Wolbachia-Induced Cytoplasmic Incompatibility Is Associated with Decreased Hira Expression in Male Drosophila

Background Wolbachia are obligate endosymbiotic bacteria that infect numerous species of arthropods and nematodes. Wolbachia can induce several reproductive phenotypes in their insect hosts including feminization, male-killing, parthenogenesis and cytoplasmic incompatibility (CI). CI is the most common phenotype and occurs when Wolbachia-infected males mate with uninfected females resulting in no or very low numbers of viable offspring. However, matings between males and females infected with the same strain of Wolbachia result in viable progeny. Despite substantial scientific effort, the molecular mechanisms underlying CI are currently unknown. Methodology/Principal Findings Gene expression studies were undertaken in Drosophila melanogaster and D. simulans which display differential levels of CI using quantitative RT-PCR. We show that Hira expression is correlated with the induction of CI and occurs in a sex-specific manner. Hira expression is significantly lower in males which induce strong CI when compared to males inducing no CI or Wolbachia-uninfected males. A reduction in Hira expression is also observed in 1-day-old males that induce stronger CI compared to 5-day-old males that induce weak or no CI. In addition, Hira mutated D. melanogaster males mated to uninfected females result in significantly decreased hatch rates comparing with uninfected crosses. Interestingly, wMel-infected females may rescue the hatch rates. An obvious CI phenotype with chromatin bridges are observed in the early embryo resulting from Hira mutant fertilization, which strongly mimics the defects associated with CI. Conclusions/Significance Our results suggest Wolbachia-induced CI in Drosophila occurs due to a reduction in Hira expression in Wolbachia-infected males leading to detrimental effects on sperm fertility resulting in embryo lethality. These results may help determine the underlying mechanism of CI and provide further insight in to the important role Hira plays in the interaction of Wolbachia and its insect host.


Introduction
Wolbachia are endosymbiotic bacteria that infect many species of arthropods and filarial nematode [1,2]. The successful spread of Wolbachia can be partially attributed to their powerful ability to alter host reproduction by mechanisms such as cytoplasmic incompatibility (CI), parthenogenesis, male killing, and feminization. CI is the most common phenotype, which is expressed as embryonic lethality when Wolbachia-infected males mate with uninfected females or with females infected with a different Wolbachia strain [3].
Although the molecular mechanism of CI has not been elucidated, several studies suggest that sperm is modified by Wolbachia during spermatogenesis. This modification prevents the paternal chromosomes from entering the anaphase of the first division, thus leading to a defect in embryogenesis except that the same Wolbachia strain is present in the egg and rescue of CI occurs resulting in hatched embryos [3][4][5]. Studies of Wolbachia during spermatogenesis in Drosophila have led to a Wolbachia-Infected spermatocyte/spermatid hypothesis, which suggests that all CIexpressing strains have Wolbachia-infected spermatocytes/spermatids (WISS+cysts) [6]. However, subsequent work in other insect hosts suggests there is little correlation between Wolbachia density in spermatocytes/spermatid and the strength of CI. For example, in the parasitic wasp Nasonia vitripennis, Wolbachia are found only in around 28% of developing sperm, but induce almost complete CI with nearly 100% embryo mortality [7,8]. This is probably because Wolbachia may produce an unknown diffusible CIinducing factor that can spread from infected to uninfected cells throughout the testis [7]. When the sperm derived from Wolbachiainfected males fertilizes uninfected eggs, the first mitotic division in the embryos is severely disrupted. As a result, when the fertilized egg proceeds to anaphase, paternal chromosomes either fail to segregate or appear as extensive chromosome bridging and fragmentation during segregation, indicating damaged or incompletely replicated chromosomes [9,10]. Tram and Sullivan observed that a delay in nuclear envelope breakdown and activation of cyclin-dependent kinase 1 (cdk1) in the male pronuclei occurred relative to that in the female pronuclei in Nasonia [11]. This delay is thought to slow down chromosome condensation in male pronuclei as cdk1 activation is required to drive chromosome condensation [12]. Recently, Landmann et al. found that CI delayed deposition of histone H3.3/H4 complex in the male pronucleus, which could be the cause of the chromosome defects present during the first mitotic division in CI embryos [13].
HIR/HIRA, a chaperone of histone H3.3, was first identified in yeast as a negative regulator of histone gene expression [14]. It contains a conserved family of proteins found in various organisms including Drosophila, Xenopus, mice and human and plays an essential role in development [15]. In Drosophila, a point mutation of Hira gene (Hira ssm , originally called sésame gene) causes female sterility. When the eggs laid by homozygous ssm females are fertilized,the formation of male pronuclei is arrested in the late chromatin decondensing stage. Therefore the paternal chromatin can not participate in the embryonic development [16]. Further studies have demonstrated that HIRA functions in replicationindependent deposition of H3.3-H4 tetramers in the male pronucleus [17]. Loss of function allele (Hira HR1 ) reveals that the HIRA has the only essential role in the assembly of paternal chromatin during male pronucleus formation, since the mutation does not affect the viability of the flies [18].
As both CI embryos and Hira mutated flies result in defects in the formation of the male pronucleus, we investigated whether the strength of Wolbachia-induced CI is correlated with Hira expression level in Drosophila flies. Our results show that in both Drosophila melanogaster and Drosophila simulans males infected by Wolbachia strains that induce strong CI, Hira expression levels are significantly decreased compared to males exhibiting no CI or Wolbachia-uninfected males. In addition, Hira expression in 1-dayold Wolbachia-infected males inducing strong CI is also significantly reduced relative to 5-day-old Wolbachia-infected males exhibiting weak or no CI. Furthermore, we demonstrate that Hira mutated male flies mimic the CI phenotype, suggesting that Wolbachiainduced CI in Drosophila may occur by reducing Hira expression in male flies. These results provide an important insight into a novel pathway in which Wolbachia interacts with its insect hosts.

