Search

Severe alpha-1-antitrypsin deficiency (AATD), phenotype PiZZ, was associated with venous thromboembolism (VTE) in a case-control study.This study aimed to determine the genetic variation in the SERPINA1 gene and a possible thrombotic risk of these variants in a population-based cohort study.The coding sequence of SERPINA1 was analyzed for the Z (rs28929474), S (rs17580), and other qualifying variants in 28,794 subjects without previous VTE (born 1923-1950, 60% women), who participated in the Malmö Diet and Cancer study (1991-1996). Individuals were followed from baseline until the first event of VTE, death, or 2018.Resequencing the coding sequence of SERPINA1 identified 84 variants in the total study population, 21 synonymous, 62 missense, and 1 loss-of-function variant. Kaplan-Meier analysis showed that homozygosity for the Z allele increased the risk of VTE whereas heterozygosity showed no effect. The S (rs17580) variant was not associated with VTE. Thirty-one rare variants were qualifying and included in collapsing analysis using the following selection criteria, loss of function, in frame deletion or non-benign (PolyPhen-2) missense variants with minor allele frequency (MAF)0.1%. Combining the rare qualifying variants with the Z variant showed that carrying two alleles (ZZ or compound heterozygotes) showed increased risk. Cox regression analysis revealed an adjusted hazard ratio of 4.5 (95% confidence interval 2.0-10.0) for combinations of the Z variant and rare qualifying variants. One other variant (rs141620200; MAF = 0.002) showed an increased risk of VTE.The SERPINA1 ZZ genotype and compound heterozygotes for severe AATD are rare but associated with VTE in a population-based Swedish study.

Background The protein C (PC) anticoagulant system has a key role in maintaining hemostatic balance. One missense (Ser219Gly) variant in the PC receptor (PROCR) was associated with venous thromboembolism (VTE) in genome-wide association studies. Objectives This study aimed to determine the thrombotic risk of rare and common PROCR variants in a large population-based cohort of middle-aged and older adults. Methods The exonic sequence of PROCR was analyzed for the Ser219Gly variant and other qualifying variants in 28,794 subjects (born 1923–1950, 60% women) without previous VTE, who participated in the Malmö Diet and Cancer study (1991–1996). Incidence of VTE was followed up until 2018. Qualifying variants were defined as loss-of-function or nonbenign (PolyPhen-2) missense variants with minor allele frequencies (MAFs) Results Re-sequencing identified 36 PROCR variants in the study population (26,210 non-VTE exomes and 2,584 VTE exomes), 11 synonymous, 22 missense, and three loss-of-function variants. Kaplan–Meier analysis of the known Ser219Gly variant (rs867186) showed that homozygosity for this variant increased the risk of disease, whereas heterozygosity showed no effect. Cox multivariate regression analysis revealed an adjusted hazard ratio (HR) of 1.5 (95% confidence interval [CI]: 1.1–2.0). Fifteen rare variants were classified as qualifying and were included in collapsing analysis (burden test and SKAT-O). They did not contribute to risk. However, a Arg113Cys missense variant (rs146420040; MAF = 0.004) showed an increased VTE risk (HR = 1.3; 95% CI: 1.0–1.9). Conclusion Homozygosity for the Ser219Gly variant and a previously identified functional PROCR variant (Arg113Cys) was associated with VTE. Other variants did not contribute to VTE.

Genome-wide association studies (GWASs) have identified genes that affect plasma von Willebrand factor (VWF) levels. ABO showed a strong effect, whereas smaller effects were seen for VWF, STXBP5, STAB2, SCARA5, STX2, TC2N, and CLEC4M. This study screened comprehensively for both common and rare variants in these eight genes by resequencing their coding sequences in 104 Swedish von Willebrand disease (VWD) patients. The common variants previously associated with the VWF level were all accumulated in the VWD patients compared to three control populations. The strongest effect was detected for blood group O coded for by the ABO gene (71 vs. 38% of genotypes). The other seven VWF level associated alleles were enriched in the VWD population compared to control populations, but the differences were small and not significant. The sequencing detected a total of 146 variants in the eight genes. Excluding 70 variants in VWF, 76 variants remained. Of the 76 variants, 54 had allele frequencies > 0.5% and have therefore been investigated for their association with the VWF level in previous GWAS. The remaining 22 variants with frequencies

