Narrow Genetic Diversity of Wolbachia Symbionts in Acrididae Grasshopper Hosts (Insecta, Orthoptera)

Bacteria of the Wolbachia genus are maternally inherited symbionts of Nematoda and numerous Arthropoda hosts. There are approximately 20 lineages of Wolbachia, which are called supergroups, and they are designated alphabetically. Wolbachia strains of the supergroups A and B are predominant in arthropods, especially in insects, and supergroup F seems to rank third. Host taxa have been studied very unevenly for Wolbachia symbionts, and here, we turn to one of largely unexplored insect families: Acrididae. On the basis of five genes subject to multilocus sequence typing, we investigated the incidence and genetic diversity of Wolbachia in 41 species belonging three subfamilies (Gomphocerinae, Oedipodinae, and Podisminae) collected in Turkey, Kazakhstan, Tajikistan, Russia, and Japan, making 501 specimens in total. Our results revealed a high incidence and very narrow genetic diversity of Wolbachia. Although only the strains belonging to supergroups A and B are commonly present in present, the Acrididae hosts here proved to be infected with supergroups B and F without A-supergroup variants. The only trace of an A-supergroup lineage was noted in one case of an inter-supergroup recombinant haplotype, where the ftsZ gene came from supergroup A, and the others from supergroup B. Variation in the Wolbachia haplotypes in Acrididae hosts within supergroups B and F was extremely low. A comprehensive genetic analysis of Wolbachia diversity confirmed specific features of the Wolbachia allelic set in Acrididae hosts. This result can help to elucidate the crucial issue of Wolbachia biology: the route(s) and mechanism(s) of Wolbachia horizontal transmission.

Acrididae grasshoppers comprise more than 10,000 species that are widespread in nearly all terrestrial landscapes. Due to the great importance of Acrididae grasshoppers as agricultural pests, much research attention has been given to this taxon.
Here, we evaluated the genetic diversity of Wolbachia in members of the Acrididae family. In particular, we screened specimens from three subfamilies (Gomphocerinae, Oedipodinae, and Podisminae) collected in Turkey, Kazakhstan, Tajikistan, Russia, and Japan and characterised Wolbachia isolates by a multilocus sequence typing (MLST) protocol that included five housekeeping genes [12]. We aimed to investigate Wolbachia incidence in Acrididae hosts and to determine whether there is a Wolbachia genetic pattern that is specific to Acrididae.

Wolbachia Occurrence in Acrididae
We examined the infection status in 41 Acrididae species and found 28 Wolbachia-infected species (Table 1). Only two species have been analysed previously: P. sapporensis [33] and P. pedestris [32]. The symbionts were found in all of the subfamilies under study: Gomphocerinae, Oedipodinae, and Podisminae. There were no noticeable differences in Wolbachia prevalence between the regions and the between specimen collection years. The estimation of infection prevalence was not the aim of our study because we mainly tested somatic tissues that may only reflect the lowest rate boundary. Nevertheless, we registered high Wolbachia prevalence in Ch. biguttulus (174/198), Ch. fallax (16/17), and P. montanus (40/50). High Wolbachia prevalence in P. sapporensis populations was determined based on gonad tissues and was reported earlier [33]; here, we screened an additional 18 specimens for infection and confirmed the previous conclusion. Sample sizes were not sufficient to make firm conclusions about other species.

