A new suite of twenty-two polymorphic microsatellite loci in the parasitic wasp , Habrobracon hebetor ( Hymenoptera : Braconidae ) : Promising molecular tools for studying the population genetics

Combining a biotin-enrichment protocol and 454GS-FLX titanium pyrosequencing technology, we characterised 22 polymorphic microsatellite loci from the parasitic wasp, Habrobracon hebetor (Say) (Hymenoptera: Braconidae), a cosmopolitan species commonly used in biological control against a wide range of both major lepidopterous pests of stored products and fi eld crops in different parts of the world. Three multiplex PCR sets were optimised and characterised across 46 H. hebetor specimens from two samples collected from millet fi elds in Niger. Two to 11 alleles were found per locus and observed heterozygosity ranged from 0.289 to 0.826. Polymorphism was detected in both samples with a similar level of observed heterozygosity (0.482 vs. 0.502) and number of alleles (4.1 vs. 3.6). Deviation from Hardy-Weinberg equilibrium was detected at the same fi ve loci in both samples and fi ve or seven more loci in each sample but was not associated with heterozygote defi ciencies. Even though evidence for linkage disequilibrium was found between a few alleles, these new loci segregated independently. The variability of the 22 loci will enable estimates of genetic diversity and structure patterns, as well as gene fl ow between H. hebetor populations at different spatial scales. Cross-species amplifi cations were successful among the six Bracon spp. tested and nine loci will be particularly appropriate for population genetic studies in B. brevicornis. * Corresponding author; e-mails: nathalie.gauthier@ird.fr, nathalie.gauthier@supagro.inra.fr INTRODUCTION Habrobracon hebetor (Say) (Hymenoptera: Braconidae) is a well-known gregarious ectoparasitoid of the larvae of a wide range of economically important moths that infest stored grains, nuts, and fruits as well as fi eld crops in various parts of the world. This parasitoid is considered to be one of the most promising biological control (BC) agents for many stored-product pests, including the Indian meal moth Plodia interpunctella (Hübner), the Mediterranean fl our moth, Ephestia kuehniella (Zeller), the rice moth Corcyra cephalonica (Stainton), the greater wax moth, Galleria mellonella L., the cotton bollworm Helicorverpa armigera (Hübner), and also the millet head miner, Heliocheilus albipunctella (de Joannis), thanks to its cosmopolitan distribution and ability to regulate populations of many destructive moth pests of stored-food commodities (Payne et al., 2011; Adarkwah et al., 2014; Ba et al., 2014; Ghimire & Philipps, 2014). The parasitoid is sold commercially as a biological control agent, whilst because it can easily be mass reared, indigenous populations have also been reared and released in the fi eld or in storage for effecEur. J. Entomol. 113: 265–269, 2016 doi: 10.14411/eje.2016.031