INTRODUCTION: Venous thromboembolism (VTE) is a complex disease that aggregates in families. Both acquired and genetic risk factors are important. Proper recognition and management of high-risk individuals are important. AREAS COVERED: The genetic risk factors for VTE, the clinical consequences, and future perspectives are summarized. Classical thrombophilia i.e. factor V Leiden (rs6025), the prothrombin G20210A mutation (rs1799963), deficiencies of antithrombin, protein C, and protein S and the recent findings from genome wide association studies (GWAS), transcriptome-wide association studies (TWAS), genetic risk score (GRS), VTE candidate genes, expression studies, animal studies, studies using next generation sequencing, pathway analysis, and clinical implications are discussed. EXPERT OPINION: Screening of inherited thrombophilia should be performed in special cases. Identification of strong risk variants might affect the management. The increasing number of genetic risk variants is likely to change management of VTE.

BACKGROUND: Inversions involving intron 22 (Inv22) of F8 are detected in approximately 45% of all severe hemophilia A patients. Diagnosis is complicated by the large size of the ~9.5 kb int22h repeated sequence which generates the inversions. Methods such as long-range PCR and inverse-shifting PCR are currently used diagnostically, but suffer from low PCR efficiencies and are difficult to standardize.OBJECTIVES: To design and validate a sensitive and robust assay for the detection of F8 int22h inversions.METHODS: Digital droplet PCR using mile-post assays was used to investigate archival DNA samples.RESULTS: The detection of linkage as a function of physical distance between loci was investigated using an anchor locus and mile-post loci located at 1, 6, 12 and 15 kb distances from the anchor locus. The proportion of linked molecules decreased with increasing distance between loci and showed 30-40% linked molecules for loci 12-15 kb apart. Mile-post assays specific for wild type and Inv22 type 1 and 2 chromosomes were then designed and optimized. All three assays showed high specificities and sensitivities, with coefficients of variation < 5% for all assays. Analysis of 106 patients and 20 carrier mothers showed complete concordance with previously known mutation status. The analysis demonstrated the robustness of the assays versus input DNA concentration (6 ng and higher) and level of fragmentation.CONCLUSIONS: Digital droplet PCR and mile-post assays can be used to detect F8 int22h inversions. The assay systems are technically simple to perform, highly efficient and robust. (Less)

The protein C anticoagulant system plays a key role in maintaining the hemostatic balance. Although several studies have identified thrombomodulin gene (THBD) variants among venous thromboembolism (VTE) patients, the role of THBD in relation to VTE in humans remains to be clarified.This study aimed to determine the thrombotic risk of rare and common THBD variants in a large population-based cohort of middle-aged and older adults.The exome sequence of THBD was analyzed for qualifying variants in 28,794 subjects (born 1923-1950, 60% women), who participated in the Malmö Diet and Cancer study (1991-1996). Patients were followed from baseline until the first event of VTE, death, or 2018. Qualifying variants were defined as loss-of-function or non-benign (PolyPhen-2) missense variants with minor allele frequency0.1%.The single common coding variant rs1042579 was not associated with incident VTE. Sixteen rare variants were classified as qualifying and included in collapsing analysis. Seven individuals with VTE compared to 24 individuals without VTE carried one qualifying variant. Cox multivariate regression analysis adjusted for age, sex, body mass index, systolic blood pressure, smoking and alcohol consumption, rs6025, rs1799963, and the top two eigenvectors from a principal components analysis showed a hazard ratio of 3.0 (95% confidence interval 1.4-6.3) for the rare qualifying variants. The distributions of qualifying variants in THBD showed a difference for individuals with and without incident VTE indicating a possible position effect.Rare qualifying THBD variants were associated with VTE, suggesting that rare variants in THBD contribute to development of VTE.

Background The heritability of venous thromboembolism (VTE) is only partially explained by variants in 17 previously VTE-associated genes. Objective This article screens for additional rare variants in the 17 genes and investigates the relative contributions of pro- and anticoagulant genes to VTE. Patients and Methods Ninety-six VTE patients from the population-based Malmö Thrombophilia Study were analysed using an AmpliSeq strategy and Ion Torrent sequencing and the variant data were compared with data from public databases. Results A total of 102 non-synonymous and 76 synonymous variants were identified. Forty-six non-synonymous variants were present in the human gene mutation database. Anticoagulant and procoagulant genes showed 14 and 22 rare non-synonymous variants, respectively. Individual patients showed varying numbers of risk factors; 13 patients had non-synonymous mutations in SERPINC1, PROC and PROS1 genes and 42 had factor V Leiden or prothrombin mutations generating a total of 47 patients with at least one of these risk factors. Ten common VTE-associated variants showed low level enrichments and no correlation to the other risk factors. The enrichment of previously identified risk factors was similar to previous studies. Determination of the nsyn/syn ratio (number of non-synonymous variants per non-synonymous site, nsyn, to the number of synonymous variants per synonymous site, syn) showed, as expected in patients, an increase of non-synonymous relative to synonymous anticoagulant variants compared with controls (nsyn/syn, 0.95 vs. 0.68). In contrast, non-synonymous procoagulant variants (nsyn/syn, 0.31 vs. 0.63) showed a decrease. We suggest that the deficit of non-synonymous variants in procoagulant genes is a novel mechanism contributing to VTE.

