Association of a New Wolbachia Strain with, and Its Effects on, Leptopilina victoriae, a Virulent Wasp Parasitic to Drosophila spp.

Abstract

Wolbachia bacteria are ubiquitous intracellular bacteria of arthropods. Often considered reproductive parasites, they can benefit certain host species. We describe a new Wolbachia strain from Leptopilina victoriae, a Drosophila wasp. The strain is closely related to Wolbachia from Culex sp. Located to the posterior poles of oocytes, it manipulates its host's reproduction by inducing a male development type of cytoplasmic incompatibility. We also report its diverse effects on the wasp's life history traits.

TEXT

The alphaproteobacterium Wolbachia pipientis is an intracellular symbiont found in many arthropod species (12). The clade containing Wolbachia pipientis is divided into 11 to 13 supergroups, depending on the study (1, 3, 18). Wolbachia bacteria are maternally transmitted reproductive parasites (19, 27). A common effect of Wolbachia is the induction of cytoplasmic incompatibility (CI). CI is observed when Wolbachia-infected males mate with either females without Wolbachia or females carrying a different Wolbachia strain. Such reproductive incompatibility gives a fitness advantage to infected females and results in an increase in Wolbachia-infected individuals in the population (27). Wolbachia can also induce male killing, thelytokous parthenogenesis, and feminization of genetic males (21, 27). In some cases, Wolbachia bacteria impose a cost on the fitness of their hosts (5, 6, 17). Conversely, Wolbachia bacteria can protect their hosts against pathogens or improve their fecundity (9, 11, 22, 24). Overall, the biological effects of Wolbachia are strain and host specific.

A Wolbachia strain, belonging to the B supergroup, has been identified in the parasitoid wasp Leptopilina victoriae (7), but little is known about its exact phylogenetic position and effects on L. victoriae. In contrast, the identity and the effects of Wolbachia in its sister species Leptopilina heterotoma have been well characterized (6, 23, 25).

In this study, we show that Wolbachia from L. victoriae constitutes a novel strain type which is related to a strain of Wolbachia from Culex sp. It induces a male development type of CI and also affects size, developmental time, and fecundity of L. victoriae. The range of Drosophila species that L. victoriae attacks is limited, and this host range appears not to be affected by Wolbachia. These results advance our understanding of the biology of L. victoriae in relation to its exceptionally well-studied Drosophila host.

Endosymbiotic bacteria of L. victoriae.

Our laboratory culture of L. victoriae (16, 20) was screened for the presence of five endosymbiotic bacteria known to influence host biology and reproduction, Wolbachia, Cardinium, Arsenophonus, Rickettsia, and Spiroplasma, as described previously (10, 15). Only one strain of Wolbachia was found to infect L. victoriae.

A new Wolbachia strain type in L. victoriae.

The five genes for multilocus sequence typing (MLST) of L. victoriae Wolbachia (gatB, coxA, hcpA, ftsZ, and fbpA) were sequenced. On the basis of a comparison to the Wolbachia MLST database (13; http://pubmlst.org/wolbachia), the alleles for gatB, coxA, and hcpA were determined to be new. The combination of alleles for all five genes in L. victoriae Wolbachia is unique and constitutes a new strain type (ST 306). The L. victoriae Wolbachia strain shares two alleles (ftsZ and fbpA) with Wolbachia from Culex sp.

Phylogenies using wsp and the concatenation of the five MLST genes show some discrepancy for the position of L. victoriae Wolbachia. Based on the wsp phylogeny, L. victoriae Wolbachia is closely related to Wolbachia from Culex murrelli, Nasonia vitripennis, and Leptopilina clavipes (Fig. 1A) while the MLST phylogeny places L. victoriae Wolbachia more closely related to Wolbachia from other Culex sp. (sequences for C. murrelli are unavailable) (Fig. 1B).

Fig 1.

Wolbachia phylogenies. (A) wsp phylogeny based on 590 bp. (B) Phylogeny based on the concatenation of gatB, coxA, hcpA, ftsZ, and fbpA genes (2,074 bp). Both phylogenies have been reconstructed by maximum likelihood. The Wolbachia A group is used as an outgroup. Wolbachia strains are indicated by their host names and strain names when available. Only bootstraps higher than 60 are shown. Each gene was assigned an allele number using the Wolbachia MLST database (gatB195, coxA182, hcpA206, ftsZ22, and fbpA4). Species are abbreviated as follows: A. encedon, Acraea encedon; A. eponina, Acraea eponina; A. vulgare, Armadillidium vulgare; E. formosa, Encarsia formosa; E. mandarina, Eurema mandarina; E. hecabe, Eurema hecabe; G. firmus, Gryllus firmus; H. bolina, Hypolimnas bolina; N. giraulti, Nasonia giraulti; N. longicomis, Nasonia longicomis; S. exempta, Spodoptera exempta; T. confusum, Trichogramma confusum; T. coeruleus, Tetrastichus coeruleus; T. deion, Trichogramma deion. The following are Drosophila species: D. simulans, D. recens, D. prosaltans, D. munda, D. innubila, D. mauritana, and D. sechellia. The following are Culex species: C. pipiens, C. pipiens pallens, C. pipiens quinquefasciatus, C. pipiens molestus, C. neomimulus, and C. quinquefasciatus.

