Multiple Horizontal Transfers of Bacteriophage WO and Host Wolbachia in Fig Wasps in a Closed Community

Wolbachia-bacteriophage WO is a good model system for studying interactions between bacteria and viruses. Previous surveys of insect hosts have been conducted via sampling from open or semi-open communities; however, no studies have reported the infection patterns of phage WO of insects living in a closed community. Figs and fig wasps form a peculiar closed community in which the Ficus tree provides a compact syconium habitat for a variety of fig wasp. Therefore, in this study, we performed a thorough survey of Wolbachia and bacteriophage WO infection patterns in a total of 1406 individuals from 23 fig wasps species living on three different fig tree species. The infection rates of Wolbachia and phage WO were 82.6% (19/23) and 39.1% (9/23), respectively. Additionally, phage WO from fig wasps showed strong insect host specificity based on orf7 sequences from fig wasps and 21 other insect species. Probably due to the physical barrier of fig syconium, most phage WO from fig wasps form a specific clade. Phylogenetic analysis showed the absence of congruence between WO and host Wolbachia, WO and insect host, as well as Wolbachia and fig wasps, suggesting that both Wolbachia and phage WO exchanged frequently and independently within the closed syconium. Thus, the infection pattern of bacteriophage WO from fig wasps appeared quite different from that in other insects living outside, although the effect and the transfer routes of phage WO are unclear, which need to be investigated in the future.


INTRODUCTION
Wolbachia (Alphaproteobacteria) are maternally inherited obligatory intracellular symbionts that are found in a wide range of arthropods and filarial nematodes (Stouthamer et al., 1999;Taylor et al., 2005), at rates ranging from 20 to 76% (Hilgenboecker et al., 2008;Fenton et al., 2011). The success of Wolbachia in achieving this high prevalence is associated with its ability to induce a variety of phenotypes, from mutualism in nematodes to various reproductive manipulations in arthropods, including cytoplasmic incompatibility (O'Neill and Karr, 1990), parthenogenesis (Stouthamer et al., 1990), male killing (Jiggins et al., 2001), feminization (Bouchon et al., 1998), and even speciation (Rokas, 2000). Bacteriophages are the most abundant organisms in the biosphere (Plett et al., 2011) and play important roles in bacterial genome evolution. The temperate phage WO was first detected in 2000 and is the only one bacteriophage known to infect Wolbachia (Masui et al., 2000). Accompanied by the widespread distribution of Wolbachia, it was reported that about 89% of Wolbachia strains are infected with WO (Bordenstein and Wernegreen, 2004). Polymerase chain reaction (PCR) amplification of the minor capsid gene orf7 has shown that the phage occurs in the majority of the parasitic A and B Wolbachia supergroups (Bordenstein and Wernegreen, 2004;Gavotte et al., 2007). Moreover, most phageinfected Wolbachia strains display low numbers of phage types, with 85% showing only one or two different phage types (Gavotte et al., 2007;Tanaka et al., 2009).
The tripartite insect host-Wolbachia-phage WO is an ideal model system for studying interactions among viruses, bacteria, and eukaryotes . The phage WO is the only known mobile genetic element that may transform the genome of Wolbachia (Metcalf and Bordenstein, 2012). In Wolbachia, prophage regions can comprise more than 20% of mobile DNA genes and account for the largest fraction of absent/divergent genes between closely related strains (Chafee et al., 2010). Some researchers have investigated the relationship between Wolbachia and WO in such insect species as Drosophila simulans, Ephestia kuehniella, Nasonia vitripennis, Culex pipiens, and Gryllus pennsylvanicus; moreover, no phylogenetic congruence between Wolbachia and WO has been shown, suggesting that the lateral transfer of WO in Wolbachia is not unusual (Masui et al., 2000;Bordenstein and Wernegreen, 2004;Chafee et al., 2010). Thus, unraveling the infection status and evolutionary dynamics of Wolbachia and WO may be the key to understanding the interactions among these organisms and methods for exploiting such interactions.
The syconia provide fig wasps with a compact habitat, in which Wolbachia horizontal transfer has been shown to be more likely to occur in this closed system than in other open and semi-open systems (Yang et al., 2012). Thus, in this study, we sought to determine the bacteriophage WO infection patterns in  Table 1.

PCR and Sequencing
The samples were first screened for Wolbachia infection by PCR amplification with the primers wsp81F (5 -TGG TCC AAT AAG TGA TGA AGA AAC-3 ) and wsp691R (5 -AAA AAT TAA ACG CTA CTC CA-3 ), which amplified a portion of the Wolbachia surface protein gene (wsp) (Zhou et al., 1998). If the amplification of wsp did not yield a sufficient band on agarose gels, another two pairs of primers were used, ftsZ-F/R for amplification of the Wolbachia cell division gene ftsZ and 16SwolF/R for amplification of the Wolbachia 16sRNA gene (O'Neill et al., 1992;Jeyaprakash and Hoy, 2000). WO was screened by the primers orf7F (5 -GTC TGG AAA GCT TAC AAA AAG-3 ) and orf7R (5 -GCT CTA TAA ATT CTC CTA T-3 ), and samples that were negative for orf7 were rechecked by another two pairs of gene primers, ORF2F/R and WD0633F/R (Masui et al., 2000). ddH 2 O was used as a blank control for all amplifications. PCR amplification was performed in a volume of 25 µL, containing 2.5 µL 10× buffer, 0.2 mM dNTPs, 0.5 µM of each primer, and 0.5 U of Trans Taq Enzyme (TransGen Biotech, Beijing, China).
Polymerase chain reaction products were purified using an EasyPure PCR purification Kit (TransGen) and then sent to Beijing Genomic Institute for sequencing. When multiple peaks appeared, the products were cloned in the pEasy-T5 vector (TransGen), and 3-7 positive clones were picked for sequencing.

