Logo of geneticsGeneticsCurrent IssueInformation for AuthorsEditorial BoardSubscribeSubmit a Manuscript
Genetics. Nov 2004; 168(3): 1575–1584.
PMCID: PMC1448766

Comparing the Linkage Maps of the Close Relatives Arabidopsis lyrata and A. thaliana


We have constructed a genetic map of Arabidopsis lyrata, a self-incompatible relative of the plant model species A. thaliana. A. lyrata is a diploid (n = 8) species that diverged from A. thaliana (n = 5) ~5 MYA. Mapping was conducted in a full-sib progeny of two unrelated F1 hybrids between two European populations of A. lyrata ssp. petraea. We used the least-squares method of the Joinmap program for map construction. The gross chromosomal differences between the two species were most parsimoniously explained with three fusions, two reciprocal translocations, and one inversion. The total map length was 515 cM, and the distances were 12% larger than those between corresponding markers in the linkage map of A. thaliana. The 72 markers, consisting of microsatellites and gene-based markers, were spaced on average every 8 cM. Transmission ratio distortion was extensive, and most distortions were specific to each reciprocal cross, suggesting cytoplasmic interactions. We estimate locations and most probable genotype frequencies of transmission ratio distorting loci (TRDL) with a Bayesian method and discuss the possible reasons for the observed distortions.

ARABIDOPSIS thaliana has become the plant molecular biology model organism partly because of its very small genome size. It has also become a focus of plant comparative genomics studies (Schmidt 2002). Since the sequencing of the A. thaliana genome (Arabidopsis Genome Initiative 2000), the genomic organization of A. thaliana has been compared with that of many other species. This work has revealed a complex history of genome-wide duplication and subsequent gene loss (Lan et al. 2000; Vision et al. 2000; Simillion et al. 2002). Even in rather remote relatives, such as rice, tomato, and potato, syntenic stretches have been found (Devos et al. 1999; Ku et al. 2000; Gebhardt et al. 2003).

Comparisons with more closely related species have shown that the rate of chromosomal evolution in the Brassicaceae is quite rapid. Kowalski et al. (1994) examined rates of chromosomal evolution between Brassica oleracea and A. thaliana and found 26 rearrangements. In a more detailed comparative mapping effort between B. nigra and A. thaliana, the conserved segments were only ~8 cM in length (Lagercrantz 1998). As the divergence time between Brassica and A. thaliana is estimated to be ~20 MY, this gives rise to a high rate of chromosomal evolution. The rapid rate of chromosomal evolution was assigned to the polyploid Brassica lineage. A comparison of A. thaliana and the even closer relative, Capsella rubella, has shown extensive microsynteny (Acarkan et al. 2000; Rossberg et al. 2001). However, the closest relatives of A. thaliana have not been mapped comparatively at the whole-genome level.

In this article we construct a genetic map of A. lyrata, the closest relative of A. thaliana (Koch et al. 1999). This species is estimated to have diverged from A. thaliana ~5 MYA (Koch et al. 2000). The synonymous divergence is 12–15% (e.g., Wright et al. 2003 and references therein). A. lyrata, like many other close relatives, has eight chromosomes (Jones 1963) while A. thaliana has five. The genome size is known to be about one and a quarter to twice as large as that of A. thaliana (T. Mitchell-Olds, unpublished results; E. Earle, unpublished results). The species can be crossed with A. thaliana (síček 1967), and the hybrids produce backcross progeny, but there seems to be no recombination within chromosomes in the hybrids (Nasrallah et al. 2000). A. lyrata is self-incompatible (Jonsell et al. 1995) and outcrossing (Schierup 1998; rkkäinen et al. 1999) and has become a species of interest for population genetics (van Treuren et al. 1997; Savolainen et al. 2000; Mitchell-Olds 2001; Clauss et al. 2003; Wright et al. 2003).

The markers in the map are largely orthologs of A. thaliana flowering time genes. In addition, we have used other genes and microsatellite loci. The goal of the map is to examine patterns of chromosomal evolution at a large-scale structural level between these closely related species, i.e., to answer questions such as: What kinds of changes have given rise to the difference of chromosome numbers of five and eight? Are these whole-arm translocations or chromosome fusions, or many smaller changes? Are there as many changes per million years between A. lyrata and A. thaliana as between A. thaliana and Brassica? Recalling that the genome size is larger than that in A. thaliana, what is the genetic size of the map? Where have the reductions of A. thaliana genome size occurred?

We used a three-generation pedigree, as we tried to avoid the possible effects of inbreeding depression in this self-incompatible species. As the mapping cross is between diverged populations of A. lyrata, we also examined the extent and possible causes of transmission ratio distortion in the cross. We used an extension of the Bayesian method of Vogl and Xu (2000) for analyzing transmission ratio distortion to an outbred full-sib family.

A further aim of the map is to provide a tool for A. lyrata genomic work. Once the patterns of the chromosomal rearrangements are known, it will be possible to examine the changes in more detail. In analyzing quantitative traits, the locations of quantitative trait loci between species can be efficiently compared. As this is an outcrossing self-incompatible species, there are no recombinant inbred lines available, but the plants genotyped for the map have been vegetatively propagated. The genotypic data and the vegetative clones are available, so the map can be used as such for QTL mapping or can be expanded for more detailed mapping.