L. victoriae Wolbachia is localized posteriorly in oocytes.

Using Hoechst 33258 staining and confocal microscopy, we detected a single, centrally located nucleus in each egg in both infected (N) and antibiotic (AB)-treated L. victoriae lines (Fig. 2, black arrows). Localized fluorescence, additional to the nucleus, was observed only in infected eggs (Fig. 2C and D, arrowheads). Strikingly, this localized signal is in the vicinity of an unidentified clear, spherical structure (Fig. 2, white arrows). This observation confirms that antibiotic treatment has eliminated Wolbachia in the AB line and suggests a preference for Wolbachia cells to cluster at the posterior pole. Such localization has also been observed in eggs of Asobara tabida and species of Nasonia (2, 4). The posterior end of insect eggs contains germ cell determinants, and, therefore, this localization of Wolbachia in L. victoriae may enhance its vertical transmission (4). While detecting the presence of Wolbachia at the posterior pole of L. victoriae oocytes is not surprising, its association with a 10-μm spherical structure is curious and merits further investigation. Immuno-localization experiments with Wolbachia-specific probes will confirm that this extra fluorescence is due to Wolbachia-specific association in L. victoriae oocytes.

Fig 2.

Wolbachia localization in L. victoriae eggs. Dissected oocytes from the AB (A and B) and N (C and D) lines of L. victoriae stained with Hoechst 33258 are shown. Black arrows point to the oocyte nucleus; white arrows point to an unidentified clear and spherical structure at the posterior pole of oocytes present in both lines. Wolbachia cells (arrowheads) are clustered posteriorly only in oocytes from the N line. Bars, 20 μm.

Wolbachia induces a male development type of cytoplasmic incompatibility in L. victoriae.

Unidirectional CI results in an offspring sex ratio biased toward males when an uninfected female and an infected male mate (21). After three generations of antibiotic treatment to remove Wolbachia and six generations on regular fly food to exclude direct effects of the antibiotic, all four crosses between Wolbachia-infected (N) and antibiotic-treated (AB) L. victoriae insects were performed. Mating behavior was observed, and the sex ratio of the resulting offspring was recorded as described by Vavre et al. (25). The experimental cross between AB females and N males yielded more than 89% male progeny (Fig. 3). This ratio is significantly skewed relative to progeny from the control cross of AB females and AB males, which yielded 48% males (Wilcoxon test [W] = 431, P = 1e−05). No significant difference in sex ratio was found between N × N and N × AB crosses (W = 203, P = 0.68), or between the control crosses N × N and AB × AB (Fig. 3 and Table 1). L. victoriae thus induces partial unidirectional CI, as some (around 10%) females are still observed in the AB × N cross.

Fig 3.

Cytoplasmic incompatibility induced by Wolbachia. The sex ratios of the offspring from four possible crosses are expressed as proportions of males. N is the L. victoriae line infected with Wolbachia. AB is the same line after antibiotic treatment. AB × AB and N × N are the control crosses and yield roughly equal numbers of male and female offspring. However, the percentage of males in the AB × N cross is close to 1. The analysis was done on 15 N × N, 20 AB × AB, 19 N × AB, and 13 AB × N crosses.

Table 1.

Life history traits of L. victoriae^a

Parameter	Value for the parameter by gender and line				Statistics
	Female AB	Female N	Male AB	Male N
Sex ratio (% females)	0.52 ± 0.19 (n = 653)	0.58 ± 0.20 (n = 381)	NA	NA	Finfection = 1.3; P = 0.26; df = 1
Survival (days)	30.7 ± 7 (n = 77)	31.6 ± 8 (n = 84)	NA	NA	x2 = 1.4; P = 0.24; df = 1
Fecundity (8-day period)	135.4 ± 43.1 (n = 14)	111.5 ± 29.5 (n = 18)	NA	NA	Finfection = 5.15; P = 0.03*; df = 1
Size (μm)b	261.3 ± 25 (n = 31)	272.7 ± 14.7 (n = 31)	268.3 ± 12.2 (n = 27)	270.5 ± 14.3 (n = 31)	Finfection = 4.82; P = 0.03*; df = 1
					Fsex = 0.49; P = 0.49; df = 1
					Finteraction = 2.09; P = 0.15; df = 1
Developmental time (days)	23.9 ± 1.4 (n = 137)	24 ± 1.8 (n = 503)	22.4 ± 1.2 (n = 504)	21.9 ± 1.5 (n = 596)	Finfection = 5.29; P = 0.02*; df = 1
					Fsex = 619; P < 2e-16***; df = 1
					Finteraction = 10.59; P = 0.001 **; df = 1

