• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of geneticsGeneticsCurrent IssueInformation for AuthorsEditorial BoardSubscribeSubmit a Manuscript
Genetics. Feb 2005; 169(2): 967–979.
PMCID: PMC1449096

Detection of Chromosomal Rearrangements Derived From Homeologous Recombination in Four Mapping Populations of Brassica napus L.


Genetic maps of Brassica napus were constructed from four segregating populations of doubled haploid lines. Each mapping population had the same male parent and used the same set of RFLP probes, facilitating the construction of a consensus map. Chromosomal rearrangements were identified in each population by molecular marker analysis and were classified as de novo homeologous nonreciprocal transpositions (HNRTs), preexisting HNRTs, and homeologous reciprocal transpositions (HRTs). Ninety-nine de novo HNRTs were identified by the presence of a few lines having duplication of a chromosomal region and loss of the corresponding homeologous region. These de novo HNRTs were more prevalent in one population that had a resynthesized B. napus as a parent. Preexisting HNRTs were identified by fragment duplication or fragment loss in many DH lines due to the segregation of HNRTs preexisting in one of the parents. Nine preexisting HNRTs were identified in the three populations involving natural B. napus parents, which likely originated from previous homeologous exchanges. The male parent had a previously described HRT between N7 and N16, which segregated in each population. These data suggest that chromosomal rearrangements caused by homeologous recombination are widespread in B. napus. The effects of these rearrangements on allelic and phenotypic diversity are discussed.

CHROMOSOMAL rearrangements have played an important role in our understanding of genetics, evolution, and speciation of many organisms (see Stebbins 1971, Rieseberg 2001, and Levin 2002 for examples in plants; Navarro and Barton 2003 and Bailey et al. 2004 for examples in humans). Chromosomal rearrangements derived from nonhomologous chromosome exchanges were first identified in corn (Zea mays) using traditional cytogenetic techniques (McClintock 1929) and have been commonly termed translocations or interchanges (Burnham 1962). In diploid organisms, translocation heterozygotes produce duplication-deletion meiotic products resulting in semisterility (Griffiths et al. 2000). In allopolyploids, gene redundancy can buffer the phenotypic effects of duplications and deficiencies caused by segregating translocations, especially for translocations that result from homeologous chromosomal exchanges. Stebbins (1947) referred to the chromosomal pairings that could lead to such exchanges as heterogenetic associations. Substitutions of entire homeologous chromosomes from such associations was recognized as a potential cause of occasional off-types in cultivated oats (Avena sativa) and in wheat (Triticum aestivum; reviewed by Huskins 1946). Osborn et al. (2003) recently described an exchange between segments of homeologous chromosomes in some genotypes of the allotetraploid Brassica napus and referred to this rearrangement as an interstitial homeologous reciprocal transposition.

B. napus is an amphidiploid species composed of homeologous A and C genomes, which are thought to have derived from the recent progenitors of extant B. rapa and B. oleracea, respectively (U 1935). In a resynthesized B. napus amphihaploid, meiotic chromosome pairing was observed, suggesting that the A and C homeologs have retained sufficient homology to form syndetic structures in the absence of true homologous chromosome pairs (Attia and Röbbelen 1986a). After chromosome doubling the amphihaploid, meiotic chromosome pairing between homologs was most prevalent, but some multivalent structures were observed (Attia and Röbbelen 1986b). Multivalent structures were also observed in natural allotetraploid B. napus, although at lower frequencies than observed in resynthesized B. napus (Newell et al. 1984; Attia and Röbbelen 1986b). These cytological observations suggest that exchanges could occur between homeologous chromosomes of B. napus.

Evidence for homeologous exchanges also has been obtained by analyzing segregating populations of B. napus with molecular markers. Sharpe et al. (1995) and Parkin et al. (1995) first documented the occurrence of homeologous nonreciprocal transpositions (HNRTs, termed nonreciprocal translocations in these articles) in B. napus. They used RFLP markers to generate genetic maps of three segregating populations of doubled haploid (DH) lines derived from reciprocal crosses between a winter and a spring B. napus cultivar (Sharpe et al. 1995) and from a resynthesized B. napus crossed to the same winter parent (Parkin et al. 1995). Unusual RFLP patterns of a few DH lines suggested that homeologous exchanges occurred during meiosis of F1 plants at low frequencies and generated de novo HNRTs. Another type of rearrangement, a homeologous reciprocal transposition (HRT), was also characterized in these three mapping populations and in three additional B. napus mapping populations (Osborn et al. 2003). All six mapping populations were derived from crosses between spring and winter parents. The HRT was identified as an exchange between homeologous regions of N7 and N16 that was inherited from the spring parent and segregated in the progeny. Molecular marker evidence for HRTs was further supported by pachytene spreads of meiotic cells from anthers from a winter × spring F1 hybrid, which illustrated possible exchange points between chromosomes.

In this study, we have generated four genetic maps of B. napus from four segregating populations of DH lines using RFLP markers. The four populations shared a common male parent, facilitating the assembly of a B. napus consensus map. The female parent of one population was a B. napus line resynthesized from B. rapa and B. oleracea, which allowed unambiguous identification of RFLP alleles from the A and C subgenomes of B. napus. Rearrangements among homeologous chromosomes (de novo HNRTs, preexisting HNRTs, and an HRT) were identified and classified on the basis of irregular segregation patterns of RFLP alleles and by matching RFLP alleles to the diploid progenitor species. The role of these rearrangements in polyploid genome evolution and their effects on allelic and phenotypic diversity of B. napus are discussed.


