INTRODUCTION
Phylosymbiosis occurs when host phylogenetic relationships parallel the community relationships of the host-associated microbiota (
1,
2). A central prediction of phylosymbiosis is that phylogenetically similar host species will exhibit lower microbiota beta diversity than distantly related hosts, in contrast to the null hypothesis that microbiota beta diversity relationships do not parallel host phylogenetic relationships (
Fig. 1A). Analyses of interspecific variation in microbiota in or on diverse body sites now frequently evaluate the presence or absence of phylosymbiosis. A growing number of examples of phylosymbiosis include the gut microbiota of
Nasonia parasitoid wasps (
1,
3),
Peromyscus deer mice (
2), American pikas (
4), 23 different bat genera (
5), dozens of mammalian species (
6), and
Cephalotes ants (
7). Examples of phylosymbiosis also occur on the skin and surfaces of animals in aquatic environments including seven different
Hydra species (
8,
9), 20 sponge species (
10), 21 major coral clades (
11), and 44 tropical reef fish species (
12).
It is important to highlight that phylosymbiosis is a measurable pattern that is assumption free with respect to process. First and foremost, it does not assume a stable evolutionary association of hosts and their microbiota or congruent splitting of ancestral, interacting organisms that may result from cospeciation or coevolution. However, it can be enhanced by such processes (
11). Second, it does not assume that microbial communities are vertically transmitted. Rather, phylosymbiosis may be driven by vertical inheritance of the microbiota or persistent host filtering of acquired microbes, and it accommodates the possibility that host-microbiota associations could change each host generation due to various diets or environments (
2,
13). Thus, phylosymbiosis is a pattern measured in a single instance of time and space, and it can theoretically change upon host exposures to a different community of environmentally acquired microbes.
As phylosymbiosis emerges as a bona fide, though not universal, trend in studies of diverse systems, a salient set of questions is emerging. What evolutionary forces (selection versus neutrality) underpin the pattern? What are the number and types of host and microbial genes that contribute to phylosymbiosis? Is phylosymbiosis more readily detectable in certain body sites or groups of host organisms? How consequential is phylosymbiosis to host biology? A model system with phylosymbiotic communities, host genetically tractability, interspecific interbreeding, axenic host rearing, and bacterial cultivability will be helpful in interrogating these long-term questions.
To test whether phylosymbiosis is consequential or not to host biology, microbiota transplantation experiments to germfree hosts can reveal whether an interspecific microbiota is helpful, harmful, or harmless relative to an intraspecific microbiota (
Fig. 1B). For example, mice humanized with microbiota by oral gavage exhibit immunodeficiencies and increased susceptibility to enteric pathogens (
14). However, transplants between distantly related hosts such as mice and humans cannot directly unravel the evolutionary processes and consequences that underpin phylosymbiosis. Comparisons between recently diverged species with phylosymbiotic microbiota are crucial for dissecting the consequences to host biology and selective pressures shaping phylosymbiosis.
We previously demonstrated that germfree
Nasonia larvae exposed to interspecific microbiota suffered significant reductions in adult survival in comparison to intraspecific microbiota (
2). However, that study did not reveal the developmental or survival defects in early life that precede and directly impact the reductions in adult survival. Moreover, crosses between two species of
Nasonia lead to F2 hybrid death in haploid male larvae in conventionally reared hybrids and germfree hybrids exposed to microbiota but not in germfree hybrids (
1). Thus, hybrid lethality is similarly contingent on the presence of gut bacteria, which suggests that hybrid breakdown results from costly host-microbiota interactions that are absent in the parental
Nasonia species exhibiting phylosymbiosis.
The parasitoid wasp genus
Nasonia (also known as the “jewel wasp”) is an evolutionary genetics model (
15,
16) well suited for understanding the impacts of host-gut microbiota symbioses and recent speciation events because it is comprised of four interfertile species,
N. vitripennis,
N. longicornis,
N. giraulti, and
N. oneida (
17,
18), that exhibit various degrees of reproductive isolation (
19). The species
N. vitripennis is estimated to have diverged from the ancestor of the three younger species 1 million years ago (mya).
N. longicornis and
N. giraulti diverged 0.4 mya, and
N. giraulti and
N. oneida diverged 0.3 mya (
20,
21). Adult
Nasonia female wasps lay their eggs within the puparia of
Sarcophaga bullata flesh flies (
17). Under laboratory conditions (constant temperature of 25°C),
Nasonia has a short generation time of approximately 14 days, in which wasp metamorphosis and development occur within the
S. bullata puparium.
Nasonia affords an ease of conventional and germfree rearing (
1,
22,
23), the ability to establish interspecies hybrids after curing of
Wolbachia (
24,
25), and the genetic advantages of haplodiploid sex determination, wherein males and females develop from unfertilized (haploid) and fertilized (diploid) eggs, respectively.