Background F8 int1h inversions (Inv1) are detected in 1-2% of severe hemophilia A (HA) patients. Long-range polymerase chain reaction (PCR) and inverse-shifting PCR have been used to diagnose these inversions. Objectives To design and validate a sensitive and robust assay for detection of F8 Inv1 inversions. Methods Archival DNA samples were investigated using mile-post assays and droplet digital PCR. Results Mile-post assays for Inv1 showing high specificities and sensitivities were designed and optimized. Analysis of four patients, two carrier mothers and 40 healthy controls showed concordance with known mutation status with one exception. One patient had a duplication involving exons 2-22 of the F8 gene instead of an Inv1 mutation. DNA mixtures with different proportions of wild type and Inv1 DNA correlated well with the observed relative linkage for both wild type and Inv1 assays and estimated the limit of detection of these assays to 2% of the rare chromosome. Conclusions The mile-post strategy has several inherent control systems. The absolute counting of target molecules by both assays enables determination of template quantity, detection of copy number variants and rare variants occurring in primer and probe annealing sites and estimation of DNA integrity through the observed linkage. The presented Inv1 mile-post analysis offers sensitive and robust detection and quantification of the F8 int1h inversions and other rearrangements involving intron 1 in patients and their mothers.

Background: The occurrence of mosaicism in hemophilia A (HA) has been investigated in several studies using different detection methods. Objectives: To characterize and compare the ability of AmpliSeq/Ion Torrent sequencing and droplet digital polymerase chain reaction (ddPCR) for mosaic detection in HA. Methods: Ion Torrent sequencing and ddPCR were used to analyze 20 healthy males and 16 mothers of sporadic HA patients. Results: An error-rate map over all coding positions and all positions reported as mutated in the F8-specific mutation database was produced. The sequencing produced a mean read depth of >1500X where >97% of positions were covered by >100 reads. Higher error frequencies were observed in positions with A or T as reference allele and in positions surrounded on both sides with C or G. Seventeen of 9319 positions had a mean substitution error frequency >1%. The ability to identify low-level mosaicism was determined primarily by read depth and error rate of each specific position. Limit of detection (LOD) was 1% require repeated testing and mononucleotide repeats with more than four repeat units need an alternative analysis strategy. Mosaicism was detected in 1 of 16 mothers and confirmed using ddPCR. Conclusions: Deep sequencing using an AmpliSeq/Ion Torrent strategy allows for simultaneous identification of disease-causing mutations in patients and mosaicism in mothers. ddPCR has high sensitivity but is hampered by the need for mutationspecific design.

von Willebrand factor (VWF) level and function are influenced by genetic variation in VWF and several other genes in von Willebrand disease type 1 (VWD1) patients. This study comprehensively screened for VWF variants and investigated the presence of ABO genotypes and common and rare VWF variants in Swedish VWD1 patients. The VWF gene was resequenced using Ion Torrent and Sanger sequencing in 126 index cases historically diagnosed with VWD. Exon 7 of the ABO gene was resequenced using Sanger sequencing. Multiplex ligation-dependent probe amplification analysis was used to investigate for copy number variants. Genotyping of 98 single nucleotide variants allowed allele frequency comparisons with public databases. Seven VWD2 mutations and 36 candidate VWD1 mutations (5 deletions, 4 nonsense, 21 missense, 1 splice, and 5 synonymous mutations) were identified. Nine mutations were found in more than one family and nine VWD1 index cases carried more than one candidate mutation. The T-allele of rs1063857 (c.2385T > C, p.Y795 = ) and blood group O were both frequent findings and contributed to disease in the Swedish VWD1 population. VWD2 mutations were found in 20 and candidate VWD1 mutations in 51 index cases out of 106 (48%). VWF mutations, a VWF haplotype, and blood group O all contributed to explain disease in Swedish VWD1 patients.

