Molecular detection and phylogenetic analyses of Wolbachia in natural populations of nine galling Aphid species

Wolbachia is one of the most abundant facultative intracellular symbionts in arthropods. It alters host biology in diverse ways, including the induction of reproductive manipulation, association of nutrient supplier and protection against pathogens. Aphids are a group of insects which exhibit interesting biological characteristics such as complex life cycles, alteration of sexual and asexual reproduction and shifts between two different hosts. Wolbachia is widely present in many orders of insects, but so far limited studies on Wolbachia in aphids have been carried out. Galling aphids are a group of aphids that induce galls on their primary host plants at specific life stage. In this study, 15 natural populations representing nine galling aphid species were analyzed for the presence of Wolbachia using species-specific primer pairs. Wolbachia presence in galling aphids was quite low and varied significantly among aphid populations. Only three of the 15 populations we analyzed had detectable Wolbachia and the overall infection rate was 20%. Two Wolbachia strains, O and B, were identified from the galling aphids Kaburagia rhusicola and Schlechtendalia chinensis. Strain O was for the first time to be found in aphids, and it is likely involved with the life stages of galling aphids living in closed microenvironments with specific survival strategies that are different from free-living aphids.

. The distribution patterns of Wolbachia in Chinese aphids are complex and varied among different species 4 . Wolbachia in aphids is underestimated primarily because of its low titer and/or high divergence of different strains. Among the 217 aphid species analyzed for Wolbachia before, only 11 were galling aphids (4.1%), and of those, only six galling aphids carried Wolbachia 3,4 . Therefore, the distribution of Wolbachia in galling aphids and its significance remain to be studied.
The objective of this study is to investigate the presence of Wolbachia in natural populations of galling aphids from China. Specifically, Wolbachia strains were screened and classified based on nine marker genes: 16S rRNA, gltA, groEL, wsp, gatB, fbpA, coxA, hcpA and ftsZ. Phylogenetic analysis was also conducted based on the sequences of the marker genes.  (Table 1). All samples were screened for the presence of Wolbachia by PCR amplification using 16S rRNA-specific primers 16S-281F/1372R. Our results showed that the presence of Wolbachia in these populations was infrequent and varied significantly among different aphid populations. Among all 15 aphid populations, only 3 populations were detected to harbor Wolbachia. These Wolbachia-carrying aphids were Kaburagia rhusicola and Schlechtendalia chinensis from the tribe Fordini, and Pemphigus yunnanensis from the tribe Pemphigini. No Wolbachia was detected in the remaining populations. Among the Wolbachia-positive species, Wolbachia was detected in four out of six samples in K. rhusicola, and in one out of nine samples in either S. chinensis or P. yunnanensis (Table 1).

Discussion
There are 16 supergroups of Wolbachia that have been identified in insects at present [3][4][5]  The relation between aphids and their facultative symbiont Wolbachia can be affected by different factors such as the ability of symbionts to spread from aphids to aphids within or across populations, the cost of infection for hosts, and aphid living environments (host plant, natural enemy pressure, or temperature) 32,33 . Compared to non-galling aphids, most life stages of galling aphids are in closed microenvironments 21,23 . Generally, hundreds to thousands of aphid individuals living in a gall are produced parthenogenetically by a single fundatrix 19,21,34 . If a fundatrix did not carry Wolbachia before she induces a gall, its offspring have basically no chance to be infected by Wolbachia in an enclosed environment. This means that Wolbachia can hardly spread across galling aphid populations. Moreover, galling aphids receive less pressure from natural enemies than non-galling aphids since they are protected by the gall wall 20,35 . Although vertical transfer is the predominant mode for Wolbachia, horizontal transfer also appeared in nature by infrequently. Horizontal transfer of Wolbachia was observed when infected and uninfected larvae of Trichogramma wasps shared the same hosts 31 . Also, Hymenopteran parasitoids of frugivorous Drosophila acquired Wolbachia through horizontal transmission with high frequency 36 . The three galling aphids in our study are host alternation, so they may readily acquire Wolbachia when they free-living on the secondary hosts by shared the same host mosses or attacked by parasitoid wasps which carried Wolbachia. All aphid samples of our study were collected from closed galls, so they are quite different in their living environments, natural enemy pressures and host plants compared to non-galling aphids 20,21,34 . This is probably the reason why they harbor Wolbachia strains or supergroups different to those in free-living aphids 3,4 .
Many marker genes have been used to detect Wolbachia in insects, but the consistency of these markers varies among different insect species [37][38][39] . The most conserved marker gene among different Wolbachia strains is the 16S rRNA gene, which also provides more consistent PCR amplification. However, because the 16S rRNA gene is highly conserved, non-target amplification occurs during PCR amplification 5 . The wsp gene evolves faster among different Wolbachia strains. Therefore, PCR amplification is not highly consistent with current primer pairs 7 . In this study, the primers for the 16S rRNA, gatB and fbpA genes could detect Wolbachia in some galling aphid populations. However, primers for the coxA, hcpA and ftsZ genes detected Wolbachia only in P. yunnanensis (Table 1). Primers for the wsp, groEL and gltA genes did not detect Wolbachia in any samples, indicating that the primer regions in these genes varied and the primers could not achieve specific PCR amplifications. Our results also revealed that Wolbachia in the galling aphids was more difficult to detect using the existing primers possibly because of the variation in the sequences of target genes in the Wolbachia strains. Thus, the selection of more efficient and specific Wolbachia gene primers is needed to advance future research on Wolbachia analyses in aphids, especially in galling aphids.