^aN and AB individuals (males and females) are compared for sex ratio, size, survival, developmental time, and realized fecundity. Fecundity was scored over an 8-day period; results show an average of the total number of offspring per female during this time. Average and standard deviation are indicated. Sex ratio data were arcsine transformed. All data except survival were compared using one- or two-way analysis of variance testing the effect of infection status (F_infection), the effect of sex (F_sex), or the interaction of infection status and sex (F_interaction). Survival curves were compared by a log rank test. n, number of insects; NA, not applicable. *, 0.05 ⩽ P ⩽ 0.01; **, 0.01 < P ⩽ 0.001; ***, P < 0.001.

^bTibia length.

Two types of CI have been described in haplo-diploid species. In the female mortality type, not only is the sex ratio distorted toward males but the eggs also die early during development, resulting in a reduced number of total offspring (2, 26). In the male development type of CI, embryos experiencing CI develop as unfertilized embryos, thus giving rise to males and a similar number of total offspring as in the controls. In L. victoriae, the total number of offspring in the AB × N cross compared to the AB × AB cross was not significantly different (W = 197, P = 0.33). This result suggests that the CI induced by L. victoriae Wolbachia is of the male development type—a condition that is similar to one described in Nasonia (2) where improper condensation of the paternal chromosome and its eventual loss leads to CI (2). Whether a similar mechanism influences the viability of L. victoriae embryos remains to be investigated.

Wolbachia effects on L. victoriae life history traits.

While Wolbachia primarily affects the reproduction of its insect host, it can also impose some costs or provide benefits to its carrier. We found that some fitness-related traits in L. victoriae are positively affected by Wolbachia. Infected males tend to develop faster than their uninfected counterparts by about half a day (Table 1). In previous studies (6, 8, 14), the developmental time has been shown either to not be affected or to be negatively affected by Wolbachia. Accelerated development of Wolbachia-infected L. victoriae males might be advantageous as they could mate with females that are first to emerge. Furthermore, Wolbachia-infected L. victoriae insects tend to be larger than their uninfected counterparts, based on the measurement of tibia length (Table 1). Thus, Wolbachia bacteria appear to confer a metabolic and developmental advantage to their L. victoriae hosts, strategies that might promote Wolbachia spreading within the population. Conversely, Wolbachia adversely affects the wasp's fitness by decreasing the total number of offspring of infected females, based on offspring emergence over an 8-day period (Table 1). A comparable result has also been described in L. heterotoma (6). Thus, in both of these closely related wasp species, Wolbachia is virulent. Sex ratio and female survival were not affected by the presence of Wolbachia (Table 1).

Finally, we examined if Wolbachia could modify L. victoriae's ability to infect a range of Drosophila species. L. victoriae is a parasitoid of Drosophila species. We found that, in lab experiments, L. victoriae is unable to parasitize Drosophila hydei (repleta group), Drosophila virilis (virilis group), Drosophila pseudoobscura (obscura group), or Drosophila willistoni (willistoni group) as not a single fly larva showed the presence of wasp upon dissection. The number of live (success of the wasp) or encapsulated (success of the fly's immune system) wasp eggs for the remaining species was recorded. Wolbachia appears to affect neither the host range of L. victoriae nor the rate at which the wasp eggs are encapsulated by their fly hosts (Mann-Whitney tests between N and AB; for Drosophila melanogaster, 41% versus 36% [W = 912, P = 0.59]; for Drosophila yakuba, 79% versus 75% [W = 280, P = 0.86]) (Fig. 4).

Fig 4.

Host range and encapsulation rate of N and AB lines of the wasp L. victoriae in eight Drosophila species. Numbers indicate the percentages of fly larvae infected by L. victoriae. The asterisks (*) indicate experiments where a significant (more than 5%) percentage of wasp eggs or larvae were encapsulated by fly hosts. Drosophila groups to which species belong are indicated. Drosophila phylogeny is based on the alignment of Adh sequences from the GenBank.

In sum, even though we did not detect clear effects of Wolbachia on the fly immune response or the host range of L. victoriae, its effects on reproduction and life history traits of L. victoriae are interesting and complex; not all traits are affected the same way. While some traits such as fecundity are adversely affected, others such as size and developmental time in males are improved. The mechanisms by which these effects of Wolbachia in L. victoriae are realized remain unclear. Our study points to the possibility that the multiplicity of effects reported here likely influences natural L. victoriae populations.

Nucleotide sequence accession numbers.

All sequences were deposited into GenBank under the following accession numbers: JQ868788 for wsp, JQ868789 for gatB, JQ868790 for coxA, JQ868791 for hcpA, JQ868792 for ftsZ, and JX127254 for fbpA.