Introduction
The Lepidoptera are one of the world's most diverse insect groups with important roles within ecological, agricultural and horticultural settings. Developing effective and reliable DNA markers (e.g. microsatellites) is especially desirable in lepidopteran insect pests for population genetic structure studies, as well as for mapping economically and agriculturally important traits (e.g. insecticide and allelochemical resistance). Recent studies involving full genome sequence analysis of the silk moth Bombyx mori have identified numerous microsatellite DNA loci (Reddy et al., Reference Reddy, Abraham and Nagaraju1999; Prasad et al., Reference Prasad, Muthulakshmi, Arunkumar, Madhu, Sreenu, Pavithra, Bose, Nagarajaram, Mita, Shimada and Nagaraju2005a,Reference Prasad, Muthulakshmi, Madhu, Archak, Mita and Nagarajub), and about 550 of these were used for mapping purposes and establishing a genetic linkage map (Miao et al., Reference Miao, Xub, Li, Li, Huang, Dai, Marino, Mills, Zeng, Mita, Jia, Zhang, Liu, Xiang, Guo, Xu, Kong, Lin, Shi, Lu, Zhang, Huang, Yasukochi, Sugasaki, Shimada, Nagaraju, Xiang, Wang, Goldsmith, Lu, Zhao and Huang2005). The use of microsatellite DNA enrichment protocols in various lepidopteran species has also led to substantial microsatellite DNA markers being developed (e.g. Daly et al., Reference Daly, Waltham, Mulley, Watts, Rosin, Kemp and Saccheri2004; Wardill et al., Reference Wardill, Scott, Graham and Zalucki2004; Zhou et al., Reference Zhou, Gu and Dorn2005). The general perception that developing and designing microsatellite DNA markers is difficult for Lepidoptera nevertheless remains (Keyghobadi et al., Reference Keyghobadi, Roland and Strobeck1999; Nève & Meglécz, Reference Nève and Meglécz2000; Zhang, Reference Zhang2004) and appears, in part, to be due to redundancy in the microsatellite DNA flanking regions across different loci (Meglécz et al., Reference Meglécz, Petenian, Danchin, D'Acier, Rasplus and Faure2004, Reference Meglécz, Anderson, Bourguet, Butcher, Caldas, Cassel-Lundhagen, d'Acier, Dawson, Faure, Fauvelot, Franck, Harper, Keyghobadi, Kluetsch, Muthulakshmi, Nagaraju, Patt, Péténian, Silvain and Wilcock2007; Van't Hof et al., Reference Van't Hof, Brakefield, Saccheri and Zwaan2007), the lack of polymorphisms (e.g. Prasad et al., Reference Prasad, Muthulakshmi, Madhu, Archak, Mita and Nagaraju2005b; Van't Hof et al., Reference Van't Hof, Zwaan, Saccheri, Daly, Bot and Brakefield2005) and possible evolutionary associations with mobile elements (Ji & Zhang, Reference Ji and Zhang2004; Meglécz et al., Reference Meglécz, Petenian, Danchin, D'Acier, Rasplus and Faure2004, Reference Meglécz, Anderson, Bourguet, Butcher, Caldas, Cassel-Lundhagen, d'Acier, Dawson, Faure, Fauvelot, Franck, Harper, Keyghobadi, Kluetsch, Muthulakshmi, Nagaraju, Patt, Péténian, Silvain and Wilcock2007; Zhang, Reference Zhang2004; Van't Hof et al. Reference Van't Hof, Brakefield, Saccheri and Zwaan2007).