From each of the A. lyrata ssp. petraea populations, Mjällom (Sweden) and Karhumäki (Russia; for more information on the populations see van Treuren et al. 1997), two random individuals were chosen to produce two independent hybrid plants (K12 × M6: KM4-2 and M3 × K10: KM6-5). These two F1 plants were crossed with each other to produce F2 progeny, using both plants as pollen donor and recipient. This crossing scheme was chosen to avoid inbreeding depression in the F2 progeny. In reciprocal 1, KM6-5 was the pollen recipient and KM4-2 was the pollen donor. In reciprocal 2, KM4-2 was the pollen recipient and KM6-5 was the pollen donor. The four grandparents, KM4-2, KM6-5, and 76-195 F2 progeny in about equal proportions from the two reciprocal crosses were scored for marker genotypes. Most markers were scored for ~90 individuals and some loci that showed transmission ratio distortion were scored for additional individuals, leading to varying numbers of missing individuals for the markers (see supplemental Table 1 at http://www.genetics.org/supplemental/).

Markers used in the map construction


Altogether 72 different markers were used (Table 1). We used microsatellites (Clauss et al. 2002) and gene-based markers (Kuittinen et al. 2002) described earlier for various Brassicaceous species and some A. thaliana microsatellites (Bell and Ecker 1994). In addition, 47 other gene-based markers were adopted in this study. For them amplification primers were obtained from the Arabidopsis cleaved amplified polymorphic sequence marker database (http://www.arabidopsis.org/), or they were designed on the basis of the A. thaliana sequence in the way described in Kuittinen et al. (2002). The fragment sizes ranged from 0.5 to 2.4 kb. Amplification was carried out using standard conditions (MgCl2 1.5–3.3 mm, annealing temperature 48°–60°). Polymorphisms in the genes were discovered either by cutting the PCR products of the F1 parents and some progeny with a set of 8–20 restriction enzymes or by sequencing the PCR product of the F1 plants. Either the progeny genotypes at a segregating site were scored by cutting the PCR products with a suitable restriction enzyme and visualizing the fragments on agarose gel or, if no suitable enzyme was available, single nucleotide polymorphisms were scored with the ABI Prism SNaPshot Multiplex kit (Applied Biosystems, Foster City, CA). Some indel polymorphisms were scored as length variation on agarose gels. For some genes with indel polymorphisms a new primer with fluorescent end label was used to produce short fragments (<350 bp) that could be run on sequencing gels. Sequencing gels were run on the ABI Prism 377 sequencer and fragments were detected with the Genescan program and their sizes determined with the Genotyper program (Applied Biosystems). Polymorphism for the ADH1 locus was detected by heteroduplex formation. PCR products were heated at 95° for 5 min, then at 65° for 1 hr, and then allowed to cool slowly to room temperature. Electrophoretic separation was then performed using a 1× MDE gel (FMC Bioproducts, Rockland, ME) run in 0.6× TBE for ~4.5 hr at 30 W at a constant temperature of 10°. Microsatellites were separated on 6% polyacrylamide or on 4% Metaphor gels and they were visualized either with silver staining or with FMBIO fluorescent gelscanner. Most microsatellites were run on sequencing gels using a fluorescent end label in one of the primers as described above for indel polymorphisms in the genes. The polymorphism could be scored in a codominant fashion at all loci except CRY2 and SLL2, which were scored as dominant markers.


In this cross a marker can segregate in only one of the parents (aa × ab, ab × aa) or in both parents (ab × ab if there are two alleles, ab × ac if there are three alleles, and ab × cd if there are four alleles). There were 15 fully informative loci; i.e., either three or four alleles were segregating as described above. These included many microsatellites and some genes in which two polymorphic sites in different configurations could be scored. Most of the markers (41) segregated in the ab × ab configuration, and 16 segregated in only one of the F1 parents. Fit of the proportions of different marker genotypes in the F2 progeny to the expected 1:1, 1:2:1, 1:1:1:1, or 1:3 ratios was tested with χ2 tests separately in both reciprocal crosses with the Joinmap program (Stam 1993). We used a locus-by-locus significance level of α = 0.05.

The four genotypes in a fully informative cross should be found in equal numbers. Processes in both gametic and zygotic stages can give rise to deviations from the expected 1:1:1:1 segregation ratio. In the case of a fully informative marker or a marker that is segregating in a backcross manner, it is possible to dissect which of the alleles present in the F1 parents is in excess or is deficient. If the bias is due to selection against some allele(s) at the gametic stage, the observed genotype frequencies should match expected frequencies that are calculated from the observed allele frequencies, assuming random union of gametes from each F1 parent. This is also expected if there is zygotic selection with additive fitness values. If the observed genotype frequencies do not match the expected, selection on the zygote(s) is indicated. χ2 tests were applied to test for deviations from equal allele frequencies in each F1 parent, if possible, and to test for deviations from the random union of gametes, given the observed allele frequencies.

Map construction:

We used the Joinmap program (version 3.0; Stam 1993) that uses a weighted least-squares method that adds markers sequentally in the map. The map was constructed with all scored markers, including those with significant transmission ratio distortion. The linkage groups (LGs) were determined using a LOD score threshold of 3.5. The Kosambi mapping function was used for calculation of the pairwise distances. For each LG, a map was constructed using thresholds of LOD 0.01 and a recombination frequency of 0.49 for pairs of markers, to use all linkage information. A threshold of 3 was used for an increase of the total χ2 for the difference between observed and calculated map distances per added degrees of freedom, and a LOD threshold of 12 was used for triplet formation. Marker order was permutated after each marker was added. For most linkage groups all markers could be mapped after this. In linkage group AL1 the marker VIP1, in linkage group AL7 the marker PRHA, and in linkage group AL4 the markers CK2a and CON were mapped after relaxing the thresholds.

