Genomics-Assisted Crop Improvement

Genomics-Assisted CropImprovement

Vol 2: Genomics Applications in Crops

Edited by

Rajeev K. VarshneyICRISAT, Patancheru, India

and

Roberto TuberosaUniversity of Bologna, Italy

A C.I.P. Catalogue record for this book is available from the Library of Congress.

ISBN 978-1-4020-6296-4 (HB)ISBN 978-1-4020-6297-1 (e-book)

Published by Springer,P.O. Box 17, 3300 AA Dordrecht, The Netherlands.

www.springer.com

Printed on acid-free paper

All Rights Reserved© 2007 SpringerNo part of this work may be reproduced, stored in a retrieval system, or transmittedin any form or by any means, electronic, mechanical, photocopying, microfilming, recordingor otherwise, without written permission from the Publisher, with the exceptionof any material supplied specifically for the purpose of being enteredand executed on a computer system, for exclusive use by the purchaser of the work.

CONTENTS

Foreword to the Series: Genomics-Assisted Crop Improvement vii

Foreword xi

Preface xiii

Color Plates xv

1. Microsatellite and SNP Markers in Wheat Breeding 1Martin W. Ganal and Marion S. Röder

2. Molecular Markers and QTL Analysis for Grain Quality Improvementin Wheat 25Domenico Lafiandra, Maria Corinna Sanguineti, Marco Maccaferriand Enzo DeAmbrogio

3. Molecular Approaches and Breeding Strategies for Drought Tolerancein Barley 51Michael Baum, Maria Von Korff, Peiguo Guo, Berhane Lakew, SripadaM. Udupa, Haitham Sayed, Wafa Choumane, Stefania Grando andSalvatore Ceccarelli

4. Molecular Markers for Gene Pyramiding and Disease ResistanceBreeding in Barley 81Wolfgang Friedt and Frank Ordon

5. Cloning Genes and QTLs for Disease Resistance in Cereals 103Beat Keller, Stéphane Bieri, Eligio Bossolini and Nabila Yahiaoui

6. Maize Breeding and Genomics: An Historical Overviewand Perspectives 129Michael Lee

7. Molecular Markers and Marker-Assisted Selection in Rice 147David J. Mackill

8. Application of Genomics for Molecular Breeding in Rice 169Nagendra K. Singh and Trilochan Mohapatra

v

vi CONTENTS

9. Marker-Assisted Selection in Sorghum 187Gebisa Ejeta and Joseph E. Knoll

10. Molecular Genetics and Breeding of Grain Legume Crops for theSemi-Arid Tropics 207Rajeev K. Varshney, David A. Hoisington, Hari D. Upadhyaya,Pooran M. Gaur, Shyam N. Nigam, Kulbhushan Saxena, Vincent Vadez,Niroj K. Sethy, Sabhyata Bhatia, Rupakula Aruna, M. V. ChannabyreGowda and Nagendra K. Singh

11. Genomics Approaches to Soybean Improvement 243Tri D. Vuong, Xiaolei Wu, MD S. Pathan, Babu Valliyodanand Henry T. Nguyen

12. Application of Genomics to Forage Crop Breeding for Quality Traits 281Thomas Lübberstedt

13. Molecular Mapping, Marker-Assisted Selection and Map-BasedCloning in Tomato 307Majid R. Foolad

14. Genomics for Improvement of Rosaceae Temperate Tree Fruit 357Pere Arús and Susan Gardiner

15. DNA Markers: Development and Application for GeneticImprovement of Coffee 399Prasad S. Hendre and Ramesh K. Aggarwal

16. Genomics of Root Nodulation in Soybean 435Kyujung Van, Moon Young Kim and Suk-Ha Lee

17. Genomics of Wheat Domestication 453Carlo Pozzi and Francesco Salamini

18. Transcriptome Analysis of The Sugarcane Genomefor Crop Improvement 483Paulo Arruda and Thaís Rezende Silva

Appendix I – List of Contributors 495

Appendix II – List of Reviewers 503

Index 505

FOREWORD TO THE SERIES: GENOMICS-ASSISTEDCROP IMPROVEMENT

Genetic markers and their application in plant breeding played a large part in myresearch career, so I am delighted to have the opportunity to write these notes toprecede the two volumes on ’Genomics-assisted crop improvement’. Although Iam not so old, I go right back to the beginning in 1923 when Karl Sax describedhow ’factors for qualitative traits’ (today’s genetic markers) could be used to selectfor ’size factors’ (today’s QTLs and genes for adaptation). But it was clear tome 40 years ago that even then plant breeders clearly understood how geneticmarkers could help them - if only they actually had the markers and understood thegenetics underlying their key traits. It was not clear to me that it was going to takeuntil the next century before marker-aided selection would become routine for cropimprovement.

In the 1960s only ’morphological’ markers were available to breeders. As aresearch student at Aberystwyth, I worked with Des Hayes at the Welsh PlantBreeding Station when he was trying to develop an F1 hybrid barley crop based ona male sterility gene linked to a DDT resistance gene. The idea was to link the malefertile allele with susceptibility and then kill the fertile plants off in segregationpopulations by dousing the field with DDT. Rachel Carson’s ’Silent Spring’ ensuredthat idea never flew.

Then I moved to the Plant Breeding Institute in Cambridge where anyone workingalongside the breeders in those early days could not help but be motivated bybreeding. Protein electrophoresis raised the first possibility of multiple neutralmarkers and we were quick to become involved in the search for new isozymemarkers in the late 1970s and early 1980s. Probably only the linkage between wheatendopeptidase and eyespot resistance was ever used by practical breeders, but wehad an immense amount of fun uncovering the genetics of a series of expensivemarkers with hardly any polymorphism, all of which needed a different visualisationtechnology!

During this same period, of course, selection for wheat bread-making qualityusing glutenin subunits was being pioneered at the PBI, and is still in use aroundthe world. These were the protein equivalent of today’s ’perfect’ or ’functional’markers for specific beneficial alleles. Such markers - although of course DNA-based, easy and economical to use, amenable to massively high throughput andavailable for all key genes in all crops - are exactly where we want to end up.

vii

viii FOREWORD TO THE SERIES

Proteins were superseded by RFLPs and in 1986 we set out to make a wheat map,only with the idea of providing breeders with the effectively infinite number ofmapped neutral markers that they had always needed. We revelled in this massivelyexpensive job, funded by a long-suffering European wheat breeding industry, ofcreating the first map with a marker technology so unwieldy that students todaywould not touch it with a bargepole, let alone plant breeders. This was, of course,before the advent of PCR, which changed everything.

The science has moved quickly and the past 20 years have seen staggeringadvances as genetics segued into genomics. We have seen a proliferation of maps,first in the major staples and later in other crops, including ’orphan’ species grownonly in developing countries. The early maps, populated with isozyme markersand RFLPs, were soon enhanced with more amenable PCR-based microsatellites,which are now beginning to give way to single nucleotide polymorphisms. Thesemaps and markers have been used, in turn, to massively extend our knowledge ofthe genetic control underlying yield and quality traits. The relatively dense mapshave allowed whole genome scans which have uncovered all regions of the genomeinvolved in the control of key adaptive traits in almost all agricultural crops of anysignificance.

More amazing is the fact that we now have the whole genome DNA sequences ofnot one but four different plant genomes - Arabidopsis, rice and poplar and sorghum.Moreover cassava, cotton, and even maize could be added to the list before thesevolumes are published. Other model genomes where sequencing has been startedinclude Aquilegia (evolutionary equidistant between rice and arabidopsis), Mimulus(for its range of variation) and Brachypodium (a small-genome relative of wheatand barley).

Two other components deserve mention. The first is synteny, the tendency forgene content and gene order to be conserved over quite distantly related genomes.Ironically, synteny emerged from comparisons between early RFLP maps andprobably would not have been observed until we had long genomic sequences tocompare had we started with PCR-based markers that require perfect DNA primersequence match. The possibility of being able to predict using genetic informationand DNA sequence gained in quite distantly related species has had a remarkableunifying effect on the research community. Ten years ago you could work away atyour own favourite crop without ever talking to researchers and breeders elsewhere.Not so today. Synteny dictates that genome researchers are part of one single globalcommunity.

The second component is the crop species and comparative databases that we alluse on a daily basis. The selfless curators, that we have all taken for granted, deservemention and ovation here because, while the rest of us have been having fun in thelab, they have been quietly collecting and collating all relevant information for usto access at the press of a button. This is a welcome opportunity to acknowledgethese unsung heroes, and of course, their sponsors.

The practical application of markers and genomics to crop improvement has beenmuch slower to emerge. While endopeptidase and the glutenin gels continue to see

FOREWORD TO THE SERIES ix

use in wheat breeding, marker-aided selection (MAS) using DNA markers has, inboth public breeding and the multinationals, emerged only in the last few yearsand examples of new varieties that have been bred using MAS are still few and farbetween. This will change, however, as the cost of marker data points continuesto plummet and the application of high-throughput methods moves the technologyfrom breeding laboratories to more competitive outsourced service providers.

The post-RFLP period and the new opportunities for deployment of economicalhigh-throughput markers are the subjects of these volumes. The first volume dealswith platforms and approaches while the second covers selected applications in arange of crop plants. The editors, Rajeev Varshney and Roberto Tuberosa, are tobe congratulated on bringing together an authorship of today’s international leadersin crop plant genomics.

The end game, where plant breeders can assemble whole genomes by manipu-lating recombination and selecting for specific alleles at all key genes for adaptationis still a very long way off. But these two volumes are a unique opportunity to takestock of exactly where we are in this exciting arena, which is poised to revolutioniseplant breeding.

Mike Gale, FRSJohn Innes Foundation Emeritus

FellowJohn Innes Centre

NorwichUnited Kingdom

FOREWORD

According to the World Bank, approximately 1.2 billion people in absolute povertylive on less than US$ 1 per day, while nearly twice this number live on less thanUS$ 2 per day. About 90% of the world’s poor comprise rural, resource-poorfarmers and their families, and landless poor who depend on agriculture for theirlivelihoods. However, poor farmers live and work in regions that continue to beimportant sources of genetic diversity. While poor farmers, in general, cannot use"modern" (high-input) crop varieties, selected for optimal performance within anarrow range of highly managed environmental conditions, the more wealthy oneshave replaced a wide range of traditional crops and varieties with a limited numberof "modern" varieties of major crops.

However, it must be said to the credit of resource-poor farmers that they planttheir own seeds (landraces) and manage their farms in a manner that allows thevarieties to evolve. They select plant types rather than varieties based on their ownobservations and specific needs. They are in a way, responsible for maintaining thegenetic diversity that is essential to the continued evolution and adaptation of plantgenotypes.

They are therefore well placed to supply formal plant breeding systems withnew genetic material that is urgently needed. Their empirical knowledge of thecharacteristics of specific plant types could help breeders identify the source ofvaluable traits for introgression into elite crop varieties. In this context, plantbreeders need to screen germplasm from regions of low-resource agriculture butrich in genetic diversity and from seed banks for traits they consider useful, and thenfind ways to introgress the desirable traits into the varieties they are developing.

In recent years, genomics, the modern science of genetics, has been providingbreeders with new tools and novel approaches to perform their tasks with highprecision and efficiency. For example, applications of molecular markers in breedingthrough marker-assisted selection (MAS) have already been demonstrated in severalcrop species to develop improved varieties with better agronomic traits and enhancedresistance or tolerance to biotic or abiotic stresses. Indeed, MAS is being usedwidely both in developed and developing countries, and is enabling breeders tomake use of unconventional plant materials.

Tremendous progress has been made in genome science in recent years. Forexample, the complete genomes for several plant species e.g., rice and sorghum,have already been sequenced, and similar efforts are underway for many other crop

xi

xii FOREWORD

species. Comparative and functional genomics approaches are helping scientiststo better understand gene functions and to more effectively tailor the desirablegenotype so that improved varieties can be released in a more timely fashion tofarmers. If these new varieties prove to be appropriate for resource-poor farmers,then farmers could use them to enhance their livelihood security. This is probablythe most obvious way to demonstrate how genomics empowers poor farmers.

