Molecular Genetics of the Narrow-leaf Lupin › files › 3241729 › ... · This thesis contains published work and work prepared for publication. Details of the journals and publications

Contributions to the Molecular Genetics of the

Narrow-leaf Lupin (Lupinus angustifolius L.) -

Mapping, Marker development and QTL analysis.

By:

Jeffrey George Boersma

BSc, MSc

This thesis is presented for the degree of Doctor of Philosophy of

The University of Western Australia

School of Earth and Geographical Sciences

Soil Science and Plant Nutrition

2007

Abstract

ABSTRACT

Narrow-leaf lupin (Lupinus angustifolius L.) was first recorded as having

been introduced into Germany during the mid-19th

century for use as green manuring

and as fodder crops. However, it was not until post World-War I that there was any

serious attempt to domesticate the species. Since that time several key domestication

genes have been incorporated to enable the species to be grown as a crop over a

range of climates, harvested as a bulk commodity and, the seed used for both animal

and human consumption. However, the recent domestication of this species has seen

a rather limited use of wild germplasm – largely as a result of the difficulty in

retaining these key domestication genes. To make the task of retaining these genes

manageable, it was decided to resort to molecular technology.

A mapping population of F8 derived recombinant inbred lines (RILs) has

previously been established by the Department of Agriculture and Food, Western

Australia, from a cross between a domesticated breeding line 83A:476 and a wild

type P27255 in narrow-leaf lupin. The parents together with 89 RILs (of a

population of 115) were subjected to DNA fingerprinting using microsatellite-

anchored fragment length polymorphism (MFLP) to rapidly generate DNA markers

for construction of a linkage map. Five hundred and twenty two unique markers of

which 21% were co-dominant, were generated and mapped. Phenotypic data for the

domestication traits: mollis (soft seeds), leucospermus (white flower and seed

colour); Lentus (reduced pod-shattering), iucundis (low alkaloid), Ku (early

flowering) and moustache pattern on seed coats; were included. Three to 7

molecular markers were identified within 5 cM of each of these domestication genes.

The anthracnose resistance gene Lanr1 was also mapped. Linkage groups were

constructed using MapManager version QTXb20, resulting in 21 linkage groups

consisting of 8 or more markers. The total map length was 1543 cM, with an

average distance of 3.2 cM between adjacent markers. This is the first

comprehensive published map for a lupin species. The map was subsequently used

to identify markers closely linked to the domestication genes suitable for conversion

into forms suitable for marker assisted selection.

Wild types of L. angustifolius require vernalisation to promote flowering.

Modern domesticated cultivars carry the early-flowering gene Ku which removes this

requirement. An MFLP marker was identified as co-segregating with the Ku gene

2

and excised from the gel. DNA sequencing showed that the marker contained a 7 bp

insertion/deletion polymorphism, as well as a single nucleotide polymorphism

(SNP). A pair of sequence-specific primers was designed and successfully converted

the size polymorphism into a simple PCR based co-dominant marker. This marker is

closely linked to the Ku gene, as it co-segregated with the Ku phenotyping in a

population consisting of 106 RILs.

Extinction of Lupinus angustifolius in the wild during periods of drought may

be avoided by production of seeds with coats that are impermeable to water,

preventing germination of a large percentage of the seed in any given year.

Domesticated cultivars of this species carry the recessive gene mollis, making the

seed coat permeable to water and, in turn promoting good crop establishment in the

year of sowing. A dominant MFLP candidate marker on the map was identified as

being tightly linked to mollis. The marker was excised from the original gel,

amplified by PCR, sequenced and extended beyond the SSR-end of the original

MseI-SSR fragment. Two SNPs were found within this extended sequence. Specific

primers were designed to create a marker 209 bp long. PCR products of these

primers run on a Single Strand Conformation Polymorphism (SSCP) gel resolved in

a co-dominant fashion. This marker will be used in marker assisted selection for

mollis when introgressing wild material into lupin breeding programs.

Wild types of narrow-leaf lupin also have seed pods that shatter upon

maturity, leading to the loss of their seeds before or during the harvest process. Two

recessive genes have been incorporated into domesticated cultivars of this species to

maximize harvest-ability of the produce. One of these genes is called lentus (le).

Two MFLP candidate markers were identified as closely linked to the le gene in the

wild x domestic RIL population. The candidate MFLP markers were isolated from

the gel, re-amplified by PCR, cloned and sequenced. The MFLP polymorphisms

were converted into dominant sequence-specific PCR-based markers. Linkage

analysis by MapManager indicated that one of the markers, LeM1, was 2.6

centiMorgans (cM) and the other, LeM2, was 1.3 cM from the gene, with both being

on the same side. The correlation between the marker genotype and the plant

phenotype for the le gene is 95% for the Australian cultivars, and approximately 36%

on wild types tested. These markers may be useful in marker assisted selection for

the le gene when introgressing wild material into lupin breeding programs.

3

The second reduced-shatter gene is called tardus (ta). The RIL marker

population was phenotyped and the gene mapped onto the previously created linkage

map (above). One of several MFLP candidate markers identified as closely linked to

the ta gene, was isolated from the gel, re-amplified by PCR, cloned and sequenced.

The MFLP polymorphism was converted into a sequence-specific PCR-based co-

dominant marker. Linkage analysis by MapManager indicated that the marker,

TaM1, was approximately 2.8 centiMorgans (cM) from the gene. The correlation

between the ta marker genotype and the plant phenotype was 100% for the 18

Australian cultivars tested, and approximately 45.5% on wild types tested. This

marker may be useful in marker assisted selection for the ta gene when introgressing

wild material into lupin breeding programs.

In this thesis I also report the first quantitative traits loci (QTL) mapped in the

same F8 RIL population of Lupinus angustifolius. Traits mapped were early vigour,

days to flowering, height at maturity and seed size. Twenty-two QTLs were found,

located on 13 linkage groups of the same map, with alleles beneficial to the crop

contributed by both parents. Early vigour was controlled by 8 QTLs on 7 linkage

groups. Time to flowering was controlled by 10 QTLs and the height at maturity

was found to be under the control of 4 QTLs. Seed size was linked to 2 QTLs. A

region linked to the Ku gene that promotes early flowering by removal of the

vernalisation requirement appeared to play a role in all 4 traits. The gene mollis

controlling soft-seededness appeared to also be linked to early vigour and; iucundis

controlling alkaloid production was linked to seed size. Five pairs of QTLs were

found to be involved in epistasis, 2 of these having an effect on early vigour and

another 3 influencing the time to opening of the first florets. Variation explained for

each trait ranged from 28% for seed size, to 88% for days to flowering. We showed

that it was possible to use this data to predict genotypes of superior progeny for these

traits under Mediterranean conditions. QTL regions were compared on a second

published linkage map and regions of conserved synteny with the model legume

Medicago truncatula high-lighted.

The work presented in this thesis demonstrates the importance of tight

linkage between markers and genes of interest. It is especially important when

dealing with genetically diverse material as found in the wild. One of the main

problems faced by molecular scientists is the phenomenon known as linkage

disequilibrium in marker populations caused by either small population size or

4

insufficient opportunity for recombination. This frequently results in the

development of markers with little or no application outside of the population in

which it was developed. Although the relatively small size of the population used in

this study exposes it to such constraints, in this case excellent and valuable results

were achieved in developing useful markers to at least 3 of the domestication traits

within a relatively short time period of less then 4 years.

v

ACKNOWLEDGMENTS

I wish to express my thanks to my supervisors, Professor Krishnapillai

Sivasithamparam (The University of Western Australia) and Dr Huaan Yang

(Department of Agriculture and Food, Western Australia (DAFWA)) for their

unwavering support and guidance over the course of this PhD research project.

Thanks are also due to the many staff at the DAFWA who encouraged my efforts,

gave advice, and contributed to my research more directly. Chief among these are

Dr Chengdao Li, Dr Mark Sweetingham, Dr Bevan Buirchell, Dr Mingpei You,

Colin Smith, Dr Nicolyn Short, Daniel Renshaw and Dr Jon Clements. Also

included is Dr Matthew Nelson from Plant Sciences at the University of Western

Australia. My thanks also to the staff manning the School of Earth and Geographical

Science offices who helped keep me on track in various ways. I also wish to

acknowledge collaboration with staff and students of the Institute of Plant Genetics

Polish Academy of Sciences, Laboratory of Structural Genomics in Poznań, Poland,

who contributed directly in a small way to my QTL analysis work. Finally, I wish to

thank the members of my family who supported and encouraged my endeavours in

achieving this milestone.

vi

PUBLICATIONS ARISING FROM THIS THESIS

The following articles arising from this thesis have been accepted for publication:

Boersma JG, Pallotta M, Li C, Buirchell BJ, Sivasithamparam K, Yang H

(2005) Construction of a genetic linkage map using MFLP and identification

of molecular markers linked to domestication genes in narrow-leafed lupin

Lupinus angustifolius L.). Cell Mol Biol Lett 10:331-344

Boersma JG, Buirchell BA, Sivasithamparam K, Yang H (2007a) Development

of a sequence-specific PCR marker linked to the Ku gene which removes the

vernalization requirement in narrow-leafed lupin. Plant Breeding 126:306-

309

Boersma JG, Buirchell BA, Sivasithamparam K, Yang H (2007b) Development

of Two Sequence-specific PCR Markers Linked to the le gene that Reduces

Pod Shattering in Narrow-leafed Lupin (Lupinus angustifolius L.). Genet Mol

Biol – In Press

Boersma JG, Buirchell BA, Sivasithamparam K, Yang H (2007c) Development

of a PCR marker tightly linked to mollis, the gene that controls seed

dormancy in Lupinus angustifolius L. Plant Breeding – In Press

The following article has been submitted for publication and is currently undergoing

review:

Boersma JG, Li C, Leśniewska K, Sivasithamparam K, Yang H (2007d)

Identification of quantitative trait loci (QTLs) influencing early vigour,

height, flowering date and seed size and their implications for breeding of

narrow-leafed lupin (Lupinus angustifolius L.). Aust J Agric Res

vii

STATEMENT OF CANDIDATE CONTRIBUTION

This thesis contains published work and work prepared for publication. Details of

the journals and publications involved have been listed on page vi. Each article

comprises one whole chapter of this thesis.

Chapters accepted for publication or submitted for publication are:

Chapter Two: Boersma et al. (2005) Cellular & Molecular Biology Letters.

Chapter Three: Boersma et al. (2007a) Plant Breeding.

Chapter Four: Boersma et al. (2007b) Genetics and Molecular Biology (In-

Press).

Chapter Six: Boersma et al. (2007c) Plant Breeding (In Press).

Chapter Seven: Boersma et al. (2007d) Australian Journal of Agricultural

Research (Under Review).

Except where duly acknowledged, all work presented in this thesis was performed by

the PhD candidate.

Dr Bevan Buirchell from the Department of Agriculture and Food (Western

Australia) supplied the germplasm used in this thesis project. Professor Krishnapillai

Sivasithamparam and Dr Hua’an Yang as PhD supervisors played an active role in

overseeing the project and reviewing manuscripts. Dr Yang advised on laboratory

procedures as necessary.

Chapter 2:

(1) Dr Margaret Pallotta provided significant support in the construction (computing)

of the map (Figure 2.1) from the markers that had been generated. She

contributed only to the computer analysis, estimated to be around 30% of the

presented information relating to the computer phase.

Chapter 7:

(1) Ms Karolina Leśniewska (Institute of Plant Genetics Polish Academy of

Sciences, Laboratory of Structural Genomics, Strzeszyńska 34 60-479, Poznań,

Poland) contributed the marker ‘mtmt_GEN_00024_04_1’ and contributed

viii

significantly to the work associated with testing of this marker, with the

exception of the actual map placement which was performed by the candidate (<

5% of the chapter).

(2) Dr Chengdao Li helped set up the data files for analysis with the program QTL

Network version 2.0 and, performed the analysis of the epistatic interaction

between two genes involved in Early Vigour of lupins (Section 7.3.6) using the

JMP statistic software (<5% of the chapter).

……………………………………… …………………………………………..

Jeffrey G Boersma Professor K. Sivasithamparam

Coordinating Supervisor

ix

TABLE OF CONTENTS

Abstract……………………………………………………………………………. i

Acknowledgments……………………………………………………….………… v

Publications arising from this thesis…………………………………………….. vi

Statement of Candidate Contribution………………………………………….. vii

Table of Contents………………………………………………………………… ix

Chapter One: Introduction

1.1 Lupins as a crop…………………………………………………….…… 2

1.2 Agronomy of Lupins………………………………………….…………. 4

1.2.1 The Domestication genes…………………………………………. 5

1.2.1.1 Low seed alkaloids…………………………………………... 5

1.2.1.2 Soft seeds……………………………………………………. 5

1.2.1.3 Non-shattering pods…………………………………………. 6

1.2.1.4 Early flowering……………………………………………… 7

1.2.2 Selection for Domestication genes ………………………………. 8

1.2.3 Disease and Pest Resistance………………………………………. 9

1.2.3.1 Brown Spot / Pleiochaeta root rot…………………………… 9

1.2.3.2 Anthracnose (Colletotrichum gloeosporoides)………..…….. 10

1.2.3.3 Phomopsis (Diaporthe toxica Williamson et al.)…………… 10

1.3 Breeding to improve lupins as a crop…………………………….…….. 12

1.3.1 Past practices………………………………………………………. 12

1.3.2 Wild genetic material……………………………………………… 12

1.3.3 Mutation breeding…………………………………………………. 12

1.3.4 Transgenic breeding……………………………………………….. 13

1.4 Molecular genetics and plant breeding………………………………… 13

1.4.1 Molecular markers………………………………………………… 13

1.4.2 Genetic mapping…………………………………………….…….. 14

1.4.2.1 Mapping and markers……………………………………….. 14

1.4.2.2 Mapping populations………………………………….…….. 15

x

1.4.2.3 Mapping……………………………………………….…….. 17

1.4.2.4 Mapping in lupins…………………………………………… 18

1.4.3 Quantitative trait loci……………………………………………… 19

1.4.4 Marker assisted breeding………………………………………….. 21

1.5 Current needs in lupin breeding……………………………………….. 23

1.6 Aims of this thesis……………………………………………………….. 23

1.7 References……………………………………………………………….. 24

Chapter Two: Construction of a genetic linkage map using MFLP,

and identification of molecular markers linked to domestication

genes in narrow-leafed lupin (Lupinus angustifolius L.)

2.1 Introduction……………………………………………………………… 34

2.2 Materials and methods………………………………………………….. 35

2.2.1 Plant material……………………………………………………… 35

2.2.2 Phenotyping……………………………………………………….. 35

2.2.3 DNA extraction……………………………………………………. 37

2.2.4 MFLP protocol…………………………………………………….. 37

2.2.5 Data analysis and map construction……………………………….. 38

2.2.6 Genome length estimates………………………………………….. 38

2.3 Results……………………………………………………………………. 39

2.3.1 Phenotyping……………………………………………………….. 39

2.3.2 Marker polymorphism and segregation…………………………… 39

2.3.3 Map construction………………………………………………….. 40

2.3.4 Mapping of domestication and anthracnose resistance genes…….. 42

2.3.5 Estimation of genome size of L. angustifolius…………………….. 43

2.4 Discussion………………………………………………………………… 46

2.5 References………………………………………………………………... 48

xi

Chapter Three: Development of a sequence-specific PCR marker

linked to the Ku gene which removes the vernalisation requirements

in narrow-leafed lupin.

3.1 Introduction……………………………………………………………… 54

3.2 Materials and Methods………………………………………………….. 55

3.2.1 Plant materials and phenotyping of Ku gene……………………… 55

3.2.2 Marker development………………………………………………. 55

3.2.3 DNA amplification and sequencing gel electrophoresis………….. 56

3.3 Results……………………………………………………………………. 56

3.3.1 Phenotyping of Ku gene…………………………………………… 56

3.3.2 DNA sequencing of the candidate MFLP marker…………………. 57

3.3.3 Marker KuHM1…………………………………………………… 58

3.4 Discussion………………………………………………………………… 60

3.5 References……………………………………………………………….. 63

Chapter Four: Development of two sequence-specific PCR

markers linked to the le gene that reduces pod shattering in narrow-

leafed lupin (Lupinus angustifolius L.)

4.1 Introduction……………………………………………………………… 66


4.2.1 Plant materials and phenotyping of le gene………………………. 67

4.2.2 Marker development……………………………………………… 68

4.2.3 Testing of converted markers…………………………………….. 68

4.2.4 Confirmation of linkage………………………………………….. 69

4.3 Results……………………………………………………………………. 69

4.3.1 Phenotyping of le gene……………………………………………. 69

4.3.2 DNA sequencing of candidate MFLP markers…………………… 69

4.3.2.1 Marker LeM1………………………………………………... 71

4.3.2.2 Marker LeM2………………………………………………… 72

xii

4.3.3 Confirmation of linkage…………………………………………… 73

4.4 Discussion………………………………………………………………… 77

4.5 References………………………………………………………………... 79

Chapter Five: Development of a sequence-specific PCR marker

linked to the tardus gene that reduces pod shattering in narrow-

leafed lupin (Lupinus angustifolius L.)

5.1 Introduction……………………………………………………………… 82


5.2.1 Plant materials and phenotyping of the ta gene…………………… 82

5.2.1.1 Field observations……………………………………………. 83

5.2.1.2 Laboratory observations……………………………………… 84

5.2.2 Marker development and mapping………………………………... 84

5.2.3 Testing of converted markers……………………………………… 85

5.2.3.1 Confirmation of linkage……………………………………… 85

5.3 Results…………………………………………………………………….. 86

5.3.1 Phenotyping of ta gene……………………………………………. 86

5.3.2 Map placement of ta and closely linked MFLP markers………….. 86

5.3.3 DNA sequencing of candidate MFLP marker…………………….. 87

5.3.3.1 Marker TaM1………………………………………………… 89

5.3.4 Confirmation of linkage…………………………………………… 89

5.4 Discussion…………………………………………………………………. 92

5.5 References………………………………………………………………… 94

xiii

Chapter Six: Development of a PCR marker tightly linked to

mollis, the gene that controls seed dormancy in Lupinus angustifolius

L.

6.1 Introduction……………………………………………………………… 98


6.2.1 Plant materials……………………………………………………... 99

6.2.2 DNA extraction for marker testing………………………………... 99

6.2.3 Phenotyping for mollis…………………………………………… 100

6.2.4 Marker development……………………………………………... 100


6.3 Results…………………………………………………………………… 102

6.3.1 Phenotyping of mollis gene………………………………………. 102

6.3.2 DNA sequencing of the candidate MFLP marker……………….. 102

6.3.3 Co-dominant marker MoA……………………………………….. 104


6.4 Discussion……………………………………………………………….. 108

6.5 References……………………………………………………………….. 110

Chapter Seven: Identification of quantitative trait loci (QTLs)

influencing early vigour, height, flowering date and seed size and

their implications for breeding of narrow-leafed lupin (Lupinus

angustifolius L.)

7.1 Introduction……………………………………………………………... 114

7.2 Materials and Methods…………………………………………………. 115

7.2.1 Plant materials……………………………………………………. 115

7.2.2 Genetic markers………………………………………………….. 116

7.2.3 Plant measurements……………………………………………… 116

7.2.4 Data analysis……………………………………………………... 117

7.3 Results…………………………………………………………………… 118

xiv

7.3.1 Marker mtmtGEN00024041……………………………………... 118

7.3.2 Genetic maps……………………………………………………... 118

7.3.3 Phenotypic variation of traits…………………………………….. 119

7.3.4 Detection of QTLs……………………………………………….. 121

7.3.4.1 Early vigour………………………………………………… 121

7.3.4.2 Height at maturity…………………………………………... 124

7.3.4.3 Days to flowering…………………………………………… 124

7.3.4.4 Seed weight…………………………………………………. 124

7.3.5 Comparison to Nelson et al. and synteny with M. truncatula……. 125

7.3.6 Superior genotype………………………………………………... 126

7.4 Discussion……………………………………………………………….. 127

7.5 References……………………………………………………………….. 131

Chapter Eight: General discussion and future directions

8.1 Introduction…………………………………………………………….. 136

8.2 Lupin breeding…………………………………………………………. 136

8.3 Marker development and marker assisted selection………………… 136

8.4 Molecular markers and mapping……………………………………... 139

8.5 Quantitative trait loci and their analysis……………………………... 140

8.6 Synteny………………………………………………………………….. 141

8.7 Future directions……………………………………………………….. 142

8.8 References………………………………………………………………. 144

Chapter Nine: Appendices

9.1 Anthracnose disease rating of RIL population and comparison

to AnM2 and nearby newly generated map markers………………... 150

9.2 Tardus shatter rating of RIL population and comparison to

nearby newly generated map markers including those of

Nelson et al. 2006……………………………………………………….. 153

xv

9.3 Domestication gene rating of RIL population (according to

parental phenotype) including seed ‘moustache’ but excluding

tardus rating……………………………………………………………. 157

9.4 Enzyme Buffers………………………………………………………… 161

9.5 PCR formulations as used in this thesis………………………………. 163

9.6 DNA extraction for MFLP – High quality but low yield…………….. 164

9.7 Example MFLP procedure (for developing a marker)……………… 165

9.8 Cloning and sequencing a DNA band…………………………………. 173

9.9 How to do mapping with MapManager……………………………….. 179

9.10 Protocol for producing MapChart chart from MapManager……….. 181

9.11 Articles published or In Press as at July 2007………………………... 188

9.12 Raw data for QTL analysis including plant heights for early vigour

and at maturity, seed weight and flowering date……………………..

214


1

Chapter One

Introduction


2

1.1 LUPINS AS A CROP

There are two centres of origin for lupins, the Mediterranean basin („Old

World‟) and the Americas („New World‟). Twelve distinct Old World species have

been described, but those from the New World are numerous and ill-defined

(Gladstones 1998).

Lupins have a long history as crop plants. Two species Lupinus albus and L.

mutabilis have been cultivated as grain legumes for at least 3000 years in the

Mediterranean basin and the South American highlands respectively (von

Sengbusch, 1953). The benefits of lupins in improving soils (and thus enhancing

yields of the following crop) have been known from antiquity, with early Greek

writers such as Theophrastsus, Varro and Columella all highlighting their values

(Gladstones 1970).

In classical Greek and Roman times L. albus was grown for both stock feed

and human consumption, having a reputation for their ability to grow on poor and

roughly cultivated land and improving these soils as a green manure (Gladstones

1970) by incorporating the standing crop into the soil. More recently, in the 1970s,

wild forms of L. angustifolius and L. cosentini were grown in the Perth citrus

orchards for the same purpose.

In general, only large-seeded lupin species have been exploited by humans.

Gladstones (1970) lists 12 species in this category, with only one (L. mutabilis)

originating from South America. Seven species have only recently been

domesticated (Gladstones 1998). The species L. angustifolius, commonly known in

Australia as „narrow-leafed lupin‟ is among them, having retained its wild

characteristics until modern times. It was referred to by the classical authors as a

„wild‟ lupin and more recently in the 18th

and 19th

century as L. varius (Gladstones

1970). One of the common names it was known by in Italy was „lupino salvatico‟

(Bauhin et al. 1651, Savi 1798) suggesting that it may have played an important role

in times of hardship (Gladstones 1970).

The first recorded introduction of narrow-leaf lupins into Germany occurred

during the mid 19th

century at or soon after the successful introduction of L. luteus,

with both crops being cultivated primarily for fodder and green-manuring (Hanelt

1960). Production declined as a result of stock losses due to lupinosis (caused by

Diaporthe toxica) in the 1860‟s and 1870‟s and the increasing availability of cheap


3

nitrogenous fertilisers during the early part of the 20th

century (Gladstones 1970).

Meanwhile, at least one large population was reported to exist in Western Australia

in the Swan valley and Gingin as early as 1880 (Gladstones 1998).

A shortage of protein feeds in Germany during World War I (1914-1918)

stimulated a renewed post-war German interest in lupins (Gladstones 1970). It was in

this time that serious attempts were made to fully domesticate narrow-leaf lupins.

Subsequently, the species became established in such diverse regions as New

Zealand, Australia, South Africa and the USA. With the development of sweet (non-

bitter), soft-seeded cultivars in1928, seed production became the main economic

focus of this crop.

However, even though the plant may no longer be used as a green manure, it

still is of benefit to following crops. Numerous studies in especially Western

Australia have shown the benefits of cropping lupins in rotation with cereals –

especially wheat, and on its own merits. Perry et al. (1998) lists the following

benefits of lupins in southern Australia:

(i) Increases the yield and quality of wheat grown in rotation by providing:

Fixed nitrogen;

A break crop for fungal disease control;

An opportunity of grass weed control;

(ii) Improved soil structure and nutrient cycling;

(iii) Diversified sources of income from sale of high protein seed;

(iv) Seed for use in integrated livestock production (pigs, sheep, cattle);

(v) High value stubbles to complement ruminant livestock production.

He also cites lupins as being responsible for encouraging the adoption of

conservation tillage where cultivation was reduced and stubble retention increased.

In 1991/1992 the narrow-leaf lupin was reported as being grown

commercially in 17 countries world-wide – including Europe, Africa and North

America and the then USSR. Australia was by far the largest producer at 1.1 million

tonnes, followed by Poland at 150 000 tonnes and the USSR at 85 000 tonnes.

Portugal produced just 33 tonnes (Swiecicki & Swiecicki 1995, Cox. 1998). In 2005

commercial lupin production (all species) in Australia was estimated to be 1.05

million tonnes after peaking at 2 million tonnes in 1999/2000, with Western

Australia alone producing more then 1.5 million tonnes (Australian Bureau of


4

Statistics (ABS) data). In 2006 (a drought year) total lupin production in Australia

was a mere 174000 tonnes, with production of lupins in Western Australia at 125000

tonnes – the lowest on record (Australian Bureau of Agricultural and Resource

Economics (ABARE), Feb 2007).

1.2 AGRONOMY OF LUPINS

Wild accessions of L. angustifolius are well adapted for survival in agricultural

environments without the intervention of man. This is facilitated by their

characteristics that:

(i) They have a mechanism to prevent the immediate germination of seeds in the

presence of moisture as may occur in a summer storm, or in years of drought

(hard seed);

(ii) Chemical defences against herbivore and insect attack (alkaloids);

(iii) Mechanisms for synchronising flowering to ensure cross-pollination and time to

fill the seed (vernalisation requirements, photo-period response) and;

(iv) A mechanism for ensuring dispersal of the seed upon maturity (shattering pods).

The above characters (and especially i, ii, iv) are undesirable in a crop because:

(a) Hard seed makes crop establishment difficult if not impossible due to poor

germination;

(b) Alkaloids are toxic to man as well as domestic animals, thus greatly devaluing

lupins as a food source.

(c) The requirement for vernalisation has a negative impact on the adaptability of the

species to warm, short season environments.

(d) A crop should not shed its seed, making it more amenable to mechanized

harvesting.


5

1.2.1 The domestication genes

The Domestication genes are those considered essential to the successful

adaptation of a wild plant type into a crop. Five domestication genes have been

incorporated into narrow-leaf lupins. A brief description and history follows:

1.2.1.1 Low seed alkaloids

Efforts to domesticate the narrow-leaf lupin started to bear fruit in Germany

in 1928/9 (Hanelt 1960, Gladstones 1970) when von Sengbusch isolated the first

natural mutant, low alkaloid plants (von Sengbusch 1931). There appears to be no

knowledge of the mechanism by which alkaloid production is restricted. The trait

was found to be simply inherited as a recessive gene. Another gene for low alkaloids

was found independently in Russia a little later (Fedotov 1932, Hackbarth and Troll

1956, Hackbarth 1957). In all, Gladstones (1970) lists three different (natural) genes

causing reduced levels of alkaloids in this species. These are iucundis ((discovered

by) von Sengbusch, 1928/29), esculentis (von Sengbusch 1928/29) and, depressus

(Fedotov 1931, Troll 1942). He also lists a fourth gene tantalus (Zachow 1967) that

had been created by means of x-ray induced mutation. In his ground-breaking work

in Western Australia (W.A.), Gladstones used the gene iucundis for breeding low-

alkaloid narrow-leaf lupins suitable for the W.A. environment. This gene has been

carried through into all modern Australian cultivars.

Modern cultivars of L. angustifolius contain <200mg alkaloids per kg and are

commonly referred to as „sweet‟ to differentiate them from the wild, high-alkaloid

forms which may contain as much as 40g per kg and are bitter to the taste. Lower

alkaloid levels are achievable but are not necessary and, the cost of controlling insect

pests – especially aphids becomes prohibitive. A typical alkaloid profile of sweet

narrow-leaf lupin consists of lupanine (42-59%), 13-hydroxylupanine (24-45%),

angustofoline (7-15%), α-isolupanine (1-1.5%) and traces of other alkaloids (<1%)

(Harris and Jago 1984, Petterson 1998).

1.2.1.2 Soft seeds

One of the survival adaptations of many plant species including that of the

wild narrow-leafed lupin (Lupinus angustifolius L.) is the production of seeds with

testa (seed coats) that are impermeable to water, inducing dormancy and preventing

full germination of the seed in any one year (Quinlivan 1967, Forbes and Wells


6

1968). This adaptation in lupins is also known as being „hard seeded‟. Dormancy is

a physical process that appears to be induced after maturity as the seed moisture

content drops below 14% and, is complete at between 9 and 11% moisture content

(Gladstones 1958). In the field, seed dormancy of L. cosentinii (formerly known as

L. varius) was found to gradually break down as a result of large daily (summer)

temperature fluctuations (Quinlivan 1966, 1968). Similar observations have also

been made for L. angustifolius. It was surmised that the fluctuating temperatures

caused the testa to crack in localised areas making it permeable, allowing a viable

seed to readily imbibe water and germinate (Quinlivan 1968).

High germination rates and plant densities in the year of sowing are

paramount for a crop species. The recessive gene mollis which is of unknown origin

(Mikolajczyk 1966; Forbes and Wells 1968) results in seeds developing a water

permeable testa at maturity, although the precise mode of action of mollis remained

unclear. Some researchers have found that the seed of soft-seeded lines and cultivars

have testa similar in thickness to hard-seeded lines (Clements et al. 2005), whereas

others found an alteration in the shapes and size of especially the outer palisade and

the hourglass cell layers of lupin seed testa (Miao et al. 2001) and that of other

legumes (Lush and Evans 1980). Recently, Garnczarska et al. (2007) determined

that the main site of water entry into the seed was through the micropyle and hilum,

suggesting that mollis has some control over how one or both these two structures

functions.

Troll first combined mollis with a low alkaloid line from von Sengbusch,

resulting in the release of the cv Müncheberger Blaue Süsslupine II in 1944

(Hackbarth and Troll 1956, Troll 1964). Not surprisingly, mollis has since

domestication of the narrow-leaf lupin been incorporated in all subsequent cultivars

of this species (Gladstones 1960; Cowling 1999).

1.2.1.3 Non-shattering pods

An essential characteristic of a modern grain crop is the ability to retain its

seeds long enough to allow mechanical harvesting at full maturity. In the genus

Lupinus, this characteristic has long been present in two species – Lupinus albus L.

and L. mutabilis (Gladstones 1967). Von Sengbusch and Zimmermann (1937) were

successful in selecting a strain of narrow-leaf lupin that had a reduced pod shatter,


7

apparently related to a large reduction in the thickness of the pod wall (Atebekova

1958), but this character proved to be unstable (Hondelmann 1984).

In 1960 Gladstones discovered two natural mutant (recessive) genes, lentus

and tardus for reduced pod-shattering which, when combined, almost eliminated all

pod shatter. One of the genes for reduced pod-shatter, known as tardus (ta), affects

the sclerenchyma strips of the dorsal and ventral pod seams, fusing the two halves to

such an extent that separation of the two halves is greatly impeded. It was

considered analogous to a gene previously found in L. luteus. The second gene

lentus (le) modified the orientation of the sclerified endocarp of the pod, resulting in

a reduction of torsional forces upon drying, and hence reduced pod shatter. This

gene was also associated with a reduction in the thickness of the pod wall, but not to

the extent of the previously found reduced shatter strain of narrow-leaf lupin of von

Sengbusch and Zimmerman (1937), leading to the conclusion that the controlling

genes were not the same. This modification was associated with a change in internal

pod pigmentation that gave the immature pods a purplish tinge and the inside surface

of mature pods a bright yellowish-brown colour (Gladstones, 1967). The gene le has

also been associated with development of a reddish pigmentation within the stem of

plants older than two months (G. Thomas, per. comm.). The first Australian cultivar

to carry both these genes was Uniharvest, released in 1971. All subsequent

Australian cultivars carry both these genes. Many European cultivars carry at least

one of the two genes.

1.2.1.4 Early flowering

Early work by Gladstones to find a gene for early flowering used as a model

the pasture legume species Trifolium subterraneum. He noted that Morley and Evans

(Evans 1959, Morley and Evans 1959) had discovered that the initiation of flowering

was controlled by three complementary processes: (1) vernalisation, (2) dark period

inhibition or long day requirement, (3) high temperature requirement. It had already

been known for some time that lupins were vernalisable long-day plants – that is, it

was possible to bring forward the date of first flowering (anthesis) by subjecting the

plants to a period of low temperatures (Gladstones 1969).

Gladstones and Hill (1969) reported the discovery of a single dominant gene

(Ku) that effectively removed all requirements for vernalisation in narrow-leaf

lupins, advancing the date of anthesis by 2 – 5 weeks depending on location and


8

sowing date, with plants sown late in colder areas showing the smallest response.

This gene had been detected in a single plant within a commercial crop of the

Swedish cultivar Borre.

In a subsequent study, Rahman and Gladstones (1972) reported that narrow-

leaf lupins responded to both periods of vernalisation and a long (24 hr) photo-

period. They found too that a 4-week vernalisation at 1-2o

C had a significant impact

on flowering in narrow-leaf lupin, reducing the time to flowering from 111.4 to 79.8

d in the cultivar Uniharvest. When the photo-period was increased from 10 h to 24 h

it also had a substantial effect, reducing the days to anthesis by 15.6 d from 53.3 to

37.7 d in a genotype possessing the gene Ku. This is supported by Dracup et al.

(1998) and Adhikari et al. (2004) who found that extending photoperiod to at least 16

h gave a significant reduction in the time to flowering. However, in their experiment,

cultivars having Ku showed a reduction of only 1 to 3 days whereas non-Ku lines

were up to 16 days earlier in reaching anthesis. In Western Australia the maximum

photo-period during April – October varies between 9.25 h in June and July and 11 h

d-1

in August, reaching a maximum of ~ 12.75 h d-1

in October

(http://www.bom.gov.au/climate/dwo/IDCJDW0600.shtml). The target date for anthesis is

early August (Nelson and Delane 1991). Thus, the impact of photoperiod would

most probably be greatly reduced, as in the earlier reported experiments of

Gladstones and Hill (1969).

Gladstones and Hill (1969) concluded that Ku greatly increased the flexibility

of lupins as a crop in that it enables them to be grown in warmer areas with

insufficient cold to satisfy the vernalisation requirements. This gene also would

allow them to be sown in areas with a short growing season as experienced in large

areas of the Western Australian grain-belt. In the last decade it was discovered that

another gene Julius (Mikolajczyk 1966) which is of unknown origin, is identical to

Ku (Cowling et al. 1998).

1.2.2 Selection for domestication genes

Until now, the only way to select for the domestication genes in lupins has

been by phenotyping and selection on the basis of physical characters. This can at

earliest be done at about 10 weeks (le gene), but could be as late as after seed harvest

(mollis). Phenotyping of tardis, is difficult and time consuming, being further

complicated if the cross also segregates for le. Ku, because it is a dominant gene

http://www.bom.gov.au/climate/dwo/IDCJDW0600.shtml


9

requires a second generation (F3) to select lines homozygous for this gene, thus

causing considerable delays in selection (as well as increasing the numbers of plants

that need to be grown by approximately 4-fold). Because of this difficulty, there has

been a reluctance to introduce new wild material into the breeding cycle. The

development of molecular markers can overcome this difficulty, enabling selection

for all of these genes within the first weeks of the F2 stage.

1.2.3 Disease and pest resistance

Diseases and pests have only been recognised as being economically

important since lupins were domesticated as a low-alkaloid crop mid-way through

the last century. Most diseases of lupins are caused by fungi or viruses, there being

no known bacterial pathogens and very few reports of significant nematode damage

(Sweetingham et al. 1998).

In Western Australia, lupin breeders have concentrated primarily on

resistance to fungal and viral diseases. Some of the major fungal diseases to have

received the breeders‟ attention are:

1.2.3.1 Brown spot / Pleiochaeta root rot.

The diseases Brown spot and Pleiochaeta root rot are caused by the same

fungus Pleiochaeta setosa. It is widespread, being present in Europe as well as in

Australia, Africa and the Americas, although it is usually less damaging overseas.

The fungus survives as spores on the stubble or is transmitted by the seed. Foliar

infection of the plant by rain-drop splash can result in the defoliation of the plants

and even death. The same organism also causes root rot of seedling lupins, with

infection usually taking place soon after germination and the appearance of lesions 3

to 4 weeks later. The roots can completely rot away, resulting in the death of the

seedling. It is considered to be the most important root pathogen of lupins in

Australia (Sweetingham et al. 1998). Resistance to both Brown spot and Pleiochaeta

root rot are under polygenic control (Cowling 1988), with no major genes having

been found for either disease. Genes providing resistance to one form of the disease

do not necessarily provide resistance to the other form.


10

1.2.3.2 Anthracnose (Colletotrichum gloeosporioides)

Anthracnose of lupins was first reported in south-eastern USA in 1939. It

was later found to be present in New Zealand, being responsible for the rapid decline

of L. arboreus in their pine forests from about 1988 and, in 1994 the first reported

outbreak of anthracnose occurred in breeders plots in Western Australia. The disease

spread rapidly causing widespread infection of commercial crops in the northern

grain-belt of W.A. and on the Eyre peninsula of South Australia in 1996, with

especially L. albus, but also L. angustifolius and L. cosentinii affected. Characteristic

symptoms of the disease are the bending over of stems and petioles with a lesion

inside the crook. These lesions can girdle and break off the stem or twist and deform

the pods. Lesions tend to be purplish or brown with masses of pink conidiospores.

Pods and seeds can also be infected (Sweetingham et al. 1998). Two strains have

been identified – vegetative compatibility groups (VCG) 1 and 2, with the more

virulent strain VCG 2 being the only one so far identified in W.A. (H Yang, pers.

comm.).

Single major resistance genes to anthracnose have been identified in each of

L. angustifolius and L. albus. Current breeding efforts are aimed at incorporating

these and any newly identified resistance genes into new cultivars. The first

anthracnose resistant cultivar of L. angustifolius, Tanjil was released in 1998, and the

first (slightly less) resistant cultivar of L. albus, Andromeda, was released in 2005 in

a bid to re-launch the species as a crop in Western Australia.

To improve the process of incorporating these genes into new varieties,

molecular genetic markers linked to these genes are being developed. Two closely

linked markers have been reported (Yang et al. 2004, You et al. 2005) for L.

angustifolius, although to date, no successful markers have been developed for the

resistance gene of L. albus.

1.2.3.3 Phomopsis (Diaporthe toxica Williamson et al.)

Phomopsis infection of lupins can result in a mycotoxicosis that can kill livestock

grazing on the stubble. This disease is a major problem in Australia and South

Africa, with the organism living parasitically in the living stems of the plant and

usually producing the deadly mycotoxin phomopsin in the plant stems after they

have senesced (Gardiner 1961, 1966). Infected plants develop dark purplish lesions

on infected stems and pods at maturity, or under conditions of stress induced by


11

frost, drought or herbicide damage. Where pod lesions develop prior to harvest, seed

may also become infected, showing a gold or brown discolouration (Sweetingham et

al. 1998). Under Australian conditions, D. toxica over-summers in colonised stubble

as both perithecia and pycnidia. The stubble can serve as a source of inoculum for at

least two years (Wood and McLean 1982). To control this disease the only practical

means is by breeding for resistance. The first resistant cultivar of L. angustifolius,

Tanjil, was released in 1998. A molecular marker linked to a major gene conferring

resistance to Phomopsis has been developed (Yang et al. 2002) to enable improved

selection of resistant genotypes.

1.3 BREEDING TO IMPROVE LUPINS AS A CROP

1.3.1 Past practices

Following on from the completion of domestication of narrow-leaf lupins in

the late 1960‟s / early 1970‟s, the main emphases have been on increased yield,

higher protein content, environmental adaptation and disease and pest resistance. A

number of the major diseases of lupins have been covered in the previous section.

