Introduction

Plum pox virus (PPV) causes devastating disease in Prunus trees (Ravelonandro et al. 2000). PPV disease, often referred to as sharka disease, affects stone fruits such as apricot, peach, plum, and cherry, which are economically important species of Prunus. Originating in Bulgaria in the early 1900s (Atanasoff 1932), PPV has spread throughout Europe (Nemeth 1994) and, as of 1999, has made its way to the Americas.

Detection of PPV in 1999 in Adams County, PA prompted the implementation of an aggressive eradication program to control its spread (Levy et al. 2000). Eradication is an extremely costly way of controlling PPV. Ultimately, the introduction of resistant cultivars of stone fruits into the orchards is the best long-term solution to virus control. Currently, there are two strategies for obtaining resistant cultivars: (1) identify natural resistance present in Prunus germplasm and introduce this resistance into commercial cultivars by standard breeding practices, and (2) engineer resistance through application of transgenic technologies. In the first case, selecting for previously identified natural resistance in breeding programs requires screening of individual trees for the resistance phenotype. This labor-intensive time-consuming process drastically slows the progress of breeding programs. However, marker-assisted selection (MAS), based on markers tightly associated with resistance, is a means to substantially streamline the breeding process. In the case of transgenic approaches to engineer resistance, a transgenic plum developed by Scorza et al. (1994) has a durable PPV coat protein-mediated resistance (Hily et al. 2004). Unfortunately, the inability to reliably regenerate transgenic plants with other Prunus species hinders the application of transgene-mediated resistance.

Peach is the most genetically characterized Prunus species; however, no known natural resistance to PPV has been identified in peach. In contrast, North American apricot cultivars, such as ‘Stark Early Orange’ and ‘Goldrich’, are resistant to PPV (Polák et al. 1995). Resistant apricot cultivars, along with the apricot’s small genome size of 294 Mb (Arumuganathan et al. 1991) and haploid number of n=8, facilitate the study of the genetics of PPV resistance. Recently, three genetic maps for apricot have been published (Hurtado et al. 2002; Vilanova et al. 2003a; Lambert et al. 2004). The maps published by Hurtado et al. (2002) and Vilanova et al. (2003a) contain mapped locations for PPV resistance, and both maps are anchored on the general Prunus genetic map (Joobeur et al. 1998). This Prunus genetic reference map is the comparative map standard for the Prunus community and contains anchor markers for integration of genetic information from all Prunus species (Aranzana et al. 2003, Dirlewanger et al. 2004). In addition, Lalli et al. (2005) published the first resistance gene map for Prunus identifying 42 resistance gene containing loci on the general Prunus map. The combined reference or resistance map serves as a resource for identification of candidate resistance genes in Prunus breeding materials and for utilization of MAS in resistance breeding.

Using ‘Goldrich’ as the donor of resistance, Hurtado et al. (2002) mapped sharka resistance (PPV resistance) as a single locus controlled trait to a linkage group corresponding to linkage group one (G1) of the general Prunus map. Using a different donor of resistance, ‘Stark Early Orange’, Vilanova et al. (2003a) similarly mapped resistance to a region of G1; however, it was not clear if the map locations for this resistance co-localized from the different donors. Many hypotheses exist for the number of genes controlling the PPV resistance trait. One hypothesis suggests that a single gene controls the resistance trait (Dicenta et al. 2000; Martinez-Gomez and Dicenta 2000). Moustafa et al. (2001a, b) suggest two genes control the trait, and Guillet-Bellanguer and Audergon (2002) suggest three genes control the trait. The research conducted by Vilanova et al. (2003a) is in agreement with the two-gene hypothesis suggested by Moustafa et al. (2001a, b).

