Introduction

Wolbachia pipientis is a common symbiont of Drosophila simulans. Wolbachia may cause a number of reproductive abnormalities, including cytoplasmic incompatibility (Yen and Barr, 1971; Hoffmann et al, 1986). In D. simulans, incompatibility may occur when an infected male mates with an uninfected female, or a female with a different strain of Wolbachia. In such a cross, there may be a dramatic reduction in the number of viable eggs produced. The reciprocal cross usually yields normal numbers of progeny, unless the parents harbor different strains of Wolbachia. The aim of this study is to explore the dynamics of a host population from New Caledonia with uninfected, singly infected, and doubly infected individuals.

D. simulans is an excellent model system for investigating the host and symbiont population genetics. D. simulans is a cosmopolitan species that may harbor five strains of Wolbachia (wHa, wNo, wRi, wAu, and wMa). A sixth strain, wKi, has been described from Tanzania, but is homosequential with wMa at both the 16S rDNA and wsp loci and is likely the same. In this study, we refer to wMa instead of wKi because wMa has nomenclatural precedence. In D. simulans, mitochondrial variation is structured into three monophyletic haplotypic groups designated siI, siII, and siIII (Baba-Aïssa et al, 1988; Solignac and Monnerot, 1986). Furthermore, mitochondrial DNA (mtDNA) appears to be non-randomly associated with the bacterial strains (Montchamp-Moreau et al, 1991; Rousset and Solignac, 1995; James and Ballard, 2000). The wHa and wNo strains have only been detected in individuals with the siI haplotype, wRi and wAu in flies with siII haplotypes, and wMa in siIII individuals. The goal of this study is to focus on the dynamics of flies that carry the siI haplotype and their wHa and/or wNo strains of Wolbachia.

It has been demonstrated both theoretically (Caspari and Watson, 1959) and empirically (Turelli and Hoffmann, 1991) that Wolbachia infections that cause incompatibility spread once they reach a threshold frequency. As a Wolbachia infection sweeps through a population, other cytoplasmic factors hitchhike with bacterial transmission. For example, Nigro and Prout (1990) started two sets of D. simulans population cages carrying two mitochondrial types (C and P), with one of the types at a frequency of 20% in one set and 80% in the other. The C type occurred in a host infected with Wolbachia whereas the P type was associated with an uninfected host. In all cages, there was a rapid increase in the frequency of the C type as the infection became predominant under unidirectional incompatibility. In the mosquito Aedes albopictus, changes in mtDNA frequencies have also been associated with a spreading Wolbachia infection in the laboratory (Kambhampati et al, 1992).

Wolbachia also causes shifts in mtDNA variation in natural populations. Incompatibility in Drosophila was first discovered in crosses between a predominantly infected population of D. simulans near Riverside, California, with various uninfected populations in northern and central California (Hoffmann et al, 1986). Initially, the infection was limited to sites south of the Tehachapi transverse range that separates the Los Angeles basin from the Central Valley. However, infected flies became increasingly common in the Central Valley after 1988 and a rapid spread north was observed from 1989 until 1994 (Turelli and Hoffmann, 1991; Turelli et al, 1992; Hoffmann and Turelli, 1997). As the infection swept through populations, the mtDNA variant initially associated with the infected D. simulans increased in frequency. Extending earlier work on incompatibility (Caspari and Watson, 1959), Turelli and colleagues (Turelli et al, 1992; Hoffmann and Turelli, 1997) developed a model with intrapopulation dynamics for the joint frequency of incompatibility types and mtDNA genotypes.

The existence of double infections in New Caledonia are somewhat of a mystery as the wNo infection type is not reported to elicit high incompatibility. To the north and east of New Caledonia, wNo is not found. Populations in Hawaii and Tahiti have very high frequencies of single infections with wHa individuals (O'Neill et al, 1992; Rousset et al, 1992; Turelli and Hoffmann, 1995). To the southwest, Australian populations are infected with wAu (Hoffmann et al, 1996). The only other known locality where double infections have been reported is in the Seychelles (Merçot et al, 1995b; Merçot and Poinsot, 1998).

Here we investigate empirically why these two strains coexist in a seemingly stable frequency in New Caledonia. Are they acting synergistically, independently, or is wNo parasitizing wHa? To address these questions, we assayed infection types in a population known to harbor double infections, identified their incompatibility phenotypes, and sequenced two polymorphic regions in the mitochondrial genome of their host. One notable finding is that wHa does not always induce strong incompatibility, as previously observed. Furthermore, infection status did not correlate with any detectable mitochondrial substructure within the siI haplotype. These data suggest that the population has reached an equilibrium where singly infected or uninfected flies arise through stochastic segregation from doubly infected mothers. We discuss these results and link them back to an inclusive strain concept for Wolbachia.