There is a general agreement that elimination of bitter types is important for the

survival of the lupin industry as a whole. Some countries, most notably Spain,

Portugal and Italy face difficulties in maintaining purity of low alkaloid lines because

of the proximity of bitter wild types that are common through that region. However,

even as recently as 1993, a bitter lupin crop was being promoted by a small group

within the European community for the production of various products including

purified alkaloids (Cowling et al. 1998).

1.3.2 Wild genetic material

Wild lupins have been exploited for resistance genes to various diseases

including Pleiochaeta setosa, Grey spot (caused by Stemphylium botryosum Wallr.,

Stemphylium solani Weber) and anthracnose. Frost tolerance is another trait

originally discovered in progeny of crosses with a wild type from Portugal (Wells

and Forbes 1982). In general, improvements in yield and disease resistance have

relied heavily on the domesticated germplasm, with only limited use of wild

accessions. The use of wild accessions has been limited primarily because of the

difficulties associated with the retention of domestication genes.


12

1.3.3 Mutation breeding

Mutations have also played a role in breeding. All of the original

domestication genes (iuc, Ku, le, ta, moll) are natural mutations that were selected in

the field. However, an induced mutant for early flowering (efl) was used in breeding

the moderately early flowering cv. Chittick in W.A. (Anon 1982). Experiments with

induced mutants having a range of restricted branching patterns were carried out in

Western Australia in the early to mid 1990‟s but have since been abandoned.

Conversely, at least one such gene „mut1‟of L. angustifolius has been incorporated in

European cultivars (Bromberek et al. 1984) where early growth cessation and rapid

maturity are required for the short summer growing season (Cowling et al. 1989).

Current research focussing on development of herbicide-tolerant cultivars is aimed at

developing cultivars tolerant to the herbicide Metribuzin, using artificially induced

mutations (Si et al. 2007) as well as naturally occurring variation as sources of

genetic improvement (Si et al. 2006).

1.3.4 Transgenic breeding

Interest in transgenic lupins has been low – probably because of widespread

public resistance to transgenic crops. Molvig et al. (1997) reported producing a

transgenic narrow-leaf lupin with enhanced seed methionine levels by incorporating

a sunflower seed albumin gene and; Jones et al. (1999) reported producing a

transgenic lupin with increased tolerance to the herbicide Basta® (Glufosinate). The

Basta-resistant lupins have not been released to the public.

1.4 MOLECULAR GENETICS AND PLANT BREEDING

1.4.1 Molecular markers

A marker may be defined as a signpost for indicating an item or object of

value and is itself of no intrinsic value. In modern genetics, molecular markers are

used to sign-post genes including (hopefully) those of agronomic value. Jones et al.

(1997) state that “Molecular markers (DNA markers) reveal neutral sites of variation

at the DNA sequence level. By „neutral‟ is meant that, unlike morphological

markers, these variations do not show themselves in the phenotype, and each might

be nothing more then a single nucleotide difference in a gene or a piece of repetitive


13

DNA”. DNA markers can be classed into two categories depending upon how the

polymorphism is revealed, namely: hybridisation-based- and PCR-based-

polymorphisms.

The most common form of hybridisation-based polymorphism are RFLPs

(restriction fragment length polymorphisms) whose sequence(s) are unique. RFLPs

are generated due to events in the DNA chromosome such as point mutations,

inversions, deletions or translocations. The RFLPs are then used as labelled probes

(usually radioactive) on membranes to which have been attached the DNA of the test

organism after it has been digested with one or more restriction enzymes (Kumar

1999). The RFLP only binds to the complementary strand, thus identifying

organisms carrying a particular allele for a given gene.

In PCR, a pair of oligonucleotide primers oriented with their 3' end pointing

towards each other, are used to copy each strand of denatured DNA bearing a

complementary sequence (to which they have annealed), with the aid of a

polymerase which adds nucleotides starting from the 3' end of the primers. The three

steps of the PCR (template denaturation, annealing of primers and enzymatic

sequence extension) are cycled a large number of times to exponentially increase the

numbers of the fragments being copied to levels detectable upon electrophoresis in a

gel. In general, the size of the fragment amplified is below 2000-3000bp (base pairs)

(Kumar 1999). The most common forms of marker generated in this fashion include

AFLPs (amplified fragment length polymorphisms), SSR (simple sequence repeats

or microsatellites) and RAPDs (randomly amplified polymorphic DNA). A recently

developed form of PCR marker is the MFLP (microsatellite fragment linked

polymorphism). The MFLP technique was derived from a combination of the AFLP

concept with SSR-anchor primers (Yang et al. 2001) and is capable of producing

DNA markers with high efficiency, with each of the detected DNA polymorphisms

including an SSR motif (Yang et al. 2001, 2002, 2004).

AFLPs and RAPDs are expressed as „dominant‟ markers only (present or

absent), whereas RFLPs are generally considered to be „co-dominant‟ markers in that

two fragments are produced – one for each allele. SSRs too are frequently expressed

as co-dominant markers whereas MFLPs may be either dominant or co-dominant in

their expression. Co-dominant markers are considered superior to dominant markers

as they enable identification of individuals heterozygous for a particular gene as well

as those that are homozygous.


14

1.4.2 Genetic mapping

1.4.2.1 Mapping and markers

PCR-based markers are increasingly popular in mapping because of the ease

with which they can be produced from even small amounts of DNA. In recent years,

the AFLP technique (Vos et al. 1995) has been widely used in plant genetic mapping

because of its high efficiency in generating large numbers of molecular markers

(Eujayl et al. 2002, Pearl et al. 2004, Yin et al. 2003). The use of SSR markers is

increasing in popularity because they occur at high frequencies in plant genomes (Li

et al. 2002) and exhibit high mutation rates (Vigouroux et al. 2002). MFLPs have

not previously been recorded as being used in genetic mapping.

1.4.2.2 Mapping populations

Selection of mapping populations is critical to successful map construction.

Because a map‟s economic significance will depend on marker-trait associations, as

many qualitatively inherited morphological traits as possible should be included in

the chosen parent lines (Staub et al. 1996).

It has long been considered that the most informative mapping population

was an F2 that had been fully characterised (Mather 1938), using a co-dominant

marker system such as RFLPs or SSRs. The main difficulties in dealing with an F2

population are that many lines (≥50%) are heterozygous and therefore cannot be

adequately characterised by dominant markers such as AFLPs (see below) and, they

have a limited life as a population because of the ongoing genetic segregation with

each generation. Consequently, they are frequently not the first choice in mapping.

The choice of mapping population therefore depends in part on the types of

markers that will be generated to produce the map, with dominant markers such as

AFLPs being best suited to Recombinant Inbred Lines (RILs) and Doubled Haploids

(DH), as they may yield inconclusive (and thus misleading) information if used on an

F2 population without doing extensive progeny testing which can also be expensive

and time consuming (Staub et al. 1996). MFLPs also come into this category as a

large percentage of the marker fragments produced are of a dominant nature (Yang et

al. 2001).

The most commonly used populations are DH and RILs. Doubled haploids

of F1 gametes are used to characterise especially cereals (e.g. Collard et al. 2005, Su


15

et al. 2006), but also Canola (Pradham et al. 2003). Recombinant inbred lines have

similar characteristics, being single plant selections of F2 progeny – usually to either

the F7 or F8 generation and are mostly used on crops for which no DH technology

(yet) exists as in lupins and lentils, although it is sometimes used as an alternative to

DH (e.g. in wheat – Röder et al. 1998).

Fine mapping of traits has frequently presented difficulties. The problem

here is that most mapping populations are too small and have had too few

opportunities for recombination to occur – being constrained both in the number of

parents (only two) and the number of generations over which recombination is

allowed to occur before fixing of the alleles begins (just one). As a consequence, the

correlation between even distant markers and a gene (the Linkage disequilibrium or

LD) can be very high so that markers apparently closely linked to a particular gene

may prove ineffective when applied to a very different population.

Recently the more advanced methods of Linkage disequilibrium (LD)

mapping and Family-based linkage (FBL) mapping have gained greater interest.

Both methods, originally developed for human genetics, are especially powerful in

detection and mapping of quantitative trait loci (QTLs) (Mackay and Powell 2007)

and suited for use in a segregating population. In essence, LD mapping relies on the

weaker linkages (i.e. larger genetic distances) to gradually break down in a large,

randomly mating population, with the end result being that only very close linkages

between genes and markers are retained. One variation of LD mapping is the

Advanced Intercross where F2 individuals are inter-mated for several generations

prior to mapping, causing LD to decay and, the precision of QTL location to be

increased (Darvasi and Soller 1995). FBL mapping is based on the establishment of

a population based on a small number of founding individuals (parents), allowing

them to freely cross over a number of generations, with a similar end result to LD

mapping. In FBL it may be possible to select individuals with extreme phenotypes

(such as individuals exhibiting a particular disease) as the parents to ensure a

population segregating for most QTL associated with traits of interest (Mackay and

Powell 2006).

Other types of mapping populations have also been developed. Bulked

segregant analysis (BSA; Michelmore et al. 2001) uses two bulked DNA samples

drawn from the segregating population of a single cross to find markers linked to a

particular trait by identifying lines identical for this trait and screening them for


16

common markers. It is similar in many ways to using near-isogenic lines (NILs) in

that both are used to screen for a particular trait, but without the problems associated

with developing an NIL (Staub et al. 1996). In tomatoes, inter-specific crosses have

resulted in production (via recurrent back-crossing) of an exotic library of

Introgression Lines (ILs) – a population consisting of a set of lines, each of which

carries a single, defined chromosome segment that originates from a donor species in

an otherwise uniform elite genetic background. It is not usually possible to produce

ordinary mapping populations from such crosses as plants produced frequently

exhibit partial sterility. Such a library is difficult to produce (Kumar 1999) and takes

about 10 generations. This population could then be used for fine mapping of QTLs

or, each line used directly for breeding (Eshed and Zamir 1995, Zamir 2001).

1.4.2.3 Mapping

Genetic mapping in its simplest definition is “putting markers in order,

indicating the relative genetic distances between them” (Jones et al. 1997). The

concept of creating a genetic map is not new, the aim commonly being to enable

improved understanding of genetic behaviour and the (efficient) selection (in crops

and stock) of superior genotypes. The earliest genetic maps were of a physical

nature – creating simple maps showing genetic linkages between morphological

traits and certain genes. One such example is that of Lindsley and Grell (1968) who

created a physical map for the fruitfly, Drosophila melanogaster. However, physical

markers to single genes are frequently difficult to find so that most early maps were

sparsely populated. In 1980, Botstein proposed the construction of a genetic linkage

map of humans using molecular markers (Botstein et al. 1980). Since then, this idea

has been implemented for a wide range of organisms, but especially to crops and

animals of commercial value.

Map construction requires the generation of suitable markers and incorporates

the calculated recombination values from all the pair-wise combinations of markers

to generate linkage groups. Traits controlled by major genes and segregating in that

population may be physically scored and mapped in the same manner as the markers

(Jones et al. 1997). The linear arrangement of these linkage groups should minimise

the number of recombination events between adjacent markers (Staub et al. 1996,

Jones et al. 1997).


17

The relationships between map distance and recombination value is

characterised by a genetic mapping function (mf). An mf is a formula expressing

quantitative relationships between distances in a linkage map using crossover

frequency. They take into account the possibility that the existence of one crossover

may influence the likelihood of another one occurring within a certain distance. This

is known as interference. Staub and Serquen (1996) note that there are several mf

that can be applied but, that the most common mapping functions were developed by

Haldane (1919) and Kosambi (1944). While Haldane‟s mf assumes an absence of

interference, Kosambi‟s assumes positive interference (i.e., fewer double

recombinants when compared to no interference). Map distances are calculated as

centiMorgan (cM), with 1cM being approximately equal to a 1% recombination rate

between the two points (Kumar 1999).

There are a number of computer programs available to help order the markers

into linkage groups. Well known programs include Linkage 1 (Suiter et al. 1983),

MapMaker (Lander et al. 1987) and MapManager-QTX (Manly et al. 2001).

Mapping programs are not always correct in how they order markers and linkage

groups. Thus, it is normal to re-check the output of the program. In the process, it is

also wise to re-score any data that appears as a double recombinant in the map, since

such events are rare and may well be indicative of a scoring error (M. Pallotta, pers.

comm.). One of the many examples of maps constructed using one of these

programs is that by Tanksley et al. (1992) who used MapMaker to construct and

compare two high density linkage maps for the potato and tomato genomes. Many

more maps of varying density have been published for both animals and plants.

Included in the cultivated plant crops are published maps for arabidopsis, banana,

barley, bean, brassica, citrus, coffee, cotton, cucumber, lentil, mustard, rice, rye,

sorghum, soybean, sugar-beet, tomato (Staub et al. 1996, Eujayl et al. 1998, Pradhan

et al. 2003, Pearl et al. 2004), with other maps having been published for tree species

including pine (Travis et al. 1998, Hayashi et al. 2001, Yin et al. 2003 ) and

eucalyptus (Grattapaglia and Sederoff 1994).

1.4.2.4 Genetic mapping in lupins

Lupinus angustifolius is a diploid species having (2n) 40 chromosomes

(Gladstones 1970). A complete genetic map of this species would therefore be

expected to have 20 linkage groups representing the (1n) chromosomes. Wolko and


18

Weeden (1994) estimated that the length of a complete map would most probably be

in excess of 1500cM – on the basis that the polyploid nature of lupins makes it likely

to be larger then that of the average diploid plant. They attempted to construct a

linkage map of an F2 population (n = 48 plants) using RAPDs as well as iso-enzymes

and were successful in linking 60 RAPD and 4 iso-enzyme markers into 17 groups

spanning a total of 986.7cM (Wolko and Weeden 1994, Kruszka and Wolko 1999).

Meanwhile Brien et al. (1999) also attempted to create a linkage map using an F2

mapping population. They generated 705 polymorphic AFLP markers that they

mapped into 52 linkage groups covering approximately 1000 cM. Later, the same

group of researchers (Scobie et al. 2002) reported having just 368 AFLP markers (In

an F8 RIL population?) in 61 linkage groups (LOD 4), with markers identified as

being linked to early maturity, reduced pod shatter, low alkaloid content, permeable

(soft) seed coat and flower colour. Clearly none of the maps were adequate – the

first and third having a very low density, the second and third having more then 2.5x

the expected number (20) of linkage groups. One probable reason for this is that

both RAPDs and AFLPs, being dominant markers fail to discriminate between

homozygous and heterozygous individuals and, failed PCR reactions may be difficult

to distinguish from a null-allele in a dominant marker, making mapping a much more

difficult exercise.

A start has also been made on physically mapping genes and molecular

markers, with the location of rDNA loci on particular chromosomes (Naganowska et

al. 2003). No comprehensive genetic map of L. angustifolius had yet been published

when work was started on this thesis (2003).

1.4.3 Quantitative trait loci

In nature a number of agronomically important traits such as yield, quality,

maturity, and resistance to biotic and abiotic stresses are considered to be polygenic,

displaying a large continuous range in variation approximating the Normal curve.

The continuous variation in phenotypic expression of such traits frequently cannot be

fitted to simple Mendelian ratios and are therefore generally considered to be the

product of the interaction of a number of gene loci (Johannsen 1909, Nilson-Ehle

1909, East 1916), commonly known as Quantitative Trait Loci (QTLs). To improve

crops all sources of superior genetics must be exploited. However, the very nature of


19

QTLs makes them difficult to select in the field and necessitates the use of closely

linked unambiguous markers.

The development of molecular genetic maps now allows researchers to

identify QTLs and make detailed analysis of their inheritance and activity, as first

achieved by Paterson et al. (1988). Several reviews have discussed mapping of

QTLs, including Tanksley (1993), Young (1996) and Kumar (1999). Some of the

main areas of discussion have been covered in the following section.

QTLs cannot be mapped in the same way as a major (single) gene because

the individual loci cannot be directly identified. The principle of QTL mapping is to

associate QTLs with molecular markers in their common inheritance (Jones et al.

1997) and; to assign individual QTLs to a chromosomal location and determine the

alleles contributed by each parent (e.g. Tanksley 1993, Paterson 1995).

Early work on the analysis of QTLs used analysis of variance (ANOVA), also

known as single point analysis, to test the relationship between traits and markers.

This approach does not require a complete molecular linkage map (Tanksley 1993).

It had limitations in that:

(i) the further a QTL is from the marker gene, the less chance it had of being

statistically significant;

(ii) the magnitude of the QTL‟s effect would normally be underestimated due

to recombination between the marker and QTL.

(iii) the likelihood of further (linked) QTLs being missed if the genome was

not completely covered.

Tanksley (1993) considered that these problems would be minimised by a having a

complete linkage map with markers at intervals of less then 15cM.

The availability of complete linkage maps and computers has resulted in

interval mapping becoming the method most commonly used (at present). Interval

mapping (IM) (Lander and Botstein 1989) considers the linkages between markers

and uses the maximum likelihood equation to give an estimate, based on the

Logarithm of Odds threshold (LOD) score, of the probability of a QTL being present

in any interval based on flanking marker information (Kumar 1999). MapManager-

QTX (Manly et al. 2001) was developed for just this purpose. This method is

particularly useful when using a map on which the markers are moderately far apart

(>20cM), but becomes inefficient at large distances (e.g. >35cM) (Tanksley 1993).


20

Interval mapping does not however, have the precision to distinguish multiple

linked QTL effects and, may result in QTLs being mapped to the wrong position

(Zeng 1993, 1994). This has led to investigating alternative methods of analysis

including the development of the Composite Interval Method (CIM). This method

combines multiple regression analyses with IM to help separate closely linked QTLs.

However, it lacks sensitivity in detecting especially smaller QTLs (Zeng 1993,

1994).

Recent work by Zhu and his co-workers in China have focused on analyses

based on a Mixed model Composite Interval Mapping (MCIM) method (Zhu 1998,

Yang and Zhu 2005). This method incorporates the use of matrices rather then IM

alone to allow mapping of QTLs with high precision, including analysis of gene

action (e.g. epistasis) and environmental interaction. An upgraded version of the

computer program became available during 2006 and is known as QTL Network

version 2.0 (Yang et al. 2005).

QTL analysis may be coupled to marker-assisted selection (below) in plant

breeding, enabling breeders to better select not only progeny but also potential

parents carrying desirable genes for a trait of interest. Kumar (1999) has tabulated

35 pest and disease organisms in 14 crops for which QTLs conferring resistance have

been mapped. In narrow-leafed lupins, it is already established that the only

resistance to Pleiochaeta setosa is polygenic in nature (Section 2, Disease and Pest

Resistance). By mapping QTLs associated with traits such as this one, the

opportunity arises to identify markers on the map closely linked to them and, to

develop them into PCR-friendly forms for implementation in breeding.

A further benefit of mapping QTLs is in exploiting synteny (conservation of

linkage). It is not uncommon for related species to have near identical coding

sequences for homologous genes (Moore et al. 1995), albeit not necessarily in

identical positions on the genome. By locating and sequencing a gene of interest it

may be possible to exploit synteny to locate that gene in a related species of interest.

1.4.4 Marker assisted breeding

To effectively and efficiently breed superior crops and livestock, it is

necessary to select parents carrying desirable alleles and combine them in the

offspring while eliminating those that are undesirable. Morphological markers are


21

infrequent and may be only loosely linked to the genes of interest, necessitating a

better method of selection – namely molecular markers.

Marker assisted selection (MAS) is based on the premise that it is possible to

infer the presence of a gene from the presence of a marker tightly linked to that gene

– usually within a distance of no more then 5 cM. This technique would be most

useful for tracing the inheritance of QTLs and major genes where the procedures for

screening the plants (as in disease resistance genes) are labour intensive, or require

plants to be screened for several traits at the same time and, in an environment not

conducive to one or more of these traits to be expressed. For MAS to be efficient,

markers linked to genes of interest should be converted into a Sequence Tagged Site

(STS) by designing primers that will allow highly specific amplification (and

visualisation) of the marker by PCR (Kumar 1999) from relatively small amounts of

DNA.

Such relationships between the gene and marker may be determined by either:

(a) phenotyping a mapping population for the trait of interest and positioning the

gene on a molecular map prior to identification of closely linked markers

(Tanksley et al. 1992);

(b) generating and phenotyping a full-sib population, with the parents differing in the

gene of interest, followed by marker generation and evaluation (Yang 2002,

2004);

(c) by repeatedly back-crossing to generate near-isogenic lines (NILs) which differ

in the presence or absence of the target gene and a small region flanking the

target gene, followed by marker generation and evaluation (Kumar 1999),

(d) A variation of the last method (c) is to generate a population of introgression

lines (IL) using a closely related species as the donor parent (Eshed and

Zamir1995) and then generating and evaluating markers.

(e) Bulk segregant analysis (Michelmore et al. 2001)

The last four methods do not require the generation of a molecular map but could

also be useful in helping to saturate a map region associated with the trait of interest

but lacking in closely linked markers. It should be noted that methods (a) – (c) can

only apply to major genes, and methods (d) and (e) could be used for both major and

minor genes (QTLs). Having a good molecular map would make the process more

efficient.


22

There are many examples in the literature of molecular markers being

developed for use in plant breeding. The crop most frequently reported is wheat (e.g.

Khabaz-Saberi et al. 2002, Collard et al. 2005, Sun et al. 2005, Distelfeld et al. 2006,

Su et al. 2006), although barley, canola, soybean, tomatoes and potatoes also are

actively researched.

Thus far, only a small number of molecular markers linked to major genes of

narrow-leaf lupins have been developed. These include:

(i) A CAPS (cleaved amplified polymorphic sequence) marker for the gene Ku

(Brien et al. 1999);

(ii) An MFLP marker linked to a gene conferring resistance to Phomopsis (Yang et

al. 2002);

(iii) Use of two MFLP markers linked to the one gene conferring resistance to

anthracnose (Yang et al. 2004; You et al. 2005).

1.5 CURRENT NEEDS IN LUPIN BREEDING

Recently there has been a recognition that the possible genetic gain for yield

and other traits including disease resistance will be increasingly limited if wild

accessions are not exploited more extensively. The reluctance to incorporate this

material can be largely attributed to the difficulties associated with the selection for

domestication traits, most of whom are recessive in nature and, of whom one or more

genes (e.g. reduced shattering (ta), early flowering) are difficult and laborious to

select for in the field.

The advent of molecular markers in lupin breeding could remove these

difficulties and radically alter the scene. However, to effectively use this technology

the lupin genome requires extensive characterisation – both morphologically and at a

molecular level. The combination of markers, genes and quantitative traits into a

molecular genetic map is therefore one of the highest priorities for ensuring steady

progress in the rate of genetic gain.

1.6 AIMS OF THIS THESIS

The aims of this thesis are therefore:

(1) To develop and characterise a mapping population of narrow-leafed lupin and; to

generate a comprehensive molecular map of lupins on which are placed all the


23

domestication genes and the previously developed disease resistance genes (and

associated markers) – Chapter 2;

(2) To identify a molecular marker closely linked to the Ku gene and, to convert this

marker into an easily implementable, PCR-friendly form – Chapter 3;

(3) To identify molecular markers closely linked to the le gene and, to convert these

markers into an easily implementable, PCR-friendly form – Chapter 4;

(4) To identify a molecular marker closely linked to the ta gene and, to convert this

marker into an easily implementable, PCR-friendly form – Chapter 5;

(5) To identify a molecular marker closely linked to the moll gene and, to convert

this marker into an easily implementable, PCR-friendly form – Chapter 6;

(6) To phenotype a mapping population for a number of complex traits and to

analyse and map the QTLs that contribute to these phenotypes – Chapter 7.

1.7 REFERENCES

Anon (1982) Registration of Lupinus angustifolius L. (narrow-leaf lupin) cv.

Chittick. J Aust Inst Agric Sci 48:122-124

Adhikari KN, Galway NW, Dracup M (2004) Extending photoperiod and

increasing illuminance hastens flowering of narrow-leafed lupin (Lupinus

angustifolius L.) In: van Santen E, Hill GD (eds) Wild and cultivated lupins

from the tropics to the poles. Proc 10th

Int Lup Conf Laugarvatn, Iceland. Int

Lupin Assn, Canterbury, New Zealand. pp 271-274

Atebekova AI (1958) The dehiscence of the pods in the genus Lupinus (Tourn.) L.

Bull Moscow Soc Nat Res 63:89

Bauhin J, Cherlero JH, Chabraeus D, Agraffenried Fr. Historiae. Plantarum

universalis. Ebroduni (1651)

Botstein D, White RL, Skolnick M, Davis RW (1980) Construction of a genetic

linkage map in man using restriction fragment length polymorphisms. Am J

Hum Genet 32:314-331

Brien SJ, Cowling WA, Potter RH, O'Brien PA, Jones RAC, Jones MGK (1999)

A molecular marker for early maturity (Ku) and marker-assisted breeding of

Lupinus angustifolius. In: Proc 11th

Aust Plant Br Conf Adelaide 2. pp. 204-

205


24

Bromberek S, Mikolajczyk J, Wiza M (1984) [Self-completing mutation of blue

lupin]. In: Proc 3rd

Int Lupin Conf. Int Lupin Assn, La Rochelle, France. pp.

566-567 (in French)

Clements JC, Dracup M, Buirchell BJ, Smith CG (2005) Variation for seed coat

and pod wall percentage and other traits in a germplasm collection and

historical cultivars of lupins. Aust J Agric Res 56:75-83

Collard BCY, Grams RA, Bovill WD, Percy CD, Jolley R, Lehmensiek A,

Wildermuth G, Sutherland MW (2005) Development of molecular markers

for crown rot resistance in wheat: mapping of QTLs for seedling resistance in

a „2-49‟ x „Janz‟ population. Plant Breed 124:532-537

Cowling WA (1988) Breeding for resistance to leaf and root rot caused by

Pleiochaeta setosa in narrow-leafed lupin. In: Proceedings of the fifth

international lupin conference PWRiL, Poznań, Poland, pp. 217-229

Cowling WA (1999) Pedigrees and characteristics of narrow-leafed lupin cultivars

released in Australia from 1967 – 1998. Agriculture Western Australia

Bulletin 4365

Cowling WA, Huyghe C, Swiecicki W (1998) Lupin Breeding. In: Gladstones JS,

Atkins C and Hamblin J (eds) Lupins as crop plants. CAB International,

University Press, Cambridge, UK. pp. 93-120

Cox B (1998) Marketing and trade (of lupins). In: Gladstones JS, Atkins C and

Hamblin J (eds) Lupins as crop plants. CAB International, University Press,

Cambridge, UK. pp. 437-454

Darvasi A, Soller M (1995) Advanced intercross lines, an experimental population

for fine genetic mapping. Genetics 164:741-745

Distelfeld A, Uauy C, Fahima T, Dubcovsky J (2006) Physical map of the wheat

high-grain protein content gene Gpc-B1 and development of a high-

throughput molecular marker. New Phytol 169:753-763

Dracup MR, Thomson R, Reader M, Kirby EJM, Shield I, Leach J (1998)

Daylength responses, flowering time and seed filling in lupins. Aust J Agric

Res 49:1047-1055

East EM (1916) Studies on size inheritance in Nicotiana. Genetics 1:164-176

Eshed Y, Zamir D (1995) An introgression line population of Lycopersicon

pennellii in the cultivated tomato enables the identification of fine mapping of

yield associated QTL. Genetics 141:1147-1162


25

Eujayl I, Baum M, Powell W, Erskine W, Pehu E (1998) A genetic linkage map of

lentil (Lens sp) based on RAPD and AFLP markers using recombinant inbred

lines. Theor Appl Genet 97:83-89

Eujayl I, Sorrels M, Baum M, Wolters P (2002) Isolation of EST-derived

microsatellite markers for genotyping the A and B genomes of wheat. Theor

Appl Genet 104:399-407

Evans LT (1959) Flower initiation in Trifolium subterraneum L. 1. Analysis of the

partial processes involved. Aust J Agric Res 10:1

Fedotov VS (1932) Preliminary results of the searches for alkaloidless lupins. Trudŷ

prikl Bot Genet Selek (Suppl) 54:38-42

Forbes I, Wells HD (1968) Hard and soft seededness in blue lupine, Lupinus

angustifolius L.: Inheritance and phenotype classification. Crop Sci 8:195-

197

Gardiner MR (1961) Lupinosis, an iron storage disease of sheep. Aust Vet J 37:135

Gardiner MR (1966) Fungus-induced toxicity in lupinosis. Brit Vet J 122:508-516

Garnczarska M, Zalewski T, Kempka M (2007) Water uptake and distribution in

germinating lupine seeds studied by magnetic resonance imaging and NMR

spectroscopy. Physiol Planta 130:23-32

Gladstones JS (1958) The influence of temperature and humidity in storage on seed

viability and hardseededness in West Australian blue lupin (Lupinus digitatus

Forsk.). Aust J Agric Res 9:171-181

Gladstones JS (1967) Selection for economic characters in Lupinus angustifolius

and L. digitatus. 1. Non-shattering pods. Aust J Exp Agric and Anim Husb

7:360 – 366

Gladstones JS (1970) Lupins as crop plants. Fld Crop Abstr 23(2):123-148

Gladstones JS. Distribution, origin, taxonomy, history and importance (of lupins)

(1998). In: Gladstones JS, Atkins C and Hamblin J (eds) Lupins as crop

plants. CAB International, University Press, Cambridge, UK. pp. 1-39

Grattapaglia D, Sederoff R (1994) Genetic linkage maps of Eucalyptus grandis and

Eucalyptus urophylla using a pseudo-testcross: Mapping strategy and RAPD

markers. Genetics 137:1121-1137

Hackbarth J (1957) [The genes of the lupin species. 2.Narrow-leafed lupin (Lupinus

angustifolius L.).] Z PflZücht 37:85-91 (in German)


26

Hackbarth J, Troll H-J (1956) [Lupins as grain legumes and fodder plants.]. In:

Kappert H and Rudorf W (eds) [Handbook of Plant Breeding, Part IV], 2nd

edn. Verlag Paul Parey, Berlin and Hamburg. pp.1-51 (in German)

Haldane JBS (1919) The combination of linkage values, and the calculation of

distance between the loci of linked factors. J Genet 8:299-309

Hanelt P (1960) [The Lupins]. A Ziemsen, Wittenberg 104 pp (in German)

Harris DJ, Jago J (1984) Report on chemical compositions of sweet lupin seed

from Western Australia. Government Chemical Laboratories, Perth, 12pp

Hayeshi E, Kondo T, Terada K, Kuramoto N, Goto Y, Okamura M, Kawasaki

H (2001) Linkage of Japanese black pine based on AFLP and RAPD markers

including markers linked to resistance against the pine gall midge. Theor


Hondelmann W (1984) The lupin – ancient and modern crop plant. Theor Appl

Genet 68:1-9

Johannsen W (1909) Elemente der exakten Erblichkeitsliehre. Fisher, Jena

Jones MGK, Wylie SJ, Berryman D, Selladurai S, Brien SJ, Li D, Yang R, Ryan

K, Li H, Li L (1999) Biotechnological tools in lupin breeding and disease

diagnosis. In: Lupin, an ancient crop for the new millennium: Proceedings of

the 9th

international lupin conference, Klink/Muritz, Germany, 20-24 June

1999. International Lupin Association, Canterbury, New Zealand 2000:111-

114

Jones N, Ougham H, Thomas H (1997) Markers and mapping: we are all

geneticists now. New Phytol 137:165-177

Khabaz-Saberi H, Graham RD, Pallotta MA, Rathjen AJ, Williams KJ (2002)

Genetic markers for manganese efficiency in durum wheat. Plant Breed

121:224-227

Kosambi DD (1944) The estimation of map distance from recombination values.

Ann Eugenics 12:172-175

Kruszka K, Wolko B (1999) Linkage maps of morphological and molecular

markers in lupin. In: Proc 9th

Int Lupin Conf, Klink / Muritz, Germany.

International Lupin Association. pp. 100-105

Kumar LS (1999) DNA markers in plant improvement: An overview. Biotech Adv

17:143-182


27

Lander ES Botstein D (1989) Mapping Mendelian factors underlying quantitative

traits using RFLP linkage maps. Genetics 121:185-199

Lander ES, Green P, Abrahamson J, Barlow A, Daly MJ, Lincoln SE, Newberg

L (1987) MAPMAKER, an interactive computer package for constructing

primary genetic linkage maps of experimental populations. Genomics 1:174-

181

Li YC, Korol AB, Fahima T, Beiles A, Nevo E (2002) Microsatellites: genomic

distribution, putative functions and mutational mechanisms: A review. Mol

Ecol 11:2453 – 2465

Lindsley DL, Grell EH (1968) Genetic variations of Drosophila melanogaster.

Carneg Inst Wash Publ 627. Washington D.C.

Lush WM, Evans LT (1980) The seed coats of cowpeas and other grain legumes:

Structure in relation to function. Fld Crop Res 3:267-286

Mackay I, Powell W (2007) Methods for linkage disequilibrium mapping in crops.

Trends Plant Sci 12(2):57-63

Manly KF, Cudmore Jr RH, Meer JM (2001) MapManager QTX, cross-platform

software for genetic mapping. Mammalian Genome 12:930-932

Mather K (1938) Measurement of linkage in heredity. Methuen and Co. London

Miao ZH, Fortune JA, Gallagher J (2001) Anatomical structure and nutritive value

of lupin seed coats. Aust J Agric Res 52:985-993

Michelmore RW, Paran I, Kesseli RV (1991) Identification of markers linked to

disease resistance genes by bulked segregant analysis: A rapid method to

detect markers in specific genomic regions using segregating populations.

Proc Natl Acad Sci USA 88:9828-9832

Mikolajczyk J (1966) Genetic studies in Lupinus angustifolius. 2. Inheritance of

some morphological characters in blue lupine. Genet Pol 7:153-180

Molvig L, Tabe LM, Eggum BO, Moore AE, Craig S, Spencer D, Higgins TJV

(1997) Enhanced methionine levels and increased nutritive value of seeds of

transgenic lupins (Lupinus angustifolius L.) expressing a sunflower seed

albumin gene. Proc Natl Acad Sci USA 94:8393-8398

Moore G, deVos KM, Wang Z, Gale MD (1995) Cereal genome evolution –

grasses, line up and form a circle. Curr Biol 5:737-739


28

Morley FHW, Evans LT (1959) Flower initiation in Trifolium subterraneum L. 2.

Limitations by vernalization, low temperatures and photoperiod in the field at

Canberra. Aust J Agric Res 10:17

Naganowska B, Doležel J, Święcicki WK (2003) Development of molecular

cytogenetics and physical mapping of ribosomal RNA genes in Lupinus. Biol

Plantar 46(2):211-215

Nelson P, Delane R (1990, 1991) Producing lupins in Western Australia. In: E.

Lawson (ed) Bulletin 4179, Department of Agriculture Western Australia

Nilsson-Ehle H (1909) Kreuzunguntersuchungen an Hafer und Weizen. Lund

Paterson AH (1995) Molecular dissection of quantitative traits: Progress and

prospects. Genome Res 5:321-323

Paterson AH, Lander ES, Hewitt JD, Peterson S, Lincoln SE, Tanksley SD

(1988) Resolution of quantitative traits into Mendelian factors, using a

complete linkage map of restriction fragment length polymorphisms. Nature

335:721-726

Pearl HM, Nagai C, Moore PH, Steiger DL, Osgood RV, Ming R (2004)

Construction of a genetic map for arabica coffee. Theor Appl Genet 108:829-

835

Perry MW, Dracup M, Nelson P, Jarvis R, Rowland I, French RJ (1998)

Agronomy and farming systems (of lupins). In: Gladstones JS, Atkins C and

Hamblin J (eds) Lupins as crop plants. CAB International, University Press,

Cambridge, UK. pp. 291-338

Petterson DS. Composition and food uses of lupins (1998) In: Gladstones JS,



Pradhan AK, Gupta V, Mukhopadhyay A, Arumugam N, Sodhi YS, Pental D

(2003) A high-density linkage map in Brassica juncea (Indian mustard) using

AFLP and RFLP markers. Theor Appl Genet 106:607-614

Quinlivan BJ (1966) The relationship between temperature fluctuations and the

softening of hard seeds of some legume species. Aust J Agric Res 12:1009-

1022

Quinlivan BJ (1967) Environmental variation in the long term pattern of hard seeds

of Lupinus varius. Aust J Exp Agric Anim Husb 7:263-265


29

Quinlivan BJ (1968) The softening of hard seeds of sand-plain lupin (Lupinus

varius L.). Aust J Agric Res 19:507-515

Rahman MS, Gladstones JS (1972) Control of lupin flower initiation by

vernalization photoperiod and temperature under controlled environment.

Aust J Exp Agric Anim Husb 12:638-645

Röder MS, Korzun V, Wendehake K, Plaschke J, Tixier, M-H, Leroy, P, Ganal

MW (1998) A Microsatellite map of wheat. Genetic 149:2007-2023

Savi G (1798) Flora pisiana. Pisa, G. Giacomelli

Scobie MN, Brien SJ, Potter RH, Jones MGK (2002) Molecular mapping and

markers in sweet narrow-leafed lupins (L. angustifolius). In: Anon (eds) Plant

breeding for the 11th

millennium; 12th

Australasian plant breeding conference,

Perth, Western Australia p. 237

Sengbusch von R (1931) [Low alkaloid lupins. 2.] Züchter 3:93-109 (in German)

Sengbusch von R (1953) [A contribution to the evolutionary history of our

cultivated food plants with special reference to individual selection]. Züchter

23:353-364 (in German)

Sengbusch von R, Zimmermann K (1937) Die Auffindung der ersten gelben und

blauen Lupinen (Lupinus luteus und Lupinus angustifolius) mit

nichtplatzenden Hülsen und die damit zusammenhängenden Probleme,

insbesondere die der Süsslupinenzüchtung. Züchter 9:57-65

Si P, Buirchell B, Sweetingham MK (2007) Seed yield and anthracnose resistance

of Tanjil mutants tolerant to Metribuzin. In: Crop updates 2007, Lupins and

pulses updates. Department of agriculture and food, Western Australia.

pp.50-51

Si P, Sweetingham MW, Buirchell BJ, Bowran DG, Piper T (2006) Genotypic

variation in metribuzin tolerance in narrow-leafed lupin (Lupinus

angustifolius L.). Aust J Exp Agric 46:85-91

Staub JE, Serquen FC, Gupta M (1996) Genetic markers, map construction, and

their application in plant breeding. HortSci 31(5):729-741

Su J-Y, Tong Y-P, Liu Q-Y, Li B, Jing R-L, Li J-Y, Li Z-S (2006) Mapping

quantitative trait loci for post-anthesis dry matter accumulation in wheat. J

Integr Plant Biol 48(8):938-944


30

Suiter KA, Wendel JF, Case JS (1983) Linkage-1: A pascal computer program for

the detection and analysis of genetic linkage. J Hered 74:203-204

Sun DJ, He ZH, Xia XC, Zhang LP, Morris CF, Appels R, Ma WJ, Wang H

(2005) A novel STS marker for polyphenol oxidase activity in bread wheat.

Mol Breed 16:209-218

Sweetingham MW, Jones RAC, Brown AGP. Diseases and pests (of lupins). In:

Gladstones JS, Atkins C and Hamblin J (eds) Lupins as crop plants. CAB

International, University Press, Cambridge, UK. (1998) 263-289

Swiecicki W, Swiecicki WK (1995) Domestication and breeding of narrow-leafed

lupin (Lupinus angustifolius L.). J Appl Genet 36:155-167

Tanksley SD (1993) Mapping polygenes. Annu Rev Genet 27:205-233

Tanksley SD, Ganal MW, Prince JP, de Vicente MC, Bonierbale MW, Broun P,

Fulton TM, Giovanni JJ, Grandillo S, Martin GB, Messeguer R, Miller

JC, Miller L, Paterson AH, Pineda O, Röder MS, Wing RA, Wu W,

Young ND (1992) High density molecular maps of the tomato and potato

genomes. Genetics 132:1141-1160

Travis SE, Ritland K, Whitham TG, Keim P (1998) A genetic linkage map of

pinyon pine (Pinus edulis) based on amplified fragment polymorphisms.

Theor Appl Genet 97:871-880

Troll H-J (1964) [Cultivation, breeding and seed problems in large seeded legumes

as contribution to the remedy of protein fodder-deficiency in particular in

reference to cultivation – competition between lupins and field peas].

