A Survey of the Bacteriophage WO in the Endosymbiotic Bacteria Wolbachia Laurent Gavotte,*1 He´le`ne Henri,* Richard Stouthamer,  Delphine Charif,* Sylvain Charlat,à Michel Boule´treau,* and Fabrice Vavre* *Laboratoire de Biome´trie et Biologie Evolutive (UMR 5558), CNRS, IFR 41, University Lyon 1, Villeurbanne, France;  Department of Entomology, University of California, Riverside; and àBiology Department, University College London, London, United Kingdom Bacteriophages are common viruses infecting prokaryotes. In addition to their deadly effect, phages are also involved in several evolutionary processes of bacteria, such as coding functional proteins potentially beneficial to them, or favoring horizontal gene transfer through transduction. The particular lifestyle of obligatory intracellular bacteria usually protects them from phage infection. However, Wolbachia, an intracellular alpha-proteobacterium, infecting diverse arthropod and nematode species and best known for the reproductive alterations it induces, harbors a phage named WO, which has recently been proven to be lytic. Here, phage infection was checked in 31 Wolbachia strains, which induce 5 different effects in their hosts and infect 25 insect species and 3 nematodes. Only the Wolbachia infecting nematodes and Trichogramma were found devoid of phage infection. All the 25 detected phages were characterized by the DNA sequence of a minor capsid protein gene. Based on all data currently available, phylogenetic analyses show a lack of congruency between Wolbachia or insect and phage WO phylogenies, indicating numerous horizontal transfers of phage among the different Wolbachia strains. The absence of relation between phage phylogeny and the effects induced by Wolbachia suggests that WO is not directly involved in these effects. Implications on phage WO evolution are discussed.

Introduction Wolbachia are maternally inherited obligatory intracellular symbionts, which infect a wide range of arthropods and filarial nematodes (Bourtzis and Miller 2003). They infect more than 17% of insect species (Bourtzis and Miller 2003) and nearly all filarial nematodes (Bandi et al. 2001). The success of Wolbachia is best explained by the variety of phenotypes they induce, which ranges from mutualism in nematodes to various reproductive alterations in arthropods, such as cytoplasmic incompatibility (CI) (O’Neill and Karr 1990), parthenogenesis induction (PI) (Stouthamer et al. 1990), feminization of genetic males (Bouchon et al. 1998), male killing (MK) (Jiggins et al. 2001), and oogenesis completion in one hymenopteran species (Dedeine et al. 2001). In addition, some strains do not induce any apparent reproductive effect (Vavre et al. 2002). The molecular targets used by Wolbachia and the mechanisms involved in their effects are not known, and the absence of any correlation between Wolbachia phylogeny and the effects they induce in hosts have led several authors to speculate on the evolution of Wolbachia-induced phenotypes (Reviewed in Stouthamer et al. 1999). Three hypotheses have been posed: 1) the effects are determined by the host’s genome, 2) transition from one effect to another is easy and is a frequent mutational event, and 3) effects are encoded by extrachromosomal elements harbored by the bacterium. No plasmid was detected in Wolbachia, but the presence of a phage, named WO, suspected for a long time (Wright et al. 1978), was confirmed recently (Masui et al. 2000). Phages are widespread viruses infecting bacteria that use the host cell molecular systems for replicating their own nucleic acid and for synthesizing their proteins. At the end of the phage infection cycle, the accumulation of phage particles in bacterial cytoplasm induces cell lysis and bacterial 1 Present address: Department of Entomology, S225 Agricultural Science Center Building—North Lexington.

Key words: Wolbachia, phage WO, insect, phylogeny. E-mail: [email protected]. Mol. Biol. Evol. 24(2):427–435. 2007 doi:10.1093/molbev/msl171 Advance Access publication November 9, 2006 Ó The Author 2006. Published by Oxford University Press on behalf of the Society for Molecular Biology and Evolution. All rights reserved. For permissions, please e-mail: [email protected]

death. However, some bacteriophages can establish a not immediately lethal association with bacteria when they enter lysogenic cycles (Lwoff 1953). In that case, a phage coding for advantageous proteins such as antibiotic resistance or toxins can be a valuable auxiliary for bacteria (Miao and Miller 1999). Phages are also implicated in transduction, a mechanism allowing genetic transfer between bacterial cells (Miller 2001). The lytic activity of the phage WO is well documented (Masui et al. 2001; Fujii et al. 2004; Gavotte et al. 2004), making its prolonged persistence in Wolbachia strains a puzzling observation. Indeed, strong selection pressures acting on endosymbiotic prokaryotes tend to eliminate parasitic DNA such as repeated DNA or phages (Andersson JO and Andersson SG 1999). WO is one of the rare reported cases of bacteriophage infection in an intracellular bacterium (Storey et al. 1989), suggesting that WO might provide some factors of importance to Wolbachia, for example, by contributing to the reproductive alterations they induce in their hosts. Like numerous intracellular symbionts, Wolbachia cannot be cultured outside of insect cells, rendering the study of WO particularly difficult and making polymerase chain reaction (PCR) surveys and sequence analyses useful methods for studying Wolbachia–phage interactions. Based on such PCR surveys, Bordenstein and Wernegreen (2004) estimated that WO infects 90% of Wolbachia strains. This estimate included 39 Wolbachia strains from A and B clade and inducing various reproductive alterations. However, for most of these phages, no sequence is available, especially for those inducing effects other than CI (Masui et al. 2000; Bordenstein and Wernegreen 2004; Gavotte et al. 2004; Sanogo and Dobson 2004; but for phage sequences from Wolbachia-inducing feminization, see Braquart-Varnier et al. 2005 and for one involved in host oogenesis, see Gavotte et al. 2004). Analysis of the diversity and the evolutionary dynamics of WO–Wolbachia associations is the first step to better understand the possible implication of WO in Wolbachia dynamics and effects on hosts, as was already proposed by some authors (Masui et al. 2000; Fujii et al. 2004). In the present study, phage WO infection was characterized