Population development:

Four DH populations were developed by crossing the RV289, TO1147, MF216, and RV128 lines with a male-fertility restorer line, P1804, produced by Bayer CropScience, Saskatoon, Saskatchewan, Canada. RV289 is a single plant selection from Hua-dbl2, a Chinese open-pollinated winter cultivar kindly provided by Jinling Meng, Huazhong Agricultural University, Wuhan, China. TO1147 is a resynthesized B. napus line created by crossing a B. rapa plant (as the female), grown from a yellow seed of cultivar Reward (Scarth et al. 1992), to an inbred B. oleracea plant, TO1000, a S5 inbred line from the rapid-cycling B. oleracea stock CrGC3-3 developed by the Crucifer Genetics Cooperative (Madison, WI). An F1 embryo from this interspecific cross, TO1147, was rescued and chromosome doubled by immersing its roots in a 0.34% colchicine solution for 1.5 hr (Song et al. 1993). MF216 is a noncanola quality DH line derived from a cross between the French winter cultivar Major and the Canadian spring canola cultivar Stellar (Ferreira et al. 1994) and has ~60% of its genome derived from Major. RV128 was developed by backcrossing alleles for spring growth habit from Westar (a Canadian spring canola cultivar) into Samourai (a French winter canola cultivar) using RFLP marker loci to select for Westar alleles at the major vernalization-requiring flowering-time QTL VFN1 (Osborn et al. 1997) and for Samourai alleles at ~85% of the remaining marker loci scattered throughout the genome (R. D. Vogelzang and T. C. Osborn, unpublished data). Each of the four female parents, RV289, TO1147, MF216, and RV128, were crossed with P1804. Two different F1 plants were used as the sources for microspores for the (RV289 × P1804) population and a single F1 plant was used for the other three populations.

Haploid plants of each population were created from their respective F1 through microspore culture (Chuong and Beversdorf 1985). They were selected to contain the male-fertility restorer gene Rf, required for the Seedlink hybrid seed production system of Bayer CropScience, by selecting for resistance to glufosinate herbicide conferred by a linked resistance gene included in the Rf construct (Mariani et al. 1992). Approximately 1000 haploid plants were treated with colchicine and male-fertile plants were self-pollinated to establish DH lines. From 162 to 170 DH lines that produced sufficient seed (>100) for future testing were genotyped for each population (Table 1). The F1-derived DH populations are referred to as HUA (RV289 × P1804), MF (MF216 × P1804), and RV (RV128 × P1804), SYN (TO1147 × P1804).

Characteristics of four genetic linkage maps and their consensus map constructed using DH lines ofB. napus, genotyped with RFLP and SSR markers

Genetic markers:

The majority of the RFLP probes included in this study had been used in previous mapping studies (Ferreira et al. 1994; Teutonico and Osborn 1994; Parkin et al. 1995; Sharpe et al. 1995; Thormann et al. 1996; Kole et al. 1997; Butruille et al. 1999) and came from three libraries: a PstI genomic DNA library and a cDNA library from B. napus cv Westar, originally designated as “WG” and “EC” probes, respectively (Thormann et al. 1994), and an EcoRI genomic DNA library from B. rapa cv Tobin (“TG” probes) that was constructed and screened as described for the Westar genomic library (Ferreira et al. 1994). Clones from the PstI and EcoRI genomic DNA libraries were labeled “pW” probes and those from the cDNA library were labeled “pX” probes, according to the nomenclature of Parkin et al. (1995) and Sharpe et al. (1995). This nomenclature was used due to its greater simplicity and ease of cross-referencing with other studies that used this nomenclature. Additional information about these probes, including DNA sequences, is available at http://osbornlab.agronomy.wisc.edu/research.html. Four cloned genes that were homologs of the Arabidopsis thaliana FLOWERING LOCUS C (FLC) gene from B. rapa (Schranz et al. 2002), and a clone of the A. thaliana FLOWERING LOCUS T gene (Kardailsky et al. 1999) were also used as probes.

The probes (120 pWs, 53 pXs, and the cloned genes) were screened by hybridization to small blots containing genomic DNA from the B. napus parental lines, B. oleracea (TO1000), and three B. rapa cultivars (Reward, Per, and R500) separately digested with HindIII and EcoRI, the enzymes used in previous studies (Ferreira et al. 1994; Thormann et al. 1996; Butruille et al. 1999). Polymorphisms observed between P1804 and the other parents were genetically mapped by hybridizing probes to mapping blots containing appropriately digested samples of the DH lines. Loci were named using the probe name with a two-letter suffix. The first letter distinguished multiple loci detected by a probe. Fragments detected in the P1804 parent were labeled alphabetically from the highest to lowest molecular weight. The second letter indicated the enzyme used to detect the RFLP locus (E, EcoRI; H, HindIII).

Map construction:

JoinMap 3.0 was used to create the genetic linkage maps (van Ooijen and Voorrips 2001). Linked loci were grouped on the basis of pairwise LOD values of 5 or 6. Locus order within the LOD groupings was determined through an optimized algorithm using three rounds of marker placement with the Kosambi mapping function (Kosambi 1944; Stam 1993). Starting with the most informative pair in the first round, markers were added in each possible position one by one. The best order was selected by goodness of fit. Markers with negative recombination fraction or large contributions to lack of fit (parameter value of 5.00) were held for the second round. Each marker not placed in the first round was tried again using the same fitting criteria, but with a more complete linkage group. Finally, the remaining markers were placed into the most likely position by minimizing the lack-of-fit parameter even though it may have exceeded the initial threshold. After positioning new loci, the ripple command with a three-locus window was used to reestimate locus orders and distances. After map construction, the data files were searched for putative double recombinants, which were verified or corrected by reexamining the RFLP autorads. Generally, after a few iterations between map construction and data verification, the third round of marker placement was not needed when reconstructing the linkage groups.

Consensus map construction:

A consensus map was constructed using individual maps from all four DH populations. The four maps were aligned by common loci using the same set of RFLP probes and previous published maps (Parkin et al. 1995; Sharpe et al. 1995; Butruille et al. 1999). Since each population had a common parent (P1804), individual linkage groups could be aligned using common alleles from P1804. In cases where different enzymes were used, common loci were inferred on the basis of the position of common flanking loci and the enzyme designation was omitted from the locus name. The consensus map was created using JoinMap 3.0 (van Ooijen and Voorrips 2001), which utilized the mean recombination frequencies and combined LOD scores to position common loci mapped in more than one population. JoinMap also was used to evaluate significant differences in recombination between common markers in each of the individual maps. MapChart (Voorrips 2002), PowerPoint, and Excel were used to draw the maps and graphically illustrate the homologous and homeologous relationships between linkage groups.