The
Nasonia male microbiota is dominated by
Providencia spp. and
Proteus spp. during second instar larval development, typically representing 81 to 96% of the bacterial community. These bacteria are rod-shaped members of the
Enterobacteriaceae family closely related to other intestinally associated insect bacteria such as
Escherichia and
Enterobacter. The dominant gut bacterial genus in
Nasonia vitripennis and
Nasonia giraulti larvae is
Providencia, while the dominant gut genus in
Nasonia longicornis larvae is
Proteus (
1). Notably in
Drosophila melanogaster, several different
Providencia strains induce variable levels of host immunogenicity (
26,
27). This variability in immune response may provide a potential pathway by which these bacteria alter
Nasonia development and survival.
Considering that much of the metabolic activity, growth, and development of
Nasonia occur before adult eclosion, the functional effects of an interspecific microbiota will likely arise during the
Nasonia life cycle following exposure in the larvae. In this study, we test the impacts of exposure of an interspecific microbiota on developing males using three
Nasonia spp. and developmental stages spanning initial embryo hatching to adulthood and reproduction. The metrics analyzed include larval growth, pupation rates, and adult fertility and longevity. The microbiota treatment used for
Nasonia exposure was heat inactivated in order to prevent bacterial overgrowth in the
Nasonia in vitro rearing system that is also used for axenic rearing. The bacterial heat inactivation prevents
Nasonia wasp exposure to microbiota under unchecked
in vitro growth conditions. Heat-inactivated bacteria are commonly used as a means to probe a host’s immune response to bacterial stimuli (
28–30) and metabolic rate (
31,
32). Thus, this design tests the impacts of early-life exposure to inactivated microbiota on host responses during important periods of growth and metamorphosis.
DISCUSSION
Phylosymbiosis describes the ecoevolutionary pattern whereby the ecological relatedness (e.g., beta diversity relationships) of host-associated microbial communities parallels the phylogeny of the host species (
1,
2). For the bacterial microbiota, phylosymbiosis is emerging as a widespread, though not universal, trend in a rising variety of animal holobionts (
2,
4,
7,
11,
33,
34). This pattern was also recently extended, for the first time, to the virome in
Nasonia parasitoid wasps (
35). These results indicate that
Nasonia evolution is associated with distinguishable microbiota and virus communities whose relationships recapitulate the wasp’s phylogeny. From this perspective, the microbiota or virome community composition is akin to a phylogenetic marker for the host species.
A key challenge in the study of phylosymbiosis is determining the evolutionary processes (selection or drift) that shape phylosymbiosis. For example, if the pattern results from inconsequential processes untethered to host fitness or performance, then hosts are not expected to benefit from a phylosymbiotic microbiota. However, if the pattern results from a selective pressure on hosts, then decreases in host fitness are expected upon exposure to interspecific or other nonnative microbiota. Selective pressures could drive the evolution of host traits that filter environmental microbiota in a specific way or facilitate vertical transmission of a host-associated microbiota. Selection on members of the microbiota to assemble in a phylosymbiotic manner may also enable the evolution of traits that optimize microbial fitness (e.g., replication) within the host. Higher microbial replication within the host may lead to increased rates of dissemination into the environment whereby members of the microbiota gain a (re)colonization advantage in the next host generation since there may be an increased likelihood of contact between hosts and microbes. This cycle between host-associated replication and environmental dissemination to other hosts could be a positive-feedback loop for microbial adaptations that contribute to the assembly of phylosymbiosis.
