Maternally inherited intracellular bacteria Wolbachia cause both parasitic and mutualistic effects on their numerous insect hosts, including manipulating the host reproductive system in order to increase the bacteria spreading in a host population, and increasing the host fitness. Here, we demonstrate that the type of Wolbachia infection determines the effect on Drosophila melanogaster egg production as a proxy for fecundity, and metabolism of juvenile hormone (JH), which acts as gonadotropin in adult insects. For this study, we used six D. melanogaster lineages carrying the nuclear background of interbred Bi90 lineage and cytoplasmic backgrounds with or without Wolbachia of different genotype variants. The wMelCS genotype of Wolbachia decreases egg production in infected D. melanogaster females in the beginning of oviposition and increases it later (from the sixth day after eclosion), whereas the wMelPop Wolbachia strain causes the opposite effect, and the wMel, wMel2 and wMel4 genotypes of Wolbachia do not show any effect on these traits compared with uninfected Bi90 D. melanogaster females. The intensity of JH catabolism negatively correlates with the fecundity level in the flies carrying both wMelCS and wMelPop Wolbachia. The JH catabolism in females infected with genotypes of the wMel group does not differ from that in uninfected females. The effects of wMelCS and wMelPop infection on egg production can be levelled by the modulation of JH titre (via precocene/JH treatment of the flies). Thus, at least one of the mechanisms promoting the effect of Wolbachia on D. melanogaster female fecundity is mediated by JH.
Wolbachia are widespread maternally transmitted endosymbiotic bacteria of invertebrates capable of affecting the host's reproduction to enhance their own spread and transmission through the host generations (O'Neill et al., 1997; Mateos et al., 2006; Werren et al., 2008; Weinert et al., 2015). Wolbachia symbionts cause cytoplasmic incompatibility in different insect species, including Drosophila melanogaster, manifested as embryonic mortality in crosses between infected males and uninfected females, often resulting in an increase in frequency of infected flies in the population (Hoffmann, 1988; Werren, 1997; Yamada et al., 2007; Ilinsky and Zakharov, 2011). However, although Wolbachia’s dramatic effects on Drosophila sperm and male fertility is well known (Serbus et al., 2008; Werren et al., 2008), its influence on female reproductive biology is slightly more puzzling. In a previous study, Charlat et al. (2004) found no effect of Wolbachia infection on female fertility: in crosses with both infected and uninfected males, infected females were not significantly more or less fertile than uninfected ones. In contrast, Fry et al. (2004) and Weeks et al. (2007) demonstrated enhanced fecundity in D. melanogaster females infected with Wolbachia compared with uninfected females. Weeks et al. (2007) suggested that fecundity-increasing types of Wolbachia infection could be polymorphic in natural Drosophila populations. We therefore studied the impact of five Wolbachia variants on female reproductive biology in D. melanogaster.
Six different genotypes of Wolbachia pipientis have been identified in D. melanogaster – wMel, wMel2, wMel3, wMel4, wMelCS and wMelCS2 (Riegler et al., 2005; Ilinsky, 2013) – plus the pathogenic wMelPop strain, which is a variant of the wMelCS genotype that causes early death of flies (Min and Benzer, 1997; Ilinsky, 2013). The genotypes differ in the level of host antiviral protection and their load in the host (Chrostek et al., 2013; Wong et al., 2015). To study the effects of various Wolbachia genotypes on D. melanogaster fitness, we have previously created a study system (Gruntenko et al., 2017) that consists of five conplastic lineages carrying the nuclear background of one wild-type lineage, Bi90, and cytoplasmic backgrounds with