This volume deals with applications of genomics towards crop improvement.I am glad to note the variety of opinions and experiences that the editors of thisvolume have gathered - eminent scientists representing international agriculturecenters, advanced research institutes and national agricultural research systems fromseveral countries - providing a treasure trove of information.

I am sure the book will be useful for the community involved in applyinggenomics research for crop improvement as well as for teachers and students toenhance their knowledge in the latest tools and approaches to genomics.

William D. DarDirector General

International Crops Research Institute for the Semi-Arid Tropics (ICRISAT)

Patancheru, Andhra PradeshIndia

PREFACE

Genomics, dealing with the collection and characterization of genes and analysisof the relationships between gene activity and cell function, is a rapidlyevolving, interdisciplinary field of study aimed at understanding and exploitingthe biological information encoded in DNA. The genomics toolbox includesautomated genetic and physical mapping, DNA sequencing, bioinformatics softwareand databases, transcriptomics, proteomics, metabolomics, and high-throughputprofiling approaches. Indeed, the past two decades have witnessed spectacularadvances in genomics. For example, at the dawn of the genomics era, Arabidopsiswas chosen as the first model genome for sequencing, which was then quicklyfollowed by the sequencing of other model genomes (rice for monocots, Medicagoand Lotus for legume crops and poplar for tree species) and crop species (soybean,cassava, sorghum, etc.). While new crops (e.g. maize, wheat, finger millet, etc.)are being added to the list for sequencing the genome or gene space, the generatedsequence data are being analyzed in parallel for characterizing the genes andvalidating their functions through comparative and functional genomics approachesincluding bioinformatics, transcriptomics, and genetical genomics. Candidate genesare becoming increasingly useful for the development of markers for assayingand understanding functional diversity, association studies, allele mining, and mostimportantly, marker-assisted selection. Therefore, genomics research has greatpotential to revolutionize the discipline of plant breeding in order to face thechallenges posed by feeding an ever-growing human population expected to top10 billion by 2050, while decreasing the environmental footprint of agriculture andpreserving the remaining biodiversity.

Several high-throughput approaches, genomics platforms, and strategies arecurrently available for applying genomics to crop breeding. However, the high costsinvested in, and associated with, genomics research currently limit the implemen-tation of genomics-assisted crop improvement, especially for autogamous and/orminor and orphan crops. This book presents a number of articles illustrating differentcontributions which genomics can offer to unravel the path from genes to phenotypesand vice versa, and how this knowledge can help to improve crops’ performanceand reduce the impact of agriculture on the environment. Each article shows howstructural and/or functional genomics can improve our capacity to unveil and deploynatural and artificial allelic variation for the benefit of plant breeders. Volume 1,entitled “Genomics Approaches and Platforms”, presents state-of-the-art genomic

xiii

xiv PREFACE

resources and platforms and also describes the strategies and approaches for effec-tively exploiting genomics research for crop improvement. Volume 2, entitled“Genomics Applications in Crops”, presents a number of case studies of importantcrop and plant species that summarize both the achievements and limitations ofgenomics research for crop improvement.

More than 90 authors, representing 16 countries from five continentshave contributed 16 chapters for Volume I and 18 chapters for Volume II(see Appendix I). The editors are grateful to all the authors, who not only provideda timely review of the published research work in their area of expertise but alsoshared their unpublished results to offer an updated view. We also appreciate theircooperation in meeting the deadlines, revising the manuscripts, and in checkingthe galley proofs. While editing this book, we received strong support from manyreviewers (see Appendix II) who provided useful suggestions for improving themanuscripts. We would like to thank our colleagues and research scholars, especiallyYogendra, Rachit, Mahender, Priti, and Spurthi working at ICRISAT for their helpin various ways. Nevertheless, we take responsibility for any errors that might havecrept in inadvertently during the editorial work.

The cooperation and encouragement received from Jacco Flipsen and NoelineGibson of Springer during various stages of the development and completion ofthis project, together with the help of Rajeshwari Pal of Integra Software Servicesfor typesetting and correcting the galley proofs, have been instrumental for thecompletion of this book and are gratefully acknowledged. We also recognize thatour editorial work took away precious time that we should have spent with ourrespective families. RKV acknowledges the help and support of his wife, Monikaand son, Prakhar (Kutkut) who allowed their time to be taken away to fulfill RKV’seditorial responsibilities in addition to research and other administrative duties atICRISAT. Similarly, RT is grateful to his wife Kay for her precious help in editingand proof-reading a number of manuscripts.

We hope that our efforts will help those working in crop genomics as wellas conventional plant breeding to better focus their research plans for cropimprovement programs. The book will also help graduate students and teachers todevelop a better understanding of this fundamental aspect of modern plant scienceresearch. Finally, we would appreciate receiving readers’ feedback on the errorsand omissions, if any, as well as their suggestions, so that a future revised andupdated edition, if planned, may prove more useful.

Rajeev K. Varshney Roberto Tuberosa

COLOR PLATES

rym15/rym15 (r) x Rym15/Rym15 (s)

rym15/Rym15 (s) Selfing

Rym15/Rym15(s):2Rym15/rym15(s):rym15/rym15(r)Resistance test

x Rym15/Rym15 (s)

rym15/Rym15 (s) Selfing

x Rym15/Rym15 (s)

P

F1

F2

F´1 BC1

F´2

rym15/rym15 (r) x Rym15/Rym15 (s)P

rym15/Rym15 (s)F1 x Rym15/Rym15 (s)

Rym15/Rym15 (s) rym15/Rym15 (s)

x Rym15/Rym15 (s)

Rym15/Rym15 (s) rym15/Rym15 (s)

x Rym15/Rym15 (s)

Rym15/Rym15 (s) rym15 /Rym15 (s)

x Rym15/Rym15 (s)

0

1

2

3

4

5

F´1 BC1

F´´1 BC2

F´´´1 BC3

Year

Conventional backcrossing Marker assisted backcrossing

Rym15/Rym15(s):2Rym15/rym15(s):rym15/rym15(r)

Resistance test

50%

75%

87.5%

93.75%

Recurrent parentgenome (%)

Plate 1. Comparison of conventional and marker-assisted backcrossing programmes for the incorporationof rym15 (from Ordon et al. 1999, mod.) (See Fig. 1, on page 88)

x

xx

Pry

m9

Rym

9ry

m9

Rym

9ry

m4

rym

4

rym

4ry

m4

Rym

4

Rym

4

Rym

4

Rym

4

rym

11

rym

11

Rym

11

Rym

11

DH

-lin

e p

rod

uct

ion

F1

rym

9R

ym9

rym

4ry

m4

Rym

4R

ym4

rym

11 R

ym11

DH

-lin

e p

rod

uct

ion

rym

9ry

m9

rym

9

Rym

9R

ym9

Rym

9

rym

4ry

m4

rym

4

rym

4ry

m4

rym

4ry

m11

rym

11

rym

11

Rym

11 R

ym11

Rym

11 mar

ker

sele

ctio

n

rym

9

rym

9

rym

9

rym

9

Rym

9

Rym

9

rym

4ry

m4

rym

4

rym

4ry

m4

rym

4

rym

11

rym

11

rym

11

rym

11

Rym

11

Rym

11

Rym

9

Rym

9

rym

4

rym

4

Rym

11

Rym

11

P´

F1´

A1´

mar

ker

sele

ctio

n

rym

9ry

m9

Rym

9R

ym9

rym

4ry

m4

rym

4ry

m4

Rym

4R

ym4

Rym

4R

ym4

rym

11ry

m11

Rym

11 R

ym11

mar

ker

sele

ctio

n

rym

9ry

m4

rym

4R

ym9

Rym

4ry

m9

Rym

4R

ym9

Rym

4 R

ym11

Rym

4ry

m11

rym

4 R

ym11

A1

rym

4ry

m11

rym

4 an

d r

ym11

fixe

dry

m4

and

rym

9fi

xed

rym

4,ry

m9

and

rym

11fi

xed

rym

Rym

= r

esis

tanc

e en

codi

ng a

llele

;=

sus

cept

ibili

ty e

ncod

ing

alle

le

Pla

te2.

Sche

me

ofpy

ram

idin

gre

sist

ance

gene

sry

m4,

rym

9an

dry

m11

bytw

oha

ploi

dyst

eps

(Wer

ner

etal

.200

5)(S

eeF

ig.

2,on

page

91)

markerselection

HVM 67 (rym9)

rym4

rym4 rym4

rym4 rym4 rym4

rym4

rym4

rym4 rym4

rym4 rym4

rym4

Rym4

rym4 rym4

rym4

rym4rym4

Rym4

Rym4 Rym4

Rym4

Rym4

Rym4 Rym4

Rym4 Rym4 Rym4

Rym4

rym4

Rym4

Rym4

Rym4 Rym4

Rym4

rym9 rym9

rym9 rym9

rym9 rym9

rym9rym9

Rym9

Rym9

Rym9

Rym9

Rym9

Rym9

Rym9 Rym9

Rym9

Rym9 Rym9

Rym9

Rym9 Rym9

Rym9

Rym9Rym9

Rym9

Rym9

Rym9

Rym9 Rym9

Rym9

Rym9

Rym9

Rym9

Rym9

Rym9

Rym9

rym11

rym11 rym11 rym11 rym11

rym11rym11

rym11

rym11rym11

Rym11

Rym4 rym9

Rym11

Rym4

Rym11

rym4

Rym11

Rym11 Rym11

Rym11

Rym11

Rym11

Rym11

Rym11

Rym11

Rym11

Rym11

Rym11

Rym11

Rym11

Rym11 Rym11

Rym11

Rym11Rym11

Rym11

Rym11

Rym11

Rym11

Rym11

Rym11 Rym11 Rym11Rym4

Rym9 rym9

DH-line production

rym4rym4 rym4rym9 rym9 Rym9rym11 Rym11

rym4 rym4 rym4

rym4

rym4

rym4

rym4rym9 rym9rym11

rym11

rym11Rym11

Rym11

Rym11

Rym9

Rym9 Rym9

marker selection

A1

xrym4 rym4Rym11Rym11 Rym11Rym9

Rym9

Rym9Rym4 Rym4

rym9

P xrym4

rym4

Rym9

Rym9

Rym4

Rym4

rym9

rym9x

rym4

rym4

Rym11

Rym11

Rym4

Rym4 rym11

rym11

rym11F1 (P´)

F1´

rym4, rym9 andrym11 fixed

rym4 and rym9fixed

rym4 and rym11fixed

Bmac 29 (rym4) HVM 03 (rym11)

Rym9

Rym9

Rym11

Rym11

Rym11

Rym11

Rym9

Rym9

* r=allele of the resistant parent, s=allele of the susceptible parent

r* ss r

s r

rym Rym= resistance encoding allele; = susceptibility encoding allele

Plate 3. Scheme of pyramiding resistance genes rym4, rym9 and rym11 by one haploidy step (Werneret al. 2005) (See Fig. 3, on page 92)

ACC

AGG

CAT

AACAAG

AAT

ACT

ACG

AAG

AC

C

s tae

p er -i

rT

AT

Extra

AG

ACC

AGG

CAT

AACAAG

AAT

ACT

ACG

AAG

AC

C

s tae

p er -i

rT

AT

Extra

C. arabica

ATAG

AC

C

A

ACG

AGC

ACT

AAT

ACC

ATG

AAC

AAGAGG

CCG

stae

per-i

rT

AT

AC

C

A

ACG

AGC

ACT

AAT

ACC

ATG

AAC

AAGAGG

CCG

stae

per-i

rT

C. canephora

1. C. arabica var. HDT2. C. arabica var. BM3. C. liberica4. C. eugenioides5. C. congensis6. C. canephora7. P. bengalensis

12

34

56

7AG

AC CGAA

CAA

GAA

TAC

CAC

GAC

TAG

CAG

GAT

GCC

G

A

B

AG

Extra

Plate 4. Relative abundance of different SSR motifs in genomes of different coffee species as revealedby: A) semi-quantitative Southern hybridization based slot-blot analysis, and B) in-silico sequenceanalysis of >1000 SSR positive clones from C. arabica and C. canephora (robusta) small-insert genomiclibraries. Note almost similar comparable pattern/relative frequencies of different SSRs across coffeespecies and also between the two approaches of evaluation (our unpublished data) (See Fig. 1, on page405)

Pla

te5.