Genetic Diversity of Wolbachia Isolates
A total of 44 Wolbachia isolates were characterised by the MLST protocol [12]; 43 were unique according to the species-site-year combination ( Table 2). We managed to amplify all five loci for 39 isolates, and for 35 isolates, we obtained unambiguous allele sequences. For many isolates, there were difficulties in obtaining complete MLST profiles. In two cases (ftsZ of i-1 and gatB of i-15), we failed to obtain amplicons, even when using nested PCR. For several isolates, we could not obtain good quality sequences; there were double chromatogram peaks for certain positions or loci, or in some cases, even repeated double peaks in the bulk of a sequence.    We reconstructed ML phylogenetic trees for each locus to examine supergroup clustering ( Figure 1, Supplementary Materials). Alleles of the analysed isolates belong to Wolbachia supergroups B and F. In particular, species belonging to Podisminae only harboured variants from supergroup B, species belonging to Oedipodinae only harboured variants from supergroup F, and species belonging to Gomphocerinae mostly harboured species from super group B with some from supergroup F. Overall, the genetic diversity of the MLST loci appeared rather low (     Only haplotype ST-299 has been reported upon earlier [46], whereas others turned out to be unique because of new alleles or new allele combinations. A new combination of previously known alleles (gatB-9, coxA-133, hcpA-6, ftsZ-106, and fbpA-197) that we designated as ST h ST -4 (see Materials and Methods) was the most frequent in the study population; it was present in seven species of Gomphocerinae and in six species of Podisminae. Haplotypes h ST -1, -2, -3, -9, and -11, which are closely related to h ST -4 (p-distance 0.0005-0.0020, namely, 1-4 mutations), were also revealed in Gomphocerinae and Podisminae species in geographically distant populations. These six haplotypes form an 'h ST -4 group' that includes a constant allele set of ftsZ-106 and fbpA-197 as well as varied but closely related (to each other) alleles of gatB, coxA, and hcpA loci. Other B-supergroup haplotypes were distantly related to h ST -4 (p-distance for ST-299: 0.0058, for h ST -6: 0.0203 and for h ST -10: 0.0323); however, they shared identical alleles with the h ST -4 group. These phenomena can be explained by intra-supergroup recombination.
Wolbachia strains belonging to supergroup F were detected in five species. Haplotypes of complete profiles (h ST -5, -7, and -8; p-distance 0.0024-0.0053) and isolates with incomplete profiles (i-1 and i-15) or with ambiguous sites (i-24 and i-25) were found to be closely related, i.e., genetic diversity was also rather low. A comparative analysis of the haplotype diversity of supergroup F retrieved from the PubMLST database showed that variants of the Acrididae isolates formed a separate cluster ( Figure S1).
To examine the relationship of the Wolbachia variants isolated from Acrididae with other Wolbachia strains, we reconstructed a phylogenetic network (Figure 3, Supplementary Materials). The dataset that was used for this analysis included Wolbachia haplotypes from different hosts that contained at least one allele that was identical or that was closely related to Wolbachia isolates from Acrididae (see Materials and Methods). Most Wolbachia isolates of Acrididae (actually the h ST -4 group) formed a separate bundle in the phylogenetic network without isolates from other hosts. Even i-16, which is an inter-supergroup recombinant, occupied a long branch in this bundle. The remaining Wolbachia haplotype h ST -10 (i-34) manifested a close relationship with isolates from Gryllidae hosts. Of note, i-34 was isolated from the Prumna primnoa population of Sakhalin in 2010, where the h ST -4 (i-33) haplotype was identified as well. Haplotypes of the F supergroup also formed a separate bundle. Only the F-supergroup Wolbachia haplotype ST-448 isolated from Teratodes monticollis (Acrididae: Teratodinae) was noticeably different in the allele set and genetic distance (p-distance 0.0293-0.0313). Therefore, the additional data from the MLST profiles based on allele similarity confirmed that the analysed Acrididae subfamilies have a specific genetic pattern of Wolbachia.
Martinez-Rodriguez and Bella [44] reported the Wolbachia MLST diversity of Ch. parallelus in Spanish and French locations. We retrieved sequences of five MLST loci deposited by [44] in GenBank, combined them with the dataset of our phylogenetic network, excluded redundant portions of sequences, and reconstructed a new phylogenetic network (Figure 4, Supplementary Materials). Half of the Wolbachia variants from Ch. parallelus matched the genetic pattern discovered in our collection for both the B-and F-supergroup haplotypes. The other half represented B-F supergroup recombinants. B-F recombinants were registered by [44]; here, we reported detailed characteristics of some variants (details in Figure 4). It is worth pointing out that the recombination occurred at all possible loci, thereby giving rise to 18 unique haplotypes in total. Martinez-Rodriguez and Bella [44] reported the Wolbachia MLST diversity of Ch. parallelus in Spanish and French locations. We retrieved sequences of five MLST loci deposited by [44] in GenBank, combined them with the dataset of our phylogenetic network, excluded redundant portions of sequences, and reconstructed a new phylogenetic network ( Figure 4, Supplementary Materials). Half of the Wolbachia variants from Ch. parallelus matched the genetic pattern discovered in our collection for both the B-and F-supergroup haplotypes. The other half represented B-F supergroup recombinants. B-F recombinants were registered by [44]; here, we reported detailed characteristics of some variants (details in Figure 4). It is worth pointing out that the recombination occurred at all possible loci, thereby giving rise to 18 unique haplotypes in total.  Branches with isolates from this study are highlighted in yellow, isolates from [44] (Cp) are in brown. The dataset of this figure differs from that in Figure 3 in the following ways: (1) the number of haplotypes was reduced (to decrease phylogenetic noise), (2) the alignment corresponds to haplotype data from [44]. Inter-supergroup recombinant haplotypes are designated as in Figure 3. See original files in Supplementary Materials.

