Effects of Different Hosts on Bacterial Communities of Parasitic Wasp Nasonia vitripennis

Parasitism is a special interspecific relationship in insects. Unlike most other ectoparasites, Nasonia vitripennis spend most of its life cycle (egg, larvae, pupae, and early adult stage) inside the pupae of flies, which is covered with hard puparium. Microbes play important roles in host development and help insect hosts to adapt to various environments. How the microbes of parasitic wasp respond to different fly hosts living in such close relationships motivated this investigation. In this study, we used N. vitripennis and three different fly pupa hosts (Lucilia sericata, Sarcophaga marshalli, and Musca domestica) to address this question, as well as to illustrate the potential transfer of bacteria through the trophic food chains. We found that N. vitripennis from different fly pupa hosts showed distinct microbiota, which means that the different fly hosts could affect the bacterial communities of their parasitic wasps. Some bacteria showed potential horizontal transfer through the trophic food chains, from the food through the fly to the parasitic wasp. We also found that the heritable endosymbiont Wolbachia could transferred from the fly host to the parasite and correlated with the bacterial communities of the corresponding parasitic wasps. Our findings provide new insight to the microbial interactions between parasite and host.


INTRODUCTION
As the most abundant animals worldwide, insects benefit greatly from their symbionts, as they help the insects in developing different strategies to adapt to various environments. With the rapid development of high-throughput sequencing technology, the composition and roles of microorganisms associated with insects have aroused extensive interest in the scientific community.
Many factors can affect the bacterial communities of insects: firstly, the species of insect hosts, which is known to shape their bacterial communities. For example, the microbial communities of honey bees and bumble bees exhibit similar patterns as the clustering of the host (Martinson et al., 2011;Koch et al., 2013). Secondly, diet can affect the hosts' bacterial diversity. Previous studies have found that the composition of intestinal microorganisms in larvae of Helicoverpa armigera (Lepidoptera, Noctuidae) is similar to that of the plant community, indicating that food is an important source of intestinal microorganisms (Priya et al., 2012). According to a study on Monochamus alternatus (Coleoptera, Cerambycidae), Enterococcus is the dominant bacterium in the intestinal tract of larvae feeding on natural food, while Lactococcus is the dominant bacterium in the intestinal tract of larvae feeding on artificial feed (Kim et al., 2017). Distantly related hosts that live in close symbiotic relationship can also maintain similar microbial communities. With varying degrees of nest sharing between them, Megalomyrmex social parasites (Solenopsidini) and their fungus-growing ant hosts from the genera Cyphomyrmex, Trachymyrmex, and Sericomyrmex share symbiont bacteria (Liberti et al., 2015). Thirdly, spatial and temporal distributions of the host could influence their bacterial compositions (Zhang and Blackwell, 2002;Falush et al., 2003;Hedlund and Staley, 2004;Papke and Ward, 2004;Martiny et al., 2006). The bacterial structure of termites shows significant differences between diverse sampling sites (Hongoh et al., 2005).
Nasonia vitripennis (Hymenoptera, Pteromalidae) is an important parasitoid whose female wasp stings, injects venom, and lays eggs in many different fly pupae, where their eggs, larvae, pupae, and early-stage adults are developed. N. vitripennis usually lives on species of the family Calliphoridae, Muscidae, and Sarcophagidae. Because of its short generation time, large offspring production, and easy rearing, Nasonia has emerged as a model organism for developmental and evolutionary genetics (Werren and Loehlin, 2009), as well as a research model for hostmicrobial community interaction studies (Dittmer et al., 2016).
Parasitism is a special symbiotic relationship in insects. Unlike other ectoparasitic wasps that usually live on the surface of their hosts, N. vitripennis spends most of its life cycle inside fly species, which are covered with puparium; thus, it lives in an enclosed environment. The microbiota from both fly species and the wasps may have more opportunities to communicate due to such an enclosed environment (Lynch, 2015). The microbial communities of N. vitripennis have been previously reported (Brooks et al., 2016;Jessica et al., 2016). However, the response of the microbial community composition of the wasps to different hosts in such an enclosed environment is still unknown. In this study, we used N. vitripennis and three different fly hosts (Lucilia sericata, Sarcophaga marshalli, and Musca domestica) with different food resources to address the following issues: (1) how the different hosts affect the bacterial communities of N. vitripennis, (2) whether there is bacterial transfer through food chains, and (3) whether the existence of the endosymbiont Wolbachia in fly species could affect the microbiota of the parasitic wasp.