After map construction the location and effect of possible transmission ratio distortion loci (TRDL) in the chromosomes were estimated by applying the Bayesian method of Vogl and Xu (2000). We use an extension of the method to outbreed full-sib data, such as in Fishman et al. (2001). It allows several multiplicatively interacting TRDL in a chromosome and can use partially informative marker information. As the segregation ratios appeared to differ among the two reciprocals, we performed a TRDL analysis separately in the two reciprocals. The maximum number of TRDL was set to four per chromosome. To obtain the posterior probabilities of the TRDLs, 10,000 simulations were run for each chromosome. The most likely position and the genotype frequencies were obtained.


Linkage groups:

Markers were first assigned to linkage groups separately in the two reciprocal data sets. With a LOD 3 criterion there were no contradictory groupings. Linkage groups were, however, fragmented to different degrees in the two reciprocals, due to low sample sizes. Recombination rates between markers did not differ significantly between the reciprocals for the vast majority of comparisons as judged with the Joinmap program. In the end, the map was constructed with both reciprocals in one data set. The markers were divided into eight linkage groups, AL1–8 (Figure 1).

Figure 1.
The A. lyrata map constructed with the Joinmap method. The homology with the A. thaliana chromosomes is indicated by color. The positions and the most probable genotype frequencies at the transmission ratio distortion loci are indicated with pie diagrams ...

Translocations and fusions:

As A. lyrata has a higher chromosome number than A. thaliana, we expected some large-scale rearrangements between the two genomes. Three breakages/fusions and two reciprocal translocations are sufficient to explain most of the differences in the maps of the two species (Figure 1). We inferred a breakage/fusion in the lower arm of chromosome 1, in the lower arm of chromosome 2, and in the lower arm of chromosome 5 of A. thaliana. AL3 consists of markers from the upper part of A. thaliana chromosome 3 and of markers from the middle part of A. thaliana chromosome 2. AL5 consists of markers from the upper part of A. thaliana chromosome 2 and the lower part of chromosome 3. This suggests a reciprocal translocation between A. thaliana chromosomes 2 and 3. AL6 is a combination of the upper arm of A. thaliana chromosome 4 and the upper arm of chromosome 5. Similarily, AL7 consists of markers in the lower part of A. thaliana chromosome 4, and markers from the middle part of A. thaliana chromosome 5. Thus a reciprocal translocation has occurred between these two chromosomes.

Marker order:

The marker order of A. lyrata, to a great extent, was similar to that of the A. thaliana recombinant inbred map (Lister and Dean 1993; http://www.arabidopsis.org/). Compared with A. thaliana, pairs of closely linked markers in AL1 were reversed in order (LHY-ATHACS, PHYA-AXR, GI-F19G10), as were ICE3 and ATPEN1 at the lower arm of AL7. In these cases, the markers were closely situated and the actual order could well be consistent with the order in A. thaliana. The five markers in the upper end of AL6 (GA1, LD, ICE7, AthDET1, CRY1) were clearly in a different order compared to A. thaliana (LD, GA1, CRY1, AthDET1, ICE7). At least one inversion (CRY1, DET1, ICE7) can be inferred from the data. There is more uncertainty in the order of the closely linked markers LD and GA1.

In one case a locus was out of its A. thaliana context. The FRI locus, situated in chromosome 4 in A. thaliana, was mapped to AL8, which is homologous to the lower arm of A. thaliana chromosome 5. It was linked to the EMF2 gene and microsatellite MFB13. Using a BLAST search we found in the A. thaliana chromosome 5 close to the EMF2 gene a genomic region with high homology to a part of the FRI gene. A hypothetical protein was annotated across this sequence.

Length of the maps:

The lengths of the linkage groups varied from 54.4 to 77.8 cM and summed to a total length of 515 cM. There was, on average, one marker every 8 cM. The total genome length is not likely to exceed this much because we selected markers at the ends of the A. thaliana linkage groups, and there are only ~13 Mb altogether between the telomeres and the outermost markers or between the markers flanking the three inferred chromosomal fusions in A. thaliana. The markers used in the A. lyrata map cover 483 cM in the five chromosomes of A. thaliana (TAIR database at http://www.arabidopsis.org/). Considering the colinear parts of the A. thaliana and A. lyrata maps, the Joinmap procedure resulted in genetic distances that were, on average, only 12% larger than those in the A. thaliana map. Since the A. thaliana map was also constructed with Joinmap, this comparison should be unbiased. The largest differences in the map distances between the two species are in AL2 and AL8, which correspond to the lower arms of A. thaliana chromosomes 1 and 5. These LGs are about twice as large in A. lyrata as in the corresponding areas in A. thaliana. AL4 was also extended to 140% compared to the homologous region in A. thaliana. The largest reductions in size were observed in the lower part of AL6 and in the upper part of AL7.