Here we report that early-life exposure of germfree Nasonia wasps to an interspecific heat-inactivated microbiota yields detrimental developmental and fitness impacts that begin early in larval development. After 3 days of daily exposure to heat-inactivated interspecific microbiota, larval growth delays in N. vitripennis and N. giraulti were most prominent during the point of peak growth in control wasps. Interestingly, in late larval development, the consequences of exposure to an interspecific microbiota subsided as larval size tended to equilibrate across the different treatment groups. However, comparable larval size this late in larval development may not be indicative of the energy storage needed at this developmental stage to successfully undergo pupation.
After microbiota exposure was halted during late larval development, recipient
N. vitripennis larvae were significantly less likely to undergo pupation if they received interspecific microbiota from either of the more distantly diverged sister species. In
N. giraulti recipient larvae, significant reductions in pupation occurred only with exposure to microbiota from their more distantly diverged sister species,
N. vitripennis (1 mya), but not from the more closely related species,
N. longicornis (400 thousand years ago [kya]). This disparity in
N. giraulti development between the microbiota donors could shed light on the evolutionary timescale in which
Nasonia wasp species develop a uniquely qualified and relatively beneficial microbiota. However, it is important to note that
N. giraulti larval exposure to the
N. longicornis microbiota did result in a reduction of pupation. Overall, these decreases in pupation from the interspecific microbiota directly translate to decreased adult survivability previously reported from the same experiment (
2). By connecting
Nasonia developmental delays with the resulting adult survival, we conclude there is a consequential impact of early-life exposure to interspecific microbiota on
Nasonia development and fitness. Thus, phylosymbiosis is adaptive because the functional interactions between host and microbiota impact metamorphosis and therefore survival.
The
Nasonia wasps were exposed with a normalized concentration of heat-inactivated microbiota from fourth instar larvae, so it is unlikely that the negative effects from an interspecific microbiota resulted from reduced nutrition conferred by the heat-inactivated microbiota, since they had comparable bacterial levels composed mostly of a few genera of
Gammaproteobacteria. These microbiota exposures were also provided alongside a nutrient-rich
Nasonia rearing medium (NRM) that, alone, has been sufficient to achieve
in vivo levels of wasp growth and development. The
in vitro rearing medium is made using the proteinaceous homogenate collected from
Sarcophaga bullata pupae, the fly host of
Nasonia wasps (
23). In the process of making the
Nasonia rearing medium, much of the pupae’s lipid contents are removed because they cause clogging in the sterile filtration process. Thus, we speculate that the interspecific differences in response to the medium control treatment group occur because
N. vitripennis larval growth is slightly hampered by this lipid-limited environment, which is compensated by the addition of heat-inactivated microbes.
N. vitripennis is a generalist species that parasitizes many different fly pupae in nature. Conversely,
N. giraulti is a specialist on
Protocalliphora bird fly pupae that are not reared in the lab. Thus, with our rearing medium, it is possible that a higher protein content artificially benefits germfree
N. giraulti and results in a slight increase in larval growth, and this benefit could be negated upon exposure to microbes and the ensuing physiological changes that occur. While the medium control group had early larval growth trends, these nonsignificant growth differences disappeared by late larval development and did not influence the rest of the wasp life cycle whereas developmental differences from interspecific microbiota exposures were present through pupation.
We hypothesize that
Nasonia larvae maintain a greater immune tolerance to intraspecific microbiota exposure than interspecific exposures. During larval development, the largest effects of the interspecific microbiota exposures occurred at a time when most of the larval energy supply was utilized on growth and storage. Any shunting of that energy flow into an activated immune response would presumably stunt growth (
36), which was seen in
Fig. 2. Prior studies indicate that
Drosophila exposure to
Providencia spp. can result in delayed larval growth and survival (
26,
37). Moreover, in
Drosophila, larvae that are unable to achieve adequate stores of lipids and other nutrients will be less likely to complete an energy-costly pupation (
38,
39). However, additional experimentation is necessary to compare immune gene expression levels of key antibacterial pathways such as Toll and Immune Deficiency to show that interspecific microbiota transplants upregulate an immune response during larval development.
