Tissue damage induced midgut stem cell proliferation and microbial dysbiosis in Spodoptera litura

Abstract

In the past decade, gut microbiota has come to the fore in search for the cause of disregulation in intestinal homeostasis. Here, we report a possible link between gut microbial dynamics and stress-inducing factors using the leaf worm moth Spodoptera litura as a model organism. Investigation reveals that S. litura exhibits dysbiosis i.e. alteration in the gut microbiota composition that might induce or suppress inflammation upon exposure to dextran sulfate sodium salt, a tissue damaging agent (DSS, 40 kD). It primarily corresponds to an expansion of the bacterial phylotypes Enterobacter sp., Pseudomonas sp., Escherichia sp. and Acinetobacter sp. belonging to subclass Gammaproteobacteria. To assess the role played by gut residents in midgut inflammation, we re-colonized the axenic insects with Pseudomonas, Enterobacter and Acinetobacter individually. We observed that Pseudomonas and Enterobacter monoassociated insects exhibit inflammatory effects like damage to gut epithelium and hyperproliferation of stem cells under stress conditions. Conversely, Acinetobacter promotes fitness in larvae and reduces inflammatory effects of DSS. However, we failed to detect phenotypic inflammatory changes like midgut epithelium damage and stem cell proliferation in axenic insects reared on DSS-supplemented diet. Our results highlight that gut commensals that apparently remain low in abundance and benign under typical conditions can exert modulatory (positive or negative) effects on host fitness in the presence of stimulator.

dextran sulfate sodium, dysbiosis, gut microbiota, homeostasis, Spodoptera litura, stem cells

INTRODUCTION

In metazoans, it has become evident that microbial communities inhabit a wide range of biological niches like intestinal tract, oral cavity, body fluids, skin (Qin et al.2010) or even specialized structures such as bacteriocytes in insects (Gottlieb et al.2008; Pandey et al.2013; Pandey and Rajagopal 2016). This multitude of bacteria is conceived to play a pivotal role in several facets of the host's life like tissue homeostasis, immune response (Hooper, Littman and Macpherson 2012; Nicholson et al.2012), gut development (Rawls, Samuel and Gordon 2004), stimulation of stem cell turnover (Jones et al.2013) and synthesis of essential amino acids in insects (Douglas 1998). It is becoming increasingly evident that constant abnormalities in the endogenous gut microbiota known as ‘dysbiosis’ have been associated with pathogenesis of chronic gastrointestinal conditions and also other allergic or infectious disorders (Sartor 2008). The significance of microbiota has been directly shown by studies on animal models where colitis failed to be induced in axenic state but rapidly emerges when animals are reconstituted with normal gut flora (Okayasu et al.1990; Nell, Suerbaum and Josenhans 2010). Nevertheless, the origin and pathogenesis of such gastroinflammatory diseases is mostly unknown as the studies in an intricate system like mammals are usually complex, expensive and must adhere to more stringent ethical guidelines. Further, the functional role of each individual bacterium remains largely unclear. Additionally, in the absence of reliable mammalian stem cell surface markers available commercially, it becomes tedious to monitor intestinal stem cells and its regulation in response to microbial, chemical or environmental insult (Barker et al.2007; Demidov et al.2007; Scoville et al.2008). Thus, it becomes important to gain an in-depth insight into such complex host–microbial interaction involved in regulation of tissue homeostasis using other suitable host model.

In vertebrates and invertebrates, gastrointestinal tract is not only involved in major digestive and absorptive functions but also protects the internal gut milieu from external environment (Sansonetti 2004). The gut epithelium of both mammals and Lepidopteran insects comprises of a single layer of mature epithelial cells with interspersed stem cells that serve to replenish intestinal epithelium (Spies and Spence 1985; Baldwin and Hakim 1991; Crosnier, Stamataki and Lewis 2006; Yen and Wright 2006). In Spodoptera, midgut epithelium of a larva comprises of a highly folded pseudostratified epithelium having three major cell types, viz.: columnar-shaped enterocytes, pear-shaped goblet cells and spherical stem cells (Cioffi 1979). The midgut epithelium is separated from the food bolus by peritrophic membrane (PM), which is a film-like semi-permeable, non-cellular structure. It protects the midgut epithelium against food abrasion and microorganisms (Cioffi 1979). Midgut stem cells respond to stress signals and undergo repeated cycles of proliferation, division and differentiation to restore midgut epithelium integrity (Hakim et al.2009). Moreover, the relative simplicity of lepidopteran larva midgut microbiota complemented with several culturable bacterial species is exemplary of the power of lepidopteran as host model (Engel and Moran 2013). Additionally, based on 16S rRNA gene sequencing and microarray analysis, the microbiota in Spodoptera includes Clostridium, Lactobacilli etc. that are also prevalent in the human gut (Tang et al.2012). Further, the possibility of supplementing the artificial solid diet provided to them with toxins, drugs, antibiotics or bacterium of interest, to decipher their functional role and interaction with the midgut is appealing. Recently, draft genome sequence of Spodoptera has been sequenced that make it genetically well suited for gut bacteria studies (Kakumani et al.2014). Together, it suggests that Spodoptera forms a technically advantageous surrogate target system to dissect host–microbe interactions. In the past, studies in Lepidopterans have provided fascinating insights into the mechanism of viral transmission from midgut to body tissue (Engelhard et al.1994; Rahman and Gopinathan 2004). However, to date, a wealth of detail on host–midgut microbiota interaction in Lepidopterans has been derived from studies in reference to insecticide Bt toxin protein (Broderick et al.2009). On the contrary, there is a paucity of information on host defense mechanism and interaction with microbiota in response to other stress-inducing factors that can induce tissue inflammation.