The noctuid moth Helicoverpa armigera has a wide geographic distribution all over the Old World (Hardwick, Reference Hardwick1965), and together with the closely related H. assulta, H. punctigera, and the New World H. zea, represents one of the world's most devastating agricultural lepidopteran pests (Hardwick, Reference Hardwick1965; Mitter et al., Reference Mitter, Poole and Matthews1993). Microsatellite markers for H. armigera have been developed by Tan et al. (Reference Tan, Chen, Zhang and Li2001), Ji et al. (Reference Ji, Zhang, Hewitt, Kang and Li2003, Reference Ji, Wu and Zhang2005) and Scott et al. (Reference Scott, Lange, Scott and Gahan2004). A sub-set of markers developed by Scott et al. (Reference Scott, Lange, Scott and Gahan2004) was applied to study Australian populations (Scott et al., Reference Scott, Wilkinson, Merritt, Scott, Lange, Schutze, Kent, Merritt, Grundy and Graham2003, Reference Scott, Lawrence, Lange, Scott, Wilkinson, Merritt, Miles, Murray and Graham2005a,Reference Scott, Wilkinson, Lawrence, Lange, Scott, Merritt, Lowe and Grahamb, Reference Scott, Lawrence, Lange, Graham, Hardwick, Rossiter, Dillon and Scott2006). Endersby et al. (Reference Endersby, Hoffmann, Mckechnie and Weeks2007) utilised microsatellite markers reported by both Ji et al. (Reference Ji, Zhang, Hewitt, Kang and Li2003) and Scott et al. (Reference Scott, Lange, Scott and Gahan2004) in further studies of Australian H. armigera. The authors reported amplification difficulties in samples collected both from Victoria and Queensland and observed varying degrees of allele dropouts (i.e. individuals scored as homozygotes once but subsequently as heterozygotes in replicate genotyping) in the markers tested, with null alleles (i.e. allele(s) that failed to amplify within an individual because of mutations at primer annealing sites) being the most likely underlying factor for these allele dropouts. Ji et al. (Reference Ji, Wu and Zhang2005) also reported the likely presence of null alleles in microsatellite markers isolated (e.g. HarSSR5 and HarSSR7), as well as the presence of multiple-copy microsatellite DNA families (fig. 1 in Ji & Zhang, Reference Ji and Zhang2004). Technical problems in various H. armigera microsatellite DNA loci developed to-date are, therefore, evident although the underlying factors have not been examined in detail.
Examples of size homoplasies in the Helicoverpa armigera EPIC markers. (a) RpL3 EPIC-DNA marker detected size homoplasy in the 145 bp amplicons from H. armigera samples AD6 (EU190361), AD4 (EU190362) and AD21 (EU190360). (b) RpS6 EPIC-DNA marker detected size homoplasy in the 269 bp alleles in H. armigera AD20 (EU190410) and AD24 (EU190411) samples. Identical bases are represented by (.), gaps inserted for alignment of sequences due to presence of indels are represented by ‘-’. SNPs detected between the homoplasious alleles are shown. Boxed regions are partial RpL/RpS exons, primer sequences are underlined.
Due to the wide distribution and migration ability of H. armigera, reliable molecular genetic markers that enable population comparisons within and between countries are required. Furthermore, such markers would be even more useful if they could be applied to related pest species within the genus (i.e. H. zea, H. assulta and H. punctigera). Exon-Primed Intron-Crossing (EPIC) PCR markers (Lessa, Reference Lessa1992; Palumbi & Baker, Reference Palumbi and Baker1994; Palumbi, Reference Palumbi, Hillis, Moritz and Mable1996) have the potential to meet these requirements. Developing EPIC markers can be challenging, especially in organisms that lack sufficient coding DNA sequence data, although EPIC markers have been described for conserved nuclear genes in a limited taxonomic range of organisms (reviewed in Palumbi, Reference Palumbi, Hillis, Moritz and Mable1996).
In additional to previously described ‘universal’ EPIC markers (Palumbi, Reference Palumbi, Hillis, Moritz and Mable1996), ribosomal protein (Rp) genes are also suitable candidates. Rp genes function as house keeping genes and are highly conserved in genomes across a wide range of organisms. In Lepidoptera, the majority of Rp genes exist as single copy genes (Lee, Reference Lee2006). Lee (Reference Lee2006) further showed that in B. mori, which represents a model organism in Lepidoptera genetics, at least 69 of the 80 Rp genes contain intron(s). The use of ribosomal protein genes as EPIC markers in population genetic studies has been reported (e.g. Gaffney, Reference Gaffney2000). Polymorphic EPIC markers based on single copy nuclear genes are robust both for use as markers in evolutionary and population genetic studies (He & Haymer, Reference He and Haymer1997; Garrick & Sunnucks, Reference Garrick and Sunnucks2006; Hubert et al., Reference Hubert, Duponchelle, Nunez, Rivera and Renno2006) and for gene mapping purposes (Lee, Reference Lee2006; Yasukochi et al., Reference Yasukochi, Ashakumary, Baba, Yoshido and Sahara2006; Pringle et al., Reference Pringle, Baxter, Webster, Papanicolaou, Lee and Jiggins2007). The occurrence of allele dropout and null alleles in EPIC markers is expected to be minimal, since primer annealing sites are specifically designed from evolutionary conserved coding regions (Palumbi, Reference Palumbi, Hillis, Moritz and Mable1996). We report here the development of H. armigera EPIC-PCR markers through comparative genomic analysis. We further provide molecular characterisation of EPIC length polymorphisms, establish Mendelian inheritance patterns and demonstrate robust PCR amplification in related pest Helicoverpa species.
