovery and/or design of good inhibitors of these proteins. However, this latter part of the strategy is not trivial, because some enzymes may have very few good inhibitors with unpredictable structures. For example, glyphosate appears to be the only good herbicidal inhibitor of 5-enolpyruvyl-shikimate-3phosphate synthase (EPSPS). We doubt whether chemists and biochemists would have found this inhibitor in a screen for inhibitors. From Genomics In Pharmaceutical Discovery To Genomics In Herbicide Discovery. As with the pharmaceutical industry, the use of genomics to elucidate targets and develop rapid screens for chemicals affecting those targets was quickly adopted by the pesticide industry (Berg et al. 1999). After altered expression of a gene of known or unknown function is found to cause lethality, there has been a tendency of researchers to rush to the patent office to make broad claims of chemicals inhibiting the gene product as herbicidal. This approach of patenting all inhibitors of molecular target sites was first used in the pre-genomics era to claim any chemical inhibiting a specific enzyme (superoxide dismutase) and thus synergizing reactive oxygen species-generating herbicides such as paraquat (Gressel and Shaaltiel 1993). The initial examiner did not accept the concept of patenting a biochemical target as sufficiently novel for patenting, but the board of appeals overruled (Steiner et al. 1992), and a precedent was seemingly set that one can essentially patent a target site. This approach was embraced by the herbicide industry, despite the lack of proven efficacy of such techniques in drug or pesticide design, as well as the high cost. The perceived potential of patenting possible future target sites for herbicide discovery, thereby eliminating competitors for entire classes of herbicides, probably added to the allure of this approach. The following example, elaborated by Berg et al. (1999), demonstrates the rapidity with which new targets can be elucidated (and patented) using genomics/bioinformatics. It also demonstrates that genomics may not have been needed from a discovery point of view. The CLA1 mutant of Arabidopsis with blocked chloroplast development was reported to be the cause of an albino phenotype (Mandel et al. 1996). At about the same time, a new pathway of plant isoprenoid biosynthesis was elucidated, whereby deoxyxylulose was found to produce isoprene units contributing to phytol and carotenoid synthesis. CLA1 was identified a year later as coding for deoxyxylulose-5-phosphate synthase (DXP synthase), a key enzyme identified in the new pathway (Lange et al. 1998), and a patent was filed claiming this new target for herbicides. Of equal significance, fosmidomycin, a known natural phytotoxin, was found to act by inhibiting deoxylulase phosphate reductoisomerase, the next enzyme in the pathway (Kuzuyama et al. 1998; Zeidler et al. 1998). More recently, the 5-keto metabolite of the commercial herbicide clomazone was found to inhibit isoprenoid synthesis by inhibition of DXP synthase (Ferhatoglu and Barrett 2006). Thus, good inhibitors of two steps in the pathway were available, and one of these had been used as a commercial herbicide (a plant metabolite of the proherbicide clomazone) for some years before its target site was known. The mixture of Arabidopsis genetics and bioinformatics

224

WEEDY AND INVASIVE PLANT GENOMICS

and good biochemistry led to discovery of a new target pathway in less than four years. However, before genomics, chemical synthesis and quantitative structure-activity relationship (QSAR) analysis apparently led to a viable herbicide for this pathway without knowing the target site. This story poses a conundrum for patented target sites. What are the legal implications when an already patented compound is found to target a newly patented target site? In the example above, there are thus far no new products since the genomic work was done and the target site patented. The question remains as to whether this process provides new modes of action and/ or new herbicides more rapidly than non-genomic approaches. Once such potential targets are identified by genomics, simple, miniaturized, high throughput screens are developed for assaying new compounds generated by combinatorial and conventional chemistry. Many plant genomics companies have combined genomics with tests to ascertain if an antisensed or cosuppressively overexpressed DNA sequence can cause phytotoxicity. This approach involves database and laboratory searches for the target enzyme and its activity, culminating in patent claims for anything that inhibits the target as well as a development of a high throughput screen to find chemicals that will provide the same result as the antisensed gene. In many cases, small genomics companies have been commissioned by agrochemical companies to identify target sites with this approach. Knocking Out Genes For Target Site Identification. If reduction in the expression of a particular gene leads to a sick or dead plant, the gene product has potential to be a herbicide target site. Several methods that are based on RNA expression can be used to partially silence a gene (Matzke et al. 2001). The DNA chosen for partial gene suppression or antisensing could be from a known gene or it can be a new open reading frame. To do this, an appropriate DNA construct is made in which the cDNA is cloned in reverse orientation and is introduced into a plant. The cDNA produces mRNA in the antisense orientation, which, in a manner not yet fully clear, partially suppresses translation of the complimentary “sense” mRNA. Höfgen et al. (1994, 1995, 1999) pioneered this approach for defining possible herbicide potential. Many such sites elucidated in this manner are described by Gressel (2002). Table 14.1 provides a partial list of herbicide target sites discovered by antisense and RNAi technologies. It is obvious that some of these would be good target sites (e.g. chlorophyll synthase), so the genomic approach has not provided much new information, although it has

Table 14.1. Some herbicide target sites identified by antisense or RNAi technologies. Target site or function

Reference

Aldolase Carbonic anhydrase Chlorophyll synthase Coproporphyrinogen oxidase Dehydroquinate dehydrase/shikimate dehydrogenase β-Cystothionine lyase Ferredoxin:NADP reductase Geranylgeranyl reductase Glutamine-semialdehyde amino transferase Sphingolipid 4-hyroxylase Thioredoxin Transketolase Uroporphyinogen decarboxylase

Haake et al. 1998 Davis et al. 2004 Aslamkhan et al. 2005 Kruse et al. 1995 Freund et al. 2003 Maimann et al. 2000 Wagner et al. 2000 Tanaka et al. 1999 Höfgen et al. 1995 Todd et al. 2004 Kurnik et al. 2003 Henkes et al. 2001 Mock and Grimm 1997

GENOMICS AND WEEDS: A SYNTHESIS

225

provided a company with intellectual property rights to the target site. In other cases (e.g. sphingolipid 4-hyroxylase), it might not be so clear that the site would be a good herbicide target site without the antisense phenotype information. The example of antisensed transketolase shows how the many effects following antisense suppression are similar to that of an active herbicide (Henkes et al. 2001). Transketolase is involved in both C6 and C5 oxidative metabolism, producing a precursor for the shikimate pathway. When the level of this enzyme was reduced more than 20%, ribulose-1,5-bisphosphate regeneration and photosynthesis were inhibited, especially under high irradiance, approximating field-level light intensities. Loss of transketolase activity led to a reduction in levels of sugars, aromatic amino acids, products of phenylpropanoid metabolism, and chloroplast pigments in the leaf midrib. However, we are unaware of the discovery of a herbicide that produces the same symptoms. Phytotoxicity symptoms caused by incompletely knocking out a gene do not mean that the enzyme is a good herbicide target. For example, antisensing ribulose 1,5 bisphosphate carboxylase/oxygenase (RUBISCO) by <85% severely reduced plant growth (Hudson et al. 1992). However, because RUBISCO constitutes 70% of chloroplast soluble protein, even if the herbicide had a high binding affinity for RUBISCO, huge amounts of the chemical would be needed to have a significant effect. For this reason, and without the use of genomics, scientists ruled this target site out many years ago. Furthermore, in some cases, reducing gene expression of a particular gene may provide no phytotoxicity symptoms, but knocking out the gene might synergize a herbicide. For example, antisensing chloroplast ascorbate peroxidase in Arabidopsis gave a normal phenotype with a 50% reduction in enzyme activity (Tarantino et al. 2005). Nevertheless, these plants were hypersensitive to paraquat, presumably because of a reduced ability to scavenge reactive oxygen species. Constitutive overexpression or antisense expression can sometimes lead to lethality that is difficult to explain from a physiological standpoint. Viewing and understanding the effects is easier when the antisensed constructs include a selectable marker and the antisense gene is under an inducible or developmental promoter. The use of such promoters facilitates propagation of antisense plants, since the plants do not die early during their growth cycle if not induced (Mock and Grimm 1997). Another way to assay for antisense is a transactivation system. A line of plants is transformed with a chimeric transcription factor under constitutive control (Molina et al. 1999). This line is crossed with various other lines carrying the antisense gene under control of an artificial promoter that is activated only when the two lines are crossed. Using this approach with antisensed protoporphyrinogen oxidase (protox), only the hybrids had necrotic symptoms similar to those caused by herbicides that inhibit protox (Molina et al. 1999). Similar necrotic symptoms were caused by anti-sensing genes encoding other enzymes in the tetrapyrrole pathway—coproprophoryinogen oxidase (Kruse et al. 1995) and uroporphyrinogen decarboxylase (Mock and Grimm 1997). Damage was caused by only a 45% suppression of the latter enzyme, which correlated with accumulation of uroporphyrin, a photodynamic compound. The inactivation of these enzymes led to a photo-induced inhibition of two other pathway enzymes, aminolevulinate dehydratase and porphybilinogen deaminase, in the antisensed plants (Mock and Grimm 1997). These results suggest that uroporphyrinogen decarboxylase would be a good target for herbicidal discovery. It is entirely possible that phytotoxins have been discovered that inhibit one or both of these enzymes, but the activity was insufficient to follow up on determination of the target site. Or, these enzymes may not have binding sites amenable to highly effective inhibitors. Considering

226

WEEDY AND INVASIVE PLANT GENOMICS

the huge numbers and structural diversity of the inhibitors of protox (Dayan and Duke 2003), the latter possibility seems unlikely. However, protox seems to be a unique target site in terms of the level of damage caused by its inhibition. The details of the complicated mechanism of action of protox inhibitors, including massive accumulation of the toxic protox product because of enzyme/substrate cellular compartmentation and an alternative oxidase (Dayan and Duke 2003), could not be predicted by genomics. Antisense approaches to herbicide target site discovery have suggested that about 1% to 2% of Arabidopsis genes encode enzymes that are potential herbicide targets (Jun et al. 2002). But, of course, Arabidopsis may not be representative of most weeds (Chapter 3). In tobacco, 20,000 randomly selected cDNAs were antisensed, and forty-six potential herbicide target sites were found (Lein et al. 2004). At the time, just half of these genes were annotated, although with sometimes vague functions. Another quarter had very poor annotation, and the last quarter had no known function. Antisensing reduces gene transcription, whereas RNAi reduces translation. RNAi methods often suppress gene expression more than antisensing. Current efforts are underway to produce and characterize large numbers of different RNAi lines (http://www.agrikola.org/index.php?o=/ agrikola/main, Hilson et al. 2004). Transcriptional Profiling To Determine Mode Of Action And Target Sites Of Known Phytotoxins. Transcriptional profiling enables the simultaneous monitoring of thousands of genes; thus, a comprehensive or “global” transcriptional response to an exogenously supplied toxicant can be readily assessed with this technology (Shaw and Morrow 2003). Transcriptional profiling has become a mainstay for clinical drug discovery efforts for the elucidation of the mechanism of action of small molecules, where it is typically used in combination with other approaches to obtain evidence for the involvement of specific cellular targets (Gerhold et al. 2002; Bharucha and Kumar 2007). Since the first published report on transcriptional profiling, microarray analysis, appeared (Schena et al. 1995), the underlying approach has remained relatively unchanged, involving the hybridization of fluorescently-labeled “target” samples to DNA “probe” sequences affixed to a solid matrix. The target samples (either cRNA or cDNA) are derived from RNA isolated from toxicant-treated vs. “mock-treated” (co-solvent plus buffer alone) populations, and differential gene transcription is determined by comparing the relative fluorescence intensities obtained from each target against a given probe feature. Commercially-prepared arrays such as the Affymetrix GeneChip™ offer advantages such as enhanced reproducibility, streamlined protocols, and probe features designed to discriminate between closely related targets. This has led to their widespread adoption in plant biology research (Busch and Lohmann 2007). Interpreting transcriptional profiling results can be a daunting task, given the large amount of information generated from a typical experiment. With respect to elucidating the mechanism of action of a given toxicant, this task is further complicated by the fact that affected genes might represent primary responses (those directly targeted by the inhibitor), secondary responses (e.g. homeostatic processes responding to the metabolic perturbation), and potentially responses associated with defense pathways involving chemosensors that respond directly to the presence of the toxicant. Moreover, some compounds affect multiple targets, thus adding an additional layer of complexity to the data (Shaw and Morrow 2003). While the use of early time points and decreased dosages can often provide a data set enriched in genes associated with the primary response(s), low-throughput, conventional experimental approaches are still required to establish a causal link. However, plant biology resources such as the Salk Homozygote T-DNA Collection (http://methylome.salk.edu/cgi-bin/homozygotes.cgi) provide a fairly

