| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Dep. of Crop and Soil Science, Oregon State Univ., Corvallis, OR 97331-3002, USA
b Pioneer Hi-Bred International, 18285 County Road 96, Woodland, CA 95695-9340, USA
c Pioneer Genetique S.A.R.L., Le Soulou, Ferme Barbara, 82700 Montech, France
d Pioneer Hi-Bred International, 7300 N.W. 62ND Avenue, P.O. Box 1004, Johnston, IA 50131-1004, USA
e ICAR-Long Ashton Research Station, Dep. of Agricultural Sciences, Univ. of Bristol, Long Ashton, Bristol, BS18 9AF, UK
f Advanta Seeds UK Ltd., Station Road, Docking, King's Lynn, Norfolk, PE31 8LS UK
g Advanta Seeds, Balcarce Research Station, Ruta 226, KM 60.3 (7620), Balcarce PCIA DE BS. AS., Argentina
h Advanta Biotechnology Department, SES-Europe NV/SA, Industriepark, Soldatenplein Z2 nr. 15, B-3300 Tienen, Belgium
* Corresponding author (steven.j.knapp{at}orst.edu)
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The development of 1089 SSR markers for cultivated sunflower (Gedil, 1999; Tang et al., 2002; Yu et al., 2002) eliminated the long-standing bottleneck caused by the scarcity of single-copy DNA markers in the public domain and supplied the critical mass of DNA markers needed to create a public reference map, unify independently developed molecular genetic linkage maps, and establish an universal linkage group nomenclature. Tang et al. (2002) constructed the first genetic linkage map for sunflower on the basis of SSR markers and the first dense public genetic linkage map on the basis of single- or low-copy DNA markers. The map was developed with RILs from a cross between confectionery and oilseed fertility restorer lines (RHA280 x RHA801) (Fick et al., 1974; Roath et al., 1981). Of the 1089 SSR markers described by Tang et al. (2002) and Yu et al. (2002), 717 were polymorphic among elite inbred lines and 408 were polymorphic in RHA280 x RHA801. The polymorphic SSR markers amplified one to three loci each and segregated for 462 SSR marker loci in RHA280 x RHA801. Of the segregating SSR marker loci, 459 coalesced into 17 linkage groups (x = 17) and three were unlinked. The genetic linkage map was 1368.3 cM long, had a mean density of 3.1 cM per locus, and supplied a genome-wide framework of sequence-based DNA markers for constructing and cross referencing new genetic linkage maps. Of the public SSR markers known to be polymorphic among elite inbred lines, 300 out of 717 have not yet been mapped (Yu et al., 2002; Tang et al., 2002). Additionally, 156 proprietary SSR markers developed for sunflower as part of the CARTISOL program (see acknowledgments) have not been screened for length polymorphisms or mapped. The public (ORS) and CARTISOL (CRT) SSR markers supply a total of 1245 SSR markers for molecular breeding and genomics research in sunflower.
SSR marker resources developed for sunflower create the basis for rapidly, efficiently, and fully integrating first generation genetic linkage maps developed by use of RFLP markers (Berry et al., 1995, 1996, 1997; Gentzbittel et al., 1995, 1999; Jan et al., 1998). While SSRs and other high-throughput, sequence-based DNA markers are supplanting RFLP markers and spawning second generation genetic linkage maps (Tang et al., 2002), the RFLP resources developed for sunflower are significant and underexploited. Collectively, 1141 RFLP loci have been mapped by means of three independent sets of proprietary probes, primarily cDNA clones (Knapp et al., 2001). Because the three RFLP maps of sunflower lack common, public domain DNA markers, quantitative trait locus (QTL) and other trait mapping results cannot be universally exploited, compared, or validated. The probe collections for the mapped RFLP markers represent an extraordinary resource for developing sequence-based DNA markers, saturating the molecular genetic linkage map of sunflower, and performing comparative analyses between sunflower, model species [e.g., Arabidopsis thaliana (L.) Heynh.], and other crop plant genomes. One goal of the present research was to integrate and cross reference the Tang et al. (2002) SSR map with the RFLP maps of Berry et al. (1997) and Jan et al. (1998) using the Gedil et al. (2001b) RFLP map as a bridge. The latter was constructed from HA370 x HA372 F2 progeny and RFLP markers from the maps of Berry et al. (1997) and Jan et al. (1998). The other goal of the present research was to increase the density and utility of the molecular genetic linkage map of cultivated sunflower by adding unmapped CRT SSR markers to the RHA280 x RHA801 map and unmapped ORS SSR markers to the HA370 x HA372 map and constructing a new genetic linkage map by means of ORS SSR markers and RILs from a cross between two elite fertility restorer lines (PHA x PHB).
