Cross-genus amplification and characterisation of microsatellite loci in the large-eared free tailed bat , Otomops ( Chiroptera : Molossidae ) from Africa and Madagascar

Primers developed for the Brazilian free-tailed bat, Tadarida brasiliensis, were successfully used to cross-amplify microsatellite loci in two Afro-tropical Otomops species. Seventy one (71) bats from two species were genotyped for two dinucleotide and four tetranucleotide loci, yielding 1 to 15 alleles per locus. For the combined sample, the observed and expected heterozygosities ranged from 0.125 to 1.000 and 0.125 to 0.919, respectively. The polymorphism information content (PIC) values were 0.295 to 0.905 (mean 0.687) for Otomops martiensseni and 0.110 to 0.797 (mean 0.442) for Otomops madagascariensis. Five O. martiensseni loci deviated significantly from Hardy-Weinberg equilibrium. These six loci provide genetic markers that will be useful in investigating the population genetic structure of Afro-Arabian O. martiensseni and Malagasy O. madagascariensis, with potential application to Asian species of Otomops and possibly other genera within the Molossidae.


INTRODUCTION
Microsatellites or short tandem repeats are popular markers in population genetic studies as they show high levels of polymorphism and are useful for estimating parameters such as gene flow, inbreeding, migration rates, population size and kinship (Selkoe and Toonen, 2006;Barker, 2002;Zane et al., 2002.As development of new microsatellite markers is relatively expensive and time-consuming (Zane et al., 2002;Abdelkrim et al., 2009), cross-amplification of microsatellites using primers developed for another species is considered a costeffective and viable option (Barbará et al., 2007).Crossspecies amplification of microsatellites has been utilised in various taxa, including plants (Datta et al., 2010;Eliott et al., 2013), insects (Chen and Dorn, 2010), fish (Dubut et al., 2010) and mammals (Kaňuch et al., 2007;Kretschmer et al., 2009;Sanvito et al., 2013).This method does, however, have limitations since the primers work best for the species for which they were developed.Loci are less likely to amplify successfully as the genetic distance between the original and target species increases, and those which amplify usually exhibit lower levels of polymorphism than in the original species (Primmer et al., 2005).Projects based on cross-species amplification should therefore be preceded by a preliminary study which assesses the ability of candidate primers to amplify suitably variable microsatellites in the target species (Schlötterer, 2000;Scribner and Pearce, 2000).
The Molossidae are one of the less studied families   within the Chiroptera, and phylogenetic and population genetic studies on this family have been based primarily on mitochondrial and nuclear sequence data (for example, Lamb et al., 2011;Ammerman et al., 2012).However, Russell et al. (2005) used microsatellites to study the population genetics of the American species, Tadarida brasiliensis, and recently Naidoo et al. (2013) reported on the utility of the primers of Russell et al. (2005) to cross-amplify polymorphic microsatellites in the molossid species, Chaerephon pumilus sensu lato from south eastern Africa.Afro-tropical members of the Old World genus Otomops, Otomops martiensseni from Africa (including the Arabian Peninsula) and Otomops madagascariensis from Madagascar, have a wide but somewhat sparse distribution throughout the region (Peterson et al., 1995;Simmons, 2005;Lamb et al., 2008).According to the 2008 IUCN (The World Conservation Union) Red List of Threatened Species, O. martiensseni has been classified globally as having a "Near Threatened" status (Mickleburgh et al., 2008).Although species-level phylogenetic and phylogeographic investigations of Otomops have been undertaken (Lamb et al., 2006(Lamb et al., , 2008)), fine-scale genetic investigations within the genus have been limited, leaving many unanswered questions including the number of species and taxonomic status of Afro-tropical individuals.