wMel, wMel2, wMel4, wMelCS and wMelPop genotype variants of Wolbachia. In our previous study, the wMelCS genotype of Wolbachia intensified dopamine metabolism in D. melanogaster, whereas wMel, wMel2 and wMel4 did not affect it, and wMelPop decreased it (Gruntenko et al., 2017). Dopamine is known to rise quickly and steeply under stress conditions, playing a significant role in the regulation of response to oxidative and heat stress, influencing survival in Drosophila (Gruntenko et al., 2004; Ueno et al., 2012; Hanna et al., 2015). Wolbachia infection is also associated with the induction of oxidative stress in Aedes aegypti (Pan et al., 2011) and two Drosophila species infected by several Wolbachia strains, including wMelCS (Wong et al., 2015). As dopamine is shown to interact with juvenile hormone (JH) in D. melanogaster females (Gruntenko and Rauschenbach, 2008; Gruntenko et al., 2012; Argue et al., 2013), we can expect wMelCS and wMelPop infection to affect female JH metabolism and reproduction. In insects, JH plays a ‘status quo’ role in larvae, providing normal growth and development and preventing premature metamorphosis, and a gonadotropic role, regulating female fertility (together with 20-hydroxyecdysone, 20E) in adults (Goodman and Granger, 2005; Riddiford, 2012; Jindra et al., 2013; Dubrovsky and Bernardo, 2014). As gonadotropins, JH and 20E induce ovarian development, initiate and maintain production of yolk proteins in the fat body and in the ovary follicular cells, and control vitellogenin uptake by oocytes (Raikhel et al., 2004). Aside from participating in the control of development and reproduction, JH and 20E are involved in the endocrine stress response (Gruntenko, Rauschenbach, 2008). The increase in the JH level under unfavourable conditions leads to the accumulation of mature eggs and ovipositional delay until the unfavourable conditions have improved (Gruntenko, Rauschenbach, 2008). In Drosophila, the JH titre negatively corresponds to the level of its degradation, and thus the latter can be regarded as an indicator of the hormone level (Gruntenko and Rauschenbach, 2008). Wolbachia is shown to establish itself in JH- and 20E-producing tissues of the host, including the fat body and the ovarian follicular cells, as demonstrated in many insect species, suggesting a specific role for the symbiont in host oogenesis, embryogenesis and moulting (Negri, 2012).
In this study, we investigated a difference in the effects of various types of Wolbachia infection on JH metabolism and reproduction of D. melanogaster. Our results suggest that at least one of the mechanisms promoting the effect of Wolbachia on female fecundity is mediated by JH.
MATERIALS AND METHODS
Twenty generations prior to the start of the experiments, the lineage Bi90 was treated with tetracycline for three generations to make Wolbachia-free lineage Bi90T. The experiments were performed on five D. melanogaster conplastic lineages (Bi90Mel, Bi90Mel2, Bi90Mel4, Bi90CS and Bi90Pop) carrying the nuclear background of inbred Bi90 lineage and cytoplasmic backgrounds with different types of Wolbachia infection (wMel, wMel2, wMel4, wMelCS and wMelPop), produced as described in Gruntenko et al. (2017) by 20 backcrosses of Bi90T males with the appropriate source of Wolbachia. Wolbachia infection status was regularly verified using PCR with primers specific to the Wolbachia 81F/691R set for the wsp gene (Braig et al., 1998) and 99F/994R for the 16SrRNA gene (O'Neill et al., 1992). The Wolbachia genotypes were identified according to Riegler et al. (2005) and Ilinsky (2013). The sixth, non-infected Bi90T lineage was used as a control.
The cultures were maintained on standard Drosophila medium (agar-agar, 7 g l−1; corn grits, 50 g l−1; dry yeast, 18 g l−1; sugar, 40 g l−1) at 25°C under a 12 h:12 h light:dark cycle, and the adults were synchronised at eclosion (flies were collected every 3–4 h).