GENOMICS AND WEEDS: A SYNTHESIS

227

straightforward means for performing follow-up experiments with sequence-characterized insertional inactivation mutants of Arabidopsis thaliana. Loss of gene expression of a cellular target should mimic the effect of an inhibitor against that target (Marton et al. 1998), thus the Salk collection provides ready-to-use seed stocks for the evaluation of candidate targets of plant growth inhibitors. While the basic methodologies for performing transcriptional profiling experiments with plants has not changed substantially over time, data interpretation has advanced significantly due to the development of sophisticated bioinformatics tools, largely patterned after those developed for the Saccharomyces, human, and mouse models (e.g. Aoki et al. 2007; van Baarlen et al. 2008). For example, up- or down-regulated genes in a data set can be organized by functional categories such as gene ontologies (Berardini et al. 2004), and the individual ontologies represented can be further analyzed for statistical significance using web-based resources such as DAVID (Dennis et al. 2003). In this manner, a large list of differentiallyregulated genes can be distilled down to a much smaller number of common biological processes, thereby greatly simplifying interpretation and hypothesis generation. A major problem with this approach is that much of the gene function annotation is “putative” by analogy, not by proof. It is too often forgotten that nature scavenges parts of one enzyme to make other enzymes with quite different functions. The sequence similarities of small portions of the different genes could lead to erroneous annotations. The recent development of tools useful for the identification of genes coexpressed with a given gene of interest is also of particular relevance to mechanism-of-action studies in plants (reviewed in Aoki et al. 2007). In contrast to model organisms such as Saccharomyces cerevisiae, Drosophila melanogaster, and Caenorhabditis elegans, a much larger percentage of the genespace of model plant species such as Arabidopsis thaliana and Oryza sativa encode proteins of unknown function, thus obscuring much of the biological significance that can be gleaned from transcriptional profiling results. Co-expression analysis makes use of the vast amount of transcriptional profiling data contained in public repositories such as GEO (http:// www.ncbi.nlm.nih.gov/geo/) and NASCArray (http://affymetrix.arabidopsis/info) to identify genes showing a similar pattern of regulation to a gene of interest under diverse conditions. Once identified, co-regulated genes whose functions are known can then be used to infer the function of the gene of interest, through “guilt by association”. Without this analysis, the primary cellular target of a plant growth inhibitor could be overlooked or de-prioritized in a data set due to a lack of information concerning its function, or alternatively, an important hypothesis may not be generated based on this lack of information. Still, the inferred functions are putative, unless experimentally verified. This brief discussion touches on merely a few aspects of the use of this technology for mechanism-of-action studies in plants. Transcriptional profiling will undoubtedly continue to serve as a powerful tool for biologists interested in determining the specific cellular targets in plants of small molecules in the foreseeable future. It can be especially useful in studies of “weediness” because it allows profiling differences between a crop and its conspecific weed, where such exist. Transcriptional profiling can assist in creating a list of candidate targets and pathways on a global scale that is unsurpassed by any other available technology, providing a foundation for the development of hypotheses concerning the putative primary cellular target or targets of a plant growth inhibitor. A robust database of transcriptome changes in plants in response to herbicides with all of the known molecular target sites when treated with doses that give similar results (e.g. the I50 dose) is not publically available, although there are a growing number of papers on herbicideinduced transcriptional changes in plants (e.g. Glombitza et al. 2004; Manfield et al. 2004;

228

WEEDY AND INVASIVE PLANT GENOMICS

Figure 14.1. Categories of the 482 genes found to be differentially expressed from three different microarray experiments in which the effects of glyphosate on both glyphosate-resistant and glyphosate-sensitive soybeans (eighteen and 464 genes, respectively) were assessed, as well as the differences between the two types of soybean without glyphosate (two genes). Twenty-seven thousand genes were assessed. (Reprinted with permission from Zhu et al. (2008), ©2008 American Chemical Society.)

Raghavan et al. 2005; Manabe et al. 2007; Zhu et al. 2008). Results with the same compound can vary considerably, depending on the herbicide dose, the method of treatment, and the time after treatment when mRNA is collected. As an example of recent study on the transcriptional response to a herbicide, the functional distribution of genes affected in soybean by glyphosate is provided in Figure 14.1 (Zhu et al. 2008). Since comparisons can be made between glyphosate-resistant and glyphosatesensitive near isogenic lines, this work was conducted on soybean. No significant transcriptomic changes were caused by glyphosate in glyphosate-resistant soybean. Overexpression To Elucidate Target Sites Of Known Phytotoxins. Increased expression of genes encoding herbicide targets usually leads to a low level of resistance. Cell culture lines with increased expression of EPSPS (Shyr et al. 1992; Steinrücken et al. 1986), glutamine synthase (Donn et al. 1984), and acetolactate synthase (Tourneur et al. 1993; Xiao et al. 1987) are resistant to the herbicides that bind to these enzymes. Therefore, putative target site overproducing mutants or transformants could assist in target site elucidation (Höfgen et al. 1994). Herbicide resistance could also result from enhanced levels of enzymes involved in metabolic degradation, uptake, or sequestration of the herbicide, which tells us nothing about the herbicide’s target site. Furthermore, for herbicides or phytotoxins that act through production of toxic compounds or metabolic intermediates (e.g. protox inhibitors, paraquat, ceremide synthase inhibitors), elevated levels of enzymes that detoxify these chemical species would provide resistance. Lastly, elevated levels of an enzyme that would result in metabolic compensation for the disruption caused by the herbicide might provide a low level of resistance. A T-DNA activation system (Hayashi et al. 1992) was used with Agrobacterium transformation of tobacco protoplasts to generate a large number of lines with overexpressed genes by using a strong enhancer derived from the CaMV 35S promoter. The cells were plated on selective media to choose resistant lines (Höfgen et al. 1994). Tagging allows the DNA conferring resistance to be isolated and cloned, so that it can be validated as the gene involved in resistance (the target gene, detoxification gene, or other gene conferring resistance). The concept was validated by selecting such cell cultures with glyphosate. The resistant calli had elevated levels of the EPSPS mRNA (by northern blots) (Höfgen et al. 1994). Libraries with amplified,

GENOMICS AND WEEDS: A SYNTHESIS

229

tagged cells from different transformation events can be generated by having an antibioticresistant selectable marker as part of the construct. The libraries can be used later with many different chemicals with unknown target sites. Höfgen (1999) reviews this method in detail. Insertional Gene “Knock-Out” Mutagenesis. Knocking out genes with tagged mutations is one approach to determine the functions of genes. This technique can randomly tag any gene in a fully sequenced genome, and then ascertain whether suppressing gene function results in a dead plant, the desired outcome of a herbicide. This is done by randomly inserting a known sequence of DNA into the genome of a plant in multitudes of separate “events.” The DNA tag can be inserted by either of two methods: •

•

Random Agrobacterium transformation or biolistic transformation. In this method, T-DNA (transfer DNA used for transformation) contains a selectable marker such as herbicide and antibiotic resistance and the “border” region, i.e. the bridge between the selectable marker and the DNA of the transformed region. The method has had limited usage because the tagged gene can be located using the sequence of the border region in only about half the mutant phenotypes that carry the selectable marker. The analysis of others is complicated due to multi-insertional events or loss of the border sequence. Transposon tagging. In this case there is a single transformation event in which the T-DNA is attached to a mobile transposable element under the control of a high expression promoter. This “jumping gene” keeps reinserting itself throughout the genome of different individuals, and the border region of the tag can be used, as above, to isolate the mutated gene.

With both of these procedures, the transformed plants are selfed, and if there are lethal or near lethal phenotypes segregating in the T2 (equivalent of F2, but from a transformational event) individuals among the progeny, a knockout or suppression can be assumed. The insertion should have been into a structural gene if it is lethal. If the phenotype is near lethal, then the insertion was probably in a regulatory element. Because the inserted “tagged” sequence is known, it can be recognized by a complimentary PCR primer and by the DNA sequences generated that should extend a bit beyond the tag. After further PCR sequencing, the gene can be deduced in species for which the complete genome sequence is known. This ability to easily find the gene of interest makes this procedure far superior to radiation or chemical mutagenesis, with which isolation of the mutated gene is vastly more complicated (e.g., Vizir et al. 1994). The methods are especially useful with fully sequenced genomes, because one does not have to sequence far beyond the border regions to know both the position and the full sequence of the adjacent DNA. Nevertheless, they are also useful when the genes are not fully sequenced. However, this requires more effort to determine the whole gene sequence and then determine function. Synteny, the similarity of sequence of genes along chromosomes of related species, assists in the physical mapping as well, where it occurs. The transposon tagging system is made up of two elements: the transposon that inserts the T-DNA into various places, and the transferase that excises the mutational element allowing the gene to remain in place. The most commonly used transposition system, the Ac/Ds (activator/dissociation) system from maize, has been introduced in Arabidopsis (Tissier et al. 1999). Arabidopsis was the organism of choice for this concept because it has a small genome that theoretically allows a rapid saturation of the genome. Arabidopsis can be cultivated on a rich medium to save some auxotrophic knockouts from death. Transposon tagging is a good choice for elucidating gene function of unknown sequences, especially in Arabidopsis.

230

WEEDY AND INVASIVE PLANT GENOMICS

Transposon systems are especially useful when transformed into species other than those in which they originated, i.e. heterologous species. The Ac/Ds system from maize is used in both dicots and monocots. It allows stable insertion, using both random insertion (for saturation) and ESTs (expressed sequence tags), to hone in on specific genes to determine what happens when the transposon jumps and causes its neighbor to suffer insertional disruption. See reviews by Federoff (2000) for background and mechanisms and Koprek et al. (2000) for applications. About 75% of Ac/Ds transpositions are sites near the original insertion. Antisense vs. Knockout. While easier to perform and less specific than the gene-directed antisense approach, knockout has limitations. If a plant has two identical or nearly identical genes making isozymes of the same enzyme, one might assume that both would be inhibited by a particular herbicide. In a random knockout, only one copy will be eliminated, allowing the plant to survive with the isozyme. Thus, this redundancy can be problematic if the herbicide can bind to identical products from more than one site. If the cDNA sequence chosen for antisense is compatible with all the similar genes encoding the isozymes, then there is a likelihood that all will be suppressed. Many evolutionarily related genes that have dissimilar functions have consensus sequences among them, and these sequences must be avoided for antisensing. The calibrational use of antisense and knock-out techniques with known herbicide targets yields inhibited or dead plants (Haake et al. 1998; Höfgen et al. 1994, 1995). A recent example from the patent literature in which this approach was used is the elucidation of ferredoxin:NADP reductase as a target through antisense technology (Wagner 2000). In this case, the target site itself was not claimed because it had been elucidated earlier. After identification of the target, an assay was developed to discover a class of compounds that inhibit this enzyme, killing plants. Many enzymes are not suitable targets for herbicide action because of the redundancy within and among gene families, as well as alternative pathways for some processes. Futhermore, inhibition of a primary non-redundant pathway that has no alternative routes does not guarantee lethality. Most herbicides rarely completely inhibit a pathway, and plants can survive with reduced carbon flow through most metabolic pathways. For example, many plants can remain alive for days, weeks, or longer under very low light (and thus very low photosynthesis). However, inhibitors of photosynthesis rapidly kill the same plants in the light largely because of photodynamic damage, even though these inhibitors do not completely block photosynthesis. Highly effective herbicides often cause the accumulation of toxic products or weaken the weeds such that other factors kill plants. Prediction of such effects, based on the place of an inhibitor in a biosynthetic pathway or other function, is difficult. Understanding why inhibitors of particular enzymes in a biosynthetic pathway are much better herbicides than inhibitors of other enzymes in the same pathway has often been elusive. Therefore, the results of incomplete suppression of gene expression achieved by antisense or overexpression cosuppression have an advantage over the knock out or deletion of genes for elucidating potential targets. Thus, if a potential target is elucidated by knockout, it might be best to ascertain if partial suppression by antisense will also validate the target.

Transgenic Crops And Weeds

Herbicide-resistant Crops. The most highly adopted transgenic crops are those with glyphosate resistance (Duke and Powles 2008). About 90% of the transgenic crop area contains herbicide-resistant varieties (James 2008). Glyphosate-resistant crops (primarily canola, cotton,

GENOMICS AND WEEDS: A SYNTHESIS

231

maize, and soybean) make up the majority of these (Dill et al. 2008). All of these glyphosateresistant crops have a glyphosate-resistant EPSPS from Agrobacterium sp. except for a few varieties of maize with an altered maize EPSPS produced by site-directed mutagenesis. New genes for glyphosate-resistance are being introduced, such as an EPSPS from a soil microbe with superior kinetic properties (Vande Berg et al. 2008) and a gene for glyphosate acetyltransferase (Castle et al. 2004; Green et al. 2008). Glufosinate-resistant crops are being grown, though considerably less than glyphosate-resistant crops (Duke and Cerdeira 2005a). Transgenic crops with multiple herbicide resistances (e.g. glyphosate resistance stacked with dicamba resistance or with resistance to acetolactate synthase inhibitors) are under development. The potential agricultural and environmental problems resulting from movement of these transgenes into weeds have been of considerable concern to agriculturalists and environmentalists (Gressel 2002; Duke and Cerdeira 2005b). Herbicide resistance transgenes are perhaps the worst-case scenario for gene flow to related weeds, if uncontained or unmitigated (Gressel 2008). The survival of the usually unfit F1 generation is facilitated by the herbicide, assisting in eventual full introgression into a wild relative of the crop. Stopping Transgene Flow. Those who oppose transgenic crops warn about the dire effects of transgene flow to wild relatives. They posit that natural ecosytems will be forever modified. Transgene flow might occur in some cases, and there is the possibility for gene flow from many crops to non-crop species, some of which occur in natural habitats (Mallory-Smith and Zapiola 2008). But, with the transgene/crop products currently on the market, most of which have been grown over broad areas for more than a decade, this has not been a problem because they do not have related weedy or wild species growing in proximity, except for canola. With herbicide-resistant canola (Brassica napus), there are apparently no large populations of weeds with a fully introgressed herbicide resistance transgene. However, in Canada, where most of herbicide-resistant canola is grown, a case of introgression of glyphosate resistance has been documented from canola to Brassica rapa (Warwick et al. 2008). The gene appeared to persist in a small proportion of the weed population in the absence of the herbicide for at least three years. Gene flow of transgenes conferring herbicide resistance to a wild relative is unlikely to confer any advantage in the wild, where no herbicides are used (Gressel 2008; Cerdeira and Duke 2006). In the absence of the herbicide, herbicide resistance transgenes are more likely to carry minor unfitness, and would probably disappear from a wild population. However, when stacked with a gene or genes that confer a trait with a clear fitness advantage in the wild (e.g. drought, insect, or pathogen resistance), herbicide resistance could enhance the probability of introgression of transgenes to wild species. In most cases, the herbicide would be especially useful to survival of the F1 hybrid, because it is often unfit, allowing it to backcross with the wild population until the gene is introgressed (Warwick et al. 2008). The next generation of transgenic crops is likely to have a variety of combinations of fitness-enhancing transgenes stacked with transgenes for herbicide resistance. The issue of gene flow to weedy relatives of the crop in which they exist has been one that presents a bigger problem to agriculture than to the environment. Maize and soybeans do not have this problem, but most other major crops do. Rice is the most problematic in this context. The world is rapidly going from hand transplanted rice to direct seeded rice (Valverde 2005). The worst, most intractable weed in direct seeded rice is a feral, weedy form of rice itself that is genetically compatible with it (Vaughan et al. 2005). Likewise sorghum has its feral form (shattercane) as well as an interbreeding relative, Sorghum halepense (Ejeta and Grenier 2005). In some wheat growing areas, a close interbreeding relative, Aegilops cylindrica, is a major