428 Gavotte et al.

using PCR survey and sequencing of a minor capsid protein in 31 new insect Wolbachia strains and 3 strains infecting nematodes. Together with data already published (Masui et al. 2000; Bordenstein and Wernegreen 2004; Gavotte et al. 2004; Sanogo and Dobson 2004; BraquartVarnier et al. 2005), our data set contains 40 phage types harbored by 52 Wolbachia strains covering the whole range of Wolbachia effects (CI, PI, MK, feminization, commensalisms, and mutualism). This study allows drawing new conclusions on the coevolutionary history of WO and Wolbachia. Materials and Methods Biological Material Table 1 reports all insect species used in this study together with the other species for which phage sequences were available in the literature (Masui et al. 2000; Bordenstein and Wernegreen 2004; Gavotte et al. 2004; Sanogo and Dobson 2004; Braquart-Varnier et al. 2005). DNA Extraction and PCR DNA extraction was carried out using Chelex resin as described in Vavre et al. (1999). PCR was performed in a 25-ll final volume reaction containing 200 lM of each dNTP, 200 nM of each primer, 0.5 units Taq DNA polymerase, and 2 ll of DNA solution. The PCR conditions were 1 min at 95 °C and then 35 cycles of 30 s at 92 °C, 40 s at specific hybridization temperature (table 2), and 1 min 15 s at 72 °C. After the cycles, a 10-min elongation time at 72 °C was realized. Amplified products were visualized under UV after 30 min migration under 100 volts current in 1% agarose gels. Enzymatic Digestion Potential number of phage types was estimated by screening all orf7 PCR products with 2 restriction enzymes. Enzymes were selected to discriminate different phage types in Asobara tabida and Leptopilina heterotoma (Gavotte et al. 2004) and proved also efficient for other bacteriophage types. Orf7 PCR products (5 ll) were digested over night using 2 restriction enzymes: DraI, and MboII. The restriction reaction was performed at 37 °C in 4-CORE B buffer (Tris– HCl: 6 mM; MgCl2: 6 mM; NaCl: 50 mM; dithiothreitol: 50 mM; pH 7.5 at 37 °C) (PROMEGA, Charbonnie`res, France). Restriction profiles were observed under UV light after 1 h 30 min migration at 100 volts current on 3.5% agarose gel containing Ethidium Bromide. Cloning and Sequencing PCR products with digestion profile corresponding to a known unique sequence were sequenced directly; all others were cloned before sequencing. PCR products were cloned using the TOPO-TA Cloning Kit (Invitrogen, Abingdon, UK) after purification by cartridge method on Concert Rapid PCR Purification System kit (Life Technologies). Purified products were inserted into a plasmid pCR

2.1-TOPO containing ampicillin resistance gene. This construction was transferred into competent Escherichia coli of TOP10. Colonies containing plasmid were selected on Luria broth (LB) medium plates (Trypton 1%, Yeast extract 0.5%, NaCl 1%, and agarose 1%) with ampicillin (25 lg/ ml), and those containing plasmids with the PCR product were discriminated by a white/blue screening. Selected clones were incubated in LB liquid medium with ampicillin (25 lg/ml). Plasmids were purified by alkaline extraction (Birnboim and Doly 1979). A specific PCR with WOF/ WOR primers was performed to select clones containing the right insert. Clones were selected using the same restriction method, and 4–10 clones were sequenced for each profile. In all cases, within a Wolbachia strain, clones sharing the same restriction profile gave all the same sequence. DNA sequences are available on GenBank under the following accession numbers: orf7: DQ380533 to DQ380547, orf2: DQ380528 to DQ380532, wsp: DQ380525 to DQ380527. Alignments and Phylogeny DNA sequences were first translated and peptide sequences aligned with MUSCLE software (Edgar 2004) with parameter by default. Using RevTrans (Wernersson and Pedersen 2003), peptide alignments were then used as a scaffold for constructing the corresponding DNA multiple alignment. GBLOCKS program (Castresana 2000) was used to select reliable wsp regions. This left 390 sites, that is, 59% of the original alignment. The orf7 informative regions were selected manually. This left 281 sites, that is, 62% of the original alignment. For both genes, maximum likelihood (ML) trees were computed using phyml (Guindon and Gascuel 2003). Substitution models were determined with MrAIC.pl (Nylander 2004). For wsp, the optimal model was GTR 1 I 1 gamma (4), that is, general time reversible with a proportion of invariant sites and 4 categories of substitution rate to estimate the alpha parameter of the gamma law, whereas the best model for orf7 was the model GTR 1 gamma (4). Five hundred replicates of Bootstraps were made with SEQBOOT from the PHYLIP package (Felsenstein 1989). Test for Bacteria–Phage Coevolution In order to test the significance of the global hypothesis of coevolution between the phages WO and the Wolbachia strains, we used the ParaFit test (Legendre et al. 2002). This test is the function of the 2 matrices of phylogenetic distances (B and C) and of the matrix of host– parasite association links (A). These 3 matrices can be combined in a fourth one: D (D 5 CA#B), which describes the host–parasite association. To test the congruence between the 2 phylogenies, a global statistic, called ParaFitGlobal, is derived from D [ParaFitGlobal 5 trace(D#D)]. By permuting at random the row of the matrix A, a distribution of the ParaFitGlobal values can be obtained. The probability for the ParaFitGlobal value obtained from the data to be larger than or equal to most of the ParaFitGlobal values obtained under permutation can then be calculated. The ParaFit test is robust to the existence of several parasites per host (Legendre et al. 2002).