Syto-11 staining of DNA was performed as described in Casper-Lindley et al. (2011) to visualise Wolbachia and host nuclei in the ovarian tissues. Female flies were examined in their reproductive peak at the age of 6 days after eclosion. Ovarian tissues were stained with Syto-11 to label Wolbachia DNA, and then the bacteria nucleoids were imaged by confocal microscopy in stage 10 oocytes. The ovaries of five mated females from each Drosophila lineage were dissected in a Petri dish in ice-cold PBS, and transferred to a large coverslip in a drop of PBS. Individual ovarioles and individual egg chambers were separated, and mature eggs were removed from the coverslip. PBS was removed with a pipette and a drop of Syto-11 (1:100; Molecular Probes, Invitrogen) was applied. Slides were placed in a dark, moist chamber and incubated at room temperature for 20–25 min. Specimens were then covered with a smaller coverslip, which was placed upon strips of tape glued to both sides of the lower coverslip to prevent the crushing of tissues. Immediately after they were analysed by laser scanning microscope (LSM 780 NLO) based on the inverted microscope AxioObserver Z1 (Carl Zeiss, Oberkochen, Germany) at the Institute of Cytology and Genetics Microscopy Center (Novosibirsk, Russia) for no more than 20–25 min.
ImageJ 2.0 Fiji software (National Institutes of Health) was used to quantify the oocyte Wolbachia titre. Before quantification, stacks of confocal images were examined to identify the deepest possible focal plane where Wolbachia were clearly visible across all samples of the replicate (Serbus et al., 2015), and images were manually processed to remove extraneous signal outside the oocyte. For fluorescence quantification, the selection tool was used to isolate the oocyte and then the brightness/contrast tool was used to set the threshold on the image to remove background noise. A mean intensity of fluorescent signals per pixel was measured for each oocyte studied. Three to eight experimental replicates were performed for all Drosophila lineages examined.
The relative amount of Wolbachia genomic DNA
DNA was extracted from the ovaries of ten 6-day-old mated females of each Drosophila lineage for each biological replicate using the CTAB DNA Extraction Protocol (Huang et al., 2000) with modification. Briefly, females were dissected in physiological saline and 10 ovaries per sample were ground in liquid nitrogen. These pools were incubated at 56°C for 1 h with 30 μl of lysis buffer (BioSilica, Russia) with 2 μl of proteinase K. Then, 450 μl of CTAB extraction buffer (20% CTAB, 100 mmol l−1 Tris–HCl pH 8.0, 20 mmol l−1 EDTA pH 8.0, 1 mol l−1 NaCl and 1.5% β-mercaptoethanol) and 450 μl of chloroform were added, shaken and centrifuged for 5 min at 2300 g. The aqueous phase was transferred to a new tube and an equal volume of isopropanol was added to precipitate DNA. The samples were then incubated for 10 min at 4°C and centrifuged for 10 min at 5900 g. Isopropanol was removed and DNA was washed with 70% ethanol, centrifuged for 5 min at 5900 g, dried, dissolved in TE buffer and stored at −20°C. Quantitative real-time PCR was performed using a CFX96 Real-Time PCR system (BioRad Laboratories) with a SYBR Green I R-402 kit (Syntol, Russia) as per the manufacturer's instructions. Each reaction was performed in triplicate with five biological replicates. Specific primers for measuring Wolbachia titer in D. melanogaster by qPCR technique were designed previously: Wolb2-F 5′-TCACAGACCTGTATTTGGTTACA-3′ and Wolb2-R 5′-ACTAAGCCCAACAGTGAACATA-3′ (I. Mazunin, personal communication). The Drosophila Rpl32 gene was used to normalise the quantitative PCR data (Rpl32-F1 5′-CAGCATACAGGCCCAAGATC-3′, Rpl32-R 5′-CGATGTTGGGCATCAGATACTG-3′) as described in Pérez-Moreno et al. (2014). The following thermal cycling protocol was applied: 3 min at 95°C, and 45 cycles of 15 s at 95°C, 15 s at 56°C, 15 s at 62°C and 5 s at 78°C. Melting curves were examined to confirm the specificity of amplified products. Cycle threshold (Ct) values were obtained using Bio-Rad CFX Manager software with default threshold settings.
For the egg production analysis, newly eclosed full-sib flies (three females and three males) carrying the same infections were placed into a vial with filter paper soaked in nutritional medium as described in Rauschenbach et al. (2014). The sample size was 20 to 40 vials (60–120 females) for each lineage under study. The nutritional medium contained 0.5% sucrose and 0.1% yeast. The flies were transferred to vials with fresh medium daily. Fecundity was calculated as the number of eggs per female per 24 h.
