Bacterial Endosymbiont Infections in Living Fossils

Molecular Ecology Resources (2014) doi: 10.1111/1755-0998.12220

Bacterial endosymbiont infections in ‘living fossils’: a case study of North American vaejovid scorpions

ROBERT W. BRYSON JR Department of Biology & Burke Museum of Natural History and Culture, University of Washington, Box 351800 Seattle, WA 98195-1800, USA

Abstract Bacterial endosymbionts are common among arthropods, and maternally inherited forms can affect the reproductive and behavioural traits of their arthropod hosts. The prevalence of bacterial endosymbionts and their role in scorpion evolution have rarely been investigated. In this study, 61 samples from 40 species of scorpion in the family Vaejovi- dae were screened for the presence of the bacterial endosymbionts Cardinium, Rickettsia, Spiroplasma and Wolbachia. No samples were infected by these bacteria. However, one primer pair specifically designed to amplify Rickettsia amplified nontarget genes of other taxa. Similar off-target amplification using another endosymbiont-specific primer was also found during preliminary screenings. Results caution against the overreliance on previously published screening primers to detect bacterial endosymbionts in host taxa and suggest that primer specificity may be higher in primers targeting nuclear rather than mitochondrial genes.

Keywords: Cardinium, DNA barcoding, mitochondrial DNA, Rickettsia, scorpion, Spiroplasma, Vaejovidae, Wolbachia Received 6 September 2012; revision received 2 December 2013; accepted 8 December 2013

mission. In the light of these interactions, endosymbionts Introduction are now regarded as major drivers of arthropod ecology Maternally inherited bacterial endosymbionts are com- and evolution (Engelstadter & Hurst 2009; Fellous et al. mon among arachnids. In spiders, for example, around 2011; Jiggins & Hurst 2011). 20% of all species may be infected by the endosymbiont The prevalence of bacterial endosymbionts and their Cardinium (Duron et al. 2008; Perlman et al. 2010; Stefa- role in scorpion evolution have rarely been investigated. nini & Duron 2012). This bacterium has also been docu- To date, three studies have documented Wolbachia infec- mented at a high frequency in Acari, Opiliones and tions in three families of scorpions (Buthidae, Suesdek- Scorpiones (Martin & Goodacre 2009). Four frequently Rocha et al. 2006; Scorpionidae, Baldo et al. 2007; and studied bacterial endosymbionts have been found in Hemiscorpiidae, Baradaran et al. 2011). The presence of arachnids, including Cardinium, Rickettsia, Spiroplasma Cardinium in scorpions has been documented once and Wolbachia (Goodacre et al. 2006; Martin & Goodacre (Hemiscorpiidae, Martin & Goodacre 2009). In this 2009). study, a diverse sampling of species in the speciose Maternally inherited bacterial endosymbionts are North American scorpion family Vaejovidae was known to affect reproductive and behavioural traits of screened for the presence of Cardinium, Rickettsia, Spiropl- their arthropod hosts (Engelstadter & Hurst 2009). The asma and Wolbachia, four bacterial endosymbionts com- effects of many of these endosymbionts such as Wolbachia monly reported in arachnids. DNA was extracted from or Cardinium can include skewing the sex ratio towards three types of scorpion tissues and, for several species, females (through male killing, feminization and parthe- from multiple individuals. A total of 61 samples were nogenesis) and cytoplasmic incompatibility, both of screened using previously published primers speciﬁcally which result in differential reproductive success of designed to amplify genes in targeted bacterial endos- infected versus noninfected females. These mechanisms ymbionts. Although not intended to be an exhaustive are strong selective pressures on bacterial endos- study, results caution against the overreliance on previ- ymbionts which are reliant upon vertical transmission ously published screening primers to detect bacterial through maternal lineages to increase their own trans- endosymbionts in host taxa and suggest that primer speciﬁcity may be higher in primers targeting nuclear Correspondence: Robert W. Bryson Jr, Fax: 206-685-3039; E-mail: [email protected] rather than mitochondrial genes.