Reduced Hira expression in males expressing strong CI
To test the correlation of CI strength with Hira expression level, we initially tested the CI strength in 1-day-old D. melanogaster males reared under uncrowded conditions [19]. The results of crossing experiments show that CI is only induced in matings between 1day-old Dmel wMel males and uninfected Dmel T females (hatch rate of 8.7861.03). In contrast, no CI is induced by Dmel wAu 1day-old males (hatch rate of 95.5561.26) ( Table 1).
To determine whether Hira expression is involved in CI level in D. melanogaster males, a quantitative RT-PCR assay was performed on 1-day-old male flies. As shown in Figure 1, Hira expression was significantly lower in wMel-infected males relative to uninfected males (Dmel wMel/Dmel T: 0.1560.05) (P,0.01). In contrast, Hira expression was ,3 fold higher in wMel-infected females relative to uninfected females (Dmel wMel/Dmel T: 2.9660.16) (P,0.05). Surprisingly, the non CI inducing wAu strain did not result in dramatically decreased Hira expression in males (Dmel wAu/Dmel T: 0.9260.03). Hira expression was 1.90 fold higher in wAu-infected females relative to uninfected females (Dmel wAu/ Dmel T: 1.9060.07) (P,0.05) ( Figure 1).
In order to determine if the correlation between CI level and Hira expression occurs in additional combinations of Wolbachia and their hosts, we examined Hira expression in D. simulans infected with wRi and wAu. Crossing experiments were undertaken with D. simulans lines and strong CI was observed to occur in matings between 1-day-old males infected with wRi and Dsim T females (hatch rate of 7.6762.18). As in D. melanogaster, the wAu strain did not induce CI in D. simulans (hatch rate of 87.6262.67) ( Table 2).

Reduced Hira expression in younger males expressing strong CI
Reynolds and Hoffmann reported that CI levels declined rapidly with increasing of male age in Drosophila strains infected by Wolbachia [20]. In order to further investigate the correlation of CI intensity with Hira expression in Wolbachia-infected male flies, we compared Hira expression levels between 1-day-old and 5-day-old males of Dmel wMel. As shown in Table 1, matings between Wolbachia-uninfected Dmel T females and 1-day-old Dmel wMel males result in significantly lower hatch rates (8.7861.03) in comparison to that of the 5-day-old Dmel wMel males (89.5161.33%) (P,0.01). However, in the reciprocal crosses between 1-day and 5-day old Wolbachia-uninfected males mated with the Dmel wMel females, there were no significant difference in hatch rates (91.6662.92% and 92.9961.36%, respectively) (P.0.05). This confirmed that in D. melanogaster, 1-day-old males induced strong CI, whereas 5-day-old males expressed no CI.
In crosses between Dsim wRi males and uninfected Dsim T females, hatch rates were also correlated with male age. When 5day-old males were used in the crosses, the hatch rate of the embryos was 50.8965.56%, which is significantly higher than that in crosses with 1-day-old males (7.6762.18%) ( Table 2). As expected, Hira expression in 5-day-old Dsim wRi males was also significantly increased compared to 1-day-old males (P,0.05). Hira expression between 1-day and 5-day old Dsim T males did not differ significantly (P.0.05) ( Figure 4) confirming a similar effect in both D. melanogaster and D. simulans.

