Symbiotic Wolbachia bacteria in coccinellid parasitoids: genetic diversity, horizontal transfer, and recombination

Parasitoids, which constitute about 25% of all insects and attack arthropods of virtually all taxa, are considered the most suitable vectors for horizontal transmission of the symbiotic bacterium Wolbachia among insects. The parasitoids studied in this article develop in the larvae and pupae of ladybirds. For the first time, Wolbachia was found in parasitic wasp species of the genus Homalotylus (Hymenoptera: Encyrtidae) and from the subfamily Tetrastichinae (Hymenoptera: Eulophidae). To characterize the Wolbachia strains, six bacterial housekeeping genes were examined and compared with previously published Wolbachia gene sequences. The same bacterial strains were found in all individuals of each species of parasitic wasps collected in different places and at different times, which indicates the absence of contamination and testifies to the heritability of the symbionts in the studied chalcids. No evidence was found that the parasitic wasps were infected with Wolbachia, identical to the symbionts of their ladybirds hosts. We found one Wolbachia strain, wHom-2, which is a product of bacterial recombination from unrelated insects, including ladybirds. The lack of correspondence between the molecular phylogenies of Wolbachia strains and mitochondrial DNA of their hosts indicates horizontal transfers of Wolbachia among parasitic wasps of the genus Homalotylus and from the subfamily Tetrastichinae.


Introduction
Wolbachia, a pandemic endosymbiont, is widespread among arthropods and affects the viability and reproduction of its hosts (Werren et al. 1995). For this reason, interactions between Wolbachia bacteria and their invertebrate hosts can serve as important drivers of competition and speciation in insects. Bacteria of the Wolbachia genus are classified into different strains within 17 phylogenetic supergroups based on phylogenetic approaches (Kaur et al. 2021). In general, these insect symbionts undergo vertical transmission from mothers to their offspring through the egg cytoplasm. As the number of Wolbachia infections studied has increased, cases of horizontal transfer have been foundidentical strains among unrelated organisms and different strains in the same species or even individuals (Vavre et al. 1999;Ahmed et al. 2016;Hou et al. 2020). Wolbachia bacteria are able to pass and survive the extracellular phase for successful horizontal transfer between organisms. It has been shown that Wolbachia bacteria can survive outside cells without reducing viability for up to 1 week, after which they were able to penetrate the cell, multiply, and create stable infections (Rasgon et al. 2006). Wolbachia can also migrate between tissues of filarial nematodes and thus has not only cellular "internal" but also "external" lifestyle (Slatko et al. 2014). The ability of Wolbachia bacteria to survive outside their host cells increases the likelihood of successful horizontal transfer of the bacteria between insect species.
A common food, especially host plants or fungi, is one of the modes of horizontal Wolbachia spread (Li et al. 2016). Another routes of horizontal transfer are predation, parasitism, and exchange of hemolymph by physical contact (Ahmed et al. 2015). Parasitoids, which constitute about 25% of all insects and attack arthropods of almost all taxa, are considered the most likely vectors for horizontal transmission of Wolbachia among insects (Werren et al. 1995).
Horizontal transmission of Wolbachia was first proven from an infected host (Drosophila simulans) to a parasitic wasp (Leptopilina boulardi), the symbiont being inherited in this new host species (Heath et al. 1999). Later, other evidences for horizontal symbiont transmission via parasitoids have been published (Huigens et al. 2004;Duron et al. 2010;Yang et al. 2013;Ahmed et al. 2015;Johannesen 2017;Qi et al. 2019). Most parasitoids develop within the body of their insect host (endoparasitism). It has been shown that in such a closed system, parasitoids are infected by their hosts' microbes (Duron et al. 2010). Moreover, multiple parasitism, when the same insect host is used by females of different parasitoid species to lay eggs, creates an opportunity for symbiotic transmission between different parasitoid species (Duron et al. 2010). Some species of parasitic wasps (Aprostocetus minutus, Catolaccus spp., Chelloneurus spp., Dendrocerus spp., Homalotyloidea dahibomii, Ooencyrtus spp., Pachyneuron spp., Prochiloneurus aegyptiacus) are hyperparasites (i.e., parasites of parasites) that are known to parasitize on Homalotylus spp. and Tetrastichinae spp. (LaSalle 1994;Ceryngier et al. 2012), thus being able to obtain bacteria from them and vice versa. Some hyperparasitoids, e.g., A. minutus, are found in hosts belonging to different orders: Hemiptera (Aphididae, Coccidae, Pseudococcidae), Coleoptera (Coccinellidae), and Hymenoptera (Aphelinidae, Braconidae, Encyrtidae, Pteromalidae) (LaSalle 1994). Although members of Hemiptera and Coleoptera are only the secondary hosts of A. minutus (i.e., larvae of A. minutus develop in larvae of other parasitic wasps, which in turn parasitize ladybirds, aphids, and coccids), they can also facilitate the transmission of symbiotic bacteria from species to species. Clearly, these modes of transmission can create new infections. The study of cases of the horizontal transfer of bacteria is important to determine the rate and manner in which new symbioses are initiated (Engelstadter and Hurst 2006).