INTRODUCTION
Habrobracon hebetor (Say) (Hymenoptera: Braconidae) is a well-known gregarious ectoparasitoid of the larvae of a wide range of economically important moths that infest stored grains, nuts, and fruits as well as fi eld crops in various parts of the world.This parasitoid is considered to be one of the most promising biological control (BC) agents for many stored-product pests, including the Indian meal moth Plodia interpunctella (Hübner), the Mediterranean fl our moth, Ephestia kuehniella (Zeller), the rice moth Corcyra cephalonica (Stainton), the greater wax moth, Galleria mellonella L., the cotton bollworm Helicorverpa armigera (Hübner), and also the millet head miner, Heliocheilus albipunctella (de Joannis), thanks to its cosmopolitan distribution and ability to regulate populations of many destructive moth pests of stored-food commodities (Payne et al., 2011;Adarkwah et al., 2014;Ba et al., 2014;Ghimire & Philipps, 2014).The parasitoid is sold commercially as a biological control agent, whilst because it can easily be mass reared, indigenous populations have also been reared and released in the fi eld or in storage for effec-complementarity of the primers and between the primers matched the quality criteria used as default parameters in Primer3.
Thirty-six primer pairs were selected.The resulting 36 potential new markers were BLASTed (http://blast.ncbi.nlm.nih.gov/Blast.cgi)against previously published markers and none showed signifi cant nucleotide identity with GenBank sequences.Consequently, they were optimised by monolocus PCR with DNA from eight females of diverse geographical origins.Genomic DNA was extracted from each individual H. hebetor using the DNeasy Tissue Kit (Qiagen).PCR amplifi cations were performed in a fi nal volume of 10 μl containing 5 μL of the Qiagen multiplex PCR Master Mix (1X) (including Taq, 200 μM of each dNTP and 1.5 mM MgCl 2 ), 2 μM of primers, 1 μL of genomic DNA, and 2.6 μL of RNase-free water.All PCRs comprised the following steps: (i) initial denaturation at 95°C for 15 min; (ii) 30 denaturation cycles at 94°C for 30 s, annealing at 57°C for 90 s, elongation at 72°C for 60 s; and (iii) fi nal elongation at 60°C for 30 min.Twenty-four of the loci provided high quality amplifi cations with unambiguous allelic patterns and apparent polymorphism scored by agarose gel electrophoresis, and were retained for multiplex PCRs with fl uorescently labelled primers.Using the Multiplex Manager v1.1 software (Holleley & Geerts, 2009), they were arranged in three multiplex PCRs that minimised the formation of duplexes and maximised the range of amplifi cation product sizes (Table 1).

Genetic analyses
To characterise the 24 loci, the 46 individuals collected in Niger were genotyped (Table 1).For analysis of microsatellite loci polymorphism, PCR amplifi cations of the extracted DNA were performed as described above.PCR products were separated on an ABI 3730 sequencer (Applied Biosystems, Montpellier, France).Allele sizes were scored against GeneScan TM 500 LIZ standard (Applied Biosystems) using GeneMapper TM v 4.0 software, and confi rmed manually.
Levels of polymorphism [i.e.number of alleles observed (Na), observed (H O ) and unbiased expected (H E ) heterozygosities, respectively], deviation from Hardy-Weinberg equilibrium (HWE) and linkage disequilibrium (LD) for each locus were estimated using GenePop 4.2 available on the web (Rousset, 2008; http:// genepop.curtin.edu.au/),both for each sample and globally over the 46 specimens.To correct for multiple comparisons, a sequential Bonferroni correction was applied for both LD and HWE tests.Where deviation from HWE was detected, departure to panmixia due to the presence of null alleles or/and scoring errors was estimated using the Micro-Checker program (van Oosterhout et al., 2004) and FreeNA package (Chapuis & Estoup, 2007).To assess the level of genetic differentiation between both samples, pairwise F ST value over all loci was computed using FreeNA (with the ENA correction for null alleles) following 1000 bootstrapping replications over loci.
Cross-species amplifi cation was tested using the three PCR multiplexes in the 61 females of six other related Bracon species, 19 B. brevicornis, 24 B. nigricans, 12 B. cephi, 4 B. lissogaster and 2 Bracon spp.PCR and genotyping conditions were as described above.

Genetic diversity, Hardy-Weinberg and linkage testing
At each locus, all 46 H. hebetor specimens provided a genotype (except for one specimen from Dantchandou at locus Heb3-08).Twenty-two loci out of the 24 tested were polymorphic, with the exception of Heb3-04 and Heb3-06).The number of alleles per locus ranged from two to 11 (average 4.667) and the observed heterozygosity over all specimens and loci from 0.289 to 0.826 (mean H O = 0.491 ± 0.218).Polymorphism was detected in the ily.We thus decided in the present study to use a combination of biotin-enrichment and 454 pyrosequencing technologies, to isolate new microsatellite loci from H. hebetor.Of twenty-four microsatellite markers tested, 22 loci were found to be polymorphic and were subsequently tested and optimised in three multiplex reactions using 46 females from two natural populations sampled in Niger, West Africa.The cross reaction of these particular markers was also tested in six closely related braconids [Bracon brevicornis (Wesmael), B. nigricans (Szépligeti), B. cephi (Gahan), B. lissogaster (Muesebeck), and two Bracon spp.], some of which are also being used as BC agents in integrated pest management schemes to combat moth pests.