JH and precocene treatment
To study the effect of JH or precocene on egg production, 1-day-old females were treated with acetone, JH or precocene as described below, and their fecundity was evaluated. For the precocene (JH inhibitor) treatment, females were collected soon after eclosion and placed in vials with standard medium (three females and three males in a vial) for 1 or 5 days. After that, flies were anaesthetised with ether, and 0.5 µl of JH-III (Sigma-Aldrich) or 0.2 µl of Precocene I (Sigma-Aldrich), diluted in acetone to a concentration of 2 or 1 mg ml−1, correspondingly, was applied to the abdomen of each female. Control females were treated with an equal amount of drug vehicle (acetone) (the treatment with pure acetone does not affect fecundity; Rauschenbach et al., 2017). The sample size was 10 vials (30 females) for each group.
JH-hydrolysing activity assay
For the JH hydrolysis measurement each fly was homogenised in ice-cold 0.1 mol l−1 sodium phosphate buffer, pH 7.4, containing 0.5 mmol l−1 phenylthiourea. The homogenates were centrifuged for 5 min at 13,300×g, and samples of the supernatant were taken for the assay. A mixture consisting of 0.1 μg of unlabelled JH-III (65%, Sigma-Aldrich) and 0.1 μg [3H]JH-III (2.2 Ci mmol−1, radiochemical purity 95%, prepared as described in Romanova et al., 2017) was used as a substrate. The reaction was carried out for 30 min and then was stopped with the addition of 5% ammonia, 50% methanol (v/v). Unhydrolysed JH was extracted with heptane. The tubes were shaken vigorously and centrifuged at 13,300 g for 10 min. Samples of both the organic and aqueous phases were placed in vials containing dioxane scintillation fluid and counted. All components of dioxane scintillation fluid were purchased from Sigma-Aldrich. The fresh fluid was prepared prior to conducting the actual experiments. Control experiments showed a linear substrate–reaction product relationship; the activity measured is proportional to the amount of the supernatant (i.e. enzyme concentration) (Gruntenko et al., 2000). Before measurement, half of the flies in each group under study were exposed to heat stress by transferring vials containing them from a 25°C incubator to a 38°C incubator for 2 h. The sample size was 20–40 flies per each control or experimental group. JH stress reactivity was calculated as the percent change in JH-hydrolyzing activity following heat stress relative to the value of the corresponding parameter obtained under normal conditions (each value obtained at 38°С was compared with the average value obtained at 25°С).
All data are presented as means±s.e.m. The false discovery rate corrections for multiple comparisons were made when appropriate. The Kolmogorov–Smirnov test was used to determine whether it was appropriate to use parametric tests. Data were subjected to arcsine or power transformation prior to analysis when appropriate. Data on fecundity (number of eggs per day per female) were analysed via two-way mixed-design ANOVA (with day after eclosion as the within-subjects factor; and infection, precocene treatment or JH treatment as the between-subjects factors). Data on the JH degradation level and the intensity of its responses to stress (JH stress reactivity) were analysed via two-way or one-way ANOVA, respectively (with infection and heat stress, or infection as the between-subjects factors, respectively). Data on the relative amount of Wolbachia were analysed via one-way ANOVA (with infection as the between-subjects factor). The comparison of the group means was performed with the Benjamini–Hochberg stepwise post hoc test. The results were considered significant at P<0.05.
The conplastic lineages differed by Wolbachia load in ovarian tissues
To find out whether various types of Wolbachia promote different effects on female host reproduction, we first tested the bacterial load of different genotypes in D. melanogaster ovaries and quantified this using ImageJ (Serbus et al., 2015). This analysis revealed that oocytes of Bi90CS females carried more bacteria nucleoids than oocytes of Bi90Mel females, but less than oocytes of Bi90Pop females (infection: F3,14=36.81, P≪0.0001; Fig. 1).