Twelve coccinellid species from 10 genera are known to be infected with Wolbachia to date, as confirmed by the study of usually one bacterial gene, either 16S rDNA or wsp or ftsZ (Kajtoch and Kotaskova 2018). Wolbachia strain characterization by multilocus sequence typing was made only for symbionts from two ladybird species: Harmonia axyridis (Goryacheva et al. 2015) and Adalia bipunctata (Shaikevich et al. 2021). The harlequin ladybird Harmonia axyridis has previously been found to be infected with two strains of Wolbachia (Goryacheva et al. 2015). Two-spot ladybird Adalia bipunctata is infected by three different Wolbachia strains (Shaikevich et al. 2021), with one of them persisting in the population in Moscow for a long time; the others were detected for the first time in 2019. In search of the sources of the newly detected bacterial strains, in this paper, we investigated parasitic insects of ladybirds.
Three hypotheses were suggested: (1) Ladybirds are infected with Wolbachia through parasitoids, (2) parasitoids are infected by Wolbachia which are symbionts of ladybirds, and (3) Wolbachia strains in parasitoids and ladybirds are unrelated and the Wolbachia infection occurred independently. To test these hypotheses, the infection with Wolbachia in parasitoids emerging from larvae (Homalotylus spp.) and pupae (Phalacrotophora spp., Tetrastichinae spp.) of ladybirds and the genetic diversity of the symbiotic bacterium were determined by commonly accepted multilocus analysis. We compared gene sequences with all known strains and performed a phylogenetic analysis of Wolbachia strains to elucidate the role of parasitoids in the distribution of symbionts among insects. As far as known, there have been no studies of bacterial exchanges between coccinellids and their parasitoids.

Collection of parasitoid specimens
Ladybirds at the fourth-instar larval and pupal stages of development were collected by visual inspection of trees and shrubs at heights of 0.5 to 2.0 m. Coccinellid larval species were identified according to the identification table (Savoiskaya 1983). A total of 1157 fourth-instar larvae and pupae of coccinellids were individually placed in 40-mm Petri dishes until adult beetles or parasitoids emerged from 117 of them. The parasitic flies and wasps were obtained from the pupae of ladybirds collected occasionally in 2017-2019 and purposely in 2020 (Table 1). Coccinellid parasitoid species were identified by nucleotide sequences of the fragment of the cox1 gene (mitochondrial DNA) and by ecological criteria (Noyes 2010;Trjapitzin 2013). Since these are gregarious parasitoids that lay multiple eggs in a single larva, we only needed to sequence DNA from one sample from each offspring to clarify the species.

Identification by molecular methods
Total DNA was extracted from at least one individual parasitoids from each host pupa and collection site by phenol-chloroform extraction according to the standard protocol (Sambrook et al. 1989). The amplification reaction with each DNA sample was performed in a volume of 25 µl using the universal Encyclo Plus PCR kit (Eurogen, Moscow) according to the manufacturer's protocol. In all cases, the corresponding negative and positive PCR controls have been carried out. Universal primers LCO1490 and HCO2198 were used to amplify the cox1 gene region (Folmer et al 1994). The new Homol-F primer developed for this work (5′-ATA ATT ATT CGT TTG GAA TTAGG-3′) in combination with HCO2198 primer was used to amplify the cox1 gene from Homalotylus spp. All PCR reactions were performed on a MiniAmp Plus amplifier (Applied Biosystems, USA). Amplification conditions were as follows: initial denaturation for 4 min 30 s at 94 °C, followed by 5 cycles (denaturation for 30 s at 94 °C, annealing for 20 s at 45 °C, and polymerization for 1 min at 72 °C) and then 35 cycles (denaturation for 30 s at 94 °C, annealing for 20 s at 55 °C, and polymerization for 1 min at 72 °C. PCR was terminated by final polymerization for 5 min at 72 °C. Wolbachia infection in parasitoids and ladybirds was first tested by PCR with primers for the fbpA gene. Five conserved Wolbachia genes, gatB, coxA, hcpA, fbpA, and ftsZ, and the highly variable wsp gene were amplified for positive samples according to MLST analysis (Baldo et al. 2006). PCR results were analyzed by electrophoresis in 1.5% agarose gel. The amplified PCR fragments corresponding to the cox1 of the parasitoid species and six different Wolbachia genes were sequenced.