In this study, we have used Spodoptera litura larvae as a model to analyze the significance of commensal bacteria in commencement or suppression of midgut inflammation during tissue injury inflicted by dextran sulfate sodium salt (DSS, 40 kD). Initially, 454 amplicon pyrosequencing analysis identified midgut bacteria that were differentially abundant in the DSS-treated and untreated larvae. We further isolated these differentially abundant midgut bacteria namely Pseudomonas, Enterobacter and Acinetobacter from conventionally reared larvae using routine microbiological techniques. We observed tissue damage and stem cell proliferation as the primary effect of feeding DSS in conventionally raised S. litura. On the contrary, axenic insect is not prone to intestinal injury inflicted by DSS. Strikingly, monoassociation of insect with specific bacterium either acts synergistically with DSS to induce midgut tissue injury or acts antagonistic to DSS action and hence prevents tissue inflammation. Thus, commensal bacteria residing in the same ecosystem behave differently during tissue injury by either assisting tissue inflammation or conferring protection to host.

MATERIALS AND METHODS

Insect rearing

Spodoptera litura larvae were reared on a standard chickpea flour-based diet under a photoperiod of 14:10 h (light:dark), 70% relative humidity and 27°C. For preparing sterile antibiotic diet, 100 μg/ml ampicillin, 50 μg/ml rifampicin, 37.5 μg/ml tetracycline and other heat sensitive components were added after autoclaving at 121°C for 15 min. For viability tests, we used 30 S. litura larvae for each treatment in three independent experiments. The chemicals included in the feeding medium were 3%–5% of dextran sulphate, sodium salt (MP Biomedicals (India) Pvt. Ltd, Pawane Village, Turbhe, Navi Mumbai, India). Spodoptera litura larvae that were still alive were transferred to fresh feeding media every day.

Establishment of flow cytometry gates

Early fourth instar larvae were surface sterilized in 70% ethanol, rinsed with sterile ringer solution (150mM NaCl, 5mM KCl, 3mM CaCl₂ and 1mM NaH₂Co₃) and dissected under aseptic conditions. Further, a longitudinal incision was given in the midgut region to remove all the food particles and peritrophic matrix. For midgut cell isolation, the protocol of Loeb and Hakim (Loeb et al.1999) with slight modifications was followed. Briefly, five cleaned midguts were microsectioned and incubated in 3 ml of Mitsuhashi and Maramorosch insect medium (M and M) for 90 min at room temperature. After incubation, midguts were flushed using wide bore pipettes and cells were sieved through 70 μm cell strainer (BD, Bioscience) into 50 ml conical tubes. Tubes were centrifuged at 400 × g, 4°C for 15 min to collect cell pellet. Subsequent to washing, cells were suspended in 1 ml of M and M insect medium. Stem and mature cell populations were separated from each other using Ficoll-Paque (GE Life Sciences) density gradient (Loeb et al.1999).

Calcein-AM (20 μM in DMSO) was used to stain purified population of stem and mature midgut cells. Staining was performed by incubating cells with freshly prepared working solution (5 μl of 20 μM solution per 1 × 10⁵ cells) in dark at room temperature for 20 min (Castagnola, Eda and Jurat-Fuentes 2011).

Stem and mature cells gate was established using FACSAria instrument (BD Biosciences). Cytometer settings were FSC (forward scatter) at 177 eV, SSC (side scatter) at 172 eV and FL1 (calcein fluorescence) at 209 eV. Threshold was set at 5000 to exclude debris. Gates were established using dot plot i.e. SSC on Y-axis for determining the cell complexity and FL1 (calcein-AM fluorescence) on X-axis for detecting fluorescence signal. Flow cytometry gates were established by conducting a minimum of three independent experiments with 1 × 10⁵ events considered for each experiment. Data were analyzed using FlowJo software.

Feeding experiments were carried out at 26°C + 1°C. For flow cytometry analyses, early fourth instar larvae were fed on control and 5% DSS- (40 kD, MP Biomedicals) supplemented diet for 3 days. Midgut cells were isolated from five fourth instar larvae of both groups for three consecutive days. Data analysis of 1 × 10⁵ cells acquired in cytometry gates was performed. The statistical significance of the data obtained from three independent experiments was calculated by conducting one-way analysis of variance (ANOVA) using SPSS (version 16.0) and PAST (version 3.03) software (Hammer, Harper and Rya 2001). ANOVA was performed at P < 0.050.

