Sex-Specific Wolbachia Infection Patterns in Populations of Polygraphus proximus Blandford (Coleoptera; Curculionidae: Scolytinae)

Simple Summary Wolbachia bacteria are the most common symbionts of insects. These bacteria are ordinarily transmitted via oocyte cytoplasm from mother to progeny, like mitochondria, and are sporadically transmitted from one species to another. The Wolbachia symbionts have evolved to be parasitic (feminization of genetic males, male-killing, parthenogenesis, and cytoplasmic incompatibility) or/and mutualistic (increasing lifespan and fecundity, providing vitamins and nutrients, defending against viruses and parasites). Here we have studied Wolbachia infection in populations of four-eyed fir bark beetle Polygraphus proximus, which is one of the most dangerous pests of Siberian fir forests. A high rate of the only wProx Wolbachia strain in P. proximus populations was found in a vastly studied territory. Surprisingly, females were more often harboring Wolbachia than males. Besides, a comparison of the Wolbachia density in individuals has revealed that females contain much more Wolbachia symbionts than males. We suppose that the difference in infection status, as well as the difference in Wolbachia load between males and females within a population, can be found in some other Wolbachia–host associations. Abstract Wolbachia symbionts are maternally inherited bacteria that are widely distributed among Arthropoda hosts. Wolbachia influence their host biology in diverse ways. They may induce reproductive abnormalities, protect hosts against pathogens and parasites, or benefit hosts through metabolic provisioning. The progeny of an infected female are ordinarily infected with Wolbachia; however, Wolbachia have no future in male host progeny because they cannot transmit the symbiont to the next generation. Here, we analyze native and invasive populations of the four-eyed fir bark beetle (Polygraphus proximus) for Wolbachia prevalence and symbiont genetic diversity. This species is a dangerous pest of Siberian fir (Abies sibirica) forests. The native range of P. proximus includes the territories of the Russian Far East, Japan, Korea, and Northeast China, whereas its invasive range includes West Siberia, with further expansion westward. Surprisingly, we revealed a difference in the patterns of Wolbachia prevalence for males and females. Infection rate and Wolbachia titers were higher in females than in males. ST-533, the only haplotype of Wolbachia supergroup B, was associated with a minimum of three out of the five described mitochondrial haplotypes.


