Isolation and Characterization of Microsatellite Loci for the Isopod Crustacean Armadillidium vulgare and Transferability in Terrestrial Isopods

Armadillidium vulgare is a terrestrial isopod (Crustacea, Oniscidea) which harbors Wolbachia bacterial endosymbionts. A. vulgare is the major model for the study of Wolbachia-mediated feminization of genetic males in crustaceans. As a consequence of their impact on host sex determination mechanisms, Wolbachia endosymbionts are thought to significantly influence A. vulgare evolution on various grounds, including population genetic structure, diversity and reproduction strategies. To provide molecular tools for examining these questions, we isolated microsatellite loci through 454 pyrosequencing of a repeat-enriched A. vulgare genomic library. We selected 14 markers and developed three polymorphic microsatellite multiplex kits. We tested the kits on two A. vulgare natural populations and found high genetic variation, thereby making it possible to investigate the impact of Wolbachia endosymbionts on A. vulgare nuclear variation at unprecedented resolution. In addition, we tested the transferability of these kits by cross-species amplification in five other terrestrial isopod species harboring Wolbachia endosymbionts. The microsatellite loci showed good transferability in particular in Armadillidium nasatum and Chaetophiloscia elongata, for which these markers represent promising tools for future genetic studies.


Introduction
Armadillidium vulgare is a terrestrial isopod (Crustacea, Oniscidea) which exhibits a worldwide distribution. A. vulgare harbors alphaproteobacterial endosymbionts of the genus Wolbachia [1,2]. These maternally-inherited, intracytoplasmic bacteria are known to manipulate host reproduction to enhance their own transmission through four different mechanisms: cytoplasmic incompatibility, thelytokous parthenogenesis, male killing and feminization [3,4]. Wolbachia endosymbionts are prevalent in terrestrial isopods [5], and A. vulgare has emerged as a major model for studying Wolbachia-mediated feminization [3,[6][7][8]. In A. vulgare, zygotes carrying Wolbachia develop a female phenotype, whatever their sex chromosome composition. In particular, genetic males harboring Wolbachia are converted into functional females. As a consequence, A. vulgare populations in which Wolbachia are present show sex ratio distortions towards females, thereby enhancing Wolbachia spread in infected populations. In addition, some A. vulgare individuals carry another feminizing factor, known as the f element, which may be a fragment of the Wolbachia genome carrying feminization information and transferred into the host nuclear genome [3,9]. Furthermore, the occurrence of multiple feminizing factors has generated genetic conflicts in this system, which resulted in the selection of A. vulgare nuclear genes resisting feminization [3,10,11]. Thus, sex determination mechanisms are very dynamic in A. vulgare, outlining the prime influence of Wolbachia. These endosymbionts are thought to impact A. vulgare evolution on various additional grounds, including population genetic structure, diversity and reproduction strategies [12].
Mitochondrial DNA markers have been used in several studies, indicating relatively high variability in A. vulgare [1,[13][14][15] as compared to other terrestrial isopod species such as Porcellionides pruinosus [16]. A more elaborate understanding of A. vulgare/ Wolbachia interactions would benefit from information on A. vulgare nuclear variation. Recently, five microsatellite markers [17] were used to investigate nuclear variation in A. vulgare populations from western France, suggesting a genetic structure compatible with isolation by distance [14]. Although polymorphic, these markers may not be in sufficient number to offer the desired resolution for detecting a possibly subtle impact of Wolbachia on A. vulgare nuclear variation, population dynamics and evolution. To provide tools for examining these questions, we isolated microsatellite loci through 454 pyrosequencing of a repeat-enriched A. vulgare genomic library. We selected markers yielding clear amplification signals and showing appropriate polymorphism levels. Next, we used the candidate loci to develop three polymorphic microsatellite multiplex kits. The transferability of these kits was tested by cross-species amplification in other terrestrial isopod species harboring Wolbachia endosymbionts [5,18].

Ethics Statement
No ethics statement was required for the described study. No specific permission was required for sampling the two A. vulgare field populations (La Crèche and Beauvoir-sur-Niort, France) because they were located in public areas. Field populations of Armadillidium nasatum (Poitiers, France) and P. pruinosus (Buxerolles, France) were sampled on private lands after the land owners gave permission to conduct sampling on the sites. None of these species is an endangered or protected species.