|
MATERIAL AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
HA370, HA372, PHA, and PHB were screened for polymorphisms with two sequence characterized amplified region (SCAR) markers (SCTO6151 and SCX20600) (Lawson et al., 1998) and 1089 public SSR markers (Gedil, 1999; Yu et al., 2002; Tang et al., 2002). SCAR marker assays were performed as described by Lawson et al. (1998). SSR genotyping assays were performed as described by Tang et al. (2002) on an ABI Prism 377 DNA Sequencer (Applied Biosystem, Perkin-Elmer, Foster City, CA) by means of polyacrylamide gels and fluorescently labeled amplicons. Filter set C and the GeneScan TAMRA500 internal standard were used for assays performed with 6FAM, TET and HEX labeled amplicons. Filter set D and the GeneScan ROX500 internal standard were used for assays performed with 6FAM, HEX, and NED labeled amplicons. The amplicons for each SSR marker were separately produced and pooled post-PCR. Three-plexes based on three different fluorophores (e.g., 6FAM, HEX, and TET) were used for polymorphism screening. Three-color multiplexes were used so that the DNA fragments produced by each SSR primer pair could be unequivocally identified. Genotypes (SSR allele lengths) were recorded by GeneScan and Genotyper and manually checked. Genotypes were only recorded for samples with fluorescent signal strengths greater than 100 fluorescence intensity units.
RHA280 and RHA801 were screened for polymorphisms by means of 156 proprietary (CARTISOL or CRT) SSR markers and 72 proprietary INDEL markers (unpublished data). The CRT SSR markers were developed from DNA sequences of clones isolated from genomic DNA libraries enriched for SSRs, essentially as described by Yu et al. (2002). The INDEL markers, hereafter ZVG INDEL markers, were developed by sequencing probes (cDNA clones) for 81 RFLP markers on the genetic linkage map of Berry et al. (1997), aligning sunflower cDNA and Arabidopsis genomic DNA sequences, predicting intron sites in sunflower genes from the presence of introns in the Arabidopsis genes, and designing flanking primers to amplify intron and flanking coding regions spanned by the primer pairs. INDEL markers were screened for length polymorphisms between RHA280 and RHA801 and genotyped in the RHA280 x RHA801 mapping population by means of agarose gels.
Genetic Linkage Map Construction
SSR genotyping assays in the mapping populations were performed by multiplexing three to 13 SSR markers per lane as described for the parent inbred line screening. The SCTO6151 SCAR marker (Lawson et al., 1998) and 117 public SSR markers (Gedil, 1999; Yu et al., 2002; Tang et al., 2002) were genotyped on 94 HA370 x HA372 F2 progeny (Gedil et al., 2001b). Genotypes for the SSR and SCAR marker loci were merged with genotypes for 80 RFLP marker loci, a resistance gene candidate (RGC) marker locus (RGC1A), and Pl1 from the map of Gedil et al. (2001a) (RGC1A was formerly designated HR-4W2). The SCTO6151 and SCX20600 SCAR markers (Lawson et al., 1998) and 243 public SSR markers (Gedil, 1999; Yu et al., 2002; Tang et al., 2002) were genotyped on 94 PHA x PHB RILs. We genotyped 91 CARTISOL SSR markers and 17 ZVG INDEL markers on 94 RHA280 x RHA801 RILs. Common prefixes were used in DNA marker names to identify the source of the loci on the various genetic maps. The prefix "ORS" was used for the public SSR marker loci developed by Gedil (1999), Yu et al. (2002), and Tang et al. (2002). The prefix "ZVG" was used for RFLP and corresponding INDEL markers from the RFLP maps described by Berry et al. (1997) and Gedil et al. (2001b). The linkage groups to which the ZVG RFLP and INDEL markers have been previously mapped (Berry et al., 1995, 1996, 1997; Gedil et al., 2001b) are noted by linkage group number suffixes, e.g., ZVG46-7 is an RFLP marker on linkage group 7. Such suffixes were limited to RFLP markers that have mapped to the same linkage groups and map positions on at least three genetic linkage maps. The prefix "UB" was used for RFLP markers from the RFLP maps described by Jan et al. (1998) and Gedil et al. (2001b). The prefix "CRT" was used for CARTISOL SSR markers (unpublished data). Linkage group number suffixes were used to identify individual loci produced by multilocus SSR markers, e.g., ORS718-1 and ORS718-3 are loci on linkage groups 1 and 3 amplified by the ORS718 SSR primer pair. If duplicated loci mapped to the same linkage group, then consecutive letters were used to identify individual loci within linkage groups, e.g., ORS814-7A and ORS814-7B are duplicated loci amplified by the ORS814 SSR primer pair on linkage group 7.