The goals of the work presented in this communication are the characterization of the genetics of PPV resistance and the development of co-dominant molecular markers for MAS in PPV resistance breeding strategies. The first goal involves determining the number of genes controlling the resistance trait and whether mapped resistance loci in populations generated from the same donor of resistance but different susceptible cultivars localize to the same region of the genome. The second goal, development of markers for MAS, relies partially upon the first goal. If mapped resistance loci, generated from the same donor of resistance but different susceptible cultivars, localize to the same region of the genome, molecular markers tightly associated with these loci will have great utility in MAS. In order to investigate the genetics of PPV resistance and develop markers for MAS, we constructed the first genetic linkage map for an apricot backcross population segregating for resistance to PPV using ‘Stark Early Orange’ as the resistance donor. This map is primarily comprised of amplified fragment length polymorphism (AFLP) markers. Simple sequence repeat (SSR) markers from the general Prunus reference map were chosen to anchor our apricot map to the general Prunus map. Four AFLP markers segregating with the PPV resistance trait, identified through bulk segregant analysis (BSA), facilitated the development of SSRs in this region. Due to the co-dominant nature of SSRs along with their high genetic transportability, the development of SSRs associated with PPV resistance in apricot could facilitate the use of MAS in breeding strategies aimed at breeding for natural resistance. Additionally, these markers can enhance fine mapping of this region and positional cloning of genes that may contribute to PPV resistance.

This map is a valuable tool for locating the regions involved in PPV resistance and the development of molecular markers with in these regions. Together with studies of synteny between peach and apricot currently being conducted and the availability of a genome sequence for peach in the near future, this map will speed the genetic studies of PPV resistance and greatly contribute to solving the problem of the spread of PPV in Prunus.

Materials and methods

Plant material

An apricot selection LE-3218 (‘Stark Early Orange’ × ‘Vestar’) was crossed as a female parent to ‘Stark Early Orange’ at the Faculty of Horticulture of Mendel University of Agriculture and Forestry in Lednice na Moravě in 1998. ‘Stark Early Orange’ is resistant to PPV. LE-3218 is susceptible to PPV. Crosses were performed by hand pollination without isolation of flowers after removing of petals and anthers from flower buds. The BC1 seeds were stratified at 5oC for 3 months and subsequent seedlings were grown in an insect-proof greenhouse. This plant material was used in BSA.

An apricot selection LE-3246 (resistant to PPV; F1 individual from ‘Stark Early Orange’ × ‘Vestar’) was crossed as a female parent to ‘Vestar’ (susceptible to PPV) at Faculty of Horticulture of Mendel University of Agriculture and Forestry in Lednice na Moravě in 1999. The BC1 seeds were stratified and the subsequent seedlings were grown in an insect-proof greenhouse. Eighty BC1 individuals were used to construct the preliminary AFLP genetic map. Due to potential out-crossing, 67 individuals were used to construct the final genetic map.

PPV inoculation and evaluation of PPV infection

When the stems of the seedlings reached about 5 mm thickness, they were inoculated with a chip-bud from a plum infected with the PPV isolate from apricot cultivar Vegama (Marcus strain; Poncarová and Komínek 1998). Seedlings were pruned directly after grafting to promote the growth of the inoculated bud. Plants without sharka symptoms on shoots growing from the inoculum bud and with negative enzyme-linked immunosorbent assay (ELISA) reaction were re-inoculated. PPV infection was evaluated over five consecutive growth periods through visual symptoms and ELISA (Polák et al. 1997). Pruning was performed at the beginning of each growth period to induce vigorous new shoots for symptom scoring. The plants, in which PPV was not detected by ELISA, were tested by reverse transcription polymerase chain reaction (RT-PCR) using the PPV specific primers P1 and P2 (Wetzel et al. 1991). Plants were classified as resistant if they did not show symptoms and positive ELISA or RT-PCR reaction in the last three growth periods evaluated.

DNA extraction

Genomic DNA was isolated from fresh apricot leaves using the hexadecyltrimethylammonium bromide (CTAB) protocol described by Eldredge et al. (1992). DNA concentrations were measured by a minifluorimeter (TKO100, Hoefer Scientific). Working solutions of genomic DNA at 100 ng/µl in TE buffer (pH 8.0) were prepared for AFLP analysis.

