Microsatellites for the Neotropical Ant, Odontomachus chelifer (Hymenoptera: Formicidae)

Abstract Odontomachus chelifer (Latreille) (Ponerinae) is a ground-dwelling, predominantly carnivorous ant whose colonies may contain multiple egg-laying queens and are potentially susceptible to border effects in the Brazilian savanna known as Cerrado. The ecology and natural history of O. chelifer is well studied, but very little is known about the genetic diversity of O. chelifer colonies. In this study, we developed microsatellite markers for the study of genetic variation in O. chelifer. We created a microsatellite-enriched library that resulted in the development and characterization of 22 markers, of which 18 were found to be polymorphic in the population studied. The mean expected heterozygosity was 0.59, whereas the mean rarified allelic richness was determined as 4.27 alleles per locus. The polymorphism level detected was similar to genetic diversity estimates found in other poneromorph ant species. The microsatellites developed here are likely to be useful for the investigation of colony structure, functional polygyny, breeding system, and population genetics in O. chelifer. Moreover, the description of O. chelifer’s genetic diversity is crucial for its conservation and maintenance of its ecological role in the Cerrado savanna.


Sampling Site
Colonies of O. chelifer were sampled in Fazenda Campininha, a Cerrado savanna reserve located in the city of Mogi-Guaçu (22°18′S e 47°11′W), in the state of São Paulo, southeastern Brazil. Via active searching, we collected 15-20 workers from 18 O. chelifer nests. Ant voucher specimens are deposited at the Museu de Zoologia da Universidade Estadual de Campinas (ZUEC, Campinas, Brazil).

DNA Extraction
DNA extraction was performed using methods described by Saghai-Maroof et al. (1984). In brief, ant individuals are solubilized in 2% CTAB solution, followed by DNA purification through extraction with chloroform/isoamyl alcohol (24:1).

Microsatellite Identification
Six workers from the same colony of O. chelifer were used to build a microsatellite-enriched library, based on protocol previously described in Billotte et al. (1999), using a hybridization-based capture with (CT) 8 and (GT) 8 biotin-linked probes, followed by recovery with streptavidin magnetic-coated beads (Promega, Madison, WI). The selected fragments were cloned into the pGEM-T vector (Promega) and transformed into competent Escherichia coli (XL1-Blue strain). Recombinant colonies were identified by colorimetric white-blue detection using X-gal. Plasmid DNA was then extracted using an alkaline lysis method and inserts were sequenced on a 3500 Genetic Analyzer sequencer (Applied Biosystems, Foster City, CA). We sequenced 103 clones and the electropherograms were edited using CLC Genomics Workbench v. 4.9 software (CLC Bio, Arhus, Denmark). Vector, adapter, and restriction site sequences were removed using Seqman (DNAStarInc, Madison, WI). Additionally, sequences were compared with entries of the public National Center for Biotechnology Information database using Blastn (Altschul et al. 1990), to eliminate possible contaminant sequences. Microsatellites were identified using the web-based SSRIT software (Temnykh et al. 2001) and primer pairs complementary to their flanking sequences were designed using Primer Select (DNAStarInc) and Primer3Plus (Untergasser et al. 2012). Amplified fragments ranging from 100 to 250 bp in size were selected for further identification of putative alleles. For each sequence, primers were designed according to the following guidelines: 1) 18-22 nucleotides in size, 2) G/C content greater than 35%, 3) T m between 45 and 65°C (maximum of 3°C difference between both primers in the pair), and 4) presence of A or T bases at the 3′ end (to reduce probability of self-pairing). We also avoided designing primers that allow the formation of homo or heterodimers. For automating the genotyping on the 3500 Genetic Analyzer sequencer (Applied Biosystems), an M13 sequence tail (5′-CACGACGTTGTAAAACGAC-3′) was added to the 5′ end of each forward primer (Schuelke 2000). Fluorescent labels 6-FAM, VIC, NED, and PET (Applied Biosystems) were used to increase genotyping efficiency.

