Bacterial supergroup‐specific “cost” of Wolbachia infections in Nasonia vitripennis

Abstract The maternally inherited endosymbiont, Wolbachia, is known to alter the reproductive biology of its arthropod hosts for its own benefit and can induce both positive and negative fitness effects in many hosts. Here, we describe the effects of the maintenance of two distinct Wolbachia infections, one each from supergroups A and B, on the parasitoid host Nasonia vitripennis. We compare the effect of Wolbachia infections on various traits between the uninfected, single A‐infected, single B‐infected, and double‐infected lines with their cured versions. Contrary to some previous reports, our results suggest that there is a significant cost associated with the maintenance of Wolbachia infections where traits such as family size, fecundity, longevity, and rates of male copulation are compromised in Wolbachia‐infected lines. The double Wolbachia infection has the most detrimental impact on the host as compared to single infections. Moreover, there is a supergroup‐specific negative impact on these wasps as the supergroup B infection elicits the most pronounced negative effects. These negative effects can be attributed to a higher Wolbachia titer seen in the double and the single supergroup B infection lines when compared to supergroup A. Our findings raise important questions on the mechanism of survival and maintenance of these reproductive parasites in arthropod hosts.


| INTRODUC TI ON
Wolbachia are maternally inherited, obligatory intracellular, endosymbionts of the order Rickettsiales (Hertig & Wolbach, 1924), which are widely found in arthropods and filarial nematodes (Bandi et al., 1998;Rousset et al., 1992;Weinert et al., 2015). To enhance their own transmission, these bacteria often alter host reproductive biology with mechanisms like male-killing, feminization, parthenogenesis, and cytoplasmic incompatibility (CI) (Werren et al., 2008). While CI leads to an increase in the number of infected individuals in the population, male-killing, and feminization shifts the offspring sex ratio towards females, which is the transmitting sex for Wolbachia. Thus, Wolbachia increases the fitness of the infected hosts, over the uninfected ones, as it increases its own rate of transmission. The vast majority of Wolbachia-host association studies reveal many negative effects on the hosts. In addition to reproductive traits, many other life-history traits like longevity and developmental time are also known to be compromised. A review of such negative effects of Wolbachia on hosts where CI is prevalent is presented in Table 1. In Trichogramma kaykai and T. deion, the infected (thelytokous) line shows reduced fecundity and adult emergence rates than the antibiotically cured (arrhenotokous) lines (Hohmann et al., 2001;Tagami et al., 2001). Leptopilina heterotoma, a Drosophila parasitoid, has adult survival rates, fecundity, and locomotor performance, of both sexes, severely compromised in Wolbachia-infected lines (Fleury et al., 2000). Larval mortality has been observed in both sexes of insecticide-resistant Wolbachia-infected lines of Culex pipiens (Duron et al., 2006). Wolbachia infections can also result in a range of behavioral changes and altered phenotypes in Aedes aegypti (Turley et al., 2009). While these cases highlight a parasitic effect of Wolbachia, there are several examples where no such effect is discernible (Hoffmann et al., 1996). Moreover, there are also examples where Wolbachia has now become a mutualist and offers specific and quantifiable benefits to its host. One such example of an obligate mutualism with Wolbachia has been reported in the common bedbug Cimex lectularius where Wolbachia, found to be localized in bacteriomes, provides essential B vitamins needed for growth and fertility (Hosokawa et al., 2010). Such examples of arthropod-Wolbachia mutualism have now been reported from various arthropod taxa (Miller et al., 2010;Pike & Kingcombe, 2009). This shift from parasitic to mutualistic effect can also happen in facultative associations as seen in Drosophila simulans, where within a span of just two decades, Wolbachia has evolved from a parasite to a mutualist (Weeks et al., 2007).