Raw Sequence Treatments
Sequence homology analysis was performed with the BLAST program in NCBI Web. Haplotypes are defined as having greater than 1.5% nucleotide diversity in the orf7 gene (Chafee et al., 2010). Wolbachia strains that were not infected by phage WO were not included in our study in order to improve readability. If multiple identical sequences were obtained from each species, we chose only one to represent the species. We reserved different sequences and removed identical sequences, yielding 34 wsp sequences and 12 orf7 sequences. The sequences have been deposited in GenBank under the following accession numbers: KT355405-KT355450.

Phylogenetic Analyses
The wsp and orf7 sequences were aligned to relevant sequences previously published on NCBI with Clustal W in BioEdit (Hall, 1999). Maximum likelihood (ML) was carried out to construct the phylogenetic tree using MEGA 6 (Tamura et al., 2013). Model selection for the ML analysis was estimated using the Akaike information criterion in Modeltest v3.7. The DNA substitution model was the general time reversible (GTR) model in which the gamma distribution and invariant sites were estimated from the data (GTR+I+G).
ML bootstrap values were generated from 1000 bootstrap replicates.

Wolbachia Infection Patterns in Fig Wasps
We screened a total of 1406 wasps of 23 species for Wolbachia and WO infection. The results are listed in Table 1 (Yang et al., 2012). After removing repeated identical sequences within each species, 34 wsp haplotypes from 19 wasp species were obtained. Besides five common Wolbachia strains (wHaw, wMel, wUni, wMors, and wCon), three other Wolbachia strains were first detected in fig wasps, named wFig-1, wFig-2 and wFig-3, respectively (Figure 1)

WO Infection Patterns Among Wolbachia Strains Associated with Fig Wasps
No bacteriophage WO was detected in fig wasps that were not infected with Wolbachia, as has been reported in previous surveys (Fujii et al., 2004;Braquart-Varnier et al., 2005;Sanogo et al., 2005;Gavotte et al., 2007

Bacteriophage WO Phylogeny
The

Phage and Wolbachia Horizontal Transfer
Comparisons of bacteriophage WO phylogeny, based on the orf7 gene, and Wolbachia phylogeny, based on the wsp gene, were performed for 13 species infected by both WO and Wolbachia (Figure 3). Obviously, no congruence was found between phage WO and its host Wolbachia phylogenies, which indicated that phages do not cospeciate with their hosts. Moreover, there was also no congruence between phylogenies between phage WO and fig wasps (Figure 4), as well as between Wolbachia and its fig wasp hosts (Figure 5). No phylogenetic correlation was found between phages and bacteria, bacteria and insect hosts, or phages and insect hosts, as in some reported insects from open and semi-open environments (Masui et al., 2000;Bordenstein and Wernegreen, 2004;Gavotte et al., 2007). The absence of an evolutionary correlation between WO and Wolbachia and between WO and insect host phylogenies indicated that WO can be transferred horizontally by itself between different Wolbachia endosymbionts or even insect hosts.
In addition to this lack of congruence between WO and Wolbachia, we found other evidence of phage WO horizontal transfer. Distant phylogenetically related Wolbachia strains (supergroup A and B) shared the same WO type. For example,   Walkerella benjamini, which was double infected by Wolbachia strains from different supergroups, harbored only one WO type. Moreover, different insect hosts from different figs (Ficus auriculata and F. benjamina), were found to harbor only one phage type.