Wissenschaftlich Zeitschrift der Karl-Marx-Universität 13:651-663 (in

German)

Vigouroux Y, Jaqueth JS, Matsuoka Y, Smith JSC, Doebley J (2002) Rate and

pattern of mutation at microsatellite loci in maize. Mol Biol Evol 19:1251-

1260

Vos P, Hogers R, Bleeker M, Reijans M, Lee T, Hornes M, Frijters A, Pot J,

Peleman J, Kuiper M., Zabeau M (1995) AFLP: A new technique for DNA

fingerprinting. Nucleic Acids Res 23:4407-4414

Wells HD, Forbes I (1982) Genetics and breeding of Lupinus for winter production

for the south of the United States of America. In: Gross R, Bunting ES (eds)

Agricultural and Nutritional aspects of lupines. Proc 1st Int lupine wrkshp.

GTZ GmbH, Eschborn, Germany. pp. 77-89


31

Wolko B, Weeden NF (1994) Linkage map of isozyme and RAPD markers for the

Lupinus angustifolius L., In: Neves-Martins, J.M. and Beirão da Costa, M.L.

(eds) Advances in Lupin Research. Proc 7th

Int Lupin Conf, Evora, Portugal.

ISA Press, Portugal. pp. 42-49

Wood P McR, McLean GD (1982) Lupins-the disease problem. W.A. J Agric

16:31-32

Yang H, Boersma JG, You M, Buirchell BJ, Sweetingham MW (2004)

Development and implementation of a sequence-specific PCR marker linked

to a gene conferring resistance to anthracnose disease in narrow-leafed lupin

(Lupinus angustifolius L.). Mol Breed 14:145-151

Yang H, Shankar M, Buirchell BJ, Sweetingham MW, Caminero C, Smith

PMC (2002) Development of molecular markers using MFLP linked to a

gene conferring resistance to Diaporthe toxica in narrow-leafed lupin

(Lupinus angustifolius L.). Theor Appl Genet 105:265-270

Yang H, Sweetingham MW, Cowling WA, Smith PMC (2001) DNA

fingerprinting based on microsatellite-anchored fragment length

polymorphisms, and isolation of sequence-specific PCR markers in lupin


Yang J, Zhu J (2005) Methods for predicting superior genotypes under multiple

environments based on QTL effects. Theor Appl Genet 110:1268-1274

Yang J, Hu CC, Ye XZ, Zhu J (2005) QTL Network 2.0. Institute of

Bioinformatics, Zhejiang University, Hangzhou, China.

(http://ibi.zju.edu.en/software/qtlnetwork)

Yin TM, Wang XR, Andersson B, Lerceteau-Köhler E. (2003) Nearly complete

genetic maps of Pinus sylvestris L. (Scots pine) constructed by AFLP marker

analysis in a full-sib family. Theor Appl Genet 106:1075-1083

You M, Boersma JG, Buirchell BJ, Sweetingham MW, Siddique KHM, Yang H

(2005) A PCR-based molecular marker applicable for marker-assisted

selection for anthracnose disease resistance in lupin breeding. Cell Mol Biol

Lett 10:123-134

Young ND (1996) QTL mapping and quantitative disease resistance in plants. Annu

Rev Phytopathol 34:479-501

Zamir D (2001) Improving plant breeding with exotic genetic libraries. Nature

Reviews – Genetics 2:983-989


32

Zeng Z-B (1993) Theoretical basis of separation of multiple linked gene effects on

mapping quantitative trait loci. Proc Natl Acad Sci USA 90:10972-10976

Zeng Z-B (1994) Precision mapping of quantitative trait loci. Genetics 136:1457-

1468

Zhu J (1998) Mixed model approaches for mapping quantitative trait loci. Hereditas

(Bejing) 20(Sup):137-138

Chapter 2: Construction of a genetic linkage map

33

Chapter Two

Construction of a genetic linkage map using MFLP, and

identification of molecular markers linked to domestication

genes in narrow-leafed lupin (Lupinus angustifolius L.)


34

2.1 INTRODUCTION

The narrow-leafed lupin (Lupinus angustifolius L.) was grown as a crop in

classical Greek and Roman times (Gladsones 1970), being used for both stock feed

and human consumption. Plant breeding in L. angustifolius started in Germany early

last century, followed by extensive work in Australia in the 1950’s (Gladstones

1998). Key domestication genes incorporated into modern cultivars of L.

angustifolius include the low-alkaloid gene iucundis (iuc), seed coat permeability

gene mollis (moll), white flower gene leucospermum (leuc), nil vernalisation

requirement gene Ku and, reduced pod shattering genes tardus (ta) and lentus (le)

(Gladstones 1970, Pate et al. 1985, Swiecicki and Swiecicki 1995). Domesticated

cultivars of L. angustifolius are currently grown in Europe, Africa and North

America (Pate et al. 1985, Swiecicki and Swiecicki 1995), but Australia remains the

world's largest producer of this species with approximately 1.5 million hectares

planted annually in Western Australia alone (Brien et al. 1999). The objectives of

modern lupin breeding programs include improved yield, higher protein content,

environmental adaptation and disease and pest resistance, particularly resistance to

anthracnose which is the most devastating known disease of lupins (Yang and

Sweetingham 1998). Recently two molecular markers linked to the anthracnose

resistance gene Lanr1, have been developed and used in marker-assisted selection for

anthracnose resistance in breeding of lupins in Australia (Yang et al. 2004, You et al.

2005).

L. angustifolius is a diploid species containing (2n) 40 chromosomes

(Gladstones 1970). Molecular genetic mapping in L. angustifolius was first

attempted in Poland (Wolko and Weeden 1994, Kruszka and Wolko 1999), with 60

RAPD markers and 4 isozyme markers mapped into 17 groups. Brien et al. (1999)

generated 705 polymorphic AFLP markers, which mapped into 52 linkage groups

covering approximately 1000cM. To date, no comprehensive genetic map of L.

angustifolius has been published.

In recent years, the AFLP technique (Vos et al. 1995) has been widely used

in plant genetic mapping because of its high efficiency in generating large numbers

of molecular markers (Eujayl et al. 2002, Li et al. 2002, Pearl et al. 2004). The use of

simple sequence repeat (SSR) markers also known as microsatellite markers, is

increasing in popularity because they occur at high frequencies in plant genomes (Li


35

et al. 2002), and exhibit high mutation rates (Vigouroux et al. 2002). The MFLP

technique was derived from a combination of the AFLP concept with SSR-anchor

primers (Yang et al. 2001) and is capable of producing DNA markers with high

efficiency, with each of the detected DNA polymorphisms including an SSR motif

(Yang et al. 2001, 2002, 2004). The objectives of this study are (1) to employ the

MFLP technique to generate large numbers of DNA markers for construction of a

genetic map of Lupinus angustifolius and (2) to identify molecular markers linked to

the domestication genes and anthracnose resistance gene Lanr1.

2.2 MATERIALS AND METHODS

2.2.1 Plant material

An F8 RIL mapping population (DxW) was developed from a cross between a

domesticated line 83A:476 (maternal) and a wild type P27255 from Morocco

(paternal). This DxW population segregates for a number of traits and domestication

genes including: permeable (soft) seed coat gene mollis (Mikolajczyk 1963); non-

shattering genes lentus and tardus (Gladstones 1960); sweetness (low alkaloid) gene

iucundis (von Sengbusch 1930); early flowering gene Ku (Gladstones and Hill 1969)

and, white flower / seed colour gene leucospermus (Hallqvist 1921). In addition, it is

known that the domesticated parent carries the gene for resistance to anthracnose,

Lanr1 (Yang et al. 2004). A total of 89 sibs were used in this study.

2.2.2 Phenotyping

Mollis (moll). This gene is recessive (moll) and allows rapid imbibition and

germination of viable seeds when soaked in water (Forbes and Wells 1968). Thirty

seeds of each RIL were placed on filter papers in a Petri plate, moistened with

deionised water, placed on the bench and maintained at room temperature. Seeds

were monitored for imbibition and emergence of the radicle (germination) as

indicative of a permeable seed coat. Lines were considered to have permeable seeds

(soft seeded) when at least 75% of viable seed imbibed within 10 days. Seeds that

had not imbibed in that period had their seed coats scarified and were monitored for a

further 10 days to confirm their viability.


36

Leucospermus (leuc). This gene controls flower colour, which is either blue

(Leuc) or white (leuc). Both flower colour and the production of a white seed coat

have previously been found to be associated with this gene (Gladstones 1970,

Hallqvist 1921). Flower colours were confirmed on rows of plants grown in the

screen-house and single plants grown in the glasshouse. Seed coat background

colours, which are either grey (Leuc) or white (leuc), were also recorded.

Lentus and Tardus (le and ta). These two genes are complementary reduced

pod-shatter genes. Each individually reduces but does not eliminate pod shattering

after ripening. When a plant possesses both reduced shattering genes (genotype lele

tata), pod shatter is largely eliminated. The le allele when homozygous has

previously been found to be associated with orange-red pigmentation in the (living)

stem and mature inner pod walls (Gladstones 1967). When the Le allele is present,

no orange-red pigmentation is produced. No such association is known to occur for

ta. Plants were grown in the glasshouse and scored for the presence of the le gene on

the basis of presence or absence of orange stem pigmentation. They were allowed to

mature and dry, and monitored for pod shatter. Pods of plants having either none or

one of the two genes eventually shattered. Those having both genes remained intact.

Iucundis (iuc). This gene controls alkaloid levels. Plants with the Iuc allele

are bitter, and plants with the iuciuc genotype are sweet (Hackbarth and von

Sengbusch 1934). Plants grown in the screen-house and glasshouse were tested for

alkaloid levels by means of Dragendorff papers (Hackbarth and von Sengbusch

1934, Rogers and Bendich 1994). Sap from leaf petioles of flowering plants was

expressed onto the Dragendorff paper and inspected. Where the paper colour

remained unchanged the plant was considered to have low alkaloid levels (sweet).

Sap from plants having high alkaloid levels (bitter) turned the paper pink.

Ku. This gene controls the need for vernalisation. Plants with the Ku

genotype do not require vernalisation, resulting in early flowering (Gladstones and

Hill 1969). Rows of RILs were grown in the screen-house over two winters and

flowering dates recorded. They were designated as either late or early based on the

timing of opening their first flowers relative to parental types.

Anthracnose disease resistance (Lanr1). Seedlings were vernalised for three

weeks at 4°C before being planted in 25cm diameter pots filled with a mixture of

river sand and potting mix. Five single plants of different RILs were selected at

random and planted per pot. All treatments were replicated 4 times. Plants were


37

inoculated with conidial spore suspension of the strain “VCG-2” of Colletotrichum

lupini (Yang and Sweetingham 1998) (previously classified as Colletotrichum

gloeosporioides) ( Nirenberg et al. 2002). Details of disease inoculation and scoring

were the same as in the other studies (Yang et al. 2004).

Moustache (mou) patterning on seed coat. The moustache pattern, also

known as eyebrow or arrowhead (C. Smith, pers. comm.) is of no known agronomic

value but has been used by Australian plant breeders in lupin seed descriptions.

Mature seed of all RILs were examined and scored for the presence or absence of a

moustache pattern.

2.2.3 DNA extraction

Total genomic DNA was extracted from 0.5g of fresh leaf material for each

RIL using the CTAB extraction method as described by Rogers and Bendich (1994).

Extracted DNA was re-suspended in TE 0.1 buffer (10 mM TrisHCl, 0.1 mM EDTA).

2.2.4 MFLP protocol

DNA from each RIL line and the two parents was digested by the restriction

enzyme Tru9I (Roche Diagnostics, Australia), an isoschizomer of MseI. The MseI-

adaptor (Vos et al. 1995) was ligated onto the restriction fragments using T4 DNA-

ligase (Roche Diagnostics Australia). To reduce and optimise the number of DNA

bands on the sequencing gel, the template DNA was further digested with a frequent

cutter HpaII (recognition site 5’-C/CGG-3’, GeneWorks Pty Ltd, Australia).

Detailed methodology for MFLPs have been described elsewhere (Yang et al. 2001).

Pre-selective MFLP reactions were set up using 10 SSR-anchor primers

(Table 2.1), each in combination with an MseI-primer with one selective nucleotide

C at the 3’ end (5’-GAT GAGTCCTGAGTAAC-3’) (Vos et al. 1995). A total of

153 selective-MFLP reactions were run using the 10 SSR-anchor primers in

combination with 16 MseI primers having two additional selective nucleotides

(MseI-Cxx, (Vos et al. 1995)). The SSR-anchor primers of the selective-MFLP

reactions were labelled with -[33

P] (Yang et al. 2001, 2002). The PCR

amplification products of the MFLP were resolved on 5% denaturing sequencing gels

and, DNA polymorphisms were detected by autoradiography (Yang et al. 2001,

2002, 2004).


38

Table 2.1. DNA sequences of SSR-anchor primers used in MFLP

Primer Name Sequences (5’ - 3’)

MF128 DVDTCTCTCTCTCTCTCa

MF129 HVHTGTGTGTGTGTGTGb

MF51 GGGAACAACAACAAC

MF42 GTCTAACAACAACAACAAC

MF43 CCTCAAGAAGAAGAAGAAG

MF62 CCCAAACAACAACAAC

MF52 GGGAAGAAGAAGAAG

MF11 GGACCTCTCTCTCTCT

MF151 CACGTCTCTCTCTCTCT

MF152 GATGCTCTCTCTCTCTC

a D=A+G+T, V=A+G+C;

b H=A+C+T

2.2.5 Data analysis and map construction

The majority of SSR markers were scored as dominant and the remainder as

co-dominant markers. Ambiguous genotypes were treated as missing data. Marker

scoring was checked at least twice. Added to the data were phenotype scores given

each RIL for the traits and anthracnose resistance gene described above under 'plant

material'. Linkage group construction was accomplished with aid of the computer

program MapManager, version QTXb20 (Manly et al. 2001). Linkage groups were

determined to consist of markers linked with a minimum LOD score of 2.5. The

order of the markers in each linkage group was initially determined by

MapManager's 'find linkage groups' tool (Linkage evaluation: Self RI; Search Link

criterion P = 0.001; Map Function: Kosambi) followed by a critical visual evaluation

of all data. The map was drawn using the computer program MapChart (Voorrips

2002).

2.2.6 Genome length estimates

The recombination length of the L. angustifolius genome was estimated using

two methods, Hulbert et al. (1988)’s method-of-moment estimator and Chakravarti et

al. (1991)’s modification of this method which corrects for upward bias related to


39

chromosome ends. The confidence intervals (p = 0.05) for the latter method were

calculated as per Gerber and Rodolphe (1994).

2.3 RESULTS

2.3.1 Phenotyping

Phenotyping of parental lines confirmed that the domesticated parent

(83A:476) bore all the domestication genes (moll, leuc, iuc, le, and Ku), whereas the

wild parent (P27255) carried the wild type genes (Moll, Leuc, Iuc, Le and ku). Parent

83A:476 is resistant to anthracnose, whereas P27255 is susceptible. Testing of all

RILs for the domestication traits and for anthracnose resistance found most of the

phenotypes segregated in a 1:1 ratio (Table 2.2). These results support earlier

findings that those traits are controlled by single major genes (Hallqvist 1921,

Gladstones 1967, Gladstones and Hill 1969, Mikolajczyk 1963). Exceptions were

the phenotypes for low alkaloids (iuc) and anthracnose resistance (lanr1), both of

which have previously also been shown to be controlled by single genes (Yang et al.

2004, von Sengbusch 1930).

2.3.2 Marker polymorphism and segregation

MFLP polymorphisms were screened from 153 primer combinations (Table

2.3). A total number of 1083 polymorphic markers were scored. Marker fragment

lengths ranged from ~50 bp to 800bp, with the majority in the range of 100 - 450 bp.

The number of polymorphic markers detected for individual primer pairs ranged

from nil to 16 with the average number being 6.7 per primer combination. There

were 522 unique marker loci. It was calculated that 21% of markers were co-

dominant and 79% dominant. Upon closer inspection it was found that

approximately 65% of the map length was covered by the first 526 markers

generated (SSR anchor primers MF128, 129, 51, 42). The coverage was increased to

over 75% by inclusion up to marker DAWA639.150 (MF43).

Chi-squared analysis revealed that 91.2% of the mapped loci segregated in

the expected 1: 1 ratio, whereas 46 (8.8%) showed distortion at P < 0.05. The

majority of these markers were clustered in Linkage Groups (LG) 2 (8), LG 6 (5),

LG 9 (14), and LG 10 (7). Distortion in LG2 was in the region of the gene for


40

anthracnose resistance (Lanr1) and in LG 9 the distortion was in the region of the

iucundis gene (iuc). Of the other 13 distorted markers, in at least 3 cases the

distortion may be due to missing data (markers (DAWA)1046.125c; 948.215;

103.05). Overall, the ratio of maternal to paternal phenotypes within the distorted

sectors was 1:3.

2.3.3 Map construction

A framework linkage map was constructed for Lupinus angustifolius (Figure

2.1). Markers and trait phenotypes were placed into 21 linkage groups (LGs)

consisting of 8 or more markers. Linkage groups have been numbered in order of

decreasing marker numbers and length. The map distance covered by the 21 LGs

was 1543cM. Linkage distance spanned by the individual LGs ranged from 19.9cM

(LG 21) to 135cM (LG 1). Linkage groups had a mean length of 73.5 and a standard

deviation of 34.5cM. The average distance between adjacent markers was 3.3cM.

The longest distance between neighbouring markers (28cM) is located on LG 1.

Some clustering of markers was noted in most LGs.


41

Table 2.2. Chi-square test on the segregation ratios of domestication traits and

anthracnose disease resistance among the 89 RILs derived from cross 83A:476 x

P27255 of Lupinus angustifolius

Traits Observed segregation

(M: P) a

Expected

segregation

2 P

Seed coat permeability 36 : 51

(moll : Moll)b, c

43.5 : 43.5 2.586 0.108

Flower colour 43 : 42

(leuc : Leuc)d, c

42.5 : 42.5 0.012 0.914

Plant alkaloid 32 : 57

(Iuc : iuc)e

44.5 : 44.5 7.022 0.008 **

Early flowering 44 : 38

(Ku : ku)f, c

41 : 41 0.439 0.508

Reduced pod shatter 1 45 : 44

(le : Le) g

44.5 : 44.5 0.011 0.916

Seed Moustache 41 : 46

(Mou : mou)h,i,c

43.5 : 43.5 0.287 0.592

Anthracnose disease 31 : 49

(R : S)j,c

40 : 40 4.050 0.044 *

aM = maternal, P = paternal numbers;

bMoll = hard-seeded coat, moll = soft-seeded seed coat;

cTotal

plant numbers do not equal 89 as a result of either inconclusive phenotyping or early plant

mortality;.dLeuc = blue flower colour, leuc = white flower colour;

eIuc = bitter, iuc = sweet;

fKu =

early flowering, ku = late flowering; gLe = shattering, le = reduced shattering;

hMou = broad

moustache , mou = narrow moustache; iMou - parental dominance not determined;

jR = resistant to

anthracnose, S = susceptible to anthracnose.


42

Table 2.3. Primer combinations in MFLP giving rise to DNA polymorphic markers

for genetic linkage mapping in Lupinus angustifolius1

SSR-

anchor

primers

MseI-primers

-CAA -CAG -CAC -CAT -CGA -CGG -CGC -CGT -CCA -CCG -CCC -CCT -CTA -CTG -CTC -CTT

MF128 142 25 32 39 55 64 75 87 96 106 115 127 133 144 155 164

MF129 178 188 191 198 204 211 nil 216 221 227 234 240 246 254 262 269

MF51 281 289 299 312 320 324 325 347 351 355 359 367 376 385 396 410

MF42 413 421 425 436 443 444 451 461 464 474 478 493 503 507 512 520

MF43 527 532 546 558 564 566 571 575 579 593 599 601 603 613 620 632

MF62 641 644 649 657 661 670 678 686 692 701 709 723 733 738 742 747

MF52 748 758 770 776 787 789 792 796 798 808 815 823 829 837 844 853

MF11 869 873 885 895 902 912 920 928 935 945 951 961 973 976 988 995

MF151 1002 nil 1008 1012 1018 1021 1025 1027 1032 1038 1040 1047 1056 1058 1061 1068

MF152 1072 1085 1089 1091 nil nil 1094 nil 1095

1 The prefix DAWA applies to all markers listed in the table;

2Numbers within each cell denote the

first scored polymorphic marker for that particular primer combination; for example, primer

combination of MF128 with MseI-CAA gave rise to 11 markers from “DAWA14” to “DAWA24”. 3Nil means that no markers could be scored.

4Redundant markers included here are not shown on the

linkage map.

2.3.4 Mapping of domestication and the anthracnose resistance genes

Except for the reduced pod-shatter gene ta for which the phenotype data was

not yet complete, all domestication genes were integrated into the molecular genetic

map (Figure 2.1, Table 2.4). The low alkaloid gene iuc was found in LG 9. Seven

molecular markers flank the iuc gene on either side at distances of less than 5 cM,

among which the closest two markers flanking the gene are DAWA632.550 (1.1 cM

to iuc) and DAWA604.335 (1.2 cM to iuc). The gene for early flowering (Ku) in

LG 17 co-segregated with marker DAWA428.290 and is flanked by a further 5

markers within 5cM. The gene for soft-seeds (moll) was located in LG 8. This gene

was found to co-segregate with marker DAWA561.180. Two more markers,

DAWA783.180 and DAWA664.290 are positioned at 3.8 and 5.6cM from moll

respectively. The le gene, one of two reduced shattering genes, was located in LG13

flanked by markers DAWA964.275 (0.6cM) and DAWA323.150 (1.2cM). The

gene leuc was located on LG 11. It is flanked on one side by marker DAWA1045.15

at a distance of 0.7cM, and on the other by marker DAWA520.350 at a distance of

1.1cM. The anthracnose resistance gene Lanr1 and marker AntjM2 (You et al.

2005) were both located in LG2 separated by a distance of 2.6cM. The next nearest


43

marker to Lanr1 is DAWA224.170 at a distance of 4.1cM and on the far side of

Lanr1 to AntjM2. The anthracnose resistance marker AntjM1 (Yang et al. 2004)

was also tried but found to be monomorphic in this population.

Table 2.4. Characteristics of 21 LGs of L. angustifolius constructed with 454 MFLP

markers based on 89 RILs.

L G

No.

LG length

(cM)

No. of

markers

Density

(markers /

cM)

No. of

gaps a Morphological Traits ;

Marker numbers b,c Distortion regions

Marker interval b, (totals)

1 137.3 45 3.0 1 758.470 (1) P d

2 124.7 33 3.8 0 Lanr1; AntjM2b1,

224.170, 1019.350c

563.130 - 824.400 (5) P;

207.325 - 1019.350c (4) P

3 123.9 39 3.2 1 972.100c - 1046.125c (3) P

4 106.2 30 3.5 0

5 95.5 26 3.7 0

6 92.4 22 4.2 0

7 89.1 15 5.9 1

8 88.0 19 4.6 1 moll; 561.180, 783.180

9 85.5 29 2.9 0 iuc; 632.550, 604.335 305.278 - 679.430 (14) P

10 74.6 27 2.8 0 mou; 479.425, 198.450c 93.090 - 41.400 (7) P

11 73.4 24 3.1 0 leuc; 520.350, 1045.150

12 72.7 22 3.0 0

13 71.5 17 4.2 0 le; 323.15, 964.275

14 87.4 20 4.4 0

15 65.8 17 3.9 0 350.050 - 742.850 (4) M

16 36.9 17 2.2 0

17 33.0 14 2.4 0 Ku; 428.290, 306.255c 229.250, 193.240 (2) P

18 31.7 9 3.5 0

19 27.1 11 2.5 0 341.230 (1) M

20 26.4 11 2.4 0 948.215c (1) M

21 19.9 7 2.8 0

Total/

mean

1543

454

3.4

4

(42)

a A distance of more than 20cM between two adjacent markers is termed a 'gap'

b The pre-fix DAWA has been omitted from marker names for ease of reading.

b1 The prefix DAWA does not apply here.

c Only the two most closely associated markers are indicated for any given trait.

d The letter P represents distortion in favour of the paternal parent. M denotes distortion in favour of

the maternal parent.

2.3.5 Estimation of genome size of L. angustifolius

According to the method of Hulbert et al. (1988), the genome size of L.

angustifolius is expected to be in the range 1596cM (LOD 7.0) to 1896cM (LOD 3.0)

with an average value of 1746cM. Alternatively, by using the methods of

Chakravarti et al. (1991) and Gerber and Rodolphe (1994), the expected genome

length was calculated as being between 1407cM (LOD 7) and 1673cM (LOD 3) with

an average length of 1540cM.


44

Fig

ure

2.1

. G

enet

ic l

ink

age

map

of

Lup

inu

s an

gu

stif

oli

us

con

stru

cted

fro

m 8

9 F

8 d

eriv

ed r

eco

mb

inan

t in

bre

d l

ines

. E

ach

mar

ker

s h

as a

fir

st n

um

ber

den

oti

ng

th

e

nu

mer

ical

ord

er o

f m

ark

ers

(Tab

le 2

.3)

from

MF

LP

gel

s, a

nd

an

oth

er n

um

ber

aft

er t

he

dec

imal

den

oti

ng t

he

size

of

the

frag

men

ts i

n b

p.

Th

e su

ffix

'c' o

n t

he

end

den

ote

s th

at t

he

mar

ker

is

co-d

om

inan

t.

DA

WA

ref

ers

to D

epar

tmen

t of

Ag

ricu

lture

Wes

tern

Au

stra

lia.

G

enet

ic d

ista

nce

s o

n t

he

left

sid

e of

the

bar

are

in

cM

(Ko

sam

bi

fun

ctio

n).

G

enes

of

inte

rest

are

lo

cate

d i

n t

he

foll

ow

ing

Lin

kag

e gro

up

s: L

G2

– A

nth

racn

ose

res

ista

nce

gen

e L

an

r1;

LG

8 –

Mo

llis

; L

G9

– I

ucu

nd

is;

LG

10

– m

ou

stac

he;

LG

11 –

Leu

cosp

erm

us;

LG

13

– L

entu

s; L

G1

7 –

Ku

. D

isto

rted

lo

ci a

re m

ark

ed w

ith

* (

P<

0.0

5)

or

** (

P<

0.0

1).


45


46

2.4 DISCUSSION

This thesis has reported the first comprehensive linkage map for Lupinus

angustifolius, with an average distance of 3.2cM between adjacent markers. We

found 522 unique marker loci of which 89% could be connected into 21 linkage

groups covering 1543cM. We have identified a weak linkage between LGs 17 and

20 as well as between some of the other linkage groups and residual fragments (data

not presented). However, it was considered that these linkages were not sufficiently

strong for mapping purposes.

Some clustering of markers was observed in most linkage groups. Clustering

of markers is not an uncommon occurrence (Keim et al. 1990, Tanksley et al. 1992,

Vallejos et al. 1992, Reiter et al. 1992) and has been suggested to occur frequently

although not exclusively in centromeric regions, possibly as a result of reduced

recombination in such regions (Tanksley et al. 1992). Clustering in our linkage

groups has mostly been consistent with this theory in that several tend to be

approximately central. Exceptions including LGs 9, 10, 11 and 12 are probably the

result of incomplete linkage groups as suggested by the residual fragments.

Distorted sib segregation of markers in RIL populations have been reported

by others as being potentially indicative of non-random sib selection or the presence

of genes linked to either the maternal or paternal parent (Yin et al. 2003), or other

distorting factors (Zamir and Tadmore 1986). However the level of distortion, about

8.8% at P < 0.05 is low (2.9% at P < 0.01) and within the range reported as normal

by other workers (e.g. Yin et al. 2003).

When compared to the expected genomic map length as calculated using the

formulae of Chakravarti et al. (1991) and Gerber and Rodolphe (1994) and the limits

set by Hulbert et al. (1988)'s formula, it could be suggested that much of the genome

has been mapped. The close proximity of markers mapped and linked to each of the

domestication genes at a distance of 2cM or less is indicative of a good coverage of

the genome within the mapped regions. This is supported by our data for the

observed distance between Lanr1 and marker AntjM2 of 2.6cM c.f. the 2.3cM

measured by You et al. (2005) for 184 RILS of a cross between the cultivars Tanjil

(resistant) and Unicrop (susceptible). However, the presence of as yet un-linked


47

fragments and several larger gaps in the LGs, suggests that the map is not yet quite

complete.

Using MFLPs to construct a linkage map has the advantage of microsatellites

as many MFLP polymorphisms can easily be converted into simple PCR based

markers desirable for routine marker implementation in MAS (Yang et al. 2001,

2002, 2004). Some benefits of mapping the genome of Lupinus angustifolius are

already being realised. We are currently in the process of developing sequence-

specific PCR primers for several of the markers associated with the domestication

genes we have mapped. With the development of PCR primers for these markers it

will be possible to make a much greater use of wild plant accessions without the need

for lengthy phenotypic screening procedures to ensure retention of these important

domestication traits. The marker AntjM2 for the anthracnose resistance gene Lanr1

is already being used (You et al. 2005). With further development of this map and

the identification of QTLs for yield and other quality traits it is anticipated that the

breeding of lupins using MAS will be greatly enhanced, resulting in the release of

superior cultivars over a reduced time scale.

Acknowledgements

The help of Dr M. You (Department of Agriculture and Food; Western Australia.) in

carrying out DNA extractions and anthracnose disease screening is gratefully

acknowledged. Mr C. Smith (DAFWA) is thanked for his advice and contribution in

sourcing and sorting references and for scanning the linkage map.


48

2.5 REFERENCES

Brien SJ, Cowling WA, Potter RH, O'Brien PA, Jones RAC, Jones MGK (1999)

A molecular marker for early maturity (Ku) and marker-assisted breeding of

Lupinus angustifolius. In: Proc. 11th

Aust Plant Br Conf (Adelaide), pp204 -

205

Chakravarti A, Lasher LK, Reefer JE (1991) A maximum-likelihood method for

estimating genome length using genetic linkage data. Genetics 128:175-182

Eujayl I, Sorrels M, Baum M, Wolters P (2002) Isolation of EST-derived

microsatellite markers for genotyping the A and B genomes of wheat. Theor



angustifolius L.: Inheritance and phenotype classification. Crop Sci 8:195-

197

Gerber S, Rodolphe F (1994) An estimation of the genome length of maritime pine

(Pinus pinaster Ait.). Theor Appl Genet 88:289-292


and L. digitatus. 1. Non-shattering pods. Aust J Exp Agric Anim Husb

7:360-366

Gladstones JS (1970) Lupins as crop plants. Fld Crop Abstr 23:123-148

Gladstones JS (1998) Distribution, origin, taxonomy, history and importance. In:

Gladstones JS, Atkins C, Hamblin J (eds) Lupins as Crop Plants: Biology,

Production and Utilization. CAB International, Oxon, U.K., pp1-39

Gladstones JS, Hill GD (1969) Selection for economic characters in Lupinus

angustifolius and L. digitatus. 2. Time of flowering. Aust J Exp Agric Anim

Husb 9:213-220

Hackbarth J, Sengbusch R von (1934) The inheritance of alkaloid freedom in

Lupinus luteus and Lupinus angustifolius. Züchter 6:249-255

Hallqvist C (1921) The inheritance of the flower colour and the seed colour in

Lupinus angustifolius. Hereditas 2: 299-363

Harborne JB (1984) Phytochemical Methods: A guide to modern techniques in

plant analysis. Chapman and Hall, London


49

Hulbert SH, Ilott TE, Legg EJ, Lincoln SE, Lander ES, Michelmore RW (1988)

Genetic analysis of the fungus, Bremia lactucae, using restriction fragment

length polymorphisms. Genetics 120:947-958

Keim P, Diers BW, Olson TC, Shoemaker RC (1990) RFLP mapping in soybean:

association between marker loci and variation in quantitative traits. Genetics

126:735-742

Kruszka K, Wolko B (1999) Linkage maps of morphological and molecular

markers in lupin. In: Proc 9th

Int Lupin Conf, Klink / Muritz, Germany.

International Lupin Association. pp. 100-105

Li YC, Korol AB, Fahima T, Beiles A, Nevo E (2002) Microsatellites: genomic

distribution, putative functions and mutational mechanisms: A review. Mol

Ecol 11:2453–2465


software for genetic mapping. Mammalian Genome 12:930-932



Nirenberg HI, Feiler U, Hagedoorn G (2002) Description of Colletotrichum lupini

comb. Nov. in modern terms. Mycologia 94:307-320

Pate JS, Williams W, Farrington P (1985) Lupin (Lupinus spp.) In: Summerfield,

R.J. and Roberts, E.H. (eds) Grain Legume Crops. Collins, London, pp37-72

Pearl HM, Nagai C, Moore PH, Steiger DL, Osgood RV, Ming R (2004)

Construction of a genetic map for arabica coffee. Theor Appl Genet 108:829-

835

Reiter RS, Williams JGK, Feldmann KA, Rafalski JA, Tingey SV, Scolnik PA

(1992) Global and local genome mapping in Arabidopsis thaliana by using

recombinant inbred lines and random amplified polymorphic DNAs. Proc

Natl Acad Sci USA 89:1477-1481

Rogers SO, Bendich AJ (1994) Extraction of Total Cellular DNA from Plants,

Algae and Fungi. In: Gelvin SB and Schilperoort RA (eds) Plant Molecular

Biology Manual, 2nd Edition. Kluwer Academic Press, Dordrecht, The

Netherlands, D1. pp1-8

Sengbusch von R (1930) Lupins low in bitter substances. Züchter 2:1-2


50

Stermitz FR, Belofsky GN, Ng D, Singer MC (1989) Quinolizidine alkaloids

obtained by Pedicularis semibarbata (Scrophulariaceae) from Lupinus

fulcratus (Leguminosae) fail to influence the specialist herbivore Euphydryas

editha (Lepidoptera). J Chem Ecol 15:2521-2530

Swiecicki W, Swiecicki WK (1995) Domestication and breeding of narrow-leafed

lupin (Lupinus angustifolius L.). J Appl Genet 36:155-167

Tanksley SD, Ganal MW, Prince JP, de Vicente MC, Bonierbale MW, Broun P,

Fulton TM, Giovanni JJ, Grandillo S, Martin GB, Messeguer R, Miller

JC, Miller L, Paterson AH, Pineda O, Röder MS, Wing RA, Wu W,

Young ND (1992) High density molecular maps of the tomato and potato

genomes. Genetics 132:1141-1160

Vallejos CE, Sakiyama NS, Chase CD (1992) A molecular marker-based linkage

map of Phaseolus vulgaris L. Genetics 131:733-740

Voorrips RE (2002) MapChart: Software for the graphical presentation of linkage

maps and QTLs. J Hered 93:177-178


Peleman J, Kuiper M, Zabeau M (1995) AFLP: A new technique for DNA

fingerprinting. Nucleic Acids Res 23:4407-4414

Wolko B, Weeden NF (1994) Linkage map of isozyme and RAPD markers for the

Lupinus angustifolius L., In: Neves-Martins, J.M. and Beirão da Costa, M.L.

(eds) Advances in Lupin Research. Proc 7th

Int Lupin Conf, Evora, Portugal.

ISA Press, Portugal. pp. 42-49




(Lupinus angustifolius L.). Mol Breed 14:145 - 151





Yang H, Sweetingham MW (1998) The taxonomy of Colletotrichum isolates

associated with lupin anthracnose. Aust J Agric Res 49:1213-1223


51





Yin TM, Wang XR, Andersson B, Lerceteau-Köhler E. (2003) Nearly complete

genetic maps of Pinus sylvestris L. (Scots pine) constructed by AFLP marker

analysis in a full-sib family. Theor Appl Genet 106:1075-1083




Lett 10:123-134

Vigouroux Y, Jaqueth JS, Matsuoka Y, Smith JSC, Doebley J (2002) Rate and

pattern of mutation at microsatellite loci in maize. Mol Biol Evol 19:1251-

1260

Zamir D, Tadmore Y (1986) Unequal segregation of nuclear genes in plants. Bot

Gaz 147:355-358

Chapter Three: Marker for Ku gene

53

Chapter Three

Development of a sequence-specific PCR marker linked to

the Ku gene which removes the vernalisation requirements

in narrow-leafed lupin.


54

3.1 INTRODUCTION

One of the climatic requirements of many wild narrow-leafed lupin (Lupinus

angustifolius L.) including those originating in the Mediterranean region is a period

of vernalisation to promote flowering (Gladstones 1970). Previous genetic studies

(Gladstones and Hill 1969) have shown that a single dominant gene Ku, originally

selected from a natural mutant of the Swedish cultivar Borre, had a major effect on

the initiation of flowering by removal of the need for vernalisation and advanced

flowering by 2 – 5 weeks (Gladstones and Hill 1969). This gene greatly increases

the flexibility of lupins as a crop in that it not only enables them to be grown in

warmer areas with insufficient cold to satisfy the vernalisation requirements, but also

allows them to be sown in areas with a short growing season as experienced in large

areas of the Western Australian wheat-belt (Gladstones and Hill 1969). Most

modern cultivars of L. angustifolius bred in Western Australia and Europe since 1971

carry this gene (Cowling 1999).

As a crop, L. angustifolius has a short history of only 80 years. Because the

gene pool in domesticated cultivars is very narrow, the inclusion of wild germplasm

into the lupin breeding program will give greater diversity and scope for increased

genetic gains by the inclusion of further desirable traits. However when a wild type

is crossed with a domesticated line it involves the risk of losing desirable

domestication genes if they are not actively selected for. Field based selection for

the Ku gene is time-consuming because a large number of progeny has to be

maintained until flowering prior to selection. Furthermore, it is impossible to

directly select plants with the homozygous early flowering genotype (KuKu) from

heterozygous early flowering individuals (Kuku) since both genotypes will exhibit

the early flowering phenotype in the field. The development of a molecular marker

linked to the Ku gene would greatly facilitate the selection of this important

domestication gene in lupin breeding programs by enabling selection at the F2 stage

and concomitant 75% reduction in plant populations in the same generation.

For marker-assisted selection (MAS) to be viable in a practical plant breeding

program, a molecular marker must possess the following characteristics (Eagles et al.

2001, Yang et al. 2004):

(1) The marker is closely linked to a gene of interest;


55

(2) The operation of the marker must be more cost-effective than conventional

glasshouse or field based selection;

(3) The marker must be amenable to large numbers of samples.

The MFLP (microsatellite-anchored fragment length polymorphism) method

is capable of producing DNA polymorphisms with high efficiency (Yang et al.

2001). Recently, MFLP markers were used to construct a molecular map of L.

angustifolius (Chapter 2). Several domestication genes were also mapped. However

MFLP markers cannot be used directly for MAS because they are too time-

consuming and expensive to implement. The aim of this research is to develop a

sequence-specific, simple PCR based marker linked to the Ku gene desirable for

MAS in lupin breeding.


3.2.1 Plant materials and phenotyping of Ku gene

An F8 recombinant inbred line (RIL) of a domestic x wild type (DxW)

population of Lupinus angustifolius was previously developed by the Department of

Agriculture and Food, Western Australia (DAFWA) using as parents lines P27255

(wild) and 83A:476 (domesticated). The resulting population from this cross has

been used to produce linkage maps for L. angustifolius (Chapter 2, Figure 2.1;

Nelson et al. 2006).

The parents and 106 RILs of this population were sown in the screen-house at

the beginning of winter as randomized plots each consisting of 3 plants. There were 3

replicates. Anthesis was determined as the first floret on the main stem of any one

plant having opened in a particular plot (Gladstones and Hill 1969, Rahman and

Gladstones 1972). The mean date for the 3 replicate plots was then used to place the

RILs and the parent lines into groups designated as either „kuku‟ (late flowering) or

„KuKu‟ (early flowering).

3.2.2 Marker development

The two parents and 89 RILs were subjected to MFLP tests involving 10

SSR-anchor primers each in combination with 16 MseI-primers (Table 2.1). MFLP

markers having the best correlation to the Ku phenotyping data were selected. DNA


56

fragments of candidate MFLP markers were isolated from MFLP gels, re-amplified

by PCR, ligated into plasmids “pGEM-T Easy Vector” (Promega) and cloned into E.

coli according to the manufacturer‟s instructions. Plasmid DNA with MFLP

fragment inserts were isolated from E. coli, and sequenced using the BigDye

Terminator system (Applied Biosystems) as described by Yang et al. (2002, 2004).

