Euphytica

, Volume 176, Issue 1, pp 49–58

Development of a co-dominant DNA marker tightly linked to gene tardus conferring reduced pod shattering in narrow-leafed lupin (Lupinus angustifolius L.)

Authors

  • Xin Li
    • School of Plant Biology, Faculty of Natural and Agricultural Sciences and Institute of AgricultureThe University of Western Australia
  • Daniel Renshaw
    • Department of Agriculture and Food Western Australia
  • Huaan Yang
    • Department of Agriculture and Food Western Australia
    • School of Plant Biology, Faculty of Natural and Agricultural Sciences and Institute of AgricultureThe University of Western Australia
Article

DOI: 10.1007/s10681-010-0212-1

Cite this article as:
Li, X., Renshaw, D., Yang, H. et al. Euphytica (2010) 176: 49. doi:10.1007/s10681-010-0212-1

Abstract

The reduced pod shattering gene tardus is one of the most important domestication genes in narrow-leafed lupin (Lupinus angustifolius L.). In development of a molecular marker linked to the tardus gene, we incorporated the concept of marker validation during the initial candidate marker identification stage. Four dominant microsatellite-anchored fragment length polymorphism (MFLP) markers were identified as candidate markers based on their banding patterns in an F8 recombination inbred line (RIL) population. One specific marker best correlating with phenotypes in the representative germplasm was selected and converted to a simple PCR-based marker. This established marker, designated as “TaLi”, is located at a distance of 1.4 cM from the tardus gene. DNA sequencing revealed six insertion/deletion sites between the non-shattering marker allele and the shattering marker allele. Validation of marker TaLi on 25 domesticated commercial cultivars and 125 accessions of the lupin core collection found a 94% marker and tardus phenotype match. Marker TaLi is the first simple PCR-based marker that can be widely used for non-shattering pod selection in narrow-leafed lupin breeding program.

Keywords

Marker-assisted selection (MAS)Lupinus angustifolius L.Sequence-specific markerMFLP

Introduction

Narrow-leafed lupin (Lupinus angustifolius L.), is an economically important legume crop cultivated in Australia. The ability to retain its seeds in pod long enough to allow mechanical harvesting at full maturity is an essential characteristic of a grain crop. However, seed pods of wild L. angustifolius shatter upon maturity. Domestication of this species in Australia began in 1960s with the discovery of two natural mutant (recessive) genes, tardus and lentus, each conditioning partially reduced pod-shattering, and only a plant possessing both two genes are fully non-shattering (Gladstones 1967). The expression of the tardus gene is influenced by environmental factors, especially the humidity and temperature, and is complicated by the presence or absence of the lentus gene. The lentus allele can be easily identified because it is 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). However, accurately phenotyping of tardus gene is challenging when introgressing new allelic variation from wild germplasm (pod-shattering) into domesticated breeding pools (non-shattering) in L.angustifolius.

Molecular markers can be developed to tag genes of interests, and then be applied in conventional breeding programs to select individual plants bearing these genes. However, the development of a molecular marker for marker-assisted selection (MAS) in plant breeding is a very challenging task. Many studies have pointed out that the effectiveness and usefulness of MAS depends on the stability of the marker in different genetic backgrounds and the distance of the marker to the actual gene (Khan et al. 2007). Also, the marker needs to be consistent with the phenotype on a wide range of breeding germplasm so that the marker can be broadly applied for MAS in breeding programs (Lin et al. 2009). In pepper, for example, a number of molecular markers were developed linked to the Restorer-of-fertility gene but were not useful due to the low phenotypic association rate (27–58%) until a more accurate marker was developed showing a prediction rate of 89% (Jo et al. 2010). Recently, Boersma et al. (2009) reported the development of three locus-specific markers flanking the tardus gene. However, these markers only had a 24–39% association between phenotypes and marker genotypes on the 33 wild accessions tested; and the best marker was a cleaved amplified polymorphic sequence (CAPS) marker, which is not cost-effective for MAS (Konieczny and Ausubel 1993). Previous attempts to convert these markers to easily applicable simple PCR markers failed.