DISCUSSION
Host-microbe-phage symbiosis comprises some of the most intimate and long-lasting associations on the planet. In order to determine the relationships between each pair of organisms, it is necessary to first determine the patterns of infection. In this study, we examined the relationships among Wolbachia and phage WO in fig wasp species within a closed system. Our data provided insights into the complex relationships of this symbiosis.
In this study, we report the detailed analysis and detection of Wolbachia and phage WO in 23 fig wasps from three different compact syconia. This system is notable because it is considered a closed community, different from other insects living in open or semi-open environments. In our study, the Wolbachia infection rate (83%) in our tested fig wasps was much higher than fig wasps in Australia (67%) and Panama (59%) (Shoemaker et al., 2002;Haine and Cook, 2005), which showed that fig wasps may have the highest known incidences of Wolbachia amongst all insects (Haine and Cook, 2005). As for bacteriophage WO, the infection rate in our tested fig wasps is 39%, which was unexpectedly lower than that in previous reports 89% (Chauvatcharin et al., 2006;Gavotte et al., 2007). The low WO infection rate (20%, 3/15) of fig wasps from F. benjamina and the high representation of fig wasps from F. benjamina in these samples (15/23) may contribute to the overall infection pattern. Interestingly, despite living in the same syconia, most Wolbachia-infected species did not harbor phage WO in F. benjamina. However, whether these fig wasps were infected previously or have never been infected is unclear. We suggest three possible explanations for this phenomenon. Firstly, the method used to screen for the presence of phage WO may not have detected all phage WO. Although several primer pairs were used to check the infection patterns repeatedly, the orf7F/R primers were not sufficiently degenerate to detect all orf7 variants. Previously, only a single WO haplotype was detected in D. simulans infected with Wolbachia strain wRi (Gavotte et al., 2007). However, the genome sequence confirmed four prophage copies in the genome (Klasson et al., 2009). Moreover, Metcalf proposed that single-gene PCR should not be used to rule out the presence of phage WO in Wolbachia (Metcalf et al., 2014). Multiple infections are common in Wolbachia; however, six out of nine WO-infected species harbored only one phage type, which was confirmed by direct sequencing of the PCR product of orf7 sequences, even when wasps were sampled from different syconia and different fig trees, indicating that they were not occasional infection events. In addition, the complete genome sequence of prophage WO (WOSol) in Wolbachia strain wSol, which infected the fig wasps Ceratosolen solmsi, also showed the presence of only one WO phage in the wSol genome (Wang et al., 2013). Furthermore, a survey of phage WO showed that 85% (28/34) of Wolbachia strains harbor only one or two different WO types (Gavotte et al., 2007).
Wolbachia spreads across hosts through both vertical and horizontal transfer. Vertical transmission is thought to be the predominant mode of Wolbachia transmission within a host , and horizontal transfer of Wolbachia has also been detected both within and among different host species in many cases (Baldo et al., 2006Frost et al., 2010). Considerable horizontal transfer of Wolbachia has been detected within F. benjamina, and researchers have proposed that the syconium may provide a platform for horizontal transfer (Yang et al., 2012). We confirmed this by analyzing the discordance between the phylogenies of Wolbachia and fig wasps from different figs (Figure 5). Similarly, we found abundant horizontal transfer of phage WO in Wolbachia associated with fig wasps, although the exact horizontal transfer routes were uncertain. From our data and previous studies, there may be several possible mechanisms for such transfer. First, we observed multiple horizontal transfer of phage WO of Wolbachia in fig wasps from two pieces of evidence: 1) there was phylogenetic discordance between Wolbachia and phage WO (Figures 2 and  3) based on the orf7 sequences, the same phage type was shared by phylogenetically distant Wolbachia strains of fig wasps from the same and different syconia. Thus, the horizontal transfer of both Wolbachia and WO was common in figs. Second, the horizontal transfer of WO was completely independent of the horizontal transfer of Wolbachia. If the horizontal transfer of WO depends on the transfer of Wolbachia, then the insects infected by the same Wolbachia strains would harbor the same WO types, and congruence between the two phylogenies would be found. However, this was not the case in our data. Instead, we found that WO could be successfully transferred by itself horizontally without its bacterial host between Wolbachia endosymbionts or insects. This has also been confirmed in other studies (Chafee et al., 2010). This independence may be possible because of the WO lysozymes, which could lyse bacterial cell walls (Fischetti, 2010). Third, the horizontal transfer of Wolbachia may facilitate the horizontal transfer of WO. The enclosed syconium provided a platform for horizontal transfer of the endosymbiont Wolbachia; this may also be beneficial for the transfer of the phage to some extent. Each of these potential mechanisms may contribute to the phage transfer.
Recently, genome sequence data shows that some mobile elements are present at sometimes high frequency in obligate intracellular bacteria, including Rickettsia, Phytoplasma, and Wolbachia (Cordaux et al., 2008;Baldridge et al., 2010;Chung et al., 2013). Bacteriophage WO is the only element that has been shown to move horizontally in obligate bacterial endosymbiont Wolbachia (Bordenstein and Wernegreen, 2004). WO is a dynamic element that has a marked effect on the genetic diversity of Wolbachia and could explain some of the interactions with Wolbachia genes and factors, without participating directly in the reproductive manipulations of Wolbachia induced in arthropods (Sanogo et al., 2005;Gavotte et al., 2007).

CONCLUSION
We provided important insights into the horizontal transfer and interactions of fig wasps, Wolbachia, and phage WO in enclosed syconia. Some evidence of how bacteriophage WO work on Wolbachia or even insect hosts is another focus of our following research. This may facilitate the use of WO DNA-delivery vectors as tools for genetic manipulation of insects in the future.

AUTHOR CONTRIBUTIONS
NXW and DWH designed the study. NXW, SSJ, and HX performed the analyses. SSJ performed experiments. NXW, SSJ, YL, and DWH wrote the manuscript. All authors revised the manuscript and approved the final version.