In
Hydra, it has previously been observed that phylosymbiosis is in part regulated by species-specific antimicrobial peptide expression. Loss-of-function experiments demonstrated that distinct antimicrobial peptide compositions were necessary to maintain phylosymbiosis across multiple
Hydra spp. (
9). Selective pressures may similarly shape phylosymbiosis in
Nasonia wasps through host immune pathways that curate and/or tolerate particular members of microbial communities over others. The impacts of selection on
Nasonia phylosymbiosis are also supported by hybrid death and phylosymbiosis breakdown previously observed in
N. vitripennis and
N. giraulti F2 hybrid larvae. The loss of immune competence in these
Nasonia hybrid larvae, as evidenced by hypermelanization and an overexpression of immune genes, coincides with 78% larval death and a marked shift in microbiota composition from the parent’s phylosymbiotic microbiota (
1). Conversely, F2 hybrid larvae from the two younger species,
N. giraulti and
N. longicornis, do not melanize, die, or show shifts from the parental microbiota.
Beyond immunity, other mechanisms can play a role in the consequences of wasp-microbiota interactions such as metabolism, developmental signaling, and mate discrimination. For example, while there was no significant impact on adult reproductive capacity from the heat-inactivated transplants, it is possible that additional rounds of microbiota exposure during development could impact adult reproduction or perhaps adult mating behavior. In summary, this research reveals that early-life exposure of interspecific microbiota impacts larval growth and pupation within closely related species of the Nasonia model system. Further, by comparing the functional impacts of the gut microbiota across species that diverged between 400,000 and 1,000,000 years ago, we have shown that phylosymbiosis is not just indicative of a recent host phylogenetic effect on microbiota composition but is also important to traits involved in early-life host growth and fitness. Because Nasonia is a tractable animal system with interfertile species, it is an ideal model for testing the genetic basis of interconnections between development, microbial symbiosis, and speciation.
MATERIALS AND METHODS
Nasonia strains and collections.
Wolbachia-uninfected
N. vitripennis AsymCx,
N. giraulti RV2x[u], and
N. longicornis NLMN8510 mated females were hosted on
S. bullata pupae and housed in glass culture tubes capped with cotton at 25°C ± 2°C in constant light, as previously described (
20). After 10 to 12 days,
S. bullata puparia were opened, and virgin
N. vitripennis or
N. giraulti female offspring were collected as black pupae. Upon adult eclosion, 200 individual virgin females were isolated and provided two
S. bullata pupae for 2 days of hosting to increase the egg deposition. As haplodiploids,
Nasonia virgin females are fecund and lay all male (haploid) offspring. After the initial 2 days of hosting, females were provided with a new
S. bullata pupa housed in a Styrofoam plug, allowing them to oviposit only on the anterior end of the host for easy embryo collection.
Germfree rearing of Nasonia.
N. vitripennis AsymCx or
N. giraulti RV2x[u] embryos were extracted with a sterile probe from
S. bullata hosts parasitized by virgin females after 12 to 24 h. Twenty to 25 embryos per host were placed on a 3-μm-pore transwell polyester membrane (
n = 12 to 15 transwells per treatment group) (Costar; Corning Incorporated, Corning, NY, USA) and sterilized twice with 70 μl of 10% bleach solution and once with 70 μl of 70% ethanol solution. The embryos were then rinsed three times with 80 μl of sterile Millipore water. After rinsing, the transwell insert was moved into a 24-well plate with 200 μl of NRM (prepared according to the NRMv2 protocol described in reference
23) in the basolateral compartment. All plates were stored in an autoclaved Tupperware box at 25°C ± 2°C under constant-light conditions for the duration of the experiment. Under sterile laminar flow, transwells were moved to new wells with 200 μl of fresh NRM every day. After 8 days, the transwells were drained of their medium on a sterile Kimwipe and moved to a clean, dry 24-well plate, and the 12 empty surrounding wells were filled with 1 ml of sterile Millipore water and 65.7 mM Tegosept solution to increase humidity and prevent fungal growth.
Heat-inactivated microbiota preparation.