Materials and methods

Drosophila lines and molecular classification of Wolbachia

Fifty-five isofemale lines of D. simulans were established from flies collected in Nouméa, New Caledonia, on 29 and 31 December 1999. DNA was isolated from these lines within 1.5 months of them being established in the laboratory. To determine the Wolbachia infection status of these lines we employed a strain specific PCR assay that generated an amplicon of specific size for each Wolbachia strain. We sequenced select lines to confirm the lines were infected with the expected Wolbachia strains.

DNA from all fly lines was extracted using the Puregene™ kit (Gentra Systems, Minneapolis, MN, USA) according to the manufacturer's protocol designated 'DNA isolation from fixed tissue'. Final DNA concentrations of fly extracts were determined using a GeneQuant spectrophotometer (Pharmacia, Alamedia, CA, USA). The control lines used in the molecular classification of Wolbachia strains and mitochondrial haplotypes have been described in detail elsewhere (Ballard, 2000a, b; James and Ballard, 2000). NC48 is infected with both wHa and wNo. TT01 carries the wHa infection, and RU07 carries the wMa infection (which differs from wNo by 1 bp in the 16S rDNA but is homosequential to wNo at the Wolbachia major surface protein [wsp] locus).

To assay for presence or absence of Wolbachia infection, Wolbachia 16S rDNA was PCR amplified following a modified protocol of O'Neill et al (1992). The thermal profile was shortened to 30 cycles, and the denaturation and annealing steps were run for 15 s each instead of 1 min. Any uninfected result was checked by running the same extraction and an independent extract of three flies from the isofemale line with primers that amplify a region of the wsp locus (Zhou et al, 1998).

To survey the population for specific Wolbachia strains, a multiplex PCR reaction that amplifies a region of the wsp locus was designed. The forward primer 81F of Zhou et al (1998) and newly designed reverse primers 463R (5'-TACCATTTTGACTACTCACAGCG-3') and 635R (5'-GATCTCTTTAGTAGCTGATAC-3') were used. The 81F primer anneals to both wHa and wNo sequence. With our protocol, 487R amplifies a 427 bp product from wHa and 658R amplifies a 570 bp product from wNo (Figure 1). The 10 l PCR reactions consisted of 10 ng template DNA, 1 l 81F, 1 l 658R, 0.35 l 487R (all primers at 10 ng/l), 1 l of 8 mM dNTP, 4.625 l ddH₂O, 1 l 10 PCR buffer with 25 mM MgCl²⁺, and 0.25 units Taq polymerase (Roche, Nutley, NJ, USA). The PCR profile was 35 cycles at 94°C for 30 s, 52°C for 15 s, and 72°C for 1 min. PCR amplicons were electrophoresed on a 1.5% agarose gel stained with ethidium bromide to visualize the size of the products. If the fly line was identified as infected with only one strain, the PCR was repeated with single pairs of primers for positive and negative verification of infections.

Figure 1.

Detection of Wolbachia infections in D. simulans. (a) Schematic diagram of PCR amplification of surface binding protein gene from Wolbachia (wsp). Primers 81F and 635R amplify a 570 bp fragment specific for wNo; primers 81F and 463R produce a 427 bp fragment specific for strain wHa. Two fragments are produced from doubly infected individuals. (b) PCR amplicons from three controls and five experimental lines on a 1.5% agarose gel. Lane 1: 100-bp DNA ladder; lane 2: TT01 (wHa control); lane 3: RU07 (wNo or wMa control); lane 4: NC48 (wHa/wNo doubly infected control); lane 5: NC103 (uninfected); lane 6: NC112 (wHa-infected); lane 7: NC117 (wMa-infected); lane 8: NC102 (doubly infected); lane 9: NC125 (doubly infected); lane 10: 100-bp DNA ladder.

Full figure and legend (34K)

We also sequenced portions of the 16S rDNA and wsp loci for lines that we included in the incompatibility assays as described in James and Ballard (2000). The NC112 line was infected with wHa, with 16S rDNA sequence identical to GenBank accession number X61769 (O'Neill et al, 1992) and wsp sequence identical to accession number AF020068 (Braig et al, 1998). The NC117 is infected with wNo, with 16S rDNA identical to number AF312372 (James and Ballard, 2000) and wsp number AF020074 (Zhou et al, 1998). The wsp sequences of each strain present in doubly infected flies from New Caledonia have previously been shown to be identical to those in the singly infected lines (James and Ballard, 2000).

Cytoplasmic incompatibility

Collection of doubly infected, singly infected, and uninfected lines from New Caledonia provided a unique opportunity to study the dynamics between symbiont strains collected from a single population. Most previous intrapopulational studies have focused on the dynamics between flies infected with a single strain of Wolbachia and uninfected flies (Hoffmann and Turelli, 1988; Turelli and Hoffmann, 1991, 1995; Turelli et al, 1992). We also determined if the incompatibility phenotype associated with singly infected wNo flies was the same as experimentally constructed lines (Merçot et al, 1995b; Merçot and Poinsot, 1998).