Transmission ratio distortion:

In reciprocals 1 and 2, there were 18 (25%) and 25 (35%) distorted markers at the locus-by-locus significance level α = 0.05 (Table 1 and supplemental Table 1 at http://www.genetics.org/supplemental/). The two reciprocals were distorted at only seven common loci. Thus, in total, there were 36 distorted markers (50%). Significant segregation distortion was found in 25 of the 72 markers tested (35%) with the χ2 tests at α = 0.05 when the data from both reciprocals were combined. The tests detected more distortion in separate data sets in spite of smaller sample sizes, obviously because the distortions were different in the two reciprocals. Distorted markers were usually linked to other similarily distorted markers so that certain chromosomal areas were distorted in the same direction, indicating a biological reason for the distortion instead of a genotyping error. If the genotype of a certain marker is mistyped frequently, the segregation ratio of this marker may be different from its neighboring markers. Hence, the algorithm for TRDL analysis would infer two spurious TRDLs with opposite effects to either side of a mistyped marker. We did not find evidence of closely spaced TRDLs of opposite effect. This indicates no major problems with genotyping in the data set. If only areas with at least two similarily distorted linked loci were taken into account, we obtained three distorted areas in reciprocal 1 and seven distorted areas in reciprocal 2. Only one of the areas (in AL6) was common to both reciprocals. The results from the Bayesian TRDL analysis were mostly congruent with those from the simple χ2 tests. With a conservative cutoff value (<50% of the runs of the posterior distribution resulted in zero TRDL in a chromosome) there was support for three TRDL in reciprocal 1 and for five TRDL in reciprocal 2 (Figure 1).

Of the 11 markers in AL6, 8 were distorted in one or the other reciprocal cross (Table 1 and supplemental Table 1 at http://www.genetics.org/supplemental/). The Bayesian method strongly suggested a TRDL at the top of this linkage group in both reciprocals (TRLD1-3, TRLD2-3), as the posterior proportion resulting in one TRDL was 69% in both cases (Figure 1). The distortions were similar in the two crosses; in both cases there was an excess of Karhumäki alleles. Conditional on the number of observed gametes, the observed and expected genotype frequencies were similar. In reciprocal 1 (KM6-5 × KM4-2) there were three additional distorted areas. The Bayesian analysis detected a TRDL close to ENR-A in AL3 with 56% posterior probability (TRDL1-1). ENR-A was the only significantly distorted marker in the region detected with χ2 tests, showing an excess of heterozygotes. In AL4 there was an excess of heterozygotes at two adjacent distorted marker loci (ATEX6 and COP1; Table 1 and supplement Table 1 at http://www.genetics.org/supplemental/). The Bayesian method did not, however, support a TRDL in this linkage group. In AL5 there was an excess of Mjällom alleles at four linked marker loci; 53% of the runs supported a model with one TRDL (TRDL1-2). The Mjällom allele from KM6-5 was especially favored, while the observed genotype frequencies followed the expected. In reciprocal 2 (KM4-2 × KM6-5), the Bayesian analysis gave support for one TRDL (65.4% posterior proportion) in AL1 (TRDL2-1). There was an excess of the Karhumäki allele from KM6-5, and comparison of observed and expected genotype frequencies revealed additionally a deficiency of homozygotes at markers close to this TRDL. Some markers at the upper part of the LG were also distorted but in another direction, and the Bayesian analysis did not detect a TRDL there. In AL2 there was a deficiency of homozygotes at two linked loci, and the one TRDL model (TRDL2-2) had strong support (51% posterior proportion). In AL4 there was an excess of Karhumäki alleles at two linked loci (CON and CK2α) with no difference in observed and expected genotype frequencies. The Bayesian analysis did not support a TRDL in this linkage group. In a group of loci in AL7 there was a deficiency of Mjällom alleles and evidence for an excess of heterozygotes. In another set of linked loci in AL7, Mjällom alleles from KM4-2 were in excess and the observed genotype frequencies matched those expected, given the gametic frequencies. The Bayesian analysis supported two TRDL in these regions with a posterior proportion of 49% (TRDL2-4 and TRDL2-5).


Chromosomal evolution:

Given that the basic chromosome number among most close relatives of A. thaliana is eight and that lower chromosome numbers in some species result from reduction of chromosome number (Jones 1963; Koch et al. 1999), the differences in chromosome number between A. thaliana and A. lyrata are most parsimoniously accounted for by three fusions in the A. thaliana lineage. Additionally, we inferred an inversion (CRY1-DET-ICE7) and two reciprocal translocations. One of the breakpoints of the other reciprocal translocation (AL6/AL7) is in an interval that is known to contain a centromere in A. thaliana (PHYC-CLC-d), and the other breakpoint is next to CRY1, which is close to the centromere. In the other reciprocal translocation (AL3/AL5) the breakpoints were also close to the centromere positions in A. thaliana.

In A. thaliana the FRI locus is on chromosome 4 but in A. lyrata it was mapped to a linkage group that is orthologous to A. thaliana chromosome 5. However, A. thaliana has a highly homologous sequence on chromosome 5, which lacks parts of the coding region, probably being a pseudogene. The nucleotide divergence between the two A. thaliana sequences was half of that between either A. thaliana sequence and the sequence of A. lyrata. This suggests that the duplication occurred recently in the A. thaliana lineage. Duplication and deletion of paralogous genes are known to have broken colinearity between different taxa and within species (e.g., Ku et al. 2000; Vision et al. 2000; Fu and Dooner 2002; Li and Gill 2002).