We tested the effects of interspecific microbial communities on host survival by transplanting heat-inactivated microbiota from three donor Nasonia species (N. vitripennis, N. giraulti, and N. longicornis) into N. vitripennis or N. giraulti male recipients. Microbiota were purified from fourth-instar larvae of each of the Nasonia donor species by homogenization of ∼100 larvae in 200 μl of sterile 1× PBS. The larval homogenate was then centrifuged at 800 rpm for 3 min to pellet large cellular debris, and the resulting supernatant was filtered through a 5- μm filter. The filtrate was centrifuged at 10,000 rpm for 3 min, and the supernatant was removed. The pellet was resuspended in sterile 1× PBS. This centrifugation step was repeated, and the pellet was resuspended in 200 μl 1× PBS. After the suspension was plated on tryptic soy agar to determine the rough microbiota concentration, it was heat inactivated by placement in a 75°C water bath for 1 h. After counting colonies on the tryptic soy agar plates, the heat-inactivated suspension was diluted in sterile 1× PBS to achieve a concentration of 5 × 106 CFU of microbiota bacteria per milliliter. This procedure was performed independently for each of the donor microbial communities 1 day before exposure. Unfortunately, the current germfree rearing technique for Nasonia makes stable, live microbiota transplantation difficult to maintain. Because antibiotics and exogenous fetal bovine serum (FBS) were removed from the methodology to maintain a more natural system, the Proteus spp. and Providencia spp. have ample nutrients for and little resistance in the medium to overproliferating within the first 8 h after exposure. The issue lies with the stagnant environment of the Nasonia rearing medium within the transwell during Nasonia development.
Transplantation of the heat-inactivated Nasonia microbiota.
After the first 24 h of germfree Nasonia rearing, the transwell inserts containing L1-stage larvae were randomly separated into four experimental groups for each recipient Nasonia genotype: a PBS negative control and three heat-inactivated N. vitripennis, N. giraulti, and N. longicornis microbiota exposure groups. For the transwell replicates in each group, 20 μl of this microbiota suspension was added directly to the transwell inserts daily during the eight transplantation days before Nasonia pupation. Nasonia rearing medium was replaced daily just before the inoculations. On days 2 to 8 of exposure, the transwell insert was drained on a sterile Kimwipe before the addition of the 20 μl of microbiota suspension. If there was any bacterial or fungal contamination in a transwell during the course of the Nasonia development, the transwell and its data were removed from the experiment.
Comparative analysis of Nasonia development.
After the replacement of Nasonia rearing medium and addition of heat-inactivated Nasonia microbiota, a picture was taken of each well under magnification using a microscope-attached AmScope MT1000 camera. Pictures of the developing Nasonia in each transwell were taken from the first day of heat-inactivated microbiota exposure to the first day of adult eclosion (∼14 to 15 days). Starting on the second day of microbiota exposure, larval length was determined using ImageJ software by measuring the anterior-to-posterior end of larvae for all larvae in the transwell. Normalized larval growth per transwell sample was calculated as the average larval length of Nasonia per transwell divided by the average larval length of the intraspecific microbiota treatment group for each treatment day. Larvae were identified as dead if they were visibly desiccated or malformed and were not included in the analysis.
On the last day of Nasonia pupation before adult eclosion (5 days after initial pupation begins), the proportion of pupated larvae was measured for each transwell. Normalized pupation per transwell was calculated as the percent pupated Nasonia per transwell divided by the average percent pupated of the intraspecific microbiota treatment group.
For each transwell, live larval counts were recorded on the third day after embryo deposition to ensure the embryos hatched. Adult counts were determined by first recording the number of remaining larvae and pupae in each transwell 20 days after embryo hatching (5 to 7 days after first adult eclosion) and then subtracting that number from the larval counts previously recorded. Normalized adult survival per transwell sample was calculated as the percent survival of Nasonia from 3 days to 20 days after embryo hatching divided by the average percent survival of the intraspecific microbiota treatment group. Larvae and pupae were scored as dead if they were visibly desiccated or malformed. Larval growth, pupation, and adult survival between the intraspecific and interspecific treatment groups were compared using a Kruskal-Wallis test with a Dunn multiple-comparison test.
Comparative analysis of adult reproductive capacity.
After initial adult eclosion from each experimental group, adult males were collected from the transwells daily and mated individually with conventional newly eclosed (<24-h) virgin females for 24 h. We measured adult male longevity by moving the individual males to a sterile glass after mating and recording the male survival status every 12 h. Adult males with deformities (walking impediments or deformed wings) that may influence mating were not used in the analysis. The mated female was then given two S. bullata pupae to parasitize. When the offspring pupae reached the black pupal development stage, the males and females were counted and a sex ratio was recorded.