AFLP amplification

AFLP analysis was performed as described by Vos et al. (1995) with a pre-amplification step using primers with one selective base each. EcoRI primers were end-labeled with γ33P-ATP (Andotek) and T4 polynucleotide kinase (Promega). PCR products were separated on 6% denaturing polyacrylamide gels, which were pre-run to reach 45°C. Four microliters of samples were loaded on to gels and electrophoresed at 70 V for 2 h and 30 min, vacuum dried, exposed to Biomax MR films (Kodak) for 1–7 days, and developed. Twelve EcoRI primers with one selective nucleotide and 23 MseI primers with three selective nucleotides were used to amplify AFLP bands. A total of 47 (EcoRI or MseI) primer combinations were used for the AFLP analysis. AFLP markers were named using the selective nucleotides in the EcoRI primer or the selective nucleotides in the MseI primer followed by numbers in descending molecular-size order.

Bulk segregant analysis

BSA was performed according to Michelmore et al. (1991). Equal amounts of pre-amplified DNA obtained from ten resistant individuals and ten susceptible individuals, respectively, of the BC1 population were pooled to constitute the resistant and susceptible bulks (Bulk+, Bulk−). Each bulk represented the alternative phenotypic states of the loci controlling the selected trait. Two bulk samples along with the parents were subjected to AFLP analysis.

Simple sequence repeat development

Cloning and sequencing of AFLP marker fragments

AFLP bands were cut from dry acrylamide gels and soaked in 100 μl of sterile distilled water overnight at 4°C to extract the DNA according to Wang et al. (2002). The extracted DNA was used as a template for PCR amplification using primers specific for each AFLP marker fragment. The amplification was carried out in a 25-μl volume containing 5 μl of extracted DNA, 90 ng of each AFLP marker fragment specific primer, 2 units of Taq DNA polymerase (Fisher Scientific), 1 X PCR Buffer (10 mM Tris-HCl pH 9.0, 50 mM KCl), 1.5 mM MgCl2, 200 μM each deoxyribonucleotide triphosphate (dNTP). PCR amplifications were programmed for 12 cycles for 30 s at 94°C, 30 s at 65–0.7°C per cycle, 1 min at 72°C and 22 cycles for 30 s at 94°C, 30 s at 56°C, 1 min at 72°C. A 5-μl aliquot of the PCR reaction was separated by electrophoresis on an agarose gel to confirm the amplification of the expected size fragments. PCR products were ligated into Promega's pGEM®-T Easy Vector and transformed into JM109 cells (Promega) according to the manufacturer's protocol. A minimum of five clones per AFLP marker fragment were sequenced using M13 universal forward and reverse primers and Applied Biosystems Big Dye® Terminator Sequencing Kit and an ABI 377 automated sequencer.