Materials and methods somatic tissues (Dobson et al. 1999; Martin & Goodacre 2009), and infection rates may be low within populations Forty species of scorpion representing seven of the 10 (Jiggins et al. 2001; Duron et al. 2008; Zug & Hammer- genera within the family Vaejovidae (following Prendini stein 2012). Therefore, DNA was also extracted from the & Wheeler 2005) were screened (Table 1). Genomic DNA gonads and surrounding tissues of 15 additional speci- was extracted from leg tissues of 39 specimens using the mens, from seven whole ﬁrst- or second-stage instars QIAGEN DNeasy Blood and Tissue Kit (Qiagen) follow- and from multiple individuals of eight species (Table 1). ing manufacturer’s recommendations. Maternally inher- In total, 61 samples were screened. ited bacterial endosymbionts may be absent from

Table 1 Vaejovid scorpions screened for the bacterial endosymbionts Cardinium, Rickettsia, Spiroplasma and Wolbachia. Taxonomy fol- lows Prendini & Wheeler (2005)

Species Locality Tissue type (n)

Paruroctonus arenicola USA: California: San Bernardino Co. Leg (1) Paruroctonus becki USA: Arizona: Mohave Co. Leg (1) Paruroctonus boreus USA: Nevada: Elko Co. Leg (1) Paruroctonus gracilior USA: Arizona: Cochise Co. Leg (1) Paruroctonus luteolus USA: California: Inyo Co. Leg (1) Paruroctonus pecos USA: New Mexico: Quay Co. Leg (1) Paruroctonus shulovi USA: Nevada: Clark Co. Leg (1) Paruroctonus silvestrii USA: California: Riverside Co. Leg (1) Paruroctonus utahensis USA: New Mexico: Sante Fe Co. Leg (1) Paruroctonus variabilis USA: California: Fresno Co. Leg (1) Paruroctonus xanthus USA: Arizona: Yuma Co. Leg (1) Pseudouroctonus apacheanus USA: Arizona: Cochise Co. Leg (1) P. apacheanus USA: Arizona: Graham Co. Instar (2) Pseudouroctonus reddelli USA: Texas: Edwards Co. Leg (1) P. reddelli USA: Texas: Bexar Co. Instar (2) Pseudouroctonus williamsi USA: California: San Diego Co. Leg (1) Serradigitus allredi USA: Arizona: Pinal Co. Leg (1) Serradigitus gertschi USA: California: Fresno Co. Leg (1) Serradigitus joshuaensis USA: Arizona: Mohave Co. Leg (1) Serradigitus miscionei USA: Arizona: Pima Co. Leg (1) Serradigitus subtilimanus USA: Arizona: Mohave Co. Leg (1) Serradigitus wupatkiensis USA: Nevada: Clark Co. Leg (1) Smeringurus mesaensis USA: Arizona: Mohave Co. Leg (1) Smeringurus vachoni USA: California: Inyo Co. Leg (1) S. vachoni USA: Nevada: Clark Co. Gonad (2) Uroctonites huachuca USA: Arizona: Santa Cruz Co. Instar (1) U. huachuca USA: Arizona: Cochise Co. Instar (1) Uroctonus franckei USA: California: Inyo Co. Leg (1) Uroctonus mordax USA: California: Marin Co. Leg (1) Vaejovis carolinianus USA: Georgia Leg (1), gonad (4) Vaejovis cashi USA: New Mexico: Hidalgo Co. Leg (1) V. cashi USA: Arizona: Cochise Co. Gonad (5) Vaejovis chisos USA: Texas: Brewster Co. Leg (1), gonad (4) Vaejovis confusus USA: Nevada: Clark Co. Leg (1) Vaejovis feti USA: Arizona: Socorro Co. Leg (1) Vaejovis franckei Mexico: Oaxaca: Cerro Corral del Piedra Leg (1) Vaejovis granulatus Mexico: Morelos: Zempoala Leg (1) Vaejovis hirsuticauda USA: Nevada: Clark Co. Leg (1) Vaejovis intermedius USA: Texas: Val Verde Co. Leg (1) Vaejovis lapidicola USA: Arizona: Coconino Co. Leg (1), instar (1) Vaejovis mumai USA: Arizona: Mohave Co. Leg (1) Vaejovis punctatus Mexico: Aguascalientes: Sierra Fria Leg (1) Vaejovis spinigerus USA: New Mexico: Hidalgo Co. Leg (1) Vaejovis tesselatus Mexico: San Luis Potosı: Alvarez Leg (1) Vaejovis waueri USA: Texas: Terrell Co. Leg (1)