Microsatellite isolation
For genomic library construction, we maximized genomic diversity by using eight A. vulgare female individuals selected from the following laboratory lines: BF (bac 377), BH (bac 366), CY (bac 291), CW (bac 49), POA (bac 42), WS (bac 45), WX (matricule 1288) and ZM (bac 47). Total genomic DNA was obtained for each individual by standard phenol-chloroform extraction [19] followed by RNase (10 mg/ml) treatment. DNA concentration was measured using a picogreen assay and equimolar amounts of the eight samples were pooled. The pooled sample was used by GenoScreen (Lille, France) to construct a microsatellite-enriched genomic library, as previously described [20]. The library was sequenced by GenoScreen in a partial 454 GS FLX sequencer run with Titanium chemistry, as previously described [20]. The resulting reads were analyzed with the QDD software [21] to identify reads containing microsatellite motifs and design primers for PCR amplification.

Locus validation and polymorphism tests
All microsatellite loci with PCR primers designed using QDD were initially tested using two A. vulgare female individuals from our laboratory line BF (matricule 2756).Total genomic DNA from the two samples was extracted as above and subjected to whole genome amplification using the illustra GenomiPhi V2 DNA Amplification Kit (GE Healthcare) to generate large enough amounts of template DNA for microsatellite testing. To reduce genotyping costs, each locus was amplified and fluorescently labeled using the M13(-21) primer genotyping protocol [22]. This PCR method uses three primers: a locus-specific forward primer with M13(-21) tail at its 59 end, a locus-specific reverse primer and a universal 6_FAM-labeled M13(-21) primer. PCR amplification was performed in 10 mL reactions, using 0.5 mM of both 6_FAM-M13(-21) and reverse primers, 0.125 mM forward primer, 0.25 U GoTaq DNA Polymerase (Promega), 1X PCR reaction buffer (Promega), 0.2 mM dNTPs (Promega) and 1 mL DNA template. PCR thermal conditions were as previously described [22]. Subsequently, 0.5 mL PCR products were added to 9 mL formamide and 0.35 mL ROX standard (Life Technologies), and resolved by electrophoresis on an ABI PRISM 3130 Genetic Analyzer. Product sizes were determined using the GeneMapper software (Applied Biosystems), followed by eye verification.
Microsatellite loci amplifying in at least one of the two tested individuals and yielding unambiguous amplification signals were further evaluated for their informativeness. Amplification success rates and number of different alleles at each locus were assessed by genotyping a panel of 24 A. vulgare individuals (12 males and 12 females) from five laboratory lines: BF (n = 5), BFog (n = 4), WXa (n = 5), BG (n = 5) and ZM (n = 5). First, the 24 samples were subjected to whole genome amplification as described above. Next, microsatellite loci were genotyped using the M13 (-21) primer protocol described above.

Multiplexing and cross-species amplification
Based on the genotyping results of the 24-individual panel, we selected 14 microsatellite markers for which locus-specific forward primers (without M13(-21) tail) were ordered with labeled dyes (6_FAM, HEX or NED) (Table 1). First, we verified amplification of the 14 markers in simplex PCR conditions on three individuals from each of two A. vulgare field populations (La Crèche and Beauvoir-sur-Niort, France, Table 2). Total genomic DNA was extracted as above. All PCR reactions were carried out using the QIAGEN multiplex PCR kit according to the manufacturer's standard microsatellite amplification protocol in a final volume of 10 mL, with an annealing temperature of 58uC and a final concentration of 0.2 mM for each primer. DNA concentrations were adjusted for all individuals between 20 and 60 ng/mL. Next, 1 mL PCR product was added to 18 mL formamide and 0.5 mL ROX standard (Life Technologies). PCR products were resolved by electrophoresis and their size determined as described above.
After simplex PCR verification, we pooled the 14 markers in three multiplex kits (Table 1) according to amplified fragment sizes and dyes to maximize efficiency and minimize costs. The multiplex kits were tested with the same three individuals used for simplex PCR reactions, using the QIAGEN multiplex PCR kit as described above. Identical results were obtained for both simplex and multiplex PCR conditions, thereby validating the use of the three multiplex kits in subsequent analyses. Polymorphism of the 14 microsatellite loci in A. vulgare field populations was evaluated by genotyping 20 individuals from each of two populations (La Crèche and Beauvoir-sur-Niort, France) using the multiplex kits.

Data analyses
To assess genetic variability and transferability of our microsatellite markers, we calculated number of alleles (Na), unbiased expected heterozygosity (He) [23], observed heterozygosity (Ho) and Fis [24] using Genetix version 4.05.2. We computed these genetic indices from two A. vulgare populations and from individuals of the five other aforementioned species. Departure from Hardy-Weinberg expectations was assessed for each microsatellite marker using exact tests (5000 permutations), as implemented in GENEPOP version 3.4 [25]. Linkage disequilibrium was assessed for each microsatellite marker using FSTAT version 2.9.3.2 [26] and with 1000 permutations. The level of significance was adjusted for multiple testing using a sequential Bonferroni correction technique [27].

