Interspecific variation and functional traits of the gut microbiome in spiders from the wild: The largest effort so far

Spiders being one of the most diverse group in phylum arthropod are of great importance due to their role as predators, silk producer, and in medicinal applications. Spiders in prey–predator relationships play a crucial role in balancing the food-chain of any ecosystem; therefore it is essential to characterize the gut microbiota of spiders collected from natural environments. In the present work, the largest effort so far has been made to characterize the gut microbiota of 35 spider species belonging to four different families using 16S amplicon targeting sequencing. Further, we compared the gut microbiota composition including endosymbiont abundance in spider species collected from different geographical locations. The results obtained revealed the presence of genera like Acinetobacter (15%), V7clade (9%), Wolbachia (8%), Pseudomonas (5%), Bacillus (6%). Although comparative analysis revealed that the gut bacterial composition in all the spider families has a similar pattern, in terms of community richness and evenness. The bacterial diversity in the spider family, Lycosidae are more diverse than in Salticidae, Tetragnathidae and Araneidae. Furthermore, it was observed that the abundance of endosymbiont genera, i.e. Wolbachia and Rickettsia, leads to shift in the abundance of other bacterial taxa and may cause sexual alterations in spider species. Moreover, predicted functional analysis based on PICRUSt2 reveals that gut microbiota of spider species were involved in functions like metabolism of carbohydrates, cofactors and vitamins, amino acids; biosynthesis of organic compounds, fatty acids, lipids etc. Based on the results obtained, it can be said that different locations do not correlate with community composition of gut microbiota in spider species collected from natural environments.


Introduction
The functioning and survival of living organisms in an ecosystem are greatly influenced by the gut microbiome they possess [1,2]. Recent research has shown that gut microbiota plays a significant role in determining mating preference [3], reproductive manipulation [4], inhibiting pathogen transmission [5,6] and insecticidal resistance [7]. Moreover, gut bacterial diversity is also linked to various physiological processes like digestion, detoxification, or nutrient supplementation [5,[8][9][10]. The emergence of Next Generation Sequencing (NGS) has revolutionised the study of gut bacterial diversity [11]. While, NGS-based gut bacterial diversity assessment and their functional pathways has been widely studied in many groups of arthropods [12], including insects [8,9,[13][14][15], scorpions [16], ticks [17], and termites [18]. So far, few species of spiders have been studied for their gut bacterial diversity [19][20][21][22][23]. The spiders used for the gut microbiome analysis in earlier studies were either reared or diet-driven which play an important role in shifting the gut microbiome composition. With more than 48,000 described species worldwide, spiders are a diverse group of predators [24] and distributed globally in all the terrestrial ecosystems [25][26][27]. Spiders are known as "liquid feeders", as they have extra oral digestion (EOD) [20] and this peculiar way of ingestion potentially affects the microbiome composition in this group. So, the survey of gut bacterial diversity and their predicted metabolic functions in spiders is of pivotal importance due to their medicinal and ecological significance.
In 2010, Wand screened 31 Chinese spider species to understand the horizontal transmission of Wolbachia between prey and predator [28]. This study detected the presence of Wolbachia in only 7 species (Eriovixia cavaleriei, Larinia argiopiformis, Araneus ventricosus, Nephila clavata, Oxyopes sertatus, Pholcus crypticolens and Coleosoma octomaculatum). Later, Vanthournout and colleagues studied the involvement of Wolbachia, Rickettisia, and Cardinium in sex ratio variation in the dwarf spider Oedothorax gibbosus [29]. Hu et al. in 2018 detected Wolbachia and Rickettsia in the spider species Nurscia albofasciata [21]. In this study, we made the first large-scale attempt to decipher the community composition of the gut microbiota in natural populations of 35 spider species belonging to four families (Araneidae, Lycosidae, Salticidae, and Tetragnathidae) using the NGS amplicon data. We used statistical methods to compare the gut microbiome of spider species. The present study will enhance our knowledge on the gut microbiome and their interfamilial relationship.