Data analysis
Nucleotide sequence chromatograms were analyzed using the DNASTAR Lasergene 6 software package (Clewley 1995;Burland 2000). The International Barcode of Life Database (BOLD) and GenBank were used to identify insect species by comparing the sequences obtained with those already known. The Wolbachia loci were compared in the database http:// pubml st. org/ wolba chia and closely related sequence types (ST) were chosen. The pairwise evolutionary divergence distances between Wolbachia species were estimated with both the five MLST genes and the concatenated sequence of these five MLST genes by using the Maximum Composite Likelihood model in MEGA v.6 (Tamura et al. 2013). A phylogenetic dendrogram based on cox1 gene was constructed in MEGA v.6 (Tamura et al. 2013) by applying the Maximum Likelihood method (ML) based on the Tamura-Nei model; statistical branch support was based on 1000 bootstrap iterations. To analyze phylogenetic relationships based on five concatenated Wolbachia alleles, two different methods were used: ML estimation using MEGA v.6 and Bayesian inferences using the MrBayes in TOPALi v2.5 (Milne et al. 2004). Bayesian tree was calculated using GTR model and were run for two million generations with a burning of 50%. The phylogenetic trees' topology obtained with the two methods were identical. Newly obtained cox1 gene sequences were registered in GenBank under the numbers (OL889887-OL889908, OL940937-OL940939) and Wolbachia genes (OL906290-OL906294).

Identification of parasitoids
Ten percent of all coccinellids collected were infected with parasitoids (Table 1). We applied the barcoding method for species identification of parasitic flies and wasps. The similarity of the obtained cox1 gene nucleotide sequences from Phalacrotophora flies to those already known is 99-100% for Phalacrotophora fasciata (OL889890-OL889900) and P. berolinensis (OL889887-OL889889). P. fasciata fly larvae were obtained from pupae of 6 ladybird species: A. bipunctata, A. decempunctata, C. septempunctata, H. axyridis, O. conglobata, and C. quatuordecimguttata. P. berolinensis was derived from 3 coccinellid species: A. bipunctata, C. septempunctata, and H. axyridis (Table 1). Two species of parasitic wasp Homalotylus spp. were derived from the larvae of 3 ladybird species: Chilocorus renipustulatus and Calvia quatuordecimguttata (Homalotylus sp. 1) and Coccinella septempunctata (Homalotylus sp. 2) ( Table 1). The cox1 sequences do not allow accurate identification of these species because the published cox1 sequences of the closest relative, Homalotylus terminalis (MH979998), which occurs in the Americas, is 94% similar to Homalotylus sp. 1 (OL940937-OL940938) and 92% similar to Homalotylus sp. 2 (OL940939). These parasitic wasps are assigned to separate species of the genus Homalotylus, as the difference between them in the analyzed cox1 gene region is 6% (Fig. 1). According to published data, two wasp species of the genus Homalotylus, namely, H. hemipterinus and H. eytelweini, parasitize on ladybirds belonging to the tribes Coccinellini and Chilocorini in Europe (Noyes 2010;Trjapitzin 2013). The GenBank and BOLD databases do not contain the cox1 sequences of these species, making it impossible to determine which species were developed from C. septempunctata and which from C. renipustulatus and C. quatuordecimguttata.