Previously published primers specifically designed to Results amplify genes in targeted bacterial endosymbionts were used in PCRs. For Wolbachia screening, the Wolbachia-spe- No amplicons were produced using the Wolbachia, Cardi- cific primer pair Wspecf/Wspecr (16S-2; Simoes~ et al. nium or Spiroplasma primers. PCRs using the Rickettsia 2011) targeting the 16S rRNA gene was used. This primer primers produced single bands of the expected target detected infections by eight of nine known Wolbachia size in eight samples: Paruroctonus silvestrii (leg tissue), supergroups (Simoes~ et al. 2011), including the super- Paruroctonus utahensis (leg tissue), Serradigitus joshuaensis groups A, B and F previously detected in scorpions (leg tissue), Vaejovis chisos (gonads), Vaejovis feti (leg tis- (Baldo et al. 2007; Baradaran et al. 2011). To screen sam- sue), Vaejovis hirsuticauda (leg tissue), Vaejovis intermedius ples for Cardinium, the primers Car-sp-F/Car-2p-R (leg tissue) and Vaejovis lapidicola (instar). Readable (Nakamura et al. 2009) targeting the 16S rRNA gene and sequence data was obtained for four of these samples gyrF/gyrR (Stefanini & Duron 2012) targeting the gyrase (S. joshuaensis, V. chisos, V. feti and V. hirsuticauda; subunit B gene were used. These primer pairs previously Appendix S1, Supporting information). BLAST results amplified Cardinium in a diverse range of other arach- revealed that these four sequences did not match the tar- nids (Nakamura et al. 2009; Stefanini & Duron 2012). For get taxa. In fact, only one sequence matched any pub- Rickettsia and Spiroplasma PCR screenings, the primer lished gene sequence in the GenBank database (the pair RICS741F/RCIT1197R, designed to amplify the cit- sequence from V. hirsuticauda was an 84% match to a rate gene of Rickettsia (Davis et al. 1998), and the primer portion of the Enterobacter lignolyticus genome). Further, pair Spits-J04/Spits-N55, designed to amplify the inter- none of the four sequences were similar to each other. genic ribosomal spacer of the Spiroplasma ixodetis group Scorpion DNA was amplified from all specimens using (von der Schulenburg et al. 2000), were used. These the scorpion-specific 16S primers which indicated that primers detected infections in a variety of spiders (Good- DNA extractions were successful. PCRs using lower acre et al. 2006). One sample of a Wolbachia and Cardini- annealing temperatures with the Wolbachia, Cardinium um double-infected spider (Grammonota sp.) was used as and Spiroplasma primers failed to produce single bands a positive control in Wolbachia and Cardinium PCRs. No matching the positive control (for Wolbachia and Cardini- positive controls were available for the Rickettsia and Spi- um) or of the expected target size (for Spiroplasma) and roplasma PCRs. To ensure DNA from each scorpion was generally resulted in amplicons with multiple bands. of sufficient quality to amplify bacterial endosymbiont DNA, additional amplifications were performed using Discussion scorpion-specific 16S primers (Bryson et al. 2013). Genes were amplified via PCR in a mix containing False positives using screening primers 6.25 lL Takara ExTaq Polymerase Premix (Takara Bio Inc.), 4.25 lL double-distilled water, 0.5 lL each primer The incidence of bacterial endosymbionts in host taxa is (10 lM) and 1.0 lL of template DNA. DNA was dena- routinely determined using PCR assays and previously tured initially at 94 °C for 2.5 min, followed by 35 cycles published screening primers. In some studies, the pres- of amplification performed under the following condi- ence of a single band of expected size is taken as evi- tions: denaturation at 94 °C for 30 s, annealing at either dence for an infection (e.g. Martin & Goodacre 2009). 60 °C (16S-2), 56 °C (Car-sp-F/Car-2p-R and gyrF/gyrR) Using this criterion, eight samples of vaejovid scorpion or 55 °C (RICS741F/RCIT1197R and Spits-J04/Spits- screened using PCR primers designed to amplify the cit- N55) for 30 s and extension at 72 °C for 1 min; this was rate gene of Rickettsia (Davis et al. 1998) appeared to test followed by a final 10 min elongation at 72 °C. Ampli- positive for this bacterial endosymbiont. However, cons were checked by electrophoresis on a 2% agarose sequences from four of these products revealed that they gel, and reactions producing single bands of the same were not the targeted Rickettsia (Appendix S1, Support- size as the positive sample (for Wolbachia and Cardinium) ing information). In this case, results from the PCR assay or of the expected size range (300–500 bp, for Rickettsia alone would lead to false positives. and Spiroplasma) were subsequently sent to the High- Preliminary PCRs using two other screening primers Throughput Genomics Unit (HGTU, University of Wash- also resulted in false positives. The primer pair Ch-F/ ington) for sequencing. Sequences were then analysed Ch-R (Zchori-Fein & Perlman 2004) was initially tested using the basic local alignment search tool (BLAST) and to screen vaejovid samples for the presence of Cardinium. ‘megablast’ filtering criterion (Altschul et al. 1990) to This primer pair produced numerous amplicons with establish that they were from the targeted bacteria. PCRs multiple bands of various sizes as well as single bands of were repeated using lower annealing temperatures (5 °C various sizes and was therefore not used in final screen- and 10 °C lower) and a subset of 20 samples for primer ing runs. However, four of the single-banded prod- pairs that produced no amplicons. ucts (amplified from Paruroctonus variabilis, Smeringurus