Microsatellite Characterization
Five workers from different nests were used to determine ideal amplification conditions for each developed marker. For all markers, we tested two touchdown PCR protocols (Don et al. 1991), with hybridization temperatures between 52 and 57°C or between 55 and 60°C, together with the following steps: 1) 94°C for 4 min; 2) 10 cycles of 94°C for 45 s, 60°C or 57°C (−0.5ºC/cycle) for 1 min, and 72°C for 1 min and 15 s; 3) 25 cycles of 94°C for 45 s, 50°C for 1 min, and 72°C for 1 min and 15 s; and 4) 72°C for 10 min. The amplification products were evaluated on a 3500 Genetic Analyzer sequencer (Applied Biosystems) following by post-sequencing analysis on Geneious prime software (v. 2019.2; Biomatters Limited, New Zealand).
Microsatellite loci with amplification patterns consistent with expected sizes and clear distinguishable peaks were further characterized for polymorphism content. To this end, we used 30 workers from different nests (Hale et al. 2012), and microsatellite loci were evaluated for occurrence of stuttering and reduced amplification of large fragments using Micro-Checker (Van Oosterhout et al. 2004). Polymorphism content (PIC) (Botstein et al. 1980) and observed and expected heterozygosity for each locus were obtained using the Microsatellites Toolkit supplement in Excel (Park 2008). Rarefied allelic richness was estimated by HP-Rare software (Kalinowski 2005). Additionally, we tested for loci adherence to frequencies expected in the Hardy-Weinberg equilibrium (HWE) using Genepop 4.7 (Rousset 2008). Linkage disequilibrium (LD) between each pair of markers was assessed using FSTAT 2.9.4 (Goudet 1995). For both HWE and LD estimates, the significance value (0.05) was corrected for multiple comparisons. The frequency of null alleles was estimated with the FreeNA software (Chapuis and Estoup 2007).

Results
The microsatellite enrichment procedure was highly efficient, with 87.85% of the sequenced clones presenting repetitive sequences. Fifty-three clones contained more than one microsatellite sequence, totaling 94 sequences. We were able to design primer pairs for 42 microsatellite loci. Twenty-two loci were successfully amplified using the touchdown PCR protocol with hybridization temperature ranging from 55 to 60°C, and all of them resulted in amplification products consistent with expected sizes (Table 1), without evidence for nonspecific amplification.
Importantly, 18 markers were found to be polymorphic. In these loci, the expected heterozygosity (H E ) (mean ± SE) was 0.59 ± 0.05, with the highest values found in the Och6 (0.89), Och3 (0.86), and Och8 (0.80) loci. Very low H E values were identified in Och88 (0.07) and Och15 (0.08) ( Table 1). Rarified allelic richness (mean ± SE) was 4.27 ± 0.53 (Table 1). The mean (±SE) PIC value was 0.53 ± 0.05 (Table 1). We found no evidence of allele stuttering or reduced amplification of fragments for any of the markers. Moreover, the frequency of null alleles ranged from 0 to 0.26735. Regarding adherence to frequencies expected at HWE, we found only seven loci at equilibrium (39%) (Table 1). Finally, it should be noted that all microsatellite loci analyzed exhibited independent segregation.

Discussion
A microsatellite-enriched library was developed for O. chelifer, resulting in 22 markers, 18 of which were found to be polymorphic in the population studied. Even though some markers displayed low polymorphic content, most proved to be highly informative (PIC > 0.5) to access genetic diversity in O. chelifer (Table 1). Diversity estimates for our markers were similar to estimates for other ant species in the subfamily Ponerinae. For instance, our results are similar to microsatellite marker allelic richness observed in Pachycondyla inversa (5-12 alleles; Trindl et al. 2004), Hypoponera opacior (9-21 alleles; Rüger et al. 2005) and Pachycondyla luteipes (2-8 alleles; Takahashi et al. 2005), even when rarefied allelic richness is considered in our work. Despite the availability of microsatellite markers for other Ponerinae or for other ant species (Butler et al. 2014), interchangeability of such markers between species is challenging. This occurs due to their high specificity, which makes cross-amplification limited even at the genus level (Barbará et al. 2007). Thus, developing specific microsatellite loci for each species is a necessity, especially in the context of studies on genetic diversity, reinforcing the importance of the markers developed here for future studies on O. chelifer genetic variation.
Eleven microsatellite loci exhibited deviations from HWE, indicating violation of one or more assumptions of the Hardy-Weinberg model, namely presence of selection, migration and/or mutation, finite population size, and non-random mating (Hartl and Clark 2010). Additionally, such deviations may arise from overlapping generations and/or the high relatedness between ant workers. Moreover, our sampling location is a fragmented Cerrado area (Christianini and Oliveira 2013), which may have reduced population size due to habitat loss and/or increased probability of inbreeding (Frankham 2010, Banks et al. 2013. Further investigation is needed to evaluate the effect of habitat fragmentation on genetic variation of O. chelifer. Although the behavior and ecology of O. chelifer have already been studied in Brazilian forests and savannas (Oliveira et al. 2017, and references therein), the microsatellite markers described here will stimulate further investigation on colony structure and breeding system in this ant species, including potential effects of habitat fragmentation, a crucial knowledge in the context of conservation of its populations and maintenance of its ecological role.

Acknowledgments
We are grateful to Gustavo M. Mori and Rebecca C. U. Ferreira for helpful comments on the manuscript, and to Luis Salles for help during field work. The final version of the manuscript was revised by Fabio Papes. The Instituto Florestal de São Paulo and the staff of the Estação Experimental de Itirapina provided logistic support during field work. The study was supported in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES, Finance Code 001). ASML and SG-N were supported by the Conselho Nacional de Desenvolvimento Científico e Tecnológico