Oomyzus scaposus wasps were derived from pupae of H. axyridis and C. septempunctata (Table 1). Their cox1 sequences (OL889901-OL889902) were 99% similar to those already known for parasitic wasps of this species (Fig. 1). The pupae of ladybirds A. bipunctata and A. decempunctata were infected by another wasp species from the subfamily Tetrastichinae (Fig. 1). The nucleotide sequences of the cox1 gene (OL889903-OL889908) do not unambiguously identify this species: the maximum similarity value was 97-98% with the cox1 sequence of Aprostocetus sp. (BIN: AEJ2350) (https:// www. bolds ystems. org). The difference in cox1 sequences in the representatives of Tetrastichinae sp. studied reaches 2.5% (Fig. 1). According to ecological data (Klausnitzer 1969;Ceryngier et al. 2012;Romanov and Matveikina 2022), this may be Aprostocetus neglectus or Tetrastichus epilachnae. Since the cox1 nucleotide sequences of A. neglectus and T. epilachnae are not available in the BOLD and GenBank databases, we cannot make a meaningful comparison. There is also no information on the host species Aprostocetus sp. (BIN: AEJ2350). Because the candidate species belong to different genera, the designation Tetrastichinae sp. will be used here (Table 1). Thus, the following coccinellid parasitoids were identified: Phalacrotophora fasciata, P. berolinensis, Homalotylus spp. (2 species), Tetrastichinae sp. (Aprostocetus neglectus or Tetrastichus epilachnae), and Oomyzus scaposus (Fig. 1).

Identification of Wolbachia in parasitoids
The scuttle flies Phalacrotophora fasciata and P. berolinensis do not have Wolbachia symbiotic bacteria. Parasitic wasps Homalotylus spp. (2 species), O. scaposus, and Tetrastichinae sp. hatched from ladybirds of different genera and species collected from distant locations (Table 1) were found to be infected with Wolbachia.
The sequences of the five bacterial housekeeping genes gatB, coxA, hcpA, ftsZ, and fbpA were determined to identify the Wolbachia isolates (Fig. 2). Five MLST loci of Wolbachia are mapped in the wMel genome in the order gatB-coxA-hcpA-ftsZ-fbpA (Reumer et al. 2010), and so they were arranged in concatenated sequences and in Table 2. Homalotylus sp. 1 emerged from Chilocorus renipustulatus and Calvia quatuordecimguttata were found to be infected with the Wolbachia strain, which was denoted wHom-1. Homalotylus sp. 2 derived from two Coccinella septempunctata larvae infected with another strain, wHom-2. All studied samples of Oomyzus scaposus were infected with the same strain of Wolbachia (wOsc-1) in both Feodosia (host Harmonia axyridis) and Moscow (host Coccinella septempunctata). Tetrastichinae sp. wasps parasitizing both Adalia bipunctata (in Moscow and Yalta) and A. decempunctata (in Moscow) were infected with a single wTet-1 strain (Table 2). Strains wHom-1 and wTet-1 belong to supergroup A; strains wHom-2 and wOsc-1 belong to supergroup B.
Genetically similar Wolbachia sequences were selected after comparing the obtained and all available Wolbachia gene sequences in the MLST database. Isolates with at least two identical alleles were selected. Additionally, all the Wolbachia gene sequences detected were blasted against those available in the GenBank. Genetically similar sequences were selected from the GenBank. Pairwise differences between concatenated sequences within supergroup B are 0-2.5%, except for strain wOsc-1, which differs from all known strains by 3.5-5.5%. Pairwise differences between concatenated sequences within supergroup A are 0.1-0.3%. Differences between supergroups A and B are 12-13%. A set of the sequences shown in Fig. 2 and Table 2 were thus selected. The geographic distribution of the selected Wolbachia strains is shown in Fig. 3.

Phylogenetic relationships of Wolbachia from supergroup A
Two closely related strains of the bacterium were found based on sequence comparisons of five Wolbachia genes: wTet-1 in Tetrastichinae sp. and wHom-1 in Homalotylus sp. 1, differing only by one G376A substitution in the coxA gene. The wTet-1 and wHom-1 sequences cluster with bacterial strains wBeva_A, wMspi_A_wspe (ST524), and wSfur_A_ YN4 (ST234) from other insects (Fig. 2). The strains in this group differ in one (0.05%) or three (0.14%) SNPs among 2079 bp concatenated sequences. The wBeva_A strain differs by one SNP in the gatB gene (G349A) from wTet-1 and two from wHom-1. The wSfur_A_YN4 strain (ST234) differs from wTet-1 by three mutations, one SNP each in the gatB, coxA, and fbpA genes. The strain wMspi_A_wspe (ST524) differs from wTet-1 in two SNPs in the gatB gene and one SNP in the fbpA gene. The pairwise differences between the sequences of single genes are shown in Table 3.