Quantization of DSS in Spodoptera litura feces

Spodoptera litura feces were collected after 3 days of DSS treatment and stored in cold acetone at –20°C until further analysis. The dried feces were rehydrated (50 mg/ml) at 4°C in 0.1 M sodium acetate buffer (pH 5.0) containing 5 mM EDTA. After 24 h, papain (Sigma) was added to the sample at 10 mg/ml and incubated at 60°C for 24 h. The mixture was then centrifuged at 3000 rpm for 30 min and the supernatant was collected. The concentration of sulfated polysaccharides was determined with a 1, 9- dimethylmethylene blue (DMB) assay. The DMB solution was prepared by dissolving 16 mg DMB, 3.04 g glycine, 2.37 g sodium chloride and 95 ml 0.1 M hydrochloric acid in 1000 ml distilled water. The pH of the final dye was adjusted to 3.0 (Farndale, Buttle and Barrett 1986). For this assay, 10 μl of each sample was mixed with 250 μl of DMB solution at room temperature and the absorbance was recorded at 525 nm. Standard curve was plotted using dilutions ranging from 0.2 μg/tube to 25 μg/tube. Fifty microgram of the sulfated polysaccharides isolated from larvae feces as well as DSS dissolved in buffer (control) were applied to 6% polyacrylamide slab gels at 200 mA for 45 min with a 0.1 M phosphate buffer (pH 11.5) containing 0.125 M formic acid (Hilborn and Anastassiadis 1969). The gel thus obtained was stained with 0.1% toluidine blue in 1% acetic acid.

Histology

Early third instar S. litura larvae were transferred to 5% DSS-supplemented diet for 8 days. These larvae were molt to form fifth instar stage after 8 days. Controls were reared on artificial lab diet from neonate stage. Larvae were starved for 1–2 h and thereafter anesthetized on ice for 15 min. Whole midgut was removed, rinsed in Ringer's solution and fixed in Bouin's fixative for 16–20 h. After fixation, the midgut was divided into three segments: anterior midgut, median midgut and posterior midgut. These specimens were embedded in paraffin wax; sections were cut to 5 μm thickness and stained with haematoxylin (Delafield's)—eosin. The stained sections were observed and photographed under Nikon Confocal light microscope. Hematoxylin and Eosin-stained midgut sections were coded for blind microscopic assessment of DSS-induced tissue damage. Histological scoring was based on three parameters namely damage to gut epithelium, damage to PM and damage to basement membrane. Damage scores for different parameters are shown in Table 1. Histological index (HI) was calculated as sum of scores assigned to different parameters. Values were added to a maximum damage score of 4. The HI was expressed as the mean ± Standard Error of Mean (SEM) for midgut section obtained from 10 larvae in each treatment from three independent experiments. Scores for tissue damage severity were analyzed using the Kruskal–Wallis test with significant difference at P < 0.001.

Table 1.

Damage score assigned to three different parameters used for calculating HI.

Feature	Score	Description
Damage to gut epithelium	0	None
	0.5	Few epithelial cells fallen into lumen
	1	Proliferating midgut epithelium
	1.5	Development of vacuoles between epithelial cells
	2	Highly corroded gut epithelium
Damage to PM	0	None
	0.5	Partially degraded PM
	1	Complete loss of PM
	0	None
Damage to basement membrane	0.5	Partially damaged basement membrane
	1	Complete loss of basement membrane

16S rRNA 454 amplicon pyrosequencing

Midgut samples were collected from two different sets. Firstly, early third instar larvae were fed with DSS-treated diet and dissected at fifth instar stage on 8th day. Secondly, neonate larvae were fed on DSS-supplemented diet till it reaches fifth instar stage and then dissected. Spodoptera litura larvae were rinsed with sterile MQ water. DNA was extracted from the midgut of fifth instar larvae (midgut from three larvae pooled together with each larvae obtained from different set of experiment) of each group using the HiPurA Insect DNA Miniprep Purification Kit according to manufacturer^’s protocol with minor modifications. The insect samples were homogenized thoroughly with a hand-held homogenizer. The insect homogenate was kept in 800 μl of lysis solution (provided with kit) containing 45 mg/ml lysozyme for 2 h at 37°C to ensure total extraction of DNA from both Gram-negative and Gram-positive bacteria. The DNA was then extracted according to kit's protocol. Purity and concentration of the DNA was analyzed by using the NanoDrop spectrophotometer (NanoDrop Technologies Inc., Wilmington, DE, USA). Further, DSS, which is a sulfated polysaccharides, acted as an inhibitor while processing of the samples for amplicon pyrosequencing. Hence, we used an inhibitor removal column to overcome the problem. After this, advance normalization has been performed to unify the amount of starting DNA across all the samples. One microliter of the sample (1 μl = 10 ng) has been used for sequencing. Insect midgut microbiota was determined by performing 454 amplicon pyrosequencing using protocol explained by Dowd et al. (2008). Briefly, 16S rRNA primers 27Fmod 5΄-AGRGTTTGATCMTGGCTCAG-3΄ and 519R 5΄-GTNTTACNGCGGCKGCTG-3΄ with a sample-specific bar code were used for DNA amplification. PCR was performed using HotStarTaq Plus Master Mix Kit (Qiagen, Valencia, CA). The PCR reaction conditions were 94°C for 3 min, followed by 28 cycles of 94°C for 30 s, 53°C for 40 s and 72°C for 1 min and after that a final elongation step at 72°C for 5 min was performed. Subsequently, Agencourt Ampure beads (Agencourt Bioscience Corporation, Massachusetts, USA) were used to purify amplicons. The amplicons sequencing was performed using Roche 454 FLX titanium instruments and reagents (MR DNA Lab, Texas).