A pair of sequence-specific primers was designed flanking the insertion/deletion site

(Yang et al. 2002, 2004; You et al. 2005). Primers were designed so that the

annealing temperature was approximately 54C based on calculations using the

nearest-neighbour model (http://www.sigmaaldrich.com).

3.2.3 DNA amplification and sequencing gel electrophoresis

Marker DNA fragments were amplified in a10 l PCR consisting of 1.5 l of

the template DNA and with one of the sequence-specific primers labelled with -33

P

as previously described by Yang et al. (2001, 2002). The selective PCR products

were separated on a 5% polyacrylamide denaturing sequencing gel (7 M urea) using

a 38x50 cm (0.4 mm in thickness) Sequi-Gen GT sequencing cell (Bio-Rad). After

electrophoresis at 55W for about 3 h, the gel was dried under vacuum on a gel drier

(Model 583, Bio-Rad). Marker bands were detected by autoradiography (Yang et al.

2002, You et al. 2005) with overnight exposure of the X-ray film on the dried gel.

The converted sequence–specific marker was tested on 106 F8 RILs of the

population derived from the DxW cross. The marker score data and the Ku

phenotyping data on the 106 F8 RILs were merged and analysed by the program

MapManager (Manly et al. 2001) to determine the genetic linkage between the

marker and the Ku gene.

3.3 RESULTS

3.3.1 Phenotyping of Ku gene

In 2005 the domesticated parent (83A:476) being early flowering (Ku),

reached anthesis 72 d after sowing. The wild parent (P27255; ku) reached anthesis

97 d after sowing. Field-testing of 106 RILs of the marker population showed that

anthesis dates could be clustered into a number of smaller groups contiguous with the

parental values (Table 3.1). There was a 5 day interval between these two groups

during which no RILs initiated flowering. Anthesis dates ranged from 69 to 84 d

http://www.sigmaaldrich.com/


57

after sowing (KuKu) and 90 to 110 d after sowing (kuku) with the majority of RILs

flowering within 1 week of either parent. Combining these groups resulted in 53

RILs being designated as „Early Flowering‟ (Ku) and another 53 RILs being

designated as „Late Flowering‟ (ku) based on the method described by Rahman and

Gladstones (1972). The segregation of plants with the KuKu allele to plants with the

kuku allele fits perfectly with the expected 1:1 ratio for the F8 progeny.

Table 3.1. Phenotyping of the early flowering gene Ku on an F8 derived RIL

population from a cross 83A:476 (KuKu) x P27255 (kuku) in Lupinus angustifolius

L.

Flowering

Date (DAS

1)

65 –

69

70 -

742

75 -

79

80 -

84

85 -

89

90 -

94

95 -

993

100 -

104

105 -

110

Numbers of

RILS 1 7 33 12 0 14 24 12 3

Total

53 KuKu

53 kuku

1DAS = days after sowing;

2The domesticated parent 83A:476 flowered on day 72;

3The wild parent

P27255 flowered on day 97. There is a discontinuity between days 84 and 90 showing the boundary

between the two major groups based on the Ku gene.

3.3.2 DNA sequencing of the candidate MFLP marker

A dominant MFLP marker designated “DAWA428.290”, originally produced

during a mapping study (Chapter 2), co-segregated with the Ku phenotyping scores

of the RILs tested. This marker was selected as a candidate for development of a

sequence-specific marker tagging the Ku gene. Marker DAWA428.290 was

generated by SSR-anchor primer MF42 (5‟-GTCTAACAACAACAACAAC-3‟) in

combination with primer MseI-CAC (5‟- GATGAGTCCTGAGTAACAC-3‟).

DNA sequencing showed that the dominant marker DAWA428.290 was a

318 bp fragment with sequences of SSR-anchor MF42 and MseI-CAC at either end

as expected (Table 3.2). A sequence-specific primer “KuHMS1” (3‟-

AGACATACCT TGTATGCGG-5‟) was designed to replace primer MseI-CAC.

PCR amplification of this marker by primer KuHMS1 in combination with the SSR-

anchor primer MF42 and using as template the DNA from the two parents and 10

randomly selected RILs revealed a pair of co-dominant bands. Apart from the band


58

produced by the 298 bp fragment of plants with the KuKu allele, plants with the kuku

allele produced a longer DNA band fragment that migrated more slowly (Figure 3.1).

DNA fragments from the ku allele were sequenced and aligned with the sequence

from the Ku allele (Table 3.2). It was revealed that the ku allele fragments contained

a 7 bp insertion, as well as a SNP of an Adenosine (A) substituting the Guanine (G)

in the Ku allele (Table 3.2).

Table 3.2. DNA sequence of the MFLP marker DAWA428.290 showing primers

giving rise to sequence-specific markers KuHM1 and KuHSNP.

The 7 bp insertion/deletion mutation site, and the SNP mutation site (G/A) are shown in bold italic.

Ku allele

ku allele

1 GATGAGTCCT GAGTAACAC1G AGACATACCT TGTATGCGGT CCTCATTGCC

AGACATACCT TGTATGCGG2T CCTCATTGCC

Ku allele

ku allele

51 AGAAATCTCA GCTTCGCTCA TAGTG -- -- -- -- AACAC CCTTCCTTTG GGCTTAGGTT

AGAAATCTCA GCTTCGCTCA TAGTG GCTTATT AACAC CCTTCCTTTG GGCTTAGGTT

Ku allele

ku allele

101 TTGTCGCTTG AGCAGTATTC ACTTATTCCT TTTTCAGTTT AGGGCACATC

TTGTCGCTTG AACAGTATTC ACTTATTCC3T TTTTCAGTTT AGGGCACATC

Ku allele

ku allele

151 GGCTTGATGT GTCTAGGATG GTTGCAGTGA TAACATACAG TGCAGTTGAC

GGCTTGATGT GTCTAGGATG GTTGCAGTGA TAACATACAG TGCAGTTGAC

Ku allele

ku allele

201 ACTGAAGTAA TTTCCTTTGC ACTTCGAACA CTAGGGTGGC GTGTTGTTGT

ACTGAAGTAA TTTCCTTTGC ACTTCGAACA CTAGGGTGGC GTGTTGTTGT

Ku allele

ku allele

251 TAGATTTCCC AACCGGTAAA CTCATTTTGT TGTCGTTATT ATTGTTTTTG4

TAGATTTCCC AACCAGTAAA CTCATTTTGT TGTCGTTATT ATTGTTTTTG

Ku allele

ku allele

301 TTGTTGTTGT TGTTAGAC

TTGTTGTTGT TGTTAGAC5

1Primer MseI-CAC (5‟-GATGAGTCCTGAGTAACAC-3‟)

2 Primer KuHMS1 (3‟-AGACATACCTTGTATGCGG-5‟).

3 Primer KuHMS2 (3‟-GAACAGTATTCATTATTCC-5‟)

4 Annealing site of primer KuHSR1 (3‟-CAAAAACAATAATAACGACAAC-5‟)

5 Annealing site of SSR-anchor primer MF42 (5‟-GTCTAACAACAACAACAAC-3‟)


59

Figure 3.1. PCR amplification products of 12 RIL plants of Lupinus angustifolius

with sequence-specific primer KuHMS1 (3‟-AGACATACCT TGTATGCGG-5‟,

labelled with -33

P) in combination with SSR-anchor primer MF42 (5‟-

GTCTAACAACAACAACAAC-3‟).

Five plants having KuKu allele (298 bp band) are the domesticated parent 83A:476 (Lane 1), A13

(Lane 5), A14 (Lane 6), A15 (Lane 7), A17 (Lane 9). The 7 other plants including the wild parent

P27253 (Lane 2), A10 (Lane 3), A12 (Lane 4), A16 (Lane 8), A18 (Lane 10), A19 (Lane 11) and A21

(Lane 12) carry the kuku allele (305 bp band).

3.3.3 Marker “KuHM1”

Because the insertion/deletion site is within the SSR-MseI fragment, a second

sequence-specific primer, KuHSR1 (3‟-CAAAAACAATAATAACGACAAC-5‟)

was designed, flanking the mutation site on the side opposite to KuHMS1. The pair

of primers successfully converted the MFLP marker into a co-dominant, sequence-

specific, simple PCR based marker. We designate this marker as “KuHM1”. Marker

KuHM1 shows a 280 bp band for the Ku allele, and a 287 bp band for the ku allele

(Figure 3.2).


60

Figure 3.2. Screening molecular marker KuHM1 on the parents and 28 F8 derived

RILs of Lupinus angustifolius using PCR with sequence-specific primer pair

KuHMS1 (3‟-AGACATACCT TGTATGCGG-5‟) and KuHSR1 (3‟-

CAAAAACAATAATAACGACAAC-5‟).

The parent plant 83A:476 (Lane 1) and 15 RILs: A64 (Lane 6), A65 (Lane 7), A69 (Lane 10), A70

(Lane 11), A73 (Lane 13), A74 (Lane 14), A75 (Lane 15), A76 (Lane 16), A77 (Lane 17), A78 (Lane

18), A83 (Lane 23), A84 (Lane 24), A86 (Lane 26), A87 (Lane 27), A93 (Lane 30) carrying the KuKu

allele produced a 280 bp band. The wild type parent P27253 (Lane 2) and 13 RILs: A61 (Lane 3),

A62 (Lane 4), A63 (Lane 5), A66 (Lane 8), A67 (Lane 9), A71 (Lane 12), A79 (Lane 19), A80 (Lane

20), A81 (Lane 21), A82 (Lane 22), A85 (Lane 25), A88 (lane 28), A92 (Lane 29) carrying the kuku

allele produced a 287 bp band.

Marker KuHM1 was tested on the parents and 106 RILs of the DxW

population. The score of marker and plant phenotypes showed perfect correlation on

all the RILs. All 53 RILs with the Ku allele developed the shorter band (280bp), and

the 53 RILs carrying the ku allele developed the longer band (287bp). Linkage

analysis by computer program MapManager suggested that the genetic distance

between the marker and the gene is less then 0.5 cM if we assume one cross-over in

that interval in a population of 107 or more.

3.4 DISCUSSION

Anthesis dates for this marker population were clustered around the parental

values, but showed signs of segregation that could be indicative of two or more genes

being involved. Rahman and Gladstones (1972) found that flowering behaviour of

non-Ku L. angustifolius cultivars was dominated by their high vernalisation

requirement and, that anthesis was also accelerated by increased photoperiod.

Higher growing temperatures, although a factor in several species did not appear to

significantly alter the time to anthesis in L. angustifolius (Rahman and Gladstones


61

1972). It is therefore possible that in this population photo-period is also a factor

influencing anthesis. Nevertheless, since vernalisation has such a major effect

(Gladstones and Hill 1969, Rahman and Gladstones 1972) we could separate the

lines into two distinct groups, classifying RILs as either KuKu or kuku with

confidence.

DAWA428.290 was originally a dominant marker in MFLP fingerprinting.

However, after sequencing we were able to directly convert this MFLP candidate

marker into co-dominant marker KuHM1. Because the two alleles both gave

excellent DNA amplification when SSR-anchor primer MF42 is used in combination

with primer KuHMS1 (Figure 3.1), it suggests that there is no difference in DNA

sequences between the two alleles at the annealing site for primer SSR-anchor primer

MF42. It is possible that a mutation exists on the cutting site for restriction enzyme

MseI on the ku allele which prevented ligation of the MseI-adaptor (Yang et al.

2001), causing the MFLP to develop only a dominant band for the Ku allele.

Marker KuHM1 reported in this paper meets the requirements for a sequence-

specific, size based simple PCR marker of low-cost, making it potentially suitable for

routine MAS in lupin breeding (You et al. 2005). The marker bands can also be

amplified by PCR, using the one sequence-specific primer KuHMS1 with the SSR-

anchor primer MF42. However, replacement of the SSR-anchor primer with the

specific primer KuHSR1 increases the specificity and makes the PCR amplification

more robust giving a significant reduction in background signal (Figure 3.2).

The tight linkage of marker KuHM1 (derived from DAWA 428.290) to the

Ku gene is indicated by the perfect correlation between plant phenotype and the

marker score on the available RIL population, giving a distance estimate of less then

0.5 cM from the gene. However there were only 106 RILs available in this instance.

A more precise estimation of the genetic distance between the marker and the gene

would require further testing on a larger population. The usefulness of this marker

for practical lupin breeding needs to be further evaluated further by testing on a wide

range of wild accessions, breeder‟s lines and cultivars.

The co-dominant resolution of sequence-specific marker KuHM1 makes it

amenable to a simple PCR-based assay for the presence of the Ku gene in lupin, able

to distinguish homozygous lines from those heterozygous for the two alleles. This

feature should make it particularly useful for MAS in a high-throughput situation

(Kumar 1999). Furthermore, this marker could be used to probe a BAC library of


62

this species to walk the chromosome to the actual gene. Successfully locating and

sequencing the Ku gene will enable development of a „perfect‟ marker that targets

the active gene sequence (e.g. Bradbury et al. 2005).

Acknowledgements

The help of Mr. D. Renshaw and Mr. C. Smith (Department of Agriculture and Food,

Western Australia) in data collection is gratefully acknowledged.


63

3.5 REFERENCES

Bradbury LMT, Henry RJ, Jin Q, Renke RJ, Waters DLE (2005) A perfect

marker for fragrance genotyping in rice. Mol Breed 16:279-283



Bulletin 4365

Dergam JA, Paiva SR, Schaeffer CE, Godinho L, Vieira F (2002)

Phylegeography and RAPD-PCR variation in Hoplias malabaricus (Bloch,

1794) (Pisces, Teleostei) in southeastern Brazil. Genet Mol Biol 25:379-387

Eagles HA, Bariana HS, Ogbonnaya FC, Rebetzke GJ, Hollamby GJ, Henry

RJ, Henschke PH, Carter M (2001) Implementation of markers in

Australian wheat breeding. Aust J Agric Res 52:1349-1356




Husb 9:213-220

Kanazin V, Talbert H, See D, DeCamp P, Nevo E, Blake T (2002) Discovery and

assay of single-nuceotide polymorphisms in barley (Hordeum vulgare). Plant

Mol Biol 48:529-537

Koebner R, Summers R (2002) The impact of molecular markers on the wheat

breeding paradigm. Cell Mol Biol Lett 7:695-702

Kumar LS (1999) DNA markers in plant improvement: An overview. Biotech Adv

17:143-182

Manly KF, Cudmore RH Jr, Meer JM (2001) MapManager QTX, cross-platform

software for genetic mapping. Mammalian Gen 12:930-932

Nelson MN, Phan HTT, Ellwood SR, Moolhuijzen PM, Bellgard M, Hane J,

Williams A, O’Lone CE, Fosu-Nyarko J, Scobie M, Cakir M, Jones

MGK, Bellgard M, Książkiewicz M, Wolko B, Barker SJ, Oliver RP,

Cowling WA (2006) The first gene-based map of Lupinus angustifolius L. –

location of domestication genes and conserved synteny with Medicago

truncatula. Theor Appl Genet 113:225-238


64

Orita M, Iwahana H, Kanazawa H, Hayashi K, Seriya T (1989) Detection of

polymorphisms of human DNA by gel electrophoresis as single-strand

conformation polymorphisms. Proc Natl Acad Sci USA 86:2766-2770

Pusch W, Wurmbach JH, Thiele H, Kostrzewa M (2002) MALDI-TOF mass

spectrometry-based SNP genotyping. Pharmacogenomics 3:537-548


vernalization, photoperiod and temperature under controlled environment.



fingerprinting based on micro-satellite anchored fragment length



Yang H, Shankar M, Buirchell BJ, Sweetingham MW, Caminero C, Smith PMC

(2002) Development of molecular markers using MFLP linked to a gene

conferring resistance to Diaporthe toxica in narrow-leafed lupin (Lupinus

angustifolius L.). Theor Appl Genet 105:265-270








Lett 10:123-134

Chapter Four: Markers for Reduced Pod Shattering in Lupin (No. 1)

65

Chapter Four

Development of two sequence-specific PCR markers linked

to the le gene that reduces pod shattering in narrow-leafed

lupin (Lupinus angustifolius L.)


66

4.1 INTRODUCTION

An essential characteristic of a modern grain crop is the ability to retain its

seeds long enough to allow mechanical harvesting at full maturity. In the genus

Lupinus, this characteristic has long been present in two species – Lupinus albus L.,

an old crop species of the Mediterranean region, and L. mutabilis (sweet) which has

been cultivated in the Peruvian Andes for over 2000 years (Gladstones 1967).

Narrow-leaf lupin (Lupinus angustifolius L.) is the most widely cultivated

grain legume crop in Australia. Originally introduced into Western Australia (W.A.)

towards the end of the 19th

century (Gladstones 1994) as a green manure crop, these

plants were essentially wild. Wild types of L. angustifolius have seed pods that

shatter upon maturity, making harvesting very difficult. Research aimed at finding

plants with non-shattering pods was reported to have begun in Germany in 1929

(Hanelt 1960). They were successful in selecting a strain that had a reduced pod

shatter (von Sengbusch and Zimmermann 1937) apparently related to a large

reduction in the thickness of the pod wall (Atebekova 1958). Domestication of this

species in W.A. began in 1960 with the discovery of two natural mutant (recessive)

genes, lentus and tardus for reduced pod-shattering (Gladstone 1967) and, in 1961 a

gene for nil-vernalisation requirement and early flowering (Gladstones 1969).

One of the genes for reduced pod-shatter, known as tardus (ta), affects the

sclerenchyma strips of the dorsal and ventral pod seams, fusing the two halves to

such an extent that separation of the two halves is greatly impeded. It was

considered analogous to a gene previously found in L. luteus. The second gene

lentus (le) modified the orientation of the sclerified endocarp of the pod, resulting in

a reduction of torsional forces upon drying, and hence reduced pod shatter. This

gene was also associated with a reduction in the thickness of the pod wall, but not to

the extent of the previously found strain of von Sengbusch and Zimmerman (1937),

leading to the conclusion that the controlling genes were not the same (Gladstones

1967). This modification was associated with a change in internal pod pigmentation

that gave the immature pods a purplish tinge and the inside surface of mature pods a

bright yellowish-brown color (Gladstones 1967). In addition, this trait is also

associated with development of a reddish pigmentation within the stem of plants

older than two months. The first Australian cultivar to carry both these genes was


67

Uniharvest, released in 1971. Apart from Fest (released 1973) all Australian and

many European cultivars released since then carry both of these genes.

Since field based selection for the le gene requires that the F2 progeny be

maintained for at least 8 weeks from sowing to ensure that this trait is retained, it

would be more efficient to use molecular markers that could be used to evaluate

progeny within the first weeks of growth. Recently, two molecular genetic maps

have been produced in L. angustifolius on which le has been mapped (Chapter 2,

Figure 2.1; Nelson et al. 2006). No marker was located within 5 centiMorgans (cM)

of the le gene on the map of Nelson et al. (2006). Conversely, several markers on

my map (Chapter 2) were closely linked to the le gene. However, the MFLP markers

in this map are not implementable in practical lupin breeding as the multiple bands

produced by their primer combinations may cause confusion, makes it impossible to

multiplex several markers for high efficiency and, is too tedious and expensive to be

implemented in a lupin breeding program for screening large numbers of plants. The

aim of this research is to develop simple, sequence-specific PCR based markers

linked to the le gene desirable for marker assisted selection (MAS) in lupin breeding.


4.2.1 Plant materials and phenotyping of le gene

A population of F8 recombinant inbred lines (RILs) of a domestic x wild type

(DxW) cross of L. angustifolius was previously developed by the Department of

Agriculture and Food, Western Australia (DAFWA) using as parents lines P27255

(wild) and 83A:476 (domesticated). Part of the resulting RIL population has

previously been used to produce linkage maps for L. angustifolius (Chapter 2, Nelson

et al. 2006).

The full population consisting of two parents and 115 RILs, was sown in a

screen-house during the winter growing season with 25-30 plants of each RIL grown

in 1.5 meter rows. Plants were either rated as LeLe (wild, shattering; pigment absent)

or lele (domesticated, reduced shattering; pigment present) based on the method

described by Gladstones (1967).


68


The parents and 89 RILs were previously subject to MFLP tests involving 10

SSR-anchor primers each in combination with 16 MseI-primers (Table 2.1). Two

MFLP markers having the best correlation to the le phenotyping data were selected.

DNA fragments from candidate MFLP markers were isolated from MFLP gels, re-

amplified by PCR, ligated into plasmids “pGEM-T Easy Vector” (Promega) and

cloned into E. coli according to the manufacturer‟s instructions. Plasmid DNA with

MFLP fragment inserts were isolated from E. coli, and sequenced using the BigDye

Terminator system (Applied Biosystems). To ensure the accuracy of DNA

sequencing, at least five clones of E. coli were used to prepare the plasmid DNA

containing the inserts, and sequenced separately. A sequence-specific primer was

designed near the MseI-end internal to the SSR-MseI fragment. (Yang et al. 2002,

2004; You et al. 2005). Primers were designed so that the annealing temperature

was approximately 54C based on calculations using the nearest-neighbor model

(http://www.sigmaaldrich.com).

4.2.3 Testing of converted markers

Testing of the converted markers was achieved by PCR, using the sequence-

specific primer in combination with the SSR anchor-primer from which the original

MFLP polymorphism was produced (Yang et al. 2001, 2002). Marker DNA

fragments were amplified in a10 l PCR consisting of 1.5 l of the template DNA

(approximately 100 ng), 0.5 unit of Taq polymerase (Fisher Biotec, Perth), 5 pmol of

each primers, 67 mM Tris-HCl (pH8.8), 2 mM MgCl2, 16.6 mM (NH4)2SO4, 0.45%

Triton X-100, 4 g gelatin, and 0.2 mM dNTPs. The sequence-specific primers were

labeled with -33

P based on a previously reported method (Yang et al. 2001, 2002,

2004). PCR was performed on a thermocycler (Hybaid DNA Express) with each

cycle comprising 30 s at 94C, 30 s at the annealing temperature (see below), and 1

min at 72C. The annealing temperature of the first cycle was 60C, and decreased

0.7C in each subsequent cycle until the temperature reached 54C. The final 25

cycles used an annealing temperature of 54C. The selective PCR products were

separated on a 5% polyacrylamide denaturing sequencing gel (7 M urea) using a

38x50 cm (0.4 mm in thickness) Sequi-Gen GT sequencing cell (Bio-Rad). After

electrophoresis at 55W for about 3 h, the gel was dried on a gel drier (Model 583,



69

Bio-Rad). Marker bands were detected by autoradiography (Yang et al. 2002, You

et al. 2004) with overnight exposure of the X-ray film to the dried gel.

4.2.4 Confirmation of linkage

The two converted sequence–specific markers were tested on all 115 F8 RILs

of the population derived from the DxW cross. The marker score data and le

phenotype data of the 115 F8 RILs were merged and analysed using “MapManager”

(Manly et al. 2001) to determine the genetic linkage between the markers and the le

gene.

Further testing of the markers was carried out on all 23 Australian historical

and current cultivars and on 36 landrace accessions from the Australian Lupin

Collection, to examine correlation of their phenotype with the le markers‟ genotypes.

Landrace accessions were selected on the basis of their geographic origins.

4.3 RESULTS

4.3.1 Phenotyping of le gene

Phenotyping on the individual plants from each of the 115 RILs revealed that

55 RIL lines were homozygous for le (reduced shatter), and 60 lines were

homozygous for Le (shattering). The segregation of RILs with the lele genotype to

plants with the LeLe genotype fits the expected 1:1 ratio (2 = 0.217, P = 0.641) for a

single gene.

4.3.2 DNA sequencing of candidate MFLP markers

Among the markers originated from 153 sets of MFLPs during a mapping

study (Chapter 2, Figure 2.1), two dominant markers designated as “DAWA323.150”

and “DAWA468.290” were identified as candidate markers for development as

sequence-specific markers tagging the le gene.

Marker DAWA323.150 mapped as1.2 cM from the le gene in the map based

on the 89 RILs (Chapter 2), was present in RILs with lele genotype, but was absent

in RILs with LeLe genotype (Figure 4.1). DNA sequencing found that MFLP marker

DAWA323.150 is a 157 bp fragment including the sequence of SSR-anchor primer

MF51 (5‟-GGGAACAACAACAAC-3‟) and the primer MseI-CGA (5‟-

GATGAGTCCTGAGTAACGA-3‟) (Table 4.1).


70

Figure 4.1. MFLP fingerprinting on 20 RILs derived from a domesticated x wild

cross in Lupinus angustifolius generated by SSR-anchor primer MF51 in

combination with primer MseI-CGA.

Eight RILs having the LeLe genotype are (DxW) 58 (Lane 2), 60 (Lane 4), 65 (Lane 6), 66 (Lane 7),

70 (Lane 9), 73 (Lane 10), 77 (Lane 13), and, 85 (Lane 19). Eleven RILs: (DxW) 57 (Lane 1), 59

(Lane 3), 64 (Lane 5), 69 (Lane 8), 74 (Lane 11), 76 (Lane 12), 78 (Lane 14), 82 (Lane 16), 83 (Lane

17), 84 (Lane 18), and 86 (Lane 20) have the genotype lele. Note that all of the 20 plants except one

(RIL 80, Lane 15) showed correct correlation between the le phenotype and the marker genotype.

Arrow indicates the candidate MFLP marker linked to the le allele at 157 bp.


giving rise to sequence-specific marker LeM1.

le allele 1 GATGAGTCCT GAGTAACGA1A CCTACCATTT G

2CCTAAACAA TATATTGTTT

le allele 51 ACTGGTTGTT GTTGTTGTTC TTCTTCTTCT TCCTCTTCTT CCTCTTCTTC

le allele 101 TTCTTCTTCT TACCATTTGC CTAAACAATA TATTGTTTAC TGGTTGTTGT

le allele 151 TGTTCCC3

1Primer MseI-CGA (5‟-GATGAGTCCTGAGTAACGA-3‟) is in italic.

2Sequence-specific primer LeBMS1 (5‟- TTAACGAACCTACCATTTG-3‟).

Note: The first 4 bp in primer LeBMS1 corresponds with the recognition site of MseI (T/TAA), while

the first nucleotide (T) was replaced by G when the MseI-adaptor was ligated onto the restriction

fragments in MFLP (Yang et al. 2001).

3Annealing site of SSR-anchor primer MF51 (3‟- GGGAACAACAACAAC-5‟)


71

Marker DAWA468.290 was also present in plants with lele genotype but absent in

plants with LeLe genotype (Figure 4.2). This marker had not been shown in the map

(Figure 2.1) as it was also calculated to be 1.2 cM from the le gene, and therefore co-

segregated with marker DAWA323.150.

Figure 4.2. MFLP fingerprinting on 20 RILs in Lupinus angustifolius generated by

SSR-anchor primer MF42 in combination with primer MseI-CCA.

RILs portrayed in this figure are the same as in Figure 4.1. Arrow indicates the candidate MFLP

marker linked to the le allele at 284 bp.

DNA sequencing revealed that marker DAWA468.290 is a 284 bp fragment

including the anchor primer sequence MF42 (3‟- GTCTAACAACAACAACAAC-

5‟) and the sequence of primer MseI-CCA (5‟- GATGAGTCCTGAGTAACCA-3‟)

(Table 4.2).

4.3.2.1 Marker “LeM1”

Based on the DNA sequence of MFLP marker DAWA323.150, a sequence-

specific primer “LeMS1” (3‟-TTAACGAACCTACCATTTG -5‟) was designed near


72

the MseI end of the original MFLP marker (Table 4.1). The dominant MFLP marker

was successfully converted into a dominant simple PCR based marker by using the

sequence-specific primer LeBMS1 in combination with SSR-anchor primer MF51.

F8 RILs with the genotype lele developed the 126 bp marker band, while plants with

genotype LeLe did not develop this band (Figure 4.3). We designated this sequence-

specific dominant marker linked to the le gene as “LeM1”.

Figure 4.3. Screening molecular marker LeM1 on the parents and 28 F8 derived

RILs of Lupinus angustifolius using PCR with sequence-specific primer LeMS1 (5‟-

TTAACGAACCTACCATTTG-3‟) and SSR anchor primer MF51.

Lines with the lele genotype which developed a 126 bp band include the domesticated parent plant

83A:476 (Lane 1) and 21 RILs: (DxW) 67 (Lane 4), 69 (Lane 5), 71 (Lane 7), 74 (Lane 9), 76 (Lane

11), 78 (Lane 13), 79 (Lane 14), 82 (Lane 17), 83 (Lane 18), 84 (Lane 19), 86 (Lane 21), 90 (Lane

23), 93 (Lane 25), 94 (Lane 26), 97 (Lane 29) 98 (Lane 30). Plants with the LeLe genotype which did

not produce the marker band are the wild type parent P27253 (Lane 2) and 11 RILs: 66 (Lane 3), 70

(Lane 6), 73 (Lane 8), 75 (Lane 10), 77 (Lane 12), 81 (Lane 16), 85 (Lane 20), 89 (Lane 22), 92 (Lane

24), 95 (Lane 27), 96 (Lane 28). RIL DxW 80 (Lane 15) also produced no band despite carrying the

lele allele.

4.3.2.2 Marker “LeM2”

A sequence-specific primer “LeMS2” (3‟- AGAAAAAGATGAATGCACG -

5‟) was designed near the MseI end based on the sequence of MFLP marker

DAWA468.290 (Table 4.2). The marker was successfully converted into a simple

PCR based marker by using primer LeMS2 in combination with SSR-anchor primer

MF42. Plants with genotype lele produced a 204 bp marker band, but the marker

fragment is absent in plants with LeLe genotype (Figure 4.4). We designated this

marker as “LeM2”.


73


giving rise to sequence-specific markers LeM2.

le allele 1 GATGAGTCCT GAGTAACCA1 C TCATAGTTTA TAAACTCTCC AATTGTTTGT

le allele 51 TATTCTCATG TATTATATCC TATCTCACAA AGAAAAAGAT GAATGCACG2A

le allele 101 GTTTTTTAGT ACAAATTTCA TCCTACACTG AATCTGTTGA AATTGAAATG

le allele 151 AATATCACAA AACATTTTGC TGCAGTGTCG GTCTTGCTTT ACCATTCGTG

le allele 201 GTGGCAACAC TTACACGTCA AGCAAAGTCA TTGATGGATG CCCCACCTAC

le allele 251 TGTGATAAAA GCTTGGTTGT TGTTGTTGTT AGAC3

1Primer MseI-CCA (5‟-GATGAGTCCTGAGTAACCA-3‟).

2 Sequence-specific primer LeMS2 (5‟-AGAAAAAGATGAATGCACG-3‟).

3Annealing site of primer MF42 (3‟-GTCTAACAACAACAACAAC-5‟).

Figure 4.4. Screening molecular marker LeM2 on the parents and 28 F8 derived

RILs of Lupinus angustifolius using PCR with sequence-specific primer LeMS2 (3‟-

AGAAAAAGATGAATGCACG-5‟) and SSR anchor primer MF42.

The length of the band fragment produced here is 204 bp. RILs portrayed in this figure are the same

as in Figure 4.3.


Markers LeM1 and LeM2 were tested on the parents and 115 RILs of the

DxW population. Marker results and RIL genotypes showed imperfect correlation,

with LeM1 having six differences and, LeM2 having three differences between the

marker result and RIL genotypes. The three RILs which showed inconsistence

between phenotype and marker genotype of LeM2 also showed the inconsistency

with LeM1. Map distances were calculated by MapManager (Manly et al. 2001) to

be 1.3 cM (LeM2) and 2.6 cM (LeM1) from the gene (Figure 4.5).


74

Figure 4.5. Genetic linkage of the two molecular markers LeM1 and LeM2 and the

reduced-pod-shatter gene lentus (le) of Lupinus angustifolius, as analysed by

MapManager.

Distances given in centiMorgans (above) between adjacent points on the linkage map are non-

cumulative

Among the 23 Australian cultivars, only the first released cultivar Uniwhite,

carries the shattering alleles LeLe, and all the other 22 cultivars have the reduced-

shattering genotype lele. The 22 cultivars having the lele genotype were correctly

scored by the markers. However, the marker genotype and Le phenotype of

Uniwhite did not match (Table 4.3).

The Australian lupin collection holds many accessions of wild and landrace

L. angustifolius collected or received from a range of locations. Most of these

accessions have been rated as either „shedding‟ (shattering; no le or ta), „reduced

shedding‟ (carrying at least one gene reducing pod shatter) or „non-shedding‟

(presumably carrying at least 2 genes for the reduction of pod shatter). Of the 36

landraces, marker LeM1 designated only six accessions as Le („shedding‟

(shattering)) and LeM2 designated seven as Le, despite all but one of them having

previously been rated as Le („Shedding‟, Australian Lupin Collection database,

2005). Only one of the accessions was designated as Le by both markers. Similarly,

both markers designated as le the one accession listed as “non-shedding”, received

from Belarus (Table 4.4). Accessions correctly designated as Le or le were collected

(or received) from eight countries including Belarus, France, Greece, Italy, Morocco,

Portugal, Spain, Syria. Accessions from the other eight source countries including

Cyprus, Chile, Germany, India, Israel, Turkey, USA and the USSR were incorrectly

genotyped using the two markers. In total, 13 out of 36 (36%) land-race accessions

showed a positive correlation between marker score and pod-shatter genotypes for

one or both markers.


75

Table 4.3. Correlation of marker score and Le phenotype of 23 Australian cultivars

of Lupinus angustifolius.

Cultivars Year of

release

Lentus

Phenotype

Presence of LeM1

marker band

Presence of LeM2

marker band

Uniwhite 1967 LeLe + +

Uniharvest 1971 lele + +

Unicrop 1973 lele + +

Marri 1976 lele + +

Illyarrie 1979 lele + +

Yandee 1980 lele + +

Chittick 1982 lele + +

Danja 1986 lele + +

Geebung 1987 lele + +

Gungurru 1988 lele + +

Yorrel 1989 lele + +

Warrah 1989 lele + +

Merrit 1991 lele + +

Myallie 1995 lele + +

Kalya 1996 lele + +

Wonga 1996 lele + +

Belara 1997 lele + +

Tallerack 1997 lele + +

Tanjil 1998 lele + +

Moonah 1998 lele + +

Quilinock 1999 lele + +

Jindalee 2000 lele + +

Mandelup 2004 lele + +

Note: Both LeM1 and LeM2 are dominant markers, where the marker band is linked to the le allele.

The one cultivar incorrectly identified by the markers, Uniwhite, is presented in bold.

+ = band present

- = band absent


76

Table 4.4. Correlation of marker score and Le phenotype of 36 Wild and Landrace

accessions of L. angustifolius from the Western Australian Lupin Collection.

Landrace

(Perth No.)

Origin Lentus

Phenotype

Presence of LeM1

marker band

Presence of LeM2

marker band

P20650 Portugal LeLe + +

P20653 Portugal LeLe + +

P20711 Italy LeLe + +

P20712 Italy LeLe + -



P20729 Russia LeLe + +

P20730 Russia LeLe + +

P20733 Germany LeLe + +

P20736 Israel LeLe + +

P21517 Israel LeLe + +

P21624 Turkey LeLe + +

P21625 Syria LeLe + +

P22609 Turkey LeLe + +

P22665 Spain LeLe - +

P22666 Spain LeLe - +

P22669 Spain LeLe + +

P22820 Portugal LeLe + -


P23051 France LeLe + -

P24025 India LeLe + +

P26042 USA LeLe + +

P26045 Greece LeLe + +

P26235 Spain LeLe + -



77

Landrace

(Perth No.)

Origin Lentus

Phenotype

Presence of LeM1

marker band

Presence of LeM2

marker band



P26436 Greece LeLe + +

P26465 Greece LeLe - +


P26675 Cyprus LeLe + +

P27055 Belarus lele + +

P27253 Morocco LeLe - -

P27434 Syria LeLe - +

P27968 Greece LeLe - +

P28003 Chile LeLe + +

Note: Both LeM1 and LeM2 are dominant markers, where presence of the marker band is linked to

the le allele. Accessions for which the Le phenotype correlates with either or both of the two

markers are presented in bold.

+ = band present

- = band absent

4.4 DISCUSSION

We have successfully developed two sequence-specific PCR markers linked

to the Le gene in L. angustifolius in this study. Both markers are simple PCR based,

are cost-effective, and could be used for MAS in lupin breeding. Previously, in

Chapter 2, marker DAWA323.150 from which LeM1 was derived, was mapped as

being only 1.2cM from the gene based on 89 RILs. This distance has now been re-

calculated as being 2.6cM from the gene on the basis of a population of 115 RILs,

including the 89 RILs previously used by them. LeM2, which had not previously

been placed on the map, has been positioned mid-way between the Le gene and

LeM1 at 1.3 cM. The genetic distance between the marker and the gene determines

the accuracy of marker assisted selection when the marker is used in the breeding

program. Since markers LeM1 and LeM2 are both on the same side of the le gene,


78

and LeM2 is closer to the gene than LeM1, it would be appropriate to use marker

LeM2 in MAS if both markers are applicable.

For MAS, it would be ideal to develop so-called “perfect Markers” where the

DNA fragments of the markers are actually on the gene of interest, in which case the

linkage between the markers and the gene is almost unbreakable (Ellis et al. 2002).

However, perfect markers are very difficult to develop. Most of the molecular

markers being implemented in practical plant breeding programs are “imperfect”

where certain genetic distances exist between the markers and the gene of interest in

the chromosomes (Staub and Serquen 1996, Gupta et al. 1999). The two markers

linked to the Le gene in lupin reported in this paper, are like-wise “imperfect”. Over

long periods of time and in the crossing in plant breeding programs, genetic

recombination may occur between the marker locus and the gene locus, and

consequently plants not having the target gene may carry the marker DNA sequence,

and vice versa (Sharp et al. 2001, You et al. 2005).

In MAS, a molecular marker can be used to screen the progeny from a cross

only if the marker is polymorphic among the parents so that the desirable allele can

be distinguished from the undesirable ones (Eagles et al. 2001). Fortunately, all the

modern Australian cultivars having the lele genotype show the target bands for

markers LeM1 and LeM2. However, testing of these markers on land-races that

were fully shattering (having the LeLe,TaTa genotype) resulted in only 13 of the 36

landrace accessions correlating the marker bands with genotype. Eleven of those 13

accessions were correctly identified by either one or the other of the two markers,

indicative of the degree of crossing over that has occurred between the marker and

the le gene in the wild even though it only occurs at an estimated 1.3 – 2.6% of the

time in any given cross. The results suggest that these markers may be of use in

MAS to screen the progeny of approximately one third of wild accessions when they

are crossed with modern Australian cultivars. This study although not exhaustive,

also suggests that these markers would be more likely useful when testing progeny of

seed material originating from Greece, Morocco (source of DxW parent P27255),

Portugal, Spain and Syria, although seed received from other sources including

collections (e.g. France, Belarus) should not be ruled out.

When markers LeM1 and LeM2 are used to screen the F2 progenies from

DxW crosses, they enable breeders to eliminate the approximately 25% of plants

which are homozygous for the LeLe allele. However, since both markers are


79

dominant, they cannot differentiate homozygous lele F2 plants from the heterozygous

Lele plants. Therefore, further selection for the lele genotype may be necessary in

later generations of the breeding cycle.

4.5 REFERENCES

Atebekova AI (1958) The dehiscence of the pods in the genus Lupinus (Tourn.) L.

Bull Moscow Soc Nat Res 63:89

Eagles HA, Bariana HS, Ogbonnaya FC, Rebetzke GJ, Hollamby GJ, Henry

RJ, Henschke PH, Carter M (2001) Implementation of markers in

Australian wheat breeding. Aust J Agric Res 52:1349-1356

Ellis MH, Spielmeyer W, Gale KR, Rebetzke GJ, Richards RA (2002) “Perfect”

markers for the Rht-B1b and Rht-D1b dwarfing genes in wheat. Theor Appl

Genet 105:1038 – 1042



7:360 – 366


and L. digitatus. 2. Time of flowering. Aust J Exp Agric Anim Husb 9:213 -

220

Gladstones JS (1994) An historical review of lupins in Australia. In: M. Dracup, J

Palta (eds) Proceedings of the first Australian lupin technical symposium.