Additionally, the sequences of the hypervariable wsp gene were determined in both specimens of Homalotylus sp. 1; their sequences are identical. The identical wsp-301 allele was found in Wolbachia from Homalotylus sp. 1 (strain wHom-1), as well as from Bicyclus evadne (wBeva_A) (KY658542) and from Megastigmus sp. (ST524) (Tables 2 and 3). No wsp gene sequences were obtained for O. scaposus and Tetrastichinae sp., and there is no wsp gene sequence for wSfur_A_YN4 strain (ST234) in the PubMLST database.

Phylogenetic relationships of Wolbachia from supergroup B
O. scaposus from H. axyridis in Feodosia and from C. septempunctata in Moscow (Table 1) are infected with the identical and unique Wolbachia strain (Fig. 2). Among the PubMLST-Wolbachia data, there are no strains matching it for three or more genes. Two genes, ftsZ and fbpA, are identical in Wolbachia from O. scaposus and ST360 from the butterfly Hipparchia autonoe (Table 2). However, a comparison of the sequences of all five genes showed that Wolbachia from O. scaposus and H. autonoe differed significantly (> 3%).
Identical Wolbachia gene sequences were found in Homalotylus sp. 2 from two Coccinella septempunctata larvae (Table 1). Two unique nucleotide substitutions C37T and T39G in the gatB gene distinguish the wHom-2 strain from the previously known gatB-9 allele of Wolbachia (Fig. 4). Apart from these two SNPs, the sequences of five genes are identical in wHom-2 and strain ST-522 (Table 3). No information on the host species identity of ST-522 is available in the PubMLST Wolbachia database.
Sequences of the Wolbachia wsp gene in wHom-2 cluster with the wsp from wAbi-3, H. axyridis_kl-34,   (Table 3). The variability of the wsp gene in this group is 1.9-3%. The wsp gene sequences in the group of strains ST-302 differ significantly (9%) from those in the groups of wHom-2, . The wsp gene sequences are not known for all Wolbachia strains analyzed by MLST and are presented in Fig. 2. The available alleles are presented in Table 3.
In the Wolbachia wMeg (CP021120) genome, genes are mapped in the order ftsZ, wsp, fbpA, hcpA, coxA, and gatB. The distance between coxA-gatB is 5.6 Kb in the wPel (AM999887) and wMeg (CP021120) genomes from Wolbachia supergroup B. In both genomes, the chromosome region comprising hcpA-coxA-gatB genes is 273 Kb and 208 Kb, respectively. The wsp-fbpA genes span 48 Kb. The ftsZ genes in both genomes are mapped about 500 Kb from both the wsp and gatB genes. Based on these data, it can be assumed that recombination of a single fbpA gene or wsp-fbpA chromosome segment (< 50 Kb) from bacteria belonging to the group of strains  into wHom-2 has occurred, followed by the emergence of mutations in the faster-evolving wsp gene. Recombination of a long fragment of hcpA-coxA-gatB (greater than 200 Kb) or two consecutive recombination events of gatB-coxA (5.6 Kb) and hcpA sites separately

Discussion
The diversity of the symbiotic bacterium Wolbachia was studied in ladybird parasitoids belonging to two different families. Wolbachia was found for the first time in two wasp species of the genus Homalotylus; the bacteria belong to different supergroups A and B. Wolbachia was also found in two species of the subfamily Tetrastichinae: Oomyzus scaposus (supergroup B) and in an unidentified species (Aprostocetus neglectus or Tetrastichus epilachnae) (supergroup A). Among the species of this subfamily, Wolbachia has previously been reported from Tetrastichus coeruleus (Reumer et al. 2010), which parasitizes the asparagus cracker Crioceris asparagi (Coleoptera: Chrysomelidae) (LaSalle 1994), but the sequences of the strain obtained differ from those of the T. coeruleus symbiont (Table 2). Wolbachia was also found in Aprostocetus sp. parasitizing on Ericerus pela coccids (Hemiptera: Coccidae) (Yang et al. 2011), but it was  Table 1 for the Wolbachia strains found in this work and in GenBank or the MLST Wolbachia database for other STs impossible to make a comparison with Wolbachia from Aprostocetus sp. as the authors did not do multilocus typing of the bacterium. The parasitoids we studied are infected with strains related to Wolbachia from other parasitoids, as well as to symbionts of genetically unrelated and ecologically disconnected insects. Closely related Wolbachia supergroup A strains were found in two species from different families: Homalotylus sp. 1 (Hymenoptera: Encyrtidae) and Tetrastichinae