• Medicago truncatula (barrel medic) has emerged as a model legume and accession A17 is the reference genotype selected for the sequencing of the genome. In the present study we compare the A17 chromosomal configuration with that of other accessions by examining pollen viability and genetic maps of intraspecific hybrids.
• Hybrids derived from crosses between M. truncatula accessions, representative of the large genetic variation within the germplasm collection, were evaluated for pollen viability using Alexander's stain. Genetic maps were generated for the following crosses: SA27063 × SA3054 (n = 94), SA27063 × A17 (n = 92), A17 × Borung (n = 99) and A17 × A20 (n = 69).
• All F1 individuals derived from crosses involving A17 showed 50% pollen viability or less. Examination of the recombination frequencies between markers of chromosomes 4 and 8 revealed an apparent genetic linkage between the lower arms of these chromosomes in genetic maps derived from A17.
• Semisterility and unexpected linkage relationship are both good indicators of a reciprocal translocation. The implications of the A17 distinctive chromosomal rearrangement on studies of genetic mapping, genome sequencing and synteny between species are discussed.
Map-based cloning, synteny studies and some genome sequencing strategies have a common approach in relying on linkage maps to represent the genome order of a given species/accession. For example, one of the strategies to efficiently sequence plant genomes is the anchored clone-by-clone strategy (Wu et al., 2004; Sasaki et al., 2005; Udvardi et al., 2005; Young et al., 2005) by which genomic segments are sequenced and positioned following the order of the segments’ markers on a genetic map. The extent to which the genetic map is a good representation of the genome depends on whether the parental lines from which the maps derive share a common chromosomal configuration.
Medicago truncatula is an annual diploid (2n = 16) self-pollinating species widely used as a model plant for legumes (Cook, 1999; Frugoli & Harris, 2001). The accession A17, a single-seed descendant line from the cultivar Jemalong, has served as the primary reference for the development of tools, methods and infrastructure for molecular, genetic, structural and functional genomic studies (Choi et al., 2004a; Mun et al., 2006) and the genome sequencing effort (Young et al., 2005).
Limited information is available on the cytogenetics of the species; fluorescent in situ hybridization (FISH) studies have been conducted with only two M. truncatula accessions: cv. Jemalong, lines A17 and J5, and R108-1 (Cerbah et al., 1999; Kulikova et al., 2001, 2004; Falistocco & Falcinelli, 2003; Choi et al., 2004a). A reciprocal translocation between chromosome 4 and 8 has been suggested as a distinctive feature between the parents of the primary mapping population derived from a cross between A17 and A20 (Choi et al., 2004a; Cannon et al., 2005; Young et al., 2005). However, evidence for this translocation has not yet been published and there was no clear indication which of the two accessions had the idiosyncratic chromosomal arrangement. This information is critical given the status of A17 as the reference genotype for M. truncatula. This paper provides strong evidence that A17 bears a reciprocal translocation involving chromosomes 4 and 8.
Materials and Methods
Accessions and hybrids
The M. truncatula Gaertn. accessions A17, A20, Borung, Cyprus, SA1499, SA3054, SA8616, SA10419, SA11734, SA18395, SA23859, SA25654, SA27063, SA27192 and SA30199 were acquired from the Genetic Resource Centre, SARDI (South Australian Research and Development Institute, Adelaide, South Australia 5001). Accession DZ315.16 was from the Institut National de la Recherche Agronomique (INRA) Montpellier, France. The hybrids described in this paper were obtained by a manual crossing procedure (Thoquet et al., 2002) and their nature confirmed by polymerase chain reaction (PCR) using DNA markers polymorphic between the parental lines. The respective F2 progeny seeds are available upon request.
Pollen viability of intraspecific hybrids
Pollen viability was scored under the light microscope; spreads of 100–500 fresh pollen grains were stained with Alexander's stain (Alexander, 1969); aborted pollen stains pale turquoise blue and nonaborted pollen stains dark blue or purple (Fig. 1b,c). At least three flowers from one to six F1 plants were scored per cross combination.