A Bacteriophage WO Survey in Wolbachia 429

Table 1 Infection of Phage WO in the Wolbachia Strains Studied Insect host

Wolbachia strain

Clade

Effecta

Phage type number

Strain origin

No. tested

No. positive

5 34 10 10 10 5 20 8 ? ? ? 25 6 ? ? ? ? ? 2 12 19 10 10 10 15 ? ? ? 13 ? 9 1 1 ? ? 20 ? 2 2 14 9 ? 1 1 14 ? 10 10 28 10 10 20

5 34 10 10 10 4 (?) 20 8 ? ? ? 25 6 ? ? ? ? ? 0 12 19 10 10 10 15 ? ? ? 7 ? 3 1 1 ? ? 20 ? 2 2 5 8 ? 1 1 14 (?) ? 0 0 0 0 0 20

10 10 10 10 40 10 10 ? 1 1 1

0 0 0 0 0 0 0 0 0 0 0

Asobara rufescens Asobara tabida

wAruf wAtab1 wAtab2 wAtab3

A A

CI CI CI oog

1 1 ? 3

Drosophila bifasciata Drosophila melanogaster

wDbif wDmel

A A

MK CI

Drosophila recens Drosophila simulans

wDres wRI

A A

CI CI

? 1 1 2 0 1

wCof wAu wNo wHa wEkue wEcauA wBol2 wLgui wLhet1 wLhet2 wLhet3

A A B A A A A A A A A

None CI CI CI CI CI Unk. CI CI CI CI

1 1 5 1 1 4 0 1 1 1 1

Muscidifurax uniraptor Nasonia giraulti Nasonia logicornis Nasonia vitripennis Pachycrepoideus dubius Protocalliphora siala Trichopria drosophilae sp. Armadillidium vulgare Adalia bipunctata Culex pipiens pipiens

wMuni wNgir wNlon wNvitA wPdub wPsia wTdro wAvul wAbipY wCpip

A A A A A A A B B B

PI CI CI CI None CI CI Fem. MK CI

2 1 3 4 2 1 1 2 1 3

Encarsia formosa Ephestia cautella Hypolimnas bolina

wEfor wEcauB wBol1

B B B

PI CI MK

1 2 2

Leptopilina clavipes Leptopilina victoriae Nasonia vitripennis Porcellio dilatatus petiti Porcelliunides pruinosus Telenomus nawai Teleogryllus taiwanemma Trichogramma brevicapillum Trichogramma cordubensis Trichogramma deion Trichogramma embryophagum Trichogramma evanescens Trichogramma kaykai

wLcla wLvic wNvitB wPpet wPprui wTnaw wTtai wTbre wTcor wTdei wTemb wTeva wTkay1 wTkay2 wTmin wTnub wTole wTpla wTpre wTsem wTsib wBmal wDimm wLsig Uninfected

B B B B B B B B B B B B B A B B B B B B B D C D nd

PI CI CI Fem. Fem. PI CI PI PI PI PI PI PI Unk. PI PI PI PI PI PI PI Mut. Mut. Mut. nd

2 1 1 1 1 ? 2 0 0 0 0 0 1

France France USA Greece Canada nd France Portugal China USA USA France Australia Australia Noumea Hawaı¨ Japan Japan Tubai-FP South Africa France The Netherlands Spain Tunisia USA USA USA USA France nd France nd Russia USA France The Netherlands Japan Tahiti–FP Moorea–FP The Netherlands Maurice Island USA nd nd Japan nd USA Portugal USA Iran France USA

0 0 0 0 0 0 0 0 0 0 0

USA Canada Yugoslavia USA Uruguay France Canada nd France France France

Ephestia kuehniella Ephestia cautella Hypolimnas bolina Leptopilina guineaensis Leptopilina heterotoma

