Conserved flanking microsatellite sequences (ReFS) differentiate between Lepidoptera species,and provide insight into microsatellite evolution

High rates of mutation and homoplasy mean that microsatellites generally are not considered to be useful molecular markers for inferring systematic relationships between species. However, an earlier pilot study suggested that conserved flanking microsatellite sequences, also known as repetitive flanking sequences (ReFS), may form a basis for a dominant marker that can differentiate between species of Lepidoptera. We present data that demonstrate that ReFS are quick and easy to use, and generate highly repeatable banding patterns from a range of Lepidoptera species. Sequence data from a subset of ReFS‐amplified bands revealed microsatellite families with flanking sequences that are more conserved within than among species: this is probably attributable to recombination‐mediated events, transposition of mobile elements or a combination of the two. Our data support the use of ReFS as dominant interspecific molecular markers, and add to the growing literature on the evolution of microsatellites in Lepidoptera.     

High similarity between flanking regions of different microsatellites detected within each of two species of Lepidoptera: Parnassius apollo and Euphydryas aurinia

Microsatellite flanking regions have been compared in two butterfly species. Several microsatellite flanking regions showed high similarity to one another among different microsatellites within a same species, but very few similarities were found between species. This can be the consequence of either duplication/multiplication events involving large regions containing microsatellites or of microsatellites imbedded in minisatellite regions. The multiplication of microsatellites might also be linked to mobile elements. Furthermore, crossing over between nonhomologous microsatellites can lead to the exchange of the flanking regions between microsatellites. The same phenomenon was observed in both studied butterfly species but not in Aphis fabae (Hemiptera), which was screened at the same time using the same protocol. These findings might explain, at least partially, why microsatellite isolation in Lepidoptera has been relatively unsuccessful so far.     

Microsatellite evolution: polarity of substitutions within repeats and neutrality of flanking sequences.

Though extensively used in a variety of disciplines, the evolutionary pattern of microsatellite sequences is still unclear. We addressed several questions relating to microsatellite evolution by analysing historically accumulated mutation events in a large set of artiodactyl (CA)n repeats, through sequence analysis of orthologous bovine and ovine loci. The substitution rate in microsatellite flanking sequences was not different from that in intron sequences, suggesting that if intron sequences in general are selectively neutral, sequences close to microsatellites are similarly so. This observation thus does not support the idea that successful heterologous amplification of microsatellites across distantly related taxa would be due to flanking sequences generally being under some form of selection. Interestingly, the substitution rate at the first nucleotide positions flanking repeats was significantly higher than in sequences further away. Moreover, the substitution rate in repeat units in the very end of microsatellites was significantly higher than that in the middle of repeat regions. Together these observations suggest a relative instability close to the boundary between repetitive and unique sequences. We present three models that potentially could explain such a feature, all involving inefficiency of mismatch repair systems.     

A test for concordance between the multilocus genealogies of genes and microsatellites in the pathogenic fungus Coccidioides immitis

Uncovering the correct phylogeny of closely related species requires analysis of multiple gene genealogies or, alternatively, genealogies inferred from the multiple alleles found at highly polymorphic loci, such as microsatellites. However, a concern in using microsatellites is that constraints on allele sizes may occur, resulting in homoplasious distributions of alleles, leading to incorrect phylogenies. Seven microsatellites from the pathogenic fungus Coccidioides immitis were sequenced for 20 clinical isolates chosen to represent the known genetic diversity of the pathogen. An organismal phylogeny for C. immitis was inferred from microsatellite-flanking sequence polymorphisms and other restriction fragment length polymorphism-containing loci. Two microsatellite genetic distances were then used to determine phylogenies for C. immitis, and the trees found by these three methods were compared. Congruence between the organismal and microsatellite phylogenies occurred when microsatellite distances were based on simple allele frequency data. However, complex mutation events at some loci made distances based on stepwise mutation models unreliable. Estimates of times of divergence for the two species of C. immitis based on microsatellites were significantly lower than those calculated from flanking sequence, most likely due to constraints on microsatellite allele sizes. Flanking-sequence insertions/deletions significantly decreased the accuracy of genealogical information inferred from microsatellite loci and caused interspecific length homoplasies at one of the seven loci. Our analysis shows that microsatellites are useful phylogenetic markers, although care should be taken to choose loci with appropriate flanking sequences when they are intended for use in evolutionary studies.     