FASTX tool kit was used for processing SFF files obtained by 454 amplicon pyrosequencing (Pearson et al.1997). Subsequent to adapter trimming, nucleotide sequences below Phred quality score of 20, homopolymer runs exceeding 6 bp, nucleotide sequence length <250 bp and reads containing ambiguous bases (N) were removed. Sequences were then denoised and chimeras removed. Subsequently, remaining reads were clustered into operational taxonomic unit (OTU) at 97% sequence similarity cutoff criteria using UPARSE pipeline (Edgar 2013). OTUs with <10 nucleotide sequence reads were included only if they met the criteria i.e. greater than three reads in total present in more than one sample (Morgan et al.2013). Each candidate OTU was assigned taxonomical status using BLASTn against a curated Greengenes database (DeSantis et al.2006).

Genus-based abundant matrix was used in a series of ecological and statistical evaluation. As the sequencing depth varies between samples, data normalization was performed by ‘Total count method.’ In the total count method, counts of a particular bacterial genus are divided by library size associated with the sample and multiplied by the mean total count across all the samples of the dataset. Normalized data were further square root transformed to reduce the contribution of abundant genus. Alpha diversity patterns: Shannon, Simpson and Chao-1 were calculated using R (http://www.r-project.org/) to estimate the species diversity and richness in a population. A heat map plot was generated using the matrix to emphasize the differential abundance of bacterial phylotypes among samples. The genus-based abundance dataset was used for calculating dissimilarity matrix across samples using Bray–Curtis distance measures. The bacterial genera enriched between any two treatments were identified by two-side Fisher's Exact test with Storey FDR multiple test correction using STAMP. A principal component analysis (PCA) based on Spearman's correlation matrix was performed to compare taxonomic profiles generated across samples.

Culture dependent isolation of bacteria from Spodoptera litura midgut

Whole midgut from individual larvae was dissected and homogenized in 1.5 ml microcentrifuge tubes containing 500 μl Ringer's solution. The homogenate obtained was diluted serially in tryptic soya broth (TSB) from 10⁻¹ to 10⁻¹⁰. Subsequently, dilutions were plated on TSB plate to obtain bacterial colonies. The plates were incubated at 30°C and surveyed consecutively for 3 days to assess development of new bacterial colonies. Unique bacterial colonies were picked based on morphological differences like color, texture, elevation, shape, size, etc. and streaked on new TSB plates. For identification, single purified colonies were boiled in 50 μl of MQ water for 15 min in a boiling water bath and later centrifuged at 10 625 × g for 10 min. A total of 10 μl of the supernatant was diluted with 90 μl of MQ water. Five microliter of DNA soup was used in a 25 μl PCR reaction mix. The 16S rRNA was amplified by using universal primers 27F and 1492R (Lane 1991). The eluted PCR product was sequenced using a 27F universal primer.

Bacterial colonization

Pseudomonas, Enterobacter and Acinetobacter were cultured by shaking overnight at 30°C in half strength TSB. The overnight grown culture was washed thoroughly and re-suspended in PBS (pH 8.0) before use. Early neonate larvae were introduced to sterile antibiotic diet for 3 days to make them axenic. The germ-free status of the S. litura larvae was validated by homogenizing the midgut under aseptic conditions and plating them on TSB agar plates. The plates were incubated overnight and observed for any sign of bacterial colonies for next 24 h. Larvae were then transferred to sterile diet augmented with 150 μl/g overnight grown culture (2.5 A₆₀₀) of bacteria for 3 days. To test for the colonization ability of bacteria, fecal matter was obtained from the larvae and homogenized in TSB media and was plated on TSB agar plates. A total of 10–15 bacterial colonies were randomly picked and analyzed in two independent experiments. DNA isolation, purification and sequence analysis of the bacterial colonies thus obtained were performed as described above. Following this, bacterium monoassociated insects were offered 5% DSS containing diet. Controls were fed on sterile diet without DSS after colonization with specific bacterium. Dissection, data acquisition and statistical analyses for flow cytometry were performed as described above. For examining midgut histopathology, bacterium monoassociated larvae were dissected after 8 days of feeding DSS-supplemented diet. The midgut sections were obtained from 10 larvae of both groups in three independent experiments.