Trichogramma nr minutum Trichogramma nubilale Trichogramma oleae Trichogramma platneri Trichogramma pretiosum Trichogramma semblidis Trichogramma sibercum Brugia malayi Dirofilaria immitis Litosomoides sigmodontis Setaria equine

Referencesb 1 1 1 1 1 2 3 4 3 3 3 4 4 1 1 1 3 3 3 3 5 6 4

3 5 5 4

7

NOTE.—Data from this study are in bold. FP, French Polynesia. a indicates bacterial effect: none, no reproductive effect detected; oog., effect on oogenesis; Unk., effect unknown; Fem., feminization; Mut., mutualism; ?, indicates unknown data; nd., not determined. b Results previously presented in (1) Gavotte et al. 2004; (2) Wu et al. 2004; (3) Bordenstein and Wernegreen 2004; (4) Masui et al. 2000; (5) Braquart-Varnier et al. 2005; (6) Sanogo and Dobson 2004; (7) by Blast on complete genome http://tools.neb.com/wolbachia/search.html.

430 Gavotte et al.

Table 2 Primers Used in the Present Work Organism Phage WO

Gene orf7 WD0633 orf2

Wolbachia

ftsZ wsp

Insect a

ITS2

Primer

Primer sequence

WOF WOR WD0633F WD0633R ORF2F ORF2R F2 R2 81F 691R ITS2F ITS2R

5#-CCC ACA TGA GCC AAT GAC GTC TG-3# 5#-CGT TCG CTC TGC AAG TAA CTC CAT TAA AAC-3# 5#-TGG GTA TCT CTT AGA TGC AAA AG-3# 5#-AAG AGC AAG GCT TTT ACA TTA GG-3# 5#-GCA GGG CTA TAT TTT GGC GAG AA-3# 5#-AAC TCC ATT AAA ACT TCC CTG GC-3# 5#-TTG CAG AGC TTG GAC TTG AA-3# 5#-CAT ATC TCC GCC ACC AGT AA-3# 5#-TGG TCC AAT AAG TGA TGA AGA AAC-3# 5#-AAA AAT TAA ACG CTA CTC CA-3# 5#-TGT GAA CTG CAG GAC ACA TG-3# 5#-AAT GCT TAA ATT TAG GGG GTA-3#

Annealing temperature

Referencesa

57 °C

1

56 °C 56 °C 55 °C

2

52 °C

3

55 °C

4

References: (1) Masui et al. 2000; (2) Holden et al. 1993; (3) Braig et al. 1998; (4) Campbell et al. 1993.

Results Phage Typing Phage presence was detected by specific PCR on the orf7 marker, a gene coding for a minor capsid protein used to detect WO (Masui et al. 2000). PCR specificity was tested on individuals cured from Wolbachia by antibiotic treatment, and WO was never detected in Wolbachia-free individuals (data not shown). Individuals showing Wolbachia infection but no signal on orf7 were tested using 2 other phage genes, the orf2 and the wd0634, and congruent results were always obtained on the 3 markers. The quality of DNA extracts tested negative for phage and bacteria was assessed by PCR using primers for the internal transcribed spaces 2 (ITS2). Samples, where we failed to amplify the ITS2, were excluded from the analysis. To determine the diversity of phage infection within each bacterial strain, the nucleotide sequence of orf7 was used. A phage type was defined by grouping all phages sharing more than 99% similarity on orf7 DNA sequences based upon pure clones sequencing repeats. Usefulness of orf7 as a phage phylogenetic marker was assessed by comparing the phylogenies obtained with orf7 and orf2 markers as in Bordenstein and Wernegreen (2004). Unfortunately, our sequences do not locate in the same gene region of the orf2, making it impossible to directly compare the 2 studies. Phylogenies were established with ML method, HKY model for orf2, and GTR for orf7 with 500 bootstrap replicates. Congruency between orf7 and orf2 phylogenies is good (fig. 1) as has been reported by Bordenstein and Wernegreen (2004). Moreover, for 3 different Wolbachia strains (wRi, wDmel, wLhet1) showing the same orf7 and orf2 sequences, a part of wd0634 gene that codes for a resolvase was also sequenced. The 3 sequences obtained were also completely identical. Thus, even though recombination can happen in phages (Bordenstein and Wernegreen 2004), the orf7 seems to be a good phylogenetic marker.

insects (wTkay1 and wTkay2) or with weak orf7 PCR signal (wTnaw and wDbip), phage infection has not been resolved (table 1). Phage WO was never detected in the nematode species studied (Litosomoides sigmodontis and Dirofilaria immitis), and no related sequences were found on the complete genome of Brugia malayi symbiont by Blast (http://tools. neb.com/wolbachia/search.html, Foster et al. 2005). Because of their particular phylogenetic position compared with Wolbachia from arthropods, these strains are not included in subsequent analyses. Wolbachia from Trichogramma were also found to be devoid of phage. The only exception is the Wolbachia infecting Trichogramma kaykai. However, a second Wolbachia strain wTkay2, different from the strains inducing PI, is also present in T. kaykai (Van Meer et al. 1999). Based on different genetic markers, this strain is identical to the Wolbachia infecting Ephestia kuehniella and could have been transferred horizontally because E. kuehniella is the host used for Trichogramma rearing. Phage Distribution Together with previously published data (Masui et al. 2000; Bordenstein and Wernegreen 2004; Gavotte et al.