Four lines from New Caledonia were chosen for incompatibility phenotype analyses. NC102 is doubly infected with wHa and wNo, NC112 carries only wHa, NC117 carries only wNo, and NC103 is uninfected. However, we do not have replicate lines within each Wolbachia infection status, and we cannot distinguish if the results presented here are specific to the lines used here or a general result for the whole population.

Our cytoplasmic incompatibility assay technique is described in James and Ballard (2000). Briefly, larvae were raised at constant temperature and density, and virgin adults were collected within a 12-h period. They were aged 3 days and then pairs were introduced in vials and left 24 h to mate. D. simulans isofemale lines were shown to mate at random (Ballard et al, in press) and we do not expect any bias in mating success among lines. Females were isolated and then placed in fresh vials for three 24-h periods. The vials from the first day were discarded, while all eggs laid on the latter two were counted within 8 h of the transfer. Between 26 and 36 h after the transfer, the number of eggs left unhatched was counted. The expression of cytoplasmic incompatibility was quantified as the number of eggs left unhatched in the second counting period divided by the total number of eggs laid. We counted between 13 and 20 pairs per cross (Table 1). Each of these 281 crosses laid, on average, 92 (2 s.e.) eggs. The uninfected flies employed as controls in this study were collected in the field free of Wolbachia infection. An alternative design would be to tetracycline treat infected flies to cure them of infection (O'Neill and Karr, 1990). A disadvantage of using tetracycline treated lines is that the antibiotic may influence the fitness of the flies.

Table 1 - Median eggs unhatched between fly lines with different bacterial strains (sample size; 25th and 75th quantiles). Each value represents the female labels from the top mated with the males to the left.

Full table (10K)

We arbitrarily assign compatibility as less than 30% eggs unhatched, incompatibility as greater than 70% eggs unhatched and partial incompatibility as any intermediate percentage. No median compatibility was greater than 8% , no incompatibility less than 97% , and partial incompatibility ranged from 30 to 42% . We present medians rather than means, because these incompatibility data are not normally distributed.

The Scheirer-Ray-Hare extension of the Kruskal-Wallis test (Sokal and Rohlf, 1995, pp 446–447) was used to test the model of incompatibility levels between each bacterial infection status. We sequentially removed the bacterial infection types that have the most complex incompatibility patterns and re-test the data to determine what effects remain significant (after James and Ballard, 2000).

mtDNA sequencing

If any infection type (single or double) in New Caledonia was in the process of increasing in frequency, the linked mtDNA genotype may also be expected increase in parallel, yielding a pattern of strong correlation between haplotype and infection type (Turelli et al, 1992). In this study, we were interested in testing whether the distinct Wolbachia strains collected in New Caledonia were associated with a specific mtDNA genotype. To investigate this question, we sequenced a region containing a variable length AT repeat that occurs in an intervening sequence between ND3 and alanine tRNA gene sequences, and a single nucleotide polymorphism (SNP) at site 5545 in an intervening sequence between COIII and the glycine tRNA (Ballard, 2000a). Although only siI flies have been collected from New Caledonia (Merçot et al, 1995b), sequencing this region would also determine if siII or siIII flies were also collected. No siII or siIII flies were found.

Based on GenBank accession number AF200834 (a complete mitochondrial sequence of the siI haplotype used as a control in this manuscript), the 3' end of the forward primer (5' ATTGACATTTTGTTGATGTAGTTT 3') aligns to position number 5471, and the reverse primer (5' TGAATATTCAATACTTTTTGAATG 3') to base 6035. The 50 l PCR reactions consisted of 4 l template DNA (10 ng/l), 2 l of each primer (25 ng/l), 5 l of 8 mM dNTP, 31.8 l ddH2O, 5 l of 10 PCR buffer with 25 mM MgCl²⁺, and 0.2 l Taq polymerase. The PCR profile included 35 cycles at 95°C for 1 min, 52°C for 1 min, and 72°C for 1 min.

To visualize amplification products, 4 l of the PCR product was run on a 1.5% agarose gel stained with ethidium bromide. The remaining PCR product was precipitated with 23 l 7.5 M ammonium acetate and 69 l cold 100% ethanol. Precipitates were washed with 200 l cold 70% ethanol, dried, and re-suspended with 25 l water. Purified PCR products were then quantified using a GeneQuant™ spectrophotometer (Pharmacia) prior to sequencing.

Sequencing reactions were carried out using 30–35 ng purified PCR product, 25 ng of primer, 4 l 1:2 TRR mix from an ABI Prism Big Dye™ Terminator Cycle Sequencing Kit, and brought to 10 l with ddH₂O. We collected double stranded sequence for all samples.