Experimental Insects
Three different fly hosts, L. sericata (Ls, Diptera: Calliphoridae), S. marshalli (Sm, Diptera: Sarcophagidae), and M. domestica (Md, Diptera: Muscidae), were trapped at the campus of Shandong Agricultural University, China, and then have been kept in the laboratory since 2012. Fly traps were baited with household pork to obtain eggs of the carnivorous L. sericata and S. marshalli, as well as wheat bran mixed with water to obtain eggs of M. domestica. Laboratory rearings were performed with pork carrion (Ca) for the carnivorous flies (L. sericata and S. marshalli), whereas M. domestica was reared on wheat bran (Wb). For larval development rearing was done at 28 • C, 50% relative humidity (RH) and 16:8 h (L:D) photoperiod. All fly adults were provided with a 10% honey solution (in water) provided ad libitum (Meister, 1962). N. vitripennis was routinely reared on M. domestica, at the same rearing conditions as descried above. Microbial diversity assays were performed starting from this lab strain, as indicated in Figure 1. Briefly, coetaneous N. vitripennis females were split into three batches. Each batch was further reared on L. sericata (NvLs), S. marshalli (NvSm), or M. domestica (NvMd) pupae for 10 generations. After 10 generations, parasitized hosts (1-day-old pupae) were isolated in 50-ml tubes to allow parasitoid development, and unparasitized hosts were used as controls, as indicated in Figure 1.

DNA Extraction, PCR Amplification, and Sequencing
Regardless of gender, fresh fly species at the pupal stage (1day-old) were collected for DNA extraction, although sex could have an impact on microbial communities (Han et al., 2017). Twenty M. domestica pupae, 10 S. marshalli pupae, and 10 L. sericata pupae were selected (members of each fly type were considered as one sample), while more than 100 N. vitripennis individuals were collected before emerging from each of the fly hosts. The fresh fly pupa hosts and parasitic wasps were sterilized in 75% ethanol for 1 min and washed in sterile water for 30 s three times. After sterilization, the samples were collected in 1.5-ml centrifuge tubes and stored in liquid nitrogen. In order to avoid bacterial contamination outside of the puparium, which could affect the bacterial composition of the wasps, after 12 days of the parasite inside the pupa, the puparium was broken with sterilized tweezers and knives. Further, N. vitripennis was transferred to sterilized 1.5-ml centrifuge tubes with small holes to maintain air circulation for the wasps to develop. After 2 days, the wasps were sterilized in 75% ethanol for 1 min and washed in sterile water three times. About 100 individuals were collected in a 1.5-ml centrifuge tube and stored in liquid nitrogen for DNA extraction. Total DNA was extracted from the fly species at the pupal stage and parasitic wasps from different fly hosts, carrion meat, and wheat bran using the OMEGA Soil DNA Extraction Kit. The extracted DNA was amplified by PCR using the bacterial 338F (5 -ACTCCTACGGGAGGCAGCAG-3 )-806R (5 -GGACTACHVGGGTWTCTAAT-3 ) primer set targeting the 16S rRNA gene. PCR conditions were as follows: pre-denaturation at 94 • C for 5 min; 35 cycles of denaturation at 94 • C for 40 s, annealing at 45-59 • C (adjusted according to different primers) for 30-60 s, and an extension at 72 • C for 1 min; and a final extension for 10 min at 72 • C. The qualified DNA samples were selected and sequenced on the Illumina MiSeq platform of Majorbio Bio-Pharm Technology Co., Ltd., Shanghai, China.
The gene coding for Wolbachia surface protein (wsp gene) was detected using the Wolbachia-specific primer set of 81F (5 -TGGTCCAATAAGTGATGAAGAAAC-3 ) and 691R (5 -AAAAATTAAACGCTACTCCA-3 ), and the amplification profiles all followed Xiao et al. (2012). It was observed that L. sericata and M. domestica were not infected with Wolbachia, while S. marshalli was divided into two groups: one group was infected with Wolbachia, while the other group was Wolbachia free.