Sampling and DNA extraction
Adult H. hebetor were collected in late summer 2014 from a pearl millet fi eld in two localities in Niger located 80 km apart: Dantchandou (13.40634°N, 2.71836°E) and Tondikoarey (13.58236°N, 1.99969°E) (Table 1).All adults from the same locality were placed alive in a tube fi lled with 80% ethanol and sexed in the laboratory using a binocular microscope.In our study, a total of 46 females consisting of 24 females H. hebetor sampled from Dantchandou and 22 females from Tondikoarey, respectively, were then stored at -20°C until molecular experiments.
Moreover, 61 females of six other related Bracon species were studied: 19 B. brevicornis originated from a laboratory massrearing initiated with founders collected in 2006 near the city of Leipzig (Germany) and 24 B. nigricans from a laboratory massrearing initiated in 2015 with founders from Sicily (Italy), 12 B. cephi and 4 B. lissogaster were collected in 2015 in the Montana (USA), and the two other Bracon spp.were collected in 2015 in the Ariège (South of France).All adults from the same locality and species were placed alive in a tube fi lled with 80% ethanol and sexed in the laboratory.Then, they were stored at -20°C until molecular experiments.
Total genomic DNA of all females collected was individually extracted with the DNeasy Blood & Tissue Kit (Qiagen, Courtaboeuf, France) according to the manufacturer's instructions.

Microsatellite library, primer design and PCR amplifi cation
Total genomic DNA from 30 pooled individual H. hebetor was similarly extracted using the DNeasy Blood & Tissue Kit (Qiagen) according to the manufacturer's instructions.Microsatellite loci were isolated at Genoscreen (Lille, France) using a biotinenrichment protocol adapted from Kijas et al. (1994).Production and pyrosequencing of enriched libraries for eight microsatellite motifs [i.e.(TG) 10 , (TC) 10 , (AAC) 8 , (AAG) 8 , (ACG) 8 , (AGG) 8 , (ACAT) 6 , (ACTC) 6 ] were carried out as described in Malausa et al. (2011).The resulting 39,189 sequences were subsequently sorted and selected using the open access QDD program (Meglécz et al., 2010).A total of 13,151 concatenated sequences longer than 100 bp with motifs displaying at least six repeats and tandem-repetition-free-fl anking regions were obtained, representing a huge source of potential markers.Nineteen to 22-bp long primers were designed using the web-based Primer3 algorithm (http://frodo.wi.mit.edu/primer3/,Rozen & Skaletsky, 2000) implemented in QDD, and satisfi ed the following criteria (i) the target microsatellite had to have at least six repeats; (ii) the resulting PCR product had to be between 100 and 420 bp long; (iii) the fl anking region had to contain at most a three-base mononucleotide stretch or two repeats of any di-hexa motif; (iv) the annealing temperature (Ta) was between 60°C and 63°C, and the difference in Ta between the forward and the reverse primer was ≤ 1.5°C; and (v) the self-  1).Only one fi xed allele was observed at locus Heb4-1 in the Tondikoarey sample (Table 1).After Bonferroni correction, signifi cant deviations from HWE were globally detected at fi ve loci (Heb2-02, Heb2-06, Heb2-03, Heb2-14, Heb2-15) in both samples, the Tondikoarey sample exhibiting deviations at seven more loci and Dantchandou at fi ve other loci (Table 1).From estimations made using the Micro-Checker and FreeNA softwares, there was no evidence of scoring errors and signifi cant frequencies (f > 0.15) of null alleles, except for three loci in the Tondikoarey sample and one in the Dantchandou sample.After Bonferroni correction, evidence for LD was detected between a few alleles but they were not the same when considering Dantchandou and Tondikoarey samples separately (data not shown).The sampling itself rather than physical linkage between the loci may explain this result.