Genetic map characteristics:

Individual genetic maps were constructed for four segregating populations of F1-derived DH lines from crosses involving three natural and one resynthesized B. napus crossed with a single natural B. napus line (Table 1). Nineteen linkage groups that probably correspond to the 19 chromosomes of B. napus were identified in each mapping population (Figure 1). Each linkage group could be assigned to either the B. rapa (A) or the B. oleracea (C) genome on the basis of the identification of alleles as coming from either the A- or the C-genome progenitors of the SYN population. Linkage groups were designated as in previous studies (Parkin et al. 1995; Sharpe et al. 1995; Butruille et al. 1999) with N1–N10 corresponding to the A genome and N11–N19 corresponding to the C genome. The four individual genetic maps showed a high level of collinearity. Some inconsistencies found in locus order among the individual genetic maps could be due to chromosomal rearrangements (e.g., N11 and N16; see below). The largest total recombinational map was detected for the SYN population, perhaps because it included more polymorphic loci and thus had more extensive genome coverage.

Figure 1.
A consensus genetic linkage map of molecular markers compiled from individual maps constructed for four segregating populations of B. napus doubled haploid lines. Marker locus names and map positions (in centimorgans) are in the first two columns of each ...

A considerable proportion of the polymorphic loci detected within the individual maps (26% to 41%) had allelic segregation that deviated significantly (P < 0.01) from the 1:1 ratio expected for a population of DH lines derived from an F1 (Table 1). All four DH populations were selected for glufosinate resistance via the microspore culture media to utilize the Seedlink hybrid production system (Goldberg et al. 1993) in subsequent experiments. The bottom of linkage group N7 was completely skewed toward the P1804 allele, indicating this as the most likely position for the glufosinate resistance transgene. The top of N3 also was skewed significantly toward the P1804 allele in all populations, perhaps because of inadvertent selection of early flowering alleles at BrFLC3 and BrFLC5 on N3. BrFLC3 and BrFLC5 are homologs of the A. thaliana flowering time gene FLC and allelic variation at BrFLC5 has been associated with flowering time in B. rapa (Schranz et al. 2002). This and other regions that had significant deviations from a 1:1 ratio also may have segregated for various genetic factors that affected the generation or selection of the doubled haploid lines used in the mapping populations (see Xu et al. 1997). In some populations skewed segregation occurred in regions that segregated for chromosomal rearrangements. These rearrangements affected the amount of intergenome heterozygosity and may have had subtle effects on the amount of seed produced (see discussion).

A consensus map was constructed from the four segregating populations of DH lines (Figure 1). Each individual map contributed complementary sets of polymorphic loci providing thorough coverage of the genome in the consensus map (Figure 1 and Table 1). On average, each DH map contributed six loci per linkage group to the consensus map. Most of the RFLP probes (88%) detected more than one segregating locus, with an average of 1.65–2.23 loci/probe detected in the individual maps and 3.02 loci/probe in the consensus map (Table 1). These multiple loci often had collinear linkage arrangements between linkage groups from the A and C genomes (Figure 2), reflecting the homeologous relationships reported previously (Parkin et al. 1995; see below).

Figure 2.
Collinearity among homeologous linkage groups in the B. napus consensus map. Linkage groups are numbered N1–N10 (A genome) and N11–N19 (C genome). Lines connect RFLP loci detected with the same probe and indicate intergenomic homeologous ...

Chromosomal rearrangements:

Several factors facilitated the detection of chromosomal rearrangements in this study. First, the DH lines used for mapping were completely homozygous, each being derived from single gametes, and their use prevented the potential misclassification of some marker genotypes as heterozygous in lines containing rearrangements. Second, by comparing four mapping populations sharing one parent we could more easily identify unusual patterns of segregation that indicated the presence of chromosomal rearrangements. Third, inclusion of a resynthesized B. napus line as one of the mapping parents allowed accurate assignment of loci to either the A or the C genome. Fourth, use of RFLP probes that hybridized to multiple, codominant markers allowed simultaneous detection of alleles at homeologous loci.

Three types of chromosomal rearrangements were detected in the four mapping populations. The first type was de novo HNRTs, which apparently were derived from occasional homeologous exchanges during meioses in the F1 plants. The second type was preexisting HNRTs within one of the mapping parents, which were inherited by the F1 plant and segregated among the DH progenies. The third type of rearrangement was a preexisting HRT between homeologous segments of linkage groups N7 and N16 in the male parent (P1804) that segregated in the four DH populations. De novo HRTs may have occurred in some DH lines that inherited both transposed chromosomal segments from a homeologous exchange, but these could not be distinguished from DH lines inheriting unrearranged homeologous chromosomes.

De novo homeologous nonreciprocal transpositions:

These HNRTs were the simplest type of chromosomal rearrangement to recognize. Individual DH lines with de novo HNRTs were missing both alleles for a given locus and had allele duplication at their homeologous loci, containing either four doses of one allele or two doses of each of two alleles (Figure 3). These distinct allelic banding patterns were most readily recognized when a single probe hybridized to, and was polymorphic for, both homeologous loci. During construction of the individual maps, it became apparent that these losses and/or duplications within individual DH lines were among the linked loci of two homeologous genomic regions and that pairing and exchange of homeologous chromosomes was the most likely explanation for these de novo HNRTs. The population involving the resynthesized B. napus parent had the highest frequency of de novo HNRTs. As a percentage of total chromosomal exchanges detected in each map, homeologous exchanges represented 1.60, 1.09, 0.49, and 0.43% in the SYN, HUA, MF, and RV populations, respectively. The highest frequencies of de novo HNRTs involved exchanges between N1 and N11, perhaps due to a greater degree of homology between these homeologous linkage groups. There was also a higher frequency of C-genome duplications than of A-genome duplications among the de novo HNRTs of the SYN population, especially for linkage groups N11 vs. N1, N18 vs. N8, and N19 vs. bottom of N9 (Figure 2).

Figure 3.
Chromatid model of homeologous exchange between N1 and N11 creating de novo HNRTs, and segregation of DH lines for RFLP loci with the exchanged region. The smaller chromosome represents N1 (A genome) and the larger chromosome represents N11 (C genome). ...