232

WEEDY AND INVASIVE PLANT GENOMICS

weed, and feral sunflowers plague sunflowers, even where wild sunflowers do not exist (Berville et al. 2005). Oilseed rape (Brassica napus) is genetically compatible with its weedy cousin Brassica rapa. Transgenic crops with herbicide resistance would be ideal to control these related weeds—for a few generations—until the transgene becomes introgressed into the weed. Two approaches have been proposed to stop transgene flow: containment (keep the transgene in the crop) and mitigation (keep the transgene from establishing and spreading if it crosses into the weed) (Gressel 2008). Containment requires knowing crop genomics, whereas mitigation requires knowing weed genomics. Containers of all types eventually spring leaks, and thus we might expect leaks when attempting transgene containment. The topic of approaches to prevention of gene flow is discussed in detail by Gressel (2002, 2008). Only a few of the methods are described here. The most commonly proposed transgene containment mechanism is to insert the transgene of choice (e.g. herbicide resistance) in the plastid (chloroplast) genome (the plastome), because chloroplasts are maternally inherited and are seldom carried by pollen. Two issues are typically ignored: plastid genes typically are conveyed to progeny in about 0.4% of cases in which there is maternal inheritance (Wang et al. 2004). Even if it were much lower than this, over vast areas, a tiny percentage still represents many chances for gene flow. Again, once escape occurs, it may not be possible to recall the gene. Furthermore, the proponents of plastome-encoded transgenes forget or ignore (Daniell 2002; Maliga 2004) that the same resistant F1 progeny can be obtained when the feral/weedy relative is the pollen parent. With dwarf grains, and tall weeds such as Green Revolution rice and feral weedy rice, gravity effectively constrains most crop × weed progeny to have the taller weed as the pollen parent. The weed is also more likely to be the pollen parent in recurrent backcrosses, and the plastid-inherited trait will end up in the weed. Genetic use restriction technology (GURTs) can be used to stop gene flow as well as to prevent use by people not paying for the technology. This so-called “terminator” technology to make the next generation sterile has neither been developed nor commercialized, apparently because of the bad publicity over the fact that farmers cannot save seed for planting the next year, as has been the case with hybrid seed for many years. Ever since the original patent was issued (Oliver et al. 1998), there have been no publications to provide information on efficacy of this approach. However, GURTs combined with a mitigation strategy could provide a failsafe technology to prevent gene flow. The simplest mitigation strategy proposes that the transgene of choice be tandemly linked with transgenes that are useful or neutral to the crop but would confer extreme unfitness on the weed or to a wild plant (Gressel 1999). The herbicide resistance gene could be flanked with genes causing dwarfing (increasing harvest index for the crop but non-competitive in the weed), non-shattering (all crop harvested, little seed bank replenishment for weed), uniform complete germination (an important crop trait, not advantageous for a weed), etc. Finding some of the genes is easy; e.g. kaurene oxidase, or other key genes in the gibberellic acid biosynthesis pathway, are highly conserved, and RNAi or antisensing causes dwarfing (Tong et al. 2007). This is not the case with other mitigation traits. For example the shattering mechanism in Brassica spp. (silique opening) (Østergaard et al. 2006) is very different from those of grains. It is clear that very different genes confer shattering in wheat, which are different from sorghum, which are different from rice (Konishi et al. 2005; Li and Gill 2006). It is the weed that shatters (dominant) and has secondary dormancy (usually dominant) or flowers prematurely (bolts) (as with conspecific sugar beet and weedy beets). Thus, it is in the weed genome that one must find the gene that can be used for mitigation (as antisense or RNAi), not in the

GENOMICS AND WEEDS: A SYNTHESIS

233

crop. The mitigation gene needed could vary between geographic regions, depending on the compatible weed in a particular area. Still, if the crop has been sequenced, with comparative sequencing using some of the newest robots that rapidly sequence short pieces of DNA or cDNA (Vega-Sanchez et al. 2007), one can rapidly find the genes that are specific to, or turned on, in the conspecific weed. Thus, again, if we understand the enemy we can deal with the enemy (Basu et al. 2004).

Weed/Crop Interference

Competition. After determining which traits provide the competitive advantage in weed/crop competition, the genetics behind these traits must be determined. Clearly, this is not a trivial matter, because these traits vary among weed species, are affected by many environmental factors, and are usually controlled by multiple genes (Basu et al. 2004). RNAi technology should be very helpful in determining which genes control what competitive traits. Other parts of this book have dealt with this topic in detail, so we only point out that it may be exceedingly difficult to translate genomic information on competition into better weed management approaches. But, technological breakthroughs generally emanate from fundamental research. Allelopathy. The second component of interference, allelopathy, is perhaps more amenable to the use of genomics for understanding and manipulating the process. Allelopathy has been very difficult to demonstrate in field situations, and as a result, has been questioned as a significant phenomenon (discussed by Dayan and Duke 2009). Just as genomics has been a powerful tool in proving the role of other phytochemicals in plant/insect interactions (Steppuhn 2004), this technology can be critical in determining the significance of allelopathy in plant/ plant interactions. Identification Of Genes Involved In Allelochemical Biosynthetic Pathways Using Functional Genomics. Genes encoding for four biosynthetic pathways leading to allelochemicals or suspected allelochemicals have been characterized: the triterpene glycoside avenacin A-1 (Qi et al. 2004), the benzoxazinone DIMBOA (Frey et al. 1997), labdane-related diterpenoids momilactone A and B (Shimura et al. 2007; Peters et al. 2006), and the benzoquinone sorgoleone (Pan et al. 2007, Baerson et al. 2008). Gene cloning efforts for avenacin A-1 employed classical forward-genetic approaches; hence, this discussion focuses on the remaining three pathways in which genomics played a critical role in gene identification. The general genomics-based strategies employed by allelopathy investigators have involved BLAST searches using the full-genome sequence for rice, EST analyses using maize and sorghum data sets, and the use of transposon-tagged maize populations. The availability of a full-genome sequence offers the possibility of performing comprehensive searches for a specific enzyme class by using a related sequence from a different species, and the identification of genes involved in momilactone biosynthesis have relied heavily on the available rice genome sequence for this purpose. For example, Sakamoto et al. (2004) used diterpene synthase sequences associated with gibberellin biosynthesis as queries against the rice genome sequence and identified four putative copalyl pyrophosphate synthase (CPS) and nine kaurene synthase (KS) genes. They further demonstrated that a subset of these sequences were transcriptionally induced by UV light as well as fungal elicitors, suggesting their involvement in the biosynthesis of diterpene phytoalexins such as momilactones A and B. A similar

234

WEEDY AND INVASIVE PLANT GENOMICS

approach was employed using the CPS and KS genes from Arabidopsis thaliana as queries for BLAST searches. Subsequent recombinant enzyme studies led to the identification of OsCPS4, that produces the syn stereoisomer of copalyl pyrophosphate, and OsKSL4, that specifically uses syn-copalyl pyrophosphate to produce syn-pimaradiene (Wilderman et al. 2004; Xu et al. 2004). OsKSL4 (also known as OsKS4) was also characterized by Otomo et al. (2004), by following up on the kaurene synthase-like sequences identified by Sakamoto et al. (2004). To identify the remaining genes involved in converting syn-pimaradiene to momilactone, the rice genome was used to identify cytochrome P450 and dehydrogenase genes closely linked to OsCPS4 and OsKSL4, based on the speculation that these genes could comprise a biosynthetic cluster located on chromosome 4 (Shimura et al. 2007). Subsequent functional studies confirmed this. Thus, all the genes required for the biosynthesis of momilactone A from geranylgeranyl pyrophosphate have now been isolated, and the stage is set for manipulation of this pathway to determine the actual role of momilactones in rice allelopathy. EST analysis involves the random sequencing of individual clones from a cDNA library. This is particularly useful when working with plant species in which genome sequences are unavailable. EST analysis is most effective when the cDNA library is generated from the tissue or cell type most actively involved in synthesizing the compound of interest, thus increasing the probability that the transcripts encoding the corresponding biosynthetic enzymes are highly represented in the resulting data set. The DIMBOA biosynthesis pathway was elucidated by using ESTs for identifying both BX6 and BX7, which catalyze the final two steps leading to the formation of DIMBOA-glucoside. BX6 encodes a 2-oxoglutarate-dependent dioxygenase (2ODD) identified by BLAST queries using known plant 2ODD sequences against a proprietary EST data set derived from maize seedling tissue libraries (Frey et al. 2003; Jonczyk et al. 2008). BX7 encodes an Omethyltransferase, which was first purified to homogeneity, then peptide sequences derived from trypsin digestion fragments were used as BLAST queries against the same EST data set used to identify BX6 (Jonczyk et al. 2008). Genes encoding enzymes for all of the remaining biosynthesis steps required for DIMBOA-glucoside production (BX1, BX2, BX3, BX4, BX5, BX8, BX9) have also been isolated using standard molecular genetics approaches, thus completing the benzoxazinone biosynthesis pathway in maize (reviewed in Jonczyk et al. 2008). Mutator (Mu) transposon-tagged populations in maize have also been used by Frey et al. (2003) to confirm the function of the above-mentioned BX6 2ODD gene in planta. Four lines containing different integrations of Mu within the BX6 coding sequence were identified by PCR screening of multiplexed DNA pools using BX6- and Mu-specific PCR primer sets (described by Brutnell 2002). The co-segregation of the Mu-disrupted BX6 loci with loss of DIMBOA production provided further compelling evidence in this study for the involvement of BX6 in the benzoxazinone biosynthesis pathway (Frey et al. 2003). EST analysis has also played a principal role in efforts to identify genes encoding enzymes involved in the biosynthesis of the allelopathic benzoquinone sorgoleone. Previous studies suggested that biosynthesis exclusively occurs in root hair cells of Sorghum spp., which produce and secrete copious amounts of sorgoleone into the surrounding rhizosphere (Czarnota et al. 2003). An EST data set was therefore generated from purified root hair cells, and BLAST queries were performed using known plant desaturase, polyketide synthase, O-methyltransferase, and cytochrome P450s in an effort to identify sequences encoding the corresponding biosynthetic enzymes (Baerson et al. 2008). Two fatty acid desaturase-like sequences identified within the data set, designated SbDES2 and SbDES3, were heterologously expressed in Saccharomyces and determined to encode

235

GENOMICS AND WEEDS: A SYNTHESIS

45000

2500 SbOMT1 SbOMT2 SbOMT3

40000 35000

2000

30000 1500

25000 20000

1000

15000 10000

500

5000 0

Relative Expression (SbOMT2)

Relative Expression (SbOMT1, SbOMT3)

A

0

6.0

Cab 2_9279 Ubi 2_9314

4000 3500

5.0

1500

2.0

1000 500

1.0

0

0.0

M at ur e

Pa ni c

Ro ot H

St em Im m at ur e Le af M at ur e Le af Sh oo tA pe x

3.0

le

4.0

2500 2000

Ro ot

3000

ai r

Relative Expression (Cab)

4500

Relative Expression (Ubi)

B

Figure 14.2. Identification of root hair-specific O-methyl transferase (OMT) sequences from sorghum by quantitative real time RT-PCR analysis. Relative expression levels were determined by quantitative real time RT-PCR using gene-specific primers. Data were normalized to an internal control (18 S rRNA), and the ∆∆CT method was used to obtain the relative expression levels for each sequence, expressed as mean ± SD. (A) Relative expression levels of three OMT genes (SbOMT1, SbOMT2, and SbOMT3) in different S. bicolor tissues. For clarity, a different y axis scale was used for displaying SbOMT2 values (right side of graph). (B) Relative expression levels of two S. bicolor control genes: chlorophyll a/b-binding protein (CAB 2_9279) and a polyubiquitin-like sequence (UBI 2_9314). For clarity, different y axis scales were used for displaying CAB 2_9279 and UBI 2_9314 values. (From Baerson et al. [2008].)

enzymes capable of converting palmitoleic acid (16:1∆9) to hexadecadienoic acid (16:2∆9,12), and hexadecadienoic acid into hexadecatrienoic acid (16:3∆9,12,15), respectively (Pan et al. 2007). Additionally, SbOMT3 has been identified as an O-methyltransferase likely involved in catalyzing the 3-O-methylation of the 5-pentadecatrienyl resorcinol sorgoleone pathway intermediate in planta (Baerson et al. 2008)(Figure 14.2). The two alkylresorcinol synthases involved in the formation of 5-pentadecatrienyl resorcinol utilizing hexadecatrienoic acid as a substrate have also been recently identified (manuscript in preparation). The remaining gene characterization work required for completely elucidating the biosynthesis pathway leading to sorgoleone should be completed in the near future. Examining Global Transcriptomic Responses To Allelochemical Exposure In Plants. Transcriptional profiling with DNA microarrays enables the simultaneous monitoring of

236

WEEDY AND INVASIVE PLANT GENOMICS

Metabolism Cell rescue, defense and virulence Unclassified proteins Cellular communication/signal tranduction mechanism Cellular transport, transport facilitation and transport routes Protein activity regulation Transcription Biogenesis of cellular components Protein fate Energy Cell cycle and DNA processing Storage protein 0

5

10

15

20

25

30

Figure 14.3. Distribution of 158 BOA-induced genes in Arabidopsis into functional categories. (From Baerson et al. [2005].)