Introduction
Bacteria of the Wolbachia genus are the most successfully inherited symbionts that inhabit numerous arthropod and nematode hosts [1][2][3]. Wolbachia are vertically transmitted through host generations via oocytes [4], and are sporadically transmitted, horizontally, to non-related hosts [5][6][7][8][9]. These symbionts may affect their host biology in different ways, from manipulating host reproduction to increasing host fitness. Thus, Wolbachia increase lifespan and fecundity [10][11][12], protect against viruses and parasites [13][14][15][16], provide vitamins and other nutrients [17,18], and suppress some mutations in their hosts [19][20][21]. They also influence host reproduction by inducing feminization of genetic males, male killing, parthenogenesis, or cytoplasmic incompatibility (CI) [22]. These phenomena promote the spread of infection in a host population. The first three phenomena lead to a shift in the sex ratio towards females in a host population [22], whereas CI is manifested by embryonic mortality in the progeny of symbiont-free females and Wolbachia-harboring males. Therefore, Wolbachia in males commonly have no future and, in the case of CI, only function in controlling the spread of the symbiont in a host population.
We studied Wolbachia infection in native and invasive populations of the four-eyed fir bark beetle Polygraphus proximus Blandford, 1894 (Coleoptera; Curculionidae: Scolytinae). This species is one of the most dangerous pests of Siberian fir (Abies sibirica) forests. Since 2014, P. proximus has been listed in the European and Mediterranean Plant Protection Organization (EPPO) Alert List (https://gd.eppo.int/taxon/POLGPR). The native range of P. proximus includes the territories of the Russian Far East, Japan, Korea, and Northeast China [23][24][25]. The secondary (invasive) range covers the territories of West Siberia [24,25], with further expansion occurring westward ( Figure 1). reproduction to increasing host fitness. Thus, Wolbachia increase lifespan and fecundity [10][11][12], protect against viruses and parasites [13][14][15][16], provide vitamins and other nutrients [17,18], and suppress some mutations in their hosts [19][20][21]. They also influence host reproduction by inducing feminization of genetic males, male killing, parthenogenesis, or cytoplasmic incompatibility (CI) [22]. These phenomena promote the spread of infection in a host population. The first three phenomena lead to a shift in the sex ratio towards females in a host population [22], whereas CI is manifested by embryonic mortality in the progeny of symbiont-free females and Wolbachia-harboring males. Therefore, Wolbachia in males commonly have no future and, in the case of CI, only function in controlling the spread of the symbiont in a host population.
We studied Wolbachia infection in native and invasive populations of the four-eyed fir bark beetle Polygraphus proximus Blandford, 1894 (Coleoptera; Curculionidae: Scolytinae). This species is one of the most dangerous pests of Siberian fir (Abies sibirica) forests. Since 2014, P. proximus has been listed in the European and Mediterranean Plant Protection Organization (EPPO) Alert List (https://gd.eppo.int/taxon/POLGPR). The native range of P. proximus includes the territories of the Russian Far East, Japan, Korea, and Northeast China [23][24][25]. The secondary (invasive) range covers the territories of West Siberia [24,25], with further expansion occurring westward ( Figure 1).
Here, we first report on Wolbachia in the four-eyed fir bark beetle and determine infection rates of the population, including the infection prevalence according to sex. To do so, we carried out Wolbachia titer by qPCR and genotyped Wolbachia isolates using a multilocus sequence typing (MLST) protocol [31]. In addition, the barcode region of mitochondrial DNA was sequenced for Insects 2020, 11, 547 3 of 12 infected and uninfected individuals collected from different localities, because Wolbachia symbionts and mitochondria are inherited together, according to the host's maternal lineage, and can therefore form cytotypes. A cytotype results from the mitotype and the genotype of an infection, meaning cytotypic analysis can reveal the evolutionary history of Wolbachia infection [6,[32][33][34][35].

Insect Collection
Adult specimens of P. proximus were collected in four regions of Russia: Sakhalin province (native range) (N = 60), Tomsk province (N = 271), Udmurtia (N = 38), and Krasnoyarsk Krai (N = 48) (invasive territories) ( Figure 1 and Table 1). In Tomsk province and Udmurtia, the beetles were collected in their parental nests, whereas in Krasnoyarsk they were collected from forest products. In Sakhalin, overwintered beetles and those intruding into the bark of a host tree were collected. The samples were stored in air-dry conditions or fixed in 96% ethanol. The primary study was designed to analyze the infection status of each individual regardless of beetle sex. However, in the course of the study, we had to change the design to specifically include sexes as defined according to forehead morphology [36]. Finally, 7 out of 13 samples were subdivided by sex, namely one from Sakhalin province (beetles forming pairs), one from Udmurtia, two from Tomsk province, and three from Krasnoyarsk krai.

DNA Extraction
The total DNA was individually extracted from whole specimens. The insects were homogenized in 200 µL of DB extraction buffer (10 mM Tris-HCl (pH 8.0), 25 mM EDTA, 0.5% SDS, 0.1 M NaCl) with 1 µL of proteinase K (AppliChem) and incubated at 56 • C for 1 h. Total DNA was purified with NaOAc, precipitated, and dissolved in 100 µL of double-distilled water.