In the endeavor to develop markers with wider applicability, Yang et al. (2008) proposed a strategy in identification of multiple candidate markers linked to the gene of interest followed by a validation step to select a best candidate marker before converting it into an implementable form. The objective of this study is to apply a marker development strategy to incorporate the validation step during the candidate marker identification step, and to develop a cost-effective molecular marker tagging the tardus gene which can be applied for MAS to a wide range of crosses in lupin breeding.

Materials and methods

Plant materials and phenotyping of tardus gene

A wild type P27255 (pod-shattering genotype Tardus/Tardus) collected from Morocco was crossed with a domesticated breeding line 83A:476 (non-shattering genotype tardus/tardus), and the progeny of this wild × domesticated (W×D) cross was advanced to F8 recombinant inbred lines (RILs) by randomly taking one plant from each line in each generation of single seed decent. This F8 RIL population was used in marker development linked to the tardus gene.

The parents and the 115 RILs were grown in a 2-m row in a screen-house. Plants were either classified as non-shattering (tardus/tardus) or shattering (Tardus/Tardus). Detailed methods of phenotyping of the tardus gene were described else where (Boersma et al. 2009).

Identification of candidate markers linked to tardus gene by MFLP

The DNA fingerprinting method microsatellite-anchored fragment length polymorphisms (MFLP) (Yang et al. 2001) was used in generating candidate molecular markers linked to the tardus gene. A total of 320 sets of MFLP fingerprints were generated by 16 AFLP primers MseI-ANN (Vos et al. 1995) in combination with 20 SSR-anchor primers listed in Table 1.
Table 1

Sequences of 20 SSR-anchor primers used in MFLP fingerprinting for generating candidate molecular markers linked to gene tardus conferring non-shattering pod in Lupinus angustifolius L

Primer name

Primer sequences (5′ → 3′)a

MF01

GTCCGAGAGAGAGAGA

MF02

GGCATGTGTGTGTGTG

MF11

GGACCTCTCTCTCTCT

MF21

CCCAAGAGAGAGAGAGAG

MF23

GGGCAGAGAGAGAGAG

MF42

GTCTAACAACAACAACAAC

MF43

CCTCAAGAAGAAGAAGAAG

MF51

GGGAACAACAACAAC

MF62

CCCAAACAACAACAAC

MF127

DBDACACACACACACA

MF128

DVDTCTCTCTCTCTCTC

MF129

HVHTGTGTGTGTGTGTG

MF151

CACGTCTCTCTCTCTCT

MF153

CCTTACACACACACAC

MF154

GAATCACACACACACA

MF155

CAACTGTGTGTGTGTG

MF201

CCCATTGTTGTTGTTG

MF202

GGAATTGTTGTTGTTG

MF203

GGGATTCTTCTTCTTCT

MF204

CCCTTTCTTCTTCTTC

aB = C+G + T, D = A+G + T, V = A+G + C, H = A+C + T

Twenty-four plants were used in each MFLP fingerprint, which included 16 representative plants from the F8 RIL population from the W×D cross, together with four domesticated cultivars and four wild types. The four cultivars, Mandelup, Tanjil, Quilinock and Myallie, were chosen to represent their wide genetic diversity based on their pedigrees; and the four wild types, P27253 (from Morocco), P20650 (Portugal), P20712 (Italy) and P21517 (Israel), were chosen to represent their different geographical origins. DNA extraction and MFLP tests followed the protocol of Yang et al. (2001). MFLP markers showing banding patterns corresponding to the tardus/Tardus phenotypes of the 16 F8 RIL population were considered as candidate markers (Yang et al. 2002; Yang et al. 2004). The candidate marker which showed best correlation between phenotypes and marker genotypes on both four domesticated cultivars and four wild types, respectively, was selected for further study to convert into a simple PCR marker for implementation for MAS (Yang et al. 2008).