Phage Diversity The phage infection status (including presence of WO and sequence on orf7) of 27 new Wolbachia strains was established. For 4 other strains, present in multiply infected

FIG. 1.—Comparison between orf2 and orf7 phylogenies for 5 Wolbachia strains.

A Bacteriophage WO Survey in Wolbachia 431

FIG. 2.—Distribution of the number of phage types found per Wolbachia strain. The high proportion of uninfected strains is due principally to Trichogramma symbionts. Data from Masui et al. (2000); Gavotte et al. (2004); Bordenstein and Wernegreen (2004); and Braquart-Varnier et al. (2005) are included.

2004; Braquart-Varnier et al. 2005), the phage infection status of 48 bacterial strains is known (fig. 2). Among these strains, 30% are devoid of phage, but this percentage is probably overestimated because most uninfected strains (12 out of 14) are from Trichogramma hosts that we extensively surveyed. Most phage-infected Wolbachia strains display low numbers of phages, 85% (28 of 34) showing only 1 or 2 different phage types. Most Wolbachia strains display identical and complete infection for all individuals tested, but 4 proved polymorphic for phage infection: wPdub infecting Pachycrepoideus dubius (7 Wolbachia-infected individuals showing phage infection out of 13 tested), wTdro infecting Trichopria drosophilae sp. (3 of 9), wLcla infecting Leptopilina clavipes (5 of 14), and wLvic infecting Leptopilina victoriae (9 of 11). Multiple Phage Infection Multiple phage infection, where a Wolbachia strain displays more than one phage type, has been observed in 12 bacterial strains: wEcauA displays 5 phage types, wNvitA 4 phage types, wNlonA, wAtab3, and wTtai 3 phage types each, and wMuni, wPdub, wAvul, wEcauB, wLcla, and wBol1 phage types each (table 1). In order to test whether phages infecting a given Wolbachia strain are more closely related to each other than to phages chosen at random, we rebuilt the observed distribution of multiple infection (1 strain with 5 phages, 1 with 4 phages, 4 with 3 phages, and 6 with 2 phages) by randomly drawing phages. Monte Carlo simulation (Raeside 1976) was done by 5000 random drawings. Between each drawing, all phages were put back because some multiply infected bacteria display common phage types. For each drawing, we calculated the average of tree branch length realized by ML method between coinfecting phages. The sampling distribution of genetic distances is normal (Kolmogorov– Smirnov test, D 5 0.0147; P 5 0.2138) with an average of 0.261 and a standard deviation of 0.029 (fig. 3). The

FIG. 3.—Distribution of the average genetic distance between phages present in a single Wolbachia strain, obtained by Monte Carlo simulation.

observed average is 0.252 with a standard deviation of 0.032. Thus, phage types coinfecting a Wolbachia strain are not more related than expected by chance (P 5 0.628). Phage Phylogeny Phage phylogeny, based on orf7, and Wolbachia phylogeny, based on wsp (Braig et al. 1998), were compared for 33 bacterial strains infecting insects (only a few strains infecting Trichogramma, devoid of phages are included, others were not included for readability) and 55 phage types (fig. 4). No congruence was found between phage and bacterial phylogenies (ParaFit test, P 5 0.1319). Thus, it appears that phages do not cospeciate often enough with their Wolbachia host to create strong phylogenetic signal, similar to the lack of congruence between the phylogenies of Wolbachia and their insect hosts. Comparisons between phage and insect host phylogenies are also not congruent (data not shown). However, a few correlations can be observed: the 2 pairs of Wolbachia wPdub/wMuni and wEkue/wTkay2 are potentially originating from bacterial horizontal transfers (Van Meer et al. 1999; Vavre et al. 1999) and display the same wsp, 16S RNA, and ftsZ sequences and also carry the same phage infection. This bacteriophage WO–Wolbachia combination was probably transferred between different insect hosts. However, absence of correlation between phage and bacterial phylogenies indicates that the bacteriophage WO can be transferred horizontally without its bacterial host between different Wolbachia strains or insects and can infect new Wolbachia hosts, as previously suggested based on smaller data sets (Masui et al. 2000; Bordenstein and Wernegreen 2004).

432 Gavotte et al.

FIG. 4.—Comparison between bacterial phylogeny on the left side (based on wsp sequence) and phage phylogeny on the right side (based on orf7 sequence) infecting these bacteria. Wolbachia strains uninfected by phage were not included for a better readability.