Mutations are not identified in ~5% of hemophilia A and 10-35% of type 1 VWD patients. The bleeding tendency also varies among patients carrying the same causative mutation, potentially indicating variants in additional genes modifying the phenotype that cannot be identified by routine single-gene analysis. The F8, F9 and VWF genes were analyzed in parallel using an AmpliSeq strategy and Ion Torrent sequencing. Targeting all exonic positions showed an average read depth of >2000X and coverage close to 100% in 24 male patients with known disease-causing mutations. Discrimination between reference alleles and alternative/indel alleles was adequate at a 25% frequency threshold. In F8, F9 and VWF there was an absolute majority of all reference alleles at allele frequencies >95% and the average alternative allele and indel frequencies never reached above 10% and 15%, respectively. In VWF, 4-5 regions showed lower reference allele frequencies; in two regions covered by the pseudogene close to the 25% cut-off for reference alleles. All known mutations, including indels, gross deletions and substitutions, were identified. Additional VWF variants were identified in three hemophilia patients. The presence of additional mutations in 2 out of 16 (12%) randomly selected hemophilia patients indicates a potential mutational contribution that may affect the disease phenotype and counseling in these patients. Parallel identification of disease-causing mutations in all three genes not only confirms the deficiency, but differentiates phenotypic overlaps and allows for correct genetic counseling.

A previous investigation of all 10 TLR genes for associations with allergic rhinitis (AR) detected a number of significant SNPs in the TLR8 locus. The associations indicated that an accumulation of rare variants could explain the signal. This study therefore searches for rare variants in the TLR8 region and also investigates the reproducibility of previous SNP associations.The TLR8 gene was resequenced in 288 AR patients from Malmö and the data were compared with publically available data. Seven previously AR-associated SNPs from TLR8 were analyzed for AR associations in 422 AR patients and 859 controls from the BAMSE cohort. The associations detected in present and previous studies were compared.Sequencing detected 13 polymorphisms (three promotor and 10 coding) among 288 AR patients. Four of the coding polymorphisms were rare (MAF1%) and three of those were novel. Two coding polymorphisms were benign missense mutations and the rest were synonymous. Comparison with 1000Genomes and Exome Aggregation Consortium data revealed no accumulation of rare variants in the AR cases. The AR association tests made using the BAMSE cohort yielded five P-values0.05. Tests of IgE levels yielded four significant SNP associations to birch pollen. Comparing results between different populations revealed opposing risk alleles, different gender effects, and response to different allergens in the different populations.Rare variants in TLR8 are not associated with AR. Comparison of present and previous association studies reveals contradictory results for common variants. Thus, no associations exist between genetic variation in TLR8 and AR.

von Willebrand factor (VWF) levels in healthy individuals and in patients with type 1 von Willebrand disease (VWD) are influenced by genetic variation in several genes, e.g. VWF, ABO, STXBP5 and CLEC4M. This study aims to screen comprehensively for CLEC4M variants and investigate their association with type 1 VWD in the Swedish population. In order to screen for CLEC4M variants, the CLEC4M gene region was re-sequenced and the polymorphic neck region was genotyped in 106 type 1 VWD patients from unrelated type 1 VWD families. Single nucleotide variants (SNV) and variable number tandem repeat (VNTR) allele and genotype frequencies were then compared with 294 individuals from the 1000Genomes project and 436 Swedish control individuals. Re-sequencing identified a total of 42 SNVs. Rare variants showed no accumulation in type 1 VWD patients and are not thought to contribute substantially to type 1 VWD. The only missense mutation (rs2277998, NP_001138379.1:p.Asp224Asn) had a higher frequency in type 1 VWD patients than in controls (4.9%). The VNTR genotypes 57 and 67 were observed at higher frequencies than expected in type 1 VWD patients (6.4% and 6.2%) and showed an increase in patients compared with controls (7.4% and 3.1%). Strong linkage disequilibrium in the CLEC4M region makes it difficult to distinguish between the effect of the missense mutation and the VNTR genotypes. In conclusion, heterozygous VNTR genotypes 57 and 67 of CLEC4M were highly enriched and are the most likely mechanism through which CLEC4M contributes to disease in the Swedish type 1 VWD population.