Conversion of selected candidate MFLP marker into a sequence-specific PCR marker

DNA fragments from the selected candidate markers were isolated from the dried MFLP gel, re-amplified by PCR, cloned and sequenced by standard procedures. A pair of sequence-specific primers was designed to flank the DNA variation sites, which were used in PCR to convert the marker into a sequence-specific PCR marker. Detailed methods of DNA cloning and marker conversion were described elsewhere (Yang et al. 2002).

Linkage confirmation between the marker and the tardus gene

The converted marker was tested on the F8 population containing 115 RILs derived from the W×D cross. The tardus phenotypic data and marker genotypic data on the 115 RILs were merged and analyzed by the software program MapManager (Manly et al. 2001) to determine the genetic linkage between the marker and the tardus gene. A linkage group corresponding to part of LG1 (Boersma et al. 2009) was established using MapDraw (Liu and Meng 2003).

Validation of converted marker on commercial cultivars and wild types of L. angustifolius

All 25 historical and current commercial cultivars of L. angustifolius released in Australia were tested with the established sequence-specific marker. All these 25 cultivars possess the non-shattering gene tardus. They are Uniwhite (released in 1967), Uniharvest (1971), Unicrop (1973), Marri (1976), Illyarrie (1979), Yandee (1980), Chittick (1982), Danja (1986), Geebung (1987), Gungurru (1988), Yorrel (1989), Warrah (1989), Merrit (1991), Myallie (1995), Kalya (1996), Wonga (1996), Belara (1997), Tallerack (1997), Tanjl (1998), Moonah (1998), Quilinock (1999), Jindalee (2000), Mandelup (2004), Coromup (2006), and Jenabillup (2008). The marker was also validated by testing on the core collection of L. angustifolius at the Australian Lupin Collection housed at Department of Agriculture and Food Western Australia. The core collection consists of 125 lupin accessions (Table 2), which was chosen from over 2000 lines of L. angustifolius based on geographic data and on DNA fingerprinting data. Phenotypic data for the tardus gene on the core collection lines were obtained from the database of the Australian Lupin Collection. The correlation of the tardus phenotype data and the marker genotype data on the commercial cultivars and core collection accessions were used to evaluate the usefulness of the established marker for MAS in lupin breeding. All plant materials used in the research were obtained from the Australian Lupin Collection housed at Department of Agriculture and Food Western Australia.
Table 2

Validation of sequence-specific PCR marker “TaLi” on the 30 representative core accessions of Lupinus angustifolius L

Accession number

Country of origin

Phenotype

The state of marker “TaLi”a

P20657

Greece

Shedding

TaLiW3

P20672

Australia

Shedding

TaLiD

P20681

Australia

Non shedding

TaLiD

P20724

Italy

Shedding

TaLiW1;TaLiW3

P20725

Italy

Shedding

TaLiW3

P20729

Russia

Shedding

TaLiD

P22661

Israel

Shedding

TaLiW3

P22664

Spain

Shedding

TaLiW1; TaLiW3

P22665

Spain

Shedding

TaLiW2

P22666

Spain

Shedding

TaLiW1

P22711

Spain

Shedding

TaLiW3

P22739

Spain

Shedding

TaLiW4

P22829

Portugal

Shedding

TaLiW3

P22831

Portugal

Shedding

TaLiW1;TaLiW4

P22836

Portugal

Shedding

TaLiW1

P22847

Morocco

Shedding

TaLiW3

P24023

Spain

Shedding

TaLiW3

P25041

Italy

Shedding

TaLiW5

P25059

France

Shedding

TaLiW5;TaLiW6

P25065

France

Shedding

TaLiW1

P25073

France

Shedding

TaLiW3

P25077

France

Shedding

TaLiW6

P26095

Greece

Shedding

TaLiW1;TaLiW3;TaLiW4

P26098

Greece

Shedding

TaLiW4

P26100

Greece

Shedding

TaLiW4

P26103

Greece

Shedding

TaLiW4

P26107

Italy

Shedding

TaLiW3

P26112

Spain

Shedding

TaLiW3;TaLiW6

P28196

Algeria

Shedding

TaLiW2;TaLiW5;TaLiW6

P29065

Belarus

Shedding

TaLiW3

aTaLiD non-shattering marker allele band, TaLiW1−6 shattering marker allelic bands based on DNA size differences