Preexisting homeologous nonreciprocal transpositions:

Molecular marker data suggested that each of the natural B. napus parents had preexisting HNRTs that segregated among the DH lines: MF216 had two HNRTs, one composed of N11 containing a transposed homeologous section of N1 [N11.N1(T)] and one composed of N13.N3(T); RV289 had two composed of N12.N2(T) and N13.N3(T); RV128 had one composed of N13.N3(T); and P1804 had four composed of N11.N1(T), N12.N2(T), N15.N5(T), and N10.N19(T). These rearrangements were more difficult to identify than de novo HNRT. The clearest evidence for preexisting HNRTs was the presence of RFLP alleles with identical fragment sizes at homeologous loci within the parental genotype (i.e., four doses of the same allele). In some cases, the identical-size fragments matched fragments from either a B. rapa or a B. oleracea genotype included on the screening blot, suggesting that they originated from the same diploid genome; however, one allele mapped to an A-genome homolog, and the other mapped to a C-genome homolog. For example, pW126 and pW247 each detected two loci, one on N5 and one on N15, which had alleles with identical fragment sizes in P1804, and these matched the fragment sizes of alleles in B. rapa cv. Reward (A genome); however, one P1804 allele mapped to N5 (A genome), and the other mapped to N15 (C genome; Figure 4). Thus, P1804 appears to have a preexisting HNRT involving a segment of N5 transposed onto N15 [N15.N5(T)]. Preexisting HNRTs were difficult to delimit because homeologous loci having identical-size alleles were interspersed with those having polymorphism at only one locus or those having different-size alleles at both homologous loci. An example of the later case are loci detected by pW125 on the transposed N5 segment of N15 and the normal N5 chromosome of P1804 (Figure 4). This and other such loci may delimit the transposed segment, having originated by a subsequent round of homeologous chromosome exchange to reintroduce a smaller homologous segment. However, in the case of pW125 the fragment sizes of both loci in P1804 matched the fragment sizes of two of the B. rapa genotypes included on our screening blots (data not shown). Thus, the HNRT on N15 may include pW125; however, its allele could be different from the N5 allele because of outcrossing subsequent to the transposition.

Figure 4.
Chromatid model of a region containing a preexisting HNRT between N5 and N15 and segregation of DH lines for RFLP loci in this region. The smaller chromosome represents N5 (A genome) and the larger chromosome represents N15 (C genome). Chromosomes of ...

Segregating loci on N1 and N11.N1(T) were particularly difficult to characterize in the MF and HUA maps. For both parents, some probes hybridized to alleles with identical fragment sizes on N1 and N11 (pW239, pW172, and pW125 in P1804; pW216 in MF216), and the fragment sizes matched those of B. rapa genotypes, further suggesting that N11 contained a segment of N1 in both parents. The preexisting HNRT segments of MF216 and P1804 may have had different origins since loci detected by pW239, pW172, and pW125 on N11.N1(T) were polymorphic between MF216 and P1804. Differences in the sizes or arrangement of these two transposed segments may have affected recombination in the region, because N11 of the MF map consisted of a cluster of tightly linked loci that was not detected in the other three genetic maps. This linkage group probably comprised the transposed N1 segment of N11 and a small portion of N11 that was not polymorphic in the other three maps. Thus, the locus order of N11 in the MF genetic map was distinct enough to exclude it during assembly of the consensus map (Figure 1). In the HUA map, we could not account for allelic relationships of several markers on N11. When P1804 alleles of these loci were each scored as a single “dominant locus, they mapped together at the top of N11, although their linkage distances could not be determined due to segregation of identical-size alleles that were probably on N1. Since these loci also were found in the MF map, they were included in the consensus map to document their presence on linkage group N11, but linkage distances are not reported (Figure 1).

Many of the de novo HNRTs occurred in genomic regions for which there was evidence of preexisting HNRTs. This may have been due to pairing and exchange of homologous segments on homeologous chromosomes. The largest number of de novo HNRTs were detected on N2 in the HUA population (Figure 2) where there was evidence of preexisting HNRTs in both parents, RV289 and P1804 [N12.N2(T)]. Only one of the F1 plants contained the N12.N2(T) from RV289, which apparently was heterozygous. All of these de novo HNRT lines originated from this F1, as determined by three probes that detected loci in the transposed region. Thus, the presence of N2 segments on both N12 chromosomes in this F1 appears to have caused frequent pairing and exchanges with the N2 chromosomes, resulting in de novo HNRTs that actually were between homologous N2 segments.

Molecular markers also provided evidence that multivalents had formed during the meioses that resulted in de novo HNRTs. Linkage groups of DH lines missing both alleles due to de novo HNRTs were inspected for additional recombination events within the same linkage group. Nearly half of the homeologous syndeses resulting in de novo HNRTs had additional exchanges with homologous chromosomes. This indicated that the chromosome with the preexisting HNRT had associated with both homologous and homeologous chromosomes in a trivalent or quadrivalent configuration during meiosis. Evidence for multivalent formation was found more frequently in the SYN F1 than in the F1's between the natural B. napus parents.

Homeologous reciprocal transposition:

Fragment sizes and marker segregation patterns provided evidence that a HRT between N7 and N16 was segregating in each of the four populations. Fragment sizes of the resynthesized parent were most useful for identifying the transposed regions because the allelic origins of TO1147 fragments were known (i.e., A or C genome). In the resynthesized parent, loci with a B. rapa (Reward; A genome) allele mapped to the N7 linkage group and loci with a B. oleracea (TO1000; C genome) allele mapped to the N16 linkage group. In P1804, many of the loci (e.g., pW150cH and pW228aH) on N16 had alleles with fragment sizes that matched those in B. rapa genotypes included on the screening blots, suggesting that the common parent, P1804, contained a B. rapa segment on N16. However, none of the N7 loci had P1804 alleles with fragment sizes matching those in TO1000. Several probes detecting loci on N7 and N16 (pW150, pW228, pW130, pW208, and pX156) hybridized to identical-size fragments in TO1147 and P1804 (i.e., monomorphic), and yet they segregated in the SYN population. For several of these loci, their genome origin was known (A genome) but they were allelic with fragments on N16 (Figure 1). In all four populations, loci on N16 showed balanced segregation for both alleles, while the N7 loci were nearly fixed for the P1804 allele, probably due to selection for the transgene conferring herbicide resistance. This fortunate coincidence greatly facilitated correct linkage group assignment of N7 and N16 loci within the four DH maps. Segregating “monomorphic” fragments were observed previously using some of the same probes in three other segregating populations of B. napus and were attributed to an interstitial homeologous reciprocal transposition between N7 and N16 in the spring-type parent of the crosses (Osborn et al. 2003).