BAC library screening with labeled AFLP marker fragments

DNA was isolated from the clones containing the AFLP marker fragments. Marker fragments were PCR-amplified in a 25-μl reaction containing 5 μl of a 1/100 dilution of plasmid DNA, 90 ng of each AFLP marker fragment specific primer, 2 units of Taq DNA polymerase (Fisher Scientific), 1 X PCR Buffer (10 mM Tris-HCl pH 9.0, 50 mM KCl), 1.5 mM MgCl2, 200 μM each dNTP. PCR amplifications were programmed as specified above. A 2-μl aliquot of the PCR reaction was separated by electrophoresis on an agarose gel in order to confirm the amplification of fragments of the expected size and DNA concentration. Approximately 10–50 ng of linear DNA was labeled with 15 μCi [α32P]dCTP using the random priming method (Feinberg and Vogelstein 1983). Labeled DNA probes were purified on Sephadex G25 columns. Labeled AFLP marker fragment probes were used in hybridizations to a BAC library from the apricot cultivar ‘Goldrich’. This library consists of 101,376 clones that are arrayed on six 22 cm2 Hybond-N+filters (Amersham, Piscataway, NJ, USA; Vilanova et al. 2003b). The library filter set was pre-hybridized for 2 h and hybridized overnight, according to Vilanova et al. (2003b). Radioactive signals were detected autoradiographically using Kodak X-OMAT Blue XB-1 film. In order to confirm the positive BACs, BAC clones identified in the first screening were inoculated into 100 μl of lysogeny broth (LB)–Chloramphenicol, and incubated at 37°C overnight. BAC clones were stamped onto Hybond-N+filters (Amersham, Piscataway, NJ, USA) placed on LB–Chloramphenicol agar plates, and incubated overnight at 37°C. The filters were removed from the agar plates and treated with a denaturing solution (1.5 M NaCl, 0.5 M NaOH) followed by a neutralizing solution (1.5 M NaCl, 0.5 M Tris, pH 7.2, 1.0 mM EDTA), rinsed with 2X SSC and the filters were baked at 80°C for 2 h in order to fix the DNA to the filters. Hybridization of these filters and detection of positive BAC clones were carried out as previously described.

Sub-cloning of confirmed positive BACs and identifying and sequencing SSRs

BAC insert DNA was purified for at least one confirmed positive BAC per AFLP marker fragment probe. Purified BAC DNA was digested with Hind III and ligated into Hind III digested, dephosphorylated p-Bluescript vector (Stratagene, La Joya, CA, USA), according to Wang et al. (2002). Sub-clones of confirmed positive BACs were selected and incubated in 100 μl LB/ampicillin overnight at 37°C. Sub-clones were stamped onto Hybond-N+filters (Amersham, Piscataway, NJ, USA), placed on LB/ampicillin agar plates, and incubated overnight at 37°C. Filters were removed from the agar plates and the colonies were denatured, neutralized, and fixed on the filters as previously described (see above). The filters were hybridized with a probe containing (CT)10, (GT)10, and (AGG)10 SSR oligonucleotides. The probes were end-labeled with [γ-33P]dATP using T4 polynucleotide kinase (Promega) according to the manufacturer’s protocol. The hybridizations were carried out at 55°C using the same hybridization conditions previously stated. The filters were washed twice with 2X SSC, 0.1% SDS and exposed to Kodak X-OMAT Blue XB-1 film. Insert DNA of BAC sub-clones identified as containing SSRs was isolated and sequenced as previously described (see above). Primers for eight SSRs were designed from the sequences flanking the SSRs using the online primer design program Primer3 (Rozen and Skaletsky 2000) and purchased from Integrated DNA Technologies (Table 1).

Table 1 SSRs derived from apricot BACs containing AFLP marker fragments segregating with PPV resistance
Full size table

SSR analysis

‘Stark Early Orange’, ‘Vestar’ and LE-3246 were screened with 117 SSR primer combinations from Aranzana et al. 2003 and the eight SSR primer combinations developed from the BACs containing the AFLP marker fragments associated with PPV resistance. PCR reactions, electrophoresis, and detection of PCR products for the ‘Vestar’ × LE-3246 population were carried out according to conditions specified in Zhebentyayeva et al. (2003). SSR primer combinations revealing polymorphism were screened in the entire mapping population.

Linkage analysis

MAPMAKER 3.0

All markers were scored as dominant markers. Loci heterozygous in LE-3246 were used for mapping. A local genetic map was constructed with mapmaker/exp 3.0 software (Lincoln et al. 1992). The mapping analysis was conducted using a 4.0 minimum logarithm of odds (LOD) score and 0.25 maximum recombination frequency (θ). The Kosambi mapping function (Kosambi 1944) was used for calculation of map distances.