Gut bacterial composition among the four spider families
To investigate the composition and diversity of gut microbes in spiders, we included 35 species of spiders belonging to four families, Araneidae (12), Lycosidae (7), Salticidae (10), and Tetragnathidae (7). However, one species, Thiania bhamoensis of the Salticidae family, was sequenced from two distinct geographic areas due to the cryptic behaviour of the species [30]. A total of 34,50,872 reads from the 16S rRNA were retained after the demultiplexing, quality filtering, and chimera removal. The average reads of each sample were 95,857 and ranged from 68004 (minimum) to 111599 (maximum). These reads were assigned to 13650 Amplicon Sequence variants (ASVs). A total of, 1282 ASVs was retained after the removal of singletons (10,154), low variance (143) and low abundance (2071) ASVs. The Venn analysis revealed 208 ASVs were shared by all the four families, while, 670 unique ASVs for Araneidae, 510 for Lycosidae, 484 for Salticidae and 590 for Tetragnathidae were recovered (Fig 1). The maximum number of ASVs (200) were shared between the spider families, Lycosidae and Salticidae while the minimum ASVs i.e. 13 were shared between Araneidae and Salticidae. Furthermore, all rarefaction curves were saturation which clearly indicated that the appropriate sequencing depth has been achieved for the taxonomic classification (S1 Fig). ASVs were assigned to taxonomic groups using the SILVA database with a 99% similarity cut-off. Moreover, the identified ASVs were distributed among 22 bacterial phyla and 150 families.
A total of 78 orders were observed in the current dataset. Among them the five major orders that are abundant in the gut of all the four spider species were Pseudomonadales (22%), Bacillales (12%), Enterobacteriales (9%), Rickettsiales (8%), and Micrococcales (5%). Further, the gut diversity of spider species belonging to family Araneidae was represented by five major orders such as Pseudomonadales, Bacillales, Enterobacteriales, Bacteroidales, and Betaproteobacteriales. They together contributed 66% to the total bacterial diversity. On other hand, the spider species belonging to the family Tetragnathidae were represented by Rickettsiales, Pseudomonadales, Rhizobiales, Enterobacteriales, and Bacillales, contributing to 80% of the total bacterial diversity. The gut bacterial diversity in spider species of family Lycosids were represented by Pseudomonadales, Bacillales, Rickettsiales, Corynebacteriales, and Micrococcales, contributed 65% of the total bacterial diversity. The family Salticidae, represented by Pseudomonadales, Bacillales, Enterobacteriales, Micrococcales and Propionibacteriales, contributed 56% to the total bacterial diversity.
The gut diversity of the members of all four spiders families were represented by 150 bacterial families. The bacterial families Moraxellaceae and Enterobacteriaceae were distributed across the four spider families, with the exception of the Lycosidae family where Enterobacteriaceae were not detected (Fig 2). In addition to this, few bacterial families were detected in specific spider family/families i.e. Bacillaceae and Prevotellaceae in Araneidae only; Rickettsiaceae and Rhizobiaceae in Tetragnathidae only; Anaplasmataceae in Tetragnathidae and Lycosidae; Pseudomonadaceae in Araneidae and Salticidae; Propionibacteriaceae in Lycosidae and Salticidae; Staphylococcaceae and Corynebacteriaceae only in Lycosidae; and Burkholderiaceae in Salticidae.

Endosymbionts abundance
We observed Rickettsia and Wolbachia as the major endosymbionts in the spider gut. These two bacterial genera have altered the other bacterial abundance in the spider species. The spider species, Orsinome vethi was detected with 77% Rickettsia and 17% Wolbachia and contributed 94% of total bacterial diversity. The spider species, Opadometa fastigata was detected with 3% Rickettsia, 32% Wolbachia, and 62% unidentified strain of the order Rickettsiales (Ac37b), which was 97% of the total bacterial diversity. Whereas, in Leucage celebesiana was detected with 88% Rickettsia, and Leucage decorata with 51% Wolbachia. Wolbachia showed an abundance of 92% in Hippasa greenalliae, and 27% in Eriovixia laglazei. This data indicated that the presence of the genus Wolbachia and Rickettsia in these species alter the abundance of other bacterial taxa except in Eriovixia laglazei (S1 Table and Fig 3).