The F2 populations SA27063 × SA3054 (n = 94), SA27063 × A17 (n = 92) and A17 × Borung (n = 99) were genotyped using primer pairs designed by the Medicago genome sequencing project (http://w.w.w.medicago.org/genome/). Genotypes for the A17 × A20 F2 population (n = 69) – used as a reference for the allocation of BACs in the sequencing project – were obtained from the file UMN_marker_genotypes_ 2005-03-14.xls available at http://www.medicago.org/genome/downloads.php#markerseqs. To avoid false positive linkage a large number of markers were excluded from the A17 × A20 analysis because of high levels of segregation distortion (Choi et al., 2004a). Linkage analyses were performed using the multipoint version 1.2 software (http://www.multiqtl.com; Institute of Evolution, Haifa University, Haifa, Israel). Multilocus ordering was conducted using the sum of recombination rates along consecutive pairs of adjacent markers and a Jackknife resampling technique (5000 iterations) for verification of the order obtained (Mester et al., 2003). Stable map segments are revealed with neighbour probabilities higher than a threshold (e.g. P = 0.9).
Hybrids derived from A17 are semisterile
Pollen viability is a good indicator of chromosomal aberrations (Griffiths et al., 1996). Heterozygotes for chromosomal rearrangements such as deletions, inversions and translocations show significant reductions in fertility because of the presence of nonfunctional gametes. Many crosses between M. truncatula accessions, representative of the large genetic variation within the germplasm collection (Skinner et al., 1999; Ellwood et al., 2006), were generated for genetic studies. Pollen grains from a sample of the hybrids and parental lines were evaluated and the proportion of nonaborted pollen grains was estimated (Alexander, 1969). The parental lines had 94–100% nonaborted pollen grains; A17 had on average 98.2% pollen viability (SE = 0.3, n = 10 plants; Fig. 1c). All F1 individuals showed approx. 100% pollen viability with the exception of those derived from crosses involving A17, which showed 50% nonaborted pollen or less (Fig. 1a,b). A 50% reduction in viable gametes (semisterility) is characteristic of heterozygotes for a reciprocal translocation (Griffiths et al., 1996). Further reductions in pollen viability, such as those observed in the A17 × SA27192 and A17 × A20 hybrids, may be a result of additional differential chromosomal arrangements between the parents.
Genetic maps derived from A17 reveal new linkage relationships
Further evidence for the segregation of a reciprocal translocation is the formation of new linkage relationships (Griffiths et al., 1996). Linkage analyses were performed using F2 populations derived from A17 and from the SA27063 × SA3054 hybrid. As expected, the SA27063 × SA3054 map included eight linkage groups corresponding to the haploid chromosome number. By contrast, seven linkage groups emerged from the analysis of the SA27063 × A17 and A17 × Borung populations. Re-evaluation of the genotype data available for the A17 × A20 population, used as a reference in the Medicago genome sequencing project, revealed that markers originally assigned to either LG4 or LG8 were clustered in a single linkage group (Fig. 2e). The matrix of neighbour frequencies provided statistical evidence for the different aspects of such clustering (Fig. 2). First, the relatively low neighbourhood frequencies observed in the SA27063 × A17 and A17 × Borung matrices (Fig. 2c,d) exposed markers with several putative neighbours making the allocation of markers in a certain order statistically unjustifiable. Alternatively, such as in the case of the A17 × A20 map (Fig. 2e), high recombination frequencies outlines the most probable order of these tightly linked markers. By contrast, data from the SA27063 × SA3054 population provided unequivocal marker order and resulted into two distinctive linkage groups (Fig. 2a,b).