PCR and Sequencing
All DNA samples were tested by PCR with HCO/LCO primers for the cytochrome c oxidase subunit I (COI) gene [37]. Wolbachia infection status was determined by conventional PCR using the W-SpecF/R primer set [38] and coxAf1/r1 [31], and additionally confirmed for random samples with primer sets hcpAf1/r1 and fbpAf1/r1 [31]. Wolbachia-negative samples and controversial samples were checked by nested PCR with ftsZunif [39] for the first round and ftsZf1/r1 for the second, and/or fbpAf2/r2 and fbpAf1/r1, respectively. The PCR reactions were performed using BioMaster HS-Taq PCR (2×) (BiolabMix, Novosibirsk, Russia) with a 20 µL volume. The thermocycler protocol was 95 • C for 5 min, followed by 35 cycles of conventional PCR, and 15 + 30 cycles of nested PCR at 95 • C for 15 s, 55 • C for 40 s (1 min for HCO/LCO), and 72 • C for 40 s, plus a final elongation step of 72 • C for 3 min. To genotype a Wolbachia isolate, we amplified five MLST genes [31]. The amplicons (COI and MLST genes) were purified with exonuclease (ExoI) enzyme (New England Biolabs), and sequenced using a BrightDye Terminator Cycle Sequencing Kit (Nimagen). Full MLST profiles of four Wolbachia isolates were deposited in the PubMLST (https://pubmlst.org/wolbachia/) database, and all MLST alleles and fragments of the COI gene were deposited in the GenBank database (Table S1).

Quantitative Real-Time PCR Assay
We designed a qPCR assay to measure the relative number of Wolbachia genomes per beetle genome. A single copy gene per Wolbachia genome fbpA, and single copy gene per beetle genome Inx2 were used to estimate the Wolbachia load in DNA samples. For a nuclear reference, the 750 bp region of the Inx2 gene was amplified by nested PCR according to Che et al. [40], and then sequenced for confirmation of the target product (deposited in GenBank MT786531). Primers for qPCR were designed using the NCBI primer-BLAST tool (https://www. ncbi.nlm.nih.gov/tools/primer-blast/index.cgi?LINK_LOC=BlastHome). The primer sets Inx2prox (F 5 -AGTGGGTGTGCTTCGTATTGT-3 , R 5 -GTCCGCCTTACAGTCCTCG-3 ), and fbpAprox (F 5 -CTGCGTGGCTGTTGGGTTTA-3 , R 5 -GCGCAGCATAGGCAATAACA-3 ) amplified 143 bp and 188 bp fragments of Inx2 and fbpA loci, respectively. Each qPCR mixture of 20 µL volume included 10 µL of ×2 qPCR HS-Taq mix (Biolabmix, Novosibirsk, Russia), 0.3 µM of each primer, and a 1 µL DNA sample that was in the range of 10-40 ng. All qPCRs were performed in duplicate on a CFX 96 Real-Time PCR Detection System (Bio-Rad Laboratories Inc., Hercules, CA, USA). Conditions of the qPCR were optimized by gradient PCR; we used 95 • C for 5 min, followed by 50 cycles of 95 • C for 10 s, and 63 • C for 30 s. Melt curves were generated to ensure specificity of the PCR product (65-95 • C with a step of 0.5 • C). Efficiencies of amplification and R 2 were determined using a 5-fold dilution of a DNA sample for both loci (Inx2prox; E 98.21%, R 2 0.9987; fbpAprox: E 101.50%, R 2 0.9989).

Data Analysis
All statistical calculations were performed using Minitab 17.1.0 software (Minitab Inc., State College, PA, USA). The sequences were analyzed using FinchTV v. 1.4.0 (Geospiza Inc., Seattle, WA, USA). The alignments were generated using the MUSCLE algorithm [41]. For phylogenetic analysis of P. proximus mtDNA, the 620 bp region of the COI gene was used (38 sequences). To identify mtDNA Insects 2020, 11, 547 5 of 12 haplogroups, we used the data from Kononov et al. [42]. Phylogenetic trees were reconstructed by the maximum likelihood algorithm in MEGA 6.0 [43], using the Tamura 3-parameter model for the COI gene and the GTR model for MLST data. For the MLST data analysis, we also used four Wolbachia isolates of Scolytinae species (Seqence Types (STs): 138, 207, 208, 210), seven STs (1,8,9,35,62,19,90) as Wolbachia supergroup references, and three STs (184, 303, 482) that had similar haplotype characteristics to ST-533 (Table S2; SM1). Each allele of ST-533 was independently analyzed for supergroup clustering to check the inter-supergroup recombination of Wolbachia haplotypes [8,9]. Wolbachia load was measured as a relative copy number (fbpA/Inx2) = 2 Cq(Inx2)−Cq(fbpA) . Because the efficiencies of both primer sets were close to 100%, the amplification differences were not taken into account in calculations.