Materials and methods
Aphid sample collection. Aphids of each population were collected from a mature gall. Fresh aphid galls were collected from host plants manually and dissected in the lab. Aphids were transferred to an Eppendorf tube from a gall, placed in 100% ethanol and then stored at − 20 °C until DNA extraction. Galls collected from each location within ten meters were treated as the same population. Since three generations of aphids are produced Bemisia tabaci 11 (KF452563)

Torotrogla cardueli (KP114111)
Torotrogla www.nature.com/scientificreports/ parthenogenetically by a mating female in a gall to achieve hundreds to thousands of offspring, aphids from a gall were treated as a clone. Galls collected from a location outside 10 m were treated as different populations. Nine galling aphid species (Aphididae: Eriosomatinae: Fordini and Pemphigini) were collected from five host trees at eight locations in Yunnan, Sichuan, Hubei and Shaanxi provinces of China. K. rhusicola was the most common species and five populations were obtained from this species. S. chinensis was the second common species with three populations collected. The other six galling aphid species included Floraphis meitanensis, Schlechtendalia peitan, Pemphigus yangcola, Pemphigus yunnanensis, Pemphigus populitransversus and Chaetogeoica folidentata, each with only one population collected (Table 1).
DnA extraction. Ten to twenty aphid individuals of a population were selected for DNA extraction.
Genomic DNA was extracted using Dzup Genomic DNA Isolation Reagent (Sangon Biotech, China) from pooled aphids of the same clone following the manufacturer's protocol and was then stored at − 20 °C until detected.
Wolbachia detection. DNA samples of 117 natural galling aphid populations were screened for the presence of Wolbachia strains. Detection was based on the amplification of the 16S rRNA gene fragment (about 1,100 bp) with the Wolbachia-specific primers 16S 281F 5′-CTA TAG CTG ATC TGA GAG GAT-3′ and 16S 1372R 5′-YGC TTC GAG TGA AAC CAA TTC-3′ (Table 3) 4 . PCR procedures were an initial step of 94 °C for 3 min, followed by 35 cycles of 94 °C for 45 s, 55 °C for 60 s, and 72 °C for 90 s and a final step of 72 °C for 10 min. Amplified DNA products were electrophoresed on agarose gels and stained. PCR products were sequenced from both directions in an ABI 3730 DNA analyzer (Applied Biosystems, Foster City, CA, USA). Sequencing results were then checked by Blast in NCBI nr database (https ://blast .ncbi.nlm.nih.gov/Blast .cgi). Only those samples, which were blasted to expected products of the specific primers, were selected for further analysis. These selected samples were further examined by PCR analyses of the genes gltA, groEL, wsp, gatB, fbpA, coxA, hcpA and ftsZ using specific primers based on previous reports (Table 3) [40][41][42][43] . PCR amplifications were carried out in 25 μl reaction volume, consisting of 1 μl DNA, 1 μl of forward and reverse primers (10 μmol/l), 12.5 μl Taq Master Mix (Sangon Biotech, China), and 9.5 μl ddH 2 O. PCR products (4 μl) were electrophoresed on a 1.5% agarose gel. Positive samples were further analyzed. www.nature.com/scientificreports/ Sequence alignment and phylogenetic analysis. Genes representative of Wolbachia strains from different supergroups were selected from the NCBI database (https ://www.ncbi.nlm.nih.gov), and used to classify Wolbachia strains detected in our aphid samples. Sequence alignments were carried out using ClustalX 1.83 44 . Maximum likelihood (ML) and Bayesian inference (BI) were used for phylogenetic analysis. The ML analysis was carried out with MEGA 6.0 and node support rates were evaluated with 1,000 bootstrap replicates. The Bayesian inference (BI) method was performed in MrBayes 3.1.2 45 . In MrBayes, four chains (one cold and three heated chains) which were run for 4 million generations. Trees were sampled every 100 generations, and the first 25% of samples were discarded as burn-in. From the remaining trees, 50% majority-rule consensus trees were generated. Posterior probabilities were computed from the remaining trees. To obtain the appropriate evolution model, the parameters were evaluated using the Akaike Information Standard (AIC) in MrModeltest 2.3 46 . Using this method, the following gene models were obtained: the HKY + G model of 16S rRNA gene, the GTR + G model of ftsZ gene, the HKY + G model of coxA gene, the GTR + I + G model of gatB gene and the HKY + G model of fbpA gene.