A17 exhibits an aberrant chromosomal arrangement
The pollen viability tests and the establishment of new linkage relationships together provide strong evidence that A17 bears a reciprocal translocation that distinguishes it from other M. truncatula accessions. The reciprocal translocation involves the long arms of chromosomes 4 and 8 (Fig. 2). Detailed cytological studies of hybrids derived from A17 are needed to determine the exact breakpoints of the translocation event. Cytogenetic studies are well developed in M. truncatula. Thus far, they been conducted in two pure inbred lines: cv. Jemalong, lines A17 and J5, and acc. R108-1 (Cerbah et al., 1999; Kulikova et al., 2001; Falistocco & Falcinelli, 2003; Choi et al., 2004a; Kulikova et al., 2004). Alternatively, denser genetic maps, cytological studies and/or sequence information from other accessions might provide enough evidence for the exact position of the translocation.
Interestingly, the genetic map derived from the cross between Jemalong line J6 and the Algerian accession DZA315.16 (Thoquet et al., 2002) shows no evidence of segregation of a reciprocal translocation. To date, no phenotypic or genotypic differences have been observed between the Jemalong lines A17, J6 and J5 which were therefore considered to be genetically identical (Thoquet et al., 2002). However, hybrids derived from the cross between A17 and DZA315.16 consistently show about 50% pollen viability (Fig. 2). Whether A17 and J6 differ in their chromosomal configuration needs to be proven; however, if this is so, the rearrangement occurred relatively late in the diversification of the species.
Since all the F1s derived from crosses not involving A17 show approximately 100% nonabortive pollen (Fig. 1) it seems that these accessions share a common chromosomal arrangement. Linkage maps derived from such crosses, as the one presented herein (SA27063 × SA3054; Fig. 2) are valuable tools to provide insights into the M. truncatula intraspecific genomic organization and evolution.
The distinctive chromosomal arrangement in A17 has significant connotations for genetic studies that involve outcrosses, such as in map-based-cloning. Semisterile F1s and further selection of F2 individuals through low germination rates and albinism bias the selection of certain gene combinations leading to skewed genetic analyses.
The implications of the aberrant chromosomal configuration on the sequencing project are noteworthy since at the present the genome sequencing has not been completed and gaps between contigs are evident. An anchored clone-by-clone strategy was chosen to efficiently sequence the M. truncatula gene-space (Young et al., 2005); the clones were initially ordered according to the position of selected markers in a genetic map derived from a A17 × A20 F2 population. The translocation accounts for the consistently poor scoring and ambiguous mapping of some markers (Zhu et al., 2002; Cannon et al., 2005) and subsequently the misallocation of the respective clones. For example, the gene-based marker SQEX derived from BAC clone AC124214 has been mapped to LG8 (Choi et al., 2004a) whereas the BAC clone is physically mapped to chromosome 4 (http://www.medicago.org/genome/cvit_bacs.php). These are small obstacles in the genome sequencing effort. However, special caution should be taken in using the genome configuration of M. truncatula A17 (i.e. physical map) for map-based cloning and comparative studies between accessions or between species.
Further, one would expect that chromosomal rearrangements, such as the reciprocal translocation, would disguise the detection of synteny among species. However, if the comparison is done by means of linkage maps (Choi et al., 2004b) the rearrangements will have a significant impact in the inferred synteny relationships only if the accessions from which the map derives share that peculiar chromosomal rearrangement. Otherwise, such as in the case of the M. truncatula primary linkage map, the map does not necessarily show the physical chromosomal composition of A17 or A20, but rather a combination of the two genomes. Therefore, the synteny relationships between M. truncatula and the other legumes are largely conserved with the exception of ambiguously allocated markers, resulting in biased observations (e.g. SQEX in MtLG8 and in PsLGVII; Choi et al., 2004b).
We thank Dr John Klingler for helpful comments and critical revision of the manuscript. This research was funded by the Grains Research and Development Corporation of Australia (GRDC), and the Department of Education, Science and Training (DEST). The research was conducted at the State Agricultural Biotechnology Centre (SABC), Murdoch University, Western Australia.