Genetic studies of chronic rhinosinusitis (CRS) have identified a total of 53 CRS-associated SNPs that were subsequently evaluated for their reproducibility in a recent study. The rs2873551 SNP in linkage disequilibrium with PARS2 showed the strongest association signal. The present study aims to comprehensively screen for rare variants in PARS2 and evaluate for accumulation of such variants in CRS-patients. Sanger sequencing and long-range PCR were used to screen for rare variants in the putative promoter region and coding sequence of 310 CRS-patients and a total of 21 variants were detected. The mutation spectrum was then compared with data from European populations of the 1000Genomes project (EUR) and the Exome Aggregation Consortium (ExAC). The CRS population showed a significant surplus of low-frequency variants compared with ExAC data. Haplotype analysis of the region showed a significant excess of rare haplotypes in the CRS population compared to the EUR population. Two missense mutations were also genotyped in the 310 CRS patients and 372 CRS-negative controls, but no associations with the disease were found. This is the first re-sequencing study in CRS research and also the first study to show an association of rare variants with the disease.

SummaryProtein S deficiency is a dominantly inherited disorder that results from mutations in the PROS1 gene. Previous sequencing of the gene failed to detect mutations in eight out of 18 investigated Swedish families, whereas segregation analyses detected large deletions in three out of the eight families. The present study investigates more thoroughly for the presence of deletions but also for other types of rearrangements. FISH analysis confirmed the existence of the three previously identified large deletions, but failed to identify any other type of rearrangement among the eight analysed families. MLPA analysis of the PROS1 gene revealed two smaller deletions covering two and four exons, respectively. Thus, deletions could be found in five out of eight families where no point mutations could be found despite sequencing of the gene. Twelve additional, not previously analysed, families were subsequently analysed using MLPA. The analysis identified two smaller deletions (3 and 4 exons). Including all PS-deficient families, i.e. also the 10 families where sequencing found a causative point mutation, deletions were identified in seven out of 30 PS-deficient families. A strategy of sequencing followed by MLPA analysis in mutation-negative families identified the causative mutation in 15 out of 18 of Swedish PS-deficient families. Most deletions were different as determined by their sizes, locations and flanking haplotypes. FISH (8 families) and MLPA analysis (20 families) failed to identify other types of rearrangements.

The evolutionary history of the common chloroplast (cp) genome of the allotetraploid Arabidopsis suecica and its maternal parent A. thaliana was investigated by sequencing 50 fragments of cpDNA, resulting in 98 polymorphic sites. The variation in the A. suecica sample was small, in contrast to that of the A. thaliana sample. The time to the most recent common ancestor (T(MRCA)) of the A. suecica cp genome alone was estimated to be about one 37th of the T(MRCA) of both the A. thaliana and A. suecica cp genomes. This corresponds to A. suecica having a MRCA between 10 000 and 50 000 years ago, suggesting that the entire species originated during, or before, this period of time, although the estimates are sensitive to assumptions made about population size and mutation rate. The data was also consistent with the hypothesis of A. suecica being of single origin. Isolation-by-distance and population structure in A. thaliana depended upon the geographical scale analysed; isolation-by-distance was found to be weak on the global scale but locally pronounced. Within the genealogical cp tree of A. thaliana, there were indications that the root of the A. suecica species is located among accessions of A. thaliana that come primarily from central Europe. Selective neutrality of the cp genome could not be rejected, despite the fact that it contains several completely linked protein-coding genes.

The level of variation and the mutation rate were investigated in an empirical study of 244 chloroplast microsatellites in 15 accessions of Arabidopsis thaliana. In contrast to SNP variation, microsatellite variation in the chloroplast was found to be common, although less common than microsatellite variation in the nucleus. No microsatellite variation was found in coding regions of the chloroplast. To evaluate different models of microsatellite evolution as possible explanations for the observed pattern of variation, the length distribution of microsatellites in the published DNA sequence of the A. thaliana chloroplast was subsequently used. By combining information from these two analyses we found that the mode of evolution of the chloroplast mononucleotide microsatellites was best described by a linear relation between repeat length and mutation rate, when the repeat lengths exceeded about 7 bp. This model can readily predict the variation observed in non-coding chloroplast DNA. It was found that the number of uninterrupted repeat units had a large impact on the level of chloroplast microsatellite variation. No other factors investigated--such as the position of a locus within the chromosome, or imperfect repeats--appeared to affect the variability of chloroplast microsatellites. By fitting the slippage models to the Genbank sequence of chromosome 1, we show that the difference between microsatellite variation in the nucleus and the chloroplast is largely due to differences in slippage rate.