The data on the relative amount of Wolbachia genomic DNA suggested that bacterial load in the ovaries of 6-day-old mated Bi90Pop females was twice as high as that of Bi90Mel and Bi90CS females (infection: F3,16=18.97, P<0.0001; Fig. 2). No significant difference in bacterial load was found between Bi90CS and Bi90Mel females.
The effects of various Wolbachia genotypes on the host fecundity
Because we found variability in the number of Wolbachia nucleoids in the ovaries of D. melanogaster lineages infected with various Wolbachia genotypes, we next tested the effects of different types of Wolbachia infection on egg production as a proxy for fecundity in comparison with uninfected controls (Fig. 3). No difference in fecundity was found between uninfected Bi90T flies and Bi90Mel, Bi90Mel2 and Bi90Mel4 flies (Fig. 3A). The egg production of Bi90CS flies was significantly lower in the beginning of the oviposition (days 3–5 after eclosion) and significantly higher from the sixth day after eclosion in comparison with that of uninfected Bi90T flies (age: F9,522=205.20, P≪0.0001; infection×age: F9,522=9.63, P≪0.0001; Fig. 3B). On the contrary, the Bi90Pop flies demonstrated an initial increase in egg production compared with uninfected Bi90T controls during the 3–5 days after the start of egg laying and a later subsequent decrease compared with Bi90T flies from the eighth day after eclosion (age: F9,612=244.37, P≪0.0001; infection×age: F9,612=10.38, P≪0.0001; Fig. 3B).
As Wolbachia is known to affect male fertility (Serbus et al., 2008; Werren et al., 2008), we tested whether the difference in the fecundity of Drosophila Bi90Mel and Bi90Pop lineages is due to the changes in male or female reproductive function (Fig. 4). Comparing the fecundity level of the uninfected Bi90T lineage, the Bi90Mel lineage and hybrids from both directions of crossing between these lineages, we found that the fecundity of females in the ♀Bi90Mel×♂Bi90T cross did not differ from that of the Bi90Mel and Bi90T lineages, whereas the ♀Bi90T×♂Bi90Mel cross showed a strong decrease in the number of eggs laid in days 3–5 after eclosion (age: F7,336=210.09, P≪0.0001; infection×age: F7,336=16.21, P≪0.0001; Fig. 4A). The comparison of fecundity levels of hybrids from both directions of crossing between the Bi90Pop and Bi90T lineages showed a similar pattern: the fecundity of females in the ♀Bi90T×♂ Bi90Pop cross was significantly lower than the fecundity of the uninfected Bi90T lineage in days 3–6 after eclosion (Fig. 4B). Two-way mixed design ANOVA (day after eclosion as the within-subjects factor; infection as the between-subjects factor) revealed significant effects for infection (F1,38=85.47, P≪0.0001) and age (F7,266=159.50, P≪0.0001). A significant interaction of these factors (F7,266=18.50, P≪0.0001) was also found. However, the most interesting finding of this experiment is that the fecundity of females in the ♀Bi90Pop×♂Bi90T cross did not differ from that of the Bi90Pop lineage, but it differed from the fecundity of the Bi90T lineage in the same way as the Bi90Pop lineage differed from the Bi90T lineage. The fecundity in the ♀Bi90Pop×♂Bi90T cross compared with the uninfected Bi90T control increased in the beginning of oviposition (days 3 and 4 after eclosion) and decreased from the seventh day after eclosion (age: F7,266=90.92, P≪0.0001; infection×age: F7,266=13.64, P≪0.0001; Fig. 4B).
Effects of various Wolbachia genotypes on JH degradation levels in D. melanogaster females under normal and heat stress conditions
To determine whether Wolbachia infection status and/or type affect female fecundity via JH, which is required for oogenesis (Goodman and Granger, 2005; Riddiford, 2012), we studied the levels of JH degradation in 1- and 6-day-old females of the Bi90Mel, Bi90CS and Bi90Pop lineages in comparison with Bi90T under normal conditions and under heat stress (Fig. 5A,C). We also tested whether such stress (38°C, 2 h) eliminates Wolbachia infection in flies, and found out that it does not. Under normal conditions, Bi90Mel flies did not differ from the uninfected Bi90T in the level of JH degradation; in Bi90CS and Bi90Pop flies, the JH degradation level, as well as the level of egg production, changed in opposite directions compared with the Bi90T control. At the very beginning of oviposition, the JH degradation level was increased in 1-day-old Bi90CS females and decreased in Bi90Pop females compared with Bi90T (infection: F3,191=11.11, P<0.0001; Fig. 5A). At the age of 6 days, Bi90CS females demonstrated a decreased JH degradation level and Bi90Pop females an increased one (infection: F3,132=24.05, P≪0.0001; Fig. 5C) under normal conditions.
