Microbiome profiling of the onion thrips, Thrips tabaci Lindeman (Thysanoptera: Thripidae)

The gut microbial community structure of adult Thrips tabaci collected from 10 different agro-climatically diverse locations of India was characterized by using the Illumina MiSeq platform to amplify the V3 region of the 16S rRNA gene of bacteria present in the sampled insects. Analyses were performed to study the bacterial communities associated with Thrips tabaci in India. The complete bacterial metagenome of T. tabaci was comprised of 1662 OTUs of which 62.25% belong to known and 37.7% of unidentified/unknown bacteria. These OTUs constituted 21 bacterial phyla of 276 identified genera. Phylum Proteobacteria was predominant, followed by Actinobacteria, Firmicutes, Bacteroidetes and Cyanobacteria. Additionally, the occurrence of the reproductive endosymbiont, Wolbachia was detected at two locations (0.56%) of the total known OTUs. There is high variation in diversity and species richness among the different locations. Alpha-diversity metrics indicated the higher gut bacterial diversity at Bangalore and lowest at Rahuri whereas higher bacterial species richness at T. tabaci samples from Imphal and lowest at Jhalawar. Beta diversity analyses comparing bacterial communities between the samples showed distinct differences in bacterial community composition of T. tabaci samples from different locations. This paper also constitutes the first record of detailed bacterial communities associated with T. tabaci. The location-wise variation in microbial metagenome profile of T. tabaci suggests that bacterial diversity might be governed by its population genetic structure, environment and habitat.


Introduction
Bacterial communities in insects play an important role in their growth, development, immunological, physiological and morphological functioning. The majority of insects are believed to harbour heritable bacterial symbionts [1] that can be pathogenic, mutualist, or commensal, with some required for survival while others are not. Across Insecta, microorganisms have been reported to positively influence many functions, including the production of essential amino acids from nutrient poor diets [2], protection against toxic agents [3][4][5], aide in the PLOS ONE | https://doi.org/10.1371/journal.pone.0223281 September 30, 2019 1 / 13 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 production of honey [6], protection against parasitoids [7], virus transmission [8], insecticide resistance [9], degradation of phytotoxins and pesticides [10]. Conversely, some are reported to negatively impact insects by causing sterility and distorting sex ratios [11,12]. Despite their influence on important metabolic processes in the host, they have not been accurately profiled due to the difficulty in isolating and culturing many of the symbionts. Identification of symbionts has been improved with the availability of next generation sequencing technology, which, bypasses the need of isolating and culturing, can detect microbes present in very low amounts, and facilitates the study of the microbial community in its natural habitat with accurate taxonomic identification and their relative abundances [13]. Characterizing the diversity of symbionts is an important first step towards understanding their importance in the life history of organisms.
There is a growing area of interest regarding the presence and influence of heritable endosymbionts on the biology and ecology of economically important insects, and on the phenotypes that influence pest status. Several studies on endosymbionts have targeted pest species that damage crops by feeding on plants and transmitting plant pathogens such as aphids [14], whiteflies [15], and thrips [16]. The studies have identified endosymbionts that influence fitness [17], virus transmission [8], host plant preferences [18], protection from biological control agents [7], and insecticide resistance [9]. Endosymbionts are also of interest as targets for future pest control strategies that could be achieved by disturbing the essential symbionts of insect pests and symbionts which contribute to their pest status [19].
Onion thrips, Thrips tabaci Lindeman (Thysanoptera: Thripidae), is a globally important polyphagous insect pest. It has been collected from approximately 300 plant species that includes economically important crops such as leek, garlic, tobacco, cabbage, pea, melon, lettuce, potato, tomato, carnation, and cotton [20,21], and causes more than $1 billion losses worldwide to onion alone [22]. In addition to damaging crops through direct feeding, T. tabaci is a vector of the Orthotospoviruses Iris yellow spot virus (IYSV) and Tomato spotted wilt virus (TSWV) [23,24]. Onion thrips exhibits three different reproductive modes (thelytoky, arrhenotoky and deuterotoky) [25], host-associated lineages, variation in virus (TSWV) vector competence [26], heteroplasmy of mtCOI haplotypes [27] and develop resistance to insecticides quickly. Variation in these phenotypes may be due to selection on standing genetic variation but could be influenced by their microbiome. Insect bacterial communities may also influence their reproduction [28,29] and virus transmission efficiency of thrips [30].
To date, limited information is available regarding the endosymbiont profiles of this species; there is one study that examined one population using culturing methods [30] or only one bacteria species was examined 25 . The objectives of this study were to identify bacterial microbiota present in T. tabaci from ten distinct locations of India using the Illumina MiSeq platform to sequence the V3 region of 16s rRNA, and examine patterns of bacterial diversity within and among these locations.