Results

Identification of multiple candidate markers linked to the tardus gene

Four sets of dominant DNA polymorphisms were identified as candidate markers linked to the tardus gene based on the banding patterns generated by 320 primer combinations on a total of 24 plants (Table 3). These candidate markers showed a clear phenotypic match with at least 15 out of 16 genotypes tested (Table 4). However, three of the candidate markers, Marker 1, 2, and 3, exhibited four “false positives” on eight representative cultivars and wild types (Table 4), indicating that they would be less useful for MAS (Yang et al. 2008). Take Marker 1 for example, the four wild types not possessing the non-shattering gene tardus showed the domesticated marker allele band, although the marker had the banding pattern completely consistent with the tardus phenotypes on all the four representative domesticated cultivars (Table 4). The candidate MFLP marker, Marker 4, which was from MFLP fingerprint produced by SSR-primer MF202 in combination with AFLP primer MseI-ACT (Fig. 1), was the only one that showed a 100% marker/phenotype match on those representative cultivars and wild types (Table 4). Therefore, “Marker 4” was selected for further study in conversion into a sequence-specific PCR marker (Table 3).
Table 3

List of candidate markers identified from MFLP fingerprints linked to tardus gene in Lupinus angustifolius L.

Candidate marker

Primer combination in MFLP

Allele linked

Approximate size (bp)

Marker 1

MseI-AGC + MF01

Wild allele

150

Marker 2

MseI-ACT + MF23

Domesticated allele

450

Marker 3

MseI-AAA + MF43

Wild allele

272

Marker 4

MseI-ACT + MF202

Wild allele

400

Table 4

The banding patterns of candidate MFLP markers linked to the tardus gene on the 24 plants in Lupinus angustifolius L.

Plant No.a

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

Phenotypeb

ta

ta

ta

ta

ta

ta

ta

ta

Ta

Ta

Ta

Ta

Ta

Ta

Ta

Ta

ta

ta

ta

ta

Ta

Ta

Ta

Ta

Marker 1

Mta c

Mta

Mta

Mta

Mta

Mta

Mta

Mta

MTa

MTa

MTa

MTa

MTa

MTa

MTa

MTa

Mta

Mta

Mta

Mta

Mta

Mta

Mta

Mta

Marker 2

Mta

Mta

Mta

Mta

Mta

Mta

Mta

Mta

MTa

MTa

MTa

MTa

MTa

MTa

MTa

MTa

MTa

Mta

MTa

MTa

MTa

MTa

Mta

MTa

Marker 3

Mta

MTa

Mta

Mta

Mta

Mta

Mta

Mta

MTa

MTa

MTa

MTa

MTa

MTa

MTa

MTa

Mta

Mta

Mta

Mta

Mta

Mta

Mta

Mta

Marker 4

Mta

MTa

Mta

Mta

Mta

Mta

Mta

Mta

MTa

MTa

MTa

MTa

MTa

MTa

MTa

MTa

Mta

Mta

Mta

Mta

MTa

MTa

MTa

MTa

aThe 24 plants were domesticated parent 83A:476 (1), RIL69 (2), RIL13 (3), RIL86 (4), RIL85 (5), RIL18 (6), RIL95 (7), RIL12 (8), RIL34 (9), RIL26 (10), RIL27 (11), RIL73 (12), RIL104 (13), RIL80 (14), RIL48 (15), RIL19 (16), domesticated cultivar Mandelup (17), Tanjil (18), Quilinock (19), Myallie (20), wild type P27253 (21), P20650 (22), P20712 (23) and P21517 (24)

bPhenotypes: ta tardus/tardus (homozygous non-shattering); Ta Tardus/Tardus (homozygous shattering)

cMarker genotypes : Mta marker allele band linked to the tardus allele; MTa marker allele band linked to the Tardus allele

https://static-content.springer.com/image/art%3A10.1007%2Fs10681-010-0212-1/MediaObjects/10681_2010_212_Fig1_HTML.gif
Fig. 1