Accession number

The 16S rRNA pyrosequencing data have been deposited in NCBI database under the accession number: SRP065112. 16S rRNA Sanger sequencing data of culturable bacteria have been deposited in NCBI database under the accession number KT991636 to KT991639.

RESULTS

Dose-dependent killing of Spodoptera litura larvae by DSS

We tested the idea that DSS could cause tissue damage and stem cell proliferation in Spodoptera litura larvae midgut. When the diet contained 5% of DSS (average MW 40 kDa), approximately 40% of the flies died within 8 days (Fig. 1, Supporting Information). Together with the killing curves for control (untreated) and diet containing 3% DSS (3% DSS), we observed a dose-dependent killing of S. litura larvae by DSS.

Quantization of sulfated polysaccharides in feces of Spodoptera litura larvae

Standard curve for DSS was plotted. A near linear graph was obtained with R² = 0.9912 (Fig. 2A, Supporting Information). The concentration of the sulfated polysaccharides in the unknown feces sample was found to be 50 μg/μl. The electrophoretic patterns of sulfated polysaccharides obtained from feces 3 days after the treatment with DSS was similar to that of DSS dissolved in buffer (Fig. 2B, Supporting Information). A significant amount of DSS was not recovered from the feces of control S. litura larvae.

Richness and diversity indices of the microbiota associated with Spodoptera litura larvae

Pyrosequencing yielded 38 901 total sequences; after quality control, 35 114 high-quality sequencing reads were obtained and utilized for further analysis. The number of curated sequences (>250 bp) obtained per treatment ranged from 2220 to 17 387 (Table 2). The rarefaction curves were plotted on normalized data for larvae exposed to different diet. Initially, as the most common species were found, rarefaction curves tend to grow rapidly; however, the curve plateau as only the rare species remain to be sampled. The rarefaction curve for all the treatments namely untreated (S. litura larvae reared on an artificial lab diet from neonate stage), 3% DSS-8 days (early third instar S. litura larvae reared on 3% DSS diet for 8 days), 5% DSS-8 days (early third instar S. litura larvae reared on 5% DSS diet for 8 days) and 3% DSS-Neonate (S. litura larvae reared on 3% DSS diet from neonate stage) reached the plateau stage indicating that sequencing depth was enough to determine the major bacterial genera (Fig. 1A). Genus-based abundance matrix was used to calculate alpha diversity indices like Shannon, Simpson and Chao-1. The comparison of biodiversity index among treatment showed that larvae reared on DSS diet from neonate stage (3% DSS-Neonate) exhibited lowest median SDI (Shannon Diversity Index) suggesting that it was poor in terms of bacterial diversity. However, early third instar larvae reared on 5% DSS diet (5% DSS-8 days) exhibited maximum SDI value when compared to normal (Table 2). Diversity indices values for all the measures between the treatments were found to be highly significant (P < 0.001; Table S1, Supporting Information).

Figure 1.

Rarefaction curve analysis based on normalized abundance data of bacterial genera at 97% sequence similarity and microbial diversity analysis. (A) (1) Untreated—S. litura larvae reared on an artificial lab diet from neonate stage; (2) 3% DSS-8 days—early third instar S. litura larvae reared on 3% DSS diet for 8 days; (3) 5% DSS-8 days—early third instar S. litura larvae reared on 5% DSS diet for 8 days; (4) 3% DSS-Neonate—S. litura larvae reared on 3% DSS diet from neonate stage; (B) Dual dendrogram of top 16 bacterial genera obtained from S. litura midgut across four different treatments. Genera and sample categories were clustered using Euclidean distance metric. Color scale is representing the relative abundance of sequence reads (normalized by square root transformation); (C) PCA (PC1 and PC2) based on normalized abundance data of bacterial genera. Treatments with similar microbial composition occupy the same spot in the scatter biplot. Untreated—S. litura larvae reared on an artificial lab diet from neonate stage, 3% DSS-8 days—early third instar S. litura larvae reared on 3% DSS diet for 8 days, 3% DSS-Neonate—S. litura larvae reared on 3% DSS diet from neonate stage and 5% DSS-8 days—early third instar S. litura larvae reared on 5% DSS diet for 8 days.

Table 2.

Measurement of alpha diversity indices based on genus-based abundant matrix from S. litura midgut samples. Untreated—S. litura larvae reared on an artificial lab diet from neonate stage, 3% DSS-8 days—early third instar S. litura larvae reared on 3% DSS diet for 8 days, 3% DSS-Neonate—S. litura larvae reared on 3% DSS diet from neonate stage and 5% DSS-8 days—early third instar S. litura larvae reared on 5% DSS diet for 8 days.