An

inte

grat

edfr

amew

ork

linka

gem

apof

robu

sta

coff

ee(C

.ca

neph

ora)

deve

lope

dus

ing

aps

eudo

-tes

tcro

ssst

rate

gyan

dD

NA

mar

kers

(AFL

P,R

APD

,SS

Rs)

.T

hem

apco

mpr

ises

11m

ajor

and

5m

inor

linka

gegr

oups

,an

dha

sa

tota

lof

374

map

ped

mar

kers

.Pr

efix

esO

p-,

AF-

inth

em

arke

rna

mes

indi

cate

sR

APD

,AFL

Pre

spec

tivel

y,w

hile

CM

-/C

R-/

CaM

-re

pres

ents

the

map

ped

SSR

s.T

hesm

all

grou

pssh

own

belo

wso

me

ofth

em

ajor

CL

Gs,

repr

esen

tth

efl

oatin

gm

arke

rsof

the

resp

ectiv

eC

LG

plac

edin

an‘A

cces

sory

grou

p/m

ap”

(unp

ublis

hed,

Hen

dre

2006

)(S

eeF

ig.

2,on

page

413)

CHAPTER 1

MICROSATELLITE AND SNP MARKERSIN WHEAT BREEDING

MARTIN W. GANAL1�∗ AND MARION S. RÖDER21TraitGenetics GmbH, Am Schwabeplan 1b, D-06466 Gatersleben, Germany2Leibniz Institute of Plant Genetics and Crop Plant Research (IPK), Corrensstr. 3, D-06466Gatersleben, Germany

Abstract: Bread wheat (Triticum aestivum L.) is one of the most important crop plants. Due toits hexaploid nature consisting of three different genomes (A, B and D) and its largegenome size of approximately 15 billion base pairs, it is also one of the most complexcrop genomes. This has rendered the use of molecular markers in wheat genomeanalysis and breeding slow and difficult. Mainly, through the use of microsatellite orSSR (simple sequence repeat) markers, wheat molecular marker analysis has gainedmomentum during the last ten years. The advantage of microsatellite markers is that theydetect an unsurpassed level of polymorphism in this recently polyploidised organismwith a generally low level of sequence variation. Furthermore, a large proportionof the microsatellite markers is genome-specific, thus amplifying a defined singleproduct from one of the three wheat genomes. Currently, 2.000 to 2.500 mappedmicrosatellite markers are available for the wheat genome. With microsatellite markers,the chromosomal position of many relevant breeding traits such as disease resistancegenes and quality traits has been identified and they are increasingly used in marker-assisted selection during wheat breeding. For the future, high expectations are being putinto another marker type that is called single nucleotide polymorphisms (SNPs) sincetheir number in the wheat genome should be much higher and cost-efficient, highlymultiplexed technologies are available for the analysis of SNP markers in plants. SNPmarker development and use are, however, still in their infancy. Based on recent results,we discuss here the advantages and disadvantages of SNPs compared to microsatellitemarkers for future wheat breeding.

1. INTRODUCTION

Hexaploid bread wheat (n = 21) consists of three genomes with a basic chromosomenumber of n = 7 for each genome. The three genomes are derived from Triticumurartu (A-genome), an unknown species from the Sitopsis section (B-genome)

∗Author for correspondence: ganal@traitgenetics.de

1

R.K. Varshney and R. Tuberosa (eds.), Genomics Assisted Crop Improvement:Vol. 2: Genomics Applications in Crops, 1–24.© 2007 Springer.

2 GANAL AND RÖDER

and Triticum tauschii (D-genome) respectively (Feldman and Levy, 2005). Thehexaploid bread wheat genome is one of the largest crop genomes with, all threegenomes combined, approximately 15 billion base pairs (Arumuganathan and Earle,1991) and each gene usually occurring in at least three copies. It is estimated thateach of the three wheat genomes contains approximately 40.000 to 50.000 genes.More than 80% of the DNA of bread wheat consists of repeated DNA sequenceswith transposons and retrotransposons representing the highest proportion (Devoset al., 2005). Intensive genetic analyses showed that synteny perturbations betweenthe three genomes are correlated with recombination rates along chromosomes(Akhunov et al., 2003a,b). Physical mapping through the use of deletion lines ledto the conclusion that the genes are not evenly distributed along the chromosomesbut are predominantly located in telomeric regions (Qi et al., 2004). However,first data on the DNA level obtained through the comparative sequencing of largeDNA regions have shown that the three genomes differ considerably at their DNAsequence level predominantly caused by the presence/absence of retrotransposonsand other repetitive sequences but also through the shuffling of genes within theindividual genomes (Wicker et al., 2001; Appels et al., 2003; Keller et al., 2005).

Over more than fifteen years, molecular markers have been used in wheat genomeanalysis. Initially, with the RFLP technology, first genetic maps of wheat have beenconstructed for all 21 chromosomes. The advantage of the RFLP technology inwheat genome analysis includes usually the detection of one copy of the respectivelocus in each of the three wheat genomes simultaneously so that very large numbersof loci can be assigned to the wheat chromosomes or specific chromosomal regionsthrough the use of deletion lines (Qi et al., 2004). Another advantage of the RFLPanalysis is that through the hybridization of probes from closely related otherGramineae, it was possible to deduce that the wheat genomes and the genomes ofother Gramineae, such as rice, maize, rye and barley share a very similar orderof genes called synteny on the chromosomal level (Van Denze et al., 1995; Galeand Devos, 1998). RFLP analysis has however, only in a few cases been used forsegregation analysis in wheat breeding except for the mapping of genes in widecrosses. This is mainly due to the fact, that the level of RFLP polymorphism isvery low in wheat breeding material. Other PCR-based marker analyses such asmultilocus systems, for example, AFLP have also not been widely used in wheatsince this system also identifies only a limited level of genetic polymorphism inbreeding material and routine analysis is being further complicated through thelarge genome size and the amount of repetitive sequences (Manifesto et al., 2001;Zhou et al., 2002).

2. MICROSATELLITE MARKERS

Only through the advent of microsatellite or SSR markers, the use of molecularmarkers has gained momentum in the last ten years since the first SSR markershave been described for hexaploid wheat (Devos et al., 1995; Röder et al., 1995;Bryan et al., 1997). Microsatellite markers are based on tandemly repeated DNA

MICROSATELLITE AND SNP MARKERS IN WHEAT BREEDING 3

sequences of short repetitive motives (e.g. poly CA, poly CT, poly AT and otherrepeated sequences of 3–5 bases). The variability of microsatellite sequences ina genome is not based on point mutations but on the variation in the number ofthe simple sequence repeats. Such variation occurs approximately ten times morefrequently through processes such as slippage during replication or unequal crossingover (Hancock, 1999). This makes SSRs the most suitable and polymorphic markersystem in species with a low level of polymorphism such as wheat.

2.1. Wheat Microsatellite Markers

The first large set of microsatellite markers for the wheat genome has beenpublished in 1998 (Röder et al., 1998b). With these markers it could be shownthat microsatellite markers have a number of characteristics that make them thecurrently best-suited marker system for the analysis of hexaploid wheat. Wheatmicrosatellite markers detect a much higher level of variability than RFLPs andother marker types (e.g. AFLPs), especially in closely related wheat germplasmand varieties such as they are used for breeding. Approximately 50% of the wheatSSR markers detect only a specific locus on one of the three genomes and thus aregenome-specific. If they amplify from more than one of the three wheat genomes,the amplified fragments are often clearly distinguishable on high resolution gels.SSR markers are multiallelic and detect up to more than 30 different alleles in thewheat germplasm for a given locus which makes them in their information contentsignificantly superior to biallelic marker systems (Plaschke et al., 1995; Röder et al.,2002). The distribution of microsatellite markers along the wheat chromosomesdoes not significantly differ from that genes, which show partial clustering at thephysical end of the chromosomes, although most of the currently used SSR markersare not in genes (Röder et al., 1998a). Finally, wheat microsatellite markers areamenable to high-throughput so that large numbers of markers can be analyzed withlarge numbers of plants. This make them the preferred marker system for markerassisted selection and mapping in wheat and for breeding (Koebner et al., 2001;Koebner and Summers, 2003).

Over the last eight years, large sets of wheat microsatellite markers have beendeveloped from various sources. At Gatersleben, most of the approximately 1.000identified SSR markers are derived from genomic clones generated from single-copy sequences out of hypomethylated regions (Röder et al., 1998b; Pestsova et al.,2000; unpublished results). Other microsatellite markers, such as the wmc markers,are derived from libraries of wheat sequences that were generated through a varietyof enrichment procedures (Song et al., 2002; Somers et al., 2004). Recently, aconsiderable number of wheat microsatellite markers have been identified frommicrosatellite motives identified within EST sequences through bioinformatic datamining (Eujayl et al., 2002; Holton et al., 2002; Gao et al., 2004; Yu et al.,2004; Peng and Lapitan, 2005). In general, the developed wheat SSR markersare of different amplification quality and have a varying number of detectedloci. Nevertheless, it becomes now clear that microsatellite markers generated

4 GANAL AND RÖDER

from microsatellite motives within ESTs are usually more highly conserved andcan be used through a wider range of germplasm (Zhang et al., 2006) but onthe other side are also significantly less polymorphic than markers derived fromgenomic sequences since EST-SSR contain on average less repeating units of themicrosatellite motif (Varshney et al., 2005).

Currently, the mapping of approximately 2.000 to 2.500 SSR markers on the21 wheat chromosomes has been published in a variety of publications (Table 1).The precise number of mapped SSR markers in wheat is at present not easilydeterminable since the published maps show considerable overlap in the used SSRmarkers and frequently in different mapping populations different loci on the threehomeologous chromosomes have been identified. Furthermore, in a number of casesonly the number of detected loci is specifically described. Finally, no comprehensivesequence comparison between the sequences used for marker generation and nomapping data integration has been performed for all markers so that it is possiblethat a considerable number of microsatellites markers that detect the same sequencehas been developed independently in different laboratories, especially in the case ofEST-SSRs mined from public databases. Figure 1 shows a map generated with morethan 1.000 unique SSR markers generating 1169 mapped microsatellite loci whichwas established in collaboration between the IPK and TraitGenetics representingcurrently the genetic map with the largest number of unique wheat SSR markers.

In wheat breeding, SSRs are increasingly being used as the marker backbonefor a variety of purposes. These include the localization of individual genesonto the 21 wheat chromosomes such as, for example, disease resistance genesor genes affecting other agriculturally important traits. A large variety ofpapers has been summarized in a review (Röder et al., 2004) and a website(http://maswheat.ucdavis.edu). Furthermore, wheat microsatellite markers have beenused for the localization of a large set of QTLs (quantitative trait loci) affectingmorphological and agronomically important traits (e.g. Perretant et al., 2000; Börneret al., 2002; Huang et al., 2003; 2004). Other examples for important mappedQTLs are loci affecting resistance against the scab disease caused by Fusarium forwhich several populations have been used (Anderson et al., 2001; Buerstmayr et al.,2002; del Blanco et al., 2003; Bai and Shaner, 2004). For these QTLs, currentlylarge SSR marker assisted selection projects are in progress. Further applicationsfor which wheat microsatellite markers have been used include the characterizationof large numbers of wheat varieties grown in Europe and North America, as wellas the characterization of wheat lines from germplasm collections for the determi-nation of genetic diversity over time (Donini et al., 2000; Christiansen et al., 2002;Huang et al., 2002; Pestsova and Röder, 2002; Röder et al., 2002). Many of theseaspects have recently been published in a number of reviews (Koebner et al., 2001,Koebner and Summers, 2003; Röder et al., 2004). Finally, it is very likely, thatwheat microsatellite markers will in the future also be used as backbone for theanchoring of a physical map of the wheat genome onto the genetic map and withinthe map-based isolation procedure for genes from the wheat genome, as it has beendescribed already in a few cases (Stein et al., 2000; Keller et al., 2005).