Statistical analyses were performed and genetic linkage maps were constructed by G-MENDEL (Holloway and Knapp, 1993) and MapMaker (Lander et al., 1987). Tests for segregation distortion were performed for each locus. Loci were grouped in the HA370 x HA372 F2 map by a likelihood odds (LOD) threshold of 3.0 and a recombination frequency threshold of 0.25. Loci were initially grouped in the PHA x PHB RIL map using a LOD threshold of 7.0 and a recombination frequency threshold of 0.25. If loci coalesced into more than one linkage group corresponding to one of the 17 known linkage groups in sunflower (Tang et al., 2002), then groups and orders were reestimated by a LOD threshold of 3.0. Several RFLP and SSR markers segregated as dominant markers. The HA370 x HA372 F2 map was constructed with dominant markers linked in coupling and unlinked dominant markers only to avert errors in the estimation of recombination frequencies and locus orders caused by dominant markers linked in repulsion (Knapp et al., 1995). Map distances (cM) were calculated by the Kosambi (1944) mapping function. RIL map distances were calculated by recombination frequency estimates corrected for multiple meioses under selfing (r), where
 |
The length of the sunflower genome was estimated by L + (2tL)/n, as proposed by Fishman et al. (2001), and
 |
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Three new genetic linkage maps were constructed by adding previously unmapped SSR and INDEL marker loci to the RHA280 x RHA801 map and previously mapped and unmapped SSR marker loci to the HA370 x HA372 and PHA x PHB maps (Fig. 1). Two SCAR markers (Lawson et al., 1998) and 117 ORS SSR markers were genotyped in HA370 x HA372, 243 ORS SSR markers were genotyped in PHA x PHB, and 91 CRT SSR and 17 ZVG INDEL markers were genotyped in RHA280 x RHA801. When merged with previously mapped RFLP and SSR marker loci, the new maps were comprised of a total of 1044 DNA marker loci (843 ORS SSR, 101 CRT SSR, 69 ZVG RFLP or INDEL, 28 UB RFLP, and three SCAR marker loci). Linkage groups were numbered on the basis of the nomenclature on the public reference map (Tang et al., 2002) and aligned by means of 200 ORS SSR and 55 ZVG RFLP-INDEL marker loci mapped in two or three mapping populations (common SSR, RFLP, or INDEL marker loci).
|
|
The HA370 x HA372 RFLP-SSR Map
The HA370 x HA372 map was constructed from 120 ORS SSR, 52 ZVG RFLP (Berry et al., 1995, 1997), 28 UB RFLP (Jan et al., 1998), and two SCAR marker loci (Lawson et al., 1998) dispersed among the 17 known linkage groups (Fig. 1). Of the SSR marker loci mapped in HA370 x HA372, 96 had been mapped in PHA x PHB, RHA280 x RHA801, or both and were selected to supply a strategically positioned framework of common SSR markers for integrating the RFLP maps of Berry et al. (1997), Gedil et al. (2001b), and Jan et al. (1998). The selected SSR marker loci mapped to the linkage groups predicted from the other two maps, and the 202 DNA marker loci coalesced into 17 linkage groups corresponding to the 17 linkage groups on the original RFLP map. The HA370 x HA372 RFLP-SSR map was 1275.4 cM long and had a mean density of 6.3 cM per locus. The number of SSR and RFLP marker loci per linkage group ranged from four on LG6 and LG12 to 22 on LG17 (Table 1). The 17 linkage groups ranged in length from 15.7 cM (LG6) to 107.7 cM (LG16) (Fig. 1), and DNA marker densities ranged from 3.8 cM per locus on LG17 to 13.8 cM per locus on LG12 (Table 1).
Genetic Mapping of SCAR Marker Loci Linked to Rust Resistance Genes
The SCAR marker SCTO6151 (Lawson et al., 1998) mapped to LG8 in both mapping populations close to a cluster of NBS-LRR resistance gene candidates (e.g., RGC1A) and the Pl1 gene for resistance to downy mildew [Plasmopara halstedii (Farl.) Berlese et de Toni] race 1 (Gedil et al., 2001a). SCTO6151 is tightly linked to the R1 gene for resistance to rust race 1 (Lawson et al., 1998). The SCAR marker SCX20600 mapped to LG13. SCX20600 is linked to the RADV gene for resistance to the Advance race of rust (Lawson et al., 1998). Using common RFLP and SSR markers (ORS799 and ZVG61-13) and comparing positions among RFLP, RFLP-SSR, and SSR maps, we found SCX20600 to map distal to the fertility restorer locus Rf1 (Fig. 1).