Frequent non-reciprocal exchange in microsatellite-containing-DNA-regions of vertebrates

Microsatellites are DNA-fragments containing short repetitive motifs with 2–10 bp. They are highly variable in most species and distributed throughout the whole genome. It is broadly accepted that their high degree of variability is closely associated with mispairing of DNA-strands during the replication phase, termed slippage, although recombination is also observed. The aim of this study is to demonstrate evidence that non-reciprocal recombination processes changing the total genomic structure are common in microsatellites and flanking regions. We sequenced DNA fragments from birds in which microsatellites are located, and analyzed the structure of the microsatellites and their flanking regions. Additionally, other data and those from literature of three microsatellite regions of primates coding for the Ataxin-2, the Huntingtin and the TATA-box binding protein were analyzed. The structures of seven avian and three primate microsatellites support the hypothesis that non-reciprocal recombination is a common process that may also contribute considerably to the variation at microsatellite loci. We conclude that results of population genetic studies that are analyzed statistically with methods based on stepwise mutation models should be interpreted with caution if no detailed information on the allelic variation of microsatellites is available.     

Survey and analysis of microsatellites in the silkworm, Bombyx mori: frequency, distribution, mutations, marker potential and their conservation in heterologous species

We studied microsatellite frequency and distribution in 21.76-Mb random genomic sequences, 0.67-Mb BAC sequences from the Z chromosome, and 6.3-Mb EST sequences of Bombyx mori. We mined microsatellites of >/=15 bases of mononucleotide repeats and >/=5 repeat units of other classes of repeats. We estimated that microsatellites account for 0.31% of the genome of B. mori. Microsatellite tracts of A, AT, and ATT were the most abundant whereas their number drastically decreased as the length of the repeat motif increased. In general, tri- and hexanucleotide repeats were overrepresented in the transcribed sequences except TAA, GTA, and TGA, which were in excess in genomic sequences. The Z chromosome sequences contained shorter repeat types than the rest of the chromosomes in addition to a higher abundance of AT-rich repeats. Our results showed that base composition of the flanking sequence has an influence on the origin and evolution of microsatellites. Transitions/transversions were high in microsatellites of ESTs, whereas the genomic sequence had an equal number of substitutions and indels. The average heterozygosity value for 23 polymorphic microsatellite loci surveyed in 13 diverse silkmoth strains having 2-14 alleles was 0.54. Only 36 (18.2%) of 198 microsatellite loci were polymorphic between the two divergent silkworm populations and 10 (5%) loci revealed null alleles. The microsatellite map generated using these polymorphic markers resulted in 8 linkage groups. B. mori microsatellite loci were the most conserved in its immediate ancestor, B. mandarina, followed by the wild saturniid silkmoth, Antheraea assama.     

Microsatellite DNA markers for rice chromosomes

We found 369 complete microsatellites, of which (CGG/GCC)_n was the most frequent, in 11 798 rice sequences in the database. Of these microsatellites, 35 out of 45 could be successfully converted into microsatellite DNA markers using sequence information in their flanking regions. Thus, the time and labor used to develop new microsatellite DNA markers could be saved by using these published sequences. Twenty eight polymorphic markers between Asominori (japonica) and IR24 (indica) have been correctly mapped on the rice genome and microsatellites appear to be randomly distributed in the rice chromosomes. Integration of these markers with the published microsatellite DNA markers showed that about 35% of the rice chromosomes were covered by the 56 microsatellite DNA markers. These microsatellites were hypervariable and were easily to assay by PCR; they were distributed to all chromosomes and therefore, one can easily select plants carrying desired chromosome regions using these microsatellite DNA markers. Thus, microsatellite maps should aid the development of new breeds of rice saving time, labor, and money.     