Incomplete MLST Profiles
As mentioned above, for some isolates, we could not obtain an amplicon or unambiguous sequences. Amplicons of three such cases in two isolates of P. sapporensis (i-40 and i-42) were cloned and sequenced. Fifteen clones of an ftsZ amplicon of i-40 represented 11 variants; among them, we regard two variants-ftsZ-106 (two clones) and ftsZ-81 with substitution A 408 (five clones)-as authentic (see the last sentence of this subsec-  Figure 3 in the following ways: (1) the number of haplotypes was reduced (to decrease phylogenetic noise), (2) the alignment corresponds to haplotype data from [44]. Inter-supergroup recombinant haplotypes are designated as in Figure 3. See original files in Supplementary Materials.

Incomplete MLST Profiles
As mentioned above, for some isolates, we could not obtain an amplicon or unambiguous sequences. Amplicons of three such cases in two isolates of P. sapporensis (i-40 and i-42) were cloned and sequenced. Fifteen clones of an ftsZ amplicon of i-40 represented 11 variants; among them, we regard two variants-ftsZ-106 (two clones) and ftsZ-81 with substitution A 408 (five clones)-as authentic (see the last sentence of this subsection) because the former is commonly found in other P. sapporensis isolates, and the latter one was assembled manually after direct amplicon sequencing. The other eight variants represented by unique clones were close to the alleles ftsZ-7, -81, and -106, but they differed by amino acid substitutions, and there was even a stop codon in clone 5. A similar pattern was observed during the screening of the hcpA and fbpA amplicons of i-42. There were seven variants that were closely related to the allele hcpA-142 with 4-7 mutations, including the variant hcpA-142 G 298 A 373 G 393 T 438 , which was manually assembled after direct amplicon sequencing. In the case of fbpA cloning, one variant matched fbpA-197, two differed by one mutation, and the others differed by 13-19 mutations. Therefore, the cloning of the PCR products revealed several variants: ftsZ-106, ftsZ-81 A 408 , and hcpA-142 G 298 A 373 G 393 T 438 in i-40 as well as fbpA-197 and probably one mutant variant of fbpA-197 in i-42. They most likely characterise the Wolbachia genome, whereas other variants quite possibly reflect Taq polymerase artefacts or denote Wolbachia genome segments that became integrated into the host genome.