Identification of a dominant DNA polymorphism associated with the non-shattering gene tardus in L. angustifolius from a MFLP fingerprint generated by SSR-anchor primer MF202 in combination with AFLP primer MseI-ACT. Eight of the testing plants contained the non-shattering gene tardus, which were from the F8 RIL derived from a W×D cross, including the domesticated parent 83A:476 (Lane 1), RIL68 (Lane 2), RIL13 (Lane 3), RIL86 (Lane 4), RIL85 (Lane 5), RIL18 (Lane 6), RIL85 (Lane 7) and RIL18 (Lane 8). The other eight plants from the same F8 RIL population were pod shattering, including RIL34 (Lane 9), RIL26 (Lane 10), RIL27 (Lane 11), RIL73 (Lane 12), RIL104 (Lane 13), RIL80 (Lane 14), RIL48 (Lane 15) and RIL19 (Lane 16). At candidate marker identification stage, DNA fingerprinting also included four cultivars with the tardus gene of Mandelup (Lane 17), Tanjil (Lane 18), Quilinock (Lane 19) and Myallie (Lane20); and four pod-shattering wild types of P27253 from Morocco (Lane 21), P20650 from Portugal (Lane 22), P20712 from Italy (Lane 23) and P21517 from Israel (Lane 24). The arrow indicates the dominant markers linked to the tardus gene

Conversion of selected candidate marker into a simple PCR based marker

DNA sequencing and alignment of the selected candidate marker indicated that the wild type allele marker band was 309 bp in length (Fig. 2). A pair of sequence-specific primers, TaLiF (5′-ATCCTACTAAATCCTGGTACAG-3′) and TaLiR (5′-GATCTGAAAAGGAATATGAAG-3′), were designed based on the marker sequences (Fig. 2). This pair of primers successfully converted the candidate MFLP marker into a co-dominant, sequence-specific PCR marker ideal for routine marker implementation (Fig. 3). This marker is designated as “TaLi”. The domestication gene allele marker band “TaLiD” was 511 bp long, and the wild type allele marker band “TaLiW” from the wild parent plant P27255 was 309 bp (Fig. 2). DNA sequencing revealed six insertion/deletion sites between the non-shattering marker allele band and the shattering marker allele band on the marker sequences we discovered (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs10681-010-0212-1/MediaObjects/10681_2010_212_Fig2_HTML.gif
Fig. 2

DNA sequences obtained during development of sequence-specific PCR marker TaLi linked to the tardus gene conferring non-shattering pod in Lupinus angustifolius L. The six insertion/deletion sites are shaded. aMseI-ACT primer (5′-GATGAGTCCTGAGTAAACT-3′) which was employed in MFLP analysis. bSequence-specific primer TaLiF (5′-ATCCTACTAAATCCTGGTACAG-3′). cAnnealing site for sequence-specific primer TaLiR (5′-GATCTGAAAAGGAATATGAAG-3′). dAnnealing site for SSR-primer MF202 (5′-GGAATTGTTGTTGTTG-3′) which was used in MFLP analysis

https://static-content.springer.com/image/art%3A10.1007%2Fs10681-010-0212-1/MediaObjects/10681_2010_212_Fig3_HTML.gif
Fig. 3