Sequence-tagged site markers for microsatellites: Simplified technique for rapidly obtaining flanking sequences

A sequencing strategy is described for the rapid recovery of DNA sequences flanking repeat sequences of microsatellites in plant nuclear genomes. Primers that represent a perfect microsatellite repeat and end in a 3′ degenerate base have been used to sequence directly from microsatellite repeats in one direction. The procedure allows the design of one flanking primer that is then used to sequence back through the repeat to define the microsatellite site precisely and also provides for the design of the second flanking primer. The strategy is versatile with various repeat sizes and different categories of microsatellites; perfect, imperfect, and compound were found to be suitable templates for analysis.     

Silene tatarica microsatellites are frequently located in repetitive DNA

The genomic distribution of microsatellites can be explained by DNA slippage, slippage like processes and base substitutions. Nevertheless, microsatellites are also frequently associated with repetitive DNA, raising the question of the relative contributions of these processes to microsatellite genesis. We show that in Silene tatarica about 50% of the microsatellites isolated by an enrichment cloning protocol are associated with repetitive DNA. Based on the flanking sequences, we distinguished seven different classes of repetitive DNA. PCR primers designed for the flanking sequences of an individual clone amplified a heterogeneous family of repetitive DNA. Despite considerable variation in the flanking sequence (pi = 0.108), the microsatellite repeats did not show any evidence for decay. Rather, we observed the emergence of a new repeat type that probably arose by mutation and was spread by replication slippage. In fact, a complete repeat type switch could be observed among the analysed clones. We propose that the analysis of microsatellite sequences embedded in repetitive DNA provides a hitherto largely unexplored tool to study microsatellite evolution.     

Causes of Size Homoplasy Among Chloroplast Microsatellites in Closely Related <Emphasis Type="Italic">Clusia</Emphasis> Species

Chloroplast DNA sequences and microsatellites are useful tools for phylogenetic as well as population genetic analyses of plants. Chloroplast microsatellites tend to be less variable than nuclear microsatellites and therefore they may not be as powerful as nuclear microsatellites for within-species population analysis. However, chloroplast microsatellites may be useful for phylogenetic analysis between closely related taxa when more conventional loci, such as ITS or chloroplast sequence data, are not variable enough to resolve phylogenetic relationships in all clades. To determine the limits of chloroplast microsatellites as tools in phylogenetic analyses, we need to understand their evolution. Thus, we examined and compared phylogenetic relationships of species within the genus Clusia, using both chloroplast sequence data and variation at seven chloroplast microsatellite loci. Neither ITS nor chloroplast sequences were variable enough to resolve relationships within some sections of the genus, yet chloroplast microsatellite loci were too variable to provide any useful phylogenetic information. Size homoplasy was apparent, caused by base substitutions within the microsatellite, base substitutions in the flanking regions, indels in the flanking regions, multiple microsatellites within a fragment, and forward/reverse mutations of repeat length resulting in microsatellites of identical base composition that were not identical by descent.     

Conservation and evolution of microsatellite loci in primate taxa

Microsatellites are promising genetic markers for the study of demographic structure and phylogenetic history in populations. However, little information exists on the molecular nature of the repeats and their flanking sequences of a same microsatellite in a large range of species. In this study, we report polymorphism and consensus sequences of eight microsatellite loci using human primers in 20 primate species. The results show size polymorphism in almost all species and microsatellites. These loci are therefore useful markers for population genetic studies between populations of the same species. Insertion/deletion events are frequent in the flanking regions, the majority concerning several contiguous bases. This is in contrast with the more usual single base pair events in non-coding regions. The ranges of allele lengths in non-human primates often show no overlap with that of human, usually due to the deletion/insertion events in the flanking sequences, producing smaller allele lengths rather than smaller numbers of repeats. The use of length of PCR product will bias the inter-species interpretation reducing the number of observable alleles and treating as the same allele very divergent molecular sequences. Caution should be used when employing microsatellites in cross-species comparisons in which the species under study are separated by significant amounts of evolutionary time: in such cases allele comparison cannot be based on lengths alone.     