Phage and Wolbachia Effects Comparing the phage phylogeny with the effects induced by their Wolbachia hosts (fig. 5) does not suggest any evident correlation between the phage phylogeny and the effect of the Wolbachia strain it infects. Indeed, the CI, MK, and PI phenotypes are scattered throughout the phylogeny, and the phages associated with these effects form paraphyletic groups. Moreover, a number of phageinfected bacteria (e.g., wPdub) do not induce any reproductive effects, whereas some uninfected Wolbachia strains are able to induce reproductive effects: all Trichogramma symbionts induce PI and wDres induces CI (Stouthamer and Kazmer 1994; Werren and Jaenike 1995). Discussion Intracellular bacteria are usually not prone to bacteriophage infection, owing both to selection for a reduction of their genome size and limited exposure to infection associated with their isolated lifestyle (Franck et al. 2002). However, we observed that infection with the bacteriophage WO is a common feature in Wolbachia because it was detected in 70% of the strains and this proportion is probably an underestimation due to the high representation of Trichogramma species in the data set. The general absence of congruence between phages and Wolbachia phylogenies demonstrates that WO is able to successfully transfer itself horizontally between different insects, with or without its bacterial host, as already

suggested by some studies on fewer Wolbachia–phage associations (Masui et al. 2000; Bordenstein and Wernegreen 2004). Random mixing of phages is also suggested by the fact that phages sharing the same Wolbachia are not more related to each other than expected by chance. Three hypotheses can be drawn for these transfers: 1) Bacteriophage WO particles can be released, under some conditions, in the proximity of insect cells infected by Wolbachia, and they can pass through the eukaryote cell wall and then initiate new infections. Parasitoid infection represents a potential route for such horizontal transfer. Whatever the result of parasitism (success or failure), the Wolbachia strain present in the ‘‘winner’’ insect comes into contact with new potential phage infection. 2) Bacteriophage WO is able to infect other bacteria than Wolbachia. If these bacteria are free living and potentially present in various environments, the possibility for Wolbachia to acquire new phages is increased compared with transmission being only possible between Wolbachia strains. It is interesting to note that the closest relative of the WO phage of Wolbachia is a phage found in the plant pathogen Xylella fastidiosa, which is transmitted by the Wolbachia-infected Glassy-winged sharpshooter (Simpson et al. 2000). 3) Previous studies observed that Wolbachia could be transmitted horizontally between insects naturally (Huigens et al. 2004) or by artificial methods (Boyle et al. 1993). Because Wolbachia is present in numerous ecologically related species, transfer of a Wolbachia strain to an insect already infected by another Wolbachia strain will establish contact among various bacterial strains with various phage types. Even

A Bacteriophage WO Survey in Wolbachia 433

FIG. 5.—Bacteriophage WO phylogeny based on orf7 sequences. Reproductive effects on Wolbachia host are reported. None, No reproductive effects known; Fem., Feminization of genetic males; and Oog., effect on oogenesis.

Could the widespread association of Wolbachia and the bacteriophage WO mean that the bacteriophage WO may be a beneficial auxiliary to Wolbachia, as is found in various other phage/bacteria couples (Miao and Miller 1999)? Among the functions potentially encoded for by the phage genome are the reproductive effects of Wolbachia on their insect host (Stouthamer et al. 1999). However, 1) no correlation was observed between phage presence or phylogeny and reproductive effects and 2) among PI-inducing Wolbachia, some are infected by phages (wLcla, wMuni, and wTnaw) but others are not (Trichogramma strains), showing that the diverse effects of Wolbachia are not due to a small number of specialized WO phages that move from one Wolbachia strain to another. This does not rule out a possible implication of WO in the effect of Wolbachia, as suggested by recent finding in mosquito where ankyrin genes within the phage genome might be involved in the induction of CI (Sinkins et al. 2005). Different phages might be involved in the induction of similar (on the basis of their phenotype) effects. Moreover, molecular mechanisms involved in reproductive effect of Wolbachia probably involve Wolbachia 3 host interaction that the phage might mediate. Finally, the phage WO could be the vector of genes involved in Wolbachia’s effect, but the frequent rearrangements occurring in the Wolbachia genomes could rapidly lead to the transfer of these genes on the bacterial chromosome itself, thereby breaking down the association between the phage and the Wolbachia effect. In conclusion, the present study suggests that the success of the bacteriophage WO mainly derives from its ability to be transferred among bacterial hosts (using a currently unknown mechanism), rather than from any beneficial effect contributed to its Wolbachia host. This situation parallels the relationship between Wolbachia and its arthropod hosts (Hurst and Mc Vean 1996). Acknowledgments

if the bacteria transferred are eliminated later, which is probably the most common case, new phage infection could have been initiated. Whatever the mechanism, all observations suggest that the bacteriophage WO can spread among Wolbachia strains through horizontal transfers of phage, indicating that WO has important infectious capacities. However, phage transfer might not be as frequent as is suggested by the absence of congruence between Wolbachia and phage phylogenies. For instance, the 3 Wolbachia strains coinfecting each individual of the parasitoid L. heterotoma do not share phage infections (Gavotte et al. 2004). Moreover, phages were not detected in 2 specific groups of Wolbachia: those infecting nematodes and Trichogramma species (Schilthuizen and Stouthamer 1997; Bandi et al. 2001). Whether these groups have lost previous infection or have never been infected is unclear, but it does show that Wolbachia strains can remain uninfected for long periods of time. The unusually high specialization of these 2 Wolbachia clades, both in terms of host range and phenotypic effects, might limit the opportunity of phage acquisition. Alternatively, lack of phage infection might also explain the high specialization of these Wolbachia, by limiting genome plasticity and gene transfers.