Treatment	Shannon_H index (mean ± standard deviation)	Simpson_1-D index (mean ± standard deviation)	Chao-1 (mean ± standard deviation) at 5%	No. of curated reads	Normalized data used for analysis total count normalized/square root transformed normalized (rounded off)
Untreated	1.69 ± 0.02	0.766 ± 0.003	26.33	17 387	8779/280
3% DSS-8 days	1.77 ± 0.02	0.771 ± 0.005	20	8649	8779/280
3% DSS-Neonate	1.34 ± 0.02	0.618 ± 0.009	14	2220	8779/231
5% DSS-8 days	2.02 ± 0.02	0.827 ± 0.003	20	6858	8779/307

Comparative bacterial diversity analyses in Spodoptera litura larvae fed on control and DSS-treated diet

The pyrosequencing dataset obtained for different treatment got clustered into 32 major bacterial genera (Table S2, Supporting Information). Further, heat map depicts that several bacterial genera exhibited notable differences between the treatments like Enterobacter and Escherichia were abundant in the larvae fed on DSS diet from the neonate stage, while Pseudomonas and Acinetobacter were prominent in larvae fed on DSS-supplemented diet for 8 days. The dual dendrogram based on Euclidean distance metric also demonstrates that samples from the treatment 3% and 5% DSS-8 days are similar in overall bacterial composition (Fig. 1B). Out of the 32 bacterial genera obtained, only 15 taxa accounting for maximum variance were further analyzed by multivariate technique called PCA. By this procedure, two highly significant components were obtained that together account about 95.69% of total data variance. The first factor explained 74.60% of data variation and the major taxa representing the first component were Enterobacter, Escherichia, Methylobacterium, Aeromicrobium, Burkholderia, Ralstonia, Sphingopyxis and Lysinimonas. Similarly, the second factor contributed 19.14% to data variation and the related taxa were Pseudomonas, Herbaspirillum and Comamonas. The third components led to 6.24% variation in total data, and the major taxa influencing the observed variation were Acinetobacter, Microbacterium, Bradyrhizobium and Bacillus. Further, scatter bi-plot depicts that a distinct bacterial phylotypes could be specified to each sample. For example, genera like Enterobacter and Escherichia were dominated in larvae fed on DSS diet from neonate stage (Fig. 1C).

A Bray–Curtis dissimilarity matrix was calculated to quantify the compositional dissimilarity among the larvae from different treatment. The values for matrix are bound between 0 and 1, where a value of 0 indicates that the two samples are similar in composition, whereas 1 indicates that they do not share any common bacterial species. A noteworthy observation was that early third instar larval population fed on DSS diet for 8 days (3% DSS-8 days or 5% DSS-8 days) exhibited minimal deviation (Bray–Curtis value 0.212–0.228) from those raised on lab diet without DSS (untreated). While, population kept on DSS diet from neonate stage (3% DSS-Neonate) exhibited maximum deviation (Bray–Curtis value 0.558–586) from other population (Table 3).

Table 3.

Pair-wise comparison of community composition between four treatments using Bray–Curtis dissimilarity matrix. Untreated—S. litura larvae reared on an artificial lab diet from neonate stage, 3% DSS-8 days—early third instar S. litura larvae reared on 3% DSS diet for 8 days, 3% DSS-Neonate- S. litura larvae reared on 3% DSS diet from neonate stage and 5% DSS-8 days—early third instar S. litura larvae reared on 5% DSS diet for 8 days.

Treatment	Untreated	3% DSS-8 days	3% DSS-Neonate	5% DSS-8 days
Untreated	0	0.21278	0.58172	0.22819
3% DSS-8 days	0.21278	0	0.55853	0.13734
3% DSS-Neonate	0.58162	0.55853	0	0.5866
5% DSS-8 days	0.22819	0.13734	0.5866	0

Further, Fisher's Exact test with Storey FDR multiple test correction demonstrated that bacterial phylotypes Enterobacter and Escherichia were significantly abundant (P < 0.05) in S. litura larvae fed on DSS diet from neonate stage as compared to other samples. Similarly, bacterial genera Pseudomonas was significantly abundant in larvae fed on 3% or 5% DSS diet for 8 days (Fig. 2A, B, C, D and E).

Figure 2.

Bioinformatics analysis of the microbiome. Extended error bar plot depicting bacterial phylotypes where Fisher's two-side exact test with Storey FDR multiple test correction produces a significant P-value > 0.05. (A) Blue bar denotes 3% DSS-8 days, brown bar denotes untreated; (B) blue bar denotes 5% DSS-8 days, brown bar denotes Untreated; (C) blue bar denotes 3% DSS-Neonate, brown bar denotes 5% DSS-8 days; (D) blue bar denotes 3% DSS-Neonate, brown bar denotes untreated; (E) blue bar denotes 3% DSS-8 days, brown bar denotes 3% DSS-Neonate. Untreated—S. litura larvae reared on an artificial lab diet from neonate stage, 3% DSS-8 days—early third instar S. litura larvae reared on 3% DSS diet for 8 days, 3% DSS-Neonate—S. litura larvae reared on 3% DSS diet from neonate stage and 5% DSS-8 days—early third instar S. litura larvae reared on 5% DSS diet for 8 days.