Genetic linkage maps:

A consensus genetic linkage map of B. napus was developed on the basis of data from four segregating populations of DH lines using publicly available RFLP probes. The 19 linkage groups found in each of the maps probably correspond to the 19 chromosomes of B. napus, but only the B. oleracea linkage groups have been associated cytogenetically with chromosomes (Howell et al. 2002). The locus order of the consensus map was generally in agreement with previous maps of B. napus, which used some of these same probes for RFLP detection (Ferreira et al. 1994; Parkin et al. 1995; Sharpe et al. 1995; Uzunova et al. 1995; Howell et al. 1996; Fray et al. 1997; Kelly et al. 1997; Butruille et al. 1999).

Use of a resynthesized B. napus as a parent in one of the populations allowed us to unambiguously assign each of the linkage groups to the A or C genome in each of the four genetic maps. Our results, in agreement with Parkin et al. (1995), suggest that the process of chromosome recombination in natural B. napus has remained relatively unchanged compared to the diploids B. rapa and B. oleracea, even though they have evolved separately since the formation of the allopolyploid. Of the 167 probes used to construct the consensus map, 147 detected more than one locus, revealing a high level of intergenomic (between A and C) duplication and intragenomic (within A and C) duplication, as has been observed in other Brassica genome studies (Kianian and Quiros 1992; Kowalski et al. 1994; Lagercrantz and Lydiate 1996; Cheung et al. 1997; Lan et al. 2000; Babula et al. 2003; Lukens et al. 2003). Large stretches of collinear marker loci, some including entire linkage groups, were found between the A- and C-genome components of B. napus. These stretches were similar to those reported previously (Lydiate et al. 1995; Parkin et al. 1995; Sharpe et al. 1995; Parkin et al. 2003) and probably reflect homeologous relationships between the A and C genomes. Similar to Kelly et al. (1997) but unlike Sharpe et al. (1995) and Howell et al. (1996), we did not detect significant differences in the number of polymorphic loci between the A and C genomes in any of the DH populations.

Chromosomal rearrangements:

The majority of DH lines appeared to derive from meiotic events in which recombination occurred only between homologous chromosomes. However, some DH lines appeared to derive from meiotic events that included some exchanges between homeologous chromosomes, resulting in de novo HNRTs. These de novo HNRTs were identified as duplications and losses of the respective homeologous RFLP loci. While they occurred throughout the genome, chromosome pairs with extensive regions of homeology had the highest frequencies of de novo HNRTs (N1–N11, N2–N12, N3–N13). A total of 99 homeologous transpositions were detected in the four DH populations, representing ~0.90% of the total recombination events. Here, and in other studies (Parkin et al. 1995; Sharpe et al. 1995; Uzunova et al. 1995), N1–N11 had the most homeologous exchange events, suggesting that these are the least divergent homeologs in the A and C genomes.

The SYN population had a higher frequency of de novo HNRTs than the populations derived from two natural B. napus parents. The frequencies observed for the MF and RV populations (0.49% and 0.43%, respectively) are similar to those detected in a natural × natural B. napus DH population examined by Sharpe et al. (1995). These results suggest that the natural B. napus subgenomes have become partially stabilized or “diploidized,” resulting in a lower rate of homeologous exchange. It is possible that homeologous recombination occurred in the natural B. napus F1's at a frequency equal to the F1 involving the SYN parent, but some of the individuals derived from these gametes did not survive or were selected against due to the uncovering of deleterious recessives masked by intergenomic heterozygosity in natural B. napus.

The observation of de novo HNRTs in DH lines of the mapping populations used in this study suggests that HNRTs also may exist in B. napus cultivars. Indeed, we found evidence of nine preexisting HNRTs in the four natural B. napus mapping parents used in our study, which were derived from oilseed cultivars. HNRTs were identified previously between N1 and N11 (Sharpe et al. 1995; Butruille et al. 1999) and between N2 and N12 (Parkin et al. 1995) in the parents used for other mapping studies. We also observed preexisting HNRTs between N1 and N11 (in MF216 and P1804) and between N2 and N12 (in P1804 and RV289), and we found evidence for other preexisting HNRTs between homeologous linkage groups N3 and N13 (MF216, RV128 and RV289), N5 and N15 (P1804), and N10 and N19 (P1804). The actual number of preexisting HNRTs may have been underestimated because we had limited numbers of diploid genotypes on our screening blots and some genomic regions were not polymorphic in the natural × natural populations.

Preexisting HNRTs could have led to incorrect assembly of linkage groups. For example, the two loci detected by each of pW126 and pW247 on N5(P) and N15(P) (Figure 4) could have been scored as single dominant loci having 3:1 segregation ratios, and this may have resulted in pseudolinkage of loci on N5 and N15 during map construction. Pseudolinkage was previously described between loci on N1 and N11 due to the presence of an HNRT (Sharpe et al. 1995). In this study, pseudolinkages were detected during assembly of three of the DH genetic maps (between loci on N1 and N11 for the MF and SYN populations and between loci on N2 and N12 for the HUA population). These loci, which initially created fused pseudolinkage groups, either were corrected to account for their true allelic relationships or were removed from the data set.

Preexisting HNRTs also could contribute to genetic variation for chromosome pairing behavior among B. napus genotypes. Jenczewski et al. (2003) recently obtained evidence for genetic control of chromosome pairing in B. napus by examining segregation for pairing behavior in haploids produced from F1 hybrids between natural B. napus lines with high and low pairing of haploids. They observed that the parental pairing patterns were inherited in a Mendelian fashion, supporting the presence of a single major gene that affects homeologous chromosomal pairing in haploids. Differences in the amount of pairing between the parental and F1 haploids indicated that minor genetic factors could segregate in addition to this putative major gene. These minor factors may include preexisting HNRTs that differed between the parents and segregated among the haploids examined by Jenczewski et al. (2003).