The negative effects of Wolbachia on their hosts are not unexpected. The presence of bacteria within a host entails sharing of nutritional and other physiological resources (Kobayashi & Crouch, 2009;Whittle et al., 2021), especially with Wolbachia, as they are obligate endosymbionts and cannot survive without cellular resources derived from their hosts (Foster et al., 2005;Slatko et al., 2010). Accordingly, Wolbachia is known to compete with the host for key resources like cholesterol and amino acids in A. aegypti (Caragata et al., 2014). The precise molecular mechanisms of many of these negative effects have not been ascertained and are generally ascribed to partitioning-off of host nutrients for its benefit, but what is clear is that Wolbachia infections can impose severe nutritional demands on their hosts (Ponton et al., 2014).
However, it is also known that Wolbachia can elicit antipathogenic responses from their hosts where the host resistance or tolerance to the infection increases (Zug & Hammerstein, 2015). For example, Wolbachia induces host methyltransferase gene Mt2 towards antiviral resistance against Sindbis virus in D. melanogaster (Bhattacharya et al., 2017). Wolbachia can utilize the immune deficiency (IMD) and Toll pathways (Pan et al., 2018) and increase reactive oxygen species (ROS) levels in Wolbachia-transfected A. aegypti mosquitoes, inhibiting the proliferation of the dengue virus (Pan et al., 2012). Such immune responses require additional allocation of resources, which can further affect other physiological traits of the host. This concept of a "cost of immunity" is wellestablished and suggests a trade-off between immunity and other life-history traits (Zuk & Stoehr, 2002). For example, elevated ROS levels negatively affect many host traits like longevity and fecundity (Dowling & Simmons, 2009;Monaghan et al., 2009;Moné et al., 2014;Selman et al., 2012). Thus, there is sufficient evidence to conclude that Wolbachia can have substantial negative effects on the overall fitness of its host.
One of the arthropod hosts infected by Wolbachia is the parasitoid wasp Nasonia vitripennis. N. vitripennis, being cosmopolitan, has been used to study Wolbachia distribution, acquisition, spread, and Wolbachia-induced reproductive manipulations (Landmann, 2019;Werren et al., 2008). However, the effect of the endosymbiont on the life-history traits of this wasp remains poorly understood with conflicting reports. N. vitripennis harbor two Wolbachia supergroup infections, one each from supergroup A and supergroup B (Perrot-Minnot et al., 1996), and the presence of these two infections has been found in all lines of N. vitripennis from continental North America to Europe (Raychoudhury et al., 2010), indicating that it has reached fixation across the distribution of its host. The two Wolbachia in N.
vitripennis together cause complete CI, but single infections of supergroup A Wolbachia cause incomplete CI while supergroup B infections still show complete CI (Perrot-Minnot et al., 1996). In some N. vitripennis lines, Wolbachia has been reported to cause enhanced fecundity (Stolk & Stouthamer, 1996), but a similar effect has not been observed in some other lines (Bordenstein & Werren, 2000).
In this study, we investigate, what, if any, are the negative effects of CI-inducing Wolbachia infections in N. vitripennis. We investigate the effects of Wolbachia infections in a recently acquired line of N.
vitripennis from the field. This line, like other N. vitripennis lines, has two Wolbachia infections, one each from the supergroup A and B.
Sequencing of the five alleles from the well-established multi-locus strain typing (MLST) system (Baldo et al., 2006)