The Molecular Genetic Linkage Map of Sunflower: Linkage Group by Linkage Group
The 81 ZVG RFLP marker loci on the framework public reference map (Berry et al., 1997; Fig. 1) were chosen to span the sunflower genome, as circumscribed by 621 RFLP marker loci on the composite RFLP map described by Berry et al. (1996). We used the ZVG RFLP framework and ZVG INDEL markers on the RHA280 x RHA801 map to assess the overlap between the genomic regions spanned by SSR markers on the HA370 x HA372, RHA280 x RHA801, PHA x PHB maps and RFLP markers on the public reference map (Fig. 1). Working left to right or vice versa in Fig. 1, the reference RFLP map (linkage groups 1A, 2A,..., 17A) and three new maps (linkage groups 1B, 2B,..., 17B, 1C, 2C,..., 17C, and 1D, 2D,..., 17D) build connections between ZVG RFLP markers (A to B), ZVG RFLP and INDEL markers (A and B to C), ORS SSR markers (B to C to D), UB RFLP markers (B to A, C, and D), and CRT SSR markers (C to A, B, and D).
The linkage group assignments and locus orders for common ZVG RFLP-INDEL markers were identical among maps A, B, and C (map D lacks ZVG RFLP and INDEL markers). The linkage group assignments for common SSR marker loci were identical, and the locus orders were nearly identical across maps B, C, and D (map A lacks SSR markers). We had a sufficient number of common SSR, RFLP, and INDEL markers to orient most of the 68 linkage groups across the four maps. Specifically, 200 ORS SSR marker loci were shared between two or three of the new maps (maps B, C, and D) and 55 ZVG RFLP or INDEL loci were shared between one or two of the new maps (maps B and C) and the reference RFLP map (map A). The locus names for common ORS SSR marker loci are shown in bold, whereas the locus names for common ZVG RFLP or INDEL marker loci are underlined and shown in bold (Fig. 1). The linkage group assignments and locus orders for common ZVG RFLP-INDEL markers were 100% identical among maps A, B, and C (RFLP and INDEL markers were lacking on map D). Linkage group assignments for common SSR marker loci were 100% identical, whereas locus orders were nearly 100% identical across maps B, C, and D (SSR markers were lacking on map A).
LG1
LG1 on map D could not be unequivocally oriented because map D only had two closely linked common SSR markers in common with maps B and C. Several SSR markers mapped distal to ZVG1-1, the upper terminal RFLP marker on map A. SSRs seem to be lacking in the region spanning ORS552 to ZVG4-1 on map B (ZVG4-1 was the lower terminal RFLP on map A).
LG2
Several SSR marker loci mapped distal to ZVG5-2, the upper terminal RFLP marker on LG2 (map A). Moreover, LG2 seems to be populated end to end with SSR marker loci, despite a 31.3 cM gap on the upper arm of map C between ORS342 and ORS229. Presumably, the gap spans a region of high recombination.
LG3
The lower half of LG3 virtually lacked SSR markers on maps B, C, and D. The only SSR marker that mapped distal to ORS488 and ORS525 on the lower ends of maps B, C, and D was ORS13 on map B. The sparse region was flanked by ZVG11-3 and ZVG13-3 on map B. Conversely, several SSR markers mapped distal to ZVG10-3 on the upper end of LG3 (ZVG10-3 was 8.0 cM downstream from the terminal RFLP marker locus). SSRs spanned an approximately 35.3 to 42.3 cM region upstream of ORS432 and ZVG10-3. The confectionery x oilseed map of LG3 (map C) was particularly dense (Table 1 and Fig. 1).
LG4
On the basis of the positions of ZVG14-4 and ZVG16-4 on LG4 of maps A and C and ORS1262 on maps B, C, and D, we concluded that an approximately 12 cM region on the lower end of the chromosome lacked SSR marker loci. The gap on the upper end of the chromosome, flanked by ORS963 and ORS785 or ZVG14-4 and ZVG15-4, was present across the four maps and undoubtedly spans a region of high recombination.
LG5
On the basis of the positions of ZVG20-5 on maps A and B and ORS1024 on maps B, C, and D, the upper approximately 10 cM of LG5 distal to ORS774 may lack SSR marker loci. Otherwise, SSRs seem to be dispersed across the entire chromosome, as demarcated by terminal RFLP marker loci on the reference RFLP map. ORS1120 mapped distal to ZVG24-5, the lower terminal RFLP marker on map A. The approximately 25.0 cM gaps spanning ORS852, 626, and 240 on the lower end of LG5 of maps B and D were absent on map C, but presumably flank a region of high recombination.