Wolbachia Infection Rates and Wolbachia Titer
Wolbachia prevalence in populations of P. proximus was initially estimated by conventional PCR, followed by symbiont negative DNA samples being rechecked by nested PCR (Table S3). Finally, we estimated Wolbachia titer by real time quantitative PCR across local collections.

Conventional PCR
The Wolbachia symbionts were found in 12 out of 13 P. proximus samples ( Table 2). The only Wolbachia-free sample was Tomarinsky-2, Sakhalin, with a sample size of N = 8. The infection rates varied among sampling sites from 39% to 77% and were not statistically different between the studied regions (Pearson's chi-square, p = 0.237).
Surprisingly, we found a great difference in Wolbachia prevalence for males and females, with 6 out of 13 samples being partially or totally subdivided based on sex. The proportion of infected females was significantly higher than males: across all collections p < 0.001; in localities, for three populations p < 0.001; for two populations p < 0.05; and for one p = 0.138.

Nested PCR
Because a significant difference in Wolbachia rates within populations based on the sex of individuals was observed, we tried to reexamine all Wolbachia negative DNA samples by nested PCR. In total, we analyzed 189 samples, some of which were checked for two loci. As a result, 69 new infected samples were found. Previously seemingly Wolbachia-free samples from Tomarinsky-2 appeared to be nearly totally infected (7 out of 8). Taking into account the results based on both conventional and nested PCR, the infection rates varied from 48% to 92%. The infection rates in Udmurtia and Sakhalin were higher than in Tomsk and Krasnoyarsk (Fisher exact test for each of the four comparisons, p < 0.05 and p < 0.001).
Overall differences between infected females and males remained significant (Fisher exact test p < 0.01); however, in terms of localities, significant differences were only found in Tomarinsky-1 (p < 0.05) and Bakcharsky (p < 0.01) ( Table 2).

Estimation of Wolbachia Titer by qPCR
The values of Wolbachia genomes per beetle genome were measured by real-time quantitative PCR. Here, we used 27 conventional-positive and 28 nested-positive samples across all regions; among them were 13 females, 20 males, and 22 samples with undetermined sex. Variation of the Wolbachia copy number (fbpA/Inx2) was in the range of 2.9 × 10 −4 to 1.73. Our results for the conventional and nested PCR were in a good agreement with the estimation of Wolbachia load by qPCR. The threshold of the two PCR approaches was about 0.018-0.020 (Figure 2). The differences between the most heavily and lowest infected females was 22-fold, and the difference was 864-fold in males. The median of the Wolbachia titer was 0.1792 in females and 0.0035 in males, which was a significant difference (U Mann-Whitney p < 0.0001). In particular, there were no females below the threshold, whereas there were seven males (detected by conventional PCR) above it. In two cases, Wolbachia infection was successfully detected by conventional PCR, although the infection titers were~0.002 and 0.005. conventional and nested PCR were in a good agreement with the estimation of Wolbachia load by qPCR. The threshold of the two PCR approaches was about 0.018-0.020 (Figure 2). The differences between the most heavily and lowest infected females was 22-fold, and the difference was 864-fold in males. The median of the Wolbachia titer was 0.1792 in females and 0.0035 in males, which was a significant difference (U Mann-Whitney p < 0.0001). In particular, there were no females below the threshold, whereas there were seven males (detected by conventional PCR) above it. In two cases, Wolbachia infection was successfully detected by conventional PCR, although the infection titers were ~0.002 and 0.005.