Results

Drosophila lines and Wolbachia infections

Of 55 isofemale lines from New Caledonia, 47 were doubly infected, four were singly infected with wHa and three were singly infected with wNo. One line was identified as uninfected by the initial screen and confirmed by further analyses.

Cytoplasmic incompatibility

The expression of incompatibility (Table 1) is shown schematically in Figure 2. All lines are self compatible (range of medians of proportion eggs unhatched is 0.00–0.04), and uninfected males can successfully reproduce with females that carry any bacteria (0.00–0.01). Doubly infected males are incompatible with uninfected females (median proportion eggs unhatched is 0.97) and partially incompatible with females that carry either wHa (0.34) or wNo (0.42) singly. Males with wHa are compatible with females doubly infected with Wolbachia (0.05). They are incompatible with females that carry wNo (1.00), and partially incompatible with uninfected females (0.33). Males that carry wNo are partially incompatible with wHa-infected females (0.30) and exhibit low, but significant, incompatibility when crossed with uninfected females (0.02).

Figure 2.

Schematic of cytoplasmic incompatibility between D. simulans lines that either carry the Wolbachia strain wHa, wNo, both, or are uninfected. Arrows go from males to females. The thickness of the line represents the level of gene flow. Thin lines represent incompatibility, thick lines represent compatibility, and dashed lines represent variable expression of incompatibility.

Full figure and legend (19K)

When all lines are considered, there are significant incompatibility differences between males and females, and there is a significant interaction between the two (Table 2a). As a consequence of this result we removed doubly infected lines. Doubly infected males are known to elicit strong incompatibility, while females are expected to be compatible with all the males in this study (Merçot et al, 1995b; Merçot and Poinsot, 1998). Table 2b shows that there are significant incompatibility differences between the remaining singly infected and uninfected males and there is a significant male-by-female interaction effect. Singly infected and uninfected females do not differ significantly in their incompatibility with singly infected or uninfected males. We then removed the singly infected wHa line because this Wolbachia strain is known to elicit incompatibility (O'Neill and Karr, 1990; James and Ballard, 2000). Table 2c shows a significant male effect between wNo infected males and uninfected females.

Table 2 - Kruskal-Wallis tests of cytoplasmic incompatibility in males and females from different lines.

Full table (11K)

mtDNA

The number of AT repeats in the intervening sequence between ND3 and the alanine tRNA ranges from 5 to 11 (Figure 3). Doubly infected flies carry 5–11 repeats, singly infected wNo carrying lines carry 8, 9, or 11 repeats and the uninfected line carries 7 repeats. The four singly infected wHa lines all carry 8 repeats as previously reported by Ballard (2000a). Nine of these lines were shown to be heteroplasmic for repeat number. To test whether wHa infection is directly linked with 8 AT repeats we assayed 18 wHa-infected Hawaiian lines: one carried 6 AT repeats, one carried 7 repeats, 12 carried 8 repeats, two carried 9 repeats, and two carried 10 repeats.

Figure 3.

The number of AT repeats sequenced from 55 lines collected in New Caledonia. The repeat region occurs in an intervening spacer region between COIII, and the glycine tRNA coding sequence.

Full figure and legend (11K)

To further explore genetic substructure of flies carrying the wHa from New Caledonia, we assayed the SNP at position 5545. Three lines had an A at this site and one a G. The polymorphism at this site was not associated with any infection type in any detectable manner. We suggest these data indicate that Wolbachia infection status is not linked with a specific mtDNA genotype in this population.

Conclusions

Wolbachia

This is the first study to report on the collection of doubly infected, singly infected and uninfected flies from a single population. Forty-seven isofemale lines were doubly infected (85.5% ; 95% confidence intervals (CI) calculated directly from the binomial distribution are (73.4 to 93.5% ), four were singly infected with wHa (7.2% ; CI 2.0 to 17.6% ), three were singly infected with wNo (5.5% ; CI 1.1 to 15.2% ), and one was uninfected (1.8% ; CI 0.0 to 2.4% ). Double infections and single wHa infections have previously been reported from New Caledonia but their relative frequencies have not been determined (Merçot et al, 1995b; Merçot and Poinsot, 1998). Doubly infected females are compatible with all males regardless of their infection status. Singly infected wNo and uninfected siI lines have both been produced by stochastic loss of the bacterial strains in the laboratory (Merçot et al, 1995a; Merçot and Poinsot, 1998), but this is the first report of these types collected from nature.