Bioinformatics and Statistical Analyses
MiSeq sequencing provided double-ended sequence data. First, according to the overlap between paired-end (PE) sequencing reads, pairs of reads were spliced (merge) into a sequence, the reads were filtered based on their quality, and the results were merged by using QIIME2 (Bolyen et al., 2019). The quality filtering standards for sequential reads are as follows. First, the reads with a mass value of less than 20 were filtered, and a 50-bp window was set. If the average mass value in the window is less than 20, the back-end bases were truncated from the window. Reads below 50 bp after quality control were filtered to remove the reads containing N bases. Then, according to the overlap relationship between PE reads, paired reads were spliced (merged) into a sequence, and the minimum overlap length was 10 bp. Then, the maximum error ratio allowed in the overlap area of the mosaicing sequence is 0.2, and the nonconforming sequence is screened. Finally, samples were differentiated according to the barcode and primer at both ends of the sequence, and the sequence direction was adjusted. The allowable mismatches of the barcode were 0, and the maximum primer mismatches were 2. According to the sequence of the two ends of the barcode and the primer sequence, the valid sequences were distinguished from the sample. Then the sequence orientation was corrected, and the optimized data were acquired. The non-repetitive sequences were extracted from the optimized sequences, which could reduce the computational complexity of the intermediate process, the single sequences without duplication were removed, and operational taxonomic units (OTUs) based on the nonrepetitive sequences (excluding the single sequence) were clustered according to 97% similarity by using UCHIME. During the clustering process, chimera sequences were removed, and the representative sequences of OTUs were obtained. All the optimized sequences were mapped to the OTU representative sequence, and the sequences with a similarity of more than 97% were selected to generate the OTU table. In order to obtain the information of the species corresponding to each OTU, the RDP classifier Bayesian algorithm was used to analyze 97% similarity of the OTU representative sequence against the SILVA ribosomal RNA gene database using a confidence threshold of 70%.
Based on 97% similarity of the OTUs or other taxonomic levels, the diversity of random sampling in the form of diversity index was calculated using mothur. Alpha diversity and beta diversity were calculated by GmT (Galaxy mothur Toolset) and QIIME2, respectively (Hiltemann et al., 2018;Bolyen et al., 2019). In order to study the similarity or differences in the bacterial composition of the samples, the number of common and unique OTUs in the samples was counted. Based on UniFrac coupled with principal coordinate analysis (PCoA) for all the samples, PCoA and Venn graphs were plotted using the R package. A heatmap plot for abundance analysis was made by using the R package. According to the classification of the composition of the samples and different grouping criteria for linear discriminant analysis (LDA), LDA effect size (LEfSe) found a significant difference in the sample division of the community or species. Differences between populations were analyzed by using one-way ANOVA. Excel was used to show the horizontal transfer of bacteria through the tertiary food chain food-host-parasitic wasp. Information about the species within or between groups was obtained by using the NetworkX. The Network Analysis Toolkit was used to calculate the node degree distribution, the diameter and the average shortest path of the network, the degree of connectivity, and closeness centrality (Hagberg et al., 2008). In Cytoscape 2.8.3, the layout was set as spring embedded, and nodes were used as connection points to reflect the connection between the sample and OTUs and to make a network diagram (Shannon et al., 2003). In order to explore the influence of Wolbachia on other bacteria, SPSS 22.0 was used to calculate the Pearson rank correlation coefficient, and Gephi 0.9.2 (Bickel, 2003) was used for figures. In addition, we took "fly species-parasitic wasp" and "carrion/wheat bran-fly species-parasitic wasp" as the whole, and the bacteria that were shared in host, parasitic wasp, and food in the food chain were figured out using Excel and iTOL software (Letunic and Bork, 2019). FastTree was used to construct the phylogenetic tree based on the maximum likelihood method by selecting the sequences corresponding to OTUs or a class of classified information at different levels. According to the cluster of orthologous groups of proteins (COG) database, the descriptions of each COG and its function were analyzed based on the eggNOG database (Huerta-Cepas et al., 2019). In addition, PICRUSt was used to obtain different levels of metabolic pathway information, as well as the abundance table at each level (Langille et al., 2013).