Isolation of culturable isolate from Spodoptera litura larvae

We next attempted to isolate these bacteria from S. litura larvae (fed on routine artificial diet) by adopting routine microbiological techniques. 16S rRNA analysis revealed the identity of these bacteria as Acinetobacter, Enterobacter, Pseudomonas and Bacillus (Table 4). Out of the four bacterial genera obtained, Acinetobacter, Enterobacter and Pseudomonas were used to study the functional role of gut microbiota during tissue inflammation in insects.

Table 4.

Phylogenetic affiliation of bacteria isolated by culturable method from midgut of S. litura based on 16S rRNA nucleotide sequence.

Bacterial division	Isolate no.	Nearest match	Accession no.a	Accession no.b	Percentage similarity
Gammaproteobacteria	SL1	Acinetobacter	KJ569367	KT991636	100%
Gammaproteobacteria	SL2	Enterobacter	KT818802	KT991637	100%
Gammaproteobacteria	SL3	Pseudomonas	KU942677	KT991638	100%
Firmicutes	SL4	Bacillus	KJ719438	KT991639	99%

a

Nucleotide sequence of 16S rRNA taken from GenBank database under their accession numbers.

b

Nucleotide sequence of 16S rRNA from this study under their accession numbers.

Monitoring tissue damage and stem cell proliferation under stress condition in conventionally raised Spodoptera litura constituting healthy gut microbiota

Histopathological structural changes of the dissected midguts from fifth instar larvae were observed by haematoxylin and eosin staining. Midgut section of healthy larvae fed on artificial lab diet displayed an ordered monolayer of mature epithelium with interspersed stem cells. The midgut epithelium is sandwiched between intact basement membrane on the hemocoel side and PM towards the lumen (Fig. 3A). However, histopathological examination revealed major signs of tissue damage in the median midgut of larvae feeding on 5% DSS-supplemented diet. Contrarily, both anterior and posterior midgut exhibit minimal histological alterations. Midgut tissue sectioning revealed phenotypic changes like degeneration of the PM (Fig. 3B) with hyperproliferation in some parts of the midgut epithelium (Fig. 3C and D), development of vacuoles between mature cells (Fig. 3E) and epithelial cell loss with basement membrane degeneration (Fig. 3F). Though the feeding period was kept fixed as 8 days for these studies and tissue damage was detected in more than 95% of insect, yet we observed a difference in the intensity of the damage. Hence, a ranking system of HI was used for evaluating the tissue damage. The Kruskal–Wallis test showed significant differences (P = 0.0001) between the HI of untreated- and DSS-treated larvae (Table 5). Subsequently, we employed flow cytometry gate for ascertaining increase in stem cells count on feeding DSS-supplemented diet. In developing this technique, two parameters were considered namely distinct-light scattering and vital staining property of each cell type on staining with calcein-AM. As observed, mature cells (red) had a complex morphology due to a dense nucleus than stem cells (blue) having a clear nucleus, thus resulting in higher SSC readings (Fig. 4A). Additionally, Calcein-AM stained the stem cells (blue) more intensely due to its high esterase activity compared to mature cells (red; Fig. 4B). Thus, dual parameters, i.e. SSC on Y-axis for determining cell complexity and FL-1 (Calcein-AM) on X-axis for detecting fluorescence emitted by cells were used to establish cytometry gates for discriminating mature (red cluster) and stem cell (blue cluster) populations (Fig. 4C). By this gating, we were able to detect 90% mature (red; Fig. 4D) and 70% purified stem cells (green; Fig. 4E) to their established gate. Flow cytometry analyses of midgut cell population ascertained increase in stem cells count on feeding DSS-supplemented diet. The average percentage of stem cell number for DSS-treated larvae on day 1 exhibited a significant 6.94% increase in stem cell count on day 1 (P < 0.001) as compared to control. Similarly, on day 2 and day 3, the difference in percentage of mean stem cell number between treated and control samples was 2.20% and 3.51%, respectively (P < 0.05 and P < 0.001; Fig. 4F).

Figure 3.

Microphotograph of midgut section of control S. litura larvae (HI = 0) and structural changes in the midgut section of S. litura larvae on feeding DSS-supplemented diet (HI = 1, 2, 3 and 4). (A) Midgut epithelium of healthy S. litura larvae depicting intact basement membrane (BM), columnar cells (CC), goblet cells (GC), peritrophic membrane (PM), clear midgut lumen (double arrow; damage score 0; HI = 0; ×100); (B) Microphotograph depicts degrading PM (black arrows, damage score = 0.5) and few damaged epithelia cells falling into lumen (damage score = 0.5; HI = 1; ×100); (C) Microphotograph depicts proliferating midgut epithelium (rectangular box, damage score = 1) and complete loss of PM (damage score 1; HI = 2; ×100; D) proliferating midgut epithelium at higher magnification (region within box in Fig. 3C; ×200); (E) Midgut section of DSS treated insect depicting increased intercellular spaces between damaged cells (black arrows, damage score = 1.5), complete loss of PM (damage score = 1) and partial damage to basement membrane (damage score = 0.5; HI = 3; ×100); (F) Microphotograph depicts corroded epithelium with mature cells shedding into lumen (red arrows), lumen filled with cell debris (double arrows, damage score = 2), ruptured PM (black arrows, damage score = 1) and damaged basement membrane (rectangular box, damage score = 1; HI = 4; ×100). Tissue sections were stained with haematoxylin and eosin stain. Microphotographs are obtained after conducting a minimum of three independent experiments with ten larvae each. Section thickness = 5μm; scale reference bar = 100 μm except in Fig. 3D where scale reference bar = 50 μm.