Incompatibility assays indicate that the two single infections (wHa and wNo) from New Caledonia are not independent of each other. The incompatibility is higher when a singly infected male is crossed with a female harboring the other strain rather than with an uninfected female. These data suggest that both male and female components contribute to the phenotypic expression of incompatibility in this system. Males harboring wHa show strong incompatibility when crossed to wNo females (1.00) but intermediate incompatibility with uninfected females (0.33). Likewise, males infected with the wNo strain show intermediate incompatibility with wHa females (0.30) but low incompatibility with uninfected females (0.08).

Our results have important implications for studying the dynamics of Wolbachia infections in D. simulans. In this study, we employ a median and the 25th and 75th quantiles because the incompatibility data are not normally distributed. Most previous studies have presented incompatibility as a mean and standard error. To facilitate direct comparisons with previous studies we compare means and standard errors in this paragraph. We observed intermediate incompatibility when males infected with wHa were crossed with uninfected females (54% 8). Poinsot and Merçot (2001) also report low incompatibility in wHa infected males from New Caledonia (57.7% ). In contrast, previous reports revealed strong incompatibly caused by wHa infected males from Hawaii and Tahiti when crossed with uninfected females (greater than 95% , O'Neill and Karr, 1990; Merçot et al, 1995b; James and Ballard, 2000). Also, we find low (22% 8) but significant incompatibility between wNo males and uninfected females. In previous studies, males carrying wNo were incompatible (78–84% ; Merçot et al, 1995b) or partially incompatible (56% ; Merçot and Poinsot, 1998) with uninfected females. Consistent with previous reports, we observed that males carrying wNo are compatible with doubly infected females (7% 3) and partially incompatible with females carrying wHa (43% 1; Merçot et al, 1995b; Merçot and Poinsot, 1998).

Assuming that the differences in incompatibility are not an artifact of the methodologies employed to conduct the assays, there are at least three explanations for the observed results. First, the recently collected Wolbachia isolates from New Caledonia may have diverged from previous collections. Sequencing additional loci may help test this hypothesis. Second, the genetic background of the host may have diverged from the other lines tested. The phenotype of incompatibility and segregation of Wolbachia is dependent on the genetic background of the host line (Boyle et al, 1993; Poinsot et al, 2000). The latter alternative may be addressed by microinjecting, or backcrossing, the wHa and wNo isolates from New Caledonia into standard hosts to standardize for host genetic background. Microinjection will permit the separation of host vs symbiont effects, as well as the interaction between the two. Backcrossing maintains the maternal cytotype such that the Wolbachia-mitochondrial interaction is maintained. Third, the wNo strain of Wolbachia may exhibit varying levels of incompatibility. James and Ballard (2000) previously reported that wMa infected males are heterogeneous in their expression of incompatibility. The wNo and wMa isolates are identical in the region of the wsp locus sequenced, and differ by a single substitution in the 16S rDNA. As a consequence it is possible that wNo and wMa are not distinct strains but sequence variants of the same strain. Resolution of this issue is not just taxonomic but also has important implications for the evolution of Wolbachia in D. simulans. Determination of strain status in Wolbachia is not simple and we discuss this nomenclatural issue more generally below.

One step to resolve these alternatives is to compare recently collected isolates with 'type' bacterial strains and fly lines. An informal meeting of biologists at Wolbachia 2000 (an international Wolbachia meeting in Crete, Greece) identified 'type' D. simulans/Wolbachia complexes (Table 3). These lines and Wolbachia sequence variants are available from the laboratories of A. A. Hoffmann (La Trobe University, Australia), H Merçot (Jacques Monod, France) and the authors.

Table 3 - Lines of D. simulans and the 'type' Wolbachia sequence variants defined by 16s rDNA and Wolbachia surface protein coding sequences.

Full table (16K)

Mitochondria

Two Wolbachia strains infect the siII haplotype. James and Ballard (2000) observed that wRi infected flies were always associated with siIIA mtDNA and wAu flies the siIIB mtDNA. In this study, all lines had the siI haplotype but Wolbachia infection status was not correlated with a specific mtDNA genotype and we hypothesize that singly infected flies arose through stochastic loss of one bacterial strain from a doubly infected mother, as has been predicted theoretically (Frank, 1998). Uninfected flies may have arisen from the loss of infection from either doubly or singly infected flies. In this study, about 7.2% of lines were wHa, 5.5% wNo, and 1.8% uninfected. We do not have estimates on the loss of infection in the field, however, Poinsot et al (2000) studied the segregation rate from doubly infected mothers to singly infected and uninfected progeny in the laboratory. On the basis of backcrossing studies in the laboratory, they estimated that doubly infected females could produce 3.5% wHa singly infected, 1.8% wNo singly infected and 0.8% uninfected eggs. Perrot-Minnot et al (1996) also observed stochastic loss of Wolbachia double infections after an artificially long diapause in Nasonia lines.