Nasonia vitripennis Microbiota Profiles in Different Host Species
A total of 754,311 16S rRNA gene sequences were obtained (accession number PRJNA479943). There were 651 OTUs, which were clustered based on 97% 16S rRNA similarity and belonged to 23 prokaryotic phyla. Rarefaction curves showed that all the samples showed enough sequencing depth to represent their bacterial diversity (Supplementary Figure S1).
Interestingly, the network showed a significant microbiota clustering of N. vitripennis and their corresponding fly hosts rather than the bacterial clustering based on fly species (Figure 2). M. domestica and Wolbachia-infected S. marshalli (SmIn) and Wolbachia-free S. marshalli (Sm) showed a closer relationship and more intersections in terms of bacterial community composition with their corresponding parasitic wasps than did L. sericata and its parasitic N. vitripennis; both showed the maximum number of unique OTUs.

Effect of Different Fly Hosts on Microbial Community of Parasitic Wasps
Wasps and their fly hosts showed different bacterial communities. At the phylum level, Proteobacteria was predominant in both L. sericata and S. marshalli whose adults were fed with carrion, while Bacteroidetes and Proteobacteria shared the predominant position in M. domestica whose adults were fed with bran. For N. vitripennis, Proteobacteria was predominant in wasps with L. sericata (NvLs) and S. marshalli (NvSm) as hosts, while Firmicutes was dominant in wasps with M. domestica (NvMd) as the host (Figure 3A). At the genus level, the differences were more defined ( Figure 3B). Providencia (41.28%) and Myroides (37.61%) were the predominant genera in L. sericata and M. domestica, respectively. For Wolbachia-free S. marshalli (Sm), Massilia (16.19%), and Sphingobium (12.85%) were the predominant genera, while for Wolbachia-infected S. marshalli (SmIn), Providencia (27.27%) and Vagococcus (16.13%) were the predominant genera, although there was no evident difference at the phylum level due to the presence or absence of Wolbachia. The bacterial communities of the parasitic wasps from different hosts were quite interesting. Gluconobacter (31.24%) and Staphylococcus (81.47%) were the dominant bacterial genera in wasps whose hosts were L. sericata (NvLs) and M. domestica (NvMd), respectively. Proteus (56.03%) and Wolbachia (59.17%) were the dominant genera in N. vitripennis whose hosts were Wolbachia-free (Sm) and Wolbachia-infected S. marshalli (SmIn), respectively ( Figure 3B).
Bacteroidetes was the predominant bacteria in carrion, which was used as food for carnivorous insects L. sericata and S. marshalli, while Proteobacteria was the predominant bacteria in wheat bran, which was used as food for the phytophagous fly M. domestica (Figure 3A). The bacterial community of N. vitripennis was quite different from its fly hosts based on PCoA (Figures 3C,D and  Supplementary Figure S2). Chao and Ace indices showed that there were significant differences in the bacterial diversity index between different samples (Supplementary Table S1). The bacterial diversity of N. vitripennis with L. sericata (NvLs) as host was significantly higher than that of the other samples. A rank-abundance graph showed that the curve length and span of N. vitripennis with L. sericata (NvLs) as host were higher than those of the other samples, indicating that the species richness FIGURE 5 | Shared bacteria of different food chains (food-pupae-wasps or pupae-wasps). The phylogenetic tree was constructed by the top 50 OTUs in each sample (internal). The blue solid circle represents the presence of the bacterial OTUs in the food chain, and the hollow circle represents the absence of bacterial OTUs in the food chain. For example, OTU153 showed that it was able to pass through the food chain of the wheat bran-M. domestica pupae-its corresponding N. vitripennis. In addition, it was also able to pass from L. sericata pupae to N. vitripennis. and species uniformity of this sample were the highest (Supplementary Figure S3).
The bacterial hierarchical clustering of the samples did not occur based on species. M. domestica and its corresponding N. vitripennis (NvMd) were close to wheat bran in a distance similar to L. sericata and its corresponding N. vitripennis (NvLs). The parasitic wasps hosted by Wolbachia-infected (SmIn) and Wolbachia-free S. marshalli (Sm) were clustered together (Figure 4).
There were 221 and 193 OTUs detected in L. sericata and M. domestica, respectively, and there were 166 and 176 OTUs in Wolbachia-infected (SmIn) and Wolbachia-free S. marshalli (Sm), respectively ( Figure 3E). Acinetobacter, Bacillus, Citrobacter, Lactococcus, Morganella, and Providencia were common in all types of fly species (Figure 3E). Massilia, Bacillus, Serratia, Providencia, Lactococcus, Staphylococcus, Morganella, Proteus, Bacteroides, Enterobacter, and Myroides were common OTUs in N. vitripennis hosted by all the fly species ( Figure 3F). The bacterial diversity of N. vitripennis with L. sericata (NvLs) as host was also higher than that of N. vitripennis inhabiting other hosts. A total of 516 and 199 OTUs were detected in N. vitripennis with L. sericata (NvLs) and M. domestica (NvMd) as hosts, respectively, while there were 142 and 115 OTUs in N. vitripennis with Wolbachia-free (NvSm) and Wolbachiainfected S. marshalli (NvSmIn) as hosts, respectively FIGURE 6 | Bacterial networks from Wolbachia-infected and Wolbachia-free N. vitripennis. The Spearman rank correlation coefficient was calculated to reflect the correlation between bacterial species. Red and green lines represent positive and negative correlations, respectively.
( Figure 3F). The results showed that bacterial diversity of the same N. vitripennis with a different fly host was significantly different.