Department of Agriculture, Western Australia, pp 1 – 38

Gupta PK, Varshney PK, Sharma PC, Ramesh B (1999) Molecular markers and

their applications in wheat breeding. Plant Breeding 118:369-390

Hanelt P (1960) The Lupins. A Ziemsen, Wittenberg, 104 pp




80







Sengbusch von R, Zimmermann K (1937) Die Auffindung der ersten gelben und

blauen Lupinen (Lupinus luteus und Lupinus angustifolius) mit

nichtplatzenden Hülsen und die damit zusammenhängenden Probleme,

insbesondere die der Süsslupinenzüchtung. Züchter 9:57-65

Sharp PJ, Johnston S, Brown G, McIntosh RA, Pallotta M, Carter M, Bariana

HS, Khatkar S, Lagudah ES, Singh RP, Khairallah M, Potter R, Jones

MGK (2001) Validation of molecular markers for wheat breeding. Aust J

Agric Res 52:1357-1366

Staub JE, Serquen FC (1996) Genetic markers, map construction, and their

application in plant breeding. HortScience 31:729-740





Yang H, Shankar M, Buirchell BJ, Sweetingham MW, Caminero C, Smith,











Lett 10:123-134

Markers for Reduced Pod Shattering in Lupin (No. 2)

81

Chapter Five

Development of a sequence-specific PCR marker linked to

the tardus gene that reduces pod shattering in narrow-leafed

lupin (Lupinus angustifolius L.)


82

5.1 INTRODUCTION

Wild types of L. angustifolius have seed pods that shatter upon maturity,

making harvesting very difficult. Two major recessive genes which together remove

pod-shattering in narrow-leafed lupin were found by Gladstones and his co-workers

(1967) and subsequently incorporated into all cultivars of this crop species in

Australia. The first gene lentus (le) modified the orientation of the sclerified

endocarp of the pod, resulting in a reduction of torsional forces upon drying, and

hence reduced pod shatter. The second gene, known as tardus (ta), affects the

sclerenchyma strips of the dorsal and ventral pod seams and was considered

analogous to a gene previously found in L. luteus. It is difficult to phenotype and is

not easily selected for visually – especially when in the presence of le.

Since the expression of the ta gene is greatly influenced by environmental

factors (particularly the humidity and temperature), and is complicated by the

presence or absence of the le gene, accurate phenotyping of the ta gene is the most

difficult task when introgressing wild material in lupin breeding. The aim of this

research is to develop simple, sequence-specific PCR based markers linked to the ta

gene suitable for marker assisted selection (MAS) in lupin breeding.


5.2.1 Plant materials and phenotyping of the ta gene

A population of F8 recombinant inbred lines (RILs) of a domestic x wild type (DxW)

cross of L. angustifolius was previously developed by the Department of Agriculture

and Food, Western Australia (DAFWA) using as parents lines P27255 (wild) and

83A:476 (domesticated). This population has previously been used to produce

linkage maps for L. angustifolius (Chapter 2; Nelson et al. 2006) and also to develop

2 markers linked to the le gene (Chapter 4).

The full population consisting of two parents and 115 RILs, was sown over

two successive winter growing seasons in a screen-house and in the open field.

In the first year just three plants were grown of each RIL in a single row in

the screen-house. Plants were grown to physiological maturity and the (still green)

tops of the plants enclosed in perforated clear plastic bread bags (holes ~ 0.5mm

diameter, 25 / cm2) to prevent seed and pod loss and aid observations as the plants

and pods dried. The second year, 25-30 plants of each RIL were grown in 1.5m long


83

rows 50cm apart, with 3 replicates. Plants were grown to maturity and observations

taken without enclosing tops or pods.

5.2.1.1 Field observations

Because this population segregates for both reduced pod-shatter genes lentus

and tardus, all RILs were rated for the presence of le (Gladstones 1967, Chapter 4) in

green plants after 10 weeks on the basis of the presence (or absence) of a red stem

pigmentation.

This population is heterogeneous for maturity as a result of varying

vernalisation and day-length requirements for flower initiation. As pods and plants

matured, observations were made either every one or two days for pod shattering.

Daily observation was especially important in hot, dry conditions (≥ 35oC). Upon

observation of splitting or complete pod shatter in any particular RIL, both the

incidence of shatter and the plant‟s condition were recorded. All observations were

dated to ensure consistency in rating when plants and pods were dry and when

shattering occurred. Any rainfall events were recorded, and re-drying of plants and

pods carefully monitored.

At the occurrence of pod shatter plants were rated according to the following criteria:

Pod condition: Plant condition: Rating:

(a) recently dry (a) completely green, fully leaved 1 (fully shattering)

(b) dry (b) have slightly green stems 2

(c) just dry 3

(c) very dry; (d) very dry 4

5 (no shattering)

Note: Pods of plants carrying neither gene for reduced pod-shatter tend to shatter spontaneously when

the plants are still green and, in some cases fully leaved (rating = 1, 2). Plants carrying both genes for

reduced pod-shatter have greatly retarded or no pod-shatter at all (rating = 5). Those with only one of

the two genes tend to be intermediate to these two extremes (rating = 3, 4).


84

5.2.1.2 Laboratory observations

Pod samples were collected from the field by carefully cutting plant stems

just below the mature (dry) pods and placing them into paper bags for later

examination. Plant materials were allowed to dry at room temperature in the

laboratory for at least another month prior to a physical and microscopic examination

of the pods.

Pods were examined for:

(i) ease of splitting:

(a) splits readily or even spontaneously

(b) difficult to split by finger pressure to the seams.

(ii) microscope examination as described by Gladstones (1967)

(a) pods split neatly, with a smooth break around the seam

(b) splitting does not always follow the seam but goes outside of it and is

jagged.

Pods of plants carrying the le gene had a yellow colouration inside and exhibited

some resistance to shattering. Those carrying ta also exhibited resistance to

shattering. When pods were forcibly split by applying pressure to the sides, those

carrying ta tended to split in a jagged fashion whereas those carrying le only or

neither gene, did not.

5.2.2 Marker development and mapping

The parents and 89 RILs were previously subject to MFLP tests involving 10

SSR-anchor primers each in combination with 16 MseI-primers (Table 2.1). The

phenotyping (ta gene) data was inserted into the map and the fit examined by

MapManager version QTXb20 (Manly et al. 2001) using the „Tools, Report‟

function. One MFLP marker having the best correlation to the ta phenotyping data

was selected. DNA fragments from candidate MFLP markers were isolated from

MFLP gels, re-amplified by PCR, ligated into plasmids “pGEM-T Easy Vector”

(Promega) and cloned into E. coli according to the manufacturer‟s instructions.


85

Plasmid DNA with MFLP fragment inserts were isolated from E. coli, and sequenced

using the BigDye Terminator system (Applied Biosystems). To ensure the accuracy

of DNA sequencing, at least five clones of E. coli were used to prepare the plasmid

DNA containing the inserts, and sequenced separately. A sequence-specific primer

was designed near the MseI-end internal to the SSR-MseI fragment. (Yang et al.

2002, 2004; You et al. 2005). Primers were designed so that the annealing

temperature was approximately 54C based on calculations using the nearest-

neighbour model (http://www.sigmaaldrich.com).

5.2.3 Testing of converted markers

Testing of the converted marker was achieved by PCR, using the sequence-

specific primer in combination with the SSR anchor-primer from which the original

MFLP polymorphism was produced (Yang et al. 2001, 2002). Marker DNA

fragments were amplified in a10 l PCR consisting of 1.5 l of the template DNA

(approximately 100 ng), 0.5 unit of Taq polymerase (Fisher Biotec, Perth), 5 pmol of

each primers, 67 mM Tris-HCl (pH8.8), 2 mM MgCl2, 16.6 mM (NH4)2SO4, 0.45%

Triton X-100, 4 g gelatin, and 0.2 mM dNTPs. The sequence-specific primers were

labelled with -33

P based on a previously reported method (Yang et al. 2001, 2002,

2004). PCR was performed on a thermocycler (Hybaid DNA Express) with each

cycle comprising 30s at 94C, 30s at the annealing temperature (see below), and 1

min at 72C. The annealing temperature of the first cycle was 60C, and decreased

0.7C in each subsequent cycle until the temperature reached 54C. The final 25

cycles used an annealing temperature of 54C. The selective PCR products were

separated on a 5% polyacrylamide denaturing sequencing gel (7 M urea) using a

38x50 cm (0.4 mm in thickness) Sequi-Gen GT sequencing cell (Bio-Rad). After

electrophoresis at 55W for about 3 h, the gel was dried on a gel drier (Model 583,

Bio-Rad). Marker bands were detected by autoradiography (Yang et al. 2002, You

et al. 2004) with overnight exposure of the X-ray film to the dried gel.

5.2.3.1 Confirmation of linkage

The two converted sequence–specific markers were tested on all 115 F8 RILs

of the population derived from the DxW cross. The marker score data and ta

phenotype data were merged and analysed using “MapManager” (Manly et al. 2001)

to determine the genetic linkage between the markers and the ta gene.



86

Further testing of the markers was carried out on 18 Australian historical and

current cultivars and on 33 wild and landrace accessions from the Australian Lupin

Collection, to examine correlation of their phenotype with the ta marker‟s genotype.

Accessions were selected on the basis of their geographic origins.

5.3 RESULTS

5.3.1 Phenotyping of ta gene

The complexity of phenotyping for ta resulted us spending 2 seasons

characterising the ta gene in the RILs involving both field and laboratory pod rating.

In the majority (113 / 115) of cases there was a good correlation between the two

methods employed, although for one RIL (line 107) results for both methods over 2

years and, a second RIL (Line 110), with only one year‟s data, were inconclusive.

Neither of these 2 RILs had been used in development of the map in Chapter 2.

Phenotyping of the 113 RILs suggested that 45 RIL lines were homozygous

for ta (reduced shatter), and 68 lines were homozygous for Ta (shattering). The

segregation of RILs with the tata genotype to plants with the TaTa genotype does not

fit the expected 1:1 ratio (2 = 4.68, P = 0.03) for a single gene. Similar outcomes

were previously observed for segments of the genetic maps developed in Chapter 2

and by Nelson et al. (2006) using this marker population.

5.3.2 Map placement of ta and closely linked MFLP markers

Among the markers originated from 153 sets of MFLPs during a mapping

study (Chapter 2), three markers ((DAWA) 169.305, 978.325, 1097.300c) were

found to be closely linked to ta. However, none of these had been placed on the

published map. Examination of the combined map of Nelson et al. (unpublished)

determined that Tardus ought to be placed in linkage group 1 in the vicinity of

marker DAWA821.140 (Figure 2.1). Based on the mapped population of 89 RILs

and using MapManager QTX (Manly et al. 2001), all three markers were determined

to be on the one side of ta, with DAWA1097.300c the closest at 2.1cM, and the two

other markers at 2.8cM from the gene. Further testing of Marker DAWA169.305 on

114 RILs (of a possible 115) of this population did not alter the distance between the

gene and this marker, there being 5 mis-matches between marker and RIL genotypes.


87

The revised order of the original and new markers in this segment and, distances

between the markers is presented in Figure 5.1.

5.3.3 DNA sequencing of candidate MFLP marker

Marker DAWA169.305 was selected for conversion into a PCR-friendly

format. DNA sequencing revealed that marker DAWA169.305 is a 306 bp fragment

including the anchor primer sequence MF128 (3‟- AGTAGCTCTCTCTCTCTCTC -

5‟) and the sequence of primer MseI-CCT (5‟- GATGAGTCCTGAGTAACCT-3‟)

(Table 5.1). MF128 is a degenerate primer. Consequently the end of the sequence

varied between analyses.


giving rise to sequence-specific marker TaM1.

Ta allele 1 GATGAGTCCT GAGTAACTT1 AAATCATTAC TTCAAATTCA CCAATCCTCT

Ta allele 51 TATGCGAAAT TTTTGTTAGA TTGGGCCTAA ACAGAGGATT GCAAATC2GT

Ta allele 101 CACACATGTG CATTCACAGA ACCCAATGGA ATGATTAGTA GATTTGCATA

Ta allele 151 TGGTTGGATT TGAGATATCA CCACACAAAT AGTGAGTCAC ACGAGCTCCT

Ta allele 201 GATTAGGGGA ATCAGTTGAG GCTAACCCTG TTTTGCTGTT CCAAATGGGT

Ta allele 251 CAAAATTATA ATGCCAAAGG AGAATCAAAA TCATAGGAGA GAGAGAGAGA

Ta allele 301 ACCAAT3

1Primer MseI-CTT (5‟- GATGAGTCCTGAGTAA CTT -3‟).

2Sequence-specific primer TaMS1 (5‟-AACAGAGGATTGCAAATC-3‟).

3Annealing site of SSR-anchor primer MF128 (3‟- AGTAGCTCTCTCTCTCTCTC -5‟).

Note: MF128 is a degenerate primer. Consequently, the sequence from 301-306 does not correspond

fully to the reverse primer sequence. Sequence ACC (bp 301 -303) alternated with CTA and CCC in

5 otherwise identical sequences.


88

Figure 5.1. Marker sequence and placement of the tardus gene in linkage group 1 of

the linkage map developed in Chapter 2.

DAWA700.150c0.0

Tardus23.8DAWA1097.300c25.9DAWA169.30526.6DAWA80.19035.1DAWA289.630c35.7DAWA821.14037.0DAWA109.260c DAWA956.29538.3DAWA749.46040.9DAWA532.48041.5DAWA377.35042.1DAWA930.135c42.8DAWA809.47544.3DAWA758.47046.4

DAWA765.18067.2DAWA523.29069.0DAWA516.29069.6

DAWA498.31075.0

LG1

Note: This segment represents the upper part of LG1in Figure 2.1 including a number of extra

markers as well as a rearrangement of the order for a better fit. All markers of the original map in this

sector are included. From DAWA516.290 (at 31.1cM in original map) onwards the linkage group is

unchanged. The new total length of LG1 is 175.8cM.


89

5.3.3.1 Marker “TaM1”

Based on the DNA sequence of MFLP marker DAWA169.305, a sequence-

specific primer “TaMS2” (3‟- AACAGAGGATTGCAAATC -5‟) was designed near

the MseI end of the original MFLP marker (Table 5.1). The dominant MFLP marker

was successfully converted into a co-dominant simple PCR based marker by using

the sequence-specific primer TaMS2 in combination with SSR-anchor primer

MF128. F8 RILs with the genotype TaTa developed the faster (shorter) marker

band(s), while only plants with the genotype tata developed the slowest band with a

length of approximately 221bp. We designated this marker as “TaM1”. No attempt

was made to determine the sequence difference(s) between the bands, although a

visual inspection of the films suggests that the length difference between each of the

bands is 1bp.

Figure 5.2. Screening of molecular marker TaM1 on 20 wild and land-race

accessions from the Australian Lupin Collection (Perth, W.A.).

All of these accessions are known to lack the ta allele. Lines carrying the sequence associated with

the reduced shatter allele ta, are in lanes 1, 2, 4, 9, 10, 12, 13, 15, 17, 18, 20. Lines carrying the Ta

allele type 1 are in lanes 3, 11, 14. Lines carrying the Ta allele type 2 are in lanes 5, 8, 19. Lines

carrying the Ta allele type 3 are in lanes 6, 7. There is one line in lane 16 carrying the heterozygous

allele combination.


The converted marker TaM1 gave results matching those of DAWA169.305.

Testing of TaM1 on a selection of 18 Australian cultivars (Table 5.2) of narrow-leaf

lupin revealed no mis-matches. When tested on wild accessions (Figure 5.2), it was

found that there were three alleles for the wild genotype, in approximately equal

numbers. Of the 33 landraces and wild types (Table 5.3), marker TaM1 designated

14 accessions as Ta (“shattering”) and 1 as ta (“non-shattering”) correctly. Lines


90

correctly identified as carrying the Ta allele were from Cyprus (1), France (1),

Greece (1), India (1), Israel (1), Italy (3), Morocco (1), Portugal (1), Spain (2), Syria

(1), Turkey (1). The one line from Belarus was correctly identified as carrying the ta

allele. Three of these lines ((P) 22829, 26235, 27055) were previously also found to

be amenable to selection for the gene lentus, with marker LeM2 (Chapter 4).

Table 5.2. Correlation of marker score and Ta phenotype of 18 Australian cultivars

of L. angustifolius.

Cultivar Year of

release

Tardus

Phenotype

TaM1 marker

rating

Marri 1976 tata tata

Illyarrie 1979 tata tata

Yandee 1980 tata tata

Chittick 1982 tata tata

Danja 1986 tata tata

Geebung 1987 tata tata

Gungurru 1988 tata tata

Yorrel 1989 tata tata

Warrah 1989 tata tata

Myallie 1995 tata tata

Kalya 1996 tata tata

Wonga 1996 tata tata

Belara 1997 tata tata

Tallerack 1997 tata tata

Moonah 1998 tata tata

Quilinock 1999 tata tata

Jindalee 2000 tata tata

Coromup 2006 tata tata


91

Table 5.3. Correlation of marker score and Ta phenotype of 33 wild and landrace

accessions of Lupinus angustifolius from the Western Australian Lupin Collection.

Landrace

(Perth No.) Origin Tardus

Phenotype

TaM1 marker

Rating

P20650 Portugal Ta - tata

P20653 Portugal Ta - Tata

P20711 Italy Ta - TaTa (2)

P20712 Italy Ta - tata

P20714 Italy Ta - tata


P20730 Russia Ta - tata

P20733 Germany Ta - tata

P20736 Israel Ta - TaTa (2)

P21625 Syria Ta - TaTa (1)

P22609 Turkey Ta - TaTa (3)

P22665 Spain Ta - tata



P22724 Spain Ta - TaTa (1)


P22829 Portugal Ta - TaTa (2)

P22842 Morocco Ta - TaTa (3)

P23051 France Ta - TaTa (3)

P24025 India Ta - TaTa (2)

P26042 USA Ta - tata

P26045 Greece Ta - tata

P26235 Spain Ta - TaTa (1)



P26436 Greece Ta - TaTa (1)


92

Landrace

(Perth No.) Origin Tardus

Phenotype

TaM1 marker

Rating

P26465 Greece Ta - tata


P26675 Cyprus Ta - TaTa (3)

P27055 Belarus tata tata

P27253 Morocco Ta - tata

P27434 Syria Ta - tata

P28003 Chile Ta - tata

Note: Accessions for which the Ta phenotype correlates with the marker are presented in bold. For

TaM1 there are 3 different wild-type alleles. The allele identified for a particular line has been

identified in brackets following the marker genotype.

5.4 DISCUSSION

The reduced-shatter gene tardus is of major importance in the domesticated

narrow-leafed lupin, being one of the two major genes having a controlling effect on

the elimination of pre-harvest shattering in pods (Gladstones 1967). The successful

phenotyping for this trait in the presence of the second major gene lentus (le)

required careful examination of plants in the field as well as a microscopic

examination of the pods after harvest. In most cases the results were clear-cut, with

only one RIL (17) giving conflicting results between field observations and pod

examination data. The conflict was resolved in favour of the pod examination that

had indicated that ta was present. This was done on two grounds: (a) The genetic

map data suggested the pod result (ta) to be the most probable correct result, (b) It is

possible in certain genotypes that ta is present but does not eliminate pod-shatter

when in the presence of le (Gladstones, pers. comm.). This outcome has allowed the

gene to be mapped into Linkage Group 1 (LG1) of my map (Chapter 2) and that of

Nelson et al. (unpublished) and the identification of three closely linked markers that

had been previously generated, but not placed in my map (Chapter 2). These results

are also indicative of the involvement of at least one more (unknown) gene in the

reduction / elimination of pod shattering in narrow-leaf lupins.

The candidate marker “DAWA169.305” was a dominant marker in the MFLP

fingerprint linked to the wild allele Ta. However, after DNA sequencing, the marker


93

was converted into a co-dominant marker showing a band for both of the Ta and the

ta alleles but with a size difference. The successful DNA amplification of co-

dominant bands in the PCR indicates that both alleles have the identical annealing

sites for the primers TaMS2 and MF128. MF128 was also one of the two primers

involved in the generation of the original candidate marker DAWA169.305 in the

MFLP. It is likely that the DNA sequence of the ta allele has a mutation at the

restriction site which causes the enzyme Tru9I to be unable to recognise it. Hence

the MseI-adaptor could not be ligated onto the ta allele and the ta allele did not show

a corresponding band in the MFLP fingerprint.

The development of a sequence-specific co-dominant PCR marker linked to

the ta gene in L. angustifolius was successful. This marker correctly identified all

the 18 Australian cultivars tested as having the ta allele. Further testing of this

marker showed that it correctly identified the phenotype of 15 out of 33 lines

selected from the Australian lupin collection. However, in a further 7 cases no PCR

reaction product was produced. Of the 33 lines (above), one had been replicated but,

only one of the two replicate samples produced a PCR product, indicating a problem

in sample preparation to be the most probable cause for the failure of those PCRs.

Thus, on the basis of the 33 successful PCRs, marker TaM1 may be useful in as

many as 45.5% of potential crosses.

The random distribution of (positive) correlation between gene and marker

suggests that potential wild parents from any region have an equal chance of being

amenable to screening by TaM1 when in a cross with a domesticated parent. The

fact that 3 of these lines were found to be amenable to selection procedures with the

(now) existing markers for both lentus (Chapter 4) and tardus could make them good

candidates for a first attempt at introducing new genetic material into the narrow-leaf

lupin breeding program. Marker TaM1, being PCR based, is cost-effective and

would therefore being eminently suited to MAS in lupin breeding, but is limited in its

application.


94

Acknowledgements:

The help of Mr. Colin Smith (DAFWA) in preliminary phenotyping work

and, the advice of Dr John Gladstones on particular aspects of phenotyping, are

gratefully acknowledged.

5.5 REFERENCES

Cowling WA, Huyghe C, Swiecicki W (1998) Lupin Breeding. In: Gladstones JS,





7:360 – 366

Gupta PK, Varshney PK, Sharma PC, Ramesh B (1999) Molecular markers and

their applications in wheat breeding. Plant Breeding 118:369-390





MGK, Książkiewicz, Wolko B, Barker SJ, Oliver RP, Cowling WA

(2006) The first gene-based map of Lupinus angustifolius L. – location of

domestication genes and conserved synteny with Medicago truncatula.

Theor Appl Genet 113(2):225 – 238





Yang H, Shankar M, Buirchell BJ, Sweetingham MW, Caminero C, Smith,





95








Lett 10:123-134

Chapter Six: Development of a marker for seed dormancy

97

Chapter Six

Development of a PCR marker tightly linked to mollis, the

gene that controls seed dormancy in

Lupinus angustifolius L.


98

6.1 INTRODUCTION

High germination rates and uniform plant densities in the year of sowing are

paramount for a crop species. One of the survival adaptations of many plant species

including that of the wild narrow-leafed lupin (Lupinus angustifolius L.) is the

production of seeds with testa (seed coats) that are impermeable to water, inducing

dormancy and preventing germination of all seed in any one year (Quinlivan 1967,

Forbes and Wells 1968). This adaptation in lupins is also known as being „hard

seeded‟. Dormancy is a physical process that appears to be induced after maturity as

the seed moisture content drops below 14% and, is complete at between 9 and 11%

moisture content (Gladstones 1958). In the field, seed dormancy of L. cosentinii

(formerly known as L. varius) was found to gradually break down as a result of large

daily (summer) temperature fluctuations (Quinlivan 1966, 1968). Similar

observations have also been made for L. angustifolius. It was surmised that the

fluctuating temperatures caused the testa to crack in localised areas making it

permeable, allowing a viable seed to readily imbibe water and germinate (Quinlivan

1968). Seeds having testa permeable to water, promoting rapid germination of

mature seed upon sowing are therefore a desirable feature for lupins grown as a seed

crop.

A single recessive gene mollis which is of unknown origin (Mikolajczyk

1966, Forbes and Wells 1968) results in seeds developing a water-permeable testa at

maturity, although the precise mode of action of mollis remains unclear. Some

researchers have found that the seed of soft-seeded lines and cultivars have testa

similar in thickness to hard-seeded lines (Clements et al. 2005), whereas others

found an alteration in the shapes and size of especially the outer palisade and the

hourglass cell layers of lupin seed testa (Miao et al. 2001) and that of other legumes

(Lush and Evans 1980). All seed crops of domesticated narrow-leafed lupin carry

this gene, mollis (Gladstones 1960; Cowling 1999).

The germplasm pool of domesticated L. angustifolius is very narrow, largely

because of its recent domestication. It is becoming increasingly obvious that there is

a need to widen the genetic base to promote gains in yield, disease resistance and

quality. Although sizeable collections are held in Australia and elsewhere, the use of

this material has been limited because of the difficulty in retaining domestication

traits such as mollis. The development of molecular markers specifically designed to


99

identify the presence of such genes will greatly improve the rate at which new

material can be introduced into the breeding program by enabling screening of the F2

seedlings and selecting only individuals carrying the gene in a homozygous

condition. It is the aim of this research to develop a molecular marker tightly linked

to mollis, enabling rapid selection for this gene when doing intra-specific crosses

with wild accessions of L. angustifolius.


6.2.1 Plant materials

The main marker population used in this study was an F8 recombinant inbred

line (RIL) of a domestic x wild type (DxW) cross of L. angustifolius (population 1)

previously developed by the Department of Agriculture and Food Western Australia

(DAFWA), using as parents lines P27255 (wild) and 83A:476 (domesticated). The

same population has also been used to produce linkage maps for L. angustifolius

(Figure 2.1; Nelson et al. 2006). The full population of this cross (115 RILs) was

used in initial testing of the developed marker. Other plant material used to test the

validity of the developed marker included 16 F2 plants (population 2) of a cross

between the Australian cultivar „Wonga‟ and the wild type P27253 and 36 landrace

(L. angustifolius) accessions from the Australian Lupin Collection, plus 24 historical

and current Australian cultivars.

6.2.2 DNA extraction for marker testing

Samples of three leaflets per plant were placed into wells of a 96-deep well

plate (Axygen, Union City, CA, USA) containing 600µl extraction buffer (Raeder

and Broda 1985) and one 4-mm diameter steel ball bearing (Consolidated Bearing

Co Pty Ltd, Perth, Australia). The leaf material was homogenised on a 2000

Geno/Grinder mill (SPEX Certiprep, Metuchen, N J, USA) for 6 min at 1300

strokes/min. The homogenised leaf material was heated in a water bath at 80oC for

30 min. After cooling down by floating on cold tap water for 10 min and chilling at -

20oC for 5 min, 300µl cold (4

oC) ammonium acetate was added to each well.

Samples were mixed and placed in the -20oC freezer to cool for a further 8 min.

Plates were centrifuged for 15 min at 3600rpm (maximum speed) in a bench top

centrifuge (Eppendorf 5804). The upper aqueous (180µl) was transferred to a new


100

block containing 300µl extraction buffer per well and mixed. DNA was precipitated

by addition of 400µl isopropanol per well, mixed and left on the bench at room

temperature for 30 – 60 min before centrifuging at 3600 rpm for 15 min and

decanting the supernatant. Pellets were washed with 150µl 70% ethanol. Samples

were left on the bench top for 10 min, centrifuged 5 min (max speed), the liquid

decanted and air dried for 10 min. Pellets were dissolved in 100µl TE0.1 buffer

(10mM Tris, 0.1mM EDTA, pH 8).

6.2.3 Phenotyping for mollis

Plants were grown in the screen-house and seed produced oven-dried at 40oC

for two weeks before being phenotyped for mollis. Twenty seeds of each RIL and 30

seeds of each F2 plant were placed on moistened filter papers in Petri-plates at room

temperature and monitored for germination over a period of 10 days (Quinlivan,

1968). Seeds that failed to germinate were scarified to pierce the seed coat (Arrieta

et al. 1994) and monitored a further 10 days. Seeds that germinated without

scarification were designated as soft (moll / moll). Those that germinated only after

scarifying were designated as hard (Moll / -). Seed that failed to germinate after

scarifying were deemed non-viable. Similar procedures were followed when

phenotyping material obtained from the Australian Lupin Collection. When a mix of

hard and soft seeds were observed (population 2), the parent was deemed to be

heterozygous, since the seed testa is of the female parent (Arrieta et al. 1994)


The two parents and 89 RILs (population 1) were subject to MFLP tests

involving 10 SSR-anchor primers each in combination with 16 MseI-primers (Table

2.1). One dominant MFLP marker having the best correlation to the mollis

phenotyping data was selected. DNA fragments from the candidate MFLP marker

was excised from MFLP gels, re-amplified by PCR, ligated into plasmids “pGEM-T

Easy Vector” (Promega, Annandale, NSW, Australia) and cloned into E. coli for

single-copy amplification, according to the manufacturer‟s instructions. Plasmid

DNA having MFLP fragment inserts were isolated from E. coli (Yang et al. 2002),

and sequenced using the BigDye Terminator system (Applied Biosystems, Foster

City, CA, USA).


101

The dominant marker developed using primers MoAMS + MF43 also gave a

back-ground signal (see Results), necessitating extension of the sequenced DNA

beyond the SSR-end of the MseI-SSR fragment to enable design of a sequence-

specific primer replacing MF43. The method of DNA extension beyond the SSR-

end is the same as reported previously (You et al. 2004). Briefly, genomic DNA of

the two parental lines was digested with restriction enzyme HpaII (recognition site

5‟-C/CGG-3‟, not found in the initial sequence), followed by ligating the restriction

fragments with a modified MseI-adaptor (Vos et al. 1995) to fit the Csp6I restriction

sites (one string is 5‟- GACGATCAGTCCTGAA -3‟, and the other string is 5‟-

GCTTCAGGACTGAT -3‟). Specific fragment amplification by PCR was achieved

using the primer “HpaII-O” (5‟- CGATCAGTCCTGAACGG -3‟) in combination

with a sequence-specific primer “MoAMS” (5‟-

TAACATCAACAAGGTGAGAATC-3‟), designed from the previously sequenced

MseI-SSR fragment. The amplified fragments were cloned and sequenced. A new

sequence-specific primer, MoASR (5‟-GAAGCATTCGATGAATTC-3‟), was

designed to use in combination with MoAMS for amplifying the marker bands

tagging the gene Mollis by PCR. We designate this marker as “MoA”.

For routine screening of marker MoA, DNA fragments were amplified in a10

l PCR consisting of 1.5 l template DNA (approximately 100 ng), 0.5 unit of Taq

polymerase (Fisher Biotec, Perth, Australia), 5 pmol each of two sequence-specific

primers, 67 mM Tris-HCl (pH8.8), 2 mM MgCl2, 16.6 mM (NH4)2SO4, 0.45% Triton

X-100, 4 g gelatin, and 0.2 mM dNTPs. Primer MoASR was labelled with -33

P

based on a previously reported method (Yang et al. 2001, 2002). PCR was

performed on a thermocycler (Hybaid DNA Express) with each cycle comprising 30

s at 94C, 30 s at the annealing temperature (see below), and 1 min at 72C. The

annealing temperature of the first cycle was 60C, and decreased 0.7C in each

subsequent cycle until the temperature reached 54C. The final 25 cycles used an

annealing temperature of 54C. DNA samples were denatured at 94C for 3 minutes

before being loaded on an SSCP gel (which contains 6% acrylamide, 5% glycerol

and10% sucrose; Gonen et al. 1999) and resolved by electrophoresis at a constant

340 volts (about 12-14 mA) for 16 h at room temperature. Marker bands were

detected by autoradiography (Yang et al. 2002, You et al. 2004) with overnight

exposure of the X-ray film to the dried gel.


102


The converted sequence–specific markers were tested on the F8 derived

population consisting of 115 RILs (population 1). The marker score data and the

mollis phenotyping data of the RILs were compared and analysed by MapManager

QTX (Manly et al. 2001) to determine the genetic linkage between the marker and

mollis. The marker was further evaluated against all Australian cultivars, a range of

wild and land-race accessions of this species obtained from the Australian Lupin

Collection and a number of hard and soft-seeded accessions from several other lupin

species. In addition, 16 F2 seedlings were tested – primarily to demonstrate the SNP

marker‟s properties in a segregating population.

6.3 RESULTS

6.3.1 Phenotyping of mollis gene

Of the DxW population (total 115 RILs), 71 RILs were found to be hard

seeded (Moll) and 44 RILs were soft seeded (moll). The segregation of plants with

the moll - moll genotype to plants with the Moll - Moll genotype deviates

significantly from the expected 1:1 ratio (2 = 6.33916, P = 0.01181) for the F8 RIL

population.

6.3.2 DNA sequencing of the candidate MFLP marker

A dominant MFLP marker designated “DAWA561.180”, originally produced

during the mapping study (Chapter 2), co-segregated with the Moll (hard seeded)

phenotyping scores of the 89 RILs investigated (Figure 6.1). This marker was

selected as a candidate for development of a sequence-specific marker tagging the

mollis gene. Marker DAWA561.180 was generated by SSR-anchor primer MF43

(5‟- CCTCAAGAAGAAGAAGAAG -3‟) in combination with primer MseI-CAT

(5‟- GATGAGTCCTGAGTAACAT-3‟).

DNA sequencing showed that the dominant marker DAWA561.180 was a

179 bp fragment with a sequence of SSR-anchor MF43 and MseI-CAT at each end as

expected (Table 6.1). A sequence-specific primer “MoAMS” (5‟-

TAACATCAACAAGGTGAGAATC-3‟) was designed to replace primer MseI-CAT.

PCR amplification of this marker by primer MoAMS in combination with the SSR-


103

anchor primer MF43 using as template the DNA from the two parents and 10

randomly selected RILs resulted in amplification of a marker band (MoA) with the

same dominant band pattern (Moll) as DAWA561.180, but with a significant

background signal for the moll (soft seed) allele.

Figure 6.1. 33

P autoradiograph of MFLP fingerprinting on 20 RILs derived from a

domesticated x wild cross in Lupinus angustifolius generated by SSR-anchor primer

MF43 in combination with primer MseI-CAT.

MFLP PCR products were resolved on a 5% denaturing sequencing gel. A dominant marker with a

sequence length of 179 bp (arrowed) shows nine RILs with the Moll / Moll genotype in lanes 2, 4, 5,

6, 12, 14, 17, 18, 19. No band was generated for moll / moll genotypes.


104

Table 6.1. DNA sequence of the candidate MFLP marker DAWA561.180, and the

sequence extended beyond the MF43-end. DNA sequences of the marker fragments

MoA amplified by pair MoAMS and MoASR contains two SNPs shown in bold and

italic.

moll allele Moll allele

GATGAGTCCT GAGTAACAT1C AACAAGGTGA GAATCAAGGT CATAGTGAAA ATCATCATCA

GATGAGTCCT GAGTAACATC AACAAGGTGA GAATC2AAGGT CAAAGTGAAA ATCATCATCA


CTCCCGAATT GTTGGTGAAG TGTTTTACCG

AATGCTTCAC GGAAGAGATT GGAGACTCGA CTCCCGAATT GTTGGTGAAG TGTTTTACCG

AATGCTTCAC GGAAGAGATT GGAGACTCGA


CAGAAGTCAG AGTAAAGTTG AAGCTGCTTT TCGGCATCGT CTTCTTCTAC TTCGGAGTCA CAGAAGTCAG AGTAAAGTTG AAGCTGCTTT TCGGCATCGT CTTCTTCTTC TTCGGAG

3TCA


TGGTGTGGGG TGAGGAATTC ATCGAATGCT TC4

TGGTGTGGGG TGAGGAATTC ATCGAATGCT TC

1Primer MseI-CAT (3‟-GATGAGTCCTGAGTAACAT -5‟).

2Primer MoAMS (3‟-TAACATCAACAAGGTGAGAATC-5‟)

3Annealing site of primer MF43 (5‟-CCTCAAGAAGAAGAAGAAG-3‟). Note: The genomic

sequence does not completely match the primer sequence at the 3‟-end.

4Annealing site of primer MoASR (5‟-GAAGCATTCGATGAATTC-3‟)

Extension of the sequence beyond that of primer MF43, resulted in the

designing of a second sequence-specific primer “MoASR” (5‟-

GAAGCATTCGATGAATTC-3‟) (Table 6.1). The PCR amplification products by

primer pair MoAMS and MoASR are 199 bp in length both for the Moll allele and

for the moll allele, but with two SNPs (Table 6.1). SNP1 was within the original

sequence at 43 bp, internal to primer MoAMS. The second SNP (SNP2) was within

the sequence of primer MF43 at 169 bp. SNP1 was a base substitution of T

(domesticated, soft-seeded) for A (wild, hard-seeded). SNP 2 had the reverse

substitution of A (domesticated, soft) for T (wild, hard). The distance between the

two SNPs was 126 bp.

6.3.3 Co-dominant marker “MoA”

On an SSCP gel, PCR products amplified by primer pair MoAMS and

MoASR showed a co-dominant segregation with the band of the moll allele moving

faster than band of the Moll allele (Figure 6.2). We designate this marker as “MoA”.


105

Figure 6.2. SSCP gel screening of molecular marker MoA on 16 F2 sibs of a cross

between the cultivar Wonga and wild-type P27253 using a 33

P labelled PCR with

sequence-specific primers MoAMS5 and MoAB9SR5.

The length of the band fragment produced here is 212 bp. Plants in lanes 1, 5, 8 are soft seeded (moll

/ moll). Plants in lanes 2, 3, 4, 6, 7, 11, 14, 15 are heterozygous hard seeded (Moll / moll) and those in

lanes 9, 10, 12, 13, and 16 are homozygous hard seeded (Moll / Moll).


MoA was tested on 115 F8 RILs (population 1), confirming that the gene and

marker are very tightly linked. Linkage analysis by computer program MapManager

QTX (Manly et al. 2001) suggested that the genetic distance between the marker and

the gene is less then 0.4 cM if we assume one cross-over in that interval in a

population of 116 or more.

Further testing of the marker on 24 Australian cultivars and on 35 wild and

Landrace accessions from the Australian Lupin Collection confirmed the score of the

marker bands to be completely consistent with the phenotype for all of these lines

(Tables 6.2, 6.3), although we found 3 alleles for the hard-seeded phenotype (data

not presented). When MoA was tested on several other Lupin species including L.

albus and L. luteus, in some instances polymorphisms did develop but they appeared

to have no correlation to soft-seededness (data not presented).


106

Table 6.2. Plant phenotype and Marker genotype scores of 24 Australian cultivars of

Lupinus angustifolius showing perfect correlation for Mollis.

Cultivar Year of

release

Plant

Phenotype

Marker

Genotype

Uniwhite 1967 moll moll

Uniharvest 1971 moll moll

Unicrop 1973 moll moll

Marri 1976 moll moll

Illyarrie 1979 moll moll

Yandee 1980 moll moll

Chittick 1982 moll moll

Danja 1986 moll moll

Geebung 1987 moll moll

Gungurru 1988 moll moll

Yorrel 1989 moll moll

Warrah 1989 moll moll

Merrit 1991 moll moll

Myallie 1995 moll moll

Kalya 1996 moll moll

Wonga 1996 moll moll

Belara 1997 moll moll

Tallerack 1997 moll moll

Tanjil 1998 moll moll

Moonah 1998 moll moll

Quilinock 1999 moll moll

Jindalee 2000 moll moll

Mandelup 2004 moll moll

Coromup 2006 moll moll

Moll = „hard‟ seeded

moll = „soft‟ seeded


107

Table 6.3. Plant phenotype and marker genotype scores of 35 wild and Landrace

accessions of L. angustifolius from the Western Australian Lupin Collection showing

perfect correlation for Mollis.

Landrace

(Perth No.)

Origin Plant

Phenotype

Marker

Genotype

P20650 Portugal Moll Moll

P20653 Portugal moll moll

P20711 Italy moll moll

P20712 Italy Moll Moll



P20729 Russia moll moll

P20730 Russia moll moll

P20733 Germany moll moll

P20736 Israel Moll Moll

P21517 Israel moll moll

P21624 Turkey Moll Moll

P21625 Syria Moll Moll

P22609 Turkey Moll Moll

P22665 Spain Moll Moll





P23051 France Moll Moll

P24025 India Moll Moll

P26042 USA moll moll

P26045 Greece moll moll




108

Landrace

(Perth No.)

Origin

Plant

Phenotype

Marker

Genotype



P26436 Greece Moll Moll

P26465 Greece Moll Moll


P26675 Cyprus Moll Moll

P27055 Belarus moll moll

P27253 Morocco Moll Moll

P27434 Syria Moll Moll

P28003 Chile moll moll

Note: Moll = „hard‟ seeded

moll = „soft‟ seeded

6.4 DISCUSSION

The MFLP marker (DAWA 561.180) identified as being linked to the gene

controlling seed dormancy (mollis) of L. angustifolius in our material was found on

the basis of an examination of the genetic map (Figure 2.1). Phenotyping of this RIL

population for mollis, gave a somewhat anomalous result in which nearly 2/3 of the

population was found to be hard-seeded (Moll). This population was previously

found to segregate in a skewed fashion for mollis by Nelson et al. (2006), although

no satisfactory reason could be given for this anomaly other then the small

population size.