Heat stress affected the levels of JH degradation in both 1-day-old (heat stress: F1,191=171.71, P≪0.0001; Fig. 5A) and 6-day-old (heat stress: F1,132=411.20, P≪0.0001; Fig. 5C) females of all lineages such that JH degradation was lower under heat stress than in normal conditions. The ability of the JH degradation activity of infected flies to respond to heat stress differed between lineages either on day 1 (infection×heat stress: F3,191=6.47, P=0.0004; Fig. 5A) or on day 6 (infection×heat stress: F3,132=10.90, P<0.0001; Fig. 5C). In particular, the stress reactivity of JH metabolic system was higher in 1-day-old Bi90CS females and lower in 1-day-old Bi90Pop females in comparison with Bi90Mel and uninfected Bi90T females of the same age (infection: F3,107=11.90, P<0.0001; Fig. 5B). In contrast, the stress reactivity of the JH metabolic system was lower in 6-day-old Bi90CS females and higher in 6-day-old Bi90Pop females in comparison with Bi90Mel and Bi90T females of the same age (infection: F3,71=16.90, P≪0.0001; Fig. 5D).
Bi90Mel females did not differ from uninfected Bi90T females in the levels of JH degradation and stress reactivity of the JH metabolic system in both ages studied (Fig. 5). Bi90CS females significantly differed from Bi90Pop females in the levels of JH degradation and stress reactivity of the JH metabolic system either on day 1 or on day 6 after eclosion (Fig. 5).
The effect of JH on the influence of Wolbachia on female fecundity
To determine whether the changes in the reproductive function of D. melanogaster Bi90Pop and Bi90CS females are associated with the altered JH level, we investigated their egg production following an artificial increase or decrease of JH titre. The decreased JH degradation level in young (just starting to lay eggs), 1-day-old Bi90Pop females (Fig. 5A) and mature (having a reproduction peak), 6-day-old Bi90CS females (Fig. 5C) compared with Bi90T females of the same age corresponds to an increased JH level in these flies (Gruntenko and Rauschenbach, 2008). Therefore, we studied the impact of the JH inhibitor precocene (Wilson et al., 1983; Argue et al., 2013) on young Bi90Pop and mature Bi90CS females to test the role of JH in the fecundity differences of these lineages. Young Bi90CS females and mature Bi90Pop females, characterized by an increased JH degradation (Fig. 5A,С) and presumably reduced JH level, were treated with a JH-III on the same days. The data are provided in Fig. 6.
The treatment with precocene decreased egg production in both young Bi90Pop females (precocene: F2,27=24.03, P<0.0001; age: F3,81=159.44, P≪0.0001; precocene×age: F6,81=14.16, P≪0.0001; Fig. 6A) and mature Bi90CS females (precocene: F2,27=29.60, P<0.0001; age: F4,108=52.14, P≪0.0001; precocene×age: F8,108=8.42, P≪0.0001; Fig. 6D) compared with the acetone-treated groups, and abolished the differences between the Wolbachia-infected and uninfected lineages.
The increase of the JH level via its application raised the egg production of both young Bi90CS females (JH: F2,27=17.16, P<0.0001; age: F3,81=165.39, P≪0.0001; JH×age: F6,81=4.98, P=0.0003; Fig. 6B) and mature Bi90Pop females (JH: F2,27=32.30, P≪0.0001; age: F4,108=24.87, P≪0.0001; JH×age: F8,108=2.93, P=0.0054; Fig. 6C) compared with the acetone-treated groups, and eliminated the differences between the Wolbachia-infected and uninfected lineages.