Materials and methods
EPIC marker design and optimisation
We used comparative genomic analysis to develop EPIC-PCR markers for H. armigera. Ribosomal protein gDNA sequences from B. mori were aligned with ESTs, cDNA and mRNA of Rp gene sequences from publicly available sequences from B. mori, Mamestra brassicae, Lonomia oblique, Papilio dardanus, Plutella xylostella, Spodoptera frugiperda, H. zea and H. armigera. Helicoverpa species dopa decarboxylase (DDC) cDNA sequences were as reported by Fang et al. (Reference Fang, Cho, Regier, Mitter, Matthews, Poole, Friedlander and Zhao1997) and aligned with B. mori DDC gDNA. Intron/exon boundaries of gDNA versus cDNA were determined using the SPIDEY tool from NCBI <http://www.ncbi.nlm.nih.gov/IEB/Research/Ostell/Spidey/>, selecting Drosophila option as the model organism for all input sequences. Introns identified in B. mori as having approximate lengths of between 80–350 bp were selected for EPIC primer design.
Primers were designed to adjacent exons of single copy nuclear genes. Primers were designed for minimal primer-dimer and heteroduplex formation, and for minimal false priming sites between primers and template DNA using the Oligo Primer Analysis Software v6.40 (Molecular Biology Insights, Inc., Cascade, Colorado, USA). Primers were tested by gradient PCR (at 45°C to 55°C; 12 increments, with 6th and 7th increments set at 50°C) in an EPPENDORF Master Cycler gradient PCR machine (5331). EPIC primers that gave distinct amplicons of expected sizes were subsequently tested for intron length polymorphisms in H. armigera samples from Australia (AD1–AD24) and China (CH1–CH24) on 6% polyacrylamide gels.
EPIC DNA marker allele characterisation
Randomly selected alleles from six EPIC markers exhibiting higher (DDC, RpL29 and RpS6) and lower (RpL11, RpL12 and RpL3) numbers of alleles were characterised by DNA sequencing from Dalmore (Australia) and China H. armigera samples using the appropriate PCR primer annealing temperature profiles (table 1). Amplicons were gel-purified, cloned and sequenced. Only indels (insertions/deletions) were scored, while single nucleotide polymorphisms (SNPs) identified were not used for allele characterisation.
EPIC markers developed from Dopa Decarboxylase (DDC) and ribosomal protein (Rp) genes for H. armigera. GCG codes used (R=G/A, K=G/T, Y=C/T, M=C/A, W=A/T, S=G/C) in degenerate primers. Primer annealing temperature (Ta), observed heterozygosity (Ho), expected heterozygosity (He) for individual primer pairs are provided. GenBank Identifier (GI) for Bombyx mori gDNA and various lepidopteran ribosomal proteins mRNA used in comparative genomic method of primer design are also provided. The number of alleles detected for the characterised EPIC-markers in five H. assulta (Hs), 20 H. zea (Hz) and 20 H. punctigera (Hp) are reported.
‘ND’, not determined, ‘†’, possible gene duplication due to detection of >2 alleles within individual H. punctigera samples.
Gene copy number, allele inheritance pattern and cross-species amplification
EPIC-PCR markers RpL12, RpL10, RpS15A, DDC, RpL29 and RpS6 were tested for Mendelian inheritance patterns in the H. armigera mapping families Ha304 or Ha929 (for details of mapping crosses, see Lee, Reference Lee2006). The single copy gene status and Mendelian inheritance patterns for all other Rp genes listed in table 1 have been ascertained by Lee (Reference Lee2006) using southern hybridization, single strand conformation polymorphism (SSCP), PCR-RFLP and PCR heteroduplex analysis. The EPIC-PCR markers DDC, RpS6, RpL29, RpL3 and RpL11 were further tested for cross-species PCR amplification in H. assulta (n=5, Ha1-5), H. punctigera (n=20, HP1-20) and H. zea (n=20, HZBR1-20) sampled from India, Australia and Brazil respectively.