Potential Bacterial Transfer via Food-Fly-Wasp Tertiary Food Chain
In this study, the parasitic wasp N. vitripennis was grown in three different fly hosts. S. marshalli and L. sericata were fed with carrion, while M. domestica was fed with wheat bran; thus, the food-pupae-wasp tertiary food chain and the pupae-wasp food chain system were good models to study the potential bacterial transmission (Figure 5 and Supplementary Table S2).
Many unique bacteria only existed in the fly species at the pupal stage to parasitic wasp food chains and were not found in food to fly food chains. Most unique OTUs were found in the L. sericata-N. vitripennis food chain and confirmed the possibility of being transmitted from fly species to N. vitripennis (Figure 5 and Supplementary Table S2). Some of the unique bacteria only existed in one of the two food chain systems, from food to fly species or from fly to wasps; these were rare bacteria appearing at a low abundance.

Effects of Wolbachia on Host Microbiota
Taking all the bacteria from Wolbachia-infected (NvSmIn) and Wolbachia-free N. vitripennis (NvSm) into account, certain large coenobium was observed in this study. The coenobium, including Proteus, Myroides, Haemophilus, Delftia, Arthrobacter, Cupriavidus, Providencia, Methylocystis, Prevotella_2, and norank_f_TM146 showed a direct relationship with Wolbachia (Figure 6). Except in the cases of Prevotella_2, Methylocystis, Arthrobacter, and norank_f_TM146, the Spearman correlation coefficient between other genera and Wolbachia showed negative correlations, suggesting that Wolbachia and these genera may be in a competitive relationship (Figure 6 and Table 1).

Function Prediction Based on 16S rRNA Gene Sequence
All the bacterial KEGG metabolic pathways were mainly classified into amino acid metabolism, carbohydrate metabolism, replication and repair, energy metabolism, glycan biosynthesis and metabolism, enzyme families, membrane transport, etc. (Figure 7). Gastric triacylglycerol lipase helps insects in digestion of fat. Comparison of the differences in gastric triacylglycerol lipase expression in different fly pupae and their corresponding N. vitripennis showed that the lipase expression in carnivorous N. vitripennis was significantly higher than that in the phytophagous M. domestica, which was fed with wheat bran. Moreover, the lipase expression in S. marshalli was significantly higher than that in M. domestica (Figure 8). Although PICRUSt was deemed not suitable for nonhuman microbiomes (Langille et al., 2013), we thought it is believable in our results.