Locus
Our aim was to test the ability of primers developed to amplify hypervariable nuclear microsatellites in the American genus T. brasiliensis (Molossidae) (Russell et al., 2005) to cross-amplify and reveal polymorphism in two Afro-Malagasy species of the molossid genus Otomops, namely O. martiensseni and O. madagascariensis.If successful, these primers may also be useful for population genetics studies on Asian species of Otomops, and possibly other genera within this pan-tropical bat family.
Consequent studies on gene-tic variation, gene flow and kinship in Otomops may prove useful in the amendment of current legislations used for the protection and conservation of this genus.
Raw allelic data were analysed and called using STRand v.2.4.59 (Toonen and Hughes, 2001;Hughes, 2006) and 1000 randomizations were performed in Micro-Checker v. 2.2.3 (Van Oosterhout et al., 2004) to check the O. martiensseni and O. madagascariensis data separately for null alleles, stuttering and large allele dropout.Additionally, FreeNA software (Chapuis and Estoup, 2007) was used to determine whether null alleles detected in the data were introducing bias in the analyses, where pairwise F ST values were calculated between O. martiensseni and O. madagascariensis with and without the excluding null alleles (ENA) method applied.The ENA method corrects for the presence of null alleles.GenAlEx 6.5b4 (Peakall andSmouse, 2006, 2012) was used to calculate the number of alleles and the observed (H O ) and expected (H E ) heterozygosities.Arlequin v.3.5.1.2(Excoffier and Lischer, 2010) was used to determine deviation from Hardy-Weinberg equilibrium (HWE) and Cervus v.3.0 (Kalinowski et al., 2007) was used to calculate polymorphism information content (PIC) values.