DNA extraction, PCR amplification, cloning and sequencing
DNA samples from Australia (AD1–AD24), China (CH1–CH24) and India (ICY1–ICY6) as reported in Behere et al. (Reference Behere, Tay, Russell, Heckel, Appelton, Kranthi and Batterham2007) were used to determine PCR robustness of EPIC markers. PCR amplification was carried out in a 25 μl reaction volume using 50–75 ng gDNA; 2.5 mm MgCl2, 2.5 μl of 10× PCR reaction buffer; 0.625 units Taq DNA polymerase (Promega, cat. #M186A); 0.5 μm of each forward and reverse primers (table 1) and 0.2 mm dNTPs. The PCR amplification profile consisted of a 5 min DNA denaturing step at 95°C (1 cycle); 35 cycles of 95°C, 50°C and 72°C (1 minute each) and a final 10 min extension at 72°C (1 cycle). All PCR products were resolved on 1.5% or 1.8% agarose gels stained with Ethidium Bromide and visualized over a UV light illuminator. PCR amplicons were gel-purified using Wizard® SV Gel and PCR Clean-Up System (Promega, cat. #A9282) according to the suppliers' manuals. Purified DNA was ligated into pGEM-T easy vector (Promega, cat. #A1380) prior to transformation in JM109 competent cells (Promega, cat. #L1001). Transformed cells were grown on LB agar plates supplemented with 0.005% Ampicillin, 0.008% X-gal (5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside) and 0.5 mm ITPG (Isopropyl β-D-1-thiogalactopyranoside) at 37°C for 16–18 h. Up to seven positively transformed colonies from each transformation experiment were lysed in 75 μl DNAse-free water (Sigma®) in individual sterile 1.7 ml Eppendorf tubes. Lysed colonies were kept frozen at −20°C until needed in PCR amplification as DNA template using universal SP6 and T7 primers in the standard 25 μl PCR amplification reaction. Sequencing of purified PCR amplicons used the universal sequencing primers SP6 and T7 and the Big Dye Terminator sequencing reaction V3.1 kit (Applied Biosystems). Post sequencing reaction clean up of samples followed the instructions as supplied by the Australian Genome Research Facility (AGRF) in Melbourne, Australia, where all samples were sequenced.
Polyacrylamide gel electrophoresis (PAGE)
Size polymorphisms were visualized on 6% PAGE gels using γ-33P dATP end-labelled forward primers following the protocol of Mohra et al. (Reference Mohra, Fellendorf, Segelbacher and Paxton2000) and Tay & Crozier (Reference Tay and Crozier2001). Gels were vacuum-dried at 70°C for 30 min prior to overnight exposure onto storage phosphor screens (Molecular Dynamics, Amersham BioSciences, cat. #63-0034-79) at room temperature and scanned using a Typhoon 8600 High Performance Laser Scanning System (Amersham Pharmacia Biotech). The numbers of alleles detected from individual loci on the 6% polyacrylamide gels were determined.
Post sequencing and PAGE EPIC DNA marker analyses
DNA sequences were analysed using the Pregap4 and Gap4 programs within the Staden Molecular DNA analysis software (Staden et al., Reference Staden, Beal and Bonfield2000). Gap4 assembled consensus sequences were imported into CLC Free Workbench 3.2.1 (CLC bio, Aarhus, Denmark) for sequence alignment. Sequence identity was determined by BLASTN search (Altschul et al., Reference Altschul, Madden, Shäfer, Zhang, Zhang, Miller and Lipman1997) against the nr DNA database deposited in GenBank. Expected levels of heterozygosity (He), observed heterozygosity (Ho) and deviation from Hardy-Weinberg equilibrium in EPIC markers DDC, RpS6, RpL3, RpL11 and RpL29 were calculated using a web version of the population genetic software Genepop v3.4 (Raymond & Rousset, Reference Raymond and Rousset1995; <http://genepop.curtin.edu.au>). We used EnzymeX v3.1 <http://mekentosj.com/enzymex/>to further characterise homoplasious alleles detected within the Rp and DDC EPIC-PCR markers.