For routine screening, a dominant marker has the disadvantages of being

unable to differentiate between homozygous moll-moll plants from heterozygous

Moll-moll individuals. By extending the original sequence we were able to convert

the marker into co-dominant form based on the presence / absence of two SNPs, and

especially SNP number 2 (Table 6.1). This SNP (number 2) was only found in plants

carrying the allele for soft seeds (moll). SNP 1 is a reversal of SNP 2 but it appears

unlikely that this reversal is significant as testing of MoA on a highly diverse

population of wild types and Landraces revealed the presence of at least 3 alleles for


109

hard-seed, possibly the result of alternative SNPs or even a deletion event within the

same sequence, although only the one allele was found for soft-seededness (moll-

moll) (data not presented). Consequently, the SNP marker could be successfully

implemented as a co-dominant by SSCP gel electrophoresis, capable of

differentiating heterozygous Moll-moll plants from homozygous moll-moll plants.

MoA correctly identified all tested accessions of L. angustifolius from the

Western Australian Lupin Collection, including at least 2 lines previously incorrectly

recorded as being hard-seeded. This indicates that linkage of this marker to mollis is

very tight. However, the failure of this marker when tested on other lupin species

may be an indication that it is not located within the gene sequence.

In conclusion; marker MoA is very tightly linked to the gene mollis in L.

angustifolius, with no sign found of the linkage having been broken in this species

across a diverse selection of both cultivated and wild types. It will be of great value

when introgressing wild material into the domesticated crop.


110

6.5 REFERENCES

Arrieta V, Besga G, Cordero, S (1994) Seed coat permeability and its inheritance

in a forage lupin (lupinus hispanicus). Euphytica 75:173-177

Clements JC, Dracup M, Buirchell BJ, Smith CG (2005) Variation for seed coat

and pod wall percentage and other traits in a germplasm collection and

historical cultivars of lupins. Aust J Agric Res 56:75-83



Bulletin 4365


angustifolius L.: Inheritance and phenotype classification. Crop Sci 8: 195-

197




Gladstones JS (1960) Lupin cultivation and breeding. J Aust Inst Agric Sci 26:19-25

Gonen D, Veenstra-VanderWeele J, Leventhal B, Cooke EH (1999) High

through-put fluorescent CE-SSCP SNP genotyping. Mol Psych 4:339-343

Lush WM, Evans LT (1980) The seed coats of cowpeas and other grain legumes:

Structure in relation to function. Fld Crop Res 3:267-286

Manly KF, Cudmore RH Jr, Meer JM (2001) MapManager QTX, cross-platform


Miao ZH, Fortune JA, Gallagher J (2001) Anatomical structure and nutritive value

of lupin seed coats. Aust J Agric Res 52:985-993




111






truncatula. Theor Appl Genet 113:225–238



1022





Raeder U, Broda P (1985) Rapid preparation of DNA from filamentous fungi. Lett

Appl Microbiol 1:17-20


Peleman J, Kuiper M, Zabeau M (1995) AFLP: A new technique for DNA

fingerprinting. Nucleic Acids Res. 23:4407-4414








Lett 10:123-134

Chapter Seven: QTL analysis of narrow-leaf lupin

113

Chapter Seven

Identification of quantitative trait loci (QTLs) influencing

early vigour, height, flowering date and seed size and their

implications for breeding of narrow-leafed lupin (Lupinus

angustifolius L.)


114

7.1 INTRODUCTION Narrow-leafed lupin (Lupinus angustifolius L.) is a relatively new crop

species. In the18th and 19th centuries it was recorded as being used as a coffee

substitute and as cattle fodder (Gladstones 1970), emerging in the 20th century as a

food seed crop, primarily for stock feed (Edwards and van Barneveld 1998), but also

for human consumption (Petterson 1998). Active domestication and breeding of this

crop started soon after World War I in Germany and Poland, with an emphasis on

early vigour and maturity (Barbacki 1952, Hanelt 1960) and low alkaloids (von

Sengbusch 1930, 1931, 1938). Later efforts were aimed at developing soft-seeded

lines (Gladstones 1958; Quinlivan 1966, 1967, 1968), early flowering and, further

reductions in pod shattering (Gladstones 1967, 1977). Most of these traits were

found to be under the control of one or two major genes, and could be selected for

relatively easily upon incorporation into the domesticated crop. Today breeding of

narrow-leaf lupins is increasingly being focused on more complex traits – especially

yield, but also seed quality and resistance to pests and diseases.

In nature, many complex plant traits such as grain yield, display a large

continuous range in variation approximating a Normal curve. The continuous

variation in phenotypic expression of such traits frequently cannot be fitted to simple

Mendelian ratios and are therefore generally considered to be the product of the

interaction of a number of gene loci (Johannsen 1909, Nilson-Ehle 1909, East 1916),

commonly known as quantitative trait loci (QTLs).

Early vigour, plant height and yield are complex traits of which little are

known in lupins. The time to flowering too is controlled by more then one gene,

although one gene (Ku) was found to have the greatest impact in the Western

Australian environment (Rahman and Gladstones 1972). Farrington and Gladstones

(1974) carried out some work on yield, primarily in assessment of the effect of

recently incorporated domestication genes. There are also some reports of the effect

of the iucundis gene (for low alkaloids) on yield, dating from the late 1930’s (von

Sengbusch 1938, 1942; Hackbarth and Troll 1960, Kress 1964). Recently, Dr J.

Clements (pers. comm.) has carried out work on the early vigour of a range of

accessions from the Australian Lupin Collection, looking at early biomass

accumulation in relation to yield. But on the whole, analysis of such traits has been


115

ignored because of a lack of suitable technology for such an assessment and, the

expression of these genes is usually further complicated by environmental and gene x

gene or epistatic interactions (Young 1996).

The development of molecular genetic maps now allows researchers to locate

QTLs on a genetic map as first achieved by Paterson et al. (1988) and, make detailed

analysis of their inheritance and activity. Provided that the map density and the

study population are adequate, this can lead to improved prediction and selection of

superior genotypes. One computer program that provides this option is QTL

Network version 2.0 (Yang et al. 2005).

A further benefit of mapping QTLs is in exploiting synteny (conservation of

linkage). It is not uncommon for related species to have near identical coding

sequences for homologous genes (Moore et al. 1995), albeit not necessarily in

identical positions on the genome. By locating and sequencing a gene of interest it

may be possible to exploit synteny to locate that gene in a related species of interest.

Recently two molecular maps have been published for narrow-leafed lupin

(Lupinus angustifolius L), the latter one also determining regions of synteny with the

model legume species, Medicago truncatula. Genome coverage is estimated to be

80% at a moderate to high density (Chapter 2; Nelson et al. 2006). These maps

allow additional genes including QTLs to be mapped with a high degree of accuracy,

opening up possibilities for marker-assisted selection in the breeding program and

the future exploitation of any synteny discovered.

In this study, we examined the genetic basis of quantitative variation in early

plant vigour, plant height at maturity, days to flowering and seed weight. QTLs

found were positioned on the published maps and regions of micro-synteny with M.

truncatula identified.

7.2 MATERIALS AND METHODS 7.2.1 Plant materials

The marker population used in this study was an F8 recombinant inbred line

(RIL) of a domestic x wild type (DxW) cross of L. angustifolius previously

developed by the Department of Agriculture and Food Western Australia (DAFWA),

using as parents lines 83A:476 (domesticated) and P27255 (wild). The same 89


116

RILs that had been used in the mapping study (Chapter 2) were grown over four

years.

7.2.2 Genetic markers Several new markers were added to the map developed in Chapter 2. The

protocols and naming details of all but one are as described in that chapter. One

marker, mtmt_GEN_00024_04_1 (referred to as mtmtGEN00024041 from here on)

is of a type not previously described. Details are as follows:

The primers for marker mtmtGEN00024041 were designed using M.

truncatula databases. Primer sequences are: (mtmtGEN00024041_L:

TTGGTGATGGATGCTGTTGT; …………………………. mtmtGEN00024041_R:

CATCGTCATCTGTGTGACCC). PCR volumes were 20µl, including 25ng

template DNA, 0.25µM of each primer, 1.25 U Taq DNA polymerase, 1.5mM

MgCl2, 2µl 1x PCR buffer (Invitrogen), 0.25 µM dNTPs (Fermentas), and 2 µg BSA

(Sigma). The marker was amplified by PCR on an MJ Research PTC-200

thermocycler over 34 cycles after an initial denaturation at 95oC for 3 min, with each

cycle consisting of: 94oC for 20 s, 53.6oC for 20 s, 72oC for 2 min. The final

extension was 5 min at 72oC. PCR products were separated in a 2% agarose gel

(19.8 x 25 cm) at 120V over 4 h alongside a 100bp ladder (Fermentas No. SM0623).

They were visualised by staining with Ethidium Bromide and exposure to a U.V.

light source.

Segregation of all new markers were observed in the same RIL population as

described in the above reference and the markers placed into the map of Chapter 2

(Figure 2.1) using MapManager QTX (Manly et al. 2001).

7.2.3 Plant measurements Data was collected for two consecutive years - 2005, 2006 (winter sown) in a

screen-house and the adjacent field respectively as randomised plots, each with three

replicates. To ensure rapid germination, all hard-seeded lines were scarified prior to

sowing. In 2005, plots were arranged into three continuous rows with each plot

consisting of 3 plants in a 10cm interval. In 2006, plots were grown as individual

rows 1.5m long consisting of ~30 plants, with rows spaced at 50cm intervals. The

sowing date in 2006 was 2 calendar days later then in 2005. Plots were irrigated as

required.


117

Plant heights were measured to the nearest 0.5cm at 10 weeks after sowing

(early vigour) and, to the nearest 1cm at maturity just prior to senescence. Three

representative plants were measured for each plot.

The date on which the first floret was fully opened was taken as the date of

flowering for that particular plot.

Seed produced in 2005 and 2006 were not harvested. Instead, seed weights

for each RIL were determined as 100 seed weights (g) of seed produced in screen-

house plots during 2002 and 2003 under similar growing conditions.

7.2.4 Data analysis Initial data analysis carried out included a simple regression analysis giving

means and standard deviations for each trait and year. Further analysis was carried

out using the program QTL Network 2.0 (Yang et al. 2005) as detailed below.

The map depicted in Figure 2.1 was used to place the QTLs, the average

marker interval on this map being 3.2cM. Marker intervals were also identified on

the map by Nelson et al. (2006) in instances where synteny had been identified with

M. truncatula. In comparing the maps we also made use of a composite map created

by merging raw data from both original maps (Nelson et al. un-published).

Mapping data of Figure 2.1 was imported into the program QTL Network 2.0

from MapManager. Input of trait data was as 2 environments (years) and 3 replicates

(except seed weights). Seed weights based on 100 seeds per RIL were analysed

separately for 2 environments, but without replication. The significance threshold

chosen for declaring a putative QTL was at P = 0.05. Data was analysed at 1cM

intervals (‘walk speed’) with 1000 permutations. Putative QTLs were separated by a

minimum of 10cM (‘filtration window’) before a decision was made that more then 1

QTL was present in any one linkage group.

QTL Network 2.0 analyses the data as multi-factorial matrices using a Mixed

Model Composite Interval Mapping (MCIM) approach (Yang & Zhu 2005) to give

the following outputs for a RIL population:

(i) Population means and variances including Genetic (G), Environmental (E)

and GxE interaction variances.

(ii) QTL positions including range, Standard Error(s) (SE), Additive gene effect

(A), Additive x Environmental (AxE) effects and the Probability (P)-value of

each.


118

(iii) Epistatic gene positions and intervals including nomination of genes acted

upon.

(iv) Heritability estimates (narrow-sense)

(v) Genotype of a predicted superior offspring.

To confirm the veracity of detected epistasis’ and the predicted superior genotypes,

we conducted a further analysis on the early vigour of plants carrying the 4 possible

gene combinations of the epistatic gene pair located on LGs 17 (CQTL 3) and 23

using the ‘fit-model’ function of the JMP statistic software (SAS Institute).

7.3 RESULTS 7.3.1 Marker mtmtGEN00024041

The amplified Marker mtmtGEN00024041 product was found to segregate in

a dominant fashion, with RILs carrying the maternal allele showing a band with a

length of approximately 950bp.

7.3.2 Genetic maps QTL data was mapped to the previously developed map (Figure 2.1) and

compared to the one by Nelson et al. (2006). Eleven new markers including

mtmtGEN00024041 were added to the map of Figure 2.1 as two linkage groups (22,

23) with a combined length of 62.7cM (Figure 7.1) increasing the total map length to

1605.7cM. These two groups were found to carry QTLs for early vigour and time to

flowering (Table 7.2).

Several QTL regions were found to have both an additive and an epistatic

effect, or were active in more then one trait. We have designated such QTLs as a

‘Common QTL’ (CQTL) and numbered these from 1 – 5. In three instances it was

unclear whether the same QTL was involved or one that was closely linked to

another on the same linkage group. These were marked with an asterisk (Table 7.2).


119

Figure 7.1. Linkage groups 22 (L) and 23 (R) showing markers (RHS) in order and

cumulative distances (LHS) in centiMorgans.

7.3.3 Phenotypic variation of traits Table 7.1 shows parental, mean, minimum, maximum and standard deviation

of trait values over 2 years for early vigour, height at maturity, days to flowering and

seed weight. The traits measured in the present study mostly exhibit continuous

variation (Figure 7.2), indicative that they are under the influence of multiple genes,

with only days to flowering being clearly divisible into two major groups clustered

around the parental values. All traits showed transgressive segregation, indicating

that alleles with positive effects were distributed among the parents. There was no

obvious increase in early vigour over the domesticated parent, although there were

some RILs exhibiting a reduction in early vigour when compared to the wild parent.

Seed weight data were not as complete and more variable, with data for the wild

parent only available for year 1 (2002).

DAWA395.040c0.0

DAWA346.0854.0

DAWA358.17520.7DAWA151.11521.3

DAWA1094.4500.0

mtmtGEN000240415.9

DAWA729.200c22.5

DAWA252.14825.6

DAWA218.29028.1DAWA701.50030.0

DAWA103.05041.4


120

Figure 7.2. Character distribution for Early Vigour, Height at Maturity, Days

to Flowering and Seed Weight. Data are grouped into intervals, based on two

years results. Note: Early Vigour is measured as plant height at 10 weeks after sowing; Seed weights are

dry weights of mature, harvested seed.

Parent A = 84A:476 (domesticated)

Parent B = P27255 (wild)


121

Table 7.1. Phenotypic variations of the four investigated traits in the mapping population and their parents over the four growing seasons (2002, 2003, 2005, 2006)

Genotype Trial Year1

Trait E. Vigour

(cm) Mature Ht. 2

(cm) DTF 3 Seed Wt 4

(g)

Parents 84A:476 (A) 1 48.6 87.1 72 16.3 2 44.8 86.2 73 13.8 P27255 (B) 1 22.6 79.3 96 11.4 2 18.3 75.4 96 - Population Mean 1 28.7 81.0 86.2 13.1 2 25.2 83.0 85.2 12.2 Range 1 18.7-48.6 59.6-100.8 69-105 9.1-18.8 2 12.2-44.8 59.4-105.3 65-102 8.0-16.8 Standard deviation 1 6.2 9.5 10.2 2.3

2 6.7 11.2 11.3 1.7

1 Early Vigour, Height at maturity, Opening of First Floret (DTF) were measured in 2005, 2006. 2 Mature Ht = Height at Maturity 3 DTF = Days To Flowering after sowing, 4 Seed weights were determined from bulk-up plots in 2002 and 2003. 24 RILs were not included in 2002 and 9 RILs were not included in 2003. 5 RILs were missing in both years. Parent B was not grown in 2003.

7.3.4 Detection of QTLs Twenty-two QTLs were found for the four traits on 13 linkage groups

including the 2 additional linkage groups (Table 7.2, Figure 7.1).

7.3.4.1 Early vigour

Eight QTLs were found to influence the early vigour of lupins (Table 7.2,

Figure 7.3a) including two epistatic loci. All but one of the gene alleles contributing

to early vigour were from the domesticated parent. Linkage groups involved were 8,

10 (2), 13, 14, 17 (2), 18 and 23. Most of these showed no environmental

interaction, the exception being the locus on LG 8 (± 0.8cm). One locus on LG14

(CQTL 1) was also found to have an epistatic effect on a locus identified as active on

LG17 (CQTL 2) and, a second locus on LG17 (CQTL 3) was found to be epistatic on

one of the loci on LG23. The range of the locus on LG8 includes Mollis, the gene

associated with soft-seededness. The range of CQTL 2 on LG17 is closely linked to,

but does not include the Ku gene that removes the vernalisation requirement for

flowering. The two loci having the greatest effects were CQTLs 1, 2 on LGs 14, 17,

together accounting for approximately 43% of the observed phenotypic variation.

The 8 QTLs in total accounted for 61% of the variation, with 58.3% being due to

additive effects and a further 2.7% due to epistatic effects (additive x additive).


122


123

Figure 7.3a: Early Vigour

Figure 7.3c: Days to Flowering

Figure 7.3b: Height at Maturity Figure 7.3d: Seed Weight

Figure 7.3. Quantitative Trait Loci showing the locus interval, range and QTL

position as well as gene effects and epistatic interactions on and between linkage

groups (LG). Note: The upper figures denote the marker numbers without the prefix (DAWA) or

the suffix (denoting the approximate marker length). The lower figures are the marker positions (cM)

as on the Linkage map (Figure 2.1 or, as in Figure 7.1. The yellow coloured bar indicates the QTL

range and the estimated position and type of QTL is indicated by the ‘button’. Epistatic genes are

linked with a line.


124

7.3.4.2 Height at maturity

Four QTLs on Linkage groups 1, 4, 8 and 17 were found to influence the

height at maturity (Figure 7.3b), three with additive effects only and a fourth that had

both an additive and GxE effect. The first three loci, including one on LG17 were

contributed by the wild parent. The fourth QTL also exhibiting GxE was contributed

by the domesticated parent. The QTL on LG17 (range 11.9 – 15.0 cM) having the

greatest impact (25.7% of phenotypic variation) overlapped the corresponding one

for early vigour (CQTL 3), with the Ku gene just 2cM outside of that range. The

QTL on LG8 (range 51.9 – 67.9) was distant from the QTL for early vigour on the

same LG (range 15.2 – 19.6). The total amount of height variation that could be

explained by these QTLs was almost 40%, with 3 of the 4 QTLs conferring increased

height coming from the wild parent.

7.3.4.3 Days to flowering

Ten QTLs on 8 LGs were found to influence the time interval from sowing to

opening of the first floret (Figure 7.3c). Six loci had additive gene effects only. The

QTL on LG17 (CQTL 2) having the largest impact (A = (-)10.56 days, explaining

81% of the observed variation) also had a small but significant genotype x

environment (GxE) interaction. This locus (range 18.8 – 20.8) was positioned

adjacent to the mapped position of the gene Ku (17.0cM). Two of the additive QTLs

(LG13, 22) were also involved in epistatic interactions and a further two pairs of

QTLs were involved in separate epistasis’. The locus on LG13 is immediately

adjacent to Lentus, one of the two reduced-pod-shatter genes incorporated into

domesticated lupin at 35.5cM. In total, 88.6% of the variation in flowering could be

accounted for.

7.3.4.4 Seed weight

Two QTLs were found having an influence on seed weight. One QTL was

positioned on the gene Ku (LG17) and also had a substantial GxE interaction

component (Figure 7.3d). The positioning of this QTL is immediately adjacent to

and overlapping CQTL 2 and, 1cM downstream of CQTL 3. The second QTL on

LG9 was positioned on the gene Iucundis associated with plant alkaloid levels (von

Sengbusch 1930, Hackbarth and von Sengbusch 1934). Increased seed size was

positively correlated with the dominant allele associated with high alkaloids carried


125

by the wild parent. Together the 2 loci accounted for between 28 and 34% of all

variation in this trait with the QTL on LG17 having approximately twice as great an

impact as the QTL on LG9.

One QTL on LG17 (CQTL 2) was found to influence both early vigour and

time to opening of the first floret. A short segment of LG17 (CQTLs 2, 3; range 10.6

– 19.8cM) was associated with all four traits

7.3.5 Comparison to Nelson et al. (2006) and synteny with M. truncatula Several regions of synteny with M. truncatula could be identified when QTL

regions were super-imposed on the map by Nelson et al. (2006).

(i) The region on LG8 adjacent to Mollis and associated with early vigour, could

be matched to LG03 of the linkage map by Nelson et al. (2006) (Map 2),

being placed between Mollis and marker Lup111a. No syntenic region for

these markers was found in M. truncatula. The second region on LG8 (map

1) associated with height at maturity, was outside of the locus found to have

some synteny.

(ii) One of the two QTLs on LG13 corresponded to a region on LG05 of Map 2

between the Le gene and 212Len, while the second could be placed near

marker LSSR05. Neither of these 2 markers was associated with synteny in

the M. truncatula genome.

(iii) Common QTL 1 on LG14, associated with both early vigour and the opening

of the first floret, could be located on LG12 of Map 2 as corresponding to the

region of Lup251 and UWA064. Lup251 had also been mapped to a position

on pseudo-chromosome 5 of M. truncatula.

(iv) The two regions on LG17adjacent to the Ku gene could be transposed onto

LG01 of Map 2 although the marker order differed with that of map 1.

CQTL 2 (associated with three out of the four traits) appears to be aligned

with markers UWA232, Lup054 and UWA214 and includes the Ku gene.

CQTL 3, associated with both early vigour and height at maturity,

corresponded to the region of markers Lup158 and VBP1 on the far side of

the Ku gene. Both of these regions have some synteny on Medicago

truncatula pseudo-chromosome 7, albeit on three widely separated loci.


126

7.3.6 Superior genotype By assuming that the ideal genotype is at either one end of the phenotypic

spectrum or the other, we predicted a “superior” genotype (Table 7.2). For

maximum early vigour, only one epistatic allele was required from the wild parent.

For height at maturity to be maximised, three out of four alleles were required from

the wild parent. The earliest flowering offspring had two additional dominant alleles

from the wild parent. Seed weight was maximised by one allele for early flowering

from the domesticated parent and the other for high alkaloids from the wild parent.

An analysis of the 4 possible genotypes for early vigour of the epistatic QTL

pair CQTL 3 on LG17, and LG23 (Table 7.2), revealed (Figure 7.4) that the 10 week

height (indicative of early vigour) ranged from a mean low of 20.5cm for plants

carrying both wild type genes (BB) to a mean maximum height of 31.9cm for plants

carrying one domesticated and one wild type gene (AB). Plants having the parental

genotype (AA) were relatively shorter, having a mean height of 28.8cm and those

with the BA genotype were intermediate (although variable) to the two parental

genotypes at 24.8cm.

Figure 7.4. The effect of two epistatic genes located on LG17 and LG23 on the early

vigour of L. angustifolius. Note: The first allele in each gene combination is located on LG17 and the second on LG 23, with ‘A’

representing the allele originating from the domesticated parent and ‘B’ that from the wild parent.


127

7.4 DISCUSSION Lupinus angustifolius is still a relatively new crop and knowledge about

relationships between various traits and crop production is limited. It is generally

accepted that in Western Australia early flowering to avoid heat induced abortion

during flowering and drought stress at grain filling are desirable to maximise grain

yield and size. The present study investigated in two trials, some of traits thought to

be of importance in grain production and harvesting in a Mediterranean environment.

A number of observations were made, relating seed size as well as early

vigour to regions near the Ku gene on LG17. We have observed that there is linkage

between Ku and early vigour (CQTL 2). However, the role of this QTL in promoting

early vigour is not so pronounced. This could be explained by the presence of

several further QTLs with effects of similar magnitude (LGs 8, 10, 13, 14, 18) and,

the negative effect on early vigour of a second, epistatic locus on LG17 (CQTL 3)

not far away from CQTL 2. The linkage to a region adjacent to Mollis (LG 8) was

surprising in that seed of all hard-seeded RILs had been scarified to ensure rapid

imbibition and germination upon sowing. Consequently, no specific measurements

had been taken of emergence dates and casual observations at the time did not reveal

any obvious differences.

The first QTL for early vigour on LG17 (CQTL 2) was also found to have a

large, significant effect on days to flowering (DTF). From previous work

(Gladstones 1970, Gladstones and Hill 1969) we know that Ku is the only gene to

have a very large effect (2 – 5 weeks) on the reduction in DTF of narrow-leaf lupins.

It therefore appears appropriate that it should have been included in this locus as

occurs on Map 2 (Nelson et al. 2006) and may indicate that some revision of the map

order in this region (Map 1 – Figure 2.1) can be expected at a future date as more

suitable markers are generated.

Other QTL genes influencing DTF cumulatively have almost as great an

effect (A, AxA) on flowering date as Ku in this experiment (Table 7.2), although the

heritability of 4 of the 8 is low. In this particular population, DTF in two RIL lines

was advanced by just 3 days in 2006 from 73 days (domesticated parent) after

sowing (DAS) to 70 days by inclusion of two QTLs from the wild parent. Current

Australian cultivars have DTF (measured at 50% of first florets open) similar to this.


128

For example, Mandelup the earliest flowering cultivar has an approximate DTF of 68

days (B. Buirchell, pers. comm.).

Four QTLs having additive effects on mature plant heights were found. Lack

of rainfall can drastically reduce the plant height at maturity in a crop, while high

rainfall situations can lead to excessive vegetative growth and lodging. Both extreme

situations can make harvest difficult or even impossible. The gene combination of

the domesticated parent tending to reduce plant height (height QTL on LG1) would

therefore be suited to a relatively high rainfall area (e.g. 400mm p.a.), with a further

reduction in height potentially available under very high rainfall (> 450mm p.a.)

conditions. In a low rainfall situation there is the opportunity to breed for increased

height using the QTLs identified on LG’s 4 and 8. The QTL on LG17 (CQTL 3)

having by far the greatest (reducing) effect on height is in the vicinity of the Ku gene

and may be involved in promoting both early vigour and early flowering. It will be

necessary to further clarify the positions of Ku and both QTLs before their usefulness

in the manipulation of plant heights can be determined.

Early flowering and pod-set allows for a longer period of seed production –

both in the extension of the flowering period and seed growth and may result in a

significant increase in seed size as found by Farrington and Gladstones (1974). It is

therefore to be expected that a QTL with close linkage to Ku was found to have a

positive influence on seed size. The second QTL associated with development of

larger seeds centred on the gene Iucundis. The recessive form of this gene, one of

several known to limit development of alkaloids in L. angustifolius (Hackbarth and

von Sengbusch 1934, Gladstones 1970), has been incorporated into all modern

Australian cultivars and is here associated with a smaller seed size. This analysis

lends weight to the suggestion by some researchers that this gene causes a reduction

in yield (von Sengbusch 1938, 1942; Hackbarth and Troll 1960, Kress 1964, Oram

1983). However, these two QTLs only explain 30% of the observed variation in seed

size, probably indicative of insufficient replication in this particular trial and, of

further QTLs having a significant effect. For example, in cereals, a large number of

QTL have been reported to be involved in yield components such as ears per plant,

grain weight and number per ear (Börner et al. 2002, Huang et al. 2003, Quarrie et

al. 2005). There is some evidence that seeds per pod may also be an important

component of yield in lupins (B. Wolko et al., pers. comm.) and, it seems likely that

this would have a (negative) correlation with seed size.


129

The number of QTLs found here for each trait are in all probability only a

subset of the total number of controlling genes as not all environments will reveal

their presence (Tanksley 1993, Young 1996). It is also possible that in this particular

cross both parents carry the same set of alleles for a particular QTL. For instance,

lines have been found exhibiting far greater early vigour then demonstrated in this

population (J. Clements, pers. comm.) and, in the Australian Lupin Collection are

many lines (of L. angustifolius) that produce very small seeds. There are also a small

number of lines tending to produce seed larger then those produced by this particular

population. To locate all of the genes involved in these traits may require

substantially more and larger populations to be generated and grown over a number

of sites and years.

It was found that several lupin QTL regions corresponded to regions of

synteny on the M. truncatula pseudo-chromosome. Of particular interest are the

QTLs associated with early vigour and days to flowering. These two traits are

important to many crops including pasture legumes in dry Mediterranean

environments. The identification of syntenic regions on the M. truncatula genome

may thus lead to improved selection of superior lines by developing the appropriate

molecular markers into forms useful across one or more related species.

The prediction of a superior genotype is an important part of plant breeding

and the selection of parents has in the past frequently been on the basis of

observations of superior performance in the field without fully understanding its

genetic basis. For example, it had long been known that in China the F1 hybrid rice

‘Shanyou 63’ was superior to both parents. A careful QTL analysis of that particular

cross using the program QTL Network 1.0, revealed the basis for the superiority of

the F1 and, that there was still the potential to further increase yield gains from the

same parental combination (Yang and Zhu 2005). The outcome of our study

confirms that in this particular parental combination it is possible to select

individuals having a higher early vigour, demonstrating the potential to improve

narrow-leaf lupins by using a similar approach. Consequently, in Table 7.2 we

inserted a column titled ‘Superior genotype’ under which we have predicted plants

genotypes that are expected to lead to superior trait expressions as based on this

experiment. However as noted earlier, the optimum for a particular trait such as

height at maturity may in fact be somewhere between the two extremes in which case

the predicted superior genotypes ought to be re-assessed.


130

Low gene heritability is a common problem in breeding. In this study, some

of the least heritable QTLs were involved in either a GxE interaction or epistasis

(Table 7.2, Figure 7.4). By developing molecular markers closely linked to these

QTLs it may be possible to more efficiently select plants having the desirable alleles,

leading to enhanced rates of genetic improvement. The need for close linkage is

high-lighted in Figure 7.4 where especially the data for plants putatively of the BA

genotype showed a large range in height well beyond that found within the other

gene combinations, suggesting that cross-overs may have occurred between the QTL

gene and the nearest associated marker. The traits (for increased) early vigour,

height at maturity and especially (a reduction in the) days to flowering would benefit

from this approach as all three have several QTLs with substantial impact but of very

low heritability.

In conclusion, this work has shown that the currently available genetic maps

of narrow-leaf lupin have opened the door of opportunity to a careful analysis of

yield component (and other) traits with a view to improved selection of parental

genotypes and accelerated breeding gains. However, the current marker population

needs to be expanded to allow for fine mapping of certain traits. This is especially

important in the region of the Ku gene where there is still confusion over the true

order of the markers and consequently, the number of QTLs controlling the key traits

of early vigour, days to flowering and seed size.

Acknowledgements

I wish to acknowledge the help of Mr D. Renshaw in seed preparation, sowing and

plant observations. Also, Dr M. Nelson for inclusion of data from collaborative

work with myself on a combined genetic map. Marker mtmtGEN00024041 was

developed as part of the Sixth European Framework Programme “New strategies to

improve grain legumes for food and feed”.


131

7.5 REFERENCES Barbacki S (1952) Lupins. Warsaw.

Börner A, Schumann E, Fürste A, Cöster H, Leithold B, Röder M, Weber W

(2002) Mapping of quantitative trait loci determining agronomic important

characters in hexaploid wheat (Triticum aestivum L.). Theor Appl Genet

105:921-936

East EM (1916) Studies on size inheritance in Nicotiana. Genetics 1:164-176

Edwards AC, van Barneveld RJ (1998) Lupins for livestock and fish. In ‘Lupins as

crop plants: biology, production and utilisation’. (Eds JS Gladstones, C

Atkins, J Hamblin) pp 385-411. (CAB International, London)

Farrington P, Gladstones JS (1974) Effect of genotype on yield of narrow-leafed

lupin (Lupinus angustifolius) and sandplain lupin (Lupinus cosentinii). Aust J

Exp Agric Anim Husb 14:742-748

Gladstones JS (1958) The influence of Temperature and humidity in storage on seed

viability and hard-seededness in the West Australian blue lupin, Lupinus

Digitatus Forsk. Aust J AgricRes 9:171-181

Gladstones JS (1960) Lupin cultivation and breeding. J Aust Inst Agric Sci 26:19-25


and L. digitatus 1. Non-shattering pods. Aust J Exp Agric Anim Husb 7:360-

366


Gladstones JS (1977) The narrow-leafed lupin in Western Australia (Lupinus

angustifolius L.). Western Australian Department of Agriculture Bulletin

3990



Husb 9:213-220

Hackbarth J, Sengbusch von R (1934) The inheritance of alkaloid freedom in

Lupinus luteus and Lupinus angustifolius. Züchter 6:249-255 [in German].

Hackbarth J, Troll HJ (1960) Cultivation and use of sweet lupins. pp 115.

(Frankfurt on Main: DLG-Verlag)

Hanelt P (1960) The lupins 104 pp. (A Ziemson, Wittenberg) [in German]


132

Huang XQ, Cöster H, Ganal MW, Röder M (2003) Advanced backcross QTL

analysis for the identification of quantitative trait alleles from wild relatives

of wheat (Triticum aestivum L.). Theor Appl Genet 106:1379-1389

Johannsen W (1909) Elemente der exakten Erblichkeitsliehre. (Fisher, Jena)

Kress H (1964) On the position and perspective of lupin breeding in the D.D.R.

Wiss Z Karl-Marx-Univ Lpz 13:638 [in German]

Mazur B, Krebber E, Tingey S (1999) Gene discovery and product development

for grain quality traits. Science 285:372-375



Moore G, deVos KM, Wang Z, Gale MD (1995) Cereal genome evolution –

grasses, line up and form a circle. Curr Biol 5:737-739



MGK, Książkiewicz M, Wolko B, Barker SJ, Oliver RP, Cowling WA

(2006) The first gene-based map of Lupinus angustifolius L. – location of

domestication genes and conserved synteny with Medicago truncatula. Theor

Appl Genet 113:225–238

Nilsson-Ehle H (1909) Kreuzunguntersuchungen an Hafer und Weizen. (Lund).

Oram RN (1983) Selection for higher seed yield in the presence of the deleterious

low alkaloid allele iucundis in Lupinus angustifolius L. Fld Crop Res 7:169-

180

Paterson AH, Lander ES, Hewitt JD, Peterson S, Lincoln SE, Tanksley SD

(1988) Resolution of quantitative traits into Mendelian factors, using a

complete linkage map of restriction fragment length polymorphisms. Nature

335:721-726

Petterson DS (1998) Composition and food uses of lupins. In ‘Lupins as crop plants:

biology, production and utilisation’. (Eds JS Gladstones, C Atkins, J

Hamblin) pp 353-384. (CAB International, London)

Quarrie SA, Steed A, Calestani C, Semikhodskii A, Lebreton C, Chinoy C,

Steele N, Pljevljakusić D, Waterman E, Wegen J, Schondelmaier J,

Habash DZ, Farmer P, Saker L, Clarkson DT, Abugalieva A,

Yessimbekova M, Turuspekov T, Abugalieva S, Tuberosa R, Sanguinet

M-C, Hollington PA, Aragués R, Royo A, Dodig D (2005) A high-density


133

genetic map of hexaploid wheat (Triticum aestivum L) from the cross Chinese

Spring x SQ1 and its use to compare QTLs for grain yield across a range of

environments. Theor Appl Genet 110:865-880



1022






vernalization photoperiod and temperature under controlled environment.


Sengbusch von R (1930) Lupins low in bitter substances. Züchter 2:1-2

Sengbusch von R (1931) Low alkaloid lupins 2. Züchter 3:93-109

Sengbusch von R (1938) The breeding of sweet lupins. Herbage Rev 6:64

Sengbusch von R (1942) Sweet lupins and oil lupins: The history of the origin of

some new crop plants. Landw Jbr 91:719 [in German]

Tanksley SD (1993) Mapping polygenes. Ann Rev Genet 27:205-233

Yang J, Zhu J (2005) Methods for predicting superior genotypes under multiple

environments based on QTL effects. Theor Appl Genet 110:1268-1274

Yang J, Hu CC, Ye XZ, Zhu J (2005) QTL Network 2.0. Institute of

Bioinformatics, Zhejiang University, Hangzhou, China.

(http://ibi.zju.edu.en/software/qtlnetwork)

Young ND (1996) QTL mapping and quantitative disease resistance in plants. Ann

Rev Phytopathol 34:479-501


135

Chapter Eight

General discussion and future directions


136

8.1 INTRODUCTION The narrow-leaf lupin (Lupinus angustifolius L.) has a relatively short history

of domestication (Chapter 1.2 – The Domestication genes). A number of major

genes known as the ‘domestication genes’ have been incorporated into the species

over the last 90 years. These genes have helped improve the seed quality, adaptation

to large scale cropping and to the Western Australian Mediterranean climate and,

allow bulk mechanical harvesting, (von Sengbusch 1931, Hackbarth and Troll 1956,

Mikolajczyk 1966, Gladstones 1967, Gladstones and Hill 1969). This has helped

lupins to become a major field crop especially in Western Australia.

8.2 LUPIN BREEDING In breeding, the selection of any desirable trait – be it in a plant or animal is

dependent on the use of a linked marker – be it a visible morphological trait or a

molecular genetic factor. The reality is that frequently, physical traits are complex

and therefore are poor indicators of genotype.

The domestication genes of narrow-leaf lupins are mostly recessive in nature

and, are not always easily phenotyped (e.g. the reduced-shatter gene tardus;

Gladstones 1967). Consequently, following domestication, very few wild genotypes

have been used as parents over the years (see Figure 8.1), resulting in a narrow gene

pool with limited scope for improvement. To widen this gene pool it is necessary to

introduce new wild material (of which large collections exist in Western Australia

and elsewhere) and, at the same time retain the domestication genes. This may best

be achieved if there are the means to readily tag these genes – either by physical or

by genetic markers. Ideally, all of these genes should be identified at the earliest

opportunity in the breeding cycle – the seedling F2. Molecular markers create almost

limitless opportunity to do so.

Figure 8.1. Pedigree relationships among the main narrow-leafed lupin cultivars released in Australia 1967-2006. Circles indicate key single crosses; diamonds represent complex crosses. Bold lines indicate crosses involving Australian cultivars. Domestication genes: moll (soft seeds), iuc (sweet), ta (non-shattering), le (non-shattering), leuc (white flowers and seeds), Ku (early flowering). All cultivars released after Unicrop (1973) in this diagram have these genes apart from Marri (lacks Ku) and Chittick (efl rather than Ku). Abbreviations: Gls-R (grey leaf spot resistant), An-R (anthracnose resistant), Ph-R (phomopsis resistant), Bs-MR (brown spot moderately resistant), MRB (moderately restricted branching), Pop. E1.1 (population E1 – first cycle of recurrent selection) (adapted by BJ Buirchell, JG Boersma from Cowling, 1999)

New Zealand Blue moll

Borre (Sweden 1947) moll, iuc

Danja 1986

Marri 1976

Yandee 1980Illyarrie 1979

Warrah 1989 Ph-R

Yorrel 1989 Ph-R

Merrit 1991 Ph-R

Gungurru 1988Ph-R

Unicrop 1973moll,ta,le,iuc,leuc,Ku

Uniharvest 1971moll,ta,le,iuc,leuc

Uniwhite 1967moll,ta,iuc,leuc

Wild type fromIsrael

Wild types from Italy

Wild types from Spain, Morocco

79A078-14-10 (high yield)

[Pop. E2.1]

[Pop. E1.1]

84A086 F1

CE2-1-1 (high yield)

naturalmutant: leuc natural

mutant: ta

naturalmutant: Ku

natural mutant: le

P20661 mutant: efl)

Rancher (USA 1965) Gls-R, An-R

Fest 1973 moll,ta,le

Belara 1997 Ph-R

Kalya 1996 Ph-R, An-MR Tallerack 1997

Bs-MR, MRB

Myallie 1995 Bs-MR

Chittick 1982 efl

65G-251 USA frost-tolerant

[Pop. E1.2]Wonga 1996

An-R, Ph-R

Sweet mutant ex Germany 1928: iuc

Moonah 1998 Ph-R

Tanjil 1998 An-R, Ph-R

Landrace with soft seeds: moll

Quilinock 1999 Ph-R

75A54-5-8

84A086-73-10

Coromup 2006

Mandelup 2004 Ph-R, AnMR

84A086-12-17


138

8.3 MARKER DEVELOPMENT AND MARKER ASSISTED

SELECTION Molecular markers that are tightly linked to desirable genes are becoming

increasingly important in many areas of genetic analysis including plant breeding.