Paternal effects of Hira mutation mimic CI in D. melanogaster
To examine the involvement of Hira expression in the CI phenotype, the crossing pattern of two Hira-mutated D. melanogaster lines (Hira ssm and Hira HR1 ) was assessed. Interestingly, we found that both Hira-mutated males (1-day-old) mated to Wolbachiauninfected Dmel T females resulted in significantly lower egg hatch rates (72.9865.10%, 74.3464.03%, respectively) relative to Dmel T males (92.4460.77%) (P,0.05) ( Table 3). However, the crosses between Hira-mutated males and wMel-infected females resulted in no significant differences of hatch rate comparing with uninfected crosses (86.8164.37%, 89.38.3468.06%, respectively) ( Table 3). In the early embryos derived from the crosses of Hiramutant males with Dmel T females, the asynchronous cleavage and chromatin bridges were observed ( Figure 5, C) which is similar to the CI phenotype in D. melanogaster ( Figure 5, B), while in the embryos derived from the uninfected flies the nuclear division was synchronous and the nuclei were evenly distributed ( Figure 5, A). This suggests that the mutation of Hira in males may mimic the CI phenotype induced by Wolbachia.
Considering that Hira is on the X chromosome, we also examined the offspring sex ratio to see whether Hira mutation impacts sperm quantity. As shown in Table 3, mating with both Hira-mutant males resulted in significantly less female progeny in the next generation. In the progeny, the ratio of females to males derived from the crosses with Hira ssm and Hira HR1 were 0.68 (60.08) and 0.60 (60.11), respectively. These sex ratios are significantly lower than those result from crosses with Dmel T males (P,0.05), where the ratio was 0.98 (60.10) ( Table 3). These results suggest that many female progeny mortality was associated with Hira mutations on one X chromosome. Therefore we conclude that Hira mutation has negative effects on sperm quantity. In addition, the presence of the wMel strain of Wolbachia in females restored the sex ratio as shown by a sex ratio of ,1 when Hira-mutated males were crossed with Dmel wMel females (Table 3).