The colinearity of the maps of these closely related species is apparently very high. High colinearity has also been found between genomes of A. thaliana and C. rubella (Brassicaceae) that have diverged 6.2–9.8 MYA (Acarkan et al. 2000). More markers could evidently reveal smaller-scale rearrangements, and only sequencing of long genomic contigs would reveal the details of genome evolution. Extensive microcolinearity was evident between sequences of C. rubella and A. thaliana (Rossberg et al. 2001), but several small-scale chromosomal rearrangements were detected between the more distant B. oleracea and A. thaliana sequences (Quiros et al. 2001).

The three fusions, two translocations, and one inversion that have occurred during the 5 MY after the lyrata-thaliana divergence (Koch et al. 2000) would give a rate estimate of 0.6 rearrangement/MY/genome. In the family Brassicaceae the rate of chromosomal evolution is exceptionally high. Kowalski et al. (1994) found 17 translocations and nine inversions between B. oleracea and A. thaliana with a map that had a resolution comparable to our map (15 cM between markers in A. thaliana). With a higher-resolution comparative map, Lagercrantz (1998) inferred that an average colinear segment is 8 cM and as many as 90 chromosomal rearrangements differentiate the genomes of B. nigra and A. thaliana. Lukens et al. (2003) obtained a somewhat larger estimate for an average colinear segment (11.6 cM) in B. oleracea, which corresponds to 3.8 Mbp in A. thaliana. With a divergence time of 14–24 MY (Yang et al. 1999; Koch et al. 2000), 26 rearrangements would result in a rate of 0.93–0.54 rearrangements/MY/genome and 90 rearrangements in 3.2–1.9 rearrangements/MY/genome. The rate of chromosomal evolution after the lyrata-thaliana divergence is thus lower than the estimate between Brassica and A. thaliana, if data with a comparable resolution are considered, and clearly lower than estimates from high-resolution mapping. The rates in comparisons among other plant taxa such as Lycopersicon-Solanum-Capsicum, Oryza-Zea-Sorghum, or Gossypium ssp. vary between 0.17 and 0.45 rearrangements/ MY (see Lagercrantz 1998, Table 3), and the estimate for A. lyrata and A. thaliana (0.6) is only slightly higher than that. However, the comparisons depend on the resolution of the map, on the criteria used for orthology, and on the estimates of divergence times.

It has been suggested that most of the chromosomal rearrangements between A. thaliana and Brassica have occurred as a result of polyploidization that took place after the divergence of Arabidopsis and Brassica (Lagercrantz and Lydiate 1996; Lagercrantz 1998; Lukens et al. 2003). The majority of rearrangements would thus have occurred in the Brassica lineage. The rather low number of chromosomal rearrangements in our study is in line with the hypothesis of a lower rate of genome evolution in the Arabidopsis lineage. Lagercrantz (1998) suggests that the ephemeral life styles of wild Brassica species would also favor fixation of chromosomal rearrangements that would be partially deleterious as heterozygotes. This should actually be true also for the weedy selfer A. thaliana. In contrast, A. lyrata is a perennial outcrosser that is likely to maintain more permanent populations. The location of A. lyrata markers in Brassica maps is currently not known. When the Brassica genome project proceeds, it will hopefully be possible to compare maps of the Arabidopsis and Brassica species in detail. This would allow defining whether the rearrangements have occurred in the lineage leading to A. lyrata or to A. thaliana.

The DNA content of A. lyrata, as measured by flow cytometry, is 125–185% as much as that in A. thaliana (T. Mitchell-Olds, unpublished results; E. Earle, unpublished results). We have regularly obtained only one sequence that corresponds to the one A. thaliana locus when developing PCR-based markers. On the basis of this evidence, gene duplications do not account for the higher genome size of A. lyrata compared to A. thaliana. Instead, this implies changes in noncoding DNA content. It has been suggested that intron sizes as well as intergenic regions may be larger in species with higher DNA content and that genes in large genomes may reside in gene-rich islands. Centromeric heterochromatin-associated sequences have often been shown to contribute to the variation in genome sizes. The reasons for genome-size differences are still not clear. There is some evidence for shorter introns in A. thaliana than in A. lyrata or Brassica (Wright et al. 2003) and shorter intergenic spacers in A. thaliana than in B. oleracea (Quiros et al. 2001). Despite the larger genome size, the length of the genetic map of A. lyrata was only slightly larger than that of A. thaliana.

Transmission ratio distortion:

Transmission ratio distortion can result from many different processes that can act either before or after fertilization (reviewed, e.g., in Korbecka et al. 2002 and in Fishman et al. 2001). Distortion of gametic frequencies can result from meiotic drive or meiotic disruptions associated with chromosomal rearrangements. Recombinant gametes may have differential fitness due to aneuploidy or epistatic interactions. Gametes may also have differential fertilization success. Selection may act after fertilization so that different zygotic genotypes have differential survival. Segregating deleterious mutations as well as male sterility and self-incompatibility alleles are known to cause segregation distortions. Much of the distortion in wide crosses is thought to be due to epistatic interactions between loci, i.e., Dobzhansky-Muller incompatibilities (e.g., Orr and Turelli 2001). With the cross design we aimed at excluding the influence of inbreeding depression or self-incompatibility alleles, but many other factors could have caused the observed distortions.