Tab

le1.

Publ

ishe

dSS

Rm

aps

inw

heat

with

the

num

ber

ofm

appe

dlo

ci

Lab

-des

igna

tor

ofm

appe

dSS

Rs

(Num

ber

ofm

appe

dSS

Rlo

ci)

Map

ping

popu

latio

nR

emar

ksR

efer

ence

gwm

(279

)O

pata

85×

W79

84(I

TM

I-po

pula

tion)

SSR

sis

olat

edfr

omge

nom

iclib

rari

esof

A,

Ban

dD

geno

mes

Röd

eret

al.(

1998

)

psp

(53)

CS

×sy

nthe

ticpo

pula

tion

SSR

sis

olat

edfr

omge

nom

iclib

rari

esof

A,

Ban

dD

geno

mes

Step

hens

onet

al.(

1998

)

gdm

(46)

ITM

I-po

pula

tion

SSR

sis

olat

edfr

omge

nom

eD

(Ae.

taus

chii

)Pe

stso

vaet

al.(

2000

)

wm

c(6

6)IT

MI-

popu

latio

nSS

Rs

isol

ated

from

enri

ched

geno

mic

libra

ries

ofA

,B

and

Dge

nom

es

Gup

taet

al.(

2002

)

Dup

W(2

2)IT

MI-

popu

latio

nSS

Rs

isol

ated

from

EST

sE

ujay

let

al.(

2002

)cf

d(8

4)C

ourt

ot×

CS

popu

latio

nSS

Rs

isol

ated

from

geno

me

D(A

e.ta

usch

ii)

Guy

omar

c’h

etal

.(20

02)

cnl,

ksum

(149

)IT

MI-

popu

latio

n,ph

ysic

alm

appi

ngSS

Rs

isol

ated

from

EST

sY

uet

al.(

2004

)

(Con

tinu

ed)

Tab

le1.

(Con

tinu

ed)

Lab

-des

igna

tor

ofm

appe

dSS

Rs

(Num

ber

ofm

appe

dSS

Rlo

ci)

Map

ping

popu

latio

nR

emar

ksR

efer

ence

wm

c,gw

m,

gdm

,cf

a,cf

d,ba

rc(1

235)

Con

sens

usm

apSS

Rs

from

vari

ous

sour

ces

Som

ers

etal

.(20

04)

cfa,

cfd,

gpw

,ks

u,gw

m,b

arc

(725

)Ph

ysic

alm

appi

ngon

nulli

-tet

raso

mic

and

dele

tion

lines

SSR

sfr

omva

riou

sso

urce

sSo

urdi

lleet

al.(

2004

)

xx-S

SR(1

01)

ITM

I-po

pula

tion

and

WSp

opan

dL

Hpo

pSS

Rs

isol

ated

from

EST

sG

aoet

al.(

2004

)

cwem

(48

asPC

R-b

ased

mar

kers

ofto

tal

of26

6eS

SRlo

ci)

Phys

ical

map

ping

onnu

lli-t

etra

som

ican

dde

letio

nlin

es

SSR

sis

olat

edfr

omE

STs

Peng

and

Lap

itan

(200

5)

barc

(347

)IT

MI-

popu

latio

n,ph

ysic

alm

appi

ngSS

Rs

isol

ated

from

shea

red

geno

mic

libra

ries

ofA

,B

and

Dge

nom

es

Song

etal

.(20

05)

hbg,

hbe,

hbd

(250

)K

itam

oe×

Mün

ster

tale

r,in

tras

peci

fic

DH

-pop

ulat

ion

SSR

sis

olat

edfr

omge

nom

iclib

rari

esan

dE

STs

Tor

ada

etal

.(20

06)

MICROSATELLITE AND SNP MARKERS IN WHEAT BREEDING 7

Xgdm0033a1.6Xgwm1223a4.1Xbcd14342.2Xksud14d5.5Xmwg938a4.5Xgwm47054.4Xgwm0033c3.4Xgwm16406.4Xmwg837a4.4Xgwm01061.9Xabc156b9.4Xgwm4583

14.0

Xgwm07892.5Xgwm03370.0Xgwm06031.5Xgwm16753.4Xgwm12916.3Xgwm4603a3.1Xgwm46651.3Xgwm08482.7Xgwm0903b0.0Xgwm0934b0.7Xgwm04580.0Xgwm10490.0Xgwm31650.9Xgwm04581.8Xgdm0019b2.7Xgwm48810.6Xgwm47801.3Xgwm48100.0Xgwm45992.1Xgwm484320.9Xcdo312a5.8Xgwm4857a5.5Xgwm48672.5Xgwm47953.6Xgwm06422.3Xksug553.1Xksug25.2Xgwm08205.6Xbcd1930a11.8Xgwm10122.3Xcmwg6955.4XAdh3.3Xgdm0126a6.5XATPasea5.2Xgwm30369.9Xgwm48530.9Xgwm44427.2Xgwm09571.3Xbcd1261a1.4Xgwm10414.9Xgwm023210.6Xgwm49130.6Xgwm46952.9Xgdm01111.4Xgwm12026.4Xksue11e5.1Xksui27b4.1Xwg241b3.9Xgwm07932.6Xgwm48780.0Xgwm4612

Xgdm0033b0.8Xgwm1223b2.2Xgwm46283.6Xksud14e5.9Xgwm01361.2Xgwm09051.6Xksud14b8.2Xgwm0033a3.1Xcdo4263.0Xgwm11042.4Xgwm30834.2Xgwm42556.6Xabc156c3.9Xgwm30940.8Xksue18D0.7Xgwm17124.3Xgwm10978.4Xgwm06912.8Xgwm14570.0Xgwm07521.3Xgwm30850.7Xcdo5806.5Xgwm15903.1Xbcd10721.2Xgwm01640.3Xgwm11480.3Xgwm15262.2Xgwm11112.2Xgwm07911.9Xgwm40070.5Xgwm4699a1.1Xgwm4090a1.6Xgwm40491.7Xgwm40675.4Xgwm07782.2Xgwm03576.2Xgwm01352.7Xcdo312b3.8Xmwg5516.6Xksuh9d7.6Xgwm063311.5Xgwm0497a23.9Xgwm40065.9Xcdo11604.7Xmwg6322.6Xgwm009914.5Xmwg912b3.8Xcdo3932.0Xgwm1139a2.7Xwg241a2.6Xksue11b12.4Xgwm0750

Xksud14a1.7Xgwm11304.6Xgwm05503.3Xgwm10784.9Xgwm41441.6Xgwm0033b1.6Xgwm0264a3.3Xgwm0911a6.8Xmwg938b2.0Xgwm30351.9Xabc156a0.0Xgwm11002.1Xgwm10282.4Xgwm44352.7Xksue18Hb4.4Xpsr6885.0Xbcd11244.1Xgwm07621.0Xgwm09262.0Xgwm00184.4Xgwm04982.9Xbcd123.1Xgwm10500.0Xgwm0903a0.7Xgwm0934a0.0Xgwm11202.9Xcdo11731.3Xgwm13690.0Xgdm0028b0.0Xgwm13750.6Xgwm09474.1Xgwm00110.8Xgwm0582b2.9Xgwm04130.0Xgwm02731.4Xgwm41584.0Xgwm4202b0.8Xgwm4085a2.9Xgwm4480a7.4Xgwm07595.5Xgwm40892.7Xgwm41061.6Xgwm0131a4.7Xgwm07842.4Xcdo63710.3Xbcd1150a13.6Xgwm040311.0Xgwm15216.2Xgwm08061.5Xbcd4428.4Xgwm40962.3Xbcd3041.6Xgdm0126b2.1Xgwm01532.1Xgwm0274a3.2Xgwm01240.0Xgwm02682.7Xksug344.6Xgwm158321.2Xbcd15626.8Xcdo118911.7Xcdo346b6.2Xgwm02592.8Xgwm08183.5Xgwm1139b1.9Xgwm13640.6Xgwm31660.6Xgwm15410.0Xgwm06592.8Xksui27a5.4Xgwm01404.4Xgwm1719

1A 1B 1D

Figure 1. Molecular linkage map of wheat based on 70 recombinant inbred lines derived of thecross Opata × W-7984, the so-called ITMI-population (International Triticeae Mapping Initiative). Themicrosatellite loci carrying the lab designators “gwm” (Gatersleben wheat microsatellite) or “gdm”(Gatersleben D-genome microsatellite) are placed in a framework of previously mapped RFLP markers

In summary, wheat SSR will still be the tool of choice for wheat genome analysisand wheat breeding over the next several years, especially if the number of usefulmarkers can be increased. Plant breeding companies are increasingly using the

8 GANAL AND RÖDER

Xgwm15392.3

Xgwm11761.5Xgwm0497b3.0Xgwm06363.0Xcdo57c1.1Xgwm17321.6Xgwm08303.2Xgdm0005a1.6Xgwm06141.3Xgwm0296b2.2Xgwm4739b2.2Xgwm4893a0.7Xgwm48493.0Xgwm05125.7Xksud188.2Xgwm07264.2Xgwm03597.4Xgwm0071a11.3Xgwm41326.6Xfba1786.2Xfbb329a2.7Xgwm119820.0Xgwm11152.4Xgwm15082.4Xgwm31010.1Xgwm00951.4Xgwm1052a1.4Xgwm46010.0Xgwm42762.7Xgwm01220.7Xgwm00100.7Xgwm02750.0Xgwm04480.0Xgwm0515a0.0Xgwm04251.0Xcdo13760.8Xgwm03392.6Xgwm13401.7Xcdo12811.6Xgwm0249b2.2Xgwm1011b2.1Xgwm08951.0Xgwm30220.0Xgwm31460.0Xgwm15780.0Xgwm10361.3Xgwm10452.6Xgwm0071b0.0Xgwm05581.4Xgwm42040.0Xgwm46202.9Xgwm4645b6.3Xgwm04736.3Xgwm03727.0Xgwm03288.6Xgwm08172.1Xgwm0912b2.1Xbcd5432.2Xgwm04454.8Xgwm0047b3.5Xksue161.9Xgwm03124.6Xgwm02942.4Xgwm07611.5Xbcd1095b10.9Xgwm45042.8Xgwm0986b1.4Xgwm30841.7Xgwm4328a2.9Xgwm48970.6Xgwm4355b2.4Xgwm1070b3.7Xgwm12566.6Xgwm03563.9Xcdo678a2.9Xgwm0991c1.9Xgwm0739b3.3Xgwm1263i2.1Xgwm0846a0.5Xgwm0311a0.5Xgwm0382a1.5Xksuh16c1.4Xgwm11510.3Xgwm02651.4Xgwm0382b6.0Xksuf11b11.5Xgwm41668.6Xgwm4105a11.1Xgwm4098

Xgwm41425.5

Xgwm11281.7Xfba280a5.9Xbcd18b3.3Xgwm17501.8Xrz694.4Xrz444b4.5Xgwm02571.6Xgwm06820.0Xgwm04291.4Xgwm4202a4.2Xgwm416711.2Xgwm0410b1.7Xmwg9502.5Xgwm01482.5Xtam72a3.1Xgwm14840.7Xcdo405b7.2Xbcd152a1.2Xgwm03740.0Xgwm09722.1Xgwm4249b3.5Xgwm06302.6Xgwm43380.0Xgwm43182.4Xksuf11c0.9Xgwm14700.0Xgwm30540.0Xgwm13940.0Xgwm11770.9Xgwm03190.0Xgwm0055c3.9Xgwm129b6.0Xwg9962.7Xgwm0055a2.0Xgwm14742.7Xgwm45745.6Xbcd11193.6Xgwm03883.8Xfbb3353.4Xgwm42161.5Xgwm43581.4Xgwm01201.4Xcdo6840.8Xgwm0912a3.6Xbcd17791.5Xgwm12491.5Xgdm01142.0Xgwm4837b0.6Xgwm0016c2.0Xgwm05013.2Xgwm0047a3.1Xfba062a1.2Xgwm0877a0.0Xgwm14130.8Xgwm30140.8Xgwm10676.8Xbcd307a10.0Xbcd1095a5.2Xgwm41222.6Xgwm49472.9Xgwm1300b3.6Xmwg546a8.0Xgwm1070a0.0Xgwm15011.9Xgwm4355a18.7Xgwm0935c3.8Xgwm0940c2.7Xcdo678b2.3Xmwg66013.0Xgwm05268.4Xgwm16911.2Xgwm48200.0Xgwm4888b4.6Xgwm13542.2Xgwm17306.4Xksud23b4.7Xcdo36a_RV2.5Xgwm13993.0Xgwm0382d0.7Xgwm0846d1.8Xgwm06191.8Xgwm30381.7Xgwm10270.0Xgwm13253.0Xgwm41103.4Xgwm42628.1Xgwm4685a4.4Xgwm4828b7.5Xgwm0739a2.0Xgwm12732.1Xbcd1231*X