| Nasonia vitripennis lines used, their Wolbachia infections, and nomenclature
The N. vitripennis NV-PU-14 line was obtained from Mohali, Punjab, India, in 2015. NV-PU-14 was cured of Wolbachia by feeding the females with 1 mg/ml tetracycline in 10 mg/ml sucrose solution for at least two generations (Breeuwer & Werren, 1990). The curing was confirmed by PCR using supergroup-specific ftsZ primers (Baldo et al., 2006), and CI crosses between the infected and uninfected lines. NV-PU-14 also served as the source strain for separating the two Wolbachia infections into single A and single B infected wasp lines.
To separate the Wolbachia supergroups, we utilized relaxed se- lines were in culture for 3 years, many of the infected lines were cured again to obtain "recently cured" lines to minimize the effects of any host divergence that might have accumulated within them.
Another N. vitripennis line, NV-KA, obtained from Bengaluru, Karnataka, India, in 2016, was similarly named wAwB(KA). The MLST sequences of the two Wolbachia strains (one each from supergroups A and B), even in wAwB(KA), were found to be identical to wAwB(PU), and were also identical to all other N. vitripennis studied across the world (Prazapati, personal communication). wAwB (KA) was also cured of Wolbachia to obtain 0(wAwB KA).
All these wasp lines were raised on Sarcophaga dux fly pupae with a generation time of 14-15 days at 25°C, 60% humidity, and a continuous daylight cycle.

| Sequential mating and sperm depletion of the males
To test the effect of Wolbachia on male reproductive traits like mating ability, individual males were assayed for the number of copulations they can perform and sperm depletion. As N. vitripennis is haplodiploid, every successful mating will result in both female and male progenies while an unsuccessful one will result in all-male progenies. The males used were obtained from virgin females hosted with one fly pupa for 24 h and were not given any external sources of nutrition (usually a mixture of sucrose in water) before the experiment.
Each male was then mated sequentially with virgin females from the same line. At the first sign of a male not completing the entire mating behavior (Jachmann & Assem, 1996), it was given a rest for half an hour and was subjected to mating again until it stopped mating altogether. The mated females were hosted after a day with one fly pupa for 24 h. The females were then removed, and the offsprings were allowed to emerge and then counted. The average number of copulations and the number of copulations before sperm depletion, were compared using the Kruskal-Wallis test with a significance level of .05. Mann-Whitney U test, with a significance level of .05, was used for comparisons between two groups.

| Host longevity, family size, and fecundity
To test whether the presence of Wolbachia has any influence on longevity, emerging wasps of both sexes were kept individually in ria vials at 25°C, without any additional nutrition. Survival following emergence was measured by counting the number of dead individuals every 6 h. The Kaplan-Meier analysis, followed by log rank statistics, was used to identify differences between the strains with a significance level of .05.
To test for the effect of Wolbachia infections on the adult family size of virgin and mated females, each female was sorted at the pupal stage and separated into individual ria vials. To enumerate the brood size of mated females, some of these virgins were offered single males from the same line and observed till mating was successful.
All the females were then hosted individually with one fly pupa for 24 h. These were kept at 25°C for the offspring to emerge, which were later counted for family size, by randomizing the ria vials in a double-blind assay. The differences between groups were compared using the Kruskal-Wallis test with a significance level of .05. The Mann-Whitney U test, with a significance level of .05, was used to compare two groups.
To investigate whether Wolbachia affects the female fecundity, emerged females were hosted with one host for 24 h. The host pupa was placed in a foam plug, so that only the head portion of the pupa was exposed and available for the females to lay eggs. The females were removed after 24 h, and the eggs laid were counted under a stereomicroscope (Leica M205 C). The differences in fecundity were compared between groups using the Kruskal-Wallis test with a significance level of .05. The fecundity difference between two groups was compared using the Mann-Whitney U test with a significance level of .05.

| Estimation of the relative density of Wolbachia infections across different developmental stages of N. vitripennis
To collect the different developmental stages, females were hosted

| The presence of Wolbachia reduces the life span of both males and females
Wolbachia can compete with the host for available nutrition, which can increase nutritional stress, resulting in a shortened life span for many hosts (Caragata et al., 2014;McMeniman et al., 2009).
Therefore, we first investigated the effect of Wolbachia infections on the survival of both male and female wasps. As Figure 1

| The presence of Wolbachia reduces the number of copulations a male can perform
Wolbachia is known to be associated with a reduction in the number of mating a male can perform in Ephestia kuehniella (Sumida et al., 2017). is evident is that the presence of Wolbachia is also associated with a reduction in the capability of a male to mate. Furthermore, by curing the infected lines again, we show that this decrease is not due to the host genotype but is an effect of the presence of Wolbachia in these lines.