In future studies we will endeavor to reconstruct the movement of Wolbachia strains and mtDNA genotypes around the world. This study gives our first glimpse into the movement of the siI mtDNA haplotype and the Wolbachia strains it harbors. It is likely that the D. melanogaster subgroup diverged in East Africa (Lachaise et al, 1986) and the simulans clade in the islands of East Africa. Doubly infected siI flies have only been collected from the Seychelles and New Caledonia suggesting that the flies traveled from the Seychelles to New Caledonia. However, additional collections between these islands is needed. New Caledonia may be the source of the wHa infection that has spread to other Pacific Islands. Lines collected in Hawaii carried a subset of the bacterial strains (only wHa).

Nomenclature

Currently, there is no established nomenclatural system and a variety of criteria have been employed to define a strain of Wolbachia. In D. simulans, strains have been designated on the basis of incompatibility phenotype (Hoffmann et al, 1986; Merçot et al, 1995b; O'Neill and Karr, 1990), 16S rDNA sequence variation as little as a single base pair change (O'Neill et al, 1992; Rousset et al, 1992; Hoffmann et al, 1996; James and Ballard, 2000), and host collection locality (Merçot and Poinsot, 1998). Lincoln et al (1998) define a 'strain' as 'a group of individuals with common physiological traits and presumed common ancestry; an infraspecific group having characteristic properties'. 'Presumed common ancestry' can be determined independently of the host by studying DNA sequence variation. In contrast, 'common physiological traits' and 'characteristic properties' are best studied in the context of a specific host genetic background.

One methodology to test 'presumed common ancestry' (Lincoln et al, 1998) from Wolbachia isolates is to construct phylogenetic hypotheses from DNA sequence data, and identify strains that are reciprocally monophyletic. The phylogenetic method is independent of host genotype but it is not always clear how selection and recombination (Schulenburg et al, 2000; Jiggins et al, 2001; Werren and Bartos, 2001) influence phylogenetic reconstruction (Slowinski and Page, 1999; Ballard, 2000c). As a consequence great care must be taken when defining a strain phylogenetically. An alternative would be to construct an arbitrary level of sequence divergence at a specific locus. However, arbitrary rules of phenetic divergence may not be biologically meaningful (Ballard et al, in press).

Wolbachia have a variety of 'physiological traits' (Lincoln et al, 1998) that are likely to be dependent on the host-genetic background. These include density in the host (Breeuwer and Werren, 1993), segregation rate (Hoffmann and Turelli, 1988; Hoffmann et al, 1990), and their influence on specific life history traits (Hoffmann and Turelli, 1997). We suggest that effects of Wolbachia on each of these traits should be studied in standardized host genetic backgrounds.

A complementary method to define a strain is 'an infraspecific group having characteristic properties' (Lincoln et al, 1998). One characteristic property of Wolbachia in D. simulans is its incompatibility phenotype. Although strains may induce strong, weak, or intermediate incompatibility, these simple phenotypic definitions are confounded by differences in host genetic background, which can greatly affect the bacterial expression of incompatibility (Poinsot and Merçot, 2001). This plasticity of phenotypic expression makes us wary of using the level of incompatibility in the definition of a strain until both host genetic background and Wolbachia density are controlled. Phenotypic definitions may become especially labile under certain environmental conditions, such as larval rearing conditions (Sinkins et al, 1995), heat shock or multiple matings (Hoffmann et al, 1986; Snook et al, 2000).

Here we investigate the effects of single and double Wolbachia infections on expression of incompatibility and mtDNA divergence of D. simulans from New Caledonia. Doubly infected females were compatible with all males in the population, explaining the high proportion of doubly infected flies. Males infected with wHa from New Caledonia showed reduced incompatibility when mated to uninfected females compared to males from Hawaii or Tahiti. Also, males carrying wNo had reduced incompatibility from studies previously reported. These data suggest that the DNA of these bacterial isolates may have diverged from those previously collected, the genetic background of the host has lead to a reduction in the phenotype of incompatibility, and/or wNo infected males, like wMa infected males, are heterogeneous in their expression of incompatibility. There was no association between mtDNA sequence polymorphism and infection type suggesting that single and uninfected flies arise from stochastic loss of bacteria strains. These points stimulate us to contemplate the factors that should be considered when designating Wolbachia strains.