sp. (Hymenoptera: Eulophidae). Moreover, Wolbachia strains are identical in the studied individuals found in Moscow and on the Black Sea coast in both taxa. Other parasitoids are also present in this group. Of those previously known, genetically related sequences of Wolbachia have been found: ST524 in the chalcid Megastigmus sp. (Hymenoptera: Torymidae) from Hunan province, China (http:// pubml st. org/ Wolba chia), ST234 in the cicada Sogatella furcifera (Hemiptera: Delphacidae) from Yunan, China (http:// pubml st. org/ Wolba chia), and wBeva-A in the butterfly Bicyclus evadne (Lepidoptera: Nymphalidae) from Africa (Duplouy and Brattström 2018). All these insects have been collected from geographically distant locations, which precludes direct contact between them. The molecular markers used (i.e., the MLST typing scheme) are the five conserved housekeeping genes of the bacterium; hence, such data can be used to study long-term evolutionary equilibrium in the Wolbachia-host system (Baldo et al. 2006;Klopfstein et al. 2018;Duplouy et al. 2020). This could explain why identical or very closely related strains were found among insects collected from geographically distant locations. However, the sequences of the hypervariable wsp gene (Baldo et al. 2006) are also identical in these Wolbachia strains. The detection of closely related strains in both Homalotylus sp. 1 and Tetrastichinae sp. in Russia and in Megastigmus sp., Sogatella furcifera in China, and Bicyclus evadne in Africa indicate the long history of this strain. However, more data on Wolbachia diversity in host-parasitoid groups are required to show whether this is a simple coincidence or whether the strain is extremely widespread in the insect world.
The analyzed sequences in supergroup B are grouped into three phylogenetic clusters. The first genetic cluster in supergroup B includes closely related Wolbachia strains from butterflies and ladybirds. This is not the first time that butterfly and beetle symbionts have been observed to be similar. Previously, another Wolbachia strain (ST19) was found sheared among other butterfly and beetle species collected in Asia, Europe, and North America (Vavre et al. 1999). Wolbachia of several parasitoids, Hyposoter horticola (Hymenoptera: Ichneumonidae), Cotesia chilonis (Hymenoptera: Braconidae, Microgastrinae), and Nasonia vitripennis (Hymenoptera: Pteromalidae) were also included in this cluster. The Wolbachia strain found in Homalotylus sp. 2 (wHom-2) parasitizing on various ladybird species differs from those known from H. horticola, C. chilonis, and N. vitripennis. The second cluster in supergroup B forms sequences from other butterfly species. Wolbachia from Oomyzus scaposus differs from all known bacteria and forms the third cluster.
The mosaic structure of the Wolbachia wHom-2 genome consists of alleles previously found in butterflies and in beetles belonging to different clusters, first and second. All butterflies and ladybirds in which these Wolbachia alleles have been found have overlapping ranges in the Palearctic (Konstantinov et al. 2009). However, Homalotylus spp. do not parasitize butterflies (Ceryngier et al. 2012), and it is still difficult to determine the route of transmission of the symbiont from butterflies. We hypothesized the recombination origin of wHom-2 strain because its fbpA gene is identical to one bacterial strain, and the hcpA gene to another bacterial strain. It is difficult to assume that six mutations in the hcpA gene or 40 mutations in the fbpA gene occurred Wolbachia strain wHom-2 is underlined. Similar color-coded nucleotides highlight sequences recombinant between different strains (see Results) simultaneously in bacteria in different hosts. So, coincident mutations must be ruled out. The identity of most of the genes and the differences in one indicate recombination at a given site rather than horizontal transfer of the entire Wolbachia (Baldo et al. 2006). It seems more likely that the parasitic wasps became infected with two strains of Wolbachia and bacteria recombined to a new strain. In this case, one strain has been existed in the body, and the second came with food. Given the specialization of the Homalotylus sp. on ladybirds, the fbpA-wsp region most likely originates from the beetle. Recombinations between Wolbachia genes were previously discovered in Megastigmus sp. in chestnut gall wasp-parasitoid communities (Hou et al. 2020).