Diversity measures
We used four diversity measures Chao1, Observed, Shannon and Simpson with the ANOVA statistical method to measure the alpha diversity in all four spider families (Fig 4). The alpha diversity of the four spider families ranged from 3-16 (Chao1, Observed), 0.05 to 1.3 (Shannon) and 0.07 to 0.16 (Simpson). The richness of four spider families based on two diversity measures (Shannon and Simpson) was significant (p < 0.05), while Chao1, Observed was nonsignificant (p> 0.05) (S2 Table). Thus, the community richness for the Lycosidae family was more diverse than that of the other three families.
The beta diversity in the gut of the four spider families based on NMDS was calculated using the Bray-Curtis index method (S2 Fig) and the dendrogram using the Unweighted Unifrac distance method (Fig 5). Both the methods showed the similar results with stress value of 0.12 (Bray-Curtis index) and 0.13 (Unweighted Unifrac distance). In both the analyses, four spider species showed the distinct differences in their gut bacterial diversity with respect to others. The Araneidae species (Cyclosa mulmeiensis, AA795) showed a high abundance of the Bacteroidetes phylum in comparison with other spider species. The Tetragnathidae species, Orsinome vethi (AA1394) and Opadometa fastigata (AA1840) showed a high abundance of Proteobacteria in compare to other spider species. The Araneidae species, Araneus mitificus (AA2217) was detected with only two phyla, i.e. Proteobacteria and Firmicutes.
Linear Discriminant Analysis (LDA) Effect Size (LEfSe) was implemented to examine the significant bacterial diversity between the species of four spider families with cut off 0.2 and Log LDA score 4 parameters. A total 48 significant features were detected, of which 14 are represented (Fig 6). The presence of the genera Rickettsia and Wolbachia in the family Tetragnathidae was a distinct difference as compared to other spider families. The genera Acinetobacter, Bacillus, and Raoultella significantly explain the differentiation of the bacterial

Comparative functional analysis
The predicted functional pathway analysis was carried out using PICRUSt2 ( Twenty eight (28) pathways were shared among the four spider families (Fig 7). The carbohydrate metabolism, cofactors and vitamin metabolism, amino acids metabolism, organic compound biosynthesis, fatty acid elongation and synthesis, heme biosynthesis and lipid biosynthesis pathways were observed in the all the four families.