Many of the de novo HNRTs that we detected were in regions containing possible preexisting HNRTs in one of the parents. This observation suggests that chromosome pairing may frequently initiate between preexisting HNRTs and their original homologs and then extend from the paired homologous segments into homeologous regions of the two chromatids where recombination could occur. For example, the homology at the top of N2 and N12.N2(T) in the HUA population may have mediated the formation of de novo HNRTs, which appeared to have extended beyond the length of the preexisting HNRT (Figure 2). Thus, HNRTs that initially involved only small segments may have been extended farther into the homeologous linkage groups via subsequent rounds of meiosis. By bringing other homeologous chromatids into closer proximity during pachytene, such pairings could also facilitate exchanges between completely homeologous chromatids, creating additional de novo NRTs.

Our observation of frequent homeologous exchanges in some regions of the B. napus genome raises the question of whether this species should be classified as a segmental allopolyploid. Stebbins (1950) defined a segmental allopolyploid as “a polyploid containing two pairs of genomes which possess in common a considerable number of homologous chromosomal segments or even whole chromosomes, but differ from each other in respect to a sufficiently large number of genes or chromosome segments, so that the different genomes produce sterility when present together at the diploid level.” Previous observations of sterility and of chromosome pairing in natural and resynthesized amphihaploids of B. napus (Attia and Röbbelen 1986a) are consistent with this definition, as are the observations reported here and elsewhere (Parkin et al. 1995; Sharpe et al. 1995) of frequent de novo HNRTs in some genomic regions. However, the frequency of de novo HNRTs is low enough to indicate that homologous chromosome pairing is greatly preferred over homeologous chromosome pairing, and thus, these homeologous associations create only slight deviations from disomic inheritance. We observed a higher frequency of de novo HNRTs when one parent had a preexisting HNRT, indicating that this condition leads to more frequent association of homeologs, and thus, greater deviation from disomic inheritance. In the HUA population, both parents had a preexisting HNRT on N12 (N12.N2), and we observed segregation approaching tetrasomic inheritance for N2 and N12.N2. This type of inheritance also was observed for these chromosomes in a different mapping population where one parent had preexisting HNRTs (Parkin et al. 1995). However, we did not observe genotypic classes having alleles only from the two parental N2 chromosomes or from the two parental N12.N2 chromosomes, which would be expected if these chromosomes exhibited complete tetrasomic inheritance. This may be due to the presence of different gene deficiencies on each homeolog that are severe enough to prevent survival when they are exposed in these genotypes.

Thus, our results indicate that some B. napus genotypes have become more homologous for certain chromosomal regions due to previous homeologous exchanges, and that progeny from matings between genotypes that have the same homogenized regions segregate in a fashion more typical of autopolyploids (polysomic inheritance) than of allopolyploids (disomic inheritance) for these regions. It is possible that B. napus will eventually become homogeneous for these rearrangements, leading to a well-defined segmental allopolyploid. This has not yet occurred on the basis of the polymorphism we observed for rearrangements among B. napus genotypes, and it may not occur if there is a selective advantage to having more intergenome heterozygosity (see below). It is also possible that some pairs of homeologous chromosomes may become more homogenized due to the cumulative effects of homeologous and homologous exchanges that alternate over many generations and create a mosaic of chromatin from both progenitor genomes. Evidence that this may occur comes from our observation that not all loci in HNRTs contained alleles with sizes identical to those in homeologous loci (e.g., pW125 on N5 and N15) and from our previous analysis of the N7 and N16 HRT, which exists as an interstitial homeologous exchange (Osborn et al. 2003). Because these rearrangements between homeologous chromosomes may be constantly evolving and difficult to delineate, we have chosen to use the more general term transpositions, which can include small interstitial segments of chromosomes, as opposed to translocations, which generally refer to larger distal segments of chromosomes.

Effects of chromosomal rearrangements on allelic and phenotypic diversity:

The homeologous chromosomal rearrangements detected in this study did not have large, obvious effects on phenotypes. Some HNRTs preexisted in cultivars that had been previously selected for seed yield and other desirable traits, and the DH lines with de novo HNRTs had no obvious defects. However, the genome homogenization caused by HNRTs could have subtle quantitative effects on phenotypes. This might occur through the effects of dosage-regulated genes that have additive effects between homeologs and convert to all one or the other homeolog in HNRT genotypes. Pires et al. (2004) reported the effects of such an exchange between N3 and N13 on variation for flowering time in resynthesized B. napus. Genome homogenization caused by HNRTs also affects the level of intergenome heterozygosity, and thus it could affect traits that show heterosis and respond to changes in heterozygosity, such as seed yield. Osborn et al. (2003) previously analyzed three populations derived from N7–N16 HRT heterozygotes and showed that the two nonparental genotypic classes (equivalent to HNRTs) had significantly lower seed yields than the parental genotypic classes in replicated field trials. Presumably, these nonparental genotypic classes had lost intergenomic heterozygosity due to chromosome assortment and contained four doses of a single allele (intragenomic homozygote) in the transposed segment. In our study, segregation distortion was observed in regions segregating for HNRTs, and in each case there was an excess of genotypes without the HNRT (i.e., those having more intergenome heterozygosity). The overall importance of genomic heterozygosity also was suggested by the bias toward intragenomic heterozygotes (i.e., two doses of two alleles) vs. intragenomic homozygotes found within de novo HNRTs throughout the genome, implying that intragenomic homozygotes may have had a lower fitness. In addition, a bias toward C-genome duplication in de novo HNRTs was also found in the SYN population. This may have been due to a higher frequency of deleterious recessive alleles derived from the open-pollinated B. rapa parent Reward than from the highly inbred B. oleracea parent TO1000.

Homeologous chromosomal transpositions could have a large impact on the genetic improvement of oilseed B. napus. If elite alleles (e.g., those controlling low erucic acid, low glucosinolates, or transgenes) are linked to or located in a transposed segment, selection for the elite allele would maintain the rearrangement and its corresponding effects (both favorable and unfavorable) in the selected lines. Such elite allele configurations may be widespread in germplasm used for cultivar development. For example, the restorer of fertility gene utilized in this study is located on an N7 chromosome containing a segment transposed from N16. Homeologous chromosomal rearrangements also could alter quantitative traits of interest through the effects of genome homogenization, as described above. Evidence for these effects could be obtained by mapping QTL in populations segregating for rearrangements. It remains to be seen to what extent the phenotypic variation in B. napus is due to effects of chromosomal rearrangements derived from homeologous recombination.