The breeding system of Arabidopsis suecica was investigated through genetic analysis of microsatellite segregation patterns in five controlled crosses as well as in 16 single-mother families collected in the wild. Analysis of single and two-locus segregations in the F2 generation following a cross clearly shows that A. suecica is reproduces sexually. The single-mother families show a high level of homozygosity corroborating earlier results indicating a high level of inbreeding. The high level of individual homozygosity is due both to a high level of selfing and to the underlying population structure.

RAPD (random amplified polymorphic DNA) is a multiplex marker system that conventionally uses single-primer PCR to amplify random DNA fragments. Because of its multiplex nature, it is frequently used in bulked segregant analysis (BSA). In view of the very large numbers of markers BSA often requires, we investigated the use of mixtures of primers as a method of increasing the number of markers available. Theoretically, if a single-primer reaction produces x bands on average, an unrestrained PCR process using a primers should produce xa2 bands. Initially, we investigated mixtures containing from one to five primers. The average number of products increased slightly from the single-primer to the multiple-primer case, whereas it was rather constant for the different multi-primer combinations. This deviation from the theoretical expectations, which we attribute to the effects of competition, shows mixtures of more than two primers to be inefficient. The properties of two-primer mixtures in which the proportions of the two primers were varied were also investigated. The intensities of most of the products were influenced by the proportions of the primers used to create the mixture. A good fit was obtained to a model in which the average competitive ability of a band is directly proportional to the probability of randomly obtaining the band-producing primer combination from the pool of primers. Using two-primer mixtures, a(a-1)/2 different two-primer mixtures can be produced. A comparison of different schemes for constructing the two-primer mixtures indicates that the degree of resampling is similar for all schemes. In conclusion, the use of two-primer mixtures is a simple but very powerful strategy in BSA as it can generate an extremely large number of markers.

More than 1100 mutations that cause hemophilia B (HB) have been identified. At the same time, specific F9 mutations are present at high frequencies in certain populations, which raise questions about the origin of HB mutations.To describe the mutation spectrum of all HB families in Sweden and investigate if mutations appearing in several families are due to independent recurrent mutations (RMs) or to a common mutation event (i.e. are identical by descent (IBD)).The registered Swedish HB population consists of patients from 86 families. Mutations were identified by resequencing and identical haplotypes were defined using 74 markers and a control population of 285 individuals. The ages of IBD mutations were estimated using ESTIAGE.Out of 77 presumably unrelated patients with substitution mutations, 47 patients (61%) had mutations in common with other patients. Haplotyping of the 47 patients showed that 24 patients had IBD mutations (51%) with estimated ages of between two and 23 generations. A majority of these patients had mild disease. Eight of the 15 mutations observed in more than one family were CT transitions in CpG sites and all eight were RMs.The association of IBD mutations with a mild phenotype is similar to what has been previously observed in hemophilia A. Noteworthy features of the mutations that are common to more than one family are the equal proportions of patients with RM and IBD mutations and the correlation between the occurrence of RMs and CT transitions at CpG sites.

A coalescent-based method was used to investigate the origins of the allotetraploid Arabidopsis suecica, using 52 nuclear microsatellite loci typed in eight individuals of A. suecica and 14 individuals of its maternal parent Arabidopsis thaliana, and four short fragments of genomic DNA sequenced in a sample of four individuals of A. suecica and in both its parental species A. thaliana and Arabidopsis arenosa. All loci were variable in A. thaliana but only 24 of the 52 microsatellite loci and none of the four sequence fragments were variable in A. suecica. We explore a number of possible evolutionary scenarios for A. suecica and conclude that it is likely that A. suecica has a recent, unique origin between 12,000 and 300,000 years ago. The time estimates depend strongly on what is assumed about population growth and rates of mutation. When combined with what is known about the history of glaciations, our results suggest that A. suecica originated south of its present distribution in Sweden and Finland and then migrated north, perhaps in the wake of the retreating ice.

Random amplified polymorphic DNA (RAPD) markers were used to estimate the level of genetic variation in Swedish accessions of the allopolyploid Arabidopsis suecica and its parental species A. thaliana and A. arenosa. The results showed clear differences among the three species with respect to the level of variation. A. arenosa was highly variable, A. thaliana showed a moderate level of variation whereas A. suecica was much less variable than the two other species. An extended analysis covering 19 Swedish populations of A. suecica corroborated the low level of variation in this species, yet 16 unique phenotypes were observed. No isolation by distance was observed. When the genetic variation was partitioned among and within populations of A. suecica, the results showed that the majority of the variation (81 %) occurred among populations. This result is interpreted as a strong indication that A. suecica is autogamous in nature.