Discussion
It is the largest effort to document the gut bacterial diversity of spiders. We studied 35 species from four families (Araneidae, Lycosidae, Salticidae, and Tetragnathidae) of spiders with their predictive functional metabolic pathways. The bacterial diversity in the gut of four families of spiders was distributed over 22 phyla and 150 families. Our study found that the most dominant phyla in the gut all spider species are Proteobacteria (61.5%) and Firmicutes (16.1%), and contributed to 78.5% of the total gut bacterial diversity. This result is well supported by previous studies and has shown that the most dominant phyla in the gut of spiders are Proteobacteria and Firmicutes [19,[21][22][23]. Furthermore, the abundance of Proteobacteria as the dominant phylum is well documented in other arthropods also [12] such as scorpions [16], ticks [17], cockroaches [31], and butterflies [15]. Other major phyla like Actinobacteria and Bacteroidetes have also been observed and contribute to 18.4% of total gut bacterial diversity.
Few differences were observed in the composition of the gut microbiome in the species studied. The genera Acinetobacter, Pseudomonas, Staphylococcus, Corynebacterium_1, Cutibacterium, and Aeromonas were detected in the gut of the four families of spiders. Whereas genus Bacillus was detected in all three families of spiders, with the exception of the family Tetragnathidae. The genus Wolbachia has been detected in Araneidae and Tetragnathidae; Rickettsia in Tetragnathidae only; Prevotella_9 in Araneidae; Micrococcus, Paracoccus and Enhydrobacter in Lycosidae; and Alishewanella, and Comamonas in Salticidae. Acinetobacter, and Pseudomonas of the phylum Proteobacteria are the most dominant taxa in the gut of the spider families as well as in other arthropods [12,[32][33][34]. The functional analysis revealed that Acinetobacter may be responsible for cofactor and vitamin metabolism in the gut of all the spider families as it has been reported to be involved in the biosynthesis of ubiquinol (https:// biocyc.org/META/NEW-IMAGE?type=PATHWAY&object=UBISYN-PWY). While the genus Pseudomonas was reported to be involved in the organophosphate degradation [35], energy and lipid metabolism [8].
In the present study, genera Bacillus and Staphylococcus of Phylum Firmicutes have been observed in all the spider families. They may be involved in the degradation of polysaccharides and aromatic compounds [36]. Members of the genus Staphylococcus were usually involved in detoxification, and defensive behaviour against natural enemies [37]. Cutibacterium in the gut of spiders may be involved in fatty acid metabolism as reported earlier in two-spotted spider mite, Tetranychus urticae (https://www.genome.jp/kegg-bin/show_pathway?pad00061). The genera Corynebacterium_1 and Aeromonas were detected in spiders and dipterans [38]. Earlier studies have revealed that the presence of Corynebacterium_1 in the gut of spiders is due to consumption of dipterans as food [39].

PLOS ONE
Gut microbiome of spiders Two endosymbiont genera, Wolbachia and Rickettsia of Phylum Proteobacteria are reported in the in the gut of spiders as well as in other arthropods such as insects [40][41][42][43] etc. These two genera are known for their reproductive manipulation ability in arthropods [44]. In the current study, Wolbachia was most abundant in 6 species, Leucage decorata, Leucage celebesiana, Opadometa fastigata, and Orsinome vethi of family Tetragnathidae, Hippasa greenalliae of Lycosidae and Eriovixia laglazei of Araneidae. While the gut of the Salticidae family members have no evidence of this genus contrary to earlier studies [45]. This genus is generally responsible for symbiotic associations, ranging from mutualism to parasitism in arthropods [46]. The genus Rickettsia has only been reported only in 2 species of the family Tetragnathidae (Orsinome vethi and Opadometa fastigata). The genus Prevotella_9 is reported in only one species of family Araneidae (Cyclosa mulmeiensis). This genus has been reported to be responsible for the larva metamorphosis into adults in the bark beetle [32]. The genera Paracoccus and Micrococcus have been reported in gut of Lycosids, ticks [47] and dipterans [48]. In the present study, two populations of Thainia bhamoensis (AA670, AA925) had few differences in the abundance of two genera Actinobacteria, and Pseudomonas. The genus Aeromonas was detected in AA925 and Klebsiella in AA670. Gut bacteria of spider families may be involved in carbohydrate metabolism, amino acid metabolism, fatty acids metabolism, cofactors and vitamin metabolism.

Conclusion
Our results showed that almost all the spider species belonging to four families had a very similar gut bacterial composition and structure. There were few exceptions like: six species (Hippasa greenalliae, Leucage decorata, Leucage celebesiana, Orsinome vethi, Opadometa fastigata, and Eriovixia laglazei) were infected by either Wolbachia or Rickettsia or both. The presence of these endosymbionts in turn affected the abundance profile of other gut bacteria in these species. Moreover the richness of gut bacterial diversity in the family Lycosidae was much higher than in Salticidae, Tetragnathidae and Araneidae. Furthermore, in order to understand the profile of microbial diversity and the relationship between host and habitat in spiders, extensive sampling and study of the gut microbiome is necessary.