Discussion
The Acrididae family includes more than 10 thousand species inhabiting different climatic zones from tropical rainforests to the subarctic zone. There is extensive information about the geographical distribution and economic significance of this group of insects and about the factors affecting the regulation of grasshopper population size in natural and anthropogenic landscapes. Nevertheless, knowledge about the symbionts of this group of insects remains meagre. Here, we presented the most detailed survey of Wolbachia diversity in Acrididae to date; 28 species out of 40 appeared to be infected by Wolbachia. On the other hand, this list, together with other studies [23,[32][33][34][35][36][37]44], comprises less than 1% of known Acrididae species. To achieve a more informative picture of Wolbachia genetic diversity in this family, it is necessary to examine other big subfamilies, and special attention should be paid to collections from warm climate zones, where the diversity is especially high. Currently, our results indicate that Wolbachia infection (i) is widespread among three subfamilies of Acrididae grasshoppers, (ii) can reach high prevalence in populations, and (iii) can be detected by means of DNA isolated from somatic tissues.
Our main result is the narrow genetic diversity of Wolbachia variants in Acrididae hosts. While Wolbachia in insects is commonly represented by the strains of supergroups A and B, the Acrididae hosts proved to be infected with supergroups B and F without A-supergroup variants. The only trace of an A-supergroup lineage was found in i-16, where the haplotype h ST -5 includes the ftsZ gene from supergroup A and other genes from the B supergroup. The variation of the B-supergroup haplotypes in Acrididae is especially low, as clearly illustrated by the phylogenetic network (Figure 3), where all the haplotypes included in the analysis showed allelic similarity with grasshopper Wolbachia. Previously, an allele set (with central haplotype ST-41) specific to butterfly hosts was characterised [46]; however, the Wolbachia variation observed here in Acrididae is much lower. The accumulation of data on Wolbachia infection in Acrididae hosts will possibly shed light on other Wolbachia variants, including A-supergroup variants. Nonetheless, it is already obvious that Acrididae hosts have a unique infection profile in the global pattern of the Wolbachia pandemic.
Active horizontal transmission (HT) of a specific set of Wolbachia strains among Acrididae grasshoppers is a reason for the observed narrow diversity. Here, we documented (i) cases (often seen in Wolbachia studies) where the same haplotype is found in different species (Table 2), (ii) findings of identical haplotype profiles in populations of Ognevia longipennis from Japan and Altai Mountains of Russia even though these populations most likely have been geographically isolated for several thousand years, and (iii) distantly related Wolbachia haplotypes within the Sakhalin population of P. primnoa (Sakhalin 2010). There is no clear understanding of the HT mechanism, although some data have been reported [22,25,26,29,47]. Potential vectors of Wolbachia HT are parasitoid wasps, red velvet mites, tachinids, entomophilic nematodes, and horsehair worms (Nematomorpha) that have wider or narrower specificity to Acrididae hosts. The next question is the nature of Wolbachia specificity to hosts. Are Acrididae species only susceptible to certain variants of Wolbachia, or does the susceptibility extend to all encountered variants?
The Wolbachia allelic diversity observed in our study is consistent with the findings about Wolbachia in Ch. parallelus studied on the Perinea peninsula [44]. A comprehensive analysis of the data generated by [44] and our results indicates that Wolbachia variants belonging to supergroups B and F are widespread in Acrididae hosts. Moreover, Ch. parallelus was found to harbour nearly every possible MLST recombinant combination of supergroups B and F (Figure 4). In our work, we were only able to reliably detect one isolate of inter-supergroup recombination. In five isolates, we obtained ambiguous signals in chromatograms. The nature of the problem for three of them (i-6, -15, and -18) was not investigated here, i.e., co-infection with B and F lineages cannot be ruled out. Two other isolates were cloned, and we came to the conclusion that the results can be explained by multi-infection with B-supergroup strains and/or by insertions of Wolbachia genes into the host's nuclear genome. The transfer of Wolbachia genes into a host nuclear genome in insects has been well documented [48][49][50][51][52], e.g., in Orthoptera [42,43].
Another essential issue of Wolbachia diversity is the concept of a bacterial species [53][54][55]. Every Wolbachia supergroup is a species candidate [6,[56][57][58][59]. This observation implies genome specificity, ecology specificity, and 'reproductive isolation', which in bacteria, take the form of a low rate or impossibility of gene exchange between strains of different supergroups/species. Indeed, recombination between strains of supergroups A and B is known to occur, albeit at a low rate [46,57,[60][61][62][63]. Recombination between supergroups F and B has been only detected in Chrysocoris stollii (Hemiptera) [62]; however, in the case of Ch. parallelus, there are 18 haplotypes [44]. Comprehensive genomic analysis of core genes indicates that the F supergroup clusters together with C and D lineages found in nematode hosts [63,64]. The F lineage has retained the genes responsible for homologous recombination (data not shown); we concluded this after the examination of the wCle genome (GenBank accession No.: AP013028). Previously, the evolution of these genes was studied in the genomes belonging to supergroups A, B, C, and D [65] but not in F. These data suggest that gene exchange between the genomes of B and F strains in Ch. parallelus may occur via homologous recombination. If rampant recombination has actually occurred in Ch. parallelus, it casts serious doubt on the idea that supergroups B and F can be considered independent bacterial species.