There are many examples for both major genes and minor genes (QTLs) including:

resistance to powdery mildew of wheat (Hartl et al. 1993, Ma et al. 1994), leaf rust

of barley (Borovkova et al. 1997), resistance to the soybean cyst nematode (e.g.

Chang et al. 1997, Concibido et al. 1994, 1996, 1997) and, resistance to late blight of

potato (Leonards-Schippers et al. 1994).

The development of molecular markers to the domestication genes of lupins

was a logical development when considering the difficulty in phenotyping plants for

genes such as tardus, and the need to wait until after harvest to identify plants with

the soft seeds allele of mollis. Ku, being a dominant gene is also awkward, requiring

a third generation to be grown to determine which F2 plants are homozygous for the

Ku allele.

In this thesis I report the development of molecular markers for: Ku (early

flowering), lentus, tardus (reduced pod shatter) and mollis (soft seeds). A marker for

iucundis was also developed but found to be unsatisfactory when tested on both

cultivars and wild accessions, hence it is not reported. Two types of MFLP markers

were converted into PCR-based forms: (i) a fragment size polymorphism type and

(ii) a single nucleotide polymorphism (SNP) type. Both marker types are amenable

to high-throughput, low-cost MAS methods.

The relative ease with which these mapped markers could be converted to

forms useful for MAS means that potentially there are many more desirable genes to

which we already have molecular markers that may be readily converted for MAS

once the association between gene(s) and markers have been established.

Testing of molecular markers associated with the domestication traits in

lupins has highlighted the molecular diversity of this species and the need for

extremely close linkage between the gene and nearest marker(s). In this study it was

shown that markers commonly be considered to be closely linked to a gene at less

then 3cM may not be close enough for all breeding purposes when introgressing wild

material into the domesticated gene pool. Thus, the finding that markers as close as


139

2.8cM from the gene tardus (Chapter 6), were correct less then 50% of the time in

identifying plants carrying the alleles for reduced pod-shatter. Similarly for the two

markers associated the with the gene lentus (Chapter 5). Even the marker developed

for mollis (Chapter 4) was applicable only to the narrow-leaf lupin, despite its

apparent perfect linkage and the belief that the mechanism for developing soft-

seededness is the same across the lupin species (Gladstones 1958; Quinlivan 1966,

1968; Arrieta et al. 1994).

The development of such highly specific PCR-based markers for the

domestication genes is however considered a major improvement for the agronomy

of the crop. Over the past 4 years, the Department of Agriculture and Food, WA

(DAFWA) has used molecular markers similar to those developed here, to screen

progeny of advanced breeding lines for the (rapid) incorporation of resistance to

especially anthracnose of lupins, but also phomopsis (both single major genes). Last

year approximately 20,000 plants were screened for one or both resistance genes,

multi-loading sequencing gels (a technique that I developed , Unpublished) so that on

average 642, and as many as 960 plants, could be tested in one gel run.

DAFWA intends to use the newly developed marker for tardus this season

(June / July 2007) to screen potential wild parents for their amenability to selection

procedures using this marker. Although only about 45% of wild accessions are

expected to be to screenable by the marker, this number may still make available as

many as 600 (out of a total of 1349 wild and landrace accessions in the Perth

collection) potential parents for crosses in which phenotyping of progeny will be

largely unnecessary. This will result in considerable savings in cost, with time and

space and, the need for experienced personnel to do the phenotyping being greatly

reduced.

The need to pre-screen potential parents to ensure the marker is valid,

highlights the desirability of developing ‘perfect’ markers to these genes (Ellis et al.

2002). This may be achieved by using the markers now developed as probes on a

Bacterial Artificial Chromosome (BAC) library (or similar) to walk the chromosome

to the gene, tracing the intervening genetic sequence via over-lap of the various

BACs. A BAC library has recently been developed for this species (Kasprzak et al.

2006) so that this option is now feasible. Currently work is under way in Poland to

find a better marker for the first known major resistance gene to anthracnose of

lupins (Yang et al. 2004, You et al. 2005) using this library (H. Yang, pers. comm.).


140

8.4 MOLECULAR MARKERS AND MAPPING Research covered in this thesis includes the first comprehensive molecular

genetic map of lupins (Chapter 2). The map is estimated to cover at least 80% of the

genome at an average marker density of 1 marker every 3.2cM. This density makes

it probable that many genes of interest may be located with accuracy and, if the

flanking markers are converted into a useful form, ensures a high degree of precision

in following the gene pathway through successive plant generations. This map is

also the only one known to have used MFLP markers in its construction. At present

there also appears to be no published work on the use of MFLP markers in crops

other then lupins, although a small number of scientists have visited the developer of

this technique to learn the procedure (H. Yang, pers. comm.).

Using MFLP markers instead of AFLPs to construct a linkage map has been

shown to be of advantage. Unlike AFLPs, many MFLP polymorphisms (frequently

carrying a micro-satellite motif and co-dominant) can be converted into simple PCR

based markers desirable for routine marker implementation in MAS (Yang et al.

2001, 2002, 2004) with an ease comparable to that of microsatellites (SSRs).

However, MFLPs are similar to AFLPs in that neither requires the development of

specific primer pairs before they can be used for mapping purposes, as do SSRs (e.g.

Röder et al. 1998). This feature of MFLPs therefore allows the rapid generation of

markers for mapping purposes and, when found to co-segregate with genes of

interest, may in many instances be readily converted into sequence-specific PCR-

friendly markers. These findings have been reinforced in this thesis with a

successful conversion of almost all markers (to the domestication genes) attempted,

into a sequence-specific, PCR-friendly form, useful for implementation (MAS) in a

breeding program.

8.5 QUANTITATIVE TRAIT LOCI AND THEIR ANALYSIS MFLP markers, when generated in large numbers and arranged into a genetic map

also open up opportunities for identification of further genes of interest – especially

minor genes (QTLs) that in combination with others have a major effect on such

complex traits as yield and disease resistance. The work in this thesis is the first

reported to map QTLs controlling the flowering, plant height and seed weight of


141

lupins. There are supporting evidence in the literature for some of these QTLs. In the

first instance, the effect of the gene iucundis on yield e.g. von Sengbusch 1938,

1942). One of the components of yield is seed size. In cereals, a longer period of

grain filling is known to promote large grains (Blacklow and Incoll 1981, Nicolas et

al. 1985, Borrell et al. 1989). A larger seed size is one outcome of the early

flowering promoted by the Ku gene of lupins. However, because the lupin is

indeterminate in its growth habit, this effect is greatly reduced and the rate of seed

filling is at least as important in determining seed size (Palta et al. 2001).

In the second instance, it has long been recognised that there are multiple genes

influencing flowering time, although none of them has as great an effect as does Ku

(Gladstones 1970, Gladstones and Hill 1969).

There is a possibility that the determination of (further) QTL gene positions

and values can be improved by increasing the size of the mapping population. Some

researchers (e.g. Lande and Thompson 1990, Zhang and Smith 1992, 1993;

Gimelfarb and Lande 1994, 1995) have recommended a population size in excess of

200 sibs when conducting QTL studies. These recommendations are on the basis of

older QTL mapping programs using linear programming methods rather then the

matrix method used here, although the impact of using this new method on numbers

required is unclear. There is some anecdotal evidence that more QTLs may be found

in a larger population. In the process of phenotyping for the tardus gene, it was

observed that there was still some variation in shattering that could not be explained

by the two genes lentus and tardus alone, suggesting the involvement of at least one

more gene (Discussion section, Chapter 6). An attempt to find this gene by means of

a QTL analysis (data not presented) failed – possibly as a consequence of the limited

population size.

It was not possible to greatly increase the number of sibs included in this

study for 2 reasons: (i) The full population for this cross is only 115 RILs, (ii) The

map on which the QTL study is centred consists of just 89 RILs. To extend the study

to the full population would have required genotyping the RILs not already included

on a framework skeleton of the map before they could be included in the analysis.

This was beyond the scope of this research project, and the improvement in outcome

is uncertain.


142

8.6 SYNTENY The benefits of finding the sequence of a particular gene may potentially be

extended to other legume plant species on the basis of synteny. As already

mentioned in the Review of Literature (Chapter 1.6), researchers working on the

‘model’ plant species for legumes, Medicago truncatula, have also discovered

several regions of micro-synteny with narrow-leaf lupins (Nelson et al. 2006).

Synteny may be exploited in plant breeding by hunting for genes using known

sequences from other species to refine the search. For example, until now how there

has been no success in locating the genes controlling flowering in legume species

(M. Nelson, pers. comm.). By walking the lupin chromosome via the BACs and

finding the lupin gene Ku that promotes early flowering (by removal of vernalisation

requirements), it may be possible to find a comparable sequence having the same

function in the model and other legume species including important crops such as

soybean. Success in locating such a gene may aid in breeding plants suited to a

range of environments, on the basis of manipulated vernalisation requirements.

At the same time, by analysing the gene sequence and associated biochemical

pathways, a better understanding may be developed of the biological mechanisms

controlling the pathways to flowering for a range of plant species. One example of

such an analysis is that of Fridman et al. (2000) who examined the factors controlling

sweetness of tomatoes. By discovering an important gene controlling the sweetness

in tomatoes, they have opened the window of opportunity to exploit this gene by

selectively cross-breeding or even genetically modifying another plant species by

introducing this gene by either an inter-specific cross or directly by Genetic

Modification (Zamir 2001).

8.7 FUTURE DIRECTIONS The development of markers to the domestication genes of lupins, especially

that for tardus, has opened a window of opportunity to dramatically increase the

level of incorporation of wild material into the gene pool of cultivated narrow-leafed

lupin. This development has the potential to result in greater rates of gain in grain

yield and quality as new gene combinations, including quantitative genes, are

introduced.


143

Further work on mapping, including QTL mapping in lupins looking at other

yield components would be beneficial in the long term. However, the current marker

population appears to either be too small or exhibit linkage disequilibrium (see

Literature Review – genetic mapping), as demonstrated by the (partial) failure of

apparently closely linked markers (to the genes iuc, le, ta) to reliably detect that wild

accessions were not carrying the domesticated gene allele. This phenomenon is not

unknown. Problems of a similar nature are well known in cereal breeding, it being

considered normal to validate the applicability of a marker to potential parents before

using it for screening (e.g. Sharp et al. 2001).

It may be of long-term benefit to create a new mapping / marker population.

This population should be developed using not only the current parents but also at

least one elite cultivar and one or more wild accessions (Mackay and Powell 2006) to

create a highly heterogeneous population segregating for major genes such as the

Phomopsis resistance genes and QTL traits of interest; displaying for example, early

vigour similar to the best discovered to date, but also retaining lines displaying very

low vigour. Retaining the parents of the current population would be an advantage in

that it increases the probability of being able to continue using (and improvement of)

the now (two) existing linkage maps for this species. It seems advisable that the

generation of new markers in such an instance be by means of MFLP, or a

comparable technique that allows markers to be easily converted for MAS as the

need arises, rather then by means of AFLPs, RAPDs or similar.

It is envisaged that the fine mapping of this species will yield markers and

gene information that will be cross-transferable to other legume species and, that

gene information may also flow in the reverse direction as the maps become more

defined.

Continued QTL analysis and gene mapping of the narrow-leaf lupin will

ultimately result in an understanding of the lupin genome comparable to that of

cereals. Using markers such as those developed using MFLPs gives breeders the

ability to rapidly incorporate and test for new genes in breeding lines, thus also

dramatically reducing the time and effort required to produce superior genotypes by

conventional breeding methods.


144

8.8 REFERENCES Arrieta V, Besga G, Cordero S (1994) Seed coat permeability and its inheritance in

a forage lupin (lupinus hispanicus). Euphytica 75:173-177

Blacklow WM, Incoll LD (1981) Nitrogen stress of winter wheat changed the

determinants of yield and the distribution of nitrogen and total dry matter

during grain filling. Aust J Plant Physiol 8:191-200

Borovkova JG, Jin Y, Steffenson B, Kilian A, Blake TK, Kleinhofs A (1997)

Identification and mapping of a leaf resistance gene in barley line Q21861.

Genome 40:236-241

Borrell AK, Incoll LD, Simpson RJ, Dalling MJ (1989) Partitioning of dry matter

and the deposition and use of stem reserves in a semi-dwarf wheat crop. Ann

Bot (Lond) 63:527-539

Chang SJC, Doubler TW, Kilo VY, Abu-Thredeih J, Prabhu R, Freire V,

Suttner R, Klein J, Schmidt ME, Gibson PT, Lightfoot DA (1997)

Association of loci underlying field resistance to soybean sudden-death

syndrome (SDS) and cyst nematode (SCN) race 3. Crop Sci 37:965-971

Concibido VC, Denny RL, Boutin SR, Hautea R, Orf JH, Young ND (1994)

DNA marker analysis of loci underlying resistance to soybean cyst nematode

(Heterodera glycines Ichinohe). Crop Sci 34:240-246

Concibido VC, Denny RL, Lange DA, Orf JH, Young ND (1996) RFLP mapping

and MAS of soybean cyst nematode resistance in PI 209332. Crop Sci

36:1643-1650

Concibido VC, Lange DA, Denny RL, Orf JH, Young ND (1997) Genome

mapping of SCN resistance genes in Peking, PI 90763, PI 88788 using DNA

markers. Crop Sci 37:258-264

Cowling (1999) Pedigrees and characteristics of narrow-leafed lupin cultivars

released in Australia from 1967 to 1998. Agriculture Western Australia,

Bulletin 4365 pp 5

Ellis MH, Spielmeyer W, Gale KR, Rebetzke GJ, Richards RA (2002) “Perfect”

markers for the Rht-B1b and Rht-D1b dwarfing genes in wheat. Theor Appl

Genet 105:1038 – 1042


145

Fridman E, Pleban T, Zamir D (2000) A recombination hotspot delimits a wild-

species qualitative trait locus for tomato sugar content to 484 bp within an

invertase gene. Proc Natl Acad Sci. USA 97:4718 – 4723

Gimelfarb A, Lande R (1994) Simulation of marker-assisted selection in hybrid

populations. Genetical Res 63:39-47

Gimelfarb A, Lande R (1995) marker-assisted selection and marker-QTL

associations in hybrid populations. Theor Appl Genet 91:522-528






7:360 – 366



Husb 9: 213-220

Hackbarth J, Troll H-J (1956) [Lupins as grain legumes and fodder plants.]. In:

Kappert H and Rudorf W (eds) [Handbook of Plant Breeding, Part IV], 2nd

edn. Verlag Paul Parey, Berlin and Hamburg. pp.1-51 (in German)

Hartl L, Weiss H, Zeller FJ, Jahoor A (1993) Use of RFLP markers for the

identification of alleles of the Pm3 locus conferring powdery mildew

resistance in wheat (Triticum aestivum L.). Theor Appl Genet 86:959-963

Kasprzak A, Šafář J, Janda J, Doležel J, Wolko B, Naganowska B (2006) The

bacterial artificial chromosome (BAC) library of the narrow-leafed lupin

(Lupinus angustifolius L.). Cell Mol Biol Lett 11:396-407

Lande R, Thompson R (1990) Efficiency of marker-assisted selection in the

improvement of quantitative traits. Genetics 124:743-756

Leonards-Schippers C, Gieffers W, Schafer-Pregl R, Ritter E, Knapp SJ (1994)

Quantitative resistance to Phtophthora infestans in potato, a case study for

mapping in an allogamous plant species. Genetics 137:67-77

Ma ZQ, Sorrels ME, Tanksley SD (1994) RFLP markers linked to powdery

mildew resistance genes Pm1, Pm2, Pm3 and Pm4 in wheat. Genome 39:830-

835


146

Mackay I, Powell W (2007) Methods for linkage disequilibrium mapping in crops.

Trends Plant Sci 12(2):57-63



Nelson, MN, Phan HTT, Ellwood SR, Moolhuijzen PM, Bellgard M, Hane J,






Nicolas ME, Lambers H, Simpson RJ, Dalling MJ (1985) Effect of drought on

metabolism and partitioning of carbon in two wheat varieties differing in

drought-tolerance. Ann Bot (Lond) 55:727-742

Palta JA, Turner NL, French RJ, Buirchell B (2001) Terminal drought and seed

yield of lupin. (visited June 19, 2007)

http://www.regional.org.au/asa/2001/1/b/paltaj.htm



1022



Röder MS, Korzun V, Wendehake K, Plaschke J, Tixier, M-H, Leroy, P, Ganal

MW (1998) A Microsatellite map of wheat. Genetics 149:2007-2023



Sengbusch von R (1942) [Sweet lupins and oil lupins. The history of the origin of

some new crop plants.] Landw Jbr 91:719-880 (in German)

Sharp PJ, Johnston S, Brown G, McIntosh RA, Pallotta M, Carter M, Bariana

HS, Khatkar S, Lagudah ES, Singh RP, Khairallah M, Potter R, Jones

MGK (2001) Validation of molecular markers for wheat breeding. Aust J

Agric Res 52:1357-1366




147














Lett 10:123-134

Zamir D (2001) Improving plant breeding with exotic genetic libraries. Nature Rev

Genetics 2:983-989

Zhang W, Smith C (1992) Computer simulations of marker-assisted selection

utilizing linkage disequilibrium. Theor Appl Genet 83:813-820

Zhang W, Smith C (1993) Simulation of marker-assisted selection utilizing linkage

disequilibrium: the effects of several additional factors. Theor Appl Genet

86:492-496


Chapter Nine

Appendices

149


9.1 ANTHRACNOSE DISEASE RATING OF RIL

POPULATION AND COMPARISON TO AnM2 AND NEARBY

NEWLY GENERATED MAP MARKERS.

RIL # Disease

rating (1-5)

Anthracnose

disease scores AnM2

DAWA

1019

DAWA

207

DAWA

224

W/D 8 1.625 A A B B B

W/D 9 5 B B B B B

W/D 10 4 B B B B B

W/D 12 4.375 B B B B B

W/D 13 3.833 B B B B B

W/D 14 3.375 B B B B B

W/D 15 1.125 A A A A A

W/D 16 3.125 B B _ B B

W/D 17 3.67 B B B B B

W/D 18 4 B B B B B

W/D 19 3.833 B B B B B

W/D 21 4 B B B B B

W/D 23 3.875 B B _ B B

W/D 24 1.25 A A A B B

W/D 25 3 B B B A A

W/D 26 1.25 A A ? ? A

W/D 27 1 A A A A A

W/D 28 3.625 B B B B B

W/D 34 1.375 A A A A A

W/D 35 4.5 B B _ A A

W/D 36 2 A A B A A

W/D 38 2 H A A A A

W/D 39 3.67 B B B B B

W/D 40 2.125 A B _ B B

W/D 41 _ _ B B B B

W/D 42 1.5 A A A B B

W/D 43 1 A A A A A

W/D 44 3.125 B B B B B

W/D 45 3.75 B B B B B

W/D 46 1.875 A B B B B

W/D 48 3.375 B B B _ B

W/D 49 2.25 A A A _ A

W/D 50 3 B B _ B B

150


RIL # Disease

rating (1-5)

Anthracnose

disease scores AnM2

DAWA

1019

DAWA

207

DAWA

224

W/D 51 3.25 B B _ B B

W/D 52 1.75 A B B B A

W/D 53 1 A A A A A

W/D 54 3.75 B B B B B

W/D 57 1 A A A A A

W/D 58 1.625 A A A A A

W/D 59 1.75 A B B A A

W/D 60* 2.375 B B B B B

W/D 64 1.875 A A B A A

W/D 65 2.875 B B B B B

W/D 66 1.375 A A A A A

W/D 69 3 B B B B B

W/D 70 3.5 B B _ B B

W/D 73 3.75 B B B B B

W/D 74 3.125 B B B B B

W/D 76 2.5 B B B B B

W/D 77 1 A A A A A

W/D 78 3 B B B B B

W/D 80 1 A B B A A

W/D 82 4.25 B B B B B

W/D 83 1 A A A A A

W/D 84 2.25 A A A A A

W/D 85 2.5 B B B B B

W/D 86 1.25 A A A A A

W/D 89 1.75 A A A A A

W/D 90 2.125 A A _ A A

W/D 92 1.125 A B B A A

W/D 93 2.5 B B B B B

W/D 94 4.25 B B A B B

W/D 95 2.67 B B A B B

W/D 97 1.375 A A _ A A

W/D 98 4.25 B B B B B

W/D 99 - _ A _ A A

W/D 100 1.125 A A _ A A

W/D 102 4.25 B B B B B

W/D 104 1.875 A A A A A

W/D 106 3.75 B A B B B

151


RIL # Disease

rating (1-5)

Anthracnose

disease scores AnM2

DAWA

1019

DAWA

207

DAWA

224

W/D 107 1 A A A A A

W/D 109 3 B B B B B

W/D 111 2.5 B B B A A

W/D 112 3.125 B B B B B

W/D 113 3.875 B B A B B

W/D 114 1 A A A A A

W/D 116 1.375 A A A A A

W/D 118 1.25 A A B A A

W/D 119 2 H A A B B

W/D 120 2.875 B B B B B

W/D 122 3.5 B B B B B

W/D 123 3.25 B B B B B

W/D 124 3 B B B B B

W/D 125 3 B B B B B

W/D 127 1.75 A A A A A

W/D 128 3.75 B B _ B B

W/D 130 1.5 B B B B B

W/D 132 2 A A _ A A

W/D 135 1.75 A B B B B

W/D 136 4 B B B B B

A 476 1.25 A A A A A

P27255 3.25 B B B B B

Marker

errors 8 12 11 11

Note: This experiment was restricted primarily to the population used in construction of the molecular

map (Chapter 2).

Missing values / data are marked with a dash.

Marker genotypes are bolded when they disagree with Disease rating.

152


9.2 TARDUS SHATTER RATING OF RIL POPULATION AND

COMPARISON TO NEARBY NEWLY GENERATED MAP

MARKERS INCLUDING THOSE OF NELSON ET AL. 2006.

RIL # Tardus

scores

DAWA

169

DAWA

978

DAWA

1097c

Lup

214

Lup

001

UWA

244

W/D 8 A A A - A A -

W/D 9 B B B B B B -

W/D 10 B B B B B B B

W/D 11 B B - - B B B

W/D 12 A A A A A A B

W/D 13 A A A A A A A



W/D 16 B B B B - - -


W/D 18 A A A A A A B

W/D 19 B B B B B B A


W/D 23 B B B - B B B


W/D 25 B B B B B - B

W/D 26 B B - B B B B



W/D 32 A - - - A A A

W/D 33 B B - - B B B


W/D 35 B B B B - - -


W/D 37 B B - - A B B

W/D 38 A A A A B A A





W/D 43 B B B B - - -

W/D 444 B B B B A B H

153


RIL # Tardus

scores

DAWA

169

DAWA

978

DAWA

1097c

Lup

214

Lup

001

UWA

244

W/D 45 A A A A A A -


W/D 47 B B - - B B B

W/D 48 B B - B B B B

W/D 49 A A A A - - -

W/D 50 B B B B - - -

W/D 51 B B B B A B B



W/D 54 A A A A B A A

W/D 55 B B - - - - -

W/D 56 A A - - - - -


W/D 58 A A A A A A -


W/D 601 A A A A - - -

W/D 61 B B - - B B B

W/D 62 B B - - B B B

W/D 63 A A - - A A A

W/D 64 B B B B - - -

W/D 65 B B B - B B -


W/D 67 B B - - B B B


W/D 70 A A A - - - -

W/D 71 B B - - B B B



W/D 75 B B - - B B B

W/D 76 B B B B B B -

W/D 77 A A - A - A A


W/D 79 B B - - - - -

W/D 80 B B B B B B -

W/D 81 A A - - A A A



W/D 84 B B - - B B B

154


RIL # Tardus

scores

DAWA

169

DAWA

978

DAWA

1097c

Lup

214

Lup

001

UWA

244



W/D 89 B A - A A A B

W/D 90 A A A - - - -

W/D 92 A A A - A A A




W/D 96 B B - - - - -



W/D 99 A A A - - - -

W/D 100 A A A - A A A

W/D 101 B B - - B B B


W/D 104 B B B B B B -

W/D 105 A A - - A A A


W/D 107 B A A B A A B

W/D 108 B B - - - - -


W/D 110 - B - - - - -



W/D 113 B B B B A - H


W/D 115 A A - - - - A

W/D 116 A A A A - - -

W/D 117 B B - - B B B




W/D 121 A A - - - -


W/D 123 A A A A A - A

W/D 124 A A A A A - A

W/D 125 B A A A B - B

155


RIL # Tardus

scores

DAWA

169

DAWA

978

DAWA

1097c

Lup

214

Lup

001

UWA

244

W/D 126 B B - - - - -

W/D 127 B B B B B - B

W/D 128 B B B B - - -

W/D 130 A A A A A - A

W/D 131 B A - - - - -

W/D 132 A A - - - - -

W/D 135 A A - - A - A

W/D 136 A B B B B - B

A 476 A A A A A A A

P27255 B B B B B B B

Marker

errors3 4 2 2 7 2 7

1RIL 60 is suspected to not be the original line 60 as that one is supposed to have died out at an earlier

stage. 2RIL 106. This line appeared to be a mixture. In 2006 one plant was shedding while the stem was

still green, whereas others did not shed before all were fully dry. 3Errors. Assuming that phenotyping is correct (bar RIL 44) and ignoring missing data (-),

discrepancies between marker and population phenotypes have been totalled. 4 It is almost certain that RIL 44 is heterozygous.

156


9.3 DOMESTICATION GENE RATING OF RIL POPULATION

(ACCORDING TO PARENTAL PHENOTYPE), INCLUDING

SEED ‘MOUSTACHE’, BUT EXCLUDING TARDUS RATING.

RIL # Alkaloid Flowering

time

Flower

colour Lentus Mollis moustache

W/D 8 B A A B A B

W/D 9 B B A B B B

W/D 10 A B B A A A

W/D 11 B B B A A -

W/D 12 B B A B B A

W/D 13 B A B A B A

W/D 14 A A A A B B

W/D 15 A A B B A A

W/D 16 A B B B B A

W/D 17 B A B A A B

W/D 18 B B A B A B

W/D 19 B B B A B B

W/D 21 B B A B B A

W/D 23 B A A A B B

W/D 24 A A B B B B

W/D 25 B B A B B A

W/D 26 B A A B A B

W/D 27 A A A B B B

W/D 28 A B B B B B

W/D 32 A B A B B -

W/D 33 B A B B B -

W/D 34 A A A B A B

W/D 35 A A B A B B

W/D 36 B B B B B A

W/D 37 A A B A B -

W/D 38 B B A B A A

W/D 39 H B B B B A

W/D 40 B B A B A A

W/D 41 B B A B B B

W/D 42 A A B A B B

W/D 43 A A A A A A

W/D 44 A B A B A A

157



time

Flower


W/D 45 A B B A B B

W/D 46 B B A B A B

W/D 47 B A B B B -

W/D 48 A B A A B B

W/D 49 B A B A B B

W/D 50 B A A A B A

W/D 51 A B B A B B

W/D 52 B A A B A A

W/D 53 A A A B B A

W/D 54 A B B A A A

W/D 55 B A A B A -

W/D 56 B A B B B -

W/D 57 B B B A A A

W/D 58 B B A B A B

W/D 59 A A B A B A

W/D 601 B A B B A A

W/D 61 B B H2 B B -

W/D 62 A B A B B -

W/D 63 A B B B A -

W/D 64 B A A A B B

W/D 65 B A A B A B

W/D 66 B B A B B A

W/D 67 B B A A B -

W/D 69 B A A A A A

W/D 70 B A B B A A

W/D 71 B B A A B -

W/D 73 B A A B B B

W/D 74 B A B A A A

W/D 75 B A B B B -

W/D 76 A A A A B A

W/D 77 A A B B B B

W/D 78 B A B A B A

W/D 79 B B? B A B -

W/D 80 B B B A A B

W/D 81 A B B B B -

W/D 82 B B B A A B

W/D 83 A A B A A B

W/D 84 B A B B H -

158



time

Flower


W/D 85 A B B B A A

W/D 86 B A A A B B

W/D 89 B B B B A A

W/D 90 B B A A B B

W/D 92 B B B B A B

W/D 93 B A B A A A

W/D 94 B A B A B B

W/D 95 A B B B B A

W/D 96 A A B B B -

W/D 97 B A B A B A

W/D 98 B A A A A A

W/D 99 A B B A B B

W/D 100 A A B A B B

W/D 101 A A B B B -

W/D 102 B B A A B B

W/D 104 A B A A A A

W/D 105 B B B B A -

W/D 106 B B B B A B

W/D 107 A B A B B A

W/D 108 A B B A B -

W/D 109 B B A B B A

W/D 110 B A B B B -

W/D 111 A A A B B B

W/D 112 B B B B B B

W/D 113 A B B A B B

W/D 114 B A B B A A

W/D 115 B B B A A -

W/D 116 B B A A A A

W/D 117 B B A B B -

W/D 118 B B A A B A

W/D 119 B A B A B A

W/D 120 B A A B A B

W/D 121 B B B A B -

W/D 122 B A B B B A

W/D 123 A B B A A B

W/D 124 A B B B B A

159



time

Flower


W/D 125 B A A B B B

W/D 126 A B A? A A -

W/D 127 B A B A A B

W/D 128 B A A B A B

W/D 130 B A B A B B

W/D 131 A B A B A -

W/D 132 B A A H B A

W/D 135 A B B A B A

W/D 136 B A A A B A

83A:476 A A A A A A

P27255 B B B B B B

Note:

The domesticated parent (A) has the following characteristics:

(i) Low in alkaloids (iuc – recessive);

(ii) Flowers early (Ku – dominant);

(iii) Has white flowers (leuc – recessive);

(iv) Has the reduced shatter gene le (recessive);

(v) Soft seeds (mollis – recessive);

(vi) Moustache pattern on seeds (mou) – gene dominance uncertain.

Missing values / data are marked with a dash. In some instances a tentative phenotype has been

inserted followed by a question mark. 1RIL 60 is suspected to not be the original line 60 as that one is supposed to have died out at an earlier

stage. 2RIL 61 was found originally to be heterozygous for flower colour. However, my notes suggest that

the white flowers were subsequently rogued out.

160


9.4 ENZYME BUFFERS

1. Boehringer Mannheim

Component (mM) Buffer “A”

Buffer “B”

Buffer “L”

Buffer “M”

Buffer “H”

T4 DNA Ligase

In MFLP

Tris-acetate 33 - - - - Tris-HCl - 10 10 10 50 66 50 Magnesium acetate 10 - - - - MgCl2 - 5 10 10 10 5 10 KCl 66 - - - - NaCl - 100 - 50 100 30 Dithioerythritol (DTE) - - 1 1 1 Dithiothreitol (DTT) 0.5 - - - - 5 ATP 1 1 2-Mercaptoethanol - 1 - - - pH 7.9 8.0 7.5 7.5 7.5 7.5 7.8

2. Fermentas (and Geneworks)

Component (mM) Buffer “Tango”

Buffer “B”

Buffer “G”

Buffer “O”

Buffer “R”

T4 DNA Ligase

In MFLP

Tris-acetate 33 - - - Tris-HCl 10 10 50 10 66 50 Magnesium acetate 10 - - - MgCl2 10 10 10 10 10 Potassium acetate 66 - - - NaCl 100 50 100 - 5 30 KCl 100 Dithioerythritol (DTE) - Dithiothreitol (DTT) - - - - 5 1 BSA 0.1

mg/ml 0.1

mg/ml 0.1

mg/ml 0.1

mg/ml

ATP - - - 1 pH 7.5 7.5 7.5 7.5 8.5 7.5 7.8 Like Like B Like M Like H

161


Enzyme Buffer in

Fermentas

T4 DNA ligase Tru1I = MseI (t/taa)

Fermentas R Use MFLP recipe

Csp6 I (G/TAC) Fermentas B Use buffer provided, but add10x ligate additive (below)

Hpa II (c/cgg) Fermentas Tango Use buffer provided, but add10x ligate additive (below)

TaqI (T/CGA) Boehringer Buffer B Boehringer Buffer B, but add10x ligate additive (below)

EcoRI (G/AATTC) Fermentas O or R Use MFLP recipe 10x ligate additive

10X stock = Molecular weight Making 10 mL stock will need:

50 mM NaCl 58.4 28.2 mg 100 mM DTT (Boehringer cat.# 197 777)

154.3 154 mg

10 mM ATP (Boehringer cat.# 519 979)

605.2 60.5 mg

250 μg/mL BSA (Boehringer cat.# 238031)

2.5 mg

Stir to dissolve, adjust pH to 7.8. Add water to 10 mL

162


9.5 PCR FORMULATIONS AS USED IN THIS THESIS

A typical 10 μL PCR includes the following: DNA 1.5 μL 33P labelled primer 0.5 μL the other primer 0.5 μL PCR mix 7.5 μL Method of 33P labelling: (1) Calculation: Label 1 μL Label 120 μL Primer (30 ng/μL) 0.333333 40 μL Sterile water 0.233333 28 μL 5X kinase buffer 0.2 24 μL T4 polynucleotide kinase (10 U/μL) 0.033333 4 μL 33P ATP 0.2 24 μL Total 120 μL

163


9.6 DNA EXTRACTION FOR MFLP - HIGH QUALITY, BUT LOW YIELD

1. Harvest about 10 to 14 full-size young leaves. Put the leaves in a mortar, add

liquid nitrogen, and grind into powder.

2. Immediately add 4 mL DNA extraction buffer, and mix.

3. Take a 1.5 mL blue tip, cut the tip off, and transfer 0.8ml of the mixture into a

1.5 mL Eppendorf tube

4. (Transfer the remaining leave slurry into a vial, label them, and store at –

20°C for future use.)

5. Add 0.5 mL phenol/chloroform / iso-amyalcohol (25:24:1), mix it with a

plastic tag (do not vortex!);

6. Centrifuge at 13,200 rpm for 30 min;

7. Take a new tube, add 5 μL RNase. Transfer 500 μL aqueous to this new

tube and mix. Incubate on bench for 5 min (not longer, as some batch of

RNase can also damage DNA).

8. Add 0.4 mL chloroform. Mix it with a plastic tag (flicking externally).

9. Centrifuge for 13,200 rpm 10 min;

10. Transfer 300 μL upper aqueous phase to new tube. Add 400 μL extraction

buffer, mix. Add 560 μL isopropanol, mix, leave on bench for 20 to 30 min,

11. Centrifuge at 11,000 rpm for 20 min. Suck away all aqueous.

12. Add 150 μL 70% ethanol. Mix by inverting a few times. Leave on bench for

1 hr, or overnight.

13. Spin 5 min. Suck away all aqueous using P200 pipette.

14. Spin 10 second. Such away all remaining aqueous using P10 pipette.

15. Open lid, leave on bench for 5 min.

16. Add 30 μL TE0.1 buffer. Leave the tubes in the fridge overnight.

17. The next day, gently mix by sucking it up and down using a cut-off yellow

tip.

164


9.7 EXAMPLE MFLP PROCEDURE

Development of molecular markers for anthracnose and phomopsis

resistance in narrow-leaf lupin. 1. Choose F8 plants for running MFLP

Note: (1) ‘Tanjil’ and ‘Unicrop’ are the two parents

(2) In this example, we need 12 plants for MFLP, but we start with 14

plants, just in case if one or two plants do not work.

(3) The plants chosen will enable you to search markers for anthracnose

resistance and phomopsis resistance at the same time.

Plant no. 5 7 Tanj 20 33 9 15 18 Uni 25 29 11 17 13 Anthracnose R R R R R R R R S S S S S S Phomopsis R R R R R S S S S S S R R R

2. Estimate DNA concentration of each plant In this example the DNA concentration is measured as in the following second table 3. Set up Restriction-Ligation (1). Prepare the adaptor: Adaptor of Yang Take the “MseI Adapter”, heat in a water bath at 95°C

for 5 min. Cool on bench for 20 min, brief spin. (2). Prepare the following “Master mix”

For each sample For 18 samples

water 6.6 μL 118.8 μL Yang’s 10 X T4 ligase buffer

2 μL 36 μL

0.5 M NaCl 1 μL 18 μL 0.5 mg/mL BSA 2 μL 36 μL Yang’s MS adopter 2 μL 36 μL Tru 9I (same as MseI) (10U/μL)

0.6 μL 10.8 μL

T4 ligase (5U/μL) 0.8 μL 14.4 μL DNA sample

Subtotal: 15.0 μL Total: 270.0 μL

165


4. Prepare the following R-L reactions • Add DNA in each tube, add water to make the volume of 5 μL. Add 15 μL

Master mix, so the final volume is 20 μL.

Lane no. /Plant DNA conc.

(ng/μL) DNA

needed (μL)

water needed (μL)

The above “Master mix”

(μL)

Final name after HaellI

1 Tanjil 400 0.7 4.3 15 HaeIII TanB1

2 20 200 1.5 3.5 15 HaeIII F8B2

3 33 150 2.0 3.0 15 HaeIII F8B3

4 9 100 3.0 2.0 15 HaeIII F8B4

5 15 150 2.0 3.0 15 HaeIII F8B5

6 18 80 4.0 1.0 15 HaeIII F8B6

7 Unicrop 150 2.0 3.0 15 HaeIII UniB7

8 25 100 3.0 2.0 15 HaeIII F8B8

9 29 150 2.0 3.0 15 HaeIII F8B9

10 11 80 4.0 1.0 15 HaeIII F8B10

11 17 80 4.0 1.0 15 HaeIII F8B11

12 13 150 2.0 3.0 15 HaeIII F8B12

13 5 100 3.0 2.0 15 HaeIII F8B13

14 7 100 3.0 2.0 15 HaeIII F8B14

Mix, spin down, incubate at 37°C in water bath for 2 hr.

5. Digest with HaeIII

To the above 20 μL R-L mix, add 80 μL TE0.1., Extract with 100 μL chloroform.

Transfer 80 μL aqueous in a new tube, add 200 μL ethanol and leave for 30 min.

Spin 15 min, pour off ethanol; spin 3 sec, and such all the ethanol. Add 20 μL Hae

III digestion mix (17 μL water, 2 μL Buffer M, 1 μL (5 U) HaeIII). Incubate at

37C for 2 hr. Add 130 TE0.1 and mix. This mix will be the “DNA” used in the

following “Pre-Selective Amplification”.

166


6. Pre-Selective MFLP Amplification MseI-primer:

MseI-C GAT GAG TCC TGA GTA A --C SSR Primer: PRIMERS:

MF203 GGGA (TTC)4T MF204 CCCT (TTC)4

• For each plant you add:

(1) 3 μL DNA sample

(2) 1 μL of SSR-primer (30 ng/μL),

(3) 1 μL of MS primer (30 ng/μL),

(4) 15 μL Fisher Biotech PCR mix

PCR tube

Template Primers (30 ng each)

Pre-amp code

PCR tube

Template Primers (30 ng each)

Pre-amp code

1 HaeIII TanB1

MS-C + MF203

MF203C TanB1

21 HaeIII TanB1

MS-C + MF204

MF204C TanB1

30g MS

2 HaeIII F8B2

MS-C + MF203

MF203C F8B2

22 HaeIII F8B2

MS-C + MF204

MF204C F8B2

30 ng SSR

3 HaeIII F8B3

MS-C + MF203

MF203C F8B3

23 HaeIII F8B3

MS-C + MF204

MF204C F8B3

4 HaeIII F8B4

MS-C + MF203

MF203C F8B4

24 HaeIII F8B4

MS-C + MF204

MF204C F8B4

5 HaeIII F8B5

MS-C + MF203

MF203C F8B5

25 HaeIII F8B5

MS-C + MF204

MF204C F8B5

6 HaeIII F8B6

MS-C + MF203

MF203C F8B6

26 HaeIII F8B6

MS-C + MF204

MF204C F8B6

7 HaeIII UniB7

MS-C + MF203

MF203C UniB7

27 HaeIII UniB7

MS-C + MF204

MF204C UniB7

8 HaeIII F8B8

MS-C + MF203

MF203C F8B8

28 HaeIII F8B8

MS-C + MF204

MF204C F8B8

9 HaeIII F8B9

MS-C + MF203

MF203C F8B9

29 HaeIII F8B9

MS-C + MF204

MF204C F8B9

10 HaeIII F8B10

MS-C + MF203

MF203C F8B10

30 HaeIII F8B10

MS-C + MF204

MF204C F8B10

11 HaeIII F8B11

MS-C + MF203

MF203C F8B11

21 HaeIII F8B11

MS-C + MF204

MF204C F8B11

12 HaeIII F8B12

MS-C + MF203

MF203C F8B12

32 HaeIII F8B12

MS-C + MF204

MF204C F8B12

13 HaeIII F8B13

MS-C + MF203

HaeIII F8B13

33 HaeIII F8B13

MS-C + MF204

MF204C F8B13

14 HaeIII F8B14

MS-C + MF203

HaeIII F8B14

34 HaeIII F8B14

MS-C + MF204

MF204C F8B14

• PCR cycles: MFLP pre-selective (94°C for 2 min. 25 cycles of 94°C 30 sec,

52°C 30 sec, 72°C 1 min, hold at 4°C)

167


• Take 10 μL of the above PCR mix, mix with 5 μL gel loading dye. Run 6%

acrylamide gel at 100V for 2 hr. If you see the “smear” bands, it means you are

successful. 7. Selective MFLP Amplification (1) 33P Primer labelling for the SSR primers Step 1. Prepare the following master mix

Components Final Volume Prepare the following:

5 X Kinase buffer 42 [γ-33P] ATP 42 T4 polynuceotide kinase (10 U/μL)

7

Sterile water 49

Total volume 140 Step 2. Prepare the following master mix

MF203 (30 ng/μL)

MF204 (30 ng/μL)

Primer stock 35 μL 35 μL The above master mix 70 μL 70 μL

Incubate the above tube at 37°C for 60 min. Then heat at 70°C to kill the enzyme.