thousands of genes, and involves the hybridization of fluorescently labeled “target” samples to DNA “probe” sequences affixed to a solid matrix (the microarray “chip”). A genome-wide transcriptional response of a plant to an exogenously supplied allelochemical can be comprehensively assessed using this technology (Duke et al. 2008). The use of transcriptional profiling in allelopathy research has been quite limited to date, with few publications appearing in the literature as of this writing. The response of Arabidopsis to the allelochemcial benzoxazolin-2(3H)-one (BOA) was elucidated using Affymetrix ATH1 GeneChip arrays (http://www.affymetrix.com), which monitor approximately 80% of the genes predicted to exist in the Arabidopsis genome (Baerson et al. 2005). Metabolism, cell rescue, and defense genes were highly represented among functional categories of genes affected (Figure 14.3). One of the major responses observed corresponded to gene families potentially associated with chemical detoxification pathways. Further experiments using a subset of these up regulated genes in real-time RT-PCR assays showed that many were also transcriptionally induced by a variety of structurally diverse xenobiotic compounds. This suggests that plants, as is the case for other organisms, rely on a panel of dedicated and coordinately regulated genes associated with a broad-spectrum xenobiotic defense response. More recently, the Affymetrix ATH1 GeneChips were used to assess the effects of fagomine, gallic acid, and rutin, allelochemicals produced by buckwheat (Fagopyrum esculentum), on twenty-day-old Arabidopsis plants (Golisz et al. 2008). As was the case for (–)-catechin, the major conclusion drawn from this study was that the compounds elicit a transcriptional response suggestive of oxidative stress, although this hypothesis was not directly tested (Golisz et al. 2008). It will be of significant interest to compare the transcriptome responses to a large number of allelochemicals once this information becomes available from future studies, as well as to compare these responses to those of other xenobiotics such as herbicides and environmental pollutants. Evolution Of Herbicide Resistance

Herbicide resistance has become a huge issue in weed science because of the rapid evolution of weeds resistant to herbicides since the early 1970s, and the introduction of transgenic,

GENOMICS AND WEEDS: A SYNTHESIS

237

herbicide-resistant crops in 1996. Genomics have been underutilized to understand and deal with the evolved weed resistance to herbicides, whereas it has been crucial in generating herbicide-resistant crops. Predicting The Propensity For Evolution Of Herbicide Resistance Via Genomics. Estimating the potential for evolution of resistance to a new herbicide should be an important factor in herbicide development. However, predicting this critical interaction between a herbicide and its many weedy target species containing huge genetic diversity is not trivial. Resistance mechanisms can be associated with target site, metabolic degradation, uptake, translocation, or sequestration, or any combination of these mechanisms. Genomic information can provide some clues regarding the probability of site of action mutations contributing to resistance. For example, if there is more than one target-mechanism gene, each encoding an enzyme that is equally inhibited by the herbicide, the likelihood of simultaneous evolution of more than one target site resistance is much lower than for a single gene-encoding enzyme. If a single gene copy is found for the target site, and sequencing many viable mutants shows that the gene has considerable plasticity, one might infer that target sitebased resistance will rapidly evolve. Most herbicides target many weed species, and the traits above might not be uniform across species. Indeed, some species seem to be predisposed to evolve resistance to several herbicide classes (e.g. Lolium spp. Conyza spp.), whereas others are not. There is very little genomic information available to explain these differences. There can also be false conjectures. When penicillin was discovered, target site mutants were quickly discovered in the lab, engendering fear of rapid evolution of resistance. Penicillin resistance did evolve, but in human and animal patients it was always a plasmid-based, nontarget site resistance. Target site resistance to penicillin remains a curiosity in a Petri dish. Similarly, insect resistance to Bt toxin is easy to generate in the laboratory, but has been slow to evolve in the field, despite predictions to the contrary (Gressel 2005). The herbicide dose used with a target species influences evolution of both the rate and mechanism of resistance. For example, low application rates of diclofop-methyl selected for multiple factor type resistance inherited in a polygenic type fashion, whereas high application rates of the same herbicide selected for target site resistance (discussed in Gressel 2002). Genomic information can be useful in understanding the details of a scenario like this, but is not yet sophisticated enough to make detailed predictions. Nevertheless, genomics may contribute to predicting the type of resistance that may evolve. If a species has a plethora of cytochrome P450 genes and no glutathione transferases, one might guess that a herbicide could be catabolized by a cytochrome P450. A good chemist who looks at the molecule can tell whether it could be P450 hydroxylated or must be conjugated or hydrolyzed. Simple experiments could just as easily imply the same. In summary, at this point, genomics has not been a highly useful tool in predicting the speed or mechanism of evolved resistance to a herbicide. Possible Uses Of Weed Genomics-based RNAi In Viruses Overcoming Herbicide Resistance. Most of the herbicide resistances that have evolved are either the result of a mutation in the protein target of the herbicide or an up regulated expression of a gene encoding an enzyme that degrades the herbicide. In many cases the resistance is heterozygous. A virus bearing an RNAi that matches the sequence of the mutational site in the target of the herbicide, or that matches the overexpressed gene for the catabolic enzyme that degrades the herbicide, could suppress herbicide resistance. In the first case, the RNAi would be lethal to the weed with the altered target site, and in the latter the RNAi would make the weed more sensitive to the herbicide.

238

WEEDY AND INVASIVE PLANT GENOMICS

Overcoming Natural Or Evolved “Phoenix” Resistance. Many perennial and other weeds have underground structures or surface rosettes with dormant buds. Contact herbicides often burn off the leaves, and the weed arises like the proverbial phoenix from the ashes from the dormant buds. A good supply of storage proteins and other nutrients are the mainstay for the ability to respond like a phoenix. A pretreatment with a virus containing RNAi for the weedspecific vegetative storage protein or the starch synthase, etc. in these tissues, might render the plant irreversibly dead after treatment with a contact herbicide, overcoming the phoenix phenomenon. This, of course, requires knowing the biochemistry of the storage compounds as well as knowing the genomics to find the requisite genes. Systemic herbicides such as glyphosate were highly effective against weeds that responded like a phoenix to contact herbicides, because these herbicides translocate to the hidden buds. Recently, a spate of such weeds have evolved resistance to glyphosate, e.g. the rosette Conyza spp. (Cerdeira and Duke 2006; Dinelli et al. 2006; Zelaya et al. 2004) and the rhizomatous Sorghum halepense (Cerdeira et al. 2007; Valverde 2007; Vila-Aiub et al. 2007, 2008), with phoenix-like symptomology. Glyphosate is no longer systemic in these biotypes, probably because a transporter has mutated or its expression has been altered. Only with physiological and genomics knowledge of what happened in these biotypes will it be possible to design RNAi or other systems to overcome this newly emerging phoenix phenomenon. Is Weed “Epigenomics” Involved In Resistance? Scientists are often prone to avoiding issues that defy explanation by common wisdom. Farmers complained that glyphosate was not controlling Amaranthus sp. despite retreatments at very high doses. Offspring from the seeds were not resistant, but cuttings from the plants were, continuously for generations of cuttings. This did not meet the academic definition of resistance, which requires inheritance of the trait through meiosis, but this definition was not helpful to the farmer who could not control weeds in his fields. Recurrent selection of seed from offspring did incrementally increase resistance and decrease variability, but the inheritance was not typically Mendelian (Zelaya and Owen 2005). These authors preferred an explanation based on polygenic inheritance, but the results are not typically polygenic either. Another possible explanation could be that the phenomenon is from epigenetic imprinting, a methylation or demethylation of a critical gene or genes modulating their expression, and conferring resistance. There is evidence that some stress-induced epigenetic effects can be “remembered” through a few generations of meiosis (Bruce et al. 2007). Thus, we may need “epigenomics,” the study of epigenetic controls of gene expression measured by expression profiling of mRNA or cDNA or DNA chips of the weed in question, but we cannot do that without genomics first. Epigenetic effects might be reversed again through viruses carrying specific plant DNA sequences that will cause specific methylations or demethylations.

Parasitic Weeds

Root attaching Orobanche and Striga species are scourges of tropical agriculture, and there are no selective herbicides for their control. The situation is no better for stem-attaching Cuscuta spp. that are more widespread but cause less devastation. Genomics have so far been of assistance in five ways to deal with these weeds: 1.

Marker assisted breeding. This has been rather effective in species that co-evolved with the Striga spp. parasites such as sorghum and cowpeas that each have different strains

GENOMICS AND WEEDS: A SYNTHESIS

239

with a low modicum of resistance. Marker-assisted breeding has allowed intelligent combining of the strains to rapidly increase the level of resistance in each (Timko et al. 2007; Grenier et al. 2007). As the markers get closer and closer to the actual genes, the genes might then be isolated and could be transferred to crops such as maize that did not evolve where parasitic weeds occurred, and have little inherent resistance. The sequences of the sorghum and the Medicago truncatula (related to cowpea) genomes may well contribute to a more rapid isolation of these partial resistance genes, both for these species (maybe overexpression will enhance resistance) as well as for transfer to other crops. 2. Herbicide-resistant crops. An inferred genomics approach was used in this case. It was proposed that if a crop has transgenic target site resistance to a systemic herbicide, the herbicide might kill the parasite underground just after attachment (Gressel 1992). But which herbicide would be best? Before PCR it would be hard to ascertain if the weeds had the requisite target sites. The best target site resistances to systemic herbicides were those inhibiting the pathways producing aromatic amino acids (glyphosate) or branched chain amino acids (ALS inhibitors). Would they work? The textbooks said that these parasites stole their organic nitrogen from the host (Press 1995). It was intuited that this view was not necessarily correct because one lab was growing parasitic weeds in in vitro culture, without amino acids in the medium (Wolf and Timko 1991). Thus, glyphosate-, ALS inhibitor-, and asulam-resistant (modified reductase) transgenic crops were obtained, and the parasitic weeds were controlled by foliar spray of the requisite herbicides on the herbicide-resistant crops (Joel et al. 1995). This was extended to other crops transgenically (Surov et al. 1998; Aviv et al. 2002). A mutant maize resistant to ALS-inhibiting herbicides has been bred for African traits and has been commercialized on a large scale. In this case, instead of spraying the crop, the herbicide is applied to the maize seed prior to planting (Kanampiu et al. 2003, 2007). This precludes the need for sprayers and allows the use of ten-fold less herbicide than would be sprayed, with the added advantage that it allows a legume to be intercropped without being harmed. 3. RNAi transmission. As an alternative to herbicides, perhaps the crop can kill the parasite by making an RNAi specific to parasite gene(s). This approach has been proposed by various groups (de Framond et al. 2007; Yoder et al. 2007) for parasitic weeds. MicroRNAs that are cleaved from RNAi pass graft junctions (Brosnan et al. 2007; Kalantidis et al. 2008), so they should get into the parasite. Crops were partially protected from nematodes (Fairbairn et al. 2007; Huang et al. 2006) and insects (Mao et al. 2007) by using RNAi sequences specific to these pests. So far, only negative data have been published for Striga; the RNAis used in the experiments suppress the same enzymes that are targeted by herbicides (de Framond et al. 2007), but the Striga remained undaunted. We hope that researchers will be able to demonstrate that this elegant approach works by selecting the right enzyme target. One group proposes to use RNAi targeted to genes involved in the formation of the haustorium of parasites that are responsible for the parasite attachment to the roots (Yoder et al. 2007). Such RNAi in crops would not kill weeds other than the parasite; a graft junction is needed to introduce the microRNA. Thus, the approach would have no “off-target” effects, or soil residues, unlike herbicides. One thing is clear—parasitic weed genomics compared with crop genomics are required to make this system work. Another approach is to introduce parasite (or other) weed-specific RNAi’s into the parasite via weed-specific viruses, bacteria, or fungi, as discussed in the following section. 4. Biocontrol. The control of weeds with weed-specific virus, bacterial, or fungal pathogens, as well as by insects, has sometimes been highly effective in non-intensive agricul-

240

5.