Results
Ribosomal protein EPIC-PCR markers
A total of 45 ribosomal protein genes from H. armigera were tested in EPIC-PCR marker design. We identified a total of 11 Rp genes with suitable, polymorphic intron lengths. A subset of these EPIC-PCR markers that exhibited comparable levels of observed heterozygosity as microsatellite DNA loci were further characterised by DNA sequence analyses and cross-species PCR amplification tests (table 1). Of the 45 Rp genes tested, six were monomorphic in the samples tested, while the remaining 28 either had intron sizes that were too large (i.e. >400 bp PCR amplicons) or failed to amplify reliably. A list of all failed or monomorphic Rp genes tested and their primer sequences are available on request.
Allele characterisation in Rp EPIC DNA markers
Examples of allelic size homoplasy (i.e. alleles sharing identical amplicon size but differing in indel and SNP composition) were detected in RpL3, RpS6 and DDC by DNA sequence characterisation (table 2). In RpL3, three 145 bp alleles from H. armigera AD4, AD6 and AD21 samples were characterised that showed the presence of four indels between AD4 (EU190362), AD6 (EU190361) and AD21 (EU190360). In RpS6, allele size homoplasy was identified for the 269 bp allele from AD24 (EU190411) and AD20 (EU190410) due to two indels at nucleotide positions 63 and 165. In DDC, size homoplasy was detected for the 207 bp allele between samples AD3 (EU190407), AD5 (EU190406), AD7 (EU190405), AD8 (EU190408) and AD1 (EU190409) due to two indels at nucleotide positions 80 and 118. Two examples of allele size homoplasy detected in RpL3 and RpS6 are provided (fig. 1). Homoplasious alleles in RpL3, RpS6 and DDC EPIC markers are all differentiable by RFLPs (table 3). No allele size homoplasy, due to indels, was detected for the remaining 11 alleles in the six EPIC markers where multiple H. armigera samples were sequenced. All alleles characterised by DNA sequencing have been deposited in GenBank (Accession Numbers EU190360–EU190424).
Helicoverpa armigera EPIC DNA markers and alleles characterised by sequencing. The numbers of clones sequenced from individual H. armigera samples are indicated. Allele size homoplasy detected in RpL3 (allele 145 bp: AD6, AD4 and AD21), RpS6 (allele 269 bp: AD20 and AD24) and DDC (allele 207: AD3, AD5, AD7, AD8 and AD1). Gaps indicated by [-], Dalmore (Australia) and China samples have the sample ID prefix ‘AD’ and ‘CH’, respectively.
Examples of expected restriction fragment length polymorphisms (RFLPs) of homoplasious alleles detected in EPIC-PCR markers of Helicoverpa armigera. For each homoplasious allele within each of the three markers, RFLP patterns generated by two different restriction enzymes are provided. RFLPs are not determined for non-homoplasious alleles. Note that visualisation of the small fragment sizes generated by DdeI restriction digest of EU190362 should be carried out using 6% polyacrylamide gel electrophoresis.
Cross-species amplification of EPIC markers
EPIC-PCR markers DDC, RpL11, RpL29, RpL3 and RpS6 were tested in H. zea (n=20), H. punctigera (n=20) and H. assulta (n=5) using conditions specified for H. armigera (table 1). No further optimisation, of PCR and 6% PAGE conditions, was necessary. The DDC-EPIC f1/r1 marker is also one of the most polymorphic markers in H. zea with four easily identifiable alleles. In two other highly polymorphic EPIC-PCR markers (RpS6 and RpL29), relatively high numbers of alleles were detected in H. punctigera (16 and 8 alleles, respectively); however, in H. zea of similar sample size, numbers of alleles from these two EPIC-PCR markers were surprisingly low (one and two alleles, respectively) (fig. 2, note that not all alleles detected were shown).
Six percent PAGE analysis of five EPIC markers in pest Helicoverpa species. Rows A, B, C and D are H. armigera, H. zea, H. punctigera and H. assulta, respectively. EPIC markers are least polymorphic in H. zea (row B) and H. assulta (row D). Note that the DDC gene in H. punctigera is potentially duplicated due to the detection of multiple (>2) bands per individual by the DDC EPIC DNA marker.