DISCUSSION
Parasitism is a symbiotic insect relationship wherein the parasite consumes its host as nutrition for development, unlike commensalism and mutualism. Nasonia is a special ectoparasite with development of more than one wasp in a single fly pupa (brood size in Nasonia ranges from 3 to 46 per host pupa); they inject their eggs through the puparium of the fly host, where the eggs hatch and then develop into larvae, pupae, and adults. Unlike other ectoparasitoids that usually live on the surface of their hosts, living under the cover of puparium makes Nasonia and its fly hosts good systems to study the interactions between insect parasites and their hosts from different perspectives.
Our results demonstrate that different hosts can affect the bacterial communities of parasitic wasps. Species is the most contributing determinant in bacterial composition of different organisms, which was confirmed by our results. Although fly food, parasitic wasps, and fly hosts had some common bacteria, such as Raoultella, Pseudomonas, Providencia, Myroides, Lactococcus, Acinetobacter, and Corynebacterium, they also showed their unique microbiota composition pattern. However, the microbiota of the parasitic wasp N. vitripennis was shaped by its corresponding host doubtlessly. From the microbiota networks of the three fly pupae and their corresponding parasitic wasps, M. domestica and Wolbachia-infected and Wolbachia-free S. marshalli showed a closer relationship and more intersections with their corresponding parasitic wasps than L. sericata and its parasitic wasps; both showed many unique OTUs, individually (Figure 2). This phenomenon needs to be investigated in our next study. Besides organism species, diet and spatial environment are also important factors that influence microbiota of an organism. It is reported that dietary specialization in mutualistic acacia-ants affects the relative abundance of the host-associated bacteria (Rubin et al., 2019). For our parasitic wasps, fly species were not only the sole diet source from egg to adult but also their closed living environment. The unique life history of the ectoparasitic wasp Nasonia also makes it a good model to study the microecology of parasite and hosts.
Different fly hosts and their corresponding N. vitripennis showed different core microbiota. Providencia, Myroides, Massilia, and Sphingobium were predominant in L. sericata, M. domestica, Wolbachia-free S. marshalli, and Wolbachiainfected S. marshalli pupae, respectively. On the other hand, bacterial communities of the same parasitic wasps from different hosts were quite interesting. Gluconobacter, Staphylococcus, Proteus, and Wolbachia were the dominant bacteria genera in wasps whose hosts were L. sericata, M. domestica, Wolbachia-free S. marshalli, and Wolbachia-infected S. marshalli, respectively. Thus, the same parasitic wasp with different hosts showed different dominant bacteria.
Moreover, some of the bacteria mentioned above were also the dominant bacteria in one or more samples. For example, Pseudomonas was one of the most abundant bacteria in Wolbachia-free S. marshalli, and Myroides was one of the most abundant bacteria in M. domestica, which are in accordance with results from a previous study by Zurek and Nayduch (2016). However, Providencia abundantly existed not only in L. sericata but also in Wolbachia-infected S. marshalli. These bacteria were the core bacteria in these samples and had important regulating effects on the physiological activity of the host. Even though there was a close relationship between the parasitic wasps and their fly pupa hosts, as N. vitripennis spends almost its whole life in the fly species, the bacterial composition of the pupae hosts and N. vitripennis showed significant differences.
Different fly hosts showed significant effects on the bacterial diversity of the parasitic wasp N. vitripennis. For this special  Table S3). Thus, different fly hosts shaped the bacterial diversity of their parasitic wasps. Brucker and Bordenstein (2012b) found that three species of Nasonia wasps, whose host was Sarcophaga bullata, were dominated by Providencia. Two possible reasons could explain this difference when compared to our results. Firstly, according to our findings, the hosts could affect the microbiota of parasitic wasps, and because different fly hosts were chosen, N. vitripennis from different hosts showed varied bacterial compositions. Secondly, different sequencing methods were used in both studies: Brucker used the first-generation sequencing method, while we used the high-through sequencing method, which can find trace bacteria; hence, our results should be more comprehensive. It is worth noting that the rearing temperature of N. vitripennis was different. Temperature can affect the microbiome of the host. Brucker used 25 • C to rear the wasp, while we used 28 • C in this study, so temperature is another possible reason for the microbial difference.