WEEDY AND INVASIVE PLANT GENOMICS

ture, especially against introduced weeds. This has been most effective when a biocontrol agent is brought from the area of origin of the weed (classical biocontrol), in cases in which the weed thrived because it immigrated without this pathogenic or insect herbivore baggage. Classical biocontrol has been far less successful with well-established global weeds. When a biocontrol agent is found that works in the lab to control major global weeds, high levels of inoculum are typically needed (inundative biocontrol). From an evolutionary viewpoint this is obvious; if a specific agent was highly virulent, both it and the weed might become extinct. Some success has been achieved (in the laboratory) by moving fungal toxin genes from one fungus to another; e.g. the NEP1 gene from a Fusarium to a Colletotrichum specific for Abutilon (Amsellem et al. 2002) after splicing a high expression promoter to it. Because the genomics of Fusarium species were not known, another effort to use the gene led to mixed successes. When the same construct was engineered into a Fusarium oxysporum forma speciales specific for the parasitic weed Orobanche (Amsellem et al. 2001), the fungus did not have enhanced virulence (Bailey et al. 2000), despite PCR and marker genes showing that the gene had been inserted (Meir et al. 2009). PCR of the non-transformed Fusarium oxysporum showed the presence of the gene as well. Subsequent analyses of all other forma speciales of Fusarium oxysporum tested had the gene, with expression at very low levels. Inserting a gene already present often leads to overexpressive cosuppression. The NEP1 gene construct did enhance the virulence of a Fusarium sp. (once classified as F. arthrosporiordes [Amsellem et al. 2001] but now known to be a new species) (John Leslie, personal communication), and the wild type did not possess the gene (Meir et al. 2009). So far, engineered RNAi has not been tested as a possible virulence-enhancing factor for biocontrol agents. RNAi clearly should be effective with weed-specific viral pathogens, because virus nucleic acids are replicated by the plant, not the virus. Even non-specific viruses could be used on condition that they be “disarmed” (mutated such that they have, by themselves, no effect on the crop) and that they contain an RNAi specific to the weed. Knowledge of weed genomics is needed to prepare constructs for this approach. Various technologies have been proposed for inoculating such viruses in field conditions, including by sand blasting (Gressel 2008). Self biocontrol—avoid the middleman. A seemingly science fiction concept based on a proposal suggested for insect control (Grigliatti et al. 2001) was to use genomics to have parasitic weeds kill themselves (Gressel and Levy 2000). It was later extrapolated to other outcrossing weeds (Gressel 2002) and for use with male sterility (Rector 2009). It would use genomic-based leads for herbicide targets described above, but would not need any chemical leads for those targets. Grigliatti et al. (2001) proposed taking any sequence that could kill a pest if expressed (e.g. antisense or RNAi segments of key genes), and splicing it to a chemically induced promoter. The construct is not transformed directly into the organism. Instead, it is transformed into a multicopy transposon, and the transposon transformed into the pest. Organisms bearing this multicopy transposon would then be released in the field. Had the construct been on a chromosome, only half the progeny would have the gene in the following generation in a diploid organism. Having the construct on a multicopy transposon results in all progeny with the trait, irrespective of ploidy. In plants this would work only with outcrossers, not with self-pollinators. Grigliatti et al. (2001) showed with insects that in five to six generations the transposon with the quiescent gene is distributed to almost 100% of the population. They then proposed to chemically activate the promoters, turning on the genes, and killing the pests. Such chemically-induced suicide

GENOMICS AND WEEDS: A SYNTHESIS

241

genes have been given the generic designation of kev genes (after Kevorkian) (Gressel 2002). Various ways were designed to use this with Striga (Gressel and Levy 2000) and other outcrossing weeds (Gressel 2002), but the drawback is finding the right promoter. Promoter genomics is a hot area of research at present, because there are no foolproof and inexpensive inducible promoters. Copper inducible promoters work in the lab, but most fields have sufficient soil copper to render the promoters constitutive. Alcohol-induced promoters can be turned on by stresses that cause glycolytic alcohol production. The only good promoters use an expensive insecticide to turn it on. Perhaps the answer will come from bat functional genomics, or the genomics of moths that hear bats (Fullard et al. 2008) and the induction will be physical. The high frequencies that bats use for echolocation must have receptors that are gene products. Such receptors could be used to activate promoters in echo-kev genes. When the multi-copy transposon has distributed itself innocuously through a weed population, the gene is turned on by the specific high frequency ultrasonic signal. A secondary effect of turning on those genes might be in scaring off some other mobile pests. While the neurophysiology of receptor cells has advanced, the receptors themselves have not been isolated, let alone sequenced (Jones and Holderied 2007; Kemmer and Vater 2001; Waters 2003).

Where Do We Go From Here?

Weeds are clearly one of the greatest enemies of food production, and understanding weed genomics is critical to effectively dealing with them at an advanced level that minimizes costs and adverse environmental impacts. There are two major challenges to arriving at this level of sophistication. The first is gaining financial support for the necessary research. Those controlling the funding of plant genomic research too often claim that funds should only go to species of economic value (crops), not understanding that important economic value can be either positive or negative. The direct economic and sometimes social and environmental costs of dealing with weeds in crops are higher than for other pest categories in most cases. Yet few plant scientists, much less administrators of scientific research funding, are aware of this. Understanding the functional genomics of a crop will be helpful in weed management, but it is only a piece of the puzzle. The second challenge will be to those of us who do find funding for weed and weed/crop genomics. Translating this genomic information into robust and reliable technologies and strategies for weed management will not be a trivial task. Integration of genomics data with information from the other “omics” will be required in most cases for meaningful progress (Yuan et al. 2008). It will require truly collaborative interactions from the laboratory to the field, involving a broad range of plant science and perhaps other disciplines.

References Amsellem Z, Kleifeld Y, Kerenyi Z, Hornok L, Goldwasser Y, Gressel J (2001) Isolation, identification, and activity of mycoherbicidal pathogens from juvenile broomrape plants. Biological Control 21, 274–284. Amsellem Z, Cohen BA, Gressel J (2002) Transgenically conferring sufficient hypervirulence on an inundative mycoherbicidal fungus for efficient weed control. Nature Biotechnology 20, 1035–1039.

242

WEEDY AND INVASIVE PLANT GENOMICS

Aoki K, Ogata T, Shibata D (2007) Approaches for extracting practical information from gene co-expression networks in plant biology. Plant and Cell Physiology 48, 381–390. Aslamkhan A, Guo L, Mulpuri R, Hoffman N, Kjemtrup S, Kearney J, Christensen CA, Davis K, Zayed A, Ascenzi R, Boyes B (2005) Fluorescent assay for identifying inhibitors of plant or fungal chlorophyll synthase useful as herbicides. U. S. Patent Application 2005130118. Aviv D, Amsellem Z, Gressel J (2002) Transformation of carrots with mutant acetolactate synthase for Orobanche (broomrape) control. Pest Management Science 58, 1187–1193. Baerson SB, Sánchez-Moreiras A, Pedrol-Bonjoch N, Schulz M, Kagan IA, Agarwal AK, Reigosa MJ, Duke SO (2005) Detoxification and transcriptome response in Arabidopsis seedlings exposed to the allelochemical benzoxazolin-2(3H)one (BOA). Journal of Biological Chemistry 280, 21867–21881. Baerson SB, Dayan FE, Rimando AM, Nanayakkara NPD, Liu C-J, Schroeder J, Fishbein M, Pan Z, Kagan IA, Pratt LH, Cordonnier-Pratt M, Duke SO (2008) A functional genomics investigation of allelochemical bioysynthesis in sorghum bicolor root hairs. Journal of Biological Chemistry 283, 3231–3247. Bailey BA, Apel-Birkhold PC, Akingbe OO, Ryan JL, O’Neill NR, Anderson JD (2000). Nep1 protein from Fusarium oxysporum enhances biological control of opium poppy by Pleospora papaveracea. Phytopathology 90, 812–818. Basu C, Halfhill MD, Mueller TC, Stewart CN Jr. (2004) Weed genomics: new tools to understand weed biology. Trends in Plant Science 9, 391–398. Berardini TZ, Mundodi S, Reiser L, Huala E, Garcia-Hernandez M, Zhang P, Mueller LA, Yoon J, Doyle A, Lander G, Moseyko N, Yoo D, Xu I, Zoeckler B, Montoya M, Miller N, Weems D, Rhee SY (2004) Functional annotation of the Arabidopsis genome using controlled vocabularies. Plant Physiology 135, 745–755. Berg D, Tietjen K, Wollweberet D, Hain R (1999) From genes to targets: impact of functional genomics on herbicide discovery. In: Proceedings of the Brighton Crop Protection Conference-Weeds, pp. 491–500. Berville A, Muller M-H, Poinso B, Serieys H (2005) Ferality—risks of gene flow between sunflower and other Helianthus species. In: Crop Ferality and Volunteerism Gressel J, ed. pp. 209–230. CRC Press, Boca Raton. Bharucha N, Kumar A (2007) Yeast genomics and drug target identification. Combinatorial Chemistry and High Throughput Screening 10, 618–634. Bouchez D, Hofte H (1998) Functional genomics in plants. Plant Physiology 118, 725–732. Brosnan CA, Mitter N, Christie M, Smith NA, Waterhouse PM, Carroll BJ (2007) Nuclear gene silencing directs reception of long-distance mRNA silencing in Arabidopsis. Proceedings of the National Academy of Sciences of the United States of America 104, 14741–14746. Bruce TJA, Matthes MC, Napier JA, Pickett JA (2007) Stressful “memories” of plants: Evidence and possible mechanisms. Plant Science 173, 603–608. Brutnell TP (2002) Transposon tagging in maize. Functional and Integrative Genomics 2,4–12. Busch W, Lohmann JU (2007) Profiling a plant: expression analysis in Arabidopsis. Current Opinion in Plant Biology 10, 136–141. Castle LA, Siehl DL, Gortaon R, Ratten PA, Chen YH, Bertain S, Cho H-J, Duck N, Wong J, Liu D, Lassner MW (2004) Discovery and directed evolution of a glyphosate tolerance gene. Science 304, 1151–1154. Cerdeira AL, Duke SO (2006) The current status and environmental impacts of glyphosate-resistant crops: A review. Journal of Environmental Quality 35, 1633–1658. Cerdeira AL, Gazziero DLP, Duke SO, Mattalo MB, Spadotto CA (2007) Review of potential environmental impacts of transgenic glyphosate-resistant soybean in Brazil. Journal of Environmental Science and Health, Part B 42, 539–549. Cole DJ, Pallett K, Rogers M (2000) Discovering new modes of action for herbicides and the impact of genomics. Pesticide Outlook 11, 223–229. Cole DJ, Rodgers MW (2000) Plant molecular biology for herbicide-tolerant crops and discovery of new herbicide targets. In: Mechanisms of Herbicide Action Cobb AH, Kirkwood RC, eds. pp. 239–278. Academic Press, Sheffield. Czarnota MA, Paul RN, Weston LA, Duke SO (2003) Anatomy of sorgoleone-secreting root hairs of sorghum species. International Journal of Plant Science 164, 861–866. Daniell H (2002) Molecular strategies for gene containment in transgenic crops. Nature Biotechnology 20, 581–586. Davis K, Zayed A, Ascenzi R, Smith M, Kurnik BS, Boyes D, Mulpuri R, Hoffman N, Kjemtrup S, Hamilton C, Woessner J, Gorlach J (2004) Identification of carbonic anhydrase as an essential enzyme for plant growth and a target for herbicides. U.S. Pat. Appl. Publ. US 2004248152 A1 20041209. Dayan FE, Duke SO (2003) Herbicides: Protoporphyrinogen oxidase inhibitors. In: Encyclopedia of Agrochemicals, Vol. 2, Plimmer JR, Gammon DW, Ragsdale NN, eds. pp. 850–863. John Wiley and Sons, New York, NY. Dayan FD, Duke SO (2009) Biological activity of allelochemicals. In: Plant-Derived Natural Products—Synthesis, Function and Application, Osbourn A, ed. Springer, in press.

GENOMICS AND WEEDS: A SYNTHESIS

243

de Framond A, Rich PJ, McMillan J, Ejeta G (2007) Effects of Striga parasitism of transgenic maize armed with RNAi constructs targeting essential S. asiatica genes. In: Integrating New Technologies for Striga Control: Ending the Witch-hunt Ejeta G, Gressel J, eds. World Scientific, Singapore. Dennis G Jr., Sherman BT, Hosack DA, Yang J, Gao W, Lane HC, Lempicki RA (2003) DAVID: Database for annotation, visualization, and integrated discovery. Genome Biology 4, 3. Dill GM, CaJacob CA, Padgette SR (2008) Glyphosate-resistant crops: adoption, use and future considerations. Pest Management Science 64, 326–331. Dinelli G, Marotti I, Bonetti A, Catizone P, Urbanco JM, Barnes J (2006) Physiological and molecular insight on the mechanisms of resistance to glyphosate in Conyza canadensis (L.) Cronq. biotypes. Pesticide Biochemistry and Physiology 86, 30–41. Donn G, Tischer E, Smith JA, Goodman HM (1984) Herbicide-resistant alfalfa cells: an example of gene amplification in plants. Journal of Molecular and Applied Genetics 2, 621–635. Duke SO, Cerdeira AL (2005a) Transgenic herbicide-resistant crops: Current status and potential for the future. Outlooks on Pest Management 16, 208–211. Duke SO, Cerdeira AL (2005b) Potential environmental impacts of herbicide-resistant crops. In: Collection of Biosafety Reviews, Vol. 2, pp. 66–143. International Centre for Genetic Engineering and Biotechnology, Trieste, Italy. Duke SO, Baerson SB, Pan Z, Kagan IA, Sánchez-Moreiras A, Reigosa MJ, Pedro-Bonjoch N, Schultz M (2008) Genomic approaches to understanding allelochemical effects on plants. In: Allelopathy Principles and Applications in Agriculture and Forestry Zen RS, Mallik AU, Luo SM, eds. pp. 157–167, Springer, New York. Duke SO, Powles SB (2008) Glyphosate: A once in a century herbicide. Pest Management Science 64, 319–325. Ejeta G, Grenier C (2005) Sorghum and its weedy hybrids. In: Crop Ferality and Volunteerism Gressel J, ed. pp. 123–135. CRC Press, Boca Raton. Fairbairn DJ, Cavallaro AS, Bernard M, Mahalinga-Iyer J, Graham MW, Botella JR (2007) Host-delivered RNAi: an effective strategy to silence genes in plant parasitic nematodes. Planta 226, 1525–1533. Federoff N (2000) Transposons and gene evolution in plants. Proceedings of the National Academy of Sciences USA 97, 7002–7007. Ferhatoglu Y, Barrett M (2006) Studies of clomazone mode of action. Pesticide Biochemistry and Physiology 85, 7–14. Freund A, Sonnewald U, Ding L (2003) Plant dehydroquinate dehydratase/shikimate dehydrogenases for use as herbicide targets and in increasing plant dry mass yield. US Patent 20030145348. Frey M, Chomet P, Glawischnig E, Stettner C, Grün S, Winklmair A, Eisenreich W, Bacher A, Meeley RB, Briggs SP, Simcox K, Gierl A (1997) Analysis of a chemical plant defense mechanism in grasses. Science 277, 696–699. Frey M, Huber K, Park WJ, Sicker D, Lindberg P, Meeley RB, Simmons CR, Yalpani N, Gierl A (2003) A 2-oxoglutaratedependent dioxygenase is integrated in DIMBOA-biosynthesis. Phytochemistry 62, 371–376. Fullard JH, Ratcliffe JM, Jacobs DS (2008) Ignoring the irrelevant: auditory tolerance of audible but innocuous sounds in the bat-detecting ears of moths. Naturwissenschaften 95, 241–245. Gerhold DL, Jensen RV, Gullans SR (2002) Better therapeutics through microarrays. Nature Genetics 32, 547–551. Gianessi LP (2008) Economic impacts of glyphosate-resistant crops. Pest Management Science 64, 346–352. Glombitza S, Dubuis PH, Thulke O, Welzl G, Bovet L, Goetz M, Affenzeller M, Geist B, Hehn A, Asnaghi C, Ernst D, Seidlitz H, Gundlach H, Mayer KF, Martinoia E, Werck-Reichart D, Mauch F, Schaeffner AR (2004) Crosstalk and differential response to abiotic and biotic stresses reflected at the transcriptionnal level of effector genes from seconday metabolism. Plant Molecular Biology 54, 817–835. Golisz A, Sugano M, Fujii Y (2008) Microarray expression profiling of Arabidopsis thaliana L. in response to allelochemicals identified in buckwheat. Journal of Experimental Botany 59, 3099–30109. Green JM, Hazel CB, Forney DR, Pugh LW (2008) New multiple-herbicide crop resistance and formulation technology to augment the utility of glyphosate. Pest Management Science 64, 332–339. Grenier C, Ibrahim Y, Haussmann BIG, Kiambi D, Ejeta G (2007) Marker assisted selection for Striga resistance in sorghum. In: Integrating New Technologies for Striga Control: Ending the Witch-hunt. Ejeta G, Gressel J, eds. pp. 159–171. World Scientific, Singapore. Gressel J (1992) The needs for new herbicide-resistant crops. In: Achievements and Developments in Combating Pesticide Resistance. Denholm I, Devonshire AL, Hollomon DW, eds. pp. 283–294. Elsevier, London. Gressel J (1999) Tandem constructs: preventing the rise of superweeds. Trends in Biotechnology 17, 361–366 Gressel J (2002) Molecular Biology of Weed Control. Taylor and Francis, London, 505 pp. Gressel J (2005) Problems in qualifying and quantifying assumptions in plant protection models: Resultant simulations can be mistaken by a factor of million. Crop Protection 24, 1007–1015. Gressel J (2008) Genetic Glass Ceilings—Transgenics for Crop Biodiversity, Johns Hopkins University Press, Baltimore, 461 pp.