We thank Stephen Dobson and Tim Vogel for suggestions for improving this paper. We thank Serge Morand, Gilsang Jeong, Greg Hurst, Chris Jiggins and Bernard Pintureau to provide a part of insect and nematodes samples used in this study. This study was partly supported by EuWol project (EU) QLK3-2000-01079 and CNRS (UMR 5558 and IFR 41). Literature Cited Andersson JO, Andersson SG. 1999. Insights into the evolutionary process of genome degradation. Curr Opin Genet Dev. 6:664–671. Bandi C, Trees AJ, Brattig NW. 2001. Wolbachia in filarial nematodes: evolutionary aspects and implications for the pathogenesis and treatment of filarial diseases. Vet Parasitol. 98: 215–238. Birnboim HC, Doly J. 1979. A rapid alkaline lysis extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513–1523. Bordenstein SR, Wernegreen JJ. 2004. Bacteriophage flux in endosymbionts (Wolbachia): infection frequency, lateral transfer, and recombination rates. Mol Biol Evol. 21:1981– 1991.

434 Gavotte et al.

Bouchon D, Rigaud T, Juchault P. 1998. Evidence for widespread Wolbachia infection in isopod crustaceans: molecular identification and host feminisation. Proc R Soc Lond B Biol Sci. 1401:1081–1090. Bourtzis K, Miller T. 2003. Insect symbiosis. Boca Raton (FL): CRC Press LLC. Boyle L, O’Neill SL, Robertson HM, Karr TL. 1993. Interspecific and intraspecific horizontal transfer of Wolbachia in Drosophila. Science. 260:1796–1799. Braig HR, Zhou W, Dobson SL, O’Neill SL. 1998. Cloning and characterization of a gene encoding the major surface protein of the bacterial endosymbiont Wolbachia pipientis. J Bacteriol. 180:2373–2378. Braquart-Varnier C, Greve P, Felix C, Martin M. 2005. Bacteriophage WO in Wolbachia infecting terrestrial isopods. Biochem Biophys Res Commun. 337:580–585. Campbell BC, Steffen-Campbell JD, Werren JH. 1993. Phylogeny of the Nasonia species complex (Hymenoptera: Pteromalidae) inferred from an internal transcribed spacer (ITS2) and 28S rDNA sequences. Insect Mol Biol. 2:225–237. Castresana J. 2000. Selection of conserved blocks from multiple alignments for their use in phylogenetic analysis. Mol Biol Evol. 17:540–552. Charif D, Lobry JR. 2006. SeqinR: a contributed package to the R project for statistical computing devoted to biological sequences retrieval and analysis. In: Bastolla U, Porto M, Roman HE, Vendruscolo M, editors. Structural approaches to sequence evolution. Springer, Berlin, Germany. Chapter 10. Dedeine F, Vavre F, Fleury F, Loppin B, Hochberg ME, Boule´treau M. 2001. Removing symbiotic Wolbachia bacteria specifically inhibits oogenesis in a parasitic wasp. Proc Natl Acad Sci USA. 98:6247–6252. Edgar RC. 2004. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 32:1792–1797. Felsenstein J. 1989. PHYLIP-Phylogeny inference package (Version 3.2). Cladistics. 5:164–166. Foster J, Ganatra M, Kamal I, et al. (23 co-authors). 2005. The Wolbachia genome of Brugia malayi: endosymbiont evolution within a human pathogenic nematode. PLoS Biol. 3:E121. Frank AC, Amiri H, Andersson SG. 2002. Genome deterioration: loss of repeated sequences and accumulation of junk DNA. Genetica. 115:1–12. Fujii Y, Kubo T, Ishikawa H, Sasaki T. 2004. Isolation and characterization of the bacteriophage WO from Wolbachia, an arthropod endosymbiont. Biochem Biophys Res Commun. 317: 1183–1188. Gavotte L, Vavre F, Henri H, Ravallec M, Stouthamer R, Boule´treau M. 2004. Diversity, distribution and specificity of WO phage infection in Wolbachia of four insect species. Insect Mol Biol. 12:147–153. Guindon S, Gascuel O. 2003. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol. 52:696–704. Holden PR, Brookfield JF, Jones P. 1993. Cloning and characterization of an ftsZ homologue from a bacterial symbiont of Drosophila melanogaster. Mol Gen Genet. 240:213–220. Huigens ME, de Almeida RP, Boons PA, Luck RF, Stouthamer R. 2004. Natural interspecific and intraspecific horizontal transfer of parthenogenesis-inducing Wolbachia in Trichogramma wasps. Proc R Soc Lond B Biol Sci. 271:509–515. Hurst LD, Mc Vean GT. 1996. Clade selection, reversible evolution and the persistence of selfish elements: the evolutionary dynamics of cytoplasmic incompatibility. Proc R Soc Lond B Biol Sci. 263:97–104. Ihaka R, Gentleman R. 1996. R: a language for data analysis and graphics. J Comput Graph Stat. 3:299–314.