Joinmap®3.0

The cross-pollinating population type of the Joinmap®3.0 (Van Ooijen and Voorrips 2001) mapping program was used to establish a genetic map that includes all marker classifications for both ‘Vestar’ and LE-3246. Chi-square analysis (available in Joinmap®3.0) was used to determine if markers departed significantly from expected Mendelian ratios. Linkage groups were established at a LOD score ≥ 4.5 and recombination fraction 0.40. The Kosambi mapping function was used for the calculation of map distances. Criteria for goodness of fit (χ 2) of markers in the map were applied as indicated by Dettori et al. (2001).

Kruskal–Wallis test

Due to the highly skewed segregation ratio observed for the PPV resistance trait, a non-parametric mapping method based on the rank sum test of Kruskal–Wallis was performed on the linkage group G1 data (generated from “Joinmap®3.0”, with the MapQTL version 3.0 program (Van Ooijen an Maliepaard 1996)).

Results

PPV segregation analysis

Monogenic, digenic, and trigenic models have been proposed for the number of genes controlling the PPV resistance trait. Moustafa et al. (2000a, b) suggest two genes control the trait and the research conducted by Vilanova et al. (2003a) is in agreement with this two-gene model. Eighty individuals from the LE-3246 × ‘Vestar’ progeny were screened for PPV resistance. Plants were classified as resistant if they did not show symptoms and/or positive ELISA or RT-PCR reaction in the last three growth periods evaluated. Thirteen individuals were found resistant and 67 susceptible. Resistant individuals were coded as heterozygous for the trait and those susceptible were coded as homozygous recessives (consistent with Vilanova et al. 2003a). The segregation ratio 1:5 (resistant/susceptible deviated significantly from the expected for a single dominant locus (1:1) with χ 2 value of 36.5 (resistant/susceptible)). These results clearly indicate that the resistance is controlled by more than a single gene. A PPV resistance locus was mapped in linkage group 1 (Fig. 1) with MAPMAKER 3.0. Several AFLP loci tightly linked to the PPV resistance locus were identified (Table 1).

Fig. 1
figure 1

Linkage group G1 genetic map for LE-3246 generated with MAPMAKER 3.0

Full size image

Bulk segregant analysis

Two hundred and seventy-one primer pair combinations were used to compare the AFLP profiles of ‘Stark Early Orange’ (S), LE-3218 (L), the resistant bulk (+), and the susceptible bulk (-). Plants were identified as resistant by the method previously described (see “PPV segregation analysis”). Eleven AFLP markers (AA/CGT, AA/GCG, AA/GGA, AT/CCT, GA/GGT, GC/CTT, GG/CGA, GT/CTT, TA/GGC, TC/CCT and TG/GCA) were detected only in ‘Stark Early Orange’ and the resistant bulk. All the markers were linked to the gene in the coupling phase. These markers were used in segregation and linkage analysis in the mapping population (LE-3246 × ‘Vestar’). Five AFLP markers (AA/GCG, AA/GGA, AT/CCT, TC/CCT and TG/GCA) detected by BSA were localized on the LE-3246 map. Three markers (GA/GGT, GC/CTT, TA/GGC) were not polymorphic in the LE-3246 × ‘Vestar’ mapping progeny, two markers (AA/CGT, GG/CGA) did not work in the mapping progeny and one marker (GT/CTT) remains independent.

Local genetic map for PPV resistance locus (gene)

The map consisted of 22 dominant AFLP markers, two dominant SSR markers, and the PPV resistance locus and covered a total genetic distance of 130.6 cM. Linkage group 1 that contains the PPV resistance locus is depicted in Fig. 1. The marker loci are distributed relatively evenly within this linkage map. None of the markers showed a skewed segregation pattern.

Cloning and sequencing of AFLP marker fragments

The four AFLP marker fragments identified from bulk segregant analysis and segregating with the PPV resistance trait were cloned into Promega’s pGEM®-T Easy Vector (Table 1). A minimum of five clones per AFLP marker fragment were sequenced and the sequences analyzed by inspection. A primary and secondary sequence were identified for the AFLP fragment EAA/MCAT clones and a primary, secondary, and tertiary sequence were identified for the AFLP fragment EAG/MCAT, EAT/MCCT, and ETC/MCCT clones. The primary sequence was assigned to the sequence that was present in a minimum of three out of the five clones sequenced per AFLP fragment. This sequence was chosen for use in all further experiments.