Genetic differentiation
The level of genetic differentiation between the samples from Dantchandou and Tondikoarey (F ST(ENA) = 0.0981) was very low and statistically non-signifi cant.In agreement with this, genetic assignment tests signifi cantly assigned all genotyped individuals to the same genetic population (results not shown).

Cross-species PCR amplifi cation
Cross-species amplifi cations were mostly successful.Twentythree loci out of the 24 selected loci from H. hebetor amplifi ed in all B. brevicornis specimens, versus 11 in 22 out of 24 of the B. nigricans specimens, 7 and 6 in at least 50% of the B. lissogaster and B. cephi specimens, respectively, and 9 and 6 in the two other Bracon spp.tested (Table 2).Interestingly, some of the loci amplifi ed in H. hebetor displayed signifi cant polymorphism in the related braconids tested, particularly in B. brevicornis with nine polymorphic loci, and B. cephi and B. lissogaster with seven polymorphic loci.Some alleles at some loci could also be species-diagnostic as they were fi xed and private.For example, locus Heb2-16 had a fi xed 226 bp allele for B. brevicornis vs. alleles of 222 bp for B. nigricans, whereas it failed to amplify using the DNA from B. lissogaster and Bracon sp.1 and provided other allele size ranges for H. hebetor and Bracon sp.2 (Table 2).

DISCUSSION
In our study, 22 polymorphic microsatellite markers were developed for the fi rst time in the benefi cial parasitic wasp, H. hebetor and − as far as we know − for closely related Braconids also.
Even though fi ve loci of the 22 characterized displayed departures to HWE, no locus displayed signifi cant heterozygote defi ciencies nor signifi cant LD.Both the mating behaviour and life history of H. hebetor promote outbreeding (in Antolin et al., 2003).But the individuals genotyped in our study were from small samples and on an agricultural crop, and may have been subject to selection and/or a possible population bottleneck.Inherent population fl uctuations may contribute to the deviations from HWE observed at some loci.As a result, these new loci are potentially valuable for investigating genetic diversity and structure patterns of H. hebetor populations from different moth hosts  (1) nnumber of individuals tested; Nsanumber of successful and unambiguous amplifi cations/locus/species; Nanumber of alleles; ASRallele size range; Nstotalnumber of successful loci amplifi ed/species; Npolnumber of polymorphic loci; "-" -irrelevant.
in different environments and at different spatial scales around the world.
The low level of genetic differentiation among samples separated by distance of ≥ 80 km suggests intense gene fl ow at a regional scale.We are currently using our suite of microsatellite loci to genotype: (1) a large number of population samples collected at various locations in Niger and bordering countries of Sub-Saharan Africa to determine the large scale spatial genetic structure of this parasitic wasp species, as well as (2) some laboratory strains that were earlier released as part of biological control programmes against moth pests of stored products in order to enable monitoring of their genetic structure and population dynamics within the agroecosystem.Preliminary results (data to be presented in full elsewhere) reveal that these new loci provide useful information on H. hebetor dispersion at the fi eld scale as it seems to be possible to locally differentiate some laboratory strains and wild strains, which will, we hope, eventually allow us to design more effi cient biocontrol strategies against the target moth pest species in question.Furthermore, some of the loci here tested appear useable in population genetics studies involving related species such as B. brevicornis, and to a lesser extent, B. cephi and B. lissogaster, and be useful in distinguishing some of them in mixed populations.

Table 1 .
Characteristics of the microsatellite loci selected in Habrobracon hebetor.Allelic diversity, heterozygosity and null allele frequencies were estimated in 46 female specimens collected from two natural populations in Niger (Dantchandou n = 24, Tondikoarey n = 22).