Screening of the sequence-specific PCR marker TaLi on the lupin lines of the core collection of L. angustifolius. The 30 lines illustrated were P20657 (Lane 1), P20672 (Lane 2), P20681 (Lane 3), P20725 (Lane 4), P20729 (Lane 5), P22661 (Lane 6), P22664 (Lane 7), P22665 (Lane 8), P22666 (Lane 9), P25041 (Lane 10), P25055 (Lane 11), P25077 (Lane 12), P26098 (Lane 13), P26100 (Lane 14), P26103 (Lane 15), P26107 (Lane 16), P26129 (Lane 17), P26170 (Lane 18), P26279 (Lane 19), P26297 (Lane 20), P26303 (Lane 21), P26314 (Lane 22), P26353 (Lane 23), P26357 (Lane 24), P26363 (Lane 25), P26382 (Lane 26), P26394 (Lane 27), P26608 (Lane 28), P27255 (Lane 29), and 83A:476 (Lane 30). The top band TaLiD is linked to the non-shattering gene tardus. Six allelic bands were identified as linked to the pod-shattering alleles, including TaLiW1, TaLiW2, TaLiW3, TaLiW4, TaLiW5 and TaLiW6. Some wild types, such as P26297 (Lane 20), P26303 (Lane 21) and P26357 (Lane 24), exhibited more than one allelic band, reflecting the genetic heterogeneity of the core collection accessions

Linkage confirmation between the established marker and the tardus gene

Analysis of the marker score data from TaLi on the F8 population containing 115 RILs derived from the W×D cross by computer program MapManager (Manly et al. 2001) indicated that marker TaLi is tightly linked to the non-shattering gene in narrow-leafed lupin, and the genetic distance between the marker and the tardus gene is 1.4 cM (LOD = 28.3) (Fig. 4).
https://static-content.springer.com/image/art%3A10.1007%2Fs10681-010-0212-1/MediaObjects/10681_2010_212_Fig4_HTML.gif
Fig. 4

A genetic linkage map of gene tardus with established markers on linkage group 1 of Lupinus angustifolius L.

Verification of converted marker TaLi on commercial cultivars and wild types of L.angustifolius

Marker TaLi was tested on all 25 historical and current commercial cultivars possessing the gene tardus, all the cultivars showed the non-shattering allele TaLiD. When the marker was tested on the core collection, a total of six wild type alleles were found (Fig. 3). The domesticated type allele was the longest of 511 bp (Fig. 3). Among the 125 accessions of the narrow-leafed lupin core collection, nine possessed the domesticated gene, which all showed the marker band TaLiD. 116 accessions were wild types, among which 15 were heterozygous showing both the TaLiD and TaLiW alleles. Therefore, the overall accurate prediction rate of marker TaLi on the cultivars and on the core collection was 94%.

Discussion

The identification of tightly linked molecular markers to traits of interest is a pre-requisite for MAS (Gupta et al. 1999; Charcosset and Moreau 2004). In this paper, we reported the development of a sequence-specific marker TaLi and its genetic location on the chromosome. Comparing to the other three markers developed by Boersma et al. (2009), it is more closely linked to tardus gene. The close linkage promises a high accuracy when the marker TaLi is employed for MAS to select the tardus gene in lupin breeding. Marker TaLi is co-dominant, enabling lupin breeders to select homozygous plants and discard heterozygous plants at the F2 stage, an outcome which is impossible to achieve by traditional selection method. Since marker score is based on DNA, it overcomes the difficulty encountered in conventional lupin breeding of mis-selection of pod shattering individuals which may naturally germinate from the previous crop. Therefore, MAS based selection for the gene tardus by marker TaLi provides significant advantages over conventional selection method in lupin breeding. This is the first simple PCR-based marker that can be used in a wide range of non-shattering pods selection in narrow-leafed lupin breeding program.

An essential requirement of large scale implementation of a molecular marker for MAS in a plant breeding program is that the marker banding pattern must consistent with the phenotypes of the target gene on a wide range of germplasm, so that the marker can be applied to a large number of crosses (Holland 2004; Yang et al. 2008). However, most molecular markers are not part of the targeted genes (Staub et al. 1996; Gupta et al. 1999), and genetic recombination can occurred between the gene and the marker. Therefore, a plant showing the desirable marker band may not necessarily possess the targeted genes (“false positive”) (Sharp et al. 2001; Yang et al. 2008); in such case the marker can not be applied for MAS to screen the progenies for crosses where such a plant is used as a parent (Snape 2004). Fortunately, marker TaLi showed the correct banding patterns on all the 25 Australian domesticated cultivars and on 116 out of 125 of the accessions in the L. angustifolius core collection, suggesting that marker TaLi is applicable to most of the W×D crosses in lupin breeding.