Identification of BAC clones and development of SSR markers from BAC sub-clones

The primary sequence for each of the four AFLP marker fragments was hybridized individually to an apricot BAC library. The results of these hybridizations are depicted in the online supplemental material. The 11 BAC clones were restriction digested with Hind III. From inspection of the restriction patterns, two of the four BAC clones that hybridized to AFLP marker fragment EAA/MCAG8b and one BAC clone for each of the other AFLP marker fragments were chosen for sub-cloning into p-Bluescript vector. Sub-clones for each of these BACs were colony lifted and hybridized to oligonucleotide probes (CT)10, (GT)10, and (AGG)10. The sub-clone inserts identified as putatively containing SSRs were end-sequenced. Eight SSRs were identified and primers were designed from the flanking sequence for each SSR (Table 1).

AFLP analysis

Forty-seven primer combinations were analyzed in the LE-3246 × ‘Vestar’ mapping population. From this analysis, 423 AFLP markers were used in the construction of the genetic map. One hundred twelve (26.5%) AFLP loci in the F1 hybrid and 112 (26.5%) in the recurrent parent ‘‘Vestar’’ segregated in a 1:1 ratio. One hundred and ninety-nine AFLP loci (47.0%) segregated in both parents in a 3:1 ratio. Deviation (P < 0.05) from the expected Mendelian ratio was observed for 74 AFLP loci (17.5%).

SSR analysis

‘Stark Early Orange’, ‘Vestar’, and LE-3246 were screened with 117 SSR primer combinations derived from the general Prunus map SSR collection and the eight SSR primer combinations developed from the BACs containing the PPV resistance associated AFLP marker fragments (Table 1). Seventy-five proved to be polymorphic (60%) and were analyzed in the entire mapping population. Only six of the eight SSRs developed from BACs containing the AFLP marker fragments segregating with PPV resistance were polymorphic. Segregation was demonstrated for 38 of the 75 SSRs (50%) exhibiting polymorphism. These SSRs were used in the construction of the genetic map. Among them, nine segregated in the recurrent parent ‘Vestar’, 18 in the F1 hybrid, and ten in both parents (Table 2). Only one locus with three alleles was detected. Eight SSRs (21%) deviated (P < 0.05) from the expected Mendelian ratio.

Table 2 SSR markers mapped in LE-3246 x ‘Vestar’ population
Full size table

Linkage map

A total of 450 genetic markers were analyzed for the genetic map construction of the apricot BC1 population (LE-3246 × ‘Vestar’). The map consists of 357 loci (330 AFLPs, 26 SSRs, and one morphological marker for PPV resistance) assigned to eight linkage groups (Figs. 1, 2, and 3). Nomenclature and orientation for the eight linkage groups is assigned according to the genetic reference map for Prunus (T × E; Joobeur et al. 1998). Twenty-two of the SSRs placed on the map anchor this map to the genetic reference map for Prunus. The average number of loci per linkage group is 45. The range of the number of loci per linkage group is 19 (G5) to 96 (G1). The map covers 522.69 cM of the apricot genome. The range of the length of each linkage group is 51.48 (G8) to 87.90 cM (G1). Linkage group G1 has the highest marker density with an average distance of 0.92 cM per marker and the least saturated linkage group G5 with an average of 3.03 cM per marker. Four gaps (>10 cM) ranging from 10.1 to 13.74 cM were observed in this map.