LG6
LG6 on map B had only one common ZVG RFLP marker and could not be oriented to map A but was oriented to map C using common SSRs. Similarly, LG6 on map D had only one common DNA marker and could not be oriented to any of the maps. LG6 has historically been one of the sparsest linkage groups in maps produced from crosses between elite inbred lines (Berry et al., 1995, 1996, 1997; Perez-Vich et al., 2002). Based on the position of ZVG26-6 on maps A and C, the SSR markers on map C seem to span the genomic regions spanned by RFLPs on map A. The gap between ORS1193 and ORS1229 on map C parallels the gap between ZVG25-6 and ZVG26-6 on map A.
LG7
On the basis of the position of ZVG29-7 on maps A and B and ORS618 on maps B and D, we concluded that several SSR marker loci (e.g., ORS400 and ORS928) mapped distal to ZVG29-7, the upper terminal RFLP marker locus on the reference RFLP map. We identified a region on the lower end of LG7 where two sets of duplicated SSR marker loci (ORS814 and ORS328) were found, a rarity in sunflower. Similar to other linkage groups, a persistent gap was present on the lower ends of LG7 on maps B, C, and D (the gap was flanked by the common SSR marker loci ORS 901 and ORS1178). SSR marker loci seem to be lacking on the lower end of LG7 distal to ZVG32-7, the lower terminal RFLP marker on map A.
LG8
SSR marker loci completely span the regions spanned by RFLP marker loci on LG8; however, as in several other linkage groups, only a few SSRs are found in the distal regions (e.g., ORS153 on map D and ORS894 on maps C and D). We identified a persistent gap (flanked by ORS1152-8 and ORS780 on maps C and D) downstream of the region harboring downy mildew and rust resistance genes on LG8 (e.g., Pl1, Pl2, and R1) (Lawson et al., 1998; Vear et al., 1997; Gentzbittel et al., 1998; Gedil et al., 2001a; Bert et al., 2001; Bouzidi et al., 2002).
LG9
Several SSR markers (e.g., ORS1265 and ORS805) on LG9 mapped distal to ZVG38-9, the upper terminal RFLP marker locus on the reference RFLP map. On the basis of the distance between ZVG40-9 and ORS442, SSR markers seem to overlap the region spanned by RFLP markers on the lower end of LG9; however, because of heterogeneity of recombination, lack of ZVG markers in the region on maps C and D, lack of SSR markers in the region on map B, and sheer density of SSR markers on map C, we could not accurately align DNA markers spanning the lower end of LG9.
LG10
LG10 was one of the densest linkage groups produced. SSR marker loci mapped close to the terminal RFLP markers ZVG43-10 and ZVG48-10 and were dispersed throughout the chromosome. The upper end of LG10 on map B was sparse compared with maps C and D. While the region was blanketed with SSR marker loci on map C, a 37.8 cM gap was present on map B.
LG11
Several SSR marker loci (e.g., ORS621 and ORS1147-11) mapped distal to the terminal RFLP marker loci on the reference map (ZVG49-11 and ZVG53-11). UB6F2 mapped 16.5 cM distal to ZVG53-11, a region lacking other DNA markers on the reference RFLP map.
LG12
The polarity of LG12 on map B could not be ascertained because maps A and B only had one RFLP marker in common (ZVG55-12). On the basis of ORS767 and ORS1085 on maps B and C and ORS911 and ORS1085 on map D, we presume that the polarities shown were correct. Furthermore, on the basis of the distance between ZVG55-12 and ORS1085 on map B and ORS502 and ORS1085 on map C, SSR marker loci seem to blanket chromosome 12. Gedil et al. (2001b) had difficulty identifying polymorphic RFLPs, and we had difficulty identifying polymorphic SSRs in HA370 x HA372. The pattern may be exemplary of other crosses between sterility maintainer lines.
LG13
ORS200 and ORS799 on LG13 mapped distal or close to the two terminal RFLP markers (ZVG58-13 and ZVG61-13) on map B. On the basis of the position of ORS200, several SSR markers mapped distal to ZVG58-13 on map C, as did the RFLP marker UB6D6 on map B. The 30.7-cM gap on map C, another region of apparently high recombination, was not closed by adding new DNA marker loci. CRT76, however, added 5.8 cM to the upper end of the chromosome.
LG14
Several SSR marker loci mapped distal to ZVG62-14, the upper terminal RFLP marker on the reference map. On the basis of the position of ZVG64-14 on maps B and C, ZVG65-14 on map B, and ORS832 on maps C and D, we speculate that ORS1260 on map C and the group of SSR marker loci distal to ORS832 on the lower end of map D fill the region spanning ZVG64-14 and ZVG65-14 on maps A and B.