244

WEEDY AND INVASIVE PLANT GENOMICS

Gressel J, Levy A (2000) Giving Striga hermonthica the DT’s. In: Breeding for Striga Resistance in Cereals. Haussmann BIG, Hess DE, Koyama ML, Grivit I, Rattunde HFW, Geiger HH, eds. pp. 207–224. Margraf Verlag, Weikersheim. Gressel J, Shaaltiel Y (1993) Synergistic herbicidal compositions comprising herbicides which generate toxic oxygen radicals and chelating agents which inhibit their detoxification. U.S. Patent, 5, 219, 825. Grigliatti TA, Pfeifer TA, Meister GA (2001) TAC-TICS: transposon-based insect control systems. In: Enhancing Biocontrol Agents and Handling Risks. Vurro M, Gressel G, Butt T, Harman GE, Pilgeram A, St. Leger RJ, Nuss DL, eds. pp. 201–216. IOS Press, Amsterdam. Haake V, Zrenner R, Sonnewald U, Stitt M (1998) A moderate decrease in plastid aldolase activity inhibits photosynthesis, alters the levels of sugars and starch, and inhibits growth of potato plants. The Plant Journal 14, 147–157. Hay JV (1998) Herbicide discovery in the 21st century—A look into the crystal ball. In: Pesticide Chemistry and Bioscience—The Food-Environment Challenge. Brooks GT, Roberts TR, eds. pp. 55–65. Royal Society of Chemistry, Cambridge, UK. Hayashi H, Czaja I, Lubenow H, Schell J, Walden R (1992) Activation of a plant gene by T-DNA tagging: auxin-independent growth in vitro. Science 258, 1350–1353. Henkes S, Sonnewald U, Badur R, Flachmann R, Stitt M (2001) A small decrease of plastid transketolase activity in antisense tobacco transformants has dramatic effects on photosynthesis and phenylpropanoid metabolism. Plant Cell 13, 535–551. Hilson P, Allemeersch J, Altmann T, Aubourg S, Avon A, Beynon J, Bhalerao RP, Bitton F, Caboche M, Cannoot B, Chardakov V, Cognet-Holliger C, Colot V, Crowe M, Darimont C, Durinck D, Eickhoff H, Falcon de Longevialle A, Farmer AE, Grant M, Kuiper MTR, Lehrach H, Léon C, Leyva A, Lundeberg J, Lurin C, Moreau Y, Nietfeld W, Paz-Ares J, Reymond P, Rouzé P, Sandberg G, Dolores Segura M, Serizet C, Tabrett A, Taconnat L, Thareau V, Van Hummelen P, Vercruysse S, Vuylsteke M, Weingartner M, Weisbeek PJ, Wirta V, Wittink FRA, Zabeau M, Small I (2004) Versatile gene-specific sequence tags for Arabidopsis functional genomics: transcript profiling and reverse genetics applications. Genome Research 14, 2176–2189. Höfgen R (1999) Molecular identification of herbicide targets apply transgenic approaches in tissue culture. Current Plant Science and Biotechnology in Agriculture 36, 495–498. Höfgen R, Axelsen KB, Kannangara CG, Schutte I, Pohlenze HD, Willmitzer L, Grimm B, von Wettstein D (1994) A visible marker for antisense mRNA expression in plants: inhibition of chlorophyll synthesis with a glutamate-1-semialdehyde aminotransferase antisense gene. Proceedings of the National Academy of Sciences USA 91, 1726–1730. Höfgen R, Laber B, Schuttke I, Klonus AK, Streber W, Pohlenz HD (1995) Repression of acetolactate synthase activity through antisense inhibition-molecular analysis and biochemical analysis of transgenic potato (Solanum-tuberosum L. cv Desiree) plants. Plant Physiology 107, 469–477. Höfgen R, Freitag J, Maimann S, Schmidt F, Hesse H (1999) Molecular approaches supporting the identification and validation of new herbicide targets. Proceedings of the Brighton Crop Protection Conference—Weeds, pp. 501–508. Huang GZ, Allen R, Davis EL, Baum TJ, Hussey RS (2006) Engineering broad root-knot resistance in transgenic plants by RNAi silencing of a conserved and essential root-knot nematode parasitism gene. Proceedings of the National Academy of Sciences USA 103, 14302–14306. Hudson GS, Evans JR, Von Caemmerer S, Arvidsson YBC, Andrews TJ (1992) Reduction of ribulose-1,5-bisphosphate carboxylase/oxygenase content by antisense RNA reduces photosynthesis in transgenic tobacco plants. Plant Physiology 98, 294–302. James C (2008) International Service for the Acquisition of Agri-biotech Applications. http://www.isaaa.org/resources/ publications/briefs/37/pptslides/Brief37slides.pdf. Joel DM, Kleifeld Y, Losner-Goshen D, Herzlinger G, Gressel J (1995) Transgenic crops against parasites. Nature 374, 220–221. Jonczyk R, Schmidt H, Osterrieder A, Fiesselmann A, Schullehner K, Haslbeck M, Sicker D, Hofmann D, Yalpani N, Simmons C, Frey M, Gierl A (2008) Elucidation of the final reactions of DIMBOA-glucoside biosynthesis in maize: characterization of Bx6 and Bx7. Plant Physiology 146, 1053–1063. Jones G, Holderied MW (2007) Bat echolocation calls: adaptation and convergent evolution. Proceedings of the Royal Society B-Biological Sciences 274, 905–912. Jun JH, Kim CS, Cho DS, Kwak JM, Ha CM, Park YS, Cho BH, Patton DA, Nam HG (2002) Random antisense cDNA mutagenesis as an efficient functional genomic approach in higher plants. Planta 214, 668–674. Kalantidis K, Schurnacher HT, Alexiadis T, Helm JM (2008) RNA silencing movement in plants. Biology of the Cell 100, 13–26. Kanampiu FK, Kabambe V, Massawe C, Jasi L, Friesen D, Ranson JK, Gressel J (2003) Multi-site, multi-season field tests demonstrate that herbicide seed-coating herbicide-resistance maize controls Striga spp. and increases yields in several African countries. Crop Protection 22, 697–706.

GENOMICS AND WEEDS: A SYNTHESIS

245

Kanampiu FK, Diallo A, Burnet M, Karaya H, Gressel J (2007) Success with the low biotech seed-coated imidazolinoneresistance maize. In: Integrating New Technologies for Striga Control: Ending the Witch-hunt. Ejeta G, Gressel J, eds. pp. 145–158. World Scientific, Singapore. Kemmer M, Vater M (2001) Cellular and subcellular distribution of AMPA-type glutamate receptor subunits and metabotropic glutamate receptor 1 alpha in the cochlear nucleus of the horseshoe bat (Rhinolophus rouxi). Hearing Research 156, 128–142. Konish S, Lin SY, Fukuta Y, Izawa T, Sasaki T, Yano M (2005) Molecular cloning of a major QTL, QSH-1, controlling seed shattering habit in rice. Plant and Cell Physiology 46, S198-S198 Koprek T, McElroy D, Louwerse J, Williams-Carrier R, Lemaux PG (2000) An efficient method for dispersing Ds elements in the barley genome as a tool for determining gene function. The Plant Journal 24, 253–263. Kruse E, Mock HP, Grimm B (1995) Reduction of coproporphyrinogen oxidase level by antisense RNA synthesis leads to deregulated gene expression of plastid proteins and affects the oxidative defense system. EMBO Journal 14, 3712–3720. Kurnik BS, Davis K, Zayed A, Ascenzi R, Harper A, Boyes D, Mulpuri B, Hoffman N, Kjemtrup S, Woessner J, Gorlach J, Hamilton C (2003) Transgenic Arabidopsis expressing antisense thioredoxin gene for herbicide screening. U.S. Pat. Appl. Publ. 2003113786. Kuzuyama T, Shimizu T, Takahashi S, Seto H (1998) Fosmidomycin, a specific inhibitor of 1-deoxy-D-xylulose 5-phosphate reductoisomerase in the nonmevalonate pathway for terpenoid biosynthesis. Tetrahedron Letters 39, 7913–7916. Lange BM, Wildung MR, McCaskill D, Croteau R (1998) A family of transketolases that directs isoprenoid biosynthesis via a mevalonate-independent pathway. Proceedings of the National Academy of Sciences USA 95, 2100–2104. Lein W, Börnke F, Reindl A, Ehrhardt T, Stitt M, Sonnewald U (2004) Target-based discovery of novel herbicides. Current Opinion in Plant Science 7, 219–225. Li W, Gill BS (2006) Multiple genetic pathways for seed shattering in the grasses. Functional and Integrative Genomics 6, 300–309. Maimann S, Wagner C, Kreft O, Zeh M, Willmitzer L, Höfgen R, Hesse H (2000) Transgenic potato plants reveal the indispensable role of cystathionine β-lyase in plant growth and development. Plant Journal 23, 747–758. Maliga P (2004) Plastid transformation in higher plants. Annual Review of Plant Biology 55, 289–313. Mallory-Smith C, Zapiola M (2008) Gene flow from glyphosate-resistant crops. Pest Management Science 64, 428–440. Manabe Y, Tinker N, Colville A, Miki B (2007) CSR1, the sole target of imidazolinone herbicide in Arabidopsis thaliana. Plant and Cell Physiology 48, 1340–1358. Mandel MA, Feldmann KA, Herrera-Estrella L, Rocha-Sosa M, Leon P (1996) CLAI, a novel gene required for chloroplast development, is highly conserved in evolution. The Plant Journal 9, 649–658. Manfield IW, Orfila C, McCartney L, Harholt J, Bernal AJ, Scheller HV, Gilmartin PM, Mikkelson JD, Knox JD, Willats WGT (2004) Novel cell wall architecture of isoxaben-habituated Arabidopsis suspension-cultured cells: Global transcript profiling and cellular analysis. Plant Journal 40, 260–275. Mao YB, Cai WJ, Wang JW, Hong GJ, Tao XY, Wang LJ, Huang YP, Chen XY (2007) Silencing a cotton bollworm P450 monooxygenase gene by plant-mediated RNAi impairs larval tolerance of gossypol. Nature Biotechnology 25, 1307–1313. Marton MJ, DeRisi JL, Bennett HA, Lyer VR, Meyer MR, Roberts CJ, Stoughton R, Burchard J, Slade D, Dai H, Bassett DE, Hartwell LH, Brown PO, Friend SH (1998) Drug target validation and identification of secondary drug target effects using DNA microarrays. Nature Medicine. 4, 1293–1301. Matzke MA, Matzke AJ, Pruss GJ, Vance VB (2001) RNA-based silencing strategies in plants. Current Opinion in Genetics and Development 11, 221–227. McLaren JS (2000) The importance of genomics to the future of crop production. Pest Management Science 56, 573–379. Meir S, Amsellem Z, Al-Ahmad H, Safran E, Gressel J (2009) Transforming a NEP1 toxin gene into two Fusarium spp. to enhance mycoherbicide activity on Orobanche—failure and success Pest Management Science 65(5), 588–595. Mock HP, Grimm B (1997) Reduction of uroporphyrinogen decarboxylase by antisense RNA expression affects activities of other enzymes involved in tetrapyrrole biosynthesis and leads to light-dependent necrosis. Plant Physiology 113, 1101–1112. Molina A, Volrath D, Guyer D, Maleck K, Ryals J, Ward E (1999) Inhibition of protoporphyrinogen oxidase expression in Arabidopsis causes a lesion-mimic that induces systemic acquired resistance. The Plant Journal 17, 667–678. Nelson DS, Bullock GC (2003) Simulating a relative environmental effect of glyphosate-resistant soybeans. Ecological Economics 45, 189–202. Oliver MJ, Quisenberry JE, Trolinder NLG, Keim DL (1998) Control of plant gene expression. U.S. Patent 5723765. Østergaard L, Kempin SA, Bies D, Klee HJ, Yanofsky MF (2006) Pod shatter resistant Brassica fruit produced by ectopic expression of the FRUITFULL gene. Plant Biotechnology Journal 4, 45–51.