168


(2) Set up Selective MFLP PCR Sequencing Gel “Tanjil – 1” • Each gel has 96 PCR samples, which is 12 plant x 8 primer combinations. • Annealing temperature of selective MFLP cycles: Starting annealing at 60°C,

decrease 0.7°C per cycle until you reach 54°C. 25 cycles annealing at 54°C. • Run 38x50 cm sequencing gel (0.4 mm thick), 5% denaturing polyacrylamide, at

55 W for 3.5 hours. Dry the gel, and expose X-ray film for 3 days.

Step 2: Set up Master mix for each group of 12 samples

MseI-primer 6.5 μL

Step 1: add 1.5 μL template DNA in 8 groups of 12 plants each 33P Labelled SSR-primer 6.5 μ 1 MF203C

TanB1 PCR mix 105 μL

2 MF203C F8B2

Step 3: Add 8.5 μL into each tube of each group (below)

3 MF203C F8B3

4 MF203C F8B4

PCR tubes Selective primers

5 MF203C F8B5

1-12 MSCAA + MF203***

6 MF203C F8B6

13-24 MSCAG + MF203***

7 MF203C UniB7

25-36 MSCAC + MF203***

8 MF203C F8B8

37-48 MSCAT + MF203***

9 MF203C F8B9

49-60 MSCGA + MF203***

10 MF203C F8B10

61-72 MSCGG + MF203***

11 MF203C F8B11

73-84 MSCGC + MF203***

12 MF203C F8B12

85-96 MSCGT + MF203***

169


Sequencing Gel “Tanjil – 2”

• Each gel has 96 PCR samples, which is 12 plant x 8 primer combinations.

• Annealing temperature of selective MFLP cycles: Starting annealing at 60°C,

decrease 0.7°C per cycle until you reach 54°C. 25 cycles annealing at 54°C.

• Run 38x50 cm sequencing gel (0.4 mm thick), 5% denaturing polyacrylamide, at

55 W for 3.5 hours. Dry the gel, and expose X-ray film for 3 days.

Step 2: Set up Master mix for each group of 12 samples

MseI-primer 6.5 μL

Step 1: add 1.5 μL template DNA in 8 group 12 plant each

33P Labelled SSR-primer 6.5 μ

1 MF203C TanB1 PCR mix 105 μL

2 MF203C F8B2

3 MF203C F8B3

Step 3: Add 8.5 μL into each tube of each group (below)

4 MF203C F8B4 PCR tubes Selective primers

5 MF203C F8B5 1-12 MSCCA + MF203***

6 MF203C F8B6 13-24 MSCCG + MF203***

7 MF203C UniB7 25-36 MSCCC + MF203***

8 MF203C F8B8 37-48 MSCCT + MF203***

9 MF203C F8B9 49-60 MSCTA + MF203***

10 MF203C F8B10 61-72 MSCTG + MF203***

11 MF203C F8B11 73-84 MSCTC + MF203***

12 MF203C F8B12 85-96 MSCTT + MF203***

Continue new MFLPs using all available primer combinations as required.

170


Addendum (1) MseI-Adaptor

MSAD1 GAC GAT GAG TCC TGA G MSAD2 TA CTC AGG ACT CAT 1. Purchase the above oligo the same way as primer. Make MSAD1 and MSAD2 at

100 pM/μL as stock.

2. Mix the above two with equal volume to form a 50 pM/μL adaptor solution. Put

50 μL such mix in each tube and store at –20 freezer (each tube enough for one

experiment).

(2) AFLP primers used in MFLP:

In Pre-selective MFLP amplification, we routinely use only one MseI primer (MS-C) MS-C GAT GAG TCC TGA GTA A –C

In Selective MFLP amplification, we use the following 16 MseI primers. Therefore,

one pre-selective MFLP will run 16 selective MFLP amplifications

MSCAA GAT GAG TCC TGA GTA A –CAA MSCAG GAT GAG TCC TGA GTA A –CAG MSCAC GAT GAG TCC TGA GTA A –CAC MSCAT GAT GAG TCC TGA GTA A –CAT MSCGA GAT GAG TCC TGA GTA A –CGA MSCGG GAT GAG TCC TGA GTA A –CGG MSCGC GAT GAG TCC TGA GTA A –CGC MSCGT GAT GAG TCC TGA GTA A –CGT MSCCA GAT GAG TCC TGA GTA A –CCA MSCCG GAT GAG TCC TGA GTA A –CCG MSCCC GAT GAG TCC TGA GTA A –CCC MSCCT GAT GAG TCC TGA GTA A –CCT MSCTA GAT GAG TCC TGA GTA A –CTA MSCTG GAT GAG TCC TGA GTA A –CTG MSCTC GAT GAG TCC TGA GTA A –CTC MSCTT GAT GAG TCC TGA GTA A –CTT

171


(3) Yang’s 10x T4 ligation buffer (Yang made using the same recipe as MFLP)

10X = MWT For 1 L For 1 mL For 10 mL

500 mM Tris 121.1 61.5 g 61.5 mg 615 mg

100 mM MgCl2.6H2O 203.3 20.3 g 20.3 mg 203 mg

100 mM DTT 154.3 15.4 g 15.4 mg 154 mg

10 mM ATP 605.2 6.05 g 6.05 mg 60.5 mg

250 μg/mL BSA 2.5 mg

pH7.8

Water till 10 mL

(4) TE0.1:

10mM Tris, 0.1 mM EDTA, pH 8.0. Autoclave. (5) Fisher Biotech PCR mix Attention: Chris Fisher Department: Fisher Biotech, P.O. Box 169, Subiaco, WA 6904 Fax no: (08) - 9322 3868 Telephone: (08) - 9322 6866

Product Name Cat. No. Pack size No. of Packs

Price

Taq polymerase TAQ-3 1000 units 1 $450 5 x polymerisation buffer

PB-10 10 mL 1 90

Note: “25 mM MgCl2" is included when you buy the Taq polymerase Procedures:

1. Take a 100 mL yellow-cap tube. Add 9.22 mL PCR grade water in using a 10-

mL sterile pipette;

2. Add 4 mL 5X polymerisation buffer;

3. Add 1.6 mL 25 mM MgCl2 ;

4. Spin the 4 tubes of Taq polymerase for 10 sec;

5. Take 200 μL, mix from the bottle, and add into each of the Taq polymerase

tubes, and gently mix. Transfer all of the enzyme into the bottle;

6. Gentle mix the bottle by shaking in circle motion;

7. Put 880 μL the mix in each of the 1.5 mL eppendorf. Store at –20°C until use.

(each eppendorf is enough for one sequencing gel of 96 MFLPs).

172


9.8 CLONING AND SEQUENCING A DNA BAND

I. PCR re-amplify a MFLP band

1. Cut the DNA band from a dried sequencing gel through the X-ray film, put the

piece of the gel into an eppendorf tube (0.5 ml); 2. Add 100 μL TE buffer (10 mM Tris, 1 mM EDTA, pH 8) to suspend the gel.

Heat at 94°C for 15 min. Brief spin. 3. Set up 2 tubes of the following PCR for each band, each tube with 50 μL, using

the following steps • Spin the DNA sample of “Gelboil” at top speed for 5 min so that you will

not collect gel residue when you take the DNA sample. • Set up the follow PCR: 7.5 μL DNA template (the supernatant of “Gelboil”), 2.5 μL MseI-primer,

2.5 μL SSR-primer 37.5 μL PCR mix

• Run PCR program “52C25CL” in the “PCRs” folder (=25 cycles at 94°C for 30 s, 52°C for 30 s, 72°C for 2 min)

II. Clean up the PCR product using Roche Diagnostics clean up kit

Roche Diagnostics Kit: “High Pure PCR Product Purification Kit” Cat. No. 1 732 668 for 50 reactions, cost $121 per kit, Or Cat. No. 1 732 676 for 250 reactions, cost $504 per kit.

1. Add 500 μL “Binding buffer” to the 100 μL PCR mixture. Mix well. 2. Load the mix into the “filter tube”, which is set on a “Collection tube”, spin 30

seconds at top speed using a bench centrifuge; 3. Pour off liquid in the collection tube. Add 500 μL “washing buffer” on the

filter tube, spin 30 s; 4. Pour off liquid in the collection tube. Add 200 μL “washing buffer” on the

filter tube, spin 30 s; 5. Take the filter tube (make sure it does not have any liquid on it) and put it on a

1.5 ml new tube as collection tube. Add 80 μL water (PCR grade) into the filter tube, leave it for 2 min, then spin 30 seconds.

III. Run a gel to check DNA concentration For each sample, take 10 μL above DNA, mix with 5 μL loading dye. Run

1.5% agarose mini gel at 70V for 1 h. One lane is 2 μL DNA mass ladder as standard. Take a photo using the digital camera, calculate DNA concentration.

173


IV. Ligate and transform E. coli using Promega’s kit

Promega’s kit: “pGEM-T Easy Vector, System I” Cat. # A1360 $242 per kit, consisting of the vector, ligase, and buffer.

Promega’s competent cells: JM109 Competent Cells, 1 ml (5 x 200 μL), Cat.# L2001, $189 per pack

1. SET WATER BATH AT 42°C 2. Make a 1.5 % agarose mini gel (1.5 g agarose, 100 ml TBE, 3 μL ethidium

bromide stock). Load 5 μL of the above DNA sample, together with the DNA mass standard, run at 70V for about 1 h.

3. Take a photo using the digital camera. Calculate the amount of DNA of each band in each sample,

4. Calculate the volume of DNA needed for each transformation.

Amount of DNA needed = (? ng vector x insert DNA band in kb / plasmid size in kb) x ratio for insert: plasmid. In our case, we use 50 ug vector, and the ratio of (insert:plasmid) is 3 For example: Band Y21 is 443 bp, and is 157 ng in the photo, which we loaded with 5 μL, Therefore it is 31.4 ng/μL The DNA needed = (50 ng vector x 0.443 kb / 3 (the plasmide pGEM-T is 3 kb) x 3 = 22 ng The volume needed = 22 / 31.4 = 0.7 μL

5. Set up ligation by combining: For example for Band Y21: ? μL DNA 1 μL vector 1 μL T4 ligase (use ONLY the T4 ligase provided by Promega!!!) 5 μL 2x butter

Add water to make the total volume of 10 μL

Mix by pipetting. Leave on bench for 1 h (or for best results, leave at 4°C overnight). HEAT AT 70°C FOR 10 MIN TO KILL THE ENZYME, leave on ice, then SPIN down.

6. Take out the competent E. coli cells (“JM109” from Promega) from freezer, and leave on ice for at least 10 min (the cells are fragile, and need to be warmed up very slowly).

7. Add 5 μL the Ligation mix into a new 1.5 ml tube. 8. Take a yellow tip, and cut off the sharp tip head, then take a tube of E.coli, and

mix the bacterial cells gently.

174


9. Add 50 μL JM109 cells into the tube, gently mix by stirring with tip (do not suck and squeeze!), place the tube on ice for 20 min.

10. Heat shock at exactly 42°C water bath for 50 seconds (do not shake); 11. Immediately put the tube on ice for 2 min. 12. Add 0.5 ml LB medium (4°C), incubate at 37°C with shaking (≈200 rpm) for

1h. 13. Take three (3) “LB-IPTG-X-Gal petri plates”. Mixing E. coli cells in the tube

by pipetting. Add 50 μL E. coli culture in the first petri plate, spread with a glass rod. Add 150 μL E. coli culture in the second petri plate, spread with a glass rod. Add all the remaining (300 μL) E. coli culture to the third petri plate, spread with a glass rod. Spread with a glass rod. Write the name of the band on the plates. Incubate the plates at 37°C overnight.

14. The next day, put the plate in a cold room for 1 h (during which the blue colour will become darker, easier for the next step (Keep it in the dark).

15. Take another LB-IPTG-X-Gal petri plate, draw 30 - 50 squares on back. Transfer 30 - 40 white colonies of E. coli for each DNA band, one on each square. (Note: colonies with slight blue colour are OK) Have a blue colony as a control. Incubate at 37°C for 20 h. Keep at cool-room until the next step.

LB medium liquid media or LB Agar media: 10 g Bacto- tryptone

5 g Bacto - yeast extract 5 g NaCl

*** do not add agar if it is LB liquid media *** do add 20g agar if it is LB Agar media Add water to 1 litre. Autoclave for 15 min

For LB ampicillin plates:

After autoclaving, allow them to cool to 50°C. Add 2 ml Ampicillin stock per litre agar medium (Ampicillin stock = 100 mg/ml). Pour 20 ml per petri plate. Store at 4°C for up to 1 month, or room temperature up to 1 wk (ampicillin goes off).

LB-IPTG-X-Gal petri plate

To LB ampicilim plates, spread 100 μL of IPTG stock (0.1g/ml), and 20 μL of X-gal Stock (=50 mg/ml). Allow 30 min for the agar to absorb before use.

Ampicillin (= 100 mg/ml) stock: Dissolve ampicillin in water. Filter to sterilise.

Store at –20°C. IPTG stock (0.1g/ml): dissolve 1 g in 10 ml water, sterilise by filtration, keep stock

at –20°C. (IPTG = Isopropyl-b-D-thiogalactopyranoside).

X-gal Stock (50 mg/ml): Dissolve 0.5 g X-Gal in 10 ml of dimethylformamide. The solution needs to be wrapped in foil to avoid light damage. Keep the solution at (-20°C) freezer. Sterilising the solution by filtration is not necessary. (X-Gal = 5’-Bromo-4-chloro-3-indolyl-b-D-galactopyranoside).

175


V. Clone PCR (Optional, especially useful when there are some unwanted DNA bands mixed in the re-amplified PCR product)

Take some E. coli cells, mix with some water. Do several E. coli clones per

DNA band. Take 1 μL of such cell suspension as template DNA, to run PCR using the same primers, PCR mix and thermocycler program as the selective PCR by which the original band was produced. Run a 1% agarose gel to check the size, keep the clones which have the right sized insert for the next step. Discard clones with inserts of the wrong size.

VI. Mini plasmid preparation using Bresatec’s kit Geneworks kit: “UltraClean Mini Plasmid Prep Kit”, Cat.# MB-12300-250,

250 preparations, $408 1. Late afternoon the day before doing plasmid extractions, put 50 ml LB

medium into a bottle, add 100 μL Ampicillin stock (= 100 mg/ml), mix. Put 6 ml into each bottle.

2. Inoculate one clone of E. coli for each bottle, incubate at 37°C for overnight WITH SHAKING.

• For the steps 3-17, do 2 tubes of plasmid extraction for each E.coli clone

3. The next morning, fill 1.8 ml (twice 0.9 ml) E. coli culture into a collection tube. Spin 30 s. Pour off supernatant.

4. Fill in another 1.8 ml (twice 0.9 ml) E. coli culture into the tube, spin 30 s. Pour off supernatant

5. Fill in another 1.8 ml (twice of 0.9 ml) E. coli culture into the tube, spin 30 s. Pour off supernatant.

6. Spin 10 s. Suck off ALL aqueous with a pipette; 7. Add 50 μL Solution 1. 8. Cut off the tip of a yellow tip, stir tip in solids until the E. coli cells are

completely suspended in the solution. (Try to minimise sucking). 9. Check Solution 2, - if precipitated, heat to dissolve. 10. Gently add 100 μL Solution 2 by pipetting onto the wall of the tube. 11. Mixing by inverting the tube twice. Leave on bench for 5 min. 12. Open the lids of all the tube, leave on bench for 5 minutes until the sticky stuff

on the lids disappears (in theory). 13. Gently add 325 μL Solution 3 by pipetting onto the wall of the tube. Mix by

inverting the tube twice. 14. Leave on bench for 30 min to 1 h 15. Spin first replicate tube of each clone at top speed for 15 min. 16. Transfer ALL of the clear liquid supernatant to a spin filter (avoid white

precipitate); 17. Spin starting from very slow, and gradually increase to top speed then stay for

30 s. Suck away of the waste aqueous in the bottom of the tube.

• Be careful here – look at the instructions!

18. Spin the second replicate tube of each clone at top speed for 15 min.

176


19. Transfer ALL of the clear liquid supernatant into the same spin filter tube from above!

20. Spin starting from very slow, and gradually increase to top speed then stay for 30 s.

21. Suck away the aqueous in the collecting tube using a P1000 pipette. 22. Add 300 μL Solution 4 to the spin filter. Leave on bench for 5 min. 23. Spin starting from very slow speed, and gradually increase to top speed then

stay for 30 s. 24. Transfer the column (without disturbing the aqueous below!!) onto a new 2 ml

tube (provided), add 50 μL water onto the column, leave for 10 min. 25. Spin 30 s. Discard the filter, close tube lid. Check the plasmid DNA – optional

Set up EcoRI restriction to 5 μL the above plasmid 1 μL 10x buffer H 1 μL EcoRI 3 μL water incubate at 37C for 1 h

Load the whole mix with 3 μL loading buffer in a 1.5% agarose gel at 60V for 40 min to check the size of insert.

If the plasmid DNA concentration is too low, increase the concentration by the following

Precipitate the remaining 45 μL plasmid DNA by adding 4.5 μL 3 M NaAc pH 5, and 100 μL 100% ethanol, leave in freezer for 1 h. Spin 15 min. Pipette away the aqueous, wash with 70% ethanol, dissolve the DNA in 10 μL water.

VII. Set up sequencing reaction

BigDyeTM mix Applied Biosystems Ph. 1800 033 747 BDT version 3.1: Cat. #: 4337455, contain BDT 800 ul, for $1520 +$30 delivery

• Typical Sequencing reaction:

The above plasmid DNA 5 μL (100 ng plamid) Primer (10 ng/μL) 1 μL BigDyeTerminator 2 μL 2.5x sequencing buffer 2 μL (if there is no sequencing butter, add 4 μL BigDyeTerminator in total)

177


Run sequencing PCR

4 min at 96°C 28 cycles 96°C 30 s, 20 s at annealing temperature (50°C for T7 and M13S; 43°C for SP6), 4 min at 60°C Hold at 4°C

Clean up sequencing PCR mix • Take a P100 pipette and set at 25μL. Take 25μL 100% ethanol by sucking up

and down then suck again to make sure it is 25μL. Mix this 25μL ethanol with the 10 μL sequencing PCR mix, and transfer into a 0.5 ml tube. Leave in freezer for 30 min (overnight is also acceptable).

• Spin at maximum speed for 15 min. Remove as much liquid as possible. • Add 50 μL 70% ethanol. Leave for 10 min. • Spin 10 min at maximum speed. Suck away as much ethanol as possible using

a P100 pipette. • Spin 30 s. Suck away remaining ethanol using a P10 pipette. • It is very important to remove ALL aqueous. • Open up the lid, put the tubes in the desiccator. • Vacuum dry for 5 min. • Close the lids, put tubes in a small plastic bag, fold it in foil paper. Send the

sample to RPH (they run the sequencing gel for you) to run gel.

178


9.9 HOW TO DO MAPPING WITH MAPMANAGER

(1) Set up data in Excel. Type data as:

a. Individual RILs in columns

b. Markers or traits in rows.

c. Score data as ‘A’ or ‘B’ (according to parentals) or ‘H’

(heterozygote). Using 1’s and 2’s is also possible (I think)

d. Missing data must be replaced with a dash.

(2) Save the Excel file as a text file in the following format:

a. Tab delimited

b. Make sure there are no missing points.

(3) Start Mapmanager

(4) Select: Import the text file (above)

a. Set edit: Put in Progeny number (the number of columns you have)

b. Delete all other

c. Press “Apply”

d. Press “OK”

(5) Select your File you wish to import (and it’s location)

(6) “Import Text” – edit the boxes

a. Data-set – tick. Do not tick any other boxes.

b. Chr – no ticks at all.

c. Locus – tick ‘Geno’ only!

d. Between Items – ‘Tab’

e. Between Records – ‘Line Feed’

f. Between Genotypes – ‘Tab’

The little box that pops up should give you:

(i) Total number of rows (e.g. 50 / 50)

(ii) Total number of progeny (e.g. 90 / 90)

179


(7) Select “Options” –

a. Tick Kosambi (or Haldane)

b. Tick Self RI (or double Haploid if applicable)

c. Tick the appropriate ‘P’ value

Note (1): each step may take up a minute or more!

Note (2): The more stringent your ’P’ values, the shorter your linkage groups, the

quicker the analysis.

(8) Select “Tools”

a. Select “Make Linkage Groups”

b. Create a file to save data

c. Linkage groups will now appear on the screen.

d. Click on each to view the group. You can view them as either map or

genotype etc – just play with the available options in the top boxes.

(9) Examine your ‘map’ and data carefully. If you high-light a particular maker,

in your options, you can select ‘Tools’ then ‘Links Report’. This will list all

markers related to this one and the strength of the linkage (LOD value). You

may now wish to go ahead with manipulating and modifying your map.

Note: Heterozygotes can be a problem if you are scoring dominant markers. For the

same reason, the scoring in this protocol is unsuitable when dealing with an F2 or

similar segregating population. In such a case a more complicated method of scoring

is required (consult the web-site of MapManager for details).

180


9.10 PROTOCOL FOR PRODUCING MAPCHART CHART

FROM MAPMANAGER.

To produce a Map from MapManager (MM) data using MapChart (MC) is

relatively easy but does require editing of the exported file for MapChart to read.

The following protocol was produced by Dr. Margaret Pallotta and has been edited

and verified by Jeff Boersma.

(1) Check that the MM file is set to “cumulative” in the ‘Options / Chromosome

Map View’ menu.

(2) Go to ‘File’ menu and select “Export” / “Map”. Select “All”.

(3) Use Excel to open the exported file (‘all files’).

(4) In the options, select “Fixed Width” column spacing. Preview the data and

move the column dividing line to the most suitable position. In the next field,

check that the “column data format” is ‘general’. Press “finish”. In Excel,

stretch the columns so that you can see the full data. The resultant data set

should appear as follows (next page).

(5) Check data again to make sure that the columns were well separated and that

numbers do not cross over into the next column.

181


Data set as first imported into Excel from MapManager

#Boersma publ ished.map() with Kosambi map function on 02/07/2006 14:16DAWA758.470 0.021442DAWA809.475 0.037072DAWA930.135c 0.043922DAWA377.350 0.049736DAWA532.480 0.05576DAWA749.460 0.105928DAWA821.140 0.118751DAWA289.630c 0.124499DAWA80.190 0.143026DAWA956.295 0.149199DAWA109.260c 0.291372DAWA765.180 0.309673DAWA523.290 0.315697DAWA516.290 0.369963DAWA498.310 0.375916DAWA400.300c 0.38173DAWA394.070c 0.405845DAWA478.500 0.471149DAWA91.195 0.5291DAWA545.140 0.535857DAWA808.500 0.564459DAWA793.405 0.576366DAWA682.240 0.582319DAWA792.425 0.588272DAWA180.240 0.594019DAWA181.210 0.606367DAWA261.060 0.625126DAWA392.190 0.64278DAWA30.050 0.701878DAWA1078.225 0.715394DAWA612.130 0.752045DAWA299.680 1.02108DAWA631.170 1.122446DAWA919.090c 1.155829DAWA789.560 1.161854DAWA486.270 1.167668DAWA339.250 1.21153DAWA194.238 1.229184DAWA627.250 1.279991DAWA101.235c 1.298072DAWA391.200 1.32341DAWA772.330c 1.350102DAWA894.120 1.356275DAWA403.208 1.376556DAWA1065.250 ---DAWA563.130c 0.006024DAWA531.125c 0.131547

182


(6) You now need to edit the file so that it appears as below:

(7) To do that, the steps are:

(a) “clear all” in the top header row that contains the export information i.e.

#Boersma published map () with Kosambi map function…

(b) Clear the third last cell (and any other like it) that contains “---“ .

(c) Re-align the map distances down 1 cell by inserting a new cell at the top.

Put a “0” (zero) in the new top cell next to the first marker for each

linkage group (was “---“originally).

(d) Add a title for each group. This needs to start with “Group” or “Chrom”.

Both of these headers are OK. Check MapChart for other naming options

if you want to have another label at the top of the chromosome images.

You can have a space after “Group” but then the rest of the chromosome /

group title cannot have spaces so you either need to put in underscores or

dashes (for example: Chrom CxS _1H), or you can enclose the name in

double quotes (for example: Chrom “CxS 1H”).

(e) The map distances need to be changed to whole numbers i.e. from

“0.021”to “2.1”. This can be done easily in Excel by creating another

column next to the first one (of distances) and using the “=cell*100”

function to generate the correct format. Then “copy” this column and do

“paste special”(select ‘values’) (so that we get values and not formulas)

into a new column next to these.

(f) Delete the two original columns of distances, keeping the “paste

special” column. Check the column and delete any zeros in line

with headers to linkage groups.

(g) Save the file as a “formatted text (space delimited)” file.

183


Data set in Excel after editing.

Group LG1 DAWA758.470 0 DAWA809.475 2.1442 DAWA930.135c 3.7072 DAWA377.350 4.3922 DAWA532.480 4.9736 DAWA749.460 5.576 DAWA821.140 10.5928 DAWA289.630c 11.8751 DAWA80.190 12.4499 DAWA956.295 14.3026 DAWA109.260c 14.9199 DAWA765.180 29.1372 DAWA523.290 30.9673 DAWA516.290 31.5697 DAWA498.310 36.9963 DAWA400.300c 37.5916 DAWA394.070c 38.173 DAWA478.500 40.5845 DAWA91.195 47.1149 DAWA545.140 52.91 DAWA808.500 53.5857 DAWA793.405 56.4459 DAWA682.240 57.6366 DAWA792.425 58.2319 DAWA180.240 58.8272 DAWA181.210 59.4019 DAWA261.060 60.6367 DAWA392.190 62.5126 DAWA30.050 64.278 DAWA1078.225 70.1878 DAWA612.130 71.5394 DAWA299.680 75.2045 DAWA631.170 102.108 DAWA919.090c 112.2446 DAWA789.560 115.5829 DAWA486.270 116.1854 DAWA339.250 116.7668 DAWA194.238 121.153 DAWA627.250 122.9184 DAWA101.235c 127.9991 DAWA391.200 129.8072 DAWA772.330c 132.341 DAWA894.120 135.0102

184


DAWA403.208 135.6275 DAWA1065.250 137.6556 Group LG2 DAWA563.130c 0 DAWA531.125c 0.6024

(8) You should now be able to open the file from MapChart. Use the option “all

files” under the file type to see the text file.

If all of the data is read correctly, MapChart will immediately produce a graph.

However, often small errors in separating columns and inserting corrected values

occur. When this happens, MapChart will give you an error message. You should

be able to access the data-set in MapChart and manually correct the data. Usually

the error is due to insufficient separation of the columns when first importing into

Excel so that one or more letters from the marker labels have ended up in the

distances column. Correct the values and try again.

Data set as opened in MapChart Group LG1 DAWA758.470 0 DAWA809.475 2.1442 DAWA930.135c 3.7072 DAWA377.350 4.3922 DAWA532.480 4.9736 DAWA749.460 5.576 DAWA821.140 10.5928 DAWA289.630c 11.8751 DAWA80.190 12.4499 DAWA956.295 14.3026 DAWA109.260c 14.9199 DAWA765.180 29.1372 DAWA523.290 30.9673 DAWA516.290 31.5697 DAWA498.310 36.9963 DAWA400.300c 37.5916 DAWA394.070c 38.173 DAWA478.500 40.5845 DAWA91.195 47.1149 DAWA545.140 52.91 DAWA808.500 53.5857 DAWA793.405 56.4459 DAWA682.240 57.6366 DAWA792.425 58.2319

185


DAWA180.240 58.8272 DAWA181.210 59.4019 DAWA261.060 60.6367 DAWA392.190 62.5126 DAWA30.050 64.278 DAWA1078.225 70.1878 DAWA612.130 71.5394 DAWA299.680 75.2045 DAWA631.170 102.108 DAWA919.090c 112.2446 DAWA789.560 115.5829 DAWA486.270 116.1854 DAWA339.250 116.7668 DAWA194.238 121.153 DAWA627.250 122.9184 DAWA101.235c 127.9991 DAWA391.200 129.8072 DAWA772.330c 132.341 DAWA894.120 135.0102 DAWA403.208 135.6275 DAWA1065.250 137.6556 Group LG2 DAWA563.130c 0 DAWA531.125c 0.6024

186


DAWA758.4700.0DAWA809.4752.1DAWA930.135c3.7DAWA377.3504.4DAWA532.4805.0DAWA749.4605.6DAWA821.14010.6DAWA289.630c11.9DAWA80.19012.4DAWA956.29514.3DAWA109.260c14.9DAWA765.18029.1DAWA523.29031.0DAWA516.29031.6DAWA498.31037.0DAWA400.300c37.6DAWA394.070c38.2DAWA478.50040.6DAWA91.19547.1DAWA545.14052.9DAWA808.50053.6DAWA793.40556.4DAWA682.24057.6DAWA792.42558.2DAWA180.24058.8DAWA181.21059.4DAWA261.06060.6DAWA392.19062.5DAWA30.05064.3DAWA1078.22570.2DAWA612.13071.5DAWA299.68075.2DAWA631.170102.1DAWA919.090c112.2DAWA789.560115.6DAWA486.270116.2DAWA339.250116.8DAWA194.238121.2DAWA627.250122.9DAWA101.235c128.0DAWA391.200129.8DAWA772.330c132.3DAWA894.120135.0DAWA403.208135.6DAWA1065.250137.7

LG1

DAWA563.130c0.0DAWA531.125c0.6

LG2

MapChart chart output

187


9.11 ARTICLES PUBLISHED OR IN PRESS AS AT JULY 2007 Publication 1 (pages 189 – 202):

Construction of a genetic linkage map using MFLP and identification of molecular

markers linked to domestication genes in narrow-leafed lupin (Lupinus angustifloius

L.) Cellular & Molecular Biology Letters 10, 331-344 (2005).

Publication 2 (pages 203 – 206):

Development of a sequence-specific PCR marker linked to the Ku gene which

removes the vernalization requirement in narrow-leafed lupin. Plant Breeding 126,

306 – 309 (2007).

Publication 3 (pages 207 – 213):

Development of two sequence-specific PCR markers linked to the le gene that

reduces pod shattering in narrow-leafed lupin (Lupinus angustifolius L.) Genetics

and Molecular Biology 30, xxxx (2007). In Press.

188


9.12 RAW DATA FOR QTL ANALYSIS INCLUDING PLANT

HEIGHTS FOR EARLY VIGOUR AND AT MATURITY, SEED

WEIGHT AND FLOWERING DATE.

RIL # 1Early

Vigour (1)

Early

Vigour (2)

Mature

Height (1)

Mature

Height (2)

2Seed

Weight (1)

Seed

Weight (2)

Flowering

DAS3 (1)

Flowering

DAS (2)

W/D 8 32.3 31.0 73.3 91.3 15.9 16.4 75 74

W/D 9 23.2 19.0 83.0 93.9 15.3 12.3 96 97

W/D 10 24.2 21.8 93.4 93.4 12.3 11.9 96 96

W/D 12 23.3 23.6 94.8 95.9 14.8 14.2 105 102

W/D 13 27.8 24.7 73.7 78.4 11.7 9.7 81 76

W/D 14 29.8 24.0 59.6 66.6 14.0 10.7 75 75

W/D 15 29.9 28.0 73.3 78.2 13.0 12.1 80 79

W/D 16 24.9 20.0 86.6 75.9 9.1 8.0 95 96

W/D 17 35.7 34.2 82.7 86.0 15.0 - 76 73

W/D 18 21.9 21.9 84.8 89.6 - 12.6 98 96

W/D 19 28.5 18.0 96.8 90.3 10.2 10.5 95 94

W/D 21 18.8 12.2 85.3 76.1 10.8 12.4 101 101

W/D 23 32.2 21.7 73.0 68.9 14.0 13.1 77 75

W/D 24 33.8 19.8 80.9 71.1 11.1 10.6 75 77

W/D 25 25.3 24.8 89.4 101.2 15.0 12.6 100 102

W/D 26 36.6 24.4 84.1 75.4 14.5 13.1 76 77

W/D 27 31.0 24.8 74.8 61.4 - - 76 74

W/D 28 21.8 17.7 83.0 84.3 9.3 9.5 94 97

W/D 34 31.9 25.1 75.7 70.7 13.9 10.2 79 78

W/D 35 39.3 25.2 73.8 66.2 - - 73 74

W/D 36 24.2 15.8 85.1 78.8 - 9.7 101 99

W/D 38 28.1 22.7 86.7 96.8 11.2 - 97 98

W/D 39 28.1 27.3 84.6 93.9 - - 93 92

W/D 40 29.8 19.7 86.3 87.7 11.5 13.5 93 96

W/D 41 18.7 12.7 71.1 72.6 - 12.1 95 97

W/D 42 31.9 29.1 71.3 70.8 13.6 10.6 77 74

W/D 43 33.7 35.6 60.9 73.4 - 12.4 73 70

W/D 44 24.4 24.2 90.8 97.8 14.0 12.0 97 97

W/D 45 25.8 14.3 87.2 80.1 - 10.3 93 97

W/D 46 21.2 20.3 79.8 79.4 10.5 - 93 93

W/D 48 27.7 14.7 91.1 75.9 10.7 10.8 95 96

W/D 49 36.1 26.0 88.6 91.0 - 14.1 78 77

W/D 50 22.3 23.1 72.4 79.7 - 9.3 83 79

214


RIL # 1Early

Vigour (1)

Early

Vigour (2)

Mature

Height (1)

Mature

Height (2)

2Seed

Weight (1)

Seed

Weight (2)

Flowering

DAS3 (1)

Flowering

DAS (2)

W/D 51 20.9 18.0 81.6 93.3 10.2 9.9 94 95

W/D 52 35.6 30.4 76.3 82.8 13.9 12.6 76 74

W/D 53 31.3 30.3 73.0 76.0 14.2 12.0 78 78

W/D 54 28.4 23.7 84.2 93.0 10.9 10.2 94 93

W/D 57 30.8 31.8 89.9 98.4 14.5 12.3 91 95

W/D 58 20.5 23.3 72.9 91.2 - 14.4 95 92

W/D 59 35.5 33.0 78.4 77.3 13.6 9.5 81 77

W/D 60 28.6 25.2 72.4 72.0 12.6 10.6 83 78

W/D 64 38.4 27.4 85.3 74.4 - 14.2 75 75

W/D 65 28.0 32.2 61.6 81.5 - 13.5 78 72

W/D 66 22.4 26.2 72.0 89.8 11.1 10.7 95 97

W/D 69 35.8 28.7 66.3 76.4 14.4 11.8 75 75

W/D 70 26.1 26.3 73.1 66.0 - 13.4 83 80

W/D 73 40.7 36.0 84.3 79.2 16.5 13.7 75 73

W/D 74 35.7 31.0 83.3 70.9 17.8 13.4 79 76

W/D 76 28.6 24.4 75.9 66.0 - 12.7 77 74

W/D 77 35.6 24.7 67.1 70.4 - 14.4 73 74

W/D 78 42.8 30.9 87.0 83.1 14.0 11.3 72 71

W/D 80 25.0 22.2 72.1 90.8 - 11.3 93 95

W/D 82 28.2 19.7 97.0 99.3 10.2 - 98 101

W/D 83 28.8 24.3 64.3 59.4 12.3 10.8 76 74

W/D 85 28.9 30.7 100.8 104.2 12.0 11.6 100 100

W/D 86 37.3 30.5 75.3 71.3 16.6 12.0 73 71

W/D 89 36.2 38.3 95.3 99.7 14.0 12.8 84 80

W/D 90 23.0 17.2 85.7 91.8 - - 97 99

W/D 92 20.9 20.0 78.4 100.1 13.2 12.9 97 97

W/D 93 40.6 33.8 75.4 67.0 14.8 13.0 69 67

W/D 94 33.5 26.7 81.2 83.8 15.9 12.8 78 75

W/D 95 20.7 14.8 81.7 88.6 11.0 11.6 96 99

W/D 97 34.5 31.5 75.8 80.6 16.2 14.6 79 78

W/D 98 39.0 42.0 73.2 89.5 17.3 16.8 76 75

W/D 99 19.8 15.0 72.4 76.2 - - 95 96

W/D 100 - 29.3 - 70.7 13.1 12.2 75

W/D 102 28.7 19.0 92.4 88.2 11.8 12.9 95 96

W/D 104 29.2 23.2 96.6 92.0 - 9.7 95 97

W/D 106 24.7 22.2 93.2 101.4 10.6 - 100 101

W/D 107 24.3 22.7 75.6 91.2 10.6 11.0 95 95

215


RIL # 1Early

Vigour (1)

Early

Vigour (2)

Mature

Height (1)

Mature

Height (2)

2Seed

Weight (1)

Seed

Weight (2)

Flowering

DAS3 (1)

Flowering

DAS (2)

W/D 109 26.2 19.2 99.9 98.7 10.2 12.5 96 98

W/D 111 27.6 22.3 75.4 62.0 14.4 12.0 79 79

W/D 112 28.7 16.9 79.1 91.6 11.3 11.8 93 92

W/D 113 27.8 23.0 83.5 86.9 11.5 11.7 91 91

W/D 114 30.1 28.8 74.1 83.3 16.7 11.5 84 79

W/D 116 25.1 21.3 92.1 105.3 - 13.6 98 98

W/D 118 24.2 19.7 96.4 99.1 10.6 11.0 103 98

W/D 119 30.6 26.9 69.1 65.1 10.6 9.9 77 71

W/D 120 37.5 35.3 83.3 75.9 18.8 14.2 75 74

W/D 122 39.2 26.1 88.5 83.8 12.6 13.9 81 79

W/D 123 25.1 21.7 94.3 93.7 11.6 13.2 94 96

W/D 124 20.4 16.6 89.8 91.9 11.1 13.3 101 99

W/D 125 36.2 36.4 82.4 95.5 16.1 14.3 77 72

W/D 127 32.7 29.2 70.7 67.9 - 12.9 75 72

W/D 128 31.8 37.0 73.6 87.3 - 15.9 79 78

W/D 130 38.1 38.5 65.4 74.3 15.0 11.8 70 65

W/D 132 30.6 27.7 80.6 79.1 - 14.5 79 78

W/D 135 26.2 16.1 99.7 89.1 10.2 10.9 100 98

W/D 136 48.6 44.8 83.4 80.0 14.6 12.0 72 71

83A:476 48.6 36.9 87.1 79.6 16.3 13.8 72 72

P27255 22.6 12.4 79.3 75.4 11.4 - 96 96

Note: In 2005, the trial was sown on May 31st and, in 2006 the trial was sown on June 2nd. Plant

heights were measured in centimetres (average of 3 reps x 3 plants); seed weights as grams per 100

seeds. Missing data are denoted by a dash (-). 1Data for year one (1) was collected in 2005 from plants grown in a screenhouse. Data for year two

(2) was collected in 2006 from plants grown in the field adjacent to the screenhouse. 2 Data for seed weight was collected in 2002 (year 1) and 2003 (year 2). 3DAS is days after sowing.

216

Documents

Molecular Genetics of the Narrow-leaf Lupin › files › 3241729 › ... · This thesis contains published work and work prepared for publication. Details of the journals and publications