Xmwg682b10.9

Xgwm4060a3.7Xgwm48264.6Xgwm10990.0Xcdo456d4.2Xbcd18a3.6Xgdm00354.4Xgwm08862.1Xgwm07211.3Xgwm07023.4Xgwm4893b3.2Xgwm47753.7Xgwm4515b2.8Xgwm0210b4.8Xgdm0005b3.4Xbcd102a3.5Xgwm04554.5Xgwm0296a0.9Xgwm4515a0.0Xgwm4739a3.9Xgwm02612.7Xgwm14180.7Xgwm48158.4Xgwm48308.9Xcdo13796.2Xgdm01076.2Xgwm08153.1Xgwm04844.0Xbcd2628.4Xgwm4901a1.6Xgwm45622.1Xgwm01025.9Xgwm09881.8Xcdo405a3.7Xgwm47560.0Xgwm49507.2Xgwm4696a0.0Xgwm40291.7Xgwm0515b2.6Xgwm16731.8Xgwm0030a0.0Xgwm0249a1.7Xgwm16280.0Xgwm12740.0Xgdm0019a2.3Xgwm40382.8Xgwm45751.2Xgwm4645a1.5Xgwm42832.6Xgwm45802.6Xgwm40750.0Xgwm48245.9Xgwm08233.5Xbcd1112.8Xgwm47596.0Xgwm14191.9Xgwm01578.6Xgwm05392.9Xtam82.0Xgwm30252.8Xgwm12043.3Xgwm4837a2.4Xgwm0790a1.0Xgwm0608b2.6Xgwm30264.8Xgwm0877b4.0Xfbb1227.3Xfbb0681.3Xgwm492310.2Xgwm12644.8Xcdo10083.6Xgdm00066.0Xgwm4603b3.1Xgwm48636.7Xgwm03497.3Xgwm03014.0Xgdm0087b6.0Xgwm03204.4Xgwm46911.2Xgwm4828a0.0Xgwm4685b1.1Xgwm4888a3.4Xgwm0991a0.0Xgwm12353.4Xgwm0846b2.0Xksuh9c1.3Xgwm11865.0Xcdo36b_RV7.3Xksuh16a10.4Xgwm0311b0.1Xgwm0382a4.1Xgwm0846c

2A 2B 2D

Figure 1. (Continued)

MICROSATELLITE AND SNP MARKERS IN WHEAT BREEDING 9

Xgwm40246.7

Xgwm4017c3.8Xgwm0757

9.8Xtam61a

6.9Xcdo460a2.9Xgwm03696.9Xbcd14284.2Xgwm30922.7Xcdo3954.4Xgwm0779c6.0Xbcd18231.7Xgwm15072.9Xmwg145.3Xgwm40183.5Xgwm0002b3.5Xcdo6383.5Xpsr903b4.1Xgwm30443.4Xgwm4819a1.6Xgwm48510.0Xgwm4328b1.1Xgwm00051.8Xgwm41610.8Xgwm4922c0.0Xgwm49141.2Xgwm0666b1.2Xgwm0030b0.0Xgwm31431.5Xgwm06743.9Xgwm16203.2Xgwm07201.3Xfba1754.5Xgwm41123.6Xgwm40124.3Xwg1770.2Xgwm11590.2Xgwm14870.2Xgwm10630.2Xgwm11210.2Xgwm11103.5Xgwm10421.7Xmwg9613.2Xgwm063811.0XATPased6.7Xgwm0032

Xbcd907c2.0Xgwm03892.9Xgwm10342.4Xgwm41812.5Xgwm4017a5.2Xgwm0533a4.3Xksug53a4.9Xgwm0779b3.8Xgwm04934.1Xcdo460c11.3Xtam61b

7.4Xgwm1037

6.1Xgwm0533b4.0Xgwm1647

8.7Xgwm1400

4.5 Xgwm13290.0 Xfba091a7.9 Xgwm41450.0 Xgwm4124c4.4 Xgwm00779a2.1 Xgwm16102.3 Xcdo1164b9.3 Xgwm4804b3.5 Xgwm4922b5.5 Xgwm41185.9 Xgwm00722.1 Xgwm05661.6 Xgwm16160.0 Xgwm02840.6 Xpsr903a0.0 Xgwm30871.2 Xgwm08458.3 Xgwm00772.5 Xgwm06851.2 Xcdo3282.2 Xgwm03761.2 Xgwm02852.5 Xgdm0120b4.1 XATPaseb6.1 Xgwm10290.8 Xcdo7182.2 Xgwm08022.2 Xgwm31442.2 Xgwm10151.3 Xgwm0131b4.2 Xgwm4357b7.3 Xgwm4940b5.9 Xgwm415513.7 Xgwm08961.3 Xgwm09384.6 Xmwg69a3.8 Xgwm46067.2 Xgwm44241.2 Xgwm43430.5 Xgwm40223.3 Xgwm08532.5 Xgwm01087.5 Xgwm13553.2 Xgwm40103.9 Xabc174a5.4 Xgwm09801.5 Xgwm07056.5 Xgwm0986a1.8 Xcdo1055.8 Xgwm15646.4 Xgwm06551.5 Xgwm13110.0 Xgwm12661.9 Xbg1315.4 Xgwm02992.0 Xfba2352.8 Xgwm0114b8.4 Xgwm47031.4 Xtam63b2.0 Xgwm05476.3 Xgwm0247

Xgwm40481.2Xgwm46790.0Xgwm4017b1.8Xgwm0071c1.4Xgwm01612.9Xgwm46984.0Xgwm47410.0Xgwm48406.2Xgwm01837.5Xgwm12434.3Xfba091c4.7Xgwm40574.0Xfba2411.7Xgwm48600.6Xgwm48854.0Xgwm0002a5.2Xksua6b16.7Xgwm4812a4.4Xgwm03411.2Xmwg6881.2Xgwm08920.8Xgwm15582.5Xgdm00723.3Xgwm4804a0.5Xgwm4819b0.5Xgwm4922a1.1Xgwm48380.0Xgwm49290.0Xgwm45493.0Xgwm04560.0Xgwm0497c0.0Xgwm00521.5Xgdm00620.8Xgwm13630.0Xgdm01280.8Xgwm14490.8Xgwm15751.4Xgwm07954.9Xgdm00084.7Xbcd1348.5Xgwm14634.1Xabc1762.9Xgwm48251.7Xgwm14163.0Xgwm10473.0Xbcd2885.6Xgwm13055.9Xgwm48002.1Xgwm47676.4Xgwm06452.4Xgwm15721.6Xgwm06642.1Xbcd22a2.3Xbcd1555a3.5Xgwm46800.0Xgwm40715.4Xbcd5155.2Xgwm03832.2Xgwm11600.0Xgwm09771.5Xksuh152.5Xgwm1300a0.8Xgwm07070.8Xgwm03140.7Xbcd3614.4Xgwm41028.4Xgwm43063.8Xfbb2696.0Xgwm12006.1Xgwm00035.4Xgwm48700.6Xgwm43824.9Xgwm1000b2.4Xgwm40562.2Xgwm4148a3.4Xgwm411313.4Xcdo4822.6Xgwm46833.7Xgwm47081.3Xgwm46770.8Xgwm4703b2.4Xmwg11b1.9Xgwm08580.0Xgwm0114a4.2Xgwm09731.7Xgdm00381.7Xgwm10882.1Xabc172a

3A 3B 3D

Figure 1. (Continued)

10 GANAL AND RÖDER

Xgwm30486.8

Xfba078b3.3Xgwm10930.8Xgwm15284.4Xgwm07810.0Xgwm0192c1.8Xgwm0165a4.9Xgwm10911.4Xgwm09291.4Xgwm06012.4Xgwm15160.0Xgwm06951.8Xgwm13180.9Xgwm15312.8Xksuf8a2.8Xcdo13874.0Xgwm00044.3Xfba3203.3Xgwm49490.0Xgwm40261.0Xgwm4350a0.0Xgwm44451.6Xgwm40630.0Xgwm44937.0Xgwm07317.8Xgwm06101.2Xgwm15861.2Xbcd402b1.2Xgwm16808.6Xgwm41359.1Xgwm03972.5Xgwm16271.7Xksug12b2.9Xwg6228.2Xgwm09371.6Xgwm089420.8Xgwm06637.1Xgwm06376.8Xgwm4265b11.6Xmwg549b3.7Xbcd1670a4.6Xgwm10819.0Xgwm43664.3Xcdo475a3.6Xmwg710c2.0Xgwm30390.0Xgwm13503.7Xgwm1179a3.0Xksud93.3Xbcd1302.0Xgwm12512.0Xgwm09593.3Xfba2313.1Xgwm15698.3Xgwm41591.3Xgwm4042a1.6Xgwm4148c2.7Xgwm01601.4Xgwm07420.0Xgwm16943.3Xgwm08323.0Xcdo545a8.2Xgwm08557.5Xgwm4357c8.3Xgwm44238.4Xgwm1258b7.9Xgwm107715.2Xgwm0350b

Xgwm13666.1

Xgwm08883.9Xcdo7954.8Xgwm09250.7Xgwm0940b0.7 Xgwm08980.6Xgwm0935b0.6 Xgwm08560.6Xgwm08912.3Xgwm09100.8Xgwm30720.7Xgwm13202.1 Xgwm07101.8Xgwm0066b0.8 Xgwm01130.7 Xgwm01072.0Xgwm17100.0Xgwm1167a0.0Xgwm09460.0Xgwm08571.4Xbcd12623.1Xgwm40730.0 Xgwm40760.0Xgwm4264b0.7Xgwm4901b0.7Xgwm43600.7Xgwm4357a1.6 Xgwm44652.5Xgwm40821.0 Xgwm44505.0Xgwm03682.8 Xfba008a5.9Xgwm05132.6Xgwm04952.3Xbcd12653.4 Xgwm0192a1.3Xgwm0165c1.0Xfba078a4.8Xcdo14013.4Xcdo1312a3.6Xgwm02514.3Xgwm09980.0 Xgwm01492.5 Xgwm10841.3 Xfbb178a0.9Xgwm46603.7Xgwm40361.5Xgwm46360.0Xgwm42604.8Xgwm09303.1Xgwm0736a4.8Xgwm31483.6 Xgwm0935d4.4Xgwm41402.5Xfbb067d7.6Xgwm05387.6Xfba177a5.4Xgdm0093b4.3 Xbcd402a

Xbcd265b4.3 Xgdm0120a6.1

Xbcd3271.1Xgdm01291.3Xgwm08191.3Xgwm30000.9Xgwm0608a0.7Xksuf8b3.1Xgwm47261.2Xgwm43460.0Xgwm46931.0Xgwm4901c1.6Xgwm44641.6Xgwm46700.5Xgwm45550.5Xgwm48660.0Xgwm41902.3Xgwm0192b1.7Xgwm0165b4.1Xgwm17063.3Xgdm00618.9Xgwm47364.7Xgdm01250.0Xgwm13021.8Xgwm31560.0Xgwm09762.7Xbcd15b6.7Xfbb178b1.1Xgwm45528.7Xgwm471615.9Xgwm400113.2Xbcd1431e17.6Xgwm48864.2Xfbb226e5.2Xgwm01945.2Xgwm11631.6Xgwm13971.5Xfba177b4.2Xgwm40836.3Xcdo9491.7Xgwm06243.5XAmy2.4Xgwm0609

4A 4B 4D

Figure 1. (Continued)

advantages of wheat microsatellite markers during marker-assisted selection andbackcrossing of important traits into elite material and the development of newvarieties (Koebner and Summers, 2003; Powell and Langridge, 2004). Specificexamples for such applications are also described in other chapters of this book.