246

WEEDY AND INVASIVE PLANT GENOMICS

Otomo K, Kanno Y, Motegi A, Kenmoku H, Yamane H, Mitsuhashi W, Oikawa H, Toshima H, Itoh H, Matsuoka H, Sassa T, Toyomasu T (2004) Diterpene cyclases responsible for the biosynthesis of phytoalexins, momilactones A, B, and oryzalexins A-F in rice. Bioscience, Biotechnology and Biochemistry 68, 2001–2006. Pan Z, Rimando AM, Baerson SB, Fishbein M, Duke SO (2007) Functional characterization of desaturases involved in the formation of the terminal double bond of an unusual 16:3∆9,12,15 fatty acid isolated from Sorghum bicolor root hairs. Journal of Biological Chemistry 282, 4326–4335. Peters RJ (2006) Uncovering the complex metabolic network underlying diterpenoid phytoalexin biosynthesis in rice and other cereal crop plants. Phytochemistry 67, 2307–2317. Press MC (1995) Carbon and nitrogen relations. In: Parasitic Plants. Press MC, Graves JD, eds. pp. 103–124. Chapman and Hall, London. Qi X, Bakht S, Leggett M, Maxwell C, Melton R, Osbourn A (2004) A gene cluster for secondary metabolism in oat— Implications for the evolution of metabolic diversity in plants. Proceedings of the National Academy of Sciences, USA 101, 8233–8238. Raghavan C, Ong EK, Dalling MJ, Stevenson TW (2005) Regulation of genes associated with auxin, ethylene, and ABA pathways by 2,4-diclorophenoxyacetic acid in Arabidopsis. Functional and Integrated Genomics 6, 60–70. Rector BG (2009) A sterile-female technique proposed for control of certain parasitic and intractable weeds: Advantages, shortcomings, and risk management. Pest Management Science 65, 596–602. Sakamoto T, Miura K, Itoh H, Tatsumi R, Ueguchi-Tanaka M, Ishiyama K, Kobayashi M, Agrawal GK, Takeda S, Abe K, Miyao A, Hirochika H, Kitano H, Ashikari M, Matsuoka M (2004) An overview of gibberellin metabolism enzyme genes and their related mutants in rice. Plant Physiology 134, 1642–1653. Schena M, Shalon D, Davis RW, Brown PO (1995) Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270, 467–470. Shaw KJ, Morrow BJ (2003) Transcriptional profiling and drug discovery. Current Opinion in Pharmacology 3, 508–512. Shimura K, Okada A, Okada K, Jikumaru Y, Ko KW, Toyomasu T, Sassa T, Hasegawa M, Kodama O, Shibuya N, Koga J, Nojiri H, Yamane H (2007) Identification of a biosynthetic gene cluster in rice for momilactones. J Biol Chem. 282, 34013–34018. Shyr YJ, Hepburn AG, Widholm JM (1992) Glyphosate selected amplification of the 5-enolpyruvylshikimate 3-phosphate synthase gene in cultured carrot cells. Molecular and General Genetics 232, 377–382. Steiner AJ, Smith WF, Smith JD (1992) Synergists for herbicidal composition. United States Board of Patent Appeals, Appeal 02–2108. Steinrücken HC, Schulz A, Amrhein N, Porter A, Fraley RT (1986) Overproduction of 5-enol pyruvylshikimate -3 -phosphate synthase in a glyphosate tolerant Petunia hybrida cell line. Archives of Biochemistry and Biophysics 244, 169–178. Steppuhn A, Gase K, Krock B, Halitschke R, Baldwin IT (2004) Nicotine’s defensive function in nature. PLoS Biology 2, 8 e217 doi:10.1371/journal.pbio.0020217. Surov T, Aviv D, Aly R, Joel DM, Goldman-Guez T, Gressel J (1998) Generation of transgenic asulam-resistant potatoes to facilitate eradication of parasitic broomrapes (Orobanche spp.) with the sul gene as the selectable marker. Theoretical and Applied Genetics 96, 132–137. Tanaka R, Oster U, Kruse E, Rudiger W, Grimm B (1999) Reduced activity of geranylgeranyl reductase leads to loss of chlorophyll and tocopherol and to partially geranylgeranylated chlorophyll in transgenic tobacco plants expressing antisense RNA for geranylgeranyl reductase. Plant Physiology 120, 695–704. Tarantino D, Vannini C, Bracale M, Campa M, Soave C, Murgia I (2005) Antisense reduction of thylakoidal ascorbate peroxidase in Arabidopsis enhances Paraquat-induced photooxidative stress and nitric oxide-induced cell death. Planta 221, 757–765. Timko MP, Gowda BS, Ouedraogo J, Ousmane B (2007) Molecular markers for the analysis of resistance to Striga gesnerioides in cowpea. In: Integrating New Technologies for Striga Control: Ending the Witch-hunt. Ejeta G, Gressel J, eds. pp. 115–128. World Scientific, Singapore. Tissier A, Marillonnet C, Klimyuk V, Patel K, Torres MA, Murphy G, Jones JD (1999) Multiple independent suppressormutator transposon insertions in Arabidopsis: a tool for functional genomics. Plant Cell 11, 1841–1852. Todd MD, Guo L, Davis K, Zayed A, Kjemtrup S, Boyes D, Warrick B, Mitchell JC, Ascernzi R, Hamilton C, Woessner B, Gorlach J, Hoffman N, Mulpuri R (2004) Identification of sphingolipid 4-hydroxylase as an essential enzyme for plant growth and a target for herbicides. U.S. Pat. Appl. Publ. US 2004265836. Tong JP, Liu XJ, Zhang SY, Li SQ, Peng XJ, Yang J, Zhu YG (2007) Identification, genetic characterization, GA response and molecular mapping of Sdt97: a dominant mutant gene conferring semi-dwarfism in rice (Oryza sativa L.). Genetical Research 89, 221–230. Tourneur C, Jouanin L, Vaucheret H (1993) Over expression of acetolactate synthase confers resistance to valine in transgenic tobacco. Plant Science 88, 159–164.

GENOMICS AND WEEDS: A SYNTHESIS

247

Valverde BE (2005) The damage by weedy rice—can feral rice remain undetected? In: Crop Ferality and Volunteerism. Gressel J, ed. pp. 279–294. CRC Press, Boca Raton. Valverde BE, Gressel J, Passalacqua S, Rodríguez JC (2007) The emerging problem of glyphosate-resistant johnsongrass (Sorghum halepense) in Argentina: An account of detection, initial spread and collaborative action for its prevention and management. Science Society of America Annual Meeting, Abstract 183. van Baarlen P, van Esse HP, Siezen RJ, Thomma PB (2008) Challenges in plant cellular pathway reconstruction based on gene expression profiling. Trends in Plant Science 13, 44–50. Vande Berg BJ, Hammer PE, Chun BL, Schouten LC, Carr B, Guo R, Peters C, Hinson TK, Beilinson V, Shekita A, Deter R, Chen Z, Samoylov V, Bryant CT, Stauffer ME, Eberle E, Moellenbeck DJ, Carozzi NB, Koziel MG, Duck NB (2008) Characterization and plant expression of glyphosate-tolerant enolpyruvylshikimate phosphate synthase. Pest Management Science 64, 340–345. Vaughan DA, Sanchez PL, Ushiki J, Kaga A, Tomooka N (2005) Asian rice and weedy rice—evolutionary perspectives. In: Crop Ferality and Volunteerism. Gressel J. ed. CRC Press, Boca Raton. Vega-Sanchez ME, Gowda M, Wang GL (2007) Tag-based approaches for deep transcriptome analysis in plants. Plant Science 173, 371–380. Vila-Aiub MM, Balbi MC, Gundel PE, Ghersa CM, Powles SB (2007) Evolution of glyphosate-resistant Johnsongrass (Sorghum halepense) in glyphosate-resistant soybean. Weed Science 55, 566–571. Vila-Aiub MM, Vidal RA, Balbi MC, Gundel PE, Trucco R, Ghersa CM (2008) Glyphosate-resistant weeds of South American cropping systems: an overview. Pest Management Science 64, 366–371. Vizir I, Anderson M, Wilson Z, Mulligan B (1994) Isolation of deficiencies in the Arabidopsis genome by gamma-irradiation of pollen. Genetics 137, 1111–1119. Wagner A (2000) The role of population size, pleiotropy and fitness effects of mutations in the evolution of overlapping gene functions. Genetics 154, 1389–1401. Wagner A, Rohl R, Grossmann K, Schmidt RM, Sonnewald U, Hajirezaei M (2000) Herbicidal compositions and processes based on ferrodoxin (sic.):NADP reductase inhibitors. U.S. Patent 6,124242. Wang T, Li Y, Shi Y, Reboud X, Darmency H, Gressel J (2004). Low frequency transmission of a plastid encoded trait in Setaria italica. Theoretical and Applied Genetics 108, 315–320. Warwick SI, Légère A, Simard MJ, James T (2008) Do escaped transgenes persist in nature? The case of an herbicide resistance transgene in a weedy Brassica rapa population. Molecular Ecology 17, 1387–1395. Waters DA (2003) Bats and moths: what is there left to learn? Physiological Entomology 28, 237–250. Wilderman PR, Xu M, Jin Y, Coates RM, Peters RJ (2004) Identification of syn-pimara-7,15-diene synthase reveals functional clustering of terpene synthases involved in rice phytoalexin/allelochemical biosynthesis. Plant Physiology 135, 2098–2105. Wolf SJ, Timko MP (1991) In vitro root culture—a novel approach to study the obligate parasite Striga asiatica (L) Kuntze. Plant Science 73, 233–242. Xiao W, Saxena P, King J, Rank G (1987) A transient duplication of the acetolactate synthase gene in a cell culture of Datura innoxia. Theoretical and Applied Genetics 74, 417–422. Xu M, Hillwig ML, Prisic S, Coates RM, Peters RJ (2004) Functional identification of rice syn-copalyl diphosphate synthase and its role in initiating biosynthesis of diterpenoid phytoalexin/allelopathic natural products. The Plant Journal 39, 309–318. Yoder JI, Reagan R, Tomilov A, Tomilova N, Torres M (2007) Host detection by root parasites: insights from transcription profiles. In: Integrating New Technologies for Striga Control: Ending the Witch hunt. Ejeta G, Gressel J, eds. pp. 33–45. World Scientific, Singapore. Yuan JS, Galbraith DW, Dai SY, Griffin P, Stewart CN Jr. (2008) Plant systems biology comes of age. Trends in Plant Science 13, 165–171. Zeidler J, Schwender J, Mueller C, Wiesner J, Weidemeyer C, Beck E, Jomaa H, Lichtenthaler HK (1998) Inhibition of the non-mevalonate 1-deoxy-D-xylulose-5-phosphate pathway of plant isoprenoid biosynthesis by fosmidomycin. Zeitschrift für Naturforschung 53, 980–986. Zelaya IA, Owen MDK, VanGessel MJ (2004) Inheritance of evolved glyphosate resistance in Conyza canadensis (L.) Cronq. Theoretical and Applied Genetics 110, 58–57. Zelaya IA, Owen MDK (2005) Differential response of Amaranthus tuberculatus (Moq ex DC) JD Sauer to glyphosate. Pest Management Science 61, 936–950. Zhu J, Patzoldt WL, Shealy RT, Vodkin LO, Cough SJ, Tranel PJ (2008) Transcriptome response to glyphosate in sensitive and resistant soybean. Journal of Agricultural and Food Chemistry 56, 6355–6363.

Index

35S promoter, see CaMV 35S promoter, 228 454 sequencing, 20, 73, 75, 157 4 HPPD (4-hydroxyphenyl-pyruvatedioxygenase), see HPPD 5-enolpyruvylshikimate 3-phosphate synthase, see EPSPS ABA, see abscisic acid ABC transporter, 5, 72, 152–156 Abiotic stress, 11, 46, 154, 167, 189, 210 Abscisic acid (ABA), 33, 116, 118, 200–203, 207 Abutilon theophrasti, 19, 66, 153, 166, 213 ACCase, 4, 17, 44, 136–138, 150 Acetohydroxyacid synthase, see ALS, see also AHAS Acetolactate synthase, see ALS Acetylation, 150, 155 Acetyl CoA carboxylase, see ACCase Acetyltransferase, 155, 231 Acnida altissima, 27, 55, see also Amaranthus tuberculatus Acroptilon repens, 19, 99–100,189, 192 Adventitious buds, 46, 113–118, 123 Aegilops cylindrica, 12, 40, 53, 54 Affymetrix, 226, 227, 236, see also microarray AFLPs, 13, 14, 86 Agrobacterium and quorum sensing, 206 Agrobacterium EPSPS, 231 Agrobacterium tumefaciens-mediated transformation, 37, 41, 46, 76, 181, 228 AHAS, 131–136, 150 Allelochemicals, 181–185, 188–192, 233, 236 Allelopathic advantage against resident species (AARS) hypothesis, 181, 187, 191 Allelopathy, 19, 42, 178, 180–192, 205–207, 236 ALS, 4–5, 37, 44–46, 48, 60–62, 64, 66–74, 149, 239, see also AHAS Alternanthera philoxeroides, 13 Amaranthus, 14, 30, 36, 45, 50–81, 238 A. hybridus, 56–65, 72–75 A. palmeri, 4, 56, 58–58, 62, 65, 69, 72, 75, 76