LG15
The new genetic linkage maps (B, C, and D) lack DNA markers from the reference RFLP map. ZVG85-15, an RFLP marker on map B but absent on map A, mapped to the lower end of LG15 on the RFLP-AFLP map of Gedil et al. (2001b), opposite of ZVG66-15, the upper terminal RFLP marker on map A; thus, on the basis of common SSR markers placed on maps B, C, and D (e.g., ORS8 and ORS151), the polarities shown for the three new maps probably match the polarity shown for map A. Finding polymorphic SSR markers for LG15 was difficult in map B. Perhaps LG15 tends to be highly monomorphic across sterility maintainer lines, as was proposed for LG12.
LG16
Map B lacked the upper two RFLP markers from map A and thus could not be used to ascertain the overlap between SSR and RFLP markers on the upper end of the chromosome. Nevertheless, on the basis of the distance between ORS899 and ZVG73-16, we speculate that SSR markers overlap the region flanked by ZVG71-16 and ZVG73-16. Long gaps were present between ORS902 and ORS807 on maps B and D.
LG17
On the basis of the position of ZVG77-17 on maps A, B, and C, SSR marker loci seem to be lacking in the approximately 14.7-cM region estimated to be upstream of ORS597 on map C. On the basis of the length of the region distal to ZVG80-17 on map B and common SSR markers on maps B, C, and D (e.g., ORS1051, ORS512, and ORS1203), SSR markers seem to span the remainder of the chromosome.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Of the individual genetic linkage maps developed thus far with single- and low-copy, probe- and sequence-specific DNA markers (e.g., RFLP or SSR markers) (Berry et al., 1995, 1996, 1997; Gentzbittel et al., 1995, 1999; Jan et al., 1998; Tang et al., 2002), the new RHA280 x RHA801 map is the densest and most complete (Fig. 1). By adding 101 SSR and 17 INDEL marker loci to the RHA280 x RHA801 map (Tang et al., 2002), we increased the number of DNA marker loci from 459 to 577 and increased the density of DNA markers from 3.1 to 2.5 cM per locus (Table 1), and produced a map as long and as dense as the 617-locus composite RFLP map described by Berry et al. (1996). The latter was constructed from nine F2 mapping populations, had a mean density of 2.3 cM per locus, and was 1472 cM long (Berry et al., 1996).
The RHA280 x RHA801 map was used to estimate the length of the genetic linkage map of sunflower (L) and the proportion of the sunflower genome circumscribed by SSR and INDEL marker loci. SSR marker loci were absent in five telomeric regions demarcated by RFLP marker loci; a 30.6-cM segment on LG1 distal to ORS552, a 31.5-cM segment on LG3 flanked by ORS13 and ZVG13-3, a 12.9-cM segment on LG4, a 10.0-cM segment on LG5, and a 14.7-cM segment on LG17. The combined length of the segments was 99.7 cM. By adding 99.7 cM to 1423.0 cM (the length of the RHA280 x RHA801 map), the sunflower genome was estimated to be 1522.7 cM long. Using the estimators proposed by Chakravarti et al. (1991) and Fishman et al. (2001), we estimated L to be in the neighborhood of 1529.6 to 1539.4 cM. Both estimates were close to the 1522.7-cM estimate produced by adding the length of reputedly missing segments to the length of the RHA280 x RHA801 map. On the basis of the three genome length estimates, we speculate that the RHA280 x RHA801 map (Fig. 1) spans approximately 93% of the sunflower genome (1423/1523 = 0.934 and 1423/1539 = 0.925). The proportion of the genome within d cM of a DNA marker locus, assuming a random distribution of DNA marker loci, was estimated to be 1 - e-2dk/L = 1 - e-2d(577)/(1539) = 0.528 on the RHA280 x RHA801 map (Chakravarti et al., 1991); thus, roughly 52.8% of the genome is within 1.0 cM and 99.9% of the genome is within 10.0 cM of a DNA marker locus on the RHA280 x RHA801 map.