In our previous work (Gruntenko et al., 2017) and the present study, we try to reveal the mechanism by which the genetic background of the endosymbiont can affect host fitness by investigating how different types of Wolbachia modulate survival and fecundity of D. melanogaster lineages with the same genetic background. Here, we report that the wMel and wMelCS genotypes of Wolbachia cause different effects on gonadotropic (JH) metabolism and egg production, differently contributing to the fitness of the host. The effect of the wMelPop strain differs from the effects of both the wMel and wMelCS genotypes. Our data suggest that the fecundity level of Drosophila lineages carrying various types of infection depends on how Wolbachia affects the reproductive biology of females. It was shown earlier that the effect of Wolbachia on host fitness and stress resistance is dependent on the genomic background of the Drosophila lineage (Fry et al., 2004; Capobianco et al., 2018). Based on these data, Capobianco et al. (2018) concluded that the effect of Wolbachia on fitness is unpredictable across the individual genetic backgrounds of host animals. We can make an addition to this statement: the effect of Wolbachia on the host fitness depends not only on the genetic background of the host, but also on that of the symbiont.
We have found that Wolbachia of different genotypes are present in Drosophila oocytes in different titres (Fig. 1). The higher load has been shown earlier for the pathogenic wMelPop strain, which overproliferates and shortens host lifespan (Min and Benzer, 1997; McGraw et al., 2002; Chrostek et al., 2013). Chrostek et al. (2013) found a higher relative amount of wMelCS Wolbachia compared with wMel in D. melanogaster males from 2 weeks after eclosion. However, they did not detect any difference in bacterial load between wMelCS- and wMel-infected males at the age of 3–6 days, which is in agreement with our data on the relative amount of Wolbachia genomic DNA in 6-day-old females (Chrostek et al., 2013). Perhaps the variation in bacterial load among different types of Wolbachia infection partly explains the diversity in the effects they caused in the host. Such a correlation between Wolbachia titre and cytoplasmic incompatibility was discovered earlier in the parasitoid wasp Asobara japonica and the mosquito Aedes albopictusi (Kraaijeveld et al., 2011; Calvitti et al., 2015).
Our results show that Wolbachia of the wMel group (wMel, wMel2 and wMel4) do not affect the fecundity of infected Drosophila lineages, whereas bacteria of the wMelCS group decrease the egg production in young females, just starting oviposition, and increase the egg production in mature females at their reproductive peak (Fig. 3). The pathogenic wMelPop strain causes the opposite changes in fecundity compared with wMelCS Wolbachia, although they are indistinguishable in terms of genetic markers (Riegler et al., 2012), except the Octamom copy number (Chrostek and Teixeira, 2015). These findings correspond well with our previous data on the effects of these Wolbachia variants on host survival under stress (Gruntenko et al., 2017); the survival of D. melanogaster lineages infected with the wMel, wMel2 and wMel4 genotypes do not differ from the survival of uninfected flies, whereas wMelCS increases the survival and wMelPop decreases it. One would expect that the positive fitness effects caused by Wolbachia infection would help to promote its spread throughout a host population; thus, there is still no explanation to the fact that the Wolbachia of the wMelCS group, having higher fecundity and stress resistance, is the rare variant in natural D. melanogaster populations compared with the bacteria of the wMel group (Riegler et al., 2005; Ilinsky and Zakharov, 2007; Nunes et al., 2008).
Previously, Fry et al. (2004) demonstrated that crosses between infected males and uninfected females resulted in a decrease in egg production, and suggested that the phenomenon could promote the spread of infection just like cytoplasmic incompatibility. Our findings support these results, showing that uninfected females crossed with males infected with wMel or wMelPop produced many fewer eggs than infected or uninfected females crossed with uninfected males (Fig. 4). We also found that the fecundity pattern of wMelPop-infected females does not depend on the infection status of males crossed with them. Therefore, Wolbachia not only modify host sperm and affect fertility of their male hosts (Serbus et al., 2008; Liu et al., 2014), but affect female reproductive biology as well. This could provide some explanation as to the maintenance of Wolbachia in D. melanogaster, which does not seem to be associated with cytoplasmic incompatibility because it is weak or absent in this species (Fry et al., 2004).
