Heredity (2003) 90, 150–156 & 2003 Nature Publishing Group All rights reserved 0018-067X/03 $25.00 www.nature.com/hdy Genetic effects on microsatellite diversity in wild emmer wheat (Triticum dicoccoides) at the Yehudiyya microsite, Israel Y-C Li1,3, T Fahima1,MSRo¨der2, VM Kirzhner1, A Beiles1, AB Korol1 and E Nevo1 1Institute of Evolution, University of Haifa, Mount Carmel, Haifa 31905, Israel; 2Institute for Plant Genetics and Crop Plant Research, Corrensstrasse 3, 06466 Gatersleben, Germany This study investigated allele size constraints and clustering, diversity. Genome B appeared to have a larger average and genetic effects on microsatellite (simple sequence repeat number (ARN), but lower variance in repeat number 2 repeat, SSR) diversity at 28 loci comprising seven types of (sARN), and smaller number of alleles per locus than genome tandem repeated dinucleotide motifs in a natural population A. SSRs with compound motifs showed larger ARN than of wild emmer wheat, Triticum dicoccoides, from a shade vs those with perfect motifs. The effects of replication slippage sun microsite in Yehudiyya, northeast of the Sea of Galilee, and recombinational effects (eg, unequal crossing over) on Israel. It was found that allele distribution at SSR loci is SSR diversity varied with SSR motifs. Ecological stresses clustered and constrained with lower or higher boundary. (sun vs shade) may affect mutational mechanisms, influen- This may imply that SSR have functional significance and cing the level of SSR diversity by both processes. natural constraints. Genetic factors, involving genome, Heredity (2003) 90, 150–156. doi:10.1038/sj.hdy.6800190 chromosome, motif, and locus significantly affected SSR Keywords: SSR variation; allelic cluster; genome effect; mutational mechanism; Triticum dicoccoides; wheat’s progenitor Introduction than expected by chance) in allele size distributions among various allele series of the same locus suggests Microsatellites, or simple sequence repeats (SSR), are the existence of evolutionary constraints on that locus ubiquitously interspersed in eukaryotic genomes (Tautz (Lehmann et al, 1996). Two alternative forces for such and Renz, 1984; Kashi et al, 1997; Kashi and Soller, 1999; nonrandom distribution of allele-size frequency have Li et al, 2002a). SSRs are among the fastest-evolving DNA been proposed: biased mutation and/or selection sequences with high mutation rates, 10À2–10À3 per locus acting on allele size (Garza et al, 1995; Dermitzakis et al, per gamete per generation (Weber and Wong, 1993), 1998). which leads to their high polymorphism in terms of The present study demonstrated constraints and repeat number. It has been suggested that replication clustering of allele size distribution in a natural popula- slippage, sister-chromatid exchanges, unequal crossing tion of wild emmer wheat, Triticum dicoccoides, and over, and gene conversion may cause SSR diversity correlation between SSR diversity and repeat length, (Tautz and Renz, 1984). Among these mutational locus chromosomal location, and genetic effects on mechanisms, replication slippage seems to play a major dinucleotide SSR diversity in a natural population of role in producing new alleles at SSR loci (Levinson and wild emmer wheat from two neighboring (a few meters Gutman, 1987; Wolff et al, 1991; Innan et al, 1997; Stephan apart) and contrasting microclimatic niches, sun, and and Kim, 1998). However, these suggestions need more shade. The microclimatic effect on SSR divergence and critical analysis across species and populations. diversity are described in a complementary paper (Li The distribution of allele sizes in SSR loci, seems to be et al, 2002b). nonrandom. For instance, alleles with long repeats were found to be more mutable than a bulk of the Ccon70 with short repeats (Crozier et al, 1999). Likewise, bimodal Materials and methods distribution of allele size was revealed at many SSR loci Wild emmer wheat, Triticum dicoccoides (Nevo et al, 2002) in some species including human (Rubinsztein et al, is the tetraploid and predominantly self-pollinated 1995), Mimulus guttatus (Awadalla and Ritland, 1997), progenitor of cultivated wheat (Zohary, 1970). This Arabidopsis (Innan et al, 1997), and the fish Sparus aurata tetraploid species contains two genomes and 28 chromo- (Dermitzakis et al, 1998). The excessive similarity (more somes (2n ¼ 4x ¼ 28, genome AABB). The plant materials used in this analysis are described in detail in Li et al (2002b). Correspondence: E Nevo, Institute of Evolution, University of Haifa, Mt Carmel, Haifa 31905, Israel. E-mail: [email protected] A total of 28 dinucleotide SSR DNA markers (one for 3Current address: Department of Plant Sciences, The University of each chromosomal arm) were chosen