Roughly 66% of the public SSR markers are known to be polymorphic among elite inbred lines (Tang et al., 2002; Yu et al., 2002) and a minimum of 821 out of 1245 SSR markers in the public and CARTISOL collections are known to be polymorphic. Of the 821 SSR markers, 606 (74%) segregated for one or more polymorphic loci and have been mapped (Fig. 1). On the basis of an analysis of a 122 single-copy SSR markers, many of the SSR markers found to be monomorphic among elite inbred lines are polymorphic among elite, exotic, and wild germplasm (Tang and Knapp, unpublished data); thus, additional SSR markers can be mapped and denser genetic linkage maps can be produced from elite x exotic and elite x wild crosses (Burke et al., 2002). Nevertheless, the SSR markers placed on the three maps described here represent the subset with the greatest utility for molecular breeding in the gene pools that dominate commercial hybrid breeding programs. Finally, because three-fourths of the most robust and polymorphic SSR markers have been mapped, a point of diminishing return has undoubtedly been reached in the process of closing gaps in and lengthening the map, e.g., the addition of 101 SSR marker loci to the RHA280 x RHA801 RIL map (Tang et al., 2002) only lengthened the map by 29.1 cM (Fig. 1). Perhaps many of the unmapped SSR markers can be mapped with progeny from a sterility maintainer x fertility restorer (B x R) cross, a combination that was not covered by our three mapping populations.
The present supply of public SSR markers seems sufficient for constructing complete genetic maps with progeny from crosses between elite inbred lines. The three maps were constructed with progeny from crosses designed to exploit contrasting patterns of DNA polymorphisms between confectionery, oilseed, fertility restorer, and sterility maintainer gene pools. Specifically, we used crosses between oilseed B-lines (HA370 x HA372), oilseed R-lines (PHA x PHB), and confectionery by oilseed R-lines (RHA280 x RHA801). Oilseed B x B and R x R crosses were used because breeders typically produce new B- and R-lines by selecting progeny from B x B and R x R crosses or by backcrossing elite B- or R-lines to segregating progeny from narrow and wide crosses; thus, the broad utility of DNA markers for molecular breeding in sunflower tends to hinge more on DNA polymorphisms in B x B and R x R crosses than in B x R, B x exotic, or R x exotic crosses, although the latter two are widely used in backcross breeding programs. Several analyses of molecular genetic diversity in sunflower have lumped elite inbred lines into B and R groups and shown that slightly less genetic diversity is present among B x B and R x R than B x R crosses (Berry et al., 1994; Gentzbittel et al., 1994; Hongtrakul et al., 1997; Paniego et al., 2002; Yu et al., 2002).
Two linkage groups on the HA370 x HA372 map (LG6 and 15) and two linkage groups on the PHA x PHB map (LG1 and 6) were sparsely populated with SSR markers and shorter than the corresponding linkage groups in RHA280 x RHA801. LG1, 6, and 15 were well covered with SSR markers on the confectionery x oilseed sunflower map. LG6 has been short in several maps and seems to be one of the least polymorphic chromosomes among elite oilseed inbred lines (Berry et al., 1995, 1997; Gedil et al., 2001b; Perez-Vich et al., 2002). LG1 was more polymorphic on the B x B than the R x R map. Conversely, LG15 was more polymorphic on the R x R than the B x B map. We speculate that LG1 is one of the least polymorphic chromosomes among elite R-lines, and that LG15 is one of the least polymorphic chromosomes among elite B- lines in sunflower. Genetic linkage maps constructed using additional B x B and R x R crosses are needed to test this hypothesis.
By using phenotypic marker loci (Pl1 and Rf1) common to autonomous RFLP linkage maps (Berry et al., 1995, 1996, 1997; Gentzbittel et al., 1995, 1999; Gedil et al., 2001a,b; Bert et al., 2001; Bouzidi et al., 2002) and mapping SCAR markers (SCTO6151 and SCX20600) linked to non-allelic rust resistant genes (R1 and RADV, respectively) (Lawson et al., 1998) in HA370 x HA372 and PHA x PHB (Fig. 1), we found that duplicated genes for resistance to rust reside in close proximity to duplicated genes for resistance to downy mildew on LG8 and 13. Bert et al. (2001) found that two nonallelic downy mildew resistance genes reside on different linkage groups in sunflower. The RFLP probe So17h3 hybridized to duplicated RFLP loci linked to Pl6 on LG8 and Pl5 on LG13 (Bert et al., 2001) (the public linkage group nomenclature is used here and throughout). Genes for resistance to downy mildew race 1 (Pl1) and 2 (Pl2) are linked to Pl6 (Vear et al., 1997) and reside on LG8 (Gedil et al., 2001a). Rf1 is linked to So17h3 and Pl5 and resides on LG13. Using common RFLP and SSR markers (ORS799 and ZVG61-13) and comparing positions across RFLP and SSR maps (Fig. 1), we found SCX20600 to map distal to the fertility restorer locus (Rf1). Thus, Pl5, RADV, and Rf1 map to the lower end of LG13 (Fig. 1).