© 2013 John Wiley & Sons Ltd 4 R. W. BRYSON mesaensis, Vaejovis carolinianus and Vaejovis spinigerus) Goodacre 2009; Stefanini & Duron 2012). However, in that matched the positive sample in size were subse- addition to Cardinium, Rickettsia, Spiroplasma and Wolba- quently sequenced and compared to similar sequences in chia, a number of other vertically acquired, endosymbi- the GenBank nucleotide database using the BLAST pro- otic bacteria are known to infect arthropods, including gram. Results revealed that products were the right gene Flavobacteria (Bacteroidetes), Arsenophonus (Proteobacte- (16S) but from off-target taxa (Appendix S1, Supporting ria), and a newly described Gamma-Proteobacteria information). All four were close matches to Bacteroide- (Goodacre & Martin 2012). The phenotypic effects of tes bacteria (96–100%) but were relatively weak matches these bacteria on their arthropod hosts are largely to Cardinium (83–84%). A previous study using this pri- unknown, but all probably strongly influence the repro- mer pair found similar results (Martin et al. 2013). ductive success of infected females (Cordaux et al. 2011; Simoes~ et al. (2011) recently published a new pair of Breeuwer et al. 2012; Goodacre & Martin 2012). Future 16S primers (16S-6) designed to screen arthropods for studies should continue to screen vaejovid scorpions to Wolbachia infections. Although these 16S-6 primers were determine whether bacterial endosymbionts are truly found to be Wolbachia-specific, or at least failed to absent. However, rather than employ repetitive PCR amplify the 16S rRNA gene in six closely related species screenings using multiple sets of primers, researchers of bacteria, these primers also provided false positives should consider the recent advances in detecting bacte- for vaejovid scorpions. Nearly all of the samples rial endosymbionts using next-generation sequencing screened during preliminary trials with the 16S-6 prim- approaches (e.g. Hirsch et al. 2012; Kautz et al. 2013). ers tested positive as evidenced by a single bright band These methods should reduce both false positives and in the agarose gel. However, all of these same samples false negatives and provide a more accurate quantifica- tested negative using the 16S-2 and the five MLST primer tion of endosymbiont prevalence. pairs (Baldo et al. 2006). Three samples amplified with As studies on bacterial endosymbionts in scorpions the 16S-6 primers were sequenced (Appendix S1, Sup- continue to grow, researchers barcoding and studying porting information). Results suggested that the 16S-6 the evolution of these ‘living fossils’ should carefully primers were amplifying the 16S gene of Sphingomonad- consider the potential genetic implications of endosymbi- aceae bacteria (sequences were 89–99% similar) probably ont infections. Although four types of bacterial endos- found on the feet of scorpions used in DNA extractions. ymbionts appeared to be absent in vaejovid scorpions in Indeed, nine samples extracted from gonads and internal this study, previous studies have documented endosym- tissues failed to produce bands when amplified using biont infections in other families of scorpions (Buthidae, the 16S-6 primers. Comparisons of the three sequences to Suesdek-Rocha et al. 2006; Scorpionidae, Baldo et al. Wolbachia 16S sequences using BLAST yielded similarities 2007; and Hemiscorpiidae, Martin & Goodacre 2009; Bar- of only 79–83%. adaran et al. 2011). It is important to note that as a mater- The false positives in this study highlight the potential nally inherited infection spreads, it carries along the for overreliance on previously published screening prim- mtDNA genotype that was initially associated with it. ers to detect the presence of bacterial endosymbionts in The genealogy of mtDNA in a host then reflects the host taxa. Interestingly, two of the three primer pairs that spread of the endosymbiotic bacteria rather than the produced false positives targeted the 16S gene. They cor- relatedness of host populations (Ballard & Rand 2005). rectly amplified the 16S gene, but of off-target bacteria. Such a pattern can lead to biases in the reconstruction of Conversely, only one of the three primer pairs targeting relationships between individuals inferred from mtDNA nuclear genes produced false positives. These results (Hurst & Jiggins 2005; Smith et al. 2012), which has led suggest that screening primers targeting slower-evolv- researchers to question the credibility of using mtDNA ing, and thus more conserved, nuclear genes in bacterial alone to study recent historical events in arthropods endosymbionts may have better primer specificity than (Hurst & Jiggins 2005). those primers targeting faster-evolving mitochondrial genes such as 16S. Acknowledgements