MICROSATELLITE AND SNP MARKERS IN WHEAT BREEDING 11

Xgwm02412.3Xgwm44202.5Xgwm10827.2Xgwm41604.3Xgwm4861b

14.0

Xgwm01540.8Xgwm0205a3.4Xgwm15741.1Xgdm0109b1.2Xgwm13198.2Xbcd1871b2.1Xgwm40436.1Xgwm10572.9Xgwm31162.8Xgwm41361.5Xgwm4879b0.7Xgwm4350c8.3Xgwm03044.5Xgwm04150.8Xgwm129a1.0Xgwm0293a11.8Xcdo7856.9Xbcd13551.9Xgwm01566.5Xgwm0639b3.5Xgwm01862.5Xgwm11913.1Xgwm1171a4.9Xbcd10884.5Xmwg5223.4Xgwm12363.6Xmwg6241.7Xgwm15916.8Xgwm0617b4.1Xbcd1235b6.4Xgwm49169.9Xbcd1830.8Xgwm0666e19.5

Xrz395b3.7Xgwm15702.4Xgwm49085.8Xgwm13422.7Xcdo1326a10.6Xgwm0982a

16.5

Xabg3918.5

Xwg114a1.4Xgwm01260.7Xgwm01797.9Xgwm0736b5.0Xgwm0595

12.3

Xmwg21125.1

Xgwm09953.1Xgwm02913.3Xgwm0410a

Xgwm02341.5Xbcd873b4.4Xgwm4841

21.4

Xgwm0293c9.2

Xbcd1871a4.1Xgwm12840.8Xgwm16564.9Xgwm10543.4Xgwm05441.3Xgwm0191a1.3Xgwm0066a0.0Xgwm05402.7Xgwm41371.4Xgwm0159b5.6Xgwm41415.3Xgwm4480b2.0Xgwm49001.1Xgwm40450.0Xgwm46591.0Xgwm4746a0.5Xgwm4945a0.3Xgwm00670.4Xgwm40193.5Xgwm0068a1.4Xgwm00671.2Xgwm30880.0Xgwm08100.7Xgwm11802.3Xwg8890.0Xgwm08433.5Xabc1644.7Xgwm02130.0Xgwm03352.0Xgwm11080.0Xgwm11652.7Xtam72c9.1Xbcd11403.6Xgwm03713.4Xgwm08311.1Xgwm04992.1Xabg473b13.7Xgwm0639c10.5Xgwm107310.0Xgwm40278.9Xgwm49053.9Xmwg9143.3Xgwm05541.8Xbcd307b3.3Xgwm10431.6Xgwm14750.8Xbcd92.9Xfba1665.1Xgwm07776.3Xfba3326.1Xcdo5044.6Xgwm04084.6Xgwm16636.0Xgwm06042.3Xcdo1326b5.6Xcdo5842.1Xgwm1246a12.5Xgwm4090b2.7Xgwm1072b12.8Xgwm0790d2.4Xgwm10163.3Xgwm42097.6Xgwm081416.5Xgwm088010.6Xgwm06053.5Xgwm12572.2Xgwm17350.7Xgwm17510.9Xgwm0118

Xgwm12525.8

Xgwm01904.9Xfba393b4.5 Xgwm15590.0 Xgwm13073.9 Xgwm0205b7.6 Xgwm03586.3 Xgdm00033.3 Xgwm0016b2.6 Xgdm00680.0 Xgwm30231.5 Xgwm15541.5 Xgwm15270.0 Xgwm31524.7 Xgwm41522.1 Xgwm0159a1.4 Xgwm47520.0 Xgwm46920.0 Xgwm48230.5 Xgwm4879a0.5 Xgwm4812b2.3 Xfbb238b3.0 Xgwm4945b3.6 Xgwm4746b2.4 Xfba1372.1 Xgwm4812c2.2 Xgwm0911b1.1 Xksud300.7 Xgwm16293.9 Xgwm09606.5 Xmwg561c4.1 Xgwm4811b7.9 Xcdo412b14.2 Xbcd18745.0 Xcdo57b7.3 Xgdm01387.9 Xgdm00991.3 Xgwm10391.1 Xgwm07000.8 Xgwm0583

5.4 Xgwm112210.2 Xgwm0639a5.4 Xgdm00431.5 Xgwm01746.1 Xgdm01532.6 Xgwm01828.0 Xgwm14620.0 Xgwm306310.4 Xgwm0121b1.0 Xgwm0271b9.7 Xbcd450a1.3 Xgwm12530.0 Xgwm02120.0 Xgwm02920.0 Xgwm30461.3 Xgwm31051.5 Xgwm422611.4 Xmwg9226.7 Xgwm08051.8 Xgwm14661.0 Xgwm1246b2.5 Xgwm30993.2 Xgwm48963.6 Xgdm01162.3 Xgwm13242.9 Xbcd11034.4 Xgwm46058.7 Xcdo346a4.0 Xgwm0982c2.1 Xgdm00633.3 Xgwm1072a9.5 Xgwm093111.1 Xgwm48350.0 Xgwm47044.8 Xgwm48227.7 Xbcd1670b13.7 Xgwm4265a5.9 Xbcd14219.0 Xgwm05650.0 Xgwm02691.0 Xgwm14771.0 Xgwm14540.0 Xgwm16600.0 Xgwm06541.4 Xgdm01180.3 Xgwm09020.3 Xbcd1971.0 Xgwm31672.1 Xgwm0272

5A 5B 5D

Figure 1. (Continued)

12 GANAL AND RÖDER

Xgwm4694

12.6

Xbcd1821b1.4Xgwm0719a1.4Xgwm04590.7Xgwm15520.9Xgwm15731.6Xgwm15401.0Xgwm0334a0.0Xcdo476b4.6Xgwm10408.0Xgwm48331.8Xksug48b6.7Xgwm1009a14.2Xksuh4c5.2Xfba085b5.6Xgwm408414.1Xcdo270b4.5 Xcdo13152.5 Xgwm1465b1.4 Xgwm129610.1Xgwm0494

6.3Xcdo14282.3Xgwm1011a1.7Xgwm07860.0Xgwm14580.0Xgwm11850.0Xgwm12930.0Xgwm30291.7Xgwm46080.0Xgwm46750.0Xgwm4501a3.8Xgwm4858a0.0Xgwm49251.4Xgwm41012.2Xfbb1923.1Xgdm0028a2.3Xbcd18600.7Xgwm30023.1Xgwm11501.6Xgwm49151.3Xgwm444312.0Xgwm057015.6Xfba0206.6Xfbb070c5.1Xksud27b2.2Xgwm10892.9Xcdo836b1.9Xgwm42513.3Xgwm04272.5Xgwm0617a7.4Xgwm016917.2Xgwm1017

Xgwm40470.0Xgwm4528b5.1Xcdo476a3.4Xgwm06130.0Xgwm09215.6Xgwm125515.0

Xrz9950.8Xgwm0790b5.2Xgdm01135.6Xgwm05182.6Xgwm08250.0Xgwm07682.9Xksuh4b2.4Xgwm05083.4Xgwm0940a0.0Xgwm0935a5.4Xgwm01932.3Xgwm03614.0Xgwm01334.0Xgwm12330.0Xgwm0644a2.1Xgwm08160.0Xgwm00700.0Xgwm0191c1.4Xbcd13832.5Xgwm06800.8Xgwm07850.8Xbcd14951.6Xgwm00881.9Xgwm4689a1.0Xgwm4124a1.3Xgwm47743.9Xgwm49411.1Xfba3283.3Xksuh14b4.8Xbcd12992.8Xbcd357b4.5Xgwm07712.8Xgwm11998.3Xgwm4858b5.4Xcdo5072.4Xgwm06260.8Xgwm14230.6Xgwm09072.5Xgwm088710.1Xgwm4861a3.5Xwg3416.9Xgwm08891.1Xgwm10762.3Xksug305.6Xgwm021919.3Xgwm148613.8Xfbb082a5.3Xgwm45061.1Xfbb070a2.7Xgwm1328a

Xgwm16620.8Xgwm09043.9Xbcd1821a2.7Xgwm4528a0.0Xgwm48621.0Xgwm44515.6Xgwm13912.5Xgdm01326.0Xgwm04695.4Xksug48a2.1Xgdm01416.9Xgdm00369.4Xfba085a2.3Xgwm15185.3Xfbb2226.7Xgwm30674.3Xgwm47022.3Xgdm01082.3Xgdm01273.1Xgwm1465a3.1Xgwm12412.3Xgwm03250.0Xgwm16300.0Xgwm03252.3Xgwm12683.1Xgwm11660.0Xgwm16540.7Xgwm07741.5Xcdo5343.7Xgdm0014a0.9Xgwm0055b0.9Xgwm47870.0Xgwm4501b0.0Xgwm4264a1.2Xgwm47681.8Xgwm46880.7Xgwm43706.4Xgwm124513.7Xgwm4858c6.1Xgwm0582a2.9Xgwm1167b18.3Xgwm076012.4Xgwm140111.9Xbcd13194.4Xgwm17494.2Xgdm00983.4Xgwm07320.0Xgwm11032.7Xksud27a0.7Xgwm1328b1.7Xgwm30517.8Xgwm15775.6Xfbb070b

6A 6B 6D

Figure 1. (Continued)

3. SINGLE NUCLEOTIDE POLYMORPHISMS OR SNP MARKERS

SNP markers are based on the variation of a specific nucleotide at a given sequenceposition between individuals. Predominantly, such variation occurs as biallelic alter-native bases or as insertion/deletions of individual or small numbers of nucleotides.SNPs have in the last years gained considerable interest due to the fact that they arethe smallest unit of genetic variation and being the basis of most genetic variationbetween individuals, they occur in virtually unlimited numbers. SNPs in codingsequences create furthermore the possibility of changes in the amino acid sequencewithin a protein (if they are not silent) and might have an effect on protein functionand thus monogenic or polygenic traits associated with the expression of suchgenes (Johnson et al., 2001). SNP analysis has been spearheaded in human genome

MICROSATELLITE AND SNP MARKERS IN WHEAT BREEDING 13

Xcdo545b2.7Xgwm4148b1.7Xgwm02331.7Xgwm04713.9Xgwm0350c5.4Xgwm0635b1.0Xgwm0666a5.4Xgwm06813.2Xgwm0735a2.1Xgwm14020.0Xgwm1258c1.6Xgwm1187a14.0Xgwm08348.8Xfba127a3.7Xgwm14102.0Xgwm30647.2Xfbb1863.2Xgwm13068.0Xcdo475b8.1Xgwm01300.0Xgwm006011.4Xabc158

7.0Xgwm1171b

6.6Xgwm47454.8Xgwm4117a3.5Xfba2486.3Xgwm46460.6Xgwm4117b4.5Xgwm4385

13.0

Xgwm15325.0

Xgdm0014c5.9

Xgwm46100.0Xgwm41262.0Xgwm40721.5Xgwm10654.4Xgwm0573b3.4Xgwm16450.0Xgwm09130.0Xgwm08901.9Xbcd10661.0Xgwm02601.0Xgwm17582.1Xgwm13030.0Xgwm08700.0Xgwm10830.0Xgwm11920.0Xgwm06315.4Xgwm4857b2.5Xgwm46045.4Xcdo9629.0Xgwm07481.0Xksug12a11.3Xgwm02768.4Xgwm4544a13.4Xgwm45070.6Xgwm4782b0.7Xgwm4544b1.7Xfba09711.8Xgwm12073.1Xgwm0746a0.3Xgwm02821.3Xgwm03322.4Xgwm00635.0Xgwm16520.7Xgwm0984b2.8Xgwm31592.1Xwg3807.5Xgwm09424.3Xgwm313419.5Xgwm06988.6Xgwm4526b1.2Xgwm4020a2.2Xgwm10660.7Xgwm10610.9Xgwm30583.7Xksuh9b6.6Xgwm0861b