Characterization of Wolbachia Isolates and Mitochondrial DNA
Complete MLST profiles were obtained for four Wolbachia isolates (SM-1) from the Tomsk, Krasnoyarsk and Sakhalin regions; all were identical and corresponded to the ST-533 haplotype. Alleles ftsZ-263 and fbpA-461 were first characterized in the PubMLST database; then, we checked them in the GenBank database using the BLASTn tool and confirmed their uniqueness. For this reason, we characterized additional isolates from populations of Udmurtia, Tomsk, and Krasnoyarsk (IDs 1940-1943; SM-1) with regard to the ftsZ and fbpA genes and did not detect any other alleles. The concatenated sequence of ST-533 and each MLST allele was clustered into supergroup B ( Figure 3 and Table S2). The ST-482 haplotype was the closest to ST-533, and there were three substitutions in the ftsZ and fbpA loci. Unfortunately, the host of ST-482 was not indicated in the PubMLST database (an isolate was not registered). The Wolbachia haplotype of Pityogenes chalcographus (Scolytinae) was also clustered into supergroup B, with a total of 72 substitutions from ST-533. Wolbachia infection of Leptopilina clavipes (parasitoid wasp) had the same hcpA-148 as ST-533 and ST-482; however, its haplotype (ST-303) was genetically closer to the isolate of P. chalcographus than that of P. proximus. The ftsZ-91 allele of Wolbachia from Hylyphantes graminicola (spider) differed from ftsZ-263 of wProx by only two substitutions, although other alleles were found to have much greater differences.

Characterization of Wolbachia Isolates and Mitochondrial DNA
Complete MLST profiles were obtained for four Wolbachia isolates (SM-1) from the Tomsk, Krasnoyarsk and Sakhalin regions; all were identical and corresponded to the ST-533 haplotype. Alleles ftsZ-263 and fbpA-461 were first characterized in the PubMLST database; then, we checked them in the GenBank database using the BLASTn tool and confirmed their uniqueness. For this reason, we characterized additional isolates from populations of Udmurtia, Tomsk, and Krasnoyarsk (IDs 1940-1943; SM-1) with regard to the ftsZ and fbpA genes and did not detect any other alleles. The concatenated sequence of ST-533 and each MLST allele was clustered into supergroup B ( Figure 3 and Table S2). The ST-482 haplotype was the closest to ST-533, and there were three substitutions in the ftsZ and fbpA loci. Unfortunately, the host of ST-482 was not indicated in the PubMLST database (an isolate was not registered). The Wolbachia haplotype of Pityogenes chalcographus (Scolytinae) was also clustered into supergroup B, with a total of 72 substitutions from ST-533. Wolbachia infection of Leptopilina clavipes (parasitoid wasp) had the same hcpA-148 as ST-533 and ST-482; however, its haplotype (ST-303) was genetically closer to the isolate of P. chalcographus than that of P. proximus. The ftsZ-91 allele of Wolbachia from Hylyphantes graminicola (spider) differed from ftsZ-263 of wProx by only two substitutions, although other alleles were found to have much greater differences. As Wolbachia infection was uniform across a vast territory, we analyzed a 620 bp barcode region of the COI gene to characterize maternal inheritance in more detail. In particular, we aimed to determine the mtDNA haplotype of infected and uninfected P. proximus specimens at all sampling sites. As references, we used data from Kononov et al. [42], where five mtDNA haplogroups have been described. The analyzed mtDNA of infected and uninfected samples of the Sakhalin and Tomsk collections belonged to haplogroup I (Figure 4). In Udmurtia they belonged to haplogroups I and II, and in Krasnoyarsk haplogroup IV was detected. Thus, we did not find dependence between the Wolbachia infection status and mtDNA haplogroup. As Wolbachia infection was uniform across a vast territory, we analyzed a 620 bp barcode region of the COI gene to characterize maternal inheritance in more detail. In particular, we aimed to determine the mtDNA haplotype of infected and uninfected P. proximus specimens at all sampling sites. As references, we used data from Kononov et al. [42], where five mtDNA haplogroups have been described. The analyzed mtDNA of infected and uninfected samples of the Sakhalin and Tomsk collections belonged to haplogroup I (Figure 4). In Udmurtia they belonged to haplogroups I and II, and in Krasnoyarsk haplogroup IV was detected. Thus, we did not find dependence between the Wolbachia infection status and mtDNA haplogroup.