Screening and Sequencing
A leg of an individual was used for DNA extraction in most cases. This approach is rather convenient because it allows the procedure to be repeated in cases of failed extraction (just take another leg) and to avoid bacterial contamination from the digestive system due to the facultative predation/cannibalism of Orthoptera species. Nonetheless, the use of somatic tissues does not permit a reliable estimation of Wolbachia prevalence in a population. Male gonad tissues were used for the screening of Podisma species collected in Hokkaido (Japan) and Kunashir (Russia), which were partially reported by Bugrov et al. [33]. Here, we added 18 specimens and present full MLST profiles for the Podisma spp. hosts. DNA extraction from each sample was performed in 0.3-0.6 mL of extraction buffer (0.1 M NaCl, 10 mM Tris-HCl (pH8.0), 25 mM EDTA, 0.5% SDS, and 0.1 mg/mL proteinase K) for 2 h at +56 • C, and DNA was then salted out with 0.5 V of 5 M potassium acetate/3 M acetic acid, after which DNA was further precipitated and dissolved in 0.2 mL of double-distillated H 2 O. Next, 1 µL of the DNA solution was used in all polymerase chain reactions (PCRs). The quality of the DNA was checked with universal primers specific to the nuclear gene of 28S rRNA (28sF3633: 5 -TACCGTGAGGGAAAGTTGAAA-3 , and 28sR4076: 5 -AGACTCCTTGGTCCGTGTTT-3 [66]) or to mitochondrial gene CO1 (LCO1490: 5 -GGTCAACAAATCATAAAGATATTGG-3 , and HCO2198: 5 -TAAACTTCAGGGTGACCAAAAAATCA-3 [67]). PCR was conducted using BioMaster HS-Taq PCR (2×) (BiolabMix, Novosibirsk, Russia) or a mix containing 3.0 mM Mg 2+ , 0.6 mM each primer, 1× PCR buffer (16 mM (NH 4 ) 2 SO 4 , 67 mM Tris-HCl pH 8.8 (at 25 • C), and 0.1% of Tween 20), and 1.0 U of Taq polymerase in a total reaction volume of 20 µL. The detection of Wolbachia infection in each specimen was performed by means of at least two MLST loci, as usually done with primer sets coxAF1/R1 and ftsZF1/R1 [12]. All five MLST loci were amplified and sequenced for Wolbachia-positive DNA samples. In cases of a weak amplicon signal or negative PCR results, the nested-PCR approach was employed [68]. External primer sets F2/R2 or F3/R3 were used for gatB, coxA, hcpA, and fbpA loci according to [https://pubmlst.org/organisms/wolbachia-spp/protocol-single-infected, accessed on 12 December 2021], primers ftsZunif1/2 for the ftsZ locus according to [6]; and the inner primers F1/R1 according to [12]. The thermal cycling conditions were as follows: initial denaturation at 95 • C 5 min, followed by 35 cycles of conventional PCR, and 15 + 30 cycles of nested PCR at 95 • C for 15 s, annealing at 55 • C for the MLST primers or at 58 • C for 28S or at 53 • C for CO1 for 40 s, elongation at 72 • C 30 s-1 min, and final elongation for 3 min. In the second round of nested PCR, we added 0.5 µL of the reaction mixture from the first round. The PCR products were visualised by agarose gel (1.0-1.5%) electrophoresis with ethidium bromide. A portion of the amplification reaction mixture was diluted as follows: 2 µL of the amplicon + 18 µL of water; then 1 µL of this solution was treated with 10 U of exonuclease I (New England Biolabs, Ipswich, MA, USA) in the supplied buffer and sequenced using the BrightDye Terminator Cycle Sequencing Kit (Nimagen, Nijmegen, The Netherlands) or BigDye Terminator v3.1 cycle sequencing Kit (Applied Biosystems, Foster City, CA, USA). Three amplicons (ftsZ of i-40; hcpA and fbpA of i-42, which yielded ambiguous sequences) were cloned in the pAL-TA vector (Evrogen, Moscow, Russia) according to the manufacturer's instructions and were sequenced with the M13 primer set. The MLST profiles of the Wolbachia isolates were deposited in the GenBank database under accession numbers MZ816445-MZ816686. Because the MLST database has not been accepting new submissions for some time, here we designated haplotypes with new combinations of alleles or new alleles as 'h ST -Number' ( Table 2).

Evolutionary Analysis
New MLST loci sequences were checked for stop codons, and sequence length was limited according to the MLST protocol for subsequent allele analysis. All of the sequences were used to reconstruct maximum likelihood (ML) phylogenetic trees of each MLST locus. As a supergroup reference, we used the alleles that had been retrieved from the following sequence types ( -448), and id-1707 (ST452)) were added to the ML tree reconstruction. Sequence alignments were generated in the MUSCLE software [69], and a nucleotide substitution model for each dataset was chosen by means of MEGA 6; statistical branch support was based on 1000 bootstrap iterations.
To expand the Wolbachia genetic diversity analysis, we created a dataset that included additional MLST profiles. The MLST profiles were chosen according to allele identity toward variants observed in Acrididae isolates ('one allele criterion', see details in refs. [46,62]. Briefly, we took an ST from the Public Databases for Molecular Typing and Microbial Genome Diversity (PubMLST) [70] if it contained the same allele as any of the loci found in our study. Because the number of alleles in our study was unique, we also included STs with the most closely related alleles. To present the phylogenetic relationships of profiles in this dataset, we reconstructed an unrooted phylogenetic network in SplitsTree4 [71] using the neighbour-net method [72]. In addition, we conducted a comprehensive phylogenetic analysis of the data on Wolbachia diversity discovered in Ch. parallelus [44] and in the above-mentioned dataset. We retrieved alleles from GenBank, assembled concatenated sequences, aligned the sequences, excluded redundant parts, and reconstructed the unrooted phylogenetic network.

Conclusions
Many Acrididae species harbour Wolbachia symbionts. Nonetheless, Wolbachia genetic diversity is rather low among these hosts: (i) there are strains of only supergroups B and F, (ii) genetic variation is narrow within each supergroup. These data indicate mas-sive Wolbachia horizontal transmission among Acrididae hosts. Specific content of Wolbachia alleles in Acrididae hosts can be used for identifying route(s) and mechanism(s) of Wolbachia horizontal transmission.