The extent of transmission ratio distortion was surprisingly high as the cross was within species. In each reciprocal cross, 25 and 35% of the loci were distorted at the 0.05 significance level, and taken together, 50% of the markers deviated from the Mendelian ratios. Remarkably, only one of the distorted regions was common to both reciprocal crosses and eight of the regions were present in only one of the crosses. The high levels of distortion found here are at the upper range of those normally found in crosses within species, and they resemble more those detected in crosses between species (reviewed in Jenczewski et al. 1997; Whitkus 1998; Fishman et al. 2001). This indicates that the Swedish Mjällom and the Russian Karhumäki populations are widely divergent. This is consistent with earlier information from molecular markers. Nei's standard genetic distance estimated from allozyme data was 0.39 and Gst based on microsatellites was 0.59 between the Karhumäki and Mjällom populations (van Treuren et al. 1997). Markers in a cross between two other European A. lyrata populations also exhibit a high level of segregation distortion (J. Leppälä et al., unpublished results). Thus strong distortion is not specific to these two populations.

The mapping population was an outbred F2 population with a maximum of four segregating alleles. We also used two reciprocal crosses. This experimental setup allowed obtaining information on different aspects of plant life cycle. With single-locus analyses segregation distortion could be assigned to the two reciprocals and to some extent to separate parents, depending on the informativeness of the loci. We could also analyze whether there was evidence for zygotic selection. The Bayesian analysis combined information from all markers and gave the most probable TRDL genotype frequencies.

In AL6, segregation distortion was similar in the two reciprocals. Mjällom alleles were selected against, and there was no indication of recessivity of the disfavored Mjällom allele (conditional on the gametic frequencies, genotypic frequencies did not deviate from those expected). This suggests that the distortion is not due to segregation of a deleterious recessive allele that was fixed in the Mjällom population. The result is consistent with gametic selection or zygotic selection with additive gene effects. Because most markers were only partially informative, we know that the distortion occurred at least in the pollen parent but cannot exclude distortion in the mother. Differential pollen-tube growth has been suggested as a reason for much of the segregation distortion observed in, e.g., Mimulus guttatus (Fishman et al. 2001). This kind of selection could be the reason for the segregation distortion in AL6, but would need additional data to be confirmed.

All the other distorted areas were specific for each reciprocal cross. The difference between reciprocals could result from interaction between cytoplasmic and nuclear factors, because the cytoplasms of the two F1 plants originated from different populations. It is expected that mitochondrial and nuclear genes originating from the same population would interact positively. In several species mitochondrial genes are known to cause male sterility while nuclear genes may restore fertility (e.g., Budar and Pelletier 2001). If the impairment of pollen function were induced by the mitochondrial genotype of the pollen parent, the Karhumäki allele should be favored in KM4-2 in reciprocal 1 and the Mjällom allele in KM6-5 in reciprocal 2. There was no evidence for this. Cytoplasmic interactions could also occur in other stages, as suggested by Tiffin et al. (2001). The two F1 plants had different nuclear genotypes, and also polymorphic nuclear factors that influence pollen, style, or endosperm functions could potentially have caused incompatibility in the cross, resulting in different distortions in the reciprocals.

Several gametic and zygotic reproductive barriers were found in a cross between Japonica and Indica rice varieties (Harushima et al. 2001, 2002). In this cross, deviations were evident not only in gametic ratios but also in many cases from expected zygotic frequencies (TRDL1-1, TRDL2-1, TRDL2-2, and TRDL2-4), suggesting selection on the zygote.

Segregation distortion exhibited different patterns, so not only are many genes responsible, but also they possibly act through different mechanisms. Further crosses, as well as data on seed and pollen fertility, would be needed to distinguish between the contribution of cytoplasmic and nuclear factors and to find out in which stage the distortion occurs.

A practical concern of transmission ratio distortion is that it can interfere with mapping. Distortion of two loci in the same direction may result in spurious linkages between them or in a decrease in the distances between them, and loci distorted in different directions may inrease the genetic distance. Strong nonadditive interactions between linked markers can influence the map construction. Single transmission ratio distortion loci (Hackett and Broadfoot 2003) or several multiplicatively interacting transmission ratio distortion loci, however, should have little influence on a dense marker map when multipoint methods are used (C. Vogl and S. Xu, unpublished results). As the marker density in the map was reasonably high, it is unlikely that transmission ratio distortion has had a major effect on the linkage map. The data contained much missing information because for some markers more individuals were sampled than for others. Missing information can result in incorrect marker order especially at small intervals (Hackett and Broadfoot 2003). Potential nonadditively interacting linked viability loci and uneven numbers of genotyped markers could have an effect on the map construction. However, we have made rather conservative inferences on the marker order. The current A. lyrata map is likely to be quite robust, because in a cross between two other A. lyrata populations—with a smaller number of markers, with practically no missing individuals, and with extensive transmission ratio distortion at different locations compared to this map—the same order and similar distances between the common markers were obtained (J. Leppälä and O. Savolainen, unpublished results).

Comparative mapping as a tool:

The almost complete colinearity of the A. lyrata map with the map of the model species A. thaliana implies that we can use the positional information from the Arabidopsis genome. For instance, after QTL mapping in A. lyrata, it will be possible to select candidate genes from the corresponding region in A. thaliana for further analysis, such as complementation or association studies. The current cross will be useful for QTL mapping as such. A set of genotyped F2 plants has been micropropagated and can be used for evaluation of the desired phenotypes. The markers can also be used in other crosses. Many of the amplification primers are likely to be applicable in close species and can be used for comparative mapping. Additional markers can easily be added to the current map to increase the density of the map or to obtain information about the location of a specific gene.