Xgwm42891.6Xgwm42543.5Xgwm09610.0Xgwm06612.6Xgwm15483.5Xgwm135623.9

Xgwm02635.7

Xgwm0569a0.0Xgwm0263r0.0Xgwm025511.6Xgwm1614

10.8Xfba042b3.4Xgwm0951

7.9Xgwm04003.2Xgwm0537

19.9

Xgwm0573a3.9Xgwm1184

6.3Xgwm30190.0Xwg1803.5Xbcd3107.8Xfbb1505.7Xgwm11730.0Xgwm00463.3Xbcd13383.4Xgwm0016a0.0Xgwm02970.1Xgwm00433.0Xgwm08710.0Xgwm16341.0Xgwm0644b1.3Xgwm15111.3Xgwm14501.3Xgwm08972.2Xgwm03332.4Xbcd1782.5Xgwm10853.3Xgwm14711.8Xbcd3492.2Xcdo5513.8Xgwm49300.6Xgwm46322.1Xgwm4124b4.6Xgwm09639.0Xgwm40352.7Xwg5143.6Xgwm09830.0Xgwm080816.1Xgwm0274b10.2Xcdo6866.3Xgwm03023.2Xgwm405913.4Xgwm4782a2.4Xgwm14982.3Xgwm07675.5Xgwm4940a2.7Xgwm13142.6Xgwm4249a6.5Xgwm45614.5Xgwm15570.7Xgwm153314.0Xksud2a7.0Xgwm114425.0Xgwm11756.1Xfbb189a1.5Xgwm4020b0.0Xgwm40464.1Xgwm46111.4Xgwm49114.3Xgwm47832.3Xgwm06114.0Xcdo4141.9Xgwm05770.9Xgwm10250.9Xgwm07450.6Xgwm0984a1.2Xgwm31193.4Xrz5082.6Xgwm31322.2Xgwm0861a4.6Xgwm146b6.7Xgwm30330.7Xgwm09360.7Xgwm03440.0Xgwm13890.7Xgwm30564.7Xgwm08833.0Xgwm07836.5Xgwm126714.1Xgwm0322

Xgwm0735b6.7

Xbcd19754.3 Xgwm1258a6.8

Xgwm4060b2.7Xgwm44833.0Xgwm31003.2Xgwm11233.8Xgwm12500.0Xgwm1000a1.6Xgwm10552.2Xgwm4042b0.0Xgwm0635a6.3Xgdm00865.7Xgwm1187b3.1Xgwm16502.2Xgwm10143.8Xgwm44292.6Xgwm16192.6Xgdm01303.9Xbcd5885.9Xgwm088513.3Xgwm4187a1.8Xgwm46097.5Xgwm12202.5Xwg8342.7Xgwm02953.1Xgwm31020.7Xgwm16720.7Xgwm47210.0Xgwm45822.8Xbcd14385.3Xmwg710b11.4Xgwm13797.9Xgwm43352.3Xgwm30620.0Xgwm10020.7Xgwm135910.3Xgwm13921.5Xgwm00442.0Xgwm41040.0Xgwm4699b2.7Xgwm45560.0Xgwm09740.7Xgwm14817.5Xgwm15871.5Xgwm06762.4Xgwm01111.9Xbcd7079.1Xgwm4811a0.0Xgwm4696b2.1Xgwm31530.0Xgwm15650.0Xgwm10440.0Xgwm15712.1Xwg7190.0Xgwm04370.6Xgwm09280.5Xgwm07801.1Xgwm10070.6Xgwm11540.5Xgdm00841.9Xgwm12423.8Xgwm1052b6.1Xcdo77510.5Xgdm00465.9Xgwm11682.6Xgwm0121a4.4Xgwm13583.1Xgdm00675.9Xfbb07911.1Xmwg9752.4Xgwm12761.2Xgwm13410.0Xgwm13352.5Xgwm0746b3.9Xwg4207.5Xfba204a1.6Xgdm015010.3Xgwm04282.3Xgwm110214.8Xgwm00377.5Xksue3b13.3Xgwm30766.1Xgwm4526a4.7Xfbb189b

7A 7B 7D

Figure 1. (Continued)

14 GANAL AND RÖDER

analysis, where meanwhile more than 2 million SNPs have been identified whichrepresent in their entirety the majority of the genetic variation within the singlecopy and expressed part of the human genome (Sachidanandam et al., 2001).

The interest in SNP markers has further accelerated through the development ofcost-efficient, high-throughput multiplexing (many markers being analyzed simul-taneously) analysis techniques based on chips or other array techniques. Up toseveral million SNPs can now be analyzed simultaneously in the human genomewith costs of a few cents per individual locus. Furthermore, individual SNPs canalso be analyzed through a variety of technologies that permit the determination ofindividual genotypes in an unsurpassed speed and accuracy such as, for examplethrough the use of fluorochrome-based analysis technologies that create data inreal-time (e.g. Taqman, Ampliflour, Invader) and/or in a quantitative fashion (Gut2001, 2004; Kwok, 2001).

Further hopes regarding the use of SNPs in routine analyses have been raisedthrough the identification of haplotypes. Haplotypes are closely linked SNPs whichoccur along a chromosome in clearly defined structures or patterns (alleles) thatextend over hundreds of base pairs or even several kilobases. The haplotype structureof SNPs alleviates the problem of scoring an extremely large number of SNPsby not requiring the analysis of each individual SNP in a genome but only of alimited number for genome coverage. Considerable efforts have been put into theidentification of the haplotype pattern of the human genome (http://hapmap.org)but at present it is still not clear how efficient the analysis of haplotyopes is for theidentification of quantitative traits (Johnson et al., 2001; Foster and Sharp, 2004).

In plants, large scale SNP development and analysis project have been performedpredominantly in diploid crop plants such as maize (Ching et al., 2002), barleyand soybean (Zhu et al., 2003) where meanwhile more than 1.000 genes withSNPs were identified each. These data have demonstrated that SNPs are present inlarge numbers in crop plants and that they share similar features (e.g. presence ashaplotypes) as in other eukaryotic species (Rafalski 2002a,b; Kahl et al., 2005).

3.1. SNP Identification in Wheat

SNP markers are usually identified through the comparative sequencing ofindividual lines or varieties or the bioinformatics analysis of EST data generatedfrom a variety of lines (Rafalski, 2002a,b). With the advent of complete genomesequences for a number of plants such as in the model organisms Arabidopsisthaliana or rice (Oryza sativa), SNPs can be readily identified in basically unlimitednumbers in single-copy DNA once such information is available.

Due to its large genome sequence, complete genome sequencing of the wheatgenome is still at least five years and probably more years away, so that theseresources will not be available in the near future. Furthermore, SNP identificationthrough comparative sequencing is not easily possible in the hexaploid wheatgenome. The main reason is the fact that with a normal primer pair usually the locusis being simultaneously amplified from the three different genomes so that even

MICROSATELLITE AND SNP MARKERS IN WHEAT BREEDING 15

when only a single fragment is identified after PCR amplification, the sequencingof that product is composed of three different sequences which make the analysisof the data at least extremely difficult. In many cases, the sequence is impossibleto determine since one of the three genomes harbours insertions/deletions pushingthe sequence out of frame for that genome. With these problems, direct sequencingis not feasible on a routine basis in wheat and other more complex approaches forSNP identification have to be used.

The most direct way of identifying SNPs in the wheat genome is the bioinformaticmining of wheat ESTs that are available in the respective EST databases. Approx-imately 500.000 ESTs from the wheat genome have been deposited in the ESTdatabases making the wheat genome one of the best-sampled plant genomes (e.g.Lazo et al., 2004). Through concerted efforts of the international wheat community,these ESTs have been sampled from a variety of lines or cultivars so that SNPsbetween these lines could be identified. One of the pitfalls of EST sequence datais that the sequence quality is usually not better than 99% or a Phred-score of20 meaning that on average 1 base out of 100 is incorrect. Thus, comparingindividual sequences to each other for SNP identification is not easily possible andfurther complicated by the presence of highly related sequences from the other twogenomes. This problem can be circumvented by the clustering of all sequence datagenerated from a wheat gene. EST clusters with a sufficiently deep sequence cover(several copies of ESTs from each of the three genomes) permit the identificationof a consensus sequence from each of the three genomes and these sequences canbe compared to similar clusters for the same gene derived from other wheat acces-sions. Since the sequences of the three genomes can be readily discriminated insuch a comparison due to their higher sequence variation, SNPs in the individualcopies derived from the three (A, B and D) genomes can be identified in a quitereliable fashion. Several years ago, an initiative has been started to analyze sequenceclusters of more than 1.000 wheat genes in that way. However, until now besidethe results of a pilot study, no detailed data regarding the outcome of that study andthe level of polymorphism have been published probably due to the complexity ofsuch an approach (Somers et al., 2003).

Another approach towards the identification of SNPs is the reduction of thehexaploid wheat genome into a diploid by the use of genome-specific primers. Theaverage sequence difference between genes in the three individual wheat genomesis in the range of a few percent. Genome-specific primers permit the amplificationof PCR products from only one specific wheat chromosome and thus make compar-ative sequencing of PCR products from individual lines feasible. Furthermore,through its inbreeding nature, such sequences from individual genomes are readilyanalyzable since only one allele should be amplified. Genome-specific primerscan be generated from clustered EST data through the generation of primers withgenome-specific 3’-ends. Since the coding sequences of wheat genes are highlyconserved, frequently one of the two genome-specific primers is being derivedfrom the less-conserved 3’-non-coding end of the respective gene. The pitfall ofthis approach is however, that not all potentially genome-specific primers do in

16 GANAL AND RÖDER

fact amplify a genome-specific product. Usually, several primer pairs have tobe tested for obtaining one good genome-specific PCR product, thus making theapproach quite laborious through the requirement of large-scale bioinformatic dataprocessing and primer testing with a yield of usually less than 50% of functionalgenome-specific primers. One advantage of wheat in this approach is that throughthe use of nullitetrasomic lines, it is quite easy to confirm the genome-specificnature of an amplification product and simultaneously assign this product to aspecific wheat chromosome. Currently, this approach is being used by the wheatHapMap-project (http://wheat.pw.usda.gov/SNP), where large numbers of poten-tially genome-specific primers have been generated and tested for SNP identificationin hexaploid wheat and its diploid and tetraploid ancestors. At TraitGenetics, wehave also used this approach in a pilot study towards the identification of SNPs inhexaploid bread wheat where we have generated more than 200 confirmed genome-specific primer pairs (unpublished results).

Further methods for the identification of SNPs in wheat are also used. For theanalysis of specific candidate genes in a number of lines or varieties with respect tothe occurrence of SNPs, it is possible to add a cloning step after the amplification ofthe respective PCR product from the three different genomes for each line and thendetermine the DNA sequence of a representative number of clones from each line.This approach creates probably the most complete data set for a given gene withoutthe need of generating genome-specific primers but it is not easily amenable to high-throughput SNP detection in large numbers of genes. Due to its low level of SNPs inwheat varieties that will be described below, approaches to identify wheat genes thatcontain SNPs within breeding germplasm will probably gain more interest in thefuture. Such approaches could be, for example, the identification of single featurepolymorphisms (SFPs) through the use of chip-technologies by means of compar-ative analysis of wheat cultivars or the use of heteroduplex analyses techniques(e.g. nuclease treatment or denaturing HPLC) for a first screening (Martins-Lopeset al., 2001). Although these techniques will require sophisticated bioinformatics todiscriminate polymorphisms between the three wheat genomes and SNPs betweenlines, they might in the future provide an important tool for the pre-selection ofgenes with useful SNPs in wheat, however, without alleviating the problem ofgenerating genome-specific PCR products at a later time. Another way of identi-fying SNPs in genome-specific sequences has been the sequencing of amplifiedmicrosatellite-flanking regions. This approach has recently also enabled the identi-fication of a number of SNPs adjacent to microsatellite sequences (Ablett et al.,2006).