References

1. Baba-Aïssa, F, Solignac, M, Dennebouy, N, David, JR (1988). Mitochondrial DNA variability in Drosophila simulans: quasi absence of polymorphism within each of the three cytoplasmic races. Heredity, 61: 419–426. | PubMed |
2. Ballard, JWO, Chernoff, B, James, AC, (in press). Divergence of mitochondrial DNA is not corroborated by nuclear DNA, morphology, or behavior in Drosophila simulans. Evolution.
3. Ballard, JWO (2000a). Comparative genomics of mitochondrial DNA in Drosophila simulans. J Mol Evol, 51: 64–75.
4. Ballard, JWO (2000b). Comparative genomics of mitochondrial DNA in members of the Drosophila melanogaster subgroup. J Mol Evol, 51: 48–63. | PubMed |
5. Ballard, JWO (2000c). When one is not enough: introgression of mitochondrial DNA in Drosophila. Mol Biol Evol, 17: 1126–1130. | PubMed | ChemPort |
6. Boyle, L, O'Neill, SL, Robertson, HM, Karr, TL (1993). Interspecific and intraspecific horizontal transfer of Wolbachia in Drosophila. Science, 260: 1796–1799. | Article | PubMed | ISI | ChemPort |
7. 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. | PubMed | ISI | ChemPort |
8. Breeuwer, JAJ, Werren, JH (1993). Cytoplasmic incompatibility and bacterial density in Nasonia vitripennis. Genetics, 135: 565–574. | PubMed | ISI | ChemPort |
9. Caspari, E, Watson, GS (1959). On the evolutionary importance of cytoplasmic sterility in mosquitoes. Evolution, 13: 568–570. | Article | ISI |
10. Frank, SA (1998). Dynamics of cytoplasmic incompatability with multiple Wolbachia infections. J Theor Biol, 192: 213–218. | Article | PubMed |
11. Hoffmann, AA, Clancy, D, Duncan, J (1996). Naturally-occurring Wolbachia infection in Drosophila simulans that does not cause cytoplasmic incompatibility. Heredity, 76: 1–8. | PubMed |
12. Hoffmann, AA, Turelli, M (1988). Unidirectional incompatibility in Drosophila simulans: inheritance, geographic variation and fitness effects. Genetics, 119: 435–444.
13. Hoffmann, AA, Turelli, M (1997). Cytoplasmic incompatibility in insects. In: O'Neill SL, Hoffmann AA, Werren JH (eds) Influential Passengers: inherited microorganisms and invertebrate reproduction, Oxford University Press: Oxford, pp 42–80.
14. Hoffmann, AA, Turelli, M, Harshman, LG (1990). Factors affecting the distribution of cytoplasmic incompatibility in Drosophila simulans. Genetics, 126: 933–948. | PubMed | ChemPort |
15. Hoffmann, AA, Turelli, M, Simmons, GM (1986). Unidirectional incompatibility between populations of Drosophila simulans. Evolution, 40: 692–701. | Article |
16. James, AC, Ballard, JWO (2000). Expression of cytoplasmic incompatibility in Drosophila simulans and its impact on infection frequencies and distribution of Wolbachia pipientis. Evolution, 54: 1661–1672. | PubMed | ISI | ChemPort |
17. Jiggins, FM, Von Der Schulenburg, JH, Hurst, GD, Majerus, ME (2001). Recombination confounds interpretations of Wolbachia evolution. Proc R Soc Lond B, 268: 1423–1427. | Article | PubMed |
18. Kambhampati, S, Rai, KS, Verleye, DM (1992). Frequencies of mitochondrial DNA haplotypes in laboratory cage populations of the mosquito, Aedes albopictus. Genetics, 132: 205–209. | PubMed |
19. Lachaise, D, David, JR, Lemeunier, F, Tsacas, L (1986). The reproductive relationships of Drosophila sechellia with D. mauritiana, D. simulans, and D. melanogaster from the Afrotropical region. Evolution, 40: 262–271.
20. Lincoln, R, Boxshall, G, Clark, P (1998). A Dictionary of Ecology, Evolution and Systematics. Cambridge University Press: Cambridge.
21. Merçot, H, Atlan, A, Jacques, M, Montchamp-Moreau, C (1995a). Sex-ratio distortion in Drosophila simulans: co-occurrence of a meiotic drive and a suppressor of drive. J Evol Bio, 8: 283–300.
22. Merçot, H, Llorente, B, Jacques, M, Atlan, A, Montchamp-Moreau, C (1995b). Variability with the Seychelles cytoplasmic incompatibility system in Drosophila simulans. Genetics, 141: 1015–1023. | PubMed | ISI | ChemPort |
23. Merçot, H, Poinsot, D (1998). Wolbachia transmission in a naturally bi-infected Drosophila simulans strain from New Caledonia. Entomol Exp Appl, 86: 97–103.
24. Montchamp-Moreau, C, Ferveur, JF, Jacques, M (1991). Geographic distribution and inheritance of three cytoplasmic incompatibility types in Drosophila simulans. Genetics, 129: 399–407. | PubMed | ChemPort |