RESULTS AND DISCUSSION
Six of nine T. brasiliensis primer pairs successfully crossamplified microsatellites in O. martiensseni and O. madagascariensis, namely TabrA10, TabrA30, TabrD10, TabrD15, TabrH6 and TabrH12.The remaining loci were not useable due to the presence of null alleles (TabrH3) or ambiguity in the peak data (TabrH2 and TabrE9) which rendered us unable to score these loci with confidence.Russell et al. (2005) also reported difficulty in amplifying TabrH2 across all Tadarida populations tested.Naidoo et al. (2013), who successfully cross-amplified six of the above loci in C. pumilus, were successful with TabrA10, TabrA30, TabrD10, TabrD15, TabrH6 and TabrE9.Repeat motifs of all loci were the same in Otomops and  (Naidoo et al., 2013) and the Argentinean population of Tadarida brasiliensis (Russell et al., 2005) are included for comparative purposes.
T. brasiliensis (Russell et al., 2005), whereas C. pumilus showed a different repeat motif for marker TabrA10 (Naidoo et al., 2013) (Table 2).Analysis of O. martiensseni data in micro-checker detected possible scoring error due to stuttering in 2 loci (TabrA10 and TabrH12) and the presence null alleles in 5 of the 6 loci tested (TabrA10, TabrA30, TabrD10, TabrH6 and TabrH12).O. madagascariensis data showed no null alleles or scoring error due to stuttering.None of the loci from either species showed any large allele dropout.To determine whether any substantial bias was introduced through the presence of the null alleles, pairwise F ST values were calculated between O. martiensseni and O. madagascariensis with (0.248) and without (0.249) the ENA algorithm.As the difference between the corrected and uncorrected estimates of genetic differentiation was not substantial, we report analyses performed on uncorrected data only.Null alleles, stuttering and large allele dropout were not reported for microsatellites crossamplified in C. pumilus (Naidoo et al., 2013).All of the O. martiensseni loci were polymorphic, with 5 to 16 (mean 10.33) alleles per locus.O. madagascariensis, however, showed lower levels of polymorphism; five of six loci were polymorphic, with polymorphism levels ranging from 1 to 7 (mean 3.5) alleles per locus.The lower level of polymorphism in O. madagascariensis is likely a reflection of the smaller sample size used for this species.Polymorphism levels in cross-amplified C. pumilus sensu lato microsatellites [9 to 15 (mean 11.7) alleles per locus] (Naidoo et al., 2013) were slightly higher than, but comparable to those of O. martiensseni.This is somewhat unexpected, as the higher divergence (RAG2 genetic distance) between T. brasiliensis and C. pumilus s.l.(4.6%) than between T. brasiliensis and O. martiensseni (3.2%) (Lamb et al., 2011) leads to an expectation of lower polymorphism in C. pumilus s.l.(Primmer et al., 2005).
Polymorphism levels in all cross-amplified microsatellites were considerably lower than those in T. brasiliensis, the species for which the primers were developed [15 to 54 (mean 36.67)alleles per locus].This is to be expected as number of amplified loci and the level of polymorphism tends to decrease with increasing genetic distance between the original and cross-amplified taxa (Primmer et al., 2005), and we are dealing here with cross-genus rather than cross-species amplification.The lower levels of polymorphism in Otomops and Chaerephon species are likely to reflect divergence which has occurred since Tadarida and Otomops (24.7 MYA) and Tadarida and Chaerephon (26.1 MYA) last shared common ancestors (Ammerman et al., 2012).
There was considerable variability in observed (H O ) and expected (H E ) heterozygosities across Otomops samples (Table 2).Consistent with expectation, the expected fractions of polymorphic offspring, as indicted by PIC values, are a little lower than the expected heterozygosities.The PIC of the O. martiensseni microsatellites ranged from 0.295 to 0.905 (mean 0.687), comparable to that of the similarly-sized sample of cross-amplified C. pumilus s.l.microsatellites, 0.51 to 0.80 (mean 0.67) (Naidoo et al., 2013).The PIC of the O. madagascariensis samples was generally lower (0.110 to 0.797 (mean 0.442), possibly due to the smaller sample size.PIC values showed some markers to be more informative than others, for example, TabrD15 was the most informative, with values of 0.905 and 0.797 in O. martiensseni and O. madagascariensis, respectively.
Markers with PIC values > 0.4 are considered moderately informative and those with values > 0.7 are considered highly informative (Hildebrand et al., 1992;Xu, 2010).Thus most of the loci tested in O. martiensseni can be deemed informative for linkage analysis, as three markers had PIC values >0.7 and two had PIC values >0.4 (Table 2).Only two markers in O. madagascariensis can be considered highly informative, that is, TabrD15 and TabrH6, but this may be due to the low sample number used.By comparison, all markers tested in C. pumilus appear to be informative, with TabrD10 having the highest PIC value (0.80) (Naidoo et al., 2013).
In summary, six of nine microsatellite markers reported for T. brasiliensis (Russell et al., 2005) have been successfully cross-amplified in two species of the molossid genus Otomops.These nuclear markers do not have as high a level of polymorphism as in the originally-studied species, T. brasiliensis, but PIC values indicate that they are sufficiently polymorphic for use in population-, colonyand individual-level genetic studies.This will allow for future work on intra-and inter-colony relationships in Otomops.Additionally, these markers may also be useful in population genetic studies on the other Otomops species, such as O. wroughtonii from southern India and O. formosus from Java.Comparison of marker statistics in cross-amplified Otomops and Chaerephon microsatellites (Naidoo et al., 2013) revealed some similarities, for example, relatively lower allele numbers and PIC values, which may be attributed to mutations which have occurred in these lineages in the 24.7 to 26.1 million years since they last shared a common ancestor with T. brasiliensis.The markers developed by Russell et al. (2005) have been successfully cross-amplified in two other molossid genera, Otomops and Chaerephon, and may therefore have the potential to be used for population genetic studies of not only Otomops, but also other poorly-studied molossid genera in the future.
AAG TGG TTG GGC GTT GTC R: GCG ATG CAC TGC CTT GAG A  TabrA30 F: AGT CGC GGG TTT GAT TCC AGT TA R: ACC CCT TCC CTT TGT TCC TTC AG  TabrD10 F: CCC CAC TCA TTT ATC CAT CCA CA  R: ATC TCG CAG CTA TTG AAG TA TabrD15 F: AGT CCT GGC TCC TAT TCT CAT TG R: CTA TCC GTC TAC CTG TCC GTC TAT TabrH6 F: ATC TCT CCA GTC CTT ACC A  R: TTT ACC CTC CAC AGT CTC A TabrH12 F: CCA TGT GAG CCA ATT CCT A  R: GTC AGG ACT CTC CAG AGA F = forward primer; R = reverse primer.Primer labels are indicated: 6-FAM, NED.

Table 1 .
Primer sequences from Russell et al. (2005) used in the cross-amplification of microsatellites in O. martiensseni and O. madagascariensis.
T a , Optimised annealing temperature; H O , observed heterozygosity; H E , expected heterozygosity; PIC, polymorphism information content.Significant deviations from Hardy-Weinberg equilibrium for each locus are indicated (*, P < 0.05 after sequential Bonferroni correction).Data for Chaerephon pumilus sensu lato