Putative in silico mapping of DNA sequences to livestock genome maps using SSLP flanking sequences

In this study, an in silico approach was developed to identify homologies existing between livestock microsatellite flanking sequences and GenBank nucleotide sequences. Initially, 1955 bovine, 1570 porcine and 1121 chicken microsatellites were downloaded and the flanking sequences were compared with the nr and dbEST databases of GenBank. A total of 74 bovine, 44 porcine and 37 chicken microsatellite flanking sequences passed our criteria and had at least one significant match to human genomic sequence, genes/expressed sequence tags (ESTs) or both. GenBank annotation and BLAT searches of the UCSC human genome assembly revealed that 38 bovine, 13 porcine and 17 chicken microsatellite flanking sequences were highly similar to known human genes. Map locations were available for 67 bovine, 44 porcine and 21 chicken microsatellite flanking sequences, providing useful links in the comparative maps of humans and livestock. In support of our approach, 112 alignments with both microsatellite and match mapping information were located in the expected chromosomal regions based on previously reported syntenic relationships. The development of this in silico mapping approach has significantly increased the number of genes and EST sequences anchored to the bovine, porcine and chicken genome maps and the number of links in various human-livestock comparative maps.     

Nucleotide sequences flanking dinucleotide microsatellites in the human, mouse and Drosophila genomes

We extracted nucleotide sequences from the EMBL database that flank dinucleotide microsatellites in the long sequenced parts of the human, mouse and drosophila genomes. Comparison of the flanking sequences showed that the microsatellites were mostly connected to the bulk of genomic DNA through conserved, highly non-random and mostly (A+T)-rich sequences having many dozens of nucleotides in length. In many cases, the connectors were mutated versions of the flanked microsatellites whose sequence pattern gradually vanished with the distance from the microsatellite center. Hence many microsatellites have hundreds rather than dozens of nucleotides in length, and their ends are diffuse. In contrast, some microsatellites containing predominantly C and/or G, did not influence their neighborhood at all. These results make us change notions about the microsatellite nature. They also indicate that the microsatellites are the dominant part of eukaryotic genomes.     

Evolutionary dynamics of duplicated microsatellites shared by sex chromosomes

Segmental duplications on sex chromosomes constitute an important proportion of recent duplications (approximately 30%). Among those, the evolution of duplicated noncoding DNA is still poorly investigated. We focus our work on repeated DNA sequences extensively used in population genetics and evolution: microsatellites. Six duplicated (CA), microsatellite loci, located on the homologous region of human sex chromosomes, were studied at the intraspecific level in Homo sapiens and by an orthologous comparison in eight primate species. At the intraspecific level, we evaluated the congruence in paralogous divergence between the flanking sequences of the six microsatellites and the approximately 2.2-kb surrounding sequences and observed that both phylogenies are congruent. At the interspecific level (8 species of primates: 54 individuals), we analyzed the sequence polymorphism and divergence of each orthologous locus for both the flanking sequence and the microsatellite. The results showed a lower divergence of flanking sequences than expected in noncoding DNA and a relative stability of the first nucleotides close to the microsatellite. The location of each CAIII locus in a Low Copy Repeated element containing duplicated VCX/Y genes (approximately 1 kb) suggested that direct or indirect selection could explain these results. Moreover, the substitution rates in the flanking sequences and in the microsatellites were correlated. Thus, the evolutionary dynamics of microsatellites seems closely linked to the variation of spontaneous mutations in the surrounding regions.     