25. Nigro, L, Prout, T (1990). Is there selection on RFLP differences in mitochondrial DNA? Genetics, 125: 551–555. | PubMed | ChemPort |
26. O'neill, SL, Giordano, R, Colbert, AM, Karr, TL, Robertson, HM (1992). 16S rRNA phylogenetic analysis of the bacterial endosymbionts associated with cytoplasmic incompatibility in insects. Proc Natl Acad Sci USA, 89: 2699–2702. | Article | PubMed | ChemPort |
27. O'neill, SL, Karr, TL (1990). Bidirectional incompatibility between conspecific populations of Drosophila simulans. Nature, 348: 178–180. | Article | PubMed | ChemPort |
28. Perrot-Minnot, MJ, Guo, LR, Werren, JH (1996). Single and double infections with Wolbachia in the parasitic wasp Nasonia vitripennis: effects on compatibility. Genetics, 143: 961–972. | PubMed | ChemPort |
29. Poinsot, D, Merçot, H (2001). Wolbachia infection from usual to naive host in Drosophila simulans (Diptera: Drosophilidae). Eur J Entomol, 98: 25–30.
30. Poinsot, D, Montchamp-Moreau, C, Merçot, H (2000). Wolbachia segregation rate in Drosophila simulans naturally bi-infected cytoplasmic lineages. Heredity, 85: 191–198. | Article |
31. Rousset, F, Solignac, M (1995). Evolution of single and double Wolbachia symbioses during speciation in the Drosophila simulans complex. Proc Natl Acad Sci USA, 92: 6389–6393. | PubMed | ChemPort |
32. Rousset, F, Vautrin, D, Solignac, M (1992). Molecular identification of Wolbachia, the agent of cytoplasmic incompatibility in Drosophila simulans, and variability in relation with host mitochondrial types. Proc R Soc Lond B,, 247: 163–168.
33. Schulenburg, JHG, Hurst, GDD, Huigens, TME, Van Meer, MMM, Jiggins, FM, Majerus, MEN (2000). Molecular evolution and phylogenetic utility of Wolbachia ftsZ and wsp gene sequences with special reference to the origin of male-killing. Mol Biol Evol, 17: 584–600. | PubMed | ChemPort |
34. Sinkins, SP, Braig, HR, O'Neill, SL (1995). Wolbachia superinfections and the expression of cytoplasmic incompatibility. Proc R Soc Lond B, 261: 325–330. | PubMed | ChemPort |
35. Slowinski, JB, Page, RDM (1999). How should species phylogenies be inferred from sequence data. Syst Biol, 48: 814–825. | PubMed |
36. Snook, RR, Cleland, SY, Wolfner, MF, Karr, TL (2000). Offsetting effects of Wolbachia infection and heat shock on sperm production in Drosophila simulans: analyses of fecundity, fertility and accessory gland proteins. Genetics, 155: 167–178. | PubMed | ChemPort |
37. Sokal, RR, Rohlf, FJ (1995). Biometry. W. H. Freeman: New York.
38. Solignac, M, Monnerot, M (1986). Race formation and introgression within Drosophila simulans, D. mauritiana and D. sechellia inferred from mitochondrial DNA analysis. Evolution, 40: 531–539.
39. Turelli, M, Hoffmann, AA (1991). Rapid spread of an inherited incompatibility factor in California Drosophila. Nature, 353: 440–442. | Article | PubMed | ISI | ChemPort |
40. Turelli, M, Hoffmann, AA (1995). Cytoplasmic incompatibility in Drosophila simulans: dynamics and parameter estimates from natural populations. Genetics, 140: 1319–1338. | PubMed | ChemPort |
41. Turelli, M, Hoffmann, AA, McKechnie, SW (1992). Dynamics of cytoplasmic incompatibility and mtDNA variation in natural Drosophila simulans populations. Genetics, 132: 713–723. | PubMed | ChemPort |
42. Werren, JH, Bartos, JD (2001). Recombination in Wolbachia. Curr Biol, 11: 431–435. | Article | PubMed | ISI | ChemPort |
43. Yen, JH, Barr, AR (1971). New hypothesis of the cause of cytoplasmic incompatibility in Culex pipiens L. Nature, 232: 657–658. | Article | PubMed | ISI | ChemPort |
44. Zhou, W, Rousset, F, O'Neill, SL (1998). Phylogeny and PCR-based classification of Wolbachia strains using wsp gene sequences. Proc R Soc Lond B, 265: 509–515. | Article | PubMed | ISI | ChemPort |

Acknowledgements

We are grateful to Ary A Hoffman, Hervé Merçot, and Sylvain Charlat for productive discussions leading to the designation of the Wolbachia type strains. We also thank Sylvain Charlat for sending us an unpublished manuscript. Kirrie Ballard helped collect flies in Nouméa. All sequencing was done at the Pritzker Laboratory of Molecular Systematics and Evolution, The Field Museum, Chicago Illinois, USA. Funds were provided by National Science Foundation Grant No. DEB-9702824 and the Field Museum Marshall Field Fund.