Fig. 2
figure 2

LE-3246 × ‘Vestar’ map: linkage groups G1, G2, G3, and G4. Genetic linkage map generated with the mapping program Joinmap 3.0. Black star indicates the Joinmap 3.0 map location for the PPV resistance trait. Black arrow indicates the MAPMAKER 3.0 map location for the PPV resistance trait

Full size image
Fig. 3
figure 3

LE-3246 × ‘Vestar’ Map: Linkage groups G5, G6, G7, and G8. Genetic linkage map generated with the mapping program Joinmap 3.0

Full size image

Fifty-nine of the 357 mapped loci (16.53%) exhibited distorted segregation and are distributed in the map as follows: G1 (14 loci), G2 (8 loci), G3 (0 loci), G4 (12 loci), G5 (1 loci), G6 (13 loci), G7 (8 loci), and G8 (three loci; Tables 3 and 4).

Table 3 Distorted loci (P < 0.05) mapped in the LE-3246 x ‘Vestar’ apricot BC1 population linkage groups G1, G2, and G4
Full size table
Table 4 Distorted Loci (P < 0.05) mapped in the LE-3246 x ‘Vestar’ apricot BC1 population. Linkage groups G5, G6, G7, and G8
Full size table

Kruskal–Wallis test

Several hypotheses for the number of genes controlling the PPV resistance trait exist (Dicenta et al. 2000; Martinez-Gomez and Dicenta 2000; Moustafa et al. 2001a, b; Guillet-Bellanguer and Audergon 2002). Additionally, the segregation ratio (1:5, resistant/susceptible) of PPV resistance in the LE-3246 × ‘Vestar’ mapping progeny deviates significantly from the single dominant gene segregation ratio of 1:1. Therefore, the Kruskal–Wallis test was performed for all markers of the LE-3246 × ‘Vestar’ linkage map. Only markers present in the upper half of linkage group G1 showed significance levels, P < 005 for the Kruskal–Wallis test statistic (Table 5; Fig. 4; see “Supplementary material”). Eleven markers showed significance levels, P < 0.00001. This suggests that a putative quantitative trait locus (QTL) for the PPV resistance trait may reside in the region of G1 between 1.5 and 40.9 cM.

Table 5 Kruskal–Wallis statistic adjusted for ties (K*) for those markers located at the LE-3246 x ‘Vestar’ map linkage group G1
Full size table
Fig. 4
figure 4

Kruskal–Wallis statistic adjusted for ties for those markers located in the LE-3246 × ‘Vestar’ map linkage group G1

Full size image

Discussion

We used AFLP markers to generate a framework genetic map for characterizing PPV resistance in the apricot BC1 population LE-3246 × ‘Vestar’. EcoRI primers with one selective nucleotide and MseI primers with three selective nucleotides were used to amplify AFLP fragments. A total of 423 AFLP polymorphic fragments were identified with 17.5% of the AFLP markers showing distorted segregation ratios in the progeny. This percentage of distorted markers is slightly higher than the published work in apricot by Vilanova et al. (2003a; 17%), Hurtado et al. (2002; 12%), and Lambert et al. (2004; 10.2%). Distribution of AFLP markers in this map is fairly uniform. Clustering of AFLP markers in centromeric regions was not observed as reported in tomato by Haanstra et al. (1999).

SSR marker detection with primers developed from peach, apricot, and almond (Sosinski et al. 2000; Cipriani et al. 1999; Testolin et al. 2000; Aranzana et al. 2002) demonstrates the high transportability of SSRs in Prunus (Dirlewanger et al. 2002). ‘Stark Early Orange’, ‘Vestar’, and LE-3246 were screened with 125 SSR primer combinations. Sixty percent of these SSRs were polymorphic. This percentage of polymorphism is slightly higher than that previously reported by Hurtado et al. (2002) and Vilanova et al. (2003a; 46% and 42%, respectively) in apricot. However, 12.5% of the SSRs primers detecting polymorphism were derived from apricot and may account for this difference.