I thank R.F. Ayrey, T. Burkhardt, M.R. Graham, R.W. Hansen, Bacterial endosymbionts in scorpions T.D. Hibbitts and K.J. McWest for sharing scorpion tissues, and None of the 61 vaejovid scorpion samples screened J.A. White for providing DNA from infected spiders. Permission to collect V. chisos was kindly provided by the U.S. National appeared to be infected with Cardinium, Rickettsia, Park Service, and I thank R. Skiles (Big Bend National Park) for Spiroplasma or Wolbachia. This remarkable absence runs his assistance. Research in Mexico was conducted under permits counter to previous work that suggests some bacterial granted by SEMARNAT to the late F. Mendoza-Quijano. R.W. endosymbionts such as Cardinium may be relatively Hansen, E.-H. Kim, and four anonymous reviewers provided common in arachnids (Duron et al. 2008; Martin & insightful comments on earlier drafts of this manuscript.

References Kautz S, Rubin BER, Moreau CS (2013) Bacterial infections across the ants: frequency and prevalence of Wolbachia, Spiroplasma, and Asaia. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ (1990) Basic local Psyche, 2013,1–11. alignment search tool. Journal of Molecular Biology, 215, 403–410. Martin OY, Goodacre SL (2009) Widespread infections by the bacterial Baldo L, Hotopp Dunning JC, Jolley KA et al. (2006) Multilocus sequence endosymbiont Cardinium in arachnids. Journal of Arachnology, 37, 106– typing system for the endosymbiont Wolbachia pipientis. Applied and 108. Environmental Microbiology, 72, 7098–7110. Martin OY, Puniamoorthy N, Gubler A, Wimmer C, Germann C, Bernas- Baldo L, Prendini L, Corthals A, Werren J (2007) Wolbachia are present in coni MV (2013) Infections with the microbe Cardinium in the Dolicho- southern African scorpions and cluster with supergroup F. Current podidae and other Empidoidea. Journal of Insect Science, 13,1–13. Microbiology, 55, 367–373. Nakamura Y, Kawai S, Yukuhiro F et al. (2009) Prevalence of Cardinium Ballard JWO, Rand DM (2005) The population biology of mitochondrial bacteria in planthoppers and spider mites and taxonomic revision of DNA and its phylogenetic implications. Annual Review of Ecology, Evo- “Candidatus Cardinium hertigii” based on detection of a new Cardini- lution, and Systematics, 36, 621–642. um group from biting midges. Applied and Environmental Microbiology, Baradaran M, Jalali A, Jolodar A (2011) Molecular diagnosis of Wolbachia 75, 6757–6763. endosymbiont from Iranian scorpion Hemiscorpius lepturus using poly- Perlman SJ, Magnus SA, Copley CR (2010) Pervasive associations merase chain reaction (PCR) amplification of 16S rDNA gene. African between Cybaeus spiders and the bacterial symbiont Cardinium. Journal Journal of Biotechnology, 10, 19802–19806. of Invertebrate Pathology, 103, 150–155. Breeuwer H, Ros VID, Groot TVM (2012) Cardinium: the next addition to Prendini L, Wheeler WC (2005) Scorpion higher phylogeny and classifi- the family of reproductive parasites. In: Manipulative Tenants: Bacteria cation, taxonomic anarchy, and standards for peer review in online Associated with Arthropods (ed. Zchori-Fein E & Bourtzis K), pp. 207– publishing. Cladistics, 21, 446–494. 