Evidence for widespread convergent evolution around human microsatellites

Microsatellites are a major component of the human genome, and their evolution has been much studied. However, the evolution of microsatellite flanking sequences has received less attention, with reports of both high and low mutation rates and of a tendency for microsatellites to cluster. From the human genome we generated a database of many thousands of (AC)_n flanking sequences within which we searched for common characteristics. Sequences flanking microsatellites of similar length show remarkable levels of convergent evolution, indicating shared mutational biases. These biases extend 25–50 bases either side of the microsatellite and may therefore affect more than 30% of the entire genome. To explore the extent and absolute strength of these effects, we quantified the observed convergence. We also compared homologous human and chimpanzee loci to look for evidence of changes in mutation rate around microsatellites. Most models of DNA sequence evolution assume that mutations are independent and occur randomly. Allowances may be made for sites mutating at different rates and for general mutation biases such as the faster rate of transitions over transversions. Our analysis suggests that these models may be inadequate, in that proximity to even very short microsatellites may alter the rate and distribution of mutations that occur. The elevated local mutation rate combined with sequence convergence, both of which we find evidence for, also provide a possible resolution for the apparently contradictory inferences of mutation rates in microsatellite flanking sequences.     

Heterogeneous Nature and Distribution of Interruptions in Dinucleotides May Indicate the Existence of Biased Substitutions Underlying Microsatellite Evolution

Some aspects of microsatellite evolution, such as the role of base substitutions, are far from being fully understood. To examine the significance of base substitutions underlying the evolution of microsatellites we explored the nature and the distribution of interruptions in dinucleotide repeats from the human genome. The frequencies that we inferred in the repetitive sequences were statistically different from the frequencies observed in other noncoding sequences. Additionally, we detected that the interruptions tended to be towards the ends of the microsatellites and 5'-3' asymmetry. In all the estimates nucleotides forming the same repetitive motif seem to be affected by different base substitution rates in AC and AG. This tendency itself could generate patterning and similarity in flanking sequences and reconcile these phenomena with the high mutation rate found in flanking sequences without invoking convergent evolution. Nevertheless, our data suggest that there is a regional bias in the substitution pattern of microsatellites. The accumulation of random substitutions alone cannot explain the heterogeneity and the asymmetry of interruptions found in this study or the relative frequency of different compound microsatellites in the human genome. Therefore, we cannot rule out the possibility of a mutational bias leading to convergent or parallel evolution in flanking sequences.     

Consensus features of microsatellite distribution: Microsatellite contents are universally correlated with recombination rates and are preferentially depressed by centromeres in multicellular eukaryotic genomes

Microsatellite DNA is highly polymorphic and informative, which makes its distribution pattern and its associations very valuable for marker applications and genomic research in evolution. Using computational and statistical approaches, based on database technology, we have demonstrated that microsatellite content is consistently and significantly 2 to 5 fold lower than the average chromosomal level in the centromeric and pericentromeric regions of the chromosomes of two plant species, Arabidopsis thaliana and Oryza sativa. We conducted a path coefficient analysis to compare the direct effect of microsatellites (from mono-nucleotide through to penta-nucleotide repeats) on recombination rates. The results revealed that tri- and penta-nucleotide microsatellites significantly influence recombination rates. In the human genome, tri-, tetra- and mono-nucleotide microsatellites, in decreasing order, make significant direct contributions to recombination rates, according to DECODE, GENTHON, and MARSHFIELD averages. Path coefficient analysis in rice and human genomes of the impact of di-nucleotide microsatellites of different motifs on recombination rates indicate that motifs with either A or T have an effect, resulting in increased recombination rates for microsatellites with motifs consisting of 50% A or T, such as AG, TC, CA, TG. Conversely, microsatellites with motifs consisting of only A & T or G & C, such as AT, TA, GC or CG, have decreased recombination rates. The extremely low microsatellite content in centromeric and pericentromeric regions, as well as the quantitative association of microsatellite sequences with the recombination rate at the genome level, suggests that purifying selection in genome evolution creates a balance between genomic polymorphisms and the biological function of sequences in a genome.     