We thank researchers at Bayer CropScience for their assistance, particularly Hieronim Polewicz for providing seed of P1804 and Debbie Doell for DH line production. We also acknowledge Jinling Meng for providing seed of Hua-dbl2, Robert Vogelzang and Enrique León for technical assistance, Lynn Hummel for greenhouse assistance, and J. Chris Pires for his critical comments on the manuscript. J.A.U. was supported in part by a Pioneer Plant Breeding Fellowship (1999-2000), and P.A.Q. was supported by a scholarship from Consejo de Desarrollo Científico y Humanístico, Universidad Central de Venezuela, government of Venezuela. Additional funds were provided by the North Central Biotechnical Initiative and the U.S. Department of Agriculture National Research Initiative grant 98-353006-6286 to T.C.O.


  • Attia, T., and G. Röbbelen, 1986. a Cytogenetic relationship within cultivated Brassica analyzed in amphihaploids from the three diploid ancestors. Can. J. Genet. Cytol. 28: 323–329.
  • Attia, T., and G. Röbbelen, 1986. b Meiotic pairing in haploids and anphidiploids of spontaneous versus synthetic origin in rape, Brassica napus L. Can. J. Genet. Cytol. 28: 323–329.
  • Babula, D., M. Kaczmarek, A. Barakat, M. Delseny, C. F. Quiros et al., 2003. Chromosomal mapping of Brassica oleracea based on ESTs from Arabidopsis thaliana: complexity of the comparative map. Mol. Gen. Genomics 268: 656–665. [PubMed]
  • Bailey, J. A., R. Baertsch, W. J. Kent, D. Haussler and E. E. Eichler, 2004. Hotspots of mammalian chromosomal evolution. Genome Biol. 5: R23. [PMC free article] [PubMed]
  • Burnham, C. R., 1962 Discussions in Cytogenetics. Burgess Publishing, Minneapolis.
  • Butruille, D. V., R. P. Guries and T. C. Osborn, 1999. Linkage analysis of molecular markers and quantitative trait loci in populations of inbred backcross lines of Brassica napus L. Genetics 153: 949–964. [PMC free article] [PubMed]
  • Cheung, W. Y., G. Champagne, N. Hubert, and B. S. Landry, 1997. Comparison of the genetic maps of Brassica napus and Brassica oleracea. Theor. Appl. Genet. 94: 569–582.
  • Chuong, P. V., and W. D. Beversdorf, 1985. High frequency embryogenesis through isolated microspore culture in Brassica napus and Brassica carinata. Plant Sci. 39: 219–226.
  • Ferreira, M. E., P. H. Williams and T. C. Osborn, 1994. RFLP mapping of Brassica napus using double haploid lines. Theor. Appl. Genet. 89: 615–621. [PubMed]
  • Fray, M. J., P. Puangsomlee, J. Goodrich, G. Coupland, E. J. Evans et al., 1997. The genetics of stamenoid petal production in oilseed rape (Brassica napus) and equivalent variation in Arabidopsis thaliana. Theor. Appl. Genet. 94: 731–736.
  • Goldberg, R. B., T. P. Beals and P. M. Sanders, 1993. Anther development: basic principles and practical applications. Plant Cell 5: 1217–1229. [PMC free article] [PubMed]
  • Griffiths, A. J. F., J. H. Miller, D. T. Suzuki, R. C. Lewontin and W. M. Gelbart, 2000 An Introduction to Genetic Analysis. W. H. Freeman, New York.
  • Howell, E. C., G. C. Braker, G. H. Jones, M. J. Kearsey, G. J. King et al., 2002. Integration of the cytogenetic and genetic linkage maps of Brassica oleracea. Genetics 161: 1225–1234. [PMC free article] [PubMed]
  • Howell, P. M., D. F. Marshall and D. J. Lydiate, 1996. Towards developing intervarietal substitution lines in Brassica napus using marker-assisted selection. Genome 39: 348–358. [PubMed]
  • Huskins, C. L., 1946. Fatuoid, speltoid and related mutations of oats and wheat. Bot. Rev. 8: 457–514.
  • Jenczewski, E., F. Eber, A. Grimaud, S. Huet, M. O. Lucas et al., 2003. PrBn, a major gene controlling homeologous pairing in oilseed rape (Brassica napus) haploids. Genetics 164: 645–653. [PMC free article] [PubMed]
  • Kardailsky, I., B. Shukla, J. H. Ahn, N. Dagenais, S. K. Christensen et al., 1999. Activation tagging of the floral inducer FT. Science 286: 1962–1965. [PubMed]
  • Kelly, A. L., A. G. Sharpe, J. H. Nixon, E. J. Evans and D. J. Lydiate, 1997. Indistinguishable patterns of recombination resulting from male and female meiosis in Brassica napus (oilseed rape). Genome 40: 49–56. [PubMed]
  • Kianian, S. F., and C. F. Quiros, 1992. Generation of a Brassica oleracea composite map: linkage arrangements among various populations and evolutionary implications. Theor. Appl. Genet. 84: 544–554. [PubMed]
  • Kole, P., R. Vogelzang and T. C. Osborn, 1997. Genetic linkage map of a Brassica rapa recombinant inbred population. J. Hered. 88: 553–557.
  • Kosambi, D. D., 1944. The estimation of map distances from recombination values. Ann. Eugen. 12: 172–175.
  • Kowalski, S. P., T. H. Lan, K. A. Feldmann and A. H. Paterson, 1994. Comparative mapping of Arabidopsis thaliana and Brassica oleracea chromosomes reveals islands of conserved organization. Genetics 138: 499–510. [PMC free article] [PubMed]
  • Lagercrantz, U., and D. J. Lydiate, 1996. Comparative genome mapping in Brassica. Genetics 144: 1903–1910. [PMC free article] [PubMed]
  • Lan, T. H., T. A. Delmonte, K. P. Reischmann, J. Hyman, S. P. Kowalski et al., 2000. An EST-enriched comparative map of Brassica oleracea and Arabidopsis thaliana. Genome Res. 10: 776–788. [PMC free article] [PubMed]
  • Levin, D. E., 2002 The Role of Chromosomal Change in Plant Evolution. Oxford University Press, New York.