224. Taylor & Francis, Boca Raton, FL. von der Schulenburg HG, Majerus TM, Dorzhu CM, Zakharov IA, Bryson RW, Riddle BR, Graham MR, Smith BT, Prendini L (2013) As old Hurst GD, Majerus ME (2000) Evolution of male-killing Spiroplasma as the hills: montane scorpions in southwestern North America reveal (Procaryotae: Mollicutes) inferred from ribosomal spacer sequences. ancient associations between biotic diversification and landscape The Journal of General and Applied Microbiology, 46,95–98. history. PLoS One, 8, e52822. Simoes~ PM, Mialdea G, Reiss D, Sagot MF, Charlat S (2011) Wolbachia Cordaux R, Bouchon D, Greve P (2011) The impact of endosymbionts detection: an assessment of standard PCR protocols. Molecular Ecology on the evolution of host sex-determination mechanisms. Trends in Resources, 11, 567–572. Genetics, 27, 332–341. Smith MA, Bertrand C, Crosby K et al. (2012) Wolbachia and DNA barcod- Davis MJ, Ying Z, Brunner BR, Pantoja A, Ferwerda FH (1998) Rickettsial ing insects: patterns, potential, and problems. PLoS One, 7, e36514. relative associated with papaya bunchy top disease. Current Microbiol- Stefanini A, Duron O (2012) Exploring the effect of the Cardinium ogy, 26,80–84. endosymbiont on spiders. Journal of Evolutionary Biology, 25, Dobson SL, Bourtzis K, Braig HR et al. (1999) Wolbachia infections are 1521–1530. distributed throughout insect somatic and germ line tissues. Insect Suesdek-Rocha L, Bertani R, da Silva-Junior PI, Selivon D (2006) The first Biochemistry and Molecular Biology, 29, 153–160. record for Wolbachia in a scorpion: the parthenogenetic yellow scorpion Duron O, Hurst GDD, Hornett EA, Josling JA, Engelstadter J (2008) High Tityus serrulatus (Scorpiones, Buthidae). Revista Iberica de Aracnologıa, incidence of the maternally inherited bacterium Cardinium in spiders. 14, 183–184. Molecular Ecology, 17, 1427–1437. Zchori-Fein E, Perlman SJ (2004) Distribution of the bacterial symbiont Engelstadter J, Hurst GDD (2009) The ecology and evolution of microbes Cardinium in arthropods. Molecular Ecology, 13, 2009–2016. that manipulate host reproduction. Annual Review of Ecology and Zug R, Hammerstein P (2012) Still a host of hosts for Wolbachia: analysis Systematics, 40, 127–149. of recent data suggests that 40% of terrestrial arthropod species are Fellous S, Duron O, Rousset F (2011) Adaptation due to symbionts and infected. PLoS One, 7, e38544. conflicts between heritable agents of biological information. Nature Reviews Genetics, 12, 663. Goodacre SL, Martin OY (2012) Modification of insect and arachnid R.W.B. designed the study, collected and analyzed the behaviours by vertically transmitted endosymbionts: infections as data, and wrote the manuscript. drivers of behavioural change and evolutionary novelty. Insects, 3, 246–261. Goodacre SL, Martin OY, Thomas CF, Hewitt GM (2006) Wolbachia and other endosymbiont infections in spiders. Molecular Ecology, 15, 517–527. Data Accessibility Hirsch J, Strohmeier S, Pfannkuchen M, Reineke A (2012) Assessment of bacterial endosymbiont diversity in Otiorhynchus spp. (Coleoptera: DNA sequences: Dryad Digital Repository http:// Curculionidae) larvae using a multitag 454 pyrosequencing approach. dx.doi.org/10.5061/dryad.40pt8 (sequences also in BMC Microbiology, 12, S6. Appendix S1). Hurst GD, Jiggins FM (2005) Problems with mitochondrial DNA as a marker in population, phylogeographic and phylogenetic studies: the effects of inherited symbionts. Proceedings of the Royal Society of London. Series B, 272, 1525–1534. Supporting Information Jiggins FM, Hurst GD (2011) Rapid insect evolution by symbiont transfer. Additional Supporting Information may be found in the online Science, 332, 185–186. version of this article: Jiggins FM, Bentley JK, Majerus MEN, Hurst GDD (2001) How many species are infected with Wolbachia? Cryptic sex ratio distorters revealed Appendix S1 Off-target sequences obtained using primers to be common by intensive sampling. Proceedings of the Royal Society of designed to specifically amplify bacterial endosymbionts. London. Series B, 268, 1123–1126.