In this investigation, we found six wild marker allele bands TaLiW1, TaLiW2, TaLiW3, TaLiW4, TaLiW5 and TaLiW6, which are allelic bands based on DNA size differences. The parental wild type allele band converted into the marker was TaLiW3. Therefore, it is evident that at least six wild marker allele bands exist in L. angustifolius apart from the domesticated allele marker band TaLiD. In this study, we found that some wild accessions in the core collection showed a mixture of two or three allele bands for marker TaLi, which is a reflection of the genetic heterogeneity of the core collections. It is important to note that shattering pod lupin accessions in the core collections showed a mixture of the domesticated allele marker band together with the wild type allele bands for marker TaLi, although the genetic heterogeneity in the wild lupin collection is expected and understandable. The possibility is the contamination of domesticated seeds into the wild accessions during planting and machine harvesting when the wild accessions were grown out after years’ storage in the cold room to keep the vitality. This is particularly evident for the two accessions P22803 and P26446, which part of the plants have white seed coat and green stem (typically domesticated characters) and other plants have dark seed coat and purple stem (typically wild characters). Another possibility is that all the plants in a wild accession showing the mixture of marker bands are all shattering, but a part of the plant may have the TaLiD marker band as “false positive” in the same way as the other nine “false positive” wild accessions identified in this study. In either case, if such wild accessions are needed in the crossing, lupin breeders are advised to screen individual plants and plants showing the TaLiD marker band should be discarded before the remaining plants are used for crossing with domesticated lines in breeding, so that TaLi can be confidently used for MAS.

Marker TaLi is a valuable resource for lupin genetic research. The recent development of a bacterial artificial chromosome (BAC) library of narrow-leafed lupin (Kasprzak et al. 2006) and the availability of genetic transformation protocols for narrow-leafed lupin (Pigeaire et al. 1997) makes DNA marker-based chromosome landing possible in this species. L. angustifolius is a diploid with 2n = 40 chromosome and an estimated DNA content of 2C = 1.89−2.07 pg (Hajdera et al. 2003; Naganowska et al. 2003). Using the conversion factor of Dolezel et al. (2003), this DNA content is equivalent to a haploid genome length of 924–1012 Mb. The total map length based on 1118 marker and trait loci is 2361.8 cM (Nelson et al. 2010). Then a very rough estimation based on the assumption of equal crossover frequency across the genome is that 1 cM is equivalent to approximately 391–428 kb of DNA in this RIL population. Marker TaLi was 1.4 cM away from the tardus gene, equivalent to approximately 547–599 kb or about 6 BAC clones in length. Therefore, chromosome walking or golfing might be possible approaches to reach the targeted gene.

A high-density, evenly distributed DNA marker genetic map is a prerequisite for a successful chromosome landing. In this paper, we applied a marker development strategy by incorporating some representative cultivars and wild types at the initial candidate marker identifications stage for the identification of best candidate markers with highest matching rate between phenotypes of the targeted gene and the marker genotypes. The principle of identifying markers with broad applicability proposed in this study is consistent with the strategy of Yang et al. (2008). However, the marker development method used in investigation is simpler than that of Yang et al. (2008), since there was no need of a separate step in candidate marker validation step. By choosing the best one from the four candidate markers, we established marker, TaLi, showed a very high correlation between tardus phenotypes and marker genotypes at 94%, which is in sharp contrast with the low correlation rates reported by Boersma et al. (2009) where no validation concept was involved during marker development, even though three markers were developed. The broad applicability of marker TaLi demonstrated the relevance and the importance for including cultivars and wild types at the candidate marker identification stage. This strategy will, therefore, be applied in our future work to identify more tightly linked markers to target genes/traits and to construct high-density maps of L. angustifolius.

Acknowledgments

This research was supported by the Grain Research and Development Corporation (Australia) project DAW0170. Xin Li is supported by a China Scholarship Council–The University of Western Australia Joint PhD Scholarship.

Copyright information

© Springer Science+Business Media B.V. 2010