Conservation of microsatellite loci within the genus Vitis

Eleven microsatellites isolated from grapevine (Vitis vinifera) were used to study the degree of conservation of these sequences across different Vitis species. Nine microsatellites were newly isolated, the remaining two (VVS2 and VVS5) came from the literature. A preliminary assay on the conservation of priming sites was carried out on 14 non-V. vinifera species, including relevant taxa for breeding. Parthenocissus quinquefolia was added as representative of a related genus. Cross-species amplification was obtained in 94% of the 176 genotype×locus tested combinations. Three microsatellite loci were then cloned and sequenced in ten species. The microsatellite repeat was found present in all cases. The repeat region was often longer in V. vinifera than in the other species. Furthermore the non-source species showed interruptions in the repeat. In spite of these constraints, which could reduce the polymorphism of microsatellites in non-source species, the results demonstrate the possibility of extending the use of microsatellite markers to wild germplasm and inter-specific hybrids. Point mutations have been found in microsatellite flanking regions and these variations have been used to investigate the genetic relationship among taxa. The Neighbor-joining tree that was obtained on the basis of ten nucleotide variations, showed that there is not a clear cut difference between American, Asian and European species and that the actual taxonomy which reflects the geographical distribution of species must most likely be revised. Moreover, in general, nucleotide variations which occur in microsatellite flanking regions provide new molecular tools for investigating the evolution of species. Received: 24 October 1999 / Accepted: 11 November 1999     

Characterization of microsatellite loci in Anopheles flavirostris,the principal malaria vector in the Philippines

Microsatellites were isolated and characterized from Anopheles flavirostris, the principal malaria vector in the Philippines. Fifty of the 150 positive clones sequenced contained mostly dinucleotide microsatellites and only 16 had trinucleotide repeats. We designed primers from the unique sequences flanking 18 microsatellite loci. Of these, 11 loci produced successful amplification and revealed high levels of polymorphism; 86 alleles were detected with allele number ranging from 2 to 16 at each locus. The high allelic variability will make these microsatellite loci very useful for taxonomic and population genetic studies.     

Characterization of new Schistosoma mansoni microsatellite loci in sequences obtained from public DNA databases and microsatellite enriched genomic libraries

In the last decade microsatellites have become one of the most useful genetic markers used in a large number of organisms due to their abundance and high level of polymorphism. Microsatellites have been used for individual identification, paternity tests, forensic studies and population genetics. Data on microsatellite abundance comes preferentially from microsatellite enriched libraries and DNA sequence databases. We have conducted a search in GenBank of more than 16,000 Schistosoma mansoni ESTs and 42,000 BAC sequences. In addition, we obtained 300 sequences from CA and AT microsatellite enriched genomic libraries. The sequences were searched for simple repeats using the RepeatMasker software. Of 16,022 ESTs, we detected 481 (3%) sequences that contained 622 microsatellites (434 perfect, 164 imperfect and 24 compounds). Of the 481 ESTs, 194 were grouped in 63 clusters containing 2 to 15 ESTs per cluster. Polymorphisms were observed in 16 clusters. The 287 remaining ESTs were orphan sequences. Of the 42,017 BAC end sequences, 1,598 (3.8%) contained microsatellites (2,335 perfect, 287 imperfect and 79 compounds). The 1,598 BAC end sequences 80 were grouped into 17 clusters containing 3 to 17 BAC end sequences per cluster. Microsatellites were present in 67 out of 300 sequences from microsatellite enriched libraries (55 perfect, 38 imperfect and 15 compounds). From all of the observed loci 55 were selected for having the longest perfect repeats and flanking regions that allowed the design of primers for PCR amplification. Additionally we describe two new polymorphic microsatellite loci.