for the analysis. Arizona, Tucson, AZ 857-19, USA. The SSR primers used in this study were described by Genetic effects on SSR diversity in wild emmer Y-C Li et al 151 Table 1 SSR motif and chromosomal locations Plaschke et al (1995) and Ro¨der et al (1995, 1998). Table 1 presents the repetitive motif, locus location, and distance Marker Motif Chromosomal from the centromere (D) in bread wheat, T. aestivum. The procedure used to detect SSR polymorphism followed Locationa Distanceb Plaschke et al (1995) and Fahima et al (1998). Fragment sizes were calculated using the Fragment Manager c GWM18 (CA)nGA(TA)k 1BS 5.7 (Pharmacia) computer program by comparing with GWM60 (CA)n 7AS 52.3 internal size standards, which were added to each lane GWM95 (AC)n 2AS 10.8 in the loading buffer. GWM99 (AC)n 1AL 95.3 GWM120 (CT)n(CA)k 2BL 23.7 Analysis of variance (ANOVA) was used to analyze GWM124 (CT)n(GT)k 1BL 82.2 genetic effects on SSR diversity. Multiple regression was GWM136 (CT)n 1AS 38.8 used to measure indirectly contributions of mutational GWM162 (CA)nAA(CA)k 3AL 91.0 mechanisms to SSR diversity in different repeat motifs. GWM169 (GA)n 6AL 49.6 The statistical analyses were performed using the GWM186 (GA)n 5AL 38.4 STATISTICA program (Statsoft, 1996). GWM218 (CT)n 3AS 30.4 GWM219 (GA)n 6BL 40.5 GWM251 (CA)n 4BL 28.9 Results GWM294 (GA)nTA(GA)k 2AL 16.1 GWM332 (GA)n 7AL 53.0 GWM340 (GA)n 3BL 131.4 SSR allele distribution with size clusters and constraints GWM361 (GA)n 6BS 11.4 The distribution of alleles at locus GWM99 showed two GWM368 (AT)n 4BS 8.7 clusters, one with small repeat numbers (7,8), the other GWM389 (CT)n(GT)k 3BS 118.4 with larger repeat numbers (21–24), with a considerable GWM408 (CA) (TA)(CA) (TA) 5BL 73.8 n k m gap of 12 repeats of (AC) between them (Figure 1). GWM415 (GA)n 5AS 25.4 GWM429 (CT)n 2BS 29.0 Similar patterns were observed at loci GWM60, GWM459 (GA)n 6AS 62.5 GWM537, and GWM577 (Figure 1). The allele-size GWM537 (CA)n(TA)k 7BS 43.6 distribution seems to have a low boundary at locus GWM540 (CT)n(CC)(CT)k 5BS 9.3 GWM577 and upper limits at locus GWM537 (Figure 1). GWM577 (CA)n(TA)k 7BL 105.0 At loci GWM415 and GWM601, alleles were limited to GWM601 (CT) 4AS 8.1 n repeat numbers 19 and 17, respectively. Allele sizes at GWM637 (CA)n 4AL 40.1 GWM95, GWM120, and GWM332a varied in the ranges aA, B: genome A and B; S, L: short and long arm. bDistance from the of 15–20, 31–36, and 17–18 repeats, respectively. These centromere (D, in cM), which was estimated according to the map of results suggest that some constraints may exist on repeat Ro¨der et al (1998), the distance of GWM 218 also referred to the map number at SSR loci. The last assumption is especially of Peng et al (2000). cm, n, and k symbolize repeat number. supported by allele distribution at GWM537 (Figure 1). Figure 1 Allele distributions with clusters and constraint at loci GWM60, GWM99, GWM537, and GWM577. Heredity Genetic effects on SSR diversity in wild emmer Y-C Li et al 152 Genetic effect on SSR variation then microsatellite loci, located farther from the centro- Analysis of variance (ANOVA) was used to test the meres should have larger diversity, because recombina- effects of genetic factors (genome, chromosome, and tion is suppressed around the centromeres (Gill et al, motif) (Table 2). The results suggested that genome (A vs 1996). In other words, the ARN and locus distance from B) significantly (Po0.05) affected the average repeat the centromere (D) may affect the genetic diversity at SSR number (ARN), with genome B showing larger ARN loci, allele number (NA) and variance in repeat number (30.2) than genome A (25.1). Genome A showed larger (s2), serving as indirect evidence in favor of one of the variance in repeat number (s2 ¼ 27.8) than genome B explanatory models. Forward stepwise regression was (s2 ¼ 19.0). The SSR ARN on chromosome 1 (including used to estimate contributions of ARN, D, and micro- 2 both 1A and 1B) were significantly (F(6,40) ¼ 3.85, Po0.01) niches to the A and s of different motifs (Table 3). The larger (ARN ¼ 42.5) than those of other chromosomes results indicated that ARN were the most important (ranged from 22 to 31). Microsatellites on chromosome 7 factor for (GA)n loci. Mutational mechanisms (repre- showed the highest variance (s2 ¼ 53.5) in repeat number sented indirectly by ARN, D, and D  ARN) could 2 2 (F(6,40) ¼ 2.72, Po0.05) compared with those on other significantly affect s at (GA)n loci (R ¼ 0.827, chromosomes. Compound SSRs, such as (CA)n(TA)k, Po0.00005). Niche could also alone (r ¼ 0.466, Po0.05) (CT)n(GT)k, and (CT)n(CA)k, appeared to have signifi- or through interaction niche  ARN (r ¼À0.538, Po0.05) 2 cantly (F(6,47) ¼ 2.88, Po0.01) larger ARN (433) than affect the s of (GA)n loci.