We are grateful to Bryan Lazzaro for his contribution to the lab work and to Patrik Waldmann for comments on the manuscript. This work was supported by the Finnish Research Council for Biosciences and Environment.


  • Acarkan, A., M. Rossberg, M. Koch and R. Schmidt, 2000. Comparative genome analysis reveals extensive conservation of genome organization for Arabidopsis thaliana and Capsella rubella. Plant J. 23: 55–62. [PubMed]
  • Arabidopsis Genome Initiative, 2000. Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408: 796–815. [PubMed]
  • Bell, C., and J. Ecker, 1994. Assignment of 30 microsatellite loci to the linkage map of Arabidopsis. Genomics 19: 137–144. [PubMed]
  • Budar, F., and G. Pelletier, 2001. Male sterility in plants: occurrence, determinism, significance and use. C. R. Acad. Sci. III 324: 543–550. [PubMed]
  • Clauss, M., H. Cobban and T. Mitchell-Olds, 2002. Cross-species microsatellite markers for elucidating population genetic structure in Arabidopsis and Arabis (Brassicaceae). Mol. Ecol. 11: 591–601. [PubMed]
  • Clauss, M., H. Cobban and T. Mitchell-Olds, 2003. Population genetics of tandem trypsin inhibitor genes in Arabidopsis species with contrasting ecology and life history. Mol. Ecol. 12: 1287–1299. [PubMed]
  • Devos, K. M., J. Beales, Y. Nagamura and T. Sasaki, 1999. Arabidopsis-rice: Will colinearity allow gene prediction across the eudicot-monocot divide? Genome Res. 9: 825–829. [PMC free article] [PubMed]
  • Fishman, L., A. J. Kelly, E. Morgan and J. H. Willis, 2001. A genetic map in the Mimulus guttatus species complex reveals transmission ratio distortion due to heterospecific interactions. Genetics 159: 1701–1716. [PMC free article] [PubMed]
  • Fu, H., and H. Dooner, 2002. Intraspecific violation of genetic colinearity and its implications in maize. Proc. Natl. Acad. Sci. USA 99: 9573–9578. [PMC free article] [PubMed]
  • Gebhardt, C., B. Walkemeier, H. Henselewski, A. Barakat, M. Delseny et al., 2003. Comparative mapping between potato (Solanum tuberosum) and Arabidopsis thaliana reveals structurally conserved domains and ancient duplications in the potato genome. Plant J. 34: 529–541. [PubMed]
  • Hackett, C. A., and L. B. Broadfoot, 2003. Effects of genotyping errors, missing values and segregation distortion in molecular marker data on the construction of linkage maps. Heredity 90: 33–38. [PubMed]
  • Harushima, Y., M. Nakagahra, M. Yano, T. Sasaki and N. Kurata, 2001. A genome-wide survey of reproductive barriers in an intraspecific hybrid. Genetics 159: 883–892. [PMC free article] [PubMed]
  • Harushima, Y., M. Nakagahra, M. Yano, T. Sasaki and N. Kurata, 2002. Diverse variation of reproductive barriers in three intraspecific rice crosses. Genetics 160: 313–322. [PMC free article] [PubMed]
  • Jenczewski, E., M. Gherardi, I. Bonnin, J. M. Prosperi, I. Olivieri et al., 1997. Insight on segregation distortions in two intraspecific crosses between annual species of Medicago (Leguminosae). Theor. Appl. Genet. 94: 682–691.
  • Jones, B. M. G., 1963 Experimental taxonomy of the genus Arabis. Ph.D. Thesis, University of Leicester, Leicester, UK.
  • Jonsell, B., K. Kustås and I. Nordal, 1995. Genetic variation in Arabis petraea, a disjunct species in northern Europe. Ecography 18: 321–332.
  • Kärkkäinen, K., H. Kuittinen, R. van Treuren, C. Vogl, S. Oikarinen et al., 1999. Genetic basis of inbreeding depression in Arabis petraea. Evolution 53: 1354–1365.
  • Koch, M., J. Bishop and T. Mitchell-Olds, 1999. Molecular systematics and evolution of Arabidopsis and Arabis. Plant Biol. 1: 529–537.
  • Koch, M., B. Haubold and T. Mitchell-Olds, 2000. Comparative evolutionary analysis of chalcone synthase and alcohol dehydrogenase loci in Arabidopsis, Arabis and related genera (Brassicaceae). Mol. Biol. Evol. 17: 1483–1498. [PubMed]
  • Korbecka, G., P. G. L. Klinkhamer and K. Vrieling, 2002. Selective embryo abortion hypothesis revisited—a molecular approach. Plant Biol. 4: 298–310.
  • Kowalski, S. P., T.-H. Lan, K. A. Feldman 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]
  • Ku, H.-M., T. Vision, J. Liu and S. D. Tanksley, 2000. Comparing sequenced segments of the tomato and Arabidopsis genomes: large-scale duplication followed by selective gene loss creates a network of synteny. Proc. Natl. Acad. Sci. USA 97: 9121–9126. [PMC free article] [PubMed]
  • Kuittinen, H., M. Aguadé, D. Charlesworth, A. De Haan, B. Lauga et al., 2002. Primers for 22 candidate genes for ecological adaptations in Brassicaceae. Mol. Ecol. Notes 2: 258–262.