Discussion
Previous studies both in vitro and in vivo have shown that Wolbachia infection may affect the expression of various host genes, including those associated with immunity, fertilization, and development [21][22][23]. For example, wMelPop strain of Wolbachia has been shown to be capable of inducing immune upregulation in Anopheles gambiae and Aedes aegypti mosquitoes [23,24]. Even in a cell line naturally infected with Wolbachia, the expression of antioxidant proteins is also upregulated [22]. However, the association of CI strength induced by Wolbachia with the gene expression level in hosts is unknown. Here, we used two lines from both D. melanogaster and D. simulans infected with different Wolbachia strains to determine if Hira gene expression is correlated with CI. Our results demonstrate that for both Dmel wMel and Dsim wRi 1day-old males, which express strong CI, the Hira expression levels are significantly decreased compared to Wolbachia-uninfected and wAu-infected males which induce either weak or no CI. Furthermore, increasing male age was correlated with increasing hatch rates (Tables 1, 2). Correspondingly, Hira expression was significantly lower in younger males (Figures 3, 4), suggesting that down regulation of Hira expression in male Drosophila might be causally linked to the CI strength.
The molecular mechanisms of abnormal embryo development in CI crosses are not fully known. Genetic and cellular evidences indicate that in CI embryos, the paternal chromosomes are improperly condensed when aligned at the metaphase plate of the first mitotic division after fertilization [9,10,25]. This could be attributed to the modification in the sperm of Wolbachia-infected  males. Riparbelli et al. have described some malformations, including abnormal mitochondria and axoneme in the sperm developed within the infected testes [7]. Additionally, the amount of sperm produced by Wolbachia-infected D. simulans males is about 40% of that in uninfected males, especially in young males expressing strong CI, indicating that Wolbachia could affect male fertility through multiple ways [26]. Studies of spermatogenesis in the parasitic wasp Nasonia vitripennis and the beetle Chelymorpha alternans revealed that Wolbachia can modify sperm despite not being present in developing sperm, suggesting that Wolbachia might alter expression and synthesis of gene products in the host, thus changing the products exported to the developing spermatids [8].
In this study, we show that Wolbachia strains that induce strong CI (wMel and wRi) significantly decrease Hira expression in male Drosophila flies, which may impair the sperm function at fertilization and result in the CI phenotype. Chromatin remodeling is extremely significant in the late stage of spermatogenesis due to the necessity of histone replacement by male-specific transition protein (TP) and later by small argininerich proteins named protamines, ensuring the compacted sperm head configuration formation and promoting sperm getting into the female reproductive tract for fertilization [27]. It is known that HIRA is a chaperone of histone variant H3.3 and is involved in a DNA replication-independent pathway of nucleosome assembly [28]. In Drosophila, H3.3 is incorporated in specific regions in the early stages of spermatogenesis and then mostly disappears in condensed spermatid nuclei just before protamine deposition [18,29]. H3.3 could play a role in spermatogenesis. Indeed, male mice carrying an impaired H3.3A gene have reduced fertility [30]. Considering that the process of spermatid differentiation is independent of DNA synthesis, the histone H3.3 chaperone HIRA might be involved in this process. In this study, we have shown that Hira mutations in young male Drosophila flies results in a significantly reduced hatch rates (Table 3) comparing with the uninfected flies, suggesting that HIRA could have an effect on male fertility by acting as a chaperone of H3.3. This is in contrast with the observations by Bonnefoy et al. [18], where they showed that Hira mutation had no effects on viability and male fertility. The contrasting results may be due to the fact that Hira mutated males do not completely lose their fertility but produce significantly less progeny.
Following fertilization, the paternal chromosomes are abnormally condensed during the first zygotic division in the embryos derived from CI crosses [4,5]. Tram and Sullivan found that in CI crosses, the nuclear envelope breakdown and Cdk1 activation are   delayed in the male pronucleus relative to those in the female pronucleus [11]. Recent studies revealed a delay in loading H3.3 onto the paternal chromosomes, possibly causing disruption of replication in the male pronucleus of CI embryos [13]. As the chaperone of H3. 3, HIRA has been demonstrated to be essential in sperm chromatin remodeling and specifically in assembling H3.3 containing nucleosomes during the formation of male pronucleus. Female flies homozygous for the null allele of Hira are sterile due to a defect in incorporated sperm nucleus decondensation [17,18]. In this regard, HIRA could be involved in the abnormal deposition of H3.3 to the paternal chromosome in CI embryos. It is possible that the low level of Hira in young Wolbachia-infected males results in a structural malformation of sperm nucleus. When this sperm fertilize an uninfected egg, HIRA chaperoned maternal H3.3 may not be deposited promptly on the male nucleus during sperm chromatin remodeling. In this study, we have shown that in Wolbachia-infected female flies, Hira expression is dramatically increased in both Dmel wMel and Dsim wRi (Figures 1, 2). High level of HIRA in the females might be able to compensate for the shortage in sperm nucleus resulted from Wolbachia infection, thus speed up the deposition of H3.3 in paternal chromosomes and rescue this defect at fertilization. Several models have been described to explain the rescue mechanism including the hypothesis that Wolbachia removes an essential component from the sperm nucleus and the same strain of Wolbachia in the egg restores this critical factor allowing embryogenesis to proceed normally. This ''titration-restitution'' model [31] can explain why CI induction can be suppressed when infected males mate with females infected with the same strain of Wolbachia [2]. We have shown here that Hira-mutants in young males lead to a higher paternal-effect embryonic lethality with a phenotype of chromatin bridges. Interestingly the egg with Wolbachia may rescue the defects and result in high hatch rates. These mimic the CI phenotype induced by Wolbachia. Furthermore, it is the sperm-carrying X chromosome in Hira-mutants that leads to embryonic lethality, since there are significantly less females produced from these crosses (Table 3). Taken together, our results suggest that Wolbachia may induce CI by regulating the expression of some key factors, such as reducing Hira expression in males, which may influence sperm fertility and cause CI phenotype. For the compatible crosses of Wolbachia-infected females with uninfected males, it is possible that higher level of Hira in females has no effects on the embryogenesis. For bidirectional incompatibility, it is likely that different Wolbachia strains differentially impact Hira expression so that females infected with one Wolbachia strain can not rescue the deficiency in the sperm caused by another Wolbachia strain.