Ethics statement
Thrips tabaci has not been notified under any act or laws and rules thereof of the Government of India as an endangered or threatened species restricting or regulating its collection and observation. Therefore, permits were not required for collecting T. tabaci for the present study.

Insect sampling
Adult T. tabaci were collected from 10 different states in India during rabi season (December-June) of the year 2013-2014. The states sampled encompassed different climatic zones in India, including temperate, tropical, and subtropical zones of the country (Fig 1).
Approximately 500 thrips were collected from each state (50 thrips from each field, 10 fields per location), and were sampled within 1-2 days from each other. Adult thrips were dislodged from onion leaves with a fine brush into a 2 ml Eppendorf tubes containing 95% ethanol. To make certain offspring from different adults were collected, a distance of 1.5 m was kept between sampled plants. Voucher specimens are located at plant protection section, ICAR-Directorate of Onion and Garlic Research, Rajgurunagar, Pune.

DNA extraction
Adult T. tabaci were surface sterilized by rinsing them with 100% ethanol. Genomic DNA was extracted from pool of 500 adult thrips from each location using a standard phenol-chloroform and ethanol precipitation method [31].
Bacterial 16S rRNA gene amplification, library preparation and sequencing 16S rRNA gene sequencing was performed at SciGenom Labs Private Limited (Cochin, India) on an Illumina MiSeq 2 x 151 platform (Illumina, Inc., San Diego, CA, USA). Base quality score distributions, average base content per read and GC distribution in the reads were used as quality check parameters for sequences obtained from the sequencer (Illumina MiSeq).

Bioinformatics and statistical analysis
To identify bacteria present in the samples sequence data needed to be filtered for quality, aligned into contigs, and blasted against existing bacterial sequence databases. An average of 361,241 paired end sequences with a length of 151bp were obtained from sequencing 10 samples (minimum 218,538; maximum: 499,241). Gut microbiome analysis was performed with QIIME2 2017.9 framework [32]. Raw sequences were quality filtered and denoised followed by chimera filtering with DADA2 [33] to obtain amplicon sequence variants (ASV). Low abundant ASVs with a total frequency less than ten were filtered out of the dataset. Taxonomy assignment was performed with a pre-trained Silva 132 99% OTUs based naïve Bayes classifier and confidence threshold of 0.7 using q2-feature-classifier plugin [34]. ASVs matching with chloroplast, mitochondria and eukaryotic sequences were removed from the database for downstream analysis. A taxonomic summary table was generated for each level of taxonomy for phylotype abundance.
Sequence alignment using MAFFT [35] was performed for all ASVs and used for the construction of phylogeny with fasttree2 [36]. Alpha rarefaction plotting was performed with a minimum and a maximum depth of 100 and 10000 respectively to identify an ideal sampling depth for further analysis. Three alpha diversity indices namely, Shannon, chao1, and simp-son_e at ideal sampling depth have been estimated. The similarity of bacterial communities between samples (beta diversity) was quantified using a metric based on phylotype abundances. The distance matrix was generated using weighted UniFrac approach [37], and a jackknife test with 100 iterations was performed to construct a consensus UPGMA (Unweighted Pair Group Method with Arithmetic Mean) tree. Principal Coordinates Analysis (PCoA) was performed to generate a 3D PCoA plot in EMPeror [38].
ANCOM compares the log ratio of the abundance of each taxon to the abundance of all the remaining taxa one at a time [39]. ANCOM (ANalysis of Composition Of Microbiomes) was performed to identify taxa with differential abundance among sample groups. ANCOM is now incorporated into the QIIME suite for metagenome analyses. For ANCOM analysis samples were grouped into four climatic zones according to Köppen classification viz; (Monatne (MON): Shrinagar; Humid subtropical (HST): Hisar, Imphal, Jabalpur, Samastipur; Tropical wet and dry (TWD): Pune, Rahuri, Chiplima and Semi-arid (SA): Bangalore and Jhalawar.

Sequencing data
The Illumina MiSeq sequencing of the V3 region of 16S rRNA gene of onion thrips from ten different locations yielded 218,538-499,241 raw reads per location. Nearly 80% of the total reads had Phred scores greater than 30 (>Q30; error-probability > = 0.001) indicating good quality of data. GC content of the reads ranged from 40 to 60% and after filtering contig length (~150bp). After quality filtering 39,976-140,603 reads per sample remained. Pre-processed reads from all samples were pooled for a total of 962,166 reads, and from them, a total of 1662 OTUs were identified (Table 1).