The second explanation could be that such a combination of genes already exists in organisms for a long time (ancestral infection) and we just discovered it for the first time in wasps. The detection of identical sequences in wHom-2 and ST522 may indicate the long-term existence of such a strain. Unfortunately, nothing is known from this strain. Phylogenetically closely related strains of maternally inherited endosymbiotic bacteria are often found in phylogenetically divergent and geographically distant insect host species (Duplouy et al. 2020). Our results are in agreement with those of Duplouy et al. (2020) that only systematic studies of Wolbachia strains from different species of insect communities, not just individual species or host-parasitoid pairs, could explain in the future how a single Wolbachia strain can occur in ecologically unrelated hosts and distant geographical regions.
Inconsistency between the phylogeny of Wolbachia strains (wHom-1, wTet-1) and those of their hosts (Homalotylus sp. 1 and Tetrastichinae sp.) can be explained by horizontal transfer of Wolbachia between species. This may be considered because these parasites coexist on the same territory (Fig. 3) and have ecological connections. Several variants of parasitoid infection can be assumed: If the infection occurred at the adult stage, the source of common Wolbachia strains may be some flowering plants or possibly aphids, as adults of Homalotylus spp. and Tetrastichinae spp. feed on nectar, sometimes on sweet secretions from aphids (Ceryngier et al. 2012). If the parasitoids were infected at the larval stage, the source of Wolbachia could have been a host (ladybird) whose Wolbachia infestation status is not yet known. The possibility of symbiotic transmission between different parasitoid species has been shown to exist when the same insect host is used by females of different parasitoid species to lay eggs (Duron et al. 2010). The source of Wolbachia may also be the larvae of Tetrastichinae sp. if we assume that, in the case of multiple parasitism, the larvae of Homalotylus sp. may have eaten their competitors. It could be an unknown hyperparasitoid.
Identical Wolbachia strains have not been found among parasitoids and their coccinellid hosts. Our results are in favor of the third tested hypothesis: Wolbachia strains in ladybirds and their parasitoids are unrelated, and the Wolbachia infection in ladybirds and their parasitoids occurred independently. The effectiveness of infecting the host from the parasite must depend on the length of time they have been together. The faster the parasite causes the host to die, the less likely it is that horizontal transfer of the bacteria will occur. The scuttle flies Phalacrotophora spp. and the wasps Homalotylus spp. and Tetrastichinae spp. are probably poor vectors because fly larvae eat their host very quickly. The wasp larvae, although feeding longer, quickly cause its death by damaging internal organs.

Conclusions
Thus, we found no evidence that ladybird parasitic wasps were infected with Wolbachia identical to the symbionts of their coccinellid hosts. Identical strains were found in all individuals of each parasitic wasp species collected in different locations and at different time points, thus indicating the absence of contamination and supporting symbiont heritability in the chalcids studied. The Wolbachia strains detected are not identical to those of ladybirds H. axyridis or A. bipunctata, supporting the hypothesis that infestation by symbionts of these parasitoids and these beetle species occurred independently. The Wolbachia wHom-2 strain, which is a recombination product of different bacteria, suggests at least the short-term presence of ladybird's Wolbachia in the wasp's body. The lack of correspondence between the phylogenetic tree topologies of Wolbachia strains and their hosts indicates the horizontal transfer of Wolbachia among parasitic wasps Homalotylus sp. and Tetrastichinae sp. Although no direct confirmation has been found, it cannot be ruled out that the exchange of bacteria takes place through a common coccinellid host or hyperparasitoids, whose symbiotic bacteria are not yet known.
The study of how symbiotic bacteria spread is of particular importance in ecology. Symbiotic bacteria can influence various aspects of host biology, including feeding behavior, sex ratio, parasitism resistance, and heat tolerance. Ultimately, these interactions shape the spatial distribution of insects and their inter-or intraspecific interactions, in some cases speciation. This leads to changes in entire insect communities and broad ecosystems. A comprehensive genetic analysis of Wolbachia alleles in coccinellid parasitoids will help to elucidate the pathways and mechanisms of Wolbachia horizontal transfer in insects.