We also tried to predict a potential horizontal transfer of bacteria through secondary or tertiary food chains. N. vitripennis inhabiting S. marshalli were infected with Wolbachia, and their corresponding N. vitripennis were infected with Wolbachia too, while when the hosts were not infected with Wolbachia, the parasitic wasps were also Wolbachia free; this phenomenon is substantial evidence for horizontal transfer of bacteria, which was validated by Brucker and Bordenstein (2012b). Taking the wheat bran-M. domestica-N. vitripennis food chain as an example, OTU102 (Sphingobacterium), OTU111 (Sphingobacterium), OTU92 (Sphingobacterium), OTU104 (Pseudomonas), OTU106 (Microbacterium), OTU146 (Actinomyces), and OTU662 (Stenotrophomonas) were found in all three, wheat bran, M. domestica, and N. vitripennis, which suggested the potential horizontal transfer from food to fly and to the parasitic wasp. These food chains are a good model to investigate horizontal transfer of bacteria through a special parasitic relationship. Some trace bacteria were also observed in this study that may play important metabolic functions in their hosts, which were shared by N. vitripennis and their corresponding fly species. While providing substantial insights on the identity, structure, and taxonomic overlap of the microbiota in different food chains, our comparative results could not provide conclusive evidence for the underlying transmission mechanisms of every bacterial symbiont, because some may be shared in nature. To illuminate and confirm the transmission mechanisms, experiments are required in the future.
Endosymbiont Wolbachia is famous for various reproductive manipulations and fitness effects on their hosts. Wolbachia can not only influence the hosts' mitochondrial DNA (Xiao et al., 2012) but also can alter hosts' microbiota composition, for example, in the terrestrial isopod Armadillidium vulgare (Isopoda, Oniscoidea) (Dittmer and Bouchon, 2018) and in adult Aedes aegypti (Diptera, Culicidae) mosquitoes (Audsley et al., 2018). In our study, we found both competitive and cooperative relationships between the microbiota. The existence of Wolbachia influenced the presence and abundance of many bacterial taxa within each host population, possibly due to competitive interactions. Proteus, Myroides, Haemophilus, Delftia, Arthrobacter, Cupriavidus, Providencia, Methylocystis, Prevotella_2, and norank_f_TM146 showed symbiotic or competitive relationships with Wolbachia. The microbiota was shaped by interactions not only between the insect hosts and its symbionts but also between the different members of the symbiotic communities (Mouton et al., 2004;Oliver et al., 2006). Competition between Wolbachia and other microorganisms for resources and space in the shared host environments decreased the abundance of some bacteria.
The parasitic wasps and fly hosts are good models to study parasitic interactions from different perspectives. Detailed research on this issues, like the role of bacteria behind successful survival of wasps in new hosts, needs our attention.

DATA AVAILABILITY STATEMENT
The datasets generated for this study have been uploaded to GenBank by the accession number PRJNA479943.

AUTHOR CONTRIBUTIONS
NW conceived the idea of study and designed the methodology. HX prepared all the experimental insect samples. RD, HX, SG, ZG, and NW participated in data analysis and wrote sections of the manuscript. All authors contributed to the article and approved the submitted version.

ACKNOWLEDGMENTS
We are grateful to professor Zhong Zhang in Shandong First Medical University & Shandong Academy of Medical Sciences for assistance in the laboratory.

SUPPLEMENTARY MATERIAL
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmicb. 2020.01435/full#supplementary-material FIGURE S1 | Sparse curve analysis of the samples.