| Wolbachia-infected males deplete their sperm reserves faster than the uninfected ones
Nasonia vitripennis males are prospermatogenic (Boivin et al., 2005), where each male emerges with their full complement of mature sperm and has not been reported to produce any more during the rest of their life span (Chirault et al., 2016). Thus, if a single male is mated sequentially with as many females as it can mate with, it should eventually run out of this full complement of sperm and produce all-male broods even after successful copulation. As Figure 2, indicates, each male did run out of sperm at the tail end of this continuous mating and produced only male progenies (shown by black dots). We looked at the number of mating done by these males before sperm depletion to see whether Wolbachia affects the sperm production in the males. As shown in Figure 3 F I G U R E 2 Wolbachia-infected males show a reduction in the number of copulations. Males from different Wolbachia infection status strains were mated sequentially until each of them stopped mating. Some of the matings had "no emergence" of progenies because of poor host quality (shown by white dots). The results show that the presence of Wolbachia is associated with the reduction in the number of copulations a male can perform. The figure also shows whether the progenies of these sequential copulations produce any daughters or not, as a measure of sperm depletion. The details of sperm depletion are shown in Figure 3. Sample sizes for the strains 0(PU), wA(PU), 0(wA PU), wB(PU), 0(wB PU), wAwB(PU), and 0(wAwB PU) were n = 7, n = 7, n = 7, n = 6, n = 5, n = 6, and n = 7, respectively.

| Wolbachia-infected females produce fewer offspring
Wolbachia is known to have a negative impact on the progeny family size of its host (Hoffmann et al., 1990;Hohmann et al., 2001). To test whether a similar effect is seen in N. vitripennis, we enumerated the family sizes for both virgin and mated females for the four different Wolbachia-infected lines and their recently cured counterparts.
As Figure 4

| Wolbachia negatively impacts the fecundity of infected females
To check whether the differences in the family sizes between the different infected lines of N. vitripennis were due to the number of eggs being laid by the females, we looked at the fecundity of both virgin and mated females across these lines. Among the virgin females The results thus suggest a negative effect of Wolbachia on egg production in females. The assay also established that the difference in family sizes can be due to the differences in the fecundity of the females.

| Relative Wolbachia density in single and multiple Wolbachia infections N. vitripennis lines
Wolbachia density has a major role to play in expressing the effects of the infection on host biology (Hoffmann et al., 1996;Min & Benzer, 1997). An increase in cellular Wolbachia density is often associated with a greater expression of their effects (Breeuwer & Werren, 1993). Thus, we estimated Wolbachia titers across the dif-

| DISCUSS ION
The results from this study (summarized in Table 2)  However, the egg to larval to pupal stage mortality could also have an effect on the brood sizes, but these were not assayed.
In most cases, these negative effects disappear with the removal and wB(PU) lines (i.e., a synergistic effect). Since the two supergroup infections are bidirectionally incompatible with each other, it is plausible that they are also competing for the host nutrition, which can further enhance the negative impacts of these infections.