Microbiome profile of T. tabaci
A total of 21 phyla, which constitute 62.25% of the total 16S rRNA gene data set, were detected from T. tabaci in this study (S1 File). The majority of bacterial contigs were of unknown origin (33%) while 4.7% were unknown bacteria. The contigs that do not have any alignment against the taxonomic database were categorized as "Unassigned". From the identifiable bacterial sequences, the majority belonged to phylum Proteobacteria, followed by Firmicutes, Actinobacteria, Bacteroidetes and Cyanobacteria (Fig 2, Fig 3). Across all locations, these five phyla comprised more than 60% of the total microbiome and more than 96% of identified microbiome of T. tabaci.

Microbiome diversity of T. tabaci at different location
Alpha-diversity indices (Shannon, Simpson-e) describe the diversity of the microbial community at each sampling location and showed that bacterial diversity is higher in Bangalore, followed by Jabalpur and that the lowest bacterial diversity is observed in Rahuri. Another alpha diversity index (Chao1) describes the species richness at each sampling location. Among ten locations, bacterial richness is found to be highest in Bangalore and lowest in the Jhalawar ( Table 2). Beta diversity considers the variations in bacterial community composition for different environments. Bacterial diversity among locations was assessed using both the weighted and unweighted UniFrac approach. UniFrac distances are based on the fraction of branch length of the 16S rRNA phylogenetic tree shared between two communities. The UPGMA tree and PCoA plots were constructed using the weighted UniFrac distance matrix. Locations were clustered based on the bacterial community structure. From PCoA plot it has been seen that there is high diversity among the bacterial communities of T. tabaci from different geographical location (Fig 4). PC-1 and PC-2 explain 61.87% and 22.1% cumulative variance in the microbiome of the T. tabaci, respectively. ANCOM analyses reveals two genera in Montane zone Chryseobacterium and Exiguobacterium found to be higher in Shrinagar location (S2 File, S3 File).  Recently, NGS has become the method of choice for insect microbiome analyses due to the ability to detect both culturable and non-culturable bacteria. The present study constitutes the first detailed examination of bacterial communities in T. tabaci using sequencing methods. In this study phyla Proteobacteria, Actinobacteria, Firmicutes and Bacteroidetes together constituted more than 90% of the total T. tabaci identified microbiome. These phyla are also reported to be the predominant in the microbiomes of other thrips species [ [49,10]. The phylum Actinobacteria, was abundant at two locations and is reported to increase metabolic versatility and the ability to exploit the wide range of nutritional resources e.g. polysaccharides like cellulose [50] and hemicelluloses in termites [51] and production of secondary metabolites with antibiotic properties [52]. The third most prevalent phylum was Firmicutes, which predominated at three locations. Bacteria in Firmicutes were also abundant in guts of termites and honeybees [53,54]. Several studies in insects and animals have shown that Firmicutes increased the ability to metabolize food resources to increase energy conversion from diet [55], and assist in the digestion of cellulose and hemicelluloses [56]. Bacteroidetes were the fourth predominant phylum in present16S rDNA data set. The members of this phyla are known for their role in the production of enzymes such as glucanase, mannanase and xylanase that aid in complex carbohydrate metabolism [57,58]. Phylum Cyanobacteria was prevalent in one location, and these bacteria are associated with high levels of protein, vitamins, and microelements. They are known to release toxins during their life that lead to concentration dependent, and speciesspecific negative effects on animal feeders [59]. These five phyla were also found to be predominant in the microbiome of the other thrips species Scirtothrips dorsalis and Hoplothrips carpathicus [16,42].
We detected the genus Wolbachia, in two out of ten locations. Wolbachia infects over 40% of all arthropods [60] and is known to manipulate sex determination mechanisms and sex ratio by cytoplasmic factors, cytoplasmic incompatibility [61,62]. This is the first report of Wolbachia in T. tabaci. Studies have looked presence of Wolbachia in different thrips species  Few plant origin endophytes such as Actinomyces, Microbacterium, and Burkholderia were recorded in the microbiome of T. tabaci. Theses endophytes might have been acquired by the insect during feeding on plants. Their occurrence in insects gut has been well documented in previous studies [64,65].
This study marks the first attempt to document the endosymbiont diversity associated with T. tabaci in India. Diversity indices showed that T. tabaci from Bangalore has the highest bacterial diversity and Imphal has the highest species richness. Further UPGMA and PCoA analyses showed bacterial communities structured by location. This spatial variation in the microbiome T. tabaci might be due to geographical location, climatic conditions and host phylogeny. In several insects, variation in the microbiome with respect to climatic conditions, the geography of habitat, and phylogeny of the insect have been documented [69,42,14].

Conclusion
This paper described the microbiome of T. tabaci collected from the different geographical locations of India using NGS approach. Findings of the present study increased understanding of microbiome of T. tabaci as well as its variation with respect to geography and climatic conditions. Though it is first report of its kind in T. tabaci, for more in depth study of T. tabaci microbiome needs to be done at different developmental stages for better understanding of its role in development and growth of T. tabaci.