Jiggins FM, Hurst GD, Schulenburg JH, Majerus ME. 2001. Two male-killing Wolbachia strains coexist within a population of the butterfly Acraea encedon. Heredity. 86:161–166. Legendre P, Desdevises Y, Bazin E. 2002. A statistical test for host-parasite coevolution. Syst Biol. 51:217–234. Lwoff A. 1953. Lysogeny. Bacteriol Rev. 17:269–337. Masui S, Kamoda S, Sasaki T, Ishikawa H. 2000. Distribution and evolution of bacteriophage WO in Wolbachia, the endosymbiont causing sexual alterations in Arthropods. J Mol Evol. 51:491–497. Masui S, Kuroiwa H, Sasaki T, Inui M, Kuroiwa T, Ishikawa H. 2001. Bacteriophage WO and virus-like particles in Wolbachia, an endosymbiont of arthropods. Biochem Biophys Res Commun. 283:1099–1104. Masui S, Sasaki T, Ishikawa H. 2000. Genes for the type IV secretion system in an intracellular symbiont, Wolbachia, a causative agent of various sexual alterations in arthropods. J Bacteriol. 182:6529–6531. Miao EA, Miller SI. 1999. Bacteriophages in the evolution of pathogen-host interactions. Proc Natl Acad Sci USA. 96: 9452–9454. Miller RV. 2001. Environmental bacteriophage-host interactions: factors contribution to natural transduction. Antonie Van Leeuwenhoek. 79:141–147. O’Neill SL, Karr TL. 1990. Bidirectional incompatibility between conspecific populations of Drosophila simulans. Nature. 6297:178–180. Paradis E, Claude J, Strimmer K. 2004. APE: analyses of phylogenetics and evolution in R language. Bioinformatics. 20: 289–290. Raeside DE. 1976. Monte Carlo principles and applications. Phys Med Biol. 21:181–197. Sanogo YO, Dobson SL. 2004. Molecular discrimination of Wolbachia in the Culex pipiens complex: evidence for variable bacteriophage hyperparasitism. Insect Mol Biol. 13:365–369. Schilthuizen M, Stouthamer R. 1997. Horizontal transmission of parthenogenesis-inducing microbes in Trichogramma wasps. Proc R Soc Lond B Biol Sci. 264:361–366. Simpson AJ, Reinac FC, Arruda P, et al. (113 co-authors). 2000. The sequence of the plant pathogen Xylella fastidiosa. Fastidiosa consortium of the organization for sequencing and analysis. Nature. 406:151–157. Sinkins SP, Walker T, Lynd AR, Steven AR, Makepeace BL, Godfray HC, Parkhill J. 2005. Wolbachia variability and host effects on crossing type in Culex mosquitoes. Nature. 7048: 257–260. Storey CC, Lusher M, Richmond SJ, Bacon J. 1989. Further characterization of a bacteriophage recovered from an avian strain of Chlamydia psittaci. J Gen Virol. 70:1321–1327. Stouthamer R, Breeuwer JA, Hurst GD. 1999. Wolbachia pipientis: microbial manipulator of arthropod reproduction. Annu Rev Microbiol. 53:71–102. Stouthamer R, Kazmer DJ. 1994. Cytogenetic of microbe associated parthenogenesis, consequences for gene flow in Trichogramma wasps. Heredity. 73:317–327. Stouthamer R, Luck RF, Hamilton WD. 1990. Antibiotics cause parthenogenetic Trichogramma (Hymenoptera/Trichogrammatidae) to revert to sex. Proc Natl Acad Sci USA. 87:2424–2427. Van Meer MM, Witteveldt J, Stouthamer R. 1999. Phylogeny of the arthropod endosymbiont Wolbachia based on the wsp gene. Insect Mol Biol. 8:399–408. Vavre F, Fleury F, Lepetit D, Fouillet P, Boule´treau M. 1999. Phylogenetic evidence for horizontal transmission of Wolbachia in host-parasitoid associations. Mol Biol Evol. 16:1711–1723. Vavre F, Fleury F, Varaldi J, Fouillet P, Bouletreau M. 2002. Infection polymorphism and cytoplasmic incompatibility in Hymenoptera-Wolbachia associations. Heredity. 88:361–365.

A Bacteriophage WO Survey in Wolbachia 435

Wernersson R, Pedersen AG. 2003. RevTrans: multiple alignment of coding DNA from aligned amino acid sequences. Nucleic Acids Res. 31:3537–3539. Werren JH, Jaenike J. 1995. Wolbachia and incompatibility in mycophagous Drosophila and their relatives. Heredity. 75:320–326. Wright JD, Sjo¨strand FS, Portreo JK, Barr AR. 1978. The ultrastructure of the Rickettsia-like microorganism Wolbachia pipientis and associated virus-like bodies in the mosquito Culex pipiens. J Ultrastruct Res. 63:79–85.

Wu M, Sun LV, Vamathevan J, et al. (30 co-authors). 2004. Phylogenomics of the reproductive parasite Wolbachia pipientis wMel: a streamlined genome overrun by mobile genetic elements. PLoS Biol. 2:327–341.

Aoife McLysaght, Associate Editor Accepted November 3, 2006