Mendelian inheritance and gene copy number
With the exception of three EPIC-PCR markers (RpL12, RpS15A and RpL10) that did not show observable length polymorphisms in the H. armigera mapping crosses Ha304 and Ha929, all remaining EPIC-PCR markers (table 1) demonstrated Mendelian inheritance patterns. Single copy gene status for individual loci was inferred from the parental alleles being present in single copy gene ratio among the F2 individuals (Lee, Reference Lee2006). No allele dropout or null alleles were detected in these EPIC-PCR markers in the mapping families.
Discussion
We have shown that alternative nuclear DNA molecular genetic markers, such as EPIC-PCR markers, can be successfully designed in the noctuid moth H. armigera, where developing microsatellite DNA markers has been difficult. Screening of intron regions detected length polymorphisms in 11 Rp genes and the nuclear gene DDC. We further carried out sequence characterisation of various alleles in three of the most polymorphic loci (DDC, RpL29 and RpS6) and in three loci that showed lower levels of allele length polymorphisms (RpL3, RpL11 and RpL12). Mutation characterisation of randomly selected alleles from these six loci indicated that sequence length polymorphisms were generally due to random insertions and deletions, although size homoplasy was detected in three instances.
DNA sequencing allows the identification and fine-scale characterisation of homoplasious alleles with respect to the presence of SNPs and/or indels. Homoplasious alleles, thus identified, could be further differentiated through RFLPs patterns (table 3) or through DNA fragment conformation analytical methods, such as SSCP coupled with denaturing gradient gel electrophoresis (Palumbi, Reference Palumbi, Hillis, Moritz and Mable1996; Estoup et al., Reference Estoup, Jarne and Corneut2002). We have provided examples of differentiating homoplasious alleles by RFLPs; however, the treatment of SNPs within introns in individual EPIC DNA loci as distinctive alleles should proceed with caution. Differentiating between real mutations and DNA polymerase introduced errors (e.g. O'Mahony et al., Reference O'Mahony, Tay and Paxton2007) would require intensive multiple sequencing and was beyond the scope of our study. Allele length homoplasies, due to SNPs and/or indels, are also common in microsatellite DNA markers, both at the flanking sequence regions and within the microsatellite DNA repeat units (reviewed in Jarne & Lagoda, Reference Jarne and Lagoda1996). Size homoplasies in microsatellite alleles have also been identified through DNA sequencing in Australian H. armigera individuals from the Dalmore population (G.T. Behere, data not shown).
Microsatellite allele size homoplasy may not significantly affect various population genetic parameter estimations such as Wright's (Reference Wright1951)F-statistics (but see O'Reilly et al. Reference O'Reilly, Canino, Bailey and Bentzen2004), relatedness estimates, the model of population isolation by distance and phylogenetic reconstruction of closely related populations (reviewed in Estoup et al., Reference Estoup, Jarne and Corneut2002). Size homoplasy can, however, lower the level of observed heterozygosity, thereby affecting the power of exact tests on genotypic linkage disequilibrium and Hardy-Weinberg equilibrium (see Estoup et al., Reference Estoup, Jarne and Corneut2002). Furthermore, strong constraints on allele size can increase size homoplasy leading to inaccurate gene flow estimates between sub-populations (Paetkau et al., Reference Paetkau, Waits, Clarkson, Craighead and Strobeck1997; Gaggiotti et al., Reference Gaggiotti, Lange, Rassmann and Gliddon1999) and can cause downward bias in F ST and/or R ST estimates of population differentiation (O'Reilly et al., Reference O'Reilly, Canino, Bailey and Bentzen2004). We observed a random pattern of indels within the intron regions of Rp and DDC EPIC markers, low frequencies of size homoplasy, the absence of null alleles in mapping crosses and Hardy-Weinberg equilibrium for all markers. These observations indicate that these EPIC markers will be useful in population and evolutionary genetic studies in H. armigera and related species. The accuracy of this assumption should be further examined based on complementary data from different types of molecular markers (Queney et al., Reference Queney, Ferrand, Weiss, Mougel and Monnerot2001; Zhang & Hewitt, Reference Zhang and Hewitt2003).