Locus identification, validation and polymorphism
Sequencing of the microsatellite-enriched library yielded 18,511 reads. The sequence dataset is available in the Dryad database at http://doi.org/10.5061/dryad.md545. Of these, 5073 (27%) reads contained microsatellite motifs according to QDD analysis. Primer pairs were designed for all loci fulfilling our criteria for primer design [20]. The 146 resulting loci comprised 93 di-, 43 tri-, 5 tetra-,1 penta-and 4 hexanucleotide repeat microsatellites with 5 to 22 repeat units (Table S1). Two A. vulgare individuals were genotyped for the 146 microsatellite loci and 41 loci were validated under our amplification conditions and criteria (Table S1). Out of the 41 loci, a first polymorphism analysis based on a 24-individual panel allowed us to identify 33 polymorphic loci (i.e. 80%) (Table  S1). Among these 33 polymorphic loci, we selected 14 loci for inclusion in multiplex kits, according to the following criteria: (i) repeat type (larger motifs favored), (ii) number of different alleles   (Table 1). We genotyped 40 A. vulgare individuals from two field populations with the three multiplex kits. No significant linkage disequilibrium between the different loci was observed after sequential Bonferroni correction. Two loci (AV0023 and AV0061) were monomorphic in the two tested populations (Table 2). However, we kept these two loci in our multiplex kits because they were polymorphic in our initial 24-individual panel, suggesting that they may be informative for other populations. The number of alleles for the 12 other loci ranged from 2 to 6 among the 40 A. vulgare individuals, with a mean of 3.9 alleles per locus. Microsatellite loci AV0002, AV0018, AV0056 and AV0099 were highly polymorphic in both populations and they may be particularly relevant markers for analyses requiring high discriminating power, e.g. to investigate paternity in A. vulgare. Observed heterozygosity levels varied from 0.1 to 1 in both populations. After sequential Bonferroni correction for multiple testing, loci AV0096 and AV0128 revealed significant excess of heterozygotes in both populations and locus AV0099 showed a significant deficit in La Crèche population. Thus, these loci departed from Hardy-Weinberg equilibrium in concerned populations, likely because of non-exhaustive population sampling. These loci may therefore turn out to be useful for future studies with a more important sampling. The other highlighted polymorphic microsatellites represent a useful set of markers to perform genetic studies on A. vulgare and to investigate the impact of Wolbachia endosymbionts on A. vulgare population genetic structure and evolution.

Locus transferability in terrestrial isopod species
The 14 newly developed markers were tested in five isopod species. Results are summarized in Table 3. Among these species, A. nasatum and C. elongata revealed high cross-species transferability with amplification success of 71% (10/14 loci) and 86% (12/14 loci), respectively. Conversely, O.asellus, P. pruinosus and P. scaber revealed moderate amplification success with 21% (3/14 loci), 29% (4/14 loci) and 43% (6/14 loci), respectively. Depending on species, mean number of alleles ranged from 1.8 to 2.8 and mean expected heterozygosity ranged from 0.40 to 0.53. Results from A. nasatum are not really surprising given the close phylogenetic relationship with A. vulgare. The important rate of successful crossamplification found in C.elongata is more surprising but microsatellite loci show a reduced number of alleles and low heterozygosity indices in C. elongata relative to A. vulgare. We detected no linkage disequilibrium whereas departure from Hardy-Weinberg expectations was detected for loci AV0056 and AV0096 in P. scaber and for loci AV0032 and AV0085 in A. nasatum. The deficit in heterozygotes observed for these loci could be explained by the quite small sampling.

Conclusions
In sum, our work highlights a large set of microsatellite markers useful for studies on A. vulgare and other terrestrial isopod species. The polymorphism of these markers now makes it possible to analyze genetic diversity, population structure and reproduction strategies of A. vulgare at unprecedented resolution. A study based on these markers is now underway to analyze the impact of Wolbachia bacterial endosymbionts on A. vulgare nuclear variation. Moreover, these microsatellite markers showed good transferability in five other terrestrial isopod species, in particular in A. nasatum and C. elongata, for which these microsatellite markers represent promising tools for future genetic studies.

Supporting Information
Table S1 Information on the 146 microsatellite loci from Armadillidium vulgare tested for inclusion in multiplex kits. For each locus the following information is provided: locus name, repeat motif, repeat number, reference sequence, forward primer, reverse primer, PCR product size in reference sequence, and results of experimental tests leading to final selection for three multiplex kits. (XLSX)