Discussion
Wolbachia symbionts in the subfamily Scolytinae have previously been found in 11 species of 8 genera [26][27][28][29][30], and many of these species are characterized by high infection rates (70-100%). Here, we first provided data on Wolbachia infection in populations of P. proximus in four distant regions of its native and invasive ranges. The overall infection rate in the studied populations was 68% (95% confidence interval, 63-73%). Surprisingly, the infection rate of females was higher than in males (79.7% vs. 61.8%; Fisher exact test p = 0.003). Moreover, most of the infected males were characterized by significantly lower Wolbachia titers than the females. Genetic diversity of Wolbachia in P. proximus populations indicated only one strain in the vast studied territory, which is associated with at least three out of five mitochondrial haplogroups.
Maternal transmission of Wolbachia implies that the symbiont has no future in males; therefore, it seemed logical that Wolbachia could be absent in males. We have not found other reports of sex-

Discussion
Wolbachia symbionts in the subfamily Scolytinae have previously been found in 11 species of 8 genera [26][27][28][29][30], and many of these species are characterized by high infection rates (70-100%). Here, we first provided data on Wolbachia infection in populations of P. proximus in four distant regions of its native and invasive ranges. The overall infection rate in the studied populations was 68% (95% confidence interval, 63-73%). Surprisingly, the infection rate of females was higher than in males (79.7% vs. 61.8%; Fisher exact test p = 0.003). Moreover, most of the infected males were characterized by significantly lower Wolbachia titers than the females. Genetic diversity of Wolbachia in P. proximus populations indicated only one strain in the vast studied territory, which is associated with at least three out of five mitochondrial haplogroups. Maternal transmission of Wolbachia implies that the symbiont has no future in males; therefore, it seemed logical that Wolbachia could be absent in males. We have not found other reports of sex-specific Wolbachia distribution in other populations, as is the case with single infected P. proximus. However, Arai et al. [44] reported, in an experiment on subdividing triple infection, a 79% transmission rate for the wHm-b strain in males of Homona magnanima (Lepidoptera: Tortricidae) versus a rate of 100% in daughters. Dutton and Sinkins [45], in double infected Aedes albopictus, found wAlbA loss in males. Further, Tortosa et al. [46] confirmed a loss of the wAlbA strain and demonstrated that wild females were double infected, whereas significant proportions of wild males were only single infected. There is a prediction of reduced expression of cytoplasmic incompatibility (CI) in males through selection of the host genome [47], and Wolbachia loss in males could be considered [46]. Therefore, if Wolbachia maintenance in a host population is based on CI, CI repression due to Wolbachia loss in males should lead to a loss of infection in a whole population. We do not have sufficient information to infer the effect(s) of Wolbachia on P. proximus biology, although the Wolbachia-inducing phenomena that lead to a sex ratio bias should be excluded, because we observed the primary sex ratio to be 1:1. There is a question as to whether Wolbachia in P. proximus is driven by CI and therefore will be lost in future, or high Wolbachia rates maintained by mutualistic effects for females, as, for instance, has been shown for Wolbachia-infected females of the coffee berry borer Hypothenemus hampei (Scolytinae), which produce more eggs than tetracycline-treated females [12].
The uninfected P. proximus males are likely the result of Wolbachia loss in ontogenesis. Most infected males are characterized by low Wolbachia density over all of the studied regions. Similar results were found for the wAlbA strain of Ae. albopictus [45,46]. In Scolytine, cases of low Wolbachia titers were found in populations of Pityogenes chalcographus [29]; however, there was no comparison between males and females. Therefore, there are few examples of sex-specific differences in Wolbachia titers, whereas there are a number of low symbiont titers in other species [29,[48][49][50][51][52][53][54][55].