Although allele homoplasy may not necessarily affect the applicability of microsatellite loci in estimating various population genetic parameters, other problems, such as presence of null alleles, allele dropout or the occurrence of multiple alleles for a specific locus within an individual (due to high copy number of microsatellite flanking sequences) can adversely affect population genetic data analysis (Endersby et al., Reference Endersby, Hoffmann, Mckechnie and Weeks2007), problems that are especially pronounced in certain organisms (e.g. Aedes egypti, Chambers et al., Reference Chambers, Meece, McGowan, Lovin, Hemme, Chadee, McAbee, Brown, Knudson and Severson2007; Anopheles gambiae, Meglécz et al., Reference Meglécz, Anderson, Bourguet, Butcher, Caldas, Cassel-Lundhagen, d'Acier, Dawson, Faure, Fauvelot, Franck, Harper, Keyghobadi, Kluetsch, Muthulakshmi, Nagaraju, Patt, Péténian, Silvain and Wilcock2007) including various lepidopteran species. In H. armigera, microsatellite loci that exhibited multiple alleles within single individuals have been detected (e.g. HaB60, HaD47, HarSSR9, HarSSR10; G.T. Behere, data not shown) and reported (Ji & Zhang, Reference Ji and Zhang2004; Grasela & McIntosh, Reference Grasela and McIntosh2005; Ji et al., Reference Ji, Wu and Zhang2005; Endersby et al., Reference Endersby, Hoffmann, Mckechnie and Weeks2007). The presence of multiple (i.e. >2) alleles from specific microsatellite loci in a diploid individual suggested that the loci may be present in multiple copies and, therefore, unsuitable for use in population genetic analysis. Multi-copy microsatellite loci, due to similarity at the flanking sequence regions in many organisms, have been linked to or suspected to have evolutionary associations with mobile elements or are due to unequal recombination events (Arcot et al., Reference Arcot, Wang, Weber, Deininger and Batzer1995; Nadir et al., Reference Nadir, Margalit, Gallily and Ben-Sasson1996; Ramsay et al., Reference Ramsay, Macaulay, Cradle, Morgante, degli Ivanissevich, Maestri, Powell and Waugh1999; Akagi et al., Reference Akagi, Yokozeki, Inagaki, Mori and Fujimura2001; Wilder & Hollocher, Reference Wilder and Hollocher2001; Meglécz et al., Reference Meglécz, Petenian, Danchin, D'Acier, Rasplus and Faure2004, Reference Meglécz, Anderson, Bourguet, Butcher, Caldas, Cassel-Lundhagen, d'Acier, Dawson, Faure, Fauvelot, Franck, Harper, Keyghobadi, Kluetsch, Muthulakshmi, Nagaraju, Patt, Péténian, Silvain and Wilcock2007; Van't Hof et al., Reference Van't Hof, Brakefield, Saccheri and Zwaan2007).
Null alleles in H. armigera microsatellite loci have represented a major challenge to investigating population genetic structures of this pest species (Ji et al., Reference Ji, Wu and Zhang2005; Endersby et al., Reference Endersby, Hoffmann, Mckechnie and Weeks2007). Anonymous nuclear DNA markers, such as microsatellite DNA markers, are especially prone to null alleles (Tay et al., Reference Tay, Miettinen and Kaitala2003; Ji et al., Reference Ji, Wu and Zhang2005) because the non-coding primer annealing sites are unlikely to be evolutionary conserved, resulting in higher frequencies of random mutations such as SNPs and indels. EPIC-PCR markers can overcome the problems of multi-copy loci and null alleles associated with microsatellite loci. However, designing EPIC-PCR markers requires several steps, including isolation of genes, determination of gene copy numbers and identification of introns of appropriate size suitable for population genetic studies. These obstacles have hindered the development of widespread lepidopteran EPIC markers in the past. With the increasing availability of ESTs for comparative genome analysis, developing EPIC markers is expected to become more achievable in many non-model organisms.