3.2. SNP Polymorphism in Wheat

The level of DNA polymorphism in hexaploid bread wheat is quite low whenlooking at fragment length polymorphism via RFLP analysis. Since point mutationsare also the basis of SNPs, it can be expected that the frequency of SNPs is alsolow in hexaploid bread wheat. First data on SNPs have demonstrated that the SNP

MICROSATELLITE AND SNP MARKERS IN WHEAT BREEDING 17

frequency in the hexaploid wheat germplasm is in the order of 1 SNP per 540base pairs (Somers et al., 2003). With that, the SNP frequency in bread wheatis at least five times lower than in maize (Ching et al., 2002; Rafalski, 2002a).Comparable levels of SNP polymorphism are being found, for example, in tomatoand soybean (Zhu et al., 2003) both being crops that have passed through severebottlenecks during domestication and breeding. For wheat, a reason for a low levelof SNPs is also that hexaploid bread wheat is a recently generated polyploid withless than 10.000 years of divergence since its generation. An SNP frequency ofone SNP per 540 base pairs would mean that on average in each sequenced genefragment from an EST approximately one SNP would be located. This is howevernot the case since in many cases SNPs are clustered in 3’-untranslated regions andintrons, so that a considerable amount of sequenced amplicons does not containSNPs. Furthermore, the currently published data are based on germplasm that hasbeen collected from all over the world and also contains highly polymorphic linessuch as Chinese Spring or synthetic wheat lines which have added diversity to thewheat germplasm through the recreation of hexaploid wheat through the artificialhybridization of Triticum turgidum with Aegilops tauschii (Caldwell et al., 2004).

For applications in plant breeding, the used molecular markers have to bepolymorphic in the respective breeding germplasm. European, North American orAustralian wheat germplasm does only contain a fraction of the entire geneticvariation of hexaploid bread wheat. Thus, for the use of SNP markers, it isimportant to determine the actual level of sequence polymorphism in such breedinggermplasm. Since SNPs are usually biallelic markers it is necessary for practicalpurposes that the identified SNPs do show a high allele frequency. The allelefrequency of the minor allele for an SNP should be at least 0.2 or 20% in orderto be useful in actual wheat marker analysis since SNP occurring in only one or afew lines are not generally useful. We have performed a pilot study regarding thelevel of SNP polymorphism in a number of wheat lines predominantly representingwell characterized European germplasm through the comparative sequencing of 202genome-specific amplicons. These data have demonstrated that only 75 (37%) ofthe sequenced genome-specific amplicons showed at least one SNPs in a set of 12wheat lines of which 4 were non-European wheat lines. An analysis of the eightEuropean wheat varieties that were selected to represent the range of the Europeanwheat germplasm based on microsatellite data, demonstrated that approximatelytwo thirds of the identified SNPs were also present in the European wheat breedingmaterial. Many of the identified SNPs were not identified in only one line butin at least 2 lines indicating a relatively high allele frequency of these SNPs inthe wheat germplasm. This is also substantiated by the fact that SNPs that werefound in the four exotic wheat lines were quite frequently also present in the wheatvarieties suggesting that with respect to SNPs at least the European wheat breedingmaterial and varieties do cover a considerable amount of the nucleotide polymor-phism that is found in the entire hexaploid wheat gene pool. The comparativesequencing approach has also shown that in case of more than 1 SNP occurringin an amplicon, these are usually present in the form of well defined haplotypes.

18 GANAL AND RÖDER

In most cases only two haplotypes were observed but in some cases up to fourhaplotypes were observed.

In summary, in only one out of 4 sequenced amplicons, one or more SNPs wereidentified in European wheat germplasm indicating that a large number of genome-specific amplicons need to be investigated in order to cover the wheat genome at areasonable density. For example, if a density of on average one informative SNP per10 centiMorgan interval between two representative European wheat lines shouldbe needed, one would need to have access to a large number of sequenced genefragments. With an estimated polymorphism level of 20% between two averagewheat European wheat lines and a genome size of approximately 3.500 to 4.000centiMorgan for the three genomes together, this would require the availability ofsequence data from nearly 10.000 genome-specific amplicons to have a sufficientlyhigh number of genes with SNPs available.

3.3. Detection Methods for SNPs in Wheat

Another point that is important concerning SNP analysis in wheat is the mode ofdetecting individual SNPs or large numbers of SNPs through multiplexed systems.Currently, there are large numbers of individual systems available that can be usedfor SNP analysis in plants (e.g. Lee et al., 2004; Giancola et al., 2006). With respectto wheat, basically no publications regarding SNP detection on a large scale existso that it is not entirely clear what system would be best-suited. Furthermore, thehexaploid wheat genome will also make SNP analysis quite difficult. Most SNPassay systems are based on the identification of individual SNPs either in individualor multiplexed systems with two major technological approaches for that purpose(Gut 2001, 2004; Kwok, 2001). One approach requires the amplification of therespective locus with a set of specific primers followed by the actual assay by meansof various detection systems. For example, the frequently used TaqMan systemis employing for this a fluorescent dye and a linked quencher, that are separatedfrom each other during the detection process. In Pyrosequencing, a minisequencingprocedure is performed with a detection primer that is a few bases upstream ordownstream of the actual SNP. Primer extension requires an assay primer that isdirectly adjacent to the SNP which will be extended by incorporating a fluorescentlylabelled didesoxynucleotide. The other approach does not require the amplificationof specific fragments via PCR. Such techniques are mostly used in a multiplexfashion in order to detect large numbers of SNPs at many different loci. Examplesfor these are chip-based systems that require the hybridization of genomic DNAonto oligonucleotides which are capable of identifying the two alleles of an SNPthrough specific hybridization to one or the other allele. The oligo-ligation assay(OLA) system catalyzes the allele-specific ligation of two oligonucleotides. Againanother technique that does not require prior amplification is the nuclease-basedInvader-technology.

Both technological approaches will have their problems with regard to SNPanalysis due to the fact that the respective SNPs have to be detected in the wheat

MICROSATELLITE AND SNP MARKERS IN WHEAT BREEDING 19

genome that contains three highly related genomes. Specifically, SNP systems thatdo not require a prior PCR step have the problem that usually all three wheatgenomes directly adjacent to a given SNPs are identical. This means that the analysisprimer will bind to more than one genome. As a result of that a codominant scoringof SNPs will be impossible or require an extremely quantitative system that in theworst case needs to be able to discriminate a 6:0 from a 5:1 and from a 4:2 ratio(6:0 - hybridization to all three genomes with the investigated genome homozygousfor allele 1; 5:1 – hybridization to all three genomes with one genome heterozygous;4:2 – hybridization to all three genomes with one genome homozygous for allele 2).At present, it is not clear whether this can be achieved with sufficient accuracy sothat in most cases SNPs detected with these non-PCR-based systems could only bescored as presence/absence and thus providing only 50% of the actual informationcompared to a codominant marker. SNP detection methods that require a PCRamplification step prior to the actual SNP assay do not face this problem becausefor the amplification of the actual locus, genome-specific primers could be used.In that way only one of the genomes will be assayed and only the three allelicstates also observed in a diploid organisms will be present so that codominantscoring will be possible. In principle, genome-specific primers already exist in caseof SNP identification via comparative sequencing but the amplified fragments forcomparative sequencing are usually large in order to obtain as much sequencinginformation as possible and for SNP analysis fragments should be amplified formost technologies that are less than 100 base pairs since larger fragments resultfrequently in unequal amplification and require the optimization of the amplificationconditions for each individual marker making high-throughput analysis difficult.

4. OUTLOOK TOWARDS THE USE OF SNPSIN WHEAT BREEDING

The future use of SNPs in wheat breeding has to be regarded from differentangles. The general use of SNPs in wheat breeding as a replacement of othermarkers (especially microsatellite markers) is still far away. Based on the firstdata towards SNP identification and SNP frequencies in wheat breeding material,it will need quite some time until a sufficiently large number of SNPs with agood allele frequency will be identified. This will require the coordinated effortsof many different laboratories since, for a reasonable genome coverage a minimumof 10.000 genome-specific amplicons have to be investigated for the presence ofSNPs. Although first efforts have been made to identify SNPs in wheat on a largescale through the wheat HapMap project (http://wheat.pw.usda.gov/SNP), it is verylikely that such an effort will require several years until completion. Furthermore, itis at present not clear what technology could be used for assaying SNPs on a largeand cost efficient scale in hexaploid wheat since all currently available technologieshave problems with SNP analysis in hexaploid species. Thus it is very likely thatover the next five years genome-wide SNP analysis will not be used widely inwheat breeding efforts.

20 GANAL AND RÖDER

In the short term, SNP markers will however gain significantly more importancefor the analysis of individual genes. With established techniques for the generation ofgenome-specific primers and comparative sequencing, SNP analysis with a varietyof techniques that involve a pre-amplification step with genome-specific primerswill certainly be used for the analysis of specific genes that have a known influenceon specific traits or have to be considered as interesting candidate genes. Examplesfor this have already been published such as the SNP analysis of the grain hardinesslocus (Giroux and Morris, 1998; Huang and Röder, 2005), the analysis of storageprotein loci which control aspects of milling and baking quality (Zhang et al., 2003;Ravel et al., 2006) and other genes (Ellis et al., 2002; Yanagisawa et al., 2003;Blake et al., 2004). It is certain that we will see more progress being made in thenear future towards the SNP analysis of such genes in wheat breeding since suchindividual analyses can be performed in a cost-efficient way on large numbers ofindividual plants once such assays have been developed and optimized. The samewill be true for markers tightly linked traits of interest in wheat breeding which areconverted to SNP markers for high-throughput analysis.

ACKNOWLEDGEMENTS

We thank all former and current members of the laboratory at the IPK that wereinvolved in the various aspects of the wheat research. At TraitGenetics, the contri-bution of Dr. Hartmut Luerßen and his staff is acknowledged for the unpublishedwork concerning SNP development. Funding for the SNP work at TraitGeneticswas provided in part through a grant from the BMBF (03i0605).

REFERENCES

Ablett G, Hill H, Henry RJ (2006) Sequence polymorphism discovery in wheat microsatellite flankingregions using pyrophosphate sequencing. Mol Breed 17:281–289

Akhunov ED, Akhunova AR, Linkiewicz AM, Dubcovsky J, Hummel D, Lazo G, Chao S, Anderson OD,David J, Qi L et al (2003a) Synteny perturbations between wheat homeologous chromosomes causedby locus duplications and deletions correlate with recombination rates along chromosome arms. ProcNatl Acad Sci USA 100:10836–10841

Akhunov ED, Goodyear AW, Geng S, Qi LL, Echalier B, Gill BS, Miftahudin, Gustafson JP, Lazo G,Chao S et al (2003b) The organization and rate of evolution of wheat genomes are correlated withrecombination rates along chromosome arms. Genome Res 13:753–763

Anderson JA, Stack RW, Liu S, Waldron BL, Fjeld AD, Coyne C, Moreno-Sevilla B, Fetch JM, SongQJ, Cregan PB, Frohberg RC (2001) DNA markers for Fusarium head blight resistance QTLs in twowheat populations. Theor Appl Genet 102:1164–1168

Appels R, Francki M, Chibbar R (2003) Advances in cereal functional genomics. Funct Integr Genomics3:1–24

Arumuganathan K, Earle ED (1991) Nuclear DNA content of some important plant species. Plant MolBiol Rep 9:208–218

Bai G, Shaner G (2004) Management and resistance in wheat and barley to Fusarium head blight. AnnuRev Phytopathol 42:135–161

Blake NK, Sherman JD, Dvorak J, Talbert LE (2004) Genome-specific primer sets for starch biosynthesisgenes in wheat. Theor Appl Genet 109:1295–1302

MICROSATELLITE AND SNP MARKERS IN WHEAT BREEDING 21

Börner A, Schumann E, Fürste A, Cöster H, Leithold B, Röder M, Weber W (2002) Mapping ofquantitative trait loci determining agronomic important characters in hexaploid wheat (Triticumaestivum L.). Theor Appl Genet 105:921–936

Bryan GJ, Collins AJ, Stephenson P, Orry A, Smith JB, Gale MD (1997) Isolation and characteri