The presence of common phenotypic and SCAR marker loci on LG8 and 13 permitted us to cross reference otherwise autonomous RFLP maps and compare mapping results between populations and laboratories. Such comparisons have heretofore not been possible on a genome- wide scale in sunflower and have been restricted to regions flanking commonly known phenotypic loci. The present research and the map of Tang et al. (2002) lays the groundwork for rectifying the problem and eliminating independent linkage group nomenclatures and creates the basis for fully integrating first generation RFLP and second generation SSR maps in sunflower. We initiated the process by adding a framework of 120 SSR marker loci to the HA370 x HA372 RFLP map. While the 81-locus public RFLP map of Berry et al. (1997) and the public SSR maps (Fig. 1) have been fully integrated, the two most significant RFLP maps (Berry et al., 1995, 1996; Gentzbittel et al., 1995, 1999) have not yet been integrated with the public SSR map.
The potential exists to create an integrated molecular genetic linkage map for sunflower comprised of 1800 or more RFLP and SSR marker loci by adding as few as 34 common, public SSR markers to each of the three proprietary RFLP maps (Berry et al., 1995, 1996; Jan et al., 1998; Gentzbittel et al., 1995, 1999). We partially integrated one of the three (Jan et al., 1998) with the public SSR map (Fig. 1). The sparse framework of UB RFLP markers on the HA370 x HA372 map facilitated the alignment of 10 linkage groups (LG2, 3, 4, 8, 9, 10, 11, 13, 16, and 17) on the public SSR map with corresponding linkage groups on the Jan et al. (1998) RFLP map. Because we only had one UB RFLP marker on each of three linkage groups (LG1, 6, and 7), the corresponding linkage groups on the Jan et al. (1998) RFLP map were identified but could not be aligned. Similarly, because four linkage groups (LG5, 12, 14, and 15) on the public SSR map lacked UB RFLP markers, the corresponding linkage groups on the Jan et al. (1998) map could not be identified.
The process of cross referencing the two maps could be completed by adding as few as 11 SSR markers to the Jan et al. (1998) RFLP map (two each to LG5, 12, 14, and 15 and one each to LG1, 6, and 7). Similarly, the other two RFLP maps of sunflower (Berry et al., 1996; Gentzbittel et al., 1999) can be cross referenced to the public SSR map by genotyping as few as 68 SSR markers (two per linkage group per map). The development of such a resource, if coupled with the public release of RFLP (cDNA) probes and DNA sequences, could dramatically affect genomics research in sunflower by instantaneously doubling the density of the molecular genetic linkage map, adding DNA markers to sparsely mapped regions of the genome, and creating expressed sequence landmarks for in silico syntenic comparisons of the sunflower map to maps developed in Arabidopsis and other plant species.
|
ACKNOWLEDGMENTS |
|---|
Received for publication March 26, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIAL AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. Yue, B. A. Vick, W. Yuan, and J. Hu Mapping One of the 2 Genes Controlling Lemon Ray Flower Color in Sunflower (Helianthus annuus L.) J. Hered., September 1, 2008; 99(5): 564 - 567. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Fonts, F. H. Andrade, M. Grondona, A. Hall, and A. J. Leon Phenological Characterization of Near-Isogenic Sunflower Families Bearing Two QTLs for Photoperiodic Response Crop Sci., July 1, 2008; 48(4): 1579 - 1585. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Kolkman, S. T. Berry, A. J. Leon, M. B. Slabaugh, S. Tang, W. Gao, D. K. Shintani, J. M. Burke, and S. J. Knapp Single Nucleotide Polymorphisms and Linkage Disequilibrium in Sunflower Genetics, September 1, 2007; 177(1): 457 - 468. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M. Wills and J. M. Burke Quantitative Trait Locus Analysis of the Early Domestication of Sunflower Genetics, August 1, 2007; 176(4): 2589 - 2599. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Yatabe, N. C. Kane, C. Scotti-Saintagne, and L. H. Rieseberg Rampant Gene Exchange Across a Strong Reproductive Barrier Between the Annual Sunflowers, Helianthus annuus and H. petiolaris Genetics, April 1, 2007; 175(4): 1883 - 1893. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Garcia-Moreno, E. M. Vera-Ruiz, J. M. Fernandez-Martinez, L. Velasco, and B. Perez-Vich Genetic and Molecular Analysis of High Gamma-Tocopherol Content in Sunflower Crop Sci., September 8, 2006; 46(5): 2015 - 2021. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Tang, A. Leon, W. C. Bridges, and S. J. Knapp Quantitative Trait Loci for Genetically Correlated Seed Traits are Tightly Linked to Branching and Pericarp Pigment Loci in Sunflower Crop Sci., February 24, 2006; 46(2): 721 - 734. [Abstract] [Full Text] [PDF]
|