A. powellii, 56, 57, 65, 67, 70, 72, 73 A. retroflexus, 56, 65, 69, 72, 73 A. rudis, 27, 45, 56, 59, 60 A. tuberculatus, 27, 28, 45, 56–77 142 Ambrosia artemisiifolia, 15, 54 Amplified fragment length polymorphism analysis, see AFLPs Alopecurus myosuroides, 17, 54, 137, 152, 221 Anisantha sterilis, 15 Antisense, 6, 7, 224–226, 230, 232, 240 Arabidopsis thaliana As a genomics model, 6, 8, 12, 13, 18, 19, 33–36, 41, 72, 76, 179 As a physiological/biochemical model, 115, 116, 121, 153, 154, 156, 171, 177, 187, 203–208, 234–236 As a weed model (or not), 11, 25–32, 132, 222–223, 226–227, 229 Artemisia A. annua, 16, 18, 188 A. biennis, 16 Atrazine, 65–66, 128, 130, 151, 153, 154 Avena, 17, 30, 36, 54, 221 Axillary buds, 100, 105, 113, 115, 116, 121 Auxin, 5, 44, 116, 118, 122, 155, 200–203, 207–209 BAC, 13, 43, 45, 47, 72–73, 76 Bacterial artificial chromosome, see BAC Barnyard grass, see Echinochloa crus-galli Beta vulgaris, 14, 46 Biocontrol, 191–192, 239–240 Bioinformatics, 6, 18, 20, 48, 223, 227 Biotic stress, 11, 46, 184, 189, 210 BLAST (basic local alignment search tool), 120, 233, 234 Brachypodium distachyon, 18, 47 Brassica, 18, 29, 30 231–232 Brassinosteroid, 201, 211 Bromoxynil, 130 Bud dormancy, 46–47, 113, 116–117, 121–123, 201

249

250

INDEX

CaMV 35S promoter, 228 Canada thistle, see Circium arvesne Capsella bursa-pastoris, 15 Catechin, 183–191, 193, 236 cDNA, 6, 11, 18–20, 44–46, 73, 75, 76, 105, 118, 123, 224, 226, 230, 233, 234, 238 Chaptalia nutans, 17 Chemical genetics, 7 Chenopodium album, 54, 130, 1665 Chloroplast, 15, 17, 65, 68, 163, 179, 201, 209, 223, 225, 232 Circium, 54 C. acaule, 15 C. arvense, 16, 44–45 Comparative genomics Compartmentation/compartmentalization, 149–151, 155–156, 226 Complementary DNA, see cDNA Computational biology, see bioinformatics Cnicin, 182, 184, 188–189, 191 Comparative genomics, 18, 19, 48, 75, 186 Contig, 18, 20, 106 Conyza canadensis, 5, 30, 54, 151, 156, 237, 238 Crepis japonica, 17 Crop breeding, using weed genes, 94 Crop mimics, 35, 41, 84, 85, 88, 91 Crop-weed hybrids, 29, 93 Cynodon, 99 Cyperus, 30, 36, 54, 99 Cytochrome P450, 5, 8, 73, 116, 120, 151–154, 188, 234, 237 Cytokinin, 116, 201, 203, 207, 208 D1 protein, 64, 65, 120, 128–131 dbEST, 15, 18, 39, 41 De-domestication, 89–92 Diclofop, 137, 150, 237 Digitaria, 54 DIMBOA, 233–234 Drug discovery, 226, see also pharmaceutical discovery Echinochloa, 13, 54, 223 Ecodormancy, 117–123 Eleusine indica, 138, 140 Endodormancy, 34, 116–123 Enolpyruvylshikimate-P-synthase, see EPSPS Epigenetics/epigenomics, 238 EPSPS, 69, 73, 138–140, 228, 231 Escherichia coli, 26, 27, 30, 139 ESTs, 15, 18, 19, 39, 41, 43–48, 75, 105, 106, 188, 234

Ethylene, 8, 119, 200–205, 209, 211, 213, 214 Euphorbia esula, 18, 19, 36, 46, 54, 113–126 Evolution of increased competitive ability (EICA) hypothesis, 190–191 Expressed sequence tags, see ESTs FISH, 17 Fitness, 64–68, 76, 99, 127–128, 131, 134, 136, 138, 139, 142, 143, 165–175, 188, 189, 194, 204, 208, 210, 214, 231 Fluorescence in situ hybridization, see FISH Forward genetics, 7, 12 Functional genomics, 5, 11, 18–20, 29, 30, 33, 108, 197, 223, 233, 241 GA, see giberellic acid Galinsoga parviflora Gametogenesis, 17 Gene duplication, 152 Gene flow, 15, 30, 42, 44, 61, 67, 70–75, 83–95, 102, 108, 173, 231–232 Gene silencing, see RNAi, see also antisense construct Gene use restriction technology (GURT), 232 Genomic diversity, 84–88 Genomic in situ hybridization, see GISH Gibberellic acid/gibberellins, 106, 116, 119, 202, 207, 214, 232, 233 Glucosyltransferase 8, 154, see also glycosyltransferase Glufosinate, 73, 231 Glutathione S-transferase, see GSTs Glycosyltransferase, 73, 152, 154–155 Glyphosate 221–223, 228–231, 238–239 Importance as an herbicide, 5, 149 Resistance, general 4–5, Amaranthus, in, 4, 45, 58, 64, 65, 68–73, 238 Conyza, in, 5, 151, 156, 238 Crops, in transgenic, 4, 221–222, 228–231, 238, 239 Lolium, in, 138 Non-target mechanisms, 149–152, 155–156 Sorghum, 221 Target-site mechanisms, 138–140, 223 Tolerance, 163–172 Goatgrass, see Aegylops GS-FLX, Roche, see 454 sequencing GSTs, 8, 152–154, 156

INDEX

Helianthus, 12, 15, 54, 73 Herbicide discovery, 6–8, 28–29, 52, 222–224 Herbicide resistance, see also non-target resistance, and target-site resistance Evolution of, 4, 53, 173, 237–238 Predicting, 7, 64, 127, 129–131, 142, 150, 226, 237 Herbivory, 166–168, 171, 184, 204, 209, 211 High throughput sequencing, see next-generation sequencing Hordeum vulgare, 11, 16, 18 HPPD, 5, 29, 142 Hybridization, 14, 40, 58–77, 86–95, 109, 179, 186, 192 Illumina Genome Analyzer (Solexa) sequencing, 20, 157, see also next-generation sequencing Imidazolinone, 131 Imperata cylindrical, 99 Interactomics, 119–121 Interference, weed-crop, 35, 42, 233–235 Inter simple sequence repeats, see ISSRs Introgression/introgressive hybridization, 61–64, 83, 90, 93, 102, 108, 179, 231–233 Ipomoea, 34, 54, 166–170 ISSRs, 15–16 Jasmonic acid, 119, 202, see also methyl jasmonate Kaurene oxidase, 232 Kaurene synthase, 233, 234 Knapweed, see Centaurea maculosa Knockdown analysis, gene, 188, 230 Knockout analysis, gene, 7, 28, 188, 189, 224, 225, 229–230 Kochia scoparia, 14, 130 Ligand, 211 Lipid, 119, 236, 155 Lolium, 54, 221, 237 L. multiflorum, 138 L. rigidum, 70, 137–138, 170 Map-based cloning, 11–13, 18, 21, 42, 73 Marker-assisted breeding, 42, 239 Messenger RNA, see mRNA Metabolism Herbicide, 4, 8, 65, 73, 153, 155, 225, 236 Plant, 19, 27, 106, 119, 164, 190, 198, 203–213

251

Metabolomics, 7, 18, 29, 189 Methylation, DNA, 238 Methyl jasmonate, 202, 213 Microarray analysis, 18–20, 37, 41, 46, 48, 76, 117–123, 151, 156, 191, 198, 211–213, 226, 228, 235 Microsatellite markers, see SSRs Mitochondria, 67, 68, 75 Models, 11, 34, 36, 48, 64, 104, 107, 108, 208, 227 Molecular ecology, 197–216 Molecular markers, 12–17, 19, 48, 73, 93, 178, see also AFLPs, ISSRs, RAPDs, RFLPs, SNPs, SSRs Mollugo verticillata, 210 mRNA, 18, 105, 118, 224, 228, 238 Multiple resistance to herbicides, 70, 149 Mutation Developmental, 114–117 Genetic, general, 7, 92, 94, 104, 107–108 Independent, 13, 15 Insertional, 33, 223, 229 Non-target-site, 150, 237 Rates of, 16 Single, 211–212 Target-site, 5, 17, 28, 61–77, 88, 127–143, 149, 163, 173, 237 Next-generation sequencing, 20, 151, 157, 231, 232, see also Illumina and 454 Nicotiana attenuata, 209–211 Nightshades, see Solanum Nitrile, 128–130 Non-target resistance, 4, 5, 65, 66, 69, 149–157 Novel weapons hypothesis (NWH), 180, 191 Open reading frame, 224 Orobanche 13, 54, 238, 240 Oryza, 54 O. longistaminata, 99, 104, 105 O. rufipogon, 84, 87, 91–95 O. sativa, 14, 15, 41–42, 72, 83–95, 104, 227 Overexpression, 115, 152–156, 225, 228–230, 239 Oxidation, 150, 152 P450, see cytochrome P450 Palmer amaranth or Palmer pigweed, see Amaranthus palmeri Paradormancy, 116–122 Paraquat, 225, 228 Parasitic weeds, 35, 238–240

252 Pathogen-associated molecular patterns (PAMPs), 206 PCR, 13–20, 48, 119, 190, 198, 229, 234–240 Pharmaceutical discovery, 223–224 Phenotypic plasticity, 25, 26, 53, 72, 75, 113, 197, 207–214 Phoenix growth, 28, 238 Phosphinothricin, see PPT Photosystem I, 142 Photosystem II, 5, 29, 37, 47, 64–66, 128–131, 143 Phytoene desaturase inhibitors, 141–142 Pigweed, see Amaranthus PLACE database, 106 Plant transformation, see transgenic/ transformation Plasmids, 18, 237 Polymerase chain reaction, see PCR Polyploid evolution, 109 Portulaca oleracea, 130 Posttranscriptional gene silencing, see RNAi PPO, see Protox PPT, 155 Promoter, 17 Proteomics, 18, 48, 189 Protox (protoporphrinogen oxydase), 4, 67–71, 225–228 PS I, see photosystem I PSII, see photosystem II PsbA, 64–65 Pyrosequencing, see 454 Quantitative PCR (Q-PCR), see real-time PCR Quantitative trait loci (QTL or QTLs), 12, 18, 37, 43, 44, 94, 104–109, 172 Random amplified polymorphic DNA analysis, see RAPDs RAPDs, 13, 88 Real-time PCR, 48, 119 Recombinant inbred line, see RIL Red rice, see Oryza Reporter genes, 115 Resistance vs. tolerance, herbicide, 163–166 Restriction fragment length polymorphism, see RFLP Reverse genetics, 7, 198 Reverse transcriptase PCR, see RT-PCR RFLP, 14–15, 18, 86 Rhizomes/rhizomatousness, 35, 42, 44, 99–109, 221, 238 Rhizosphere, 154, 181, 205–206, 209, 234

INDEX

Ribulose 1,5-bisphosphate carboxylase oxygenase, see Rubisco RIL, 45, 75–76, 104, 185, 186 RNAi, 6, 7, 28, 108, 224, 226, 231, 234–236, 239–240 RNA interference, see RNAi Roundup, Roundup-Ready, see glyphosate RT-PCR, 119, 235, 236 Rubisco 6, 209–210, 225 RuBPCase, see Rubisco Saccharum spontaneum, 99, 103 Safeners, 8, 153–156, 222 Salicylic acid, 202 Secondary metabolites, 153–155, 180, 183, 186, 189, 205 Seed germination, 12, 25–26, 35, 42, 53, 60, 94, 182–183, 202, 212, 232 Selection pressure Genetic targets, 73 Herbicide, 3, 53, 64, 67, 69, 71, 93, 127–128, 136–137, 164–165 Human, 58, 61, 63, 107, 178, 191 Natural, 166–167, 197, 204, 211, 214 Recurrent, 69, 238 Senecio vulgaris, 4, 13, 64, 128, 163 Setaria, 14, 17, 54, 140, 221 Shikimate, 138–139, 202, 225 Shoot apical meristem (SAM), 113–115, 208 Signal transduction, 7, 119, 150–151, 156, 201, 213 Silene latifolia, 13 Simple sequence repeats, see SSRs, 14–15, 19, 42, 47, 86–87, 104 Single nucleotide polymorphisms, see SNPs SNPs, 16–17, 19 Solanum spp., 36–38, 54, 67 Solanum nigrum, 13, 212–215 Sorghum, 54 S. bicolor, 18, 29, 30, 43–44, 101–109, 231–235, 238, 239 S. halepense, 18, 30, 43–44, 99–109, 221, 231, 238 S. propinqum, 18, 44, 101–109 Sorgoleone, 29, 233–235 Stress tolerance, 151, 207–211 Striga, 36, 54, 238, 239, 241 Sulfonylurea, 131–134 Systems biology, 20, 151, 156, 157 TAIR, 13, 33 Taraxacum officinale, 210

INDEX

Target-site resistance, 65–66, 72–73, 127–143, 149 T-DNA, 28, 76, 226–229 The Arabidopsis Information Network, see TAIR Traits of weeds, 21, 26, 35 Transcriptional profiling, 151, 211, 214, 226, 227, 235, 236 Transcriptome/transcriptomomics, 46, 47, 75, 105, 109, 118–119, 123, 185, 198, 200, 211, 227, 228, 235, 236 Transfer DNA, see T-DNA Transgene flow or escape, 23, 77, 93, 102, 108, 231–233 Transgenic plants, 29, 40, 46, 68, 76–77, 102, 108, 115, 149, 152–157, 188–189, 202, 211, 221–222, 230–232, 236, 239 Transposon, 7, 229–230, 233–234, 240–241

253

Triazine, 4, 45, 64–66, 70, 128–129 Triticum T. aestivum, 12, 14, 37, 39–40 T. dicoccoides, 14 Tubulin, 140–141 Vacuole, 153, 155–156 Volatile organic compounds (VOCs), 201, 202, 204–205 Xanthium strumarium, 16, 54, 132 YAC (yeast artificial chromosome), 13, 18 Waterhemp, see Amaranthus rudis and A. tuberculatus Weed genes, 214