Previous reports have suggested a direct correlation between
Wolbachia density and the level of CI (Breeuwer & Werren, 1993;Dutton & Sinkins, 2004;Ikeda et al., 2003;Noda et al., 2001;Ruang-Areerate & Kittayapong, 2006). Our results also suggest that the cost of Wolbachia maintenance is correlated with the density of the Wolbachia titer and hence shows a milder intensity of the effect of CI. Our results also confirm these previous reports of the positive correlation between Wolbachia abundance and the level of CI induced not only in N. vitripennis  but also in other insect taxa as well Kondo et al., 2002).
wB (  shows incomplete CI ( Figure S1). Thus, higher levels of Wolbachia in wB(PU) than in wA(PU) can also explain the more severe effects in wB(PU) than wA(PU).
The negative fitness effects of CI-inducing Wolbachia, and nutritional competition raises important questions on the maintenance of these endosymbionts over long evolutionary time scales.
Theoretical studies indicate that evolution towards mutualism can aid the long-term persistence of these maternally inherited reproductive parasites (Prout, 1994;Turelli, 1994 (Zug & Hammerstein, 2015). Host suppressor alleles have been identified, which confer resistance against feminizing (Rigaud et al., 1999) and male-killing Wolbachia (Hornett et al., 2006). However, no such host genetic factors have been found for CI-inducing Wolbachia, especially in N. vitripennis. Therefore, a possible explanation for the maintenance of these multiple infections then comes from the high efficiency of transmission of these infections in N. vitripennis, which is nearly 100% (Breeuwer & Werren, 1990). Theoretical studies also suggest that even in the presence of selective pressures, multiple infections are maintained and transmitted owing to the fitness advantages conferred and CI (Vautrin et al., 2008).
Another possibility can be that these Wolbachia infections in N.
vitripennis are relatively recent, the evidence of which comes from the rapid spread of Wolbachia in populations of N. vitripennis across North America and Europe (Raychoudhury et al., 2010). These recent infections, although bearing a cost on the host at present, might eventually lead to the evolution of host resistance against them.
Our results indicate supergroup B to be a "stronger" Wolbachia than supergroup A and any competition for nutritional resources and niche habituation between them should drive out supergroup A Wolbachia. Moreover, wA(PU) has milder effects on females with the reduction in longevity being the only pronounced negative effect.
Therefore, the continuation of this supergroup infection is difficult to explain. One possibility could be the supergroup A infection conferring mutualistic effects on the host. This strain is closely related to other supergroup A Wolbachia strains like wMel in D. melanogaster and wHa, wAu, and wRi in D. simulans (Díaz-Nieto et al., 2021). wMel in D. melanogaster and wHa, wAu, and wRi in D. simulans are known to provide defense against viral infections to their hosts (Bhattacharya et al., 2017;Pimentel et al., 2021;Teixeira et al., 2008). The continued presence of supergroup A Wolbachia in N. vitripennis could be due to such defenses against viral infections, but this hypothesis remains to be tested.
The higher cost of maintenance of supergroup B Wolbachia can be an attribute of the CI phenotype induced by supergroup B Wolbachia. Complete CI (i.e., nearly 100%) are rare events reported mainly for supergroup B Wolbachia in Culex pipiens, Aedes aegypti (Sinkins et al., 2005;Xi et al., 2005), and N. vitripennis ( Figure S1 and Bordenstein et al., 2006). This essentially means that nearly the entire sperm complement of each male has the Wolbachia-induced CI-inducing Wolbachia is known to have negative effects on various physiological traits in the vast majority of its host population (summarized in Table 1). The present study also suggests such effects, or a "cost," associated with the maintenance of Wolbachia infection in N.
vitripennis. This is in contrast to the previous reports suggesting positive fitness effects (Stolk & Stouthamer, 1996) and no fitness effects (Bordenstein & Werren, 2000) of Wolbachia on N. vitripennis. However, the strain used are all from India and the negative effects seen can be unique to these lines. Although the lines used here have the same or very similar Wolbachia as far as sequence uniformity is concerned across the five MLST alleles, other lines from other continents need to be analyzed to confirm whether this effect is ubiquitous in N. vitripennis.

ACK N OWLED G M ENTS
We thank the Indian Institute of Science Education and Research (IISER) Mohali for the funding and graduate fellowship for AT.
Partial funding was obtained from grant no. BT/PR14278/ BRB/10/1417/2015, Department of Biotechnology, Government of India, awarded to Nagaraj Guru Prasad and RR.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The raw data for all the experiments have been archived at Dryad with https://doi.org/10.5061/dryad.w0vt4b8s9, https://datad ryad.