The use of intron sequences in phylogenetic inferences has been demonstrated (Hedin & Maddison, Reference Hedin and Maddison2001; Allen & Omland, Reference Allen and Omland2003; Kawakita et al., Reference Kawakita, Sota, Ascher, Ito, Tanaka and Kato2003; Fujita et al., Reference Fujita, Engstrom, Starkey and Shaffer2004), and applying EPIC markers to population genetic structure investigations has been reported in many organisms including aquatic organisms and insects (Palumbi & Baker, Reference Palumbi and Baker1994; He & Haymer, Reference He and Haymer1997; Villablanca et al., Reference Villablanca, Roderick and Palumbi1998; Berrebi et al., Reference Berrebi, Boissin, Fang and Cattaneo-Berrebi2005, Reference Berrebi, Retif, Fang and Zhang2006; Hubert et al., Reference Hubert, Duponchelle, Nunez, Rivera and Renno2006; see also Roderick, Reference Roderick1996; Zhang & Hewitt, Reference Zhang and Hewitt2003); however, its use in population genetics of Lepidoptera is yet to be demonstrated. Similar substitution rates between microsatellite DNA flanking regions and introns have been reported (Brohede & Ellegren, Reference Brohede and Ellegren1999). Mutation rates of H. armigera EPIC-PCR intron regions should be compared with the mutation rates of H. armigera microsatellite DNA loci not affected by issues such as multi-copy alleles, the presence of null alleles or allele dropout. Similar mutation rates between EPIC DNA intron regions and non-coding simple sequence repeat regions (i.e. neutral microsatellite loci regions) will further support the suitability of EPIC markers as alternative markers for investigating H. armigera population genetic structures. Through H. armigera family crosses, Mendelian inheritance patterns in DDC, RpL29 and RpS6 EPIC markers were confirmed in this study. Furthermore, Lee (Reference Lee2006) has shown by southern hybridization, heteroduplex conformation, and EPIC-PCR analyses that the majority of Rp genes analysed in this study (table 1) also existed as single copy genes. The single copy gene status of RpL12, RpS10 and RpS15A has not been determined, as these EPIC markers were monomorphic in the H. armigera family crosses. These Rp genes are likely to exist as single copy genes, as shown by chromosome synteny analysis in B. mori, Ostrinia nubilalis and Heliothis virescens (Lee, Reference Lee2006). Nevertheless, utilising these three EPIC markers in population and evolutionary genetic studies should proceed with caution, as these genes may be affected by the presence of pseudogenes and/or gene duplication, both of which are known to affect some Rp genes (Lee, Reference Lee2006) while duplication of the DDC gene and surrounding region has been reported in spiders (Hedin & Maddison, Reference Hedin and Maddison2001) and Drosophila (Eveleth & Marsh, Reference Eveleth and Marsh1986) and suspected, based on 6% PAGE, in H. punctigera (table 1, fig. 2).
We have shown that five of the EPIC-PCR markers developed for H. armigera also amplified in H. assulta, H. punctigera and H. zea, and the numbers of alleles detected were generally higher in H. punctigera. That all H. assulta samples were found to be monomorphic in three of the five EPIC markers characterised was unexpected, although this is likely to reflect the small sample size (table 1). Low numbers of alleles were also detected in H. zea, as compared with H. armigera and H. punctigera, all of which consisted of similar numbers of individuals. This may reflect the finding of Behere et al. (Reference Behere, Tay, Russell, Heckel, Appelton, Kranthi and Batterham2007) that H. zea populations in North and South America were likely the results of H. armigera founder events at approximately 1.5 million years ago. Microsatellite DNA markers remained the marker system of choice and will continue to be powerful genetic markers in evolutionary, molecular ecological and population genetic studies for the majority of organisms. However, in organisms where the development and utilisation of microsatellite markers have consistently been problematic, EPIC DNA markers offer an alternative option. EPIC markers developed in this study will be valuable to the investigation of H. armigera population genetic structure, mating behaviour and population evolutionary history, thereby providing a much needed basis to better understand the population dynamics of one of the most significant lepidopteran pests in the world.
Acknowledgements
This project was supported by The Australian Research Council (ARC) through its funding of the Special Research Centre CESAR (Centre for Environmental Stress and Adaptation Research) and funding from The State Government of Victoria, Australia to WTT. GTB was supported by the Melbourne International Research Scholarship (MIRS) and Melbourne International Fee Remission Scholarship (MIFRS). SFL and WTT were supported by the Max-Planck-Gesellschaft. Ary Hoffman, Steve McKechnie, Adam Williams, Tamar Stzal and Nancy Endersby provided helpful discussions during the course of this study. Helicoverpa samples were kindly provided by Nancy Endersby, Stephen Cameron, Keshav Kranthi, Yidong Wu and Derek Russell.