We identified four AFLP marker fragments as segregating with PPV resistance through bulk segregant analysis. SSR primers aprigms 17–24 were developed from BACs containing the AFLP fragments segregating with PPV resistance (Table 1). Two of these SSR primers, aprigms 18 and aprigms 24, detected polymorphisms that map to G1. This work presents the first genetic map based on an apricot BC1 population. It consists of 357 loci (330 AFLPs, 26 SSRs, and 1 morphological marker for PPV resistance) assigned to eight linkage groups (Figs. 1, 2, and 3). Since 22 of the mapped SSRs are shared in common with genetic reference map for Prunus (T × E; Joobeur et al. 1998), we employed the nomenclature and orientation for the eight linkage groups of this reference map. The map covers 522.7 cM of the apricot genome with an average distance of 1.46 cM/marker. The length of the map is consistent with the apricot maps published by Hurtado et al.(2002), Vilanova et al. (2003a), and Lambert et al. (2004) and the average distance per marker is less than all other maps published for apricot (Hurtado et al. 2002; Vilanova et al. 2003; Lambert et al. 2004).

Approximately 17% of the mapped loci exhibited distorted segregation with G1 and G6 exhibiting the largest number of distorted loci. Vilanova et al. (2003a) observed similar results in apricot.

A PPV resistance locus was mapped in linkage group 1 (Fig. 1) with MAPMAKER3.0. PPV resistance also mapped to linkage group G1 using the Joinmap3.0 program; however, the map locations are not the same (Fig. 1). The location for PPV resistance mapped in Joinmap for the BC1 population for LE-3246 × ‘Vestar’ (Fig. 3) and the location tentatively placed in the ‘Lito’ × ‘Lito’ F2 populations appear to be the same location (Vilanova et al. 2003a). Additionally, Romero et al. (2005) performed the Kruskal–Wallis test for mapping PPV as a putative QTL in the ‘Lito’ × ‘Lito’ population. Markers in the upper half of linkage group G1 were significant for the Kruskal–Wallis test statistic (between 12 and 38.4 cM). The same test was performed for all markers of the LE-3246 × ‘Vestar’ linkage group G1. Markers present in the upper half of linkage group G1 were significant for the Kruskal–Wallis test statistic (P < 0.005; between 0.3 and 40.9 cM; Table 5; Fig. 4; see “Supplementary material”). The highest level of significance was detected in the same region in both mapping populations and indicates that a putative locus controlling PPV resistance may exist in this region. The most interesting observation is that the non-parametric mapping location for PPV resistance in ‘Lito’ × ‘Lito’ (Romero et al. 2005) and LE-3246 × ‘Vestar’ is very similar to the mapped location for PPV resistance in the LE-3246 × ‘Vestar’ progeny map constructed with MAPMAKER 3.0 (Fig. 1). Additionally, two SSRs, aprigms18(1) and EPDCU5100(1), were significant for the Kruskal–Wallis test statistic and may be useful for MAS for breeding for PPV resistance. Two different strategies employed to map the PPV resistance trait have generated two unique map locations that may putatively represent two genes controlling the PPV resistance trait. Further evaluation of these loci will be necessary to characterize the genetic control of the PPV resistance trait.

PPV resistance is a complex trait controlled by at least two genes. In an effort to characterize PPV resistance, we constructed the first genetic linkage map for an apricot backcross population segregating for resistance to PPV. The map is comprised of AFLP markers along with SSR markers from the general Prunus reference map that anchor our apricot map to the general Prunus map. This also facilitates the comparison of our map with other genetic maps for Prunus. Four AFLP markers segregating with the PPV resistance trait, identified through BSA, facilitated the development of SSRs in this region. Due to the co-dominant nature of SSRs along with their high genetic transportability, the development of SSRs associated with PPV resistance in apricot could facilitate the use of MAS in breeding strategies aimed at breeding for natural resistance. Additionally, these markers can serve as landmarks to aid in fine mapping of this region as well as enable positional cloning of genes that may contribute to PPV resistance. Mapping the PPV resistance trait and the development of molecular markers associated with trait will be extremely valuable for identifying candidate resistance genes when the genome sequence for peach becomes available.