Our findings support and extend previous studies showing positive correlations between an increase/decrease in JH degradation (accompanied by a decrease/increase in JH level) and a decrease/increase in the fecundity of D. melanogaster females (Gruntenko and Rauschenbach, 2008; Gruntenko et al., 2010; Rauschenbach et al., 2014). Thus, in young wMelCS-infected females, the fecundity decrease (compared with uninfected females) is accompanied by an increase in JH degradation activity; and vice versa, in mature wMelCS-infected females, the fecundity increase is accompanied by a decrease in JH degradation (Figs 2 and 4). The increase/decrease in the fecundity of young/mature wMelPop-infected females also correlates with decrease/increase in the JH degradation in these females (Figs 2 and 4). At the same time, females infected with the wMel genotype of Wolbachia do not differ from uninfected females in terms of both JH metabolism and fecundity level (Figs 2 and 4). Importantly, our data indicate that the effect of Wolbachia infection on the stress response of the JH metabolic system also depends on the type of infection and age of the flies. The results of the influence of wMelCS and wMelPop on JH metabolism correspond well with the data by Zheng et al. (2011) and Liu et al. (2014), who demonstrated an interaction between Wolbachia and the JH signalling pathway in the testes of male D. melanogaster. Zheng et al. (2011) found a 10-fold increase in the transcription level of the gene for JH-induced protein 26 (JhI-26) caused by Wolbachia infection. JhI-26 expression is shown to be triggered rapidly and specifically by JH-ΙΙΙ (Dubrovsky et al., 2000). Liu et al. (2014) showed that Wolbachia significantly increased the expressions of genes that play key roles in the JH signalling pathway: Jhamt (encoding JH acid methyltransferase, a key regulatory enzyme of JH biosynthesis) and Met (encoding JH receptor). These data support our conclusion concerning the influence of Wolbachia on JH signalling in Drosophila.
JH/precocene treatment of young and mature females infected with wMelCS and wMelPop Wolbachia produced a pronounced fecundity response in both treated lineages (Fig. 6). This treatment was aimed at a shift in the JH level in young or mature wMelCS- and wMelPop-infected females to the level typical for uninfected controls. Egg production curves from infected females with a corrected JH level coincide with the curves from uninfected flies both in the beginning and at the peak of oviposition. It is worth noting that the influence of JH on the egg production of wMelCS- and wMelPop-infected females is strain-specific in contrast to the JH effect on stress resistance: the same JH treatment induces a decrease in the survival under heat stress in Drosophila females infected with wMelCS, wMel and wMelPop as well as in uninfected controls (Rauschenbach et al., 2018). These results indicate that JH plays a mediating role in the effect of Wolbachia on host fecundity.
In conclusion, we found that the influence of Wolbachia on female reproductive biology in D. melanogaster depends on the genetic background of the symbiont, and JH mediates the Wolbachia modulation of female fecundity.
We thank Dr Luis Teixeira for sharing the valuable iso wMelPop strain, and Darya Pirozhkova for assistance with the quantification of the relative amount of Wolbachia genomic DNA.
Conceptualization: N.E.G., Y.Y.I., I.Y.R.; Methodology: Y.Y.I.; Validation: P.N.M.; Investigation: E.K.K., N.V.A., O.V.A., E.V.B., R.A.B.; Data curation: P.N.M., I.Y.R.; Writing - original draft: N.E.G.; Writing - review & editing: N.E.G., Y.Y.I., P.N.M., I.Y.R.; Supervision: N.E.G., I.Y.R.; Project administration: N.E.G.; Funding acquisition: I.Y.R.
This work was supported by the Russian Foundation for Basic Research grant 16-04-00060 and by The Ministry of Science and Higher Education of the Russian Federation (Budgeted Project 0324-2019-0041).
The authors declare no competing or financial interests.