Material and methods
The specimens were collected from the six Indian states of Assam, Chhattisgarh, Karnataka, Kerala, Odisha and West Bengal (Table 1). We used two methods of collection: by hand picking method under the wood logs, foliage of leaves, and stones; and by sweep net collection from the vegetation. Each specimen was placed in a separate 15 ml tube. After three to four hours, each specimen was transferred to 100% alcohol to be stored at -20˚C. All specimens were morphologically identified using published literature and available taxonomic keys. The species included in this study are not endangered or protected.

DNA isolation and 16srRNA amplicon sequencing
Before dissection, each specimen was washed three times with Milli-Q water to remove the environmental contamination. After washing, the abdomen of ten specimens of each species was removed and pooled to isolate the DNA. The DNA was isolated from pooled specimens of each species using a DNeasy Blood & Tissue Kit (Qiagen) as per manufacturer protocol.
The quantity and quality of the extracted DNA was quantified by Qubit 2.0 Fluorometer (Q32866, Thermofisher) and by agarose gel electrophoresis (Cell BioScience Alphalmager MINI). PCR reaction (Polymerase Chain Reaction) was carried out for the amplification of 36 extracted DNA of 35 spider species (one species has two geographical replicates) using primer sets of V3-V4 region of 16srRNA. The reaction mixture (25 μL) includes both forward and reverse primer (1 μl of each), Taq DNA polymerase (0.5 μl), dNTPs (1 μl), 10 × buffer (2.5 μl), DNA template, Milli-Q water. The reaction cycle involved the following steps: 5 min at 98˚C followed by 35 cycles for 30 s at 98˚C (denaturation), 45 s at 53˚C (annealing), and 72˚C for 45 s (elongation), and 7 min at 72˚C (final extension). The PCR products were visualized by agarose gels. The sequencing of V3-V4 region of 16srRNA was carried out on the Illumina MiSeq Platform. The raw reads were submitted under the Bio-Project ID PRJNA638522 with accession number SAMN15196477 to SAMN15196488, SAMN15580727 to SAMN15580750 to The National Center for Biotechnology Information (NCBI) GenBank Portal.

Bioinformatics and statistical analyses
The paired end raw reads (4441444) of 36 specimens from 35 species with minimum (121255) and maximum (125368) were generated. These reads were merged into single reads in QIIME2 (ver. 2019.10) using demultiplexing [49]. These single reads were quality filtered, trimmed, de-noised and merged by using a DADA2 pipeline in QIIME2 [50]. The chimeric reads were identified and removed and non-chimeric reads were grouped and assigned to Amplicon Sequence Variants (ASVs). QIIME2 q2-feature-classifier plugin was used for the taxonomic classification of these ASVs based on 99% similarity on SILVA database (version 132). The taxonomy file and feature tables were generated for the downstream analysis.
Further analysis was carried out on The MicrobiomeAanalyst (web based tool), using Marker Data Profiling (MDP) and recovered 13650 ASVs [51]. These ASVs were again subjected to singleton removal, low variance and low abundance features with a final recovery of 1282 ASVs. The unfiltered data was used to calculate the diversity measures. The Observed, Chao1, Shannon and Simpson diversity measures with T-test/ANOVA statistical methods were used for alpha diversity to calculate the diversity and richness. The PERMANOVA based statistical method for Bray Curtis and Ward's linkage based method for unweighted UniFrac distance measure were used for beta diversity to calculate the similarity of gut bacterial diversity in spider species. Nonmetric Multidimensional Scaling (NMDS) and Dendrogram were used for beta diversity analysis. We have used linear discriminant analysis (LDA) of effect size with cut off 0.2 and Log LDA score 4 parameters to discover biomarkers that differed among four spider families. Colour was used to indicate the different spider families.
A web based tool jvenn was used to view the unique and shared ASVs between the four spider families [52]. To predict the functional metabolic pathways between spider families, The Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) [53] based on Kyoto Encyclopedia of Genes and Genomes (KEGG) database was used [54][55][56]. We used Stamp software [57] for the plotting of predicted metabolic pathway between the four spider families.