Fly lines
Fly lines were kept on a standard corn diet at a constant temperature of 25uC, with 8L : 16D (light : dark) cycle and were reared under non-crowded condition (200610 eggs per 50 ml vial of media in 150 ml conical flask) [19]. The following Drosophila strains were used in the study: Wolbachia-infected Dmel wMel (D. melanogaster Brisbane nuclear background with introgressed wMel from YW), Dmel wAu (D. melanogaster Brisbane nuclear background with injected wAu), Dsim wRi (D. simulans naturally infected with wRi), and Dsim wAu (D. simulans Coffs Habour). Cured Dmel wMel and Dsim wRi were subsequently generated by tetracycline treatment following established protocols [32] and designated Dmel T and Dsim T, respectively. D. melanogaster of Hira ssm (a point mutant of Hira) [17] and Hira HR1 (a loss of function Hira allele) [18] were provided kindly by Dr. Loppin B. at the Université Claude Bernard Lyon I, France. These two Hira mutated lines were confirmed to be Wolbachia-free by PCR (data not shown).

CI Assays
CI tests were performed as previously described by Yamada et al. [19]. All the crossing schemes including expected compatible and incompatible crosses in this study are shown in Tables 1, 2, and 3. In all crosses, adult virgins were collected and crosses were undertaken with 30 females (3-4 days old) and 20 males (either 1day or 5-day-old) at 25uC in bottles upturned on agar/grape juice plastic Petri dishes (ca.4cm 2 ). After mating for around 10 h, all males were removed from the bottle to avoid diminishing CI  Table 3. Effect of Hira-mutant males (1-day-old) on egg hatch rates and progeny sex ratio.

Quantitative RT-PCR Assay
Quantitative reverse transcription PCR (qRT-PCR) was performed to determine the relative Hira gene expression level in different fly lines. Total RNA was extracted from adults (males or females) using Trizol (Invitrogen). DNA contamination was removed with RQ1 DNase (Promega). The first-strand cDNA was synthesized from 2 mg of total RNA using reverse transcriptase (RT) (Promega) and oligo dT15 primer (Takara) at 42uC for 1.5 h. Special primers were designed based on flybase for Hira and the reference gene rp49 as following: HiraF (59-ATGCGGCTCCT-TAAGCCGGC-39) and HiraR (59-ATCTTCGGAACATCCG-CATCG-39); rp49F (59-CGGTTACGGATCGAACAAGC-39) and rp49R (59-CTTGCGCTTCTTGGAGGAGA-39). qPCR was performed using a Miniopticon system (BioRad) with a Platinum SYBR Green qPCR superMix (Takara). The reaction volume was 20 ml, containing 10 ml SYBR Premix Ex Taq (26), 0.15 ml of forward and reverse primer (20 mM), respectively, 7.7 ml ddH 2 O and 2 ml of cDNA template diluted by 10-fold. The qPCR procedure was consisted of 95uC for 2 min, followed by 95uC for 10 s, 61uC for 15 s and 72uC for 10 s per cycle for 40 cycles, then a melting curve analysis was carried out by a slow increase (0.2uC/ s) from 55uC to 98uC, in purpose of examining if there were primer-dimers or nonspecific amplification. The relative expression ratio of Hira gene for samples A to B was calibrated against rp49 gene using the 2 2DDCT calculation method: DDCT = (CT Hira 2 CT rp49 ) sampleA 2(CT Hira 2CT rp49 ) sampleB .

Immunofluorescence
Embryos were collected every 15 minutes and immersed in a 50% bleach solution for 2-3 mins to remove the chorion. Next they were washed in embryo wash buffer (0.7% NaCl, 0.05% Triton X-100) and fixed by vigorous shaking in a 1:1 heptane/ methanol mix for 30s. Embryos were washed three times with methanol and stored in methanol at 220uC until use. Embryos were washed 5 min in 220uC acetone and then three times (10 min each) in TBST (50 mM Tris-HCl, pH 7.4, 50 mM NaCl, 0.02% sodium azide, 0.1% Triton X-100). The samples were stained with DAPI (Beyotime, China) solution for 15 min at room temperature, then washed using TBST for 10 min. Embryos were mounted in the mounting medium (10% PBS, 90% glycerol). The slides were observed and photographed using a Leica DM 4000B fluorescence microscope (Leica, Germany).

Statistical Analysis
Results are presented as means 6 SE (n = 3). Differences among means were analyzed by one-way analysis of variance (one-way ANOVA). Differences were regarded as statistically significant when P,0.05.