  • Lukens, L., F. Zou, D. J. Lydiate, I. A. P. Parkin and T. C. Osborn, 2003. Comparison of a Brassica oleracea genetic map with the genome of Arabidopsis thaliana. Genetics 164: 359–372. [PMC free article] [PubMed]
  • Lydiate, D., P. Dale, U. Lagercrantz, I. Parkin and P. Howell, 1995. Selecting the optimum genetic background for transgenic varieties, with examples from Brassica. Euphytica 85: 351–358.
  • Mariani, C., V. Gossele, M. De Beuckeleer, M. De Block, R. B. Goldberg et al., 1992. A chimeric ribonuclease–inhibitor gene restores fertility to male sterile plants. Nature 357: 384–387.
  • McClintock, B., 1929. Chromosome morphology of Zea mays. Science 69: 629–630. [PubMed]
  • Navarro, A., and N. H. Barton, 2003. Chromosomal speciation and molecular divergence–accelerated evolution in rearranged chromosomes. Science 300: 321–324. [PubMed]
  • Newell, C. A., M. L. Rhodas and D. L. Bidney, 1984. Cytogenetic analysis of plants regenerated from tissue explants and mesophyll protoplasts of winter rape, Brassica napus L. Can. J. Genet. Cytol. 26: 752–761.
  • Osborn, T. C., C. Kole, I. A. P. Parkin, M. Kuiper, D. J. Lydiate et al., 1997. Comparison of flowering time genes in Brassica rapa, B. napus and Arabidopsis thaliana. Genetics 146: 1123–1129. [PMC free article] [PubMed]
  • Osborn, T. C., D. V. Butruille, A. G. Sharpe, K. J. Pickering, I. A. P. Parkin et al., 2003. Detection and effects of a homeologous reciprocal transposition in Brassica napus. Genetics 165: 1569–1577. [PMC free article] [PubMed]
  • Parkin, I. A. P., A. G. Sharpe, D. J. Keith and D. J. Lydiate, 1995. Identification of the A and C genomes of amphidiploid Brassica napus (oilseed rape). Genome 38: 1122–1131. [PubMed]
  • Parkin, I. A. P., A. G. Sharpe and D. J. Lydiate, 2003. Patterns of genome duplication within the Brassica napus genome. Genome 46: 291–303. [PubMed]
  • Pires, J. C., J. Zhao, M. E. Schranz, E. J. León, P. A. Quijada et al., 2004. Flowering time divergence and genomic rearrangements in resynthesized Brassica polyploids (Brassicaceae). Biol. J. Linn. Soc. 82: 675–688.
  • Rieseberg, L. H., 2001. Chromosomal rearrangements and speciation. Trends. Ecol. Evol. 16: 351–358. [PubMed]
  • Scarth, R., S. R. Rimmer and P. B. E. McVetty, 1992. Reward summer turnip rape. Can. J. Plant Sci. 72: 839–840.
  • Schranz, M. E., P. Quijada, S. B. Sung, L. Lukens, R. Amasino et al., 2002. Characterization and effects of the replicated flowering time gene FLC in Brassica rapa. Genetics 162: 1457–1468. [PMC free article] [PubMed]
  • Sharpe, A. G., I. A. P. Parkin, D. J. Keith and D. J. Lydiate, 1995. Frequent nonreciprocal translocations in the amphidiploid genome of oilseed rape (Brassica napus). Genome 38: 1112–1121. [PubMed]
  • Song, K., T. C. Osborn and P. H. Williams, 1993. Development of synthetic Brassica amphidiploids by reciprocal hybridization and comparison to natural amphidiploids. Theor. Appl. Genet. 89: 885–894. [PubMed]
  • Stam, P., 1993. Construction of integrated genetic linkage maps by means of a new computer package: JoinMap. Plant J. 3: 739–744.
  • Stebbins, G. L., 1947. Types of polyploids: their classification and significance. Adv. Genet. 1: 403–429. [PubMed]
  • Stebbins, G. L., 1950 Variation and Evolution in Plants. Columbia University Press, New York.
  • Stebbins, G. L., 1971 Chromosome Evolution in Higher Plants. Edward Arnold, London.
  • Teutonico, R. A., and T. C. Osborn, 1994. Mapping of RFLP and qualitative trait loci in Brassica rapa and comparison to the linkage maps of B. napus, B. oleracea, and Arabidopsis thaliana. Theor. Appl. Genet. 89: 885–894. [PubMed]
  • Thormann, C. E., M. E. Ferreira, L. E. A. Camargo, J. G. Tivang and T. C. Osborn, 1994. Comparison of RFLP and RAPD markers to estimating genetic relationships within and among cruciferous species. Theor. Appl. Genet. 88: 973–980. [PubMed]
  • Thormann, C. E., J. Romero, J. Mantet and T. C. Osborn, 1996. Mapping loci controlling the concentrations of erucic and linolenic acids in seed oil of Brassica napus L. Theor. Appl. Genet. 93: 282–286. [PubMed]
  • U, N., 1935. Genome analysis in Brassica with special reference to the experimental formation of B. napus and peculiar mode of fertilization. Jpn. J. Bot. 7: 389–452.
  • Uzunova, M., W. Ecke, K. Weissleder and G. Röbbelen, 1995. Mapping the genome of rapeseed (Brassica napus L.). I. Construction of an RFLP linkage map and localization of QTLs for seed glucosinolate content. Theor. Appl. Genet. 90: 194–204. [PubMed]
  • Van Ooijen, J. W., and R. E. Voorrips, 2001 JoinMap Version 3.0: Software for the Calculation of Genetic Linkage Maps. Plant Research International, Wageningen, The Netherlands.
  • Voorrips, R. E., 2002. MapChart: software for the graphical presentation of linkage maps and QTLs. J. Hered. 93: 77–78. [PubMed]
  • Xu, Y., L. Zhu, J. Xiao, N. Huang and S. R. McCouch, 1997. Chromosomal regions associated with segregation distortion of molecular markers in F2, backcross, doubled haploid, and recombinant inbred populations in rice (Oryza sativa L.) Mol. Gen. Genet. 253: 535–545. [PubMed]

Articles from Genetics are provided here courtesy of Genetics Society of America
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...