  • Lagercrantz, U., 1998. Comparative mapping between Arabidopsis thaliana and Brassica nigra indicates that Brassica genomes have evolved through extensive genome replication accompanied by chromosome fusions and frequent rearrangements. Genetics 150: 1217–1228. [PMC free article] [PubMed]
  • Lagercrantz, U., and D. 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]
  • Li, W., and B. S. Gill, 2002. The colinearity of the Sh2/A1 orthologous region in rice, sorghum and maize is interrupted and accompanied by genome expansion in the Triticaceae. Genetics 160: 1153–1162. [PMC free article] [PubMed]
  • Lister, C., and C. Dean, 1993. Recombinant inbred lines for mapping RFLP and phenotypic markers in Arabidopsis thaliana. Plant J. 4: 745–750.
  • Lukens, L., F. Zou, D. Lydiate, I. Parkin and T. Osborn, 2003. Comparison of a Brassica oleracea genetic map with the genome of Arabidopsis thaliana. Genetics 164: 359–372. [PMC free article] [PubMed]
  • Měsíček, J., 1967. The chromosome morphology of Arabidopsis thaliana (L.) Heynh. and some remarks on the problem of Hylandra suecica (Fr.) Love. Folia Geobotanica Phytotaxonomica 2: 433–436.
  • Mitchell-Olds, T., 2001. Arabidopsis thaliana and its wild relatives: a model system for ecology and evolution. Trends Ecol. Evol. 16: 693–700.
  • Nasrallah, M. E., K. Yogeeswaran, S. Snyder and J. B. Nasrallah, 2000. Arabidopsis species hybrids in the study of species differences and evolution of amphiploidy in plants. Plant Physiol. 124: 1605–1614. [PMC free article] [PubMed]
  • Orr, A., and M. Turelli, 2001. The evolution of postzygotic isolation: accumulating Dobzhansky-Muller incompatibilities. Evolution 55: 1085–1094. [PubMed]
  • Quiros, C., F. Grellet, J. Sadowski, T. Suzuki, G. Li et al., 2001. Arabidopsis and Brassica comparative genomics: sequence, structure and gene content in the ABI1-Rps2-Ck1 chromosomal segment and related regions. Genetics 157: 1321–1330. [PMC free article] [PubMed]
  • Rossberg, M., K. Theres, A. Acarkan, R. Herrero, T. Schmitt et al., 2001. Comparative sequence analysis reveals extensive microcolinearity in the lateral suppressor regions of the tomato, Arabidopsis, and Capsella genomes. Plant Cell 13: 979–988. [PMC free article] [PubMed]
  • Savolainen, O., C. H. Langley, B. P. Lazzaro and H. Fréville, 2000. Contrasting patterns of nucleotide polymorphism at the alcohol dehydrogenase locus in the outcrossing Arabidopsis lyrata and the selfing Arabidopsis thaliana. Mol. Biol. Evol. 17: 645–655. [PubMed]
  • Schierup, M. H., 1998. The effect of enzyme heterozygosity on growth in a strictly outcrossing species, the self-incompatible Arabis petraea (Brassicaceae). Hereditas 128: 21–31.
  • Schmidt, R., 2002. Plant genome evolution: lessons from comparative genomics at the DNA level. Plant Mol. Biol. 48: 21–37. [PubMed]
  • Simillion, C., K. Vandepoele, M. C. Van Montagu, M. Zabeau and Y. Van De Peer, 2002. The hidden duplication past of Arabidopsis thaliana. Proc. Natl. Acad. Sci. USA 99: 13627–13632. [PMC free article] [PubMed]
  • Stam, P., 1993. Construction of integrated genetic linkage maps by means of a new computer package: JoinMap. Plant J. 3: 739–744.
  • Tiffin, P. M. S. Olson and L. C. Moyle, 2001. Asymmetrical crossing barriers in angiosperms. Proc. R. Soc. Lond. Ser. B 268: 861–867. [PMC free article] [PubMed]
  • van Treuren, R., H. Kuittinen, K. Kärkkäinen, E. Baena-Gonzalez and O. Savolainen, 1997. Evolution of microsatellites in Arabis petraea and Arabis lyrata, outcrossing relatives of Arabidopsis thaliana. Mol. Biol. Evol. 14: 220–229. [PubMed]
  • Vision, T. J., D. G. Brown and S. D. Tanksley, 2000. The origins of genomic duplications in Arabidopsis. Science 290: 2114–2117. [PubMed]
  • Vogl, C., and S. Xu, 2000. Multipoint mapping of viability and segregation distorting loci using molecular markers. Genetics 155: 1439–1447. [PMC free article] [PubMed]
  • Whitkus, R., 1998. Genetics of adaptive radiation in Hawaiian and Cook Islands species of Tetramolopium (Asteraceae). II. Genetic linkage map and its implications for interspecific breeding barriers. Genetics 150: 1209–1216. [PMC free article] [PubMed]
  • Wright, S., B. Lauga and D. Charlesworth, 2003. Rates and patterns of molecular evolution in inbred and outbred Arabidopsis. Mol. Biol. Evol. 19: 1407–1420. [PubMed]
  • Yang, Y.-W., K.-N. Lai, P.-Y. Tai and W.-H. Li, 1999. Rates of nucleotide substitution in angiosperm mitochondrial DNA sequences and dates of divergence between Brassica and other angiosperm lineages. J. Mol. Evol. 48: 597–604. [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...