Figure 4.

Establishment of flow cytometry gates for S. litura larvae midgut cells. (A) Representative plot for comparing side scatter (SSC) of midgut mature (red) and stem cells (blue); (B) Representative plot for comparing fluorescence intensity (calcein-AM fluorescence) of midgut mature (red) and stem cells (blue); (C) Plot comparing midgut mature and stem cells in S. litura larvae. Midgut mature cells (red cluster) have low fluorescence intensity (low calcein-AM fluorescence) and high SSC as compared to midgut stem cells (blue cluster) having high fluorescence (high calcein-AM fluorescence) and low SSC; Establishment of flow cytometry gates for discriminating S. litura larvae calcein-AM stained midgut cells (D) mature (red) and (E) stem (green) were stained independently with calcein-AM and detected using a fluorescence filter; (F) S. litura larvae fed on the DSS diet exhibited increase in stem cell count as compared to control. Feeding experiments were performed using 5% DSS in artificial solid diet. Flow cytometry gates were established by conducting a minimum of three independent experiments with 1 × 10⁵ events considered for each experiment. Midgut cells were isolated from five larvae of control and DSS treated groups. 1 × 10⁵ cells were counted in a flow cytometer after each day for three successive days. The numbers above the bar represent mean percentage of stem cells from three independent experiments. Error bars equates to standard error of mean. N = larvae fed on the artificial diet, T = larvae fed on DSS treated diet; asterisk sign presents the significant P value of each day—P* < 0.05, P*** < 0.001.

Table 5.

Comparative HI of midgut tissue damage induced by DSS with P-values. Feeding experiments were performed using 5% DSS in artificial solid diet. The midgut sections were obtained from 10 larvae in each treatment from three independent experiments. HI (in parenthesis) have been summarized as Mean ± SEM. N = larvae fed on the artificial diet, T = larvae fed on DSS-treated diet; N_S = axenic larvae fed on the sterile diet, T_S = axenic larvae fed on DSS-treated sterile diet; N_P = larvae monoassociated with Pseudomonas fed on the sterile diet, T_P = larvae monoassociated with Pseudomonas fed on DSS-treated sterile diet; N_E = larvae monoassociated with Enterobacter fed on the sterile diet, T_E = larvae monoassociated with Enterobacter fed on DSS-treated sterile diet; N_A = larvae monoassociated with Acinetobacter fed on the sterile diet, T_A = larvae monoassociated with Acinetobacter fed on DSS-treated sterile diet. Values in bold represent significant P-values, NS = non-significant and empty cells represent repeated P-values.

Treatment	T (2.3 ± 0.29)	NS (0.2 ± 0.11)	TS (0.37 ± 0.14)	NP (0.33 ± 0.12)	TP (2.2 ± 0.25)	NE (0.27 ± 0.19)	TE (1.6 ± 0.1)	NA (0.13 ± 0.09)	TA (0.27 ± 0.12)
N (0.2 ± 0.11)	P = 0.0001	NS	NS	NS	P = 0.00005	NS	P = 0.0001	NS	NS
T (2.3 ± 0.29)		P = 0.0001	P = 0.0006	P = 0.0003	P = 1	P = 0.0002	NS	P = 0.00009	P = 0.0002
NS (0.2 ± 0.11)			NS	NS	P = 0.00005	NS	P = 0.0001	NS	NS
TS (0.37 ± 0.14)				NS	P = 0.0001	NS	P = 0.0007	NS	NS
NP (0.33 ± 0.12)					P = 0.00007	NS	P = 0.0003	NS	NS
TP (2.2 ± 0.25)						P = 0.00006	NS	P = 0.00004	P = 0.00006
NE (0.27 ± 0.19)							P = 0.0002	NS	NS
TE (1.6 ± 0.1)								P = 0.00008	P = 0.0002
NA (0.13 ± 0.09)									NS

Repression of DSS-induced tissue damage by administration of antibiotics

Further, we intend to investigate the role played by gut microbiota in DSS-induced tissue damage. To examine, axenic early third instar larvae were transferred to sterile diet with or without DSS for 8 days. Tissue sectioning did not reveal major inflammatory changes in midgut epithelium of axenic insect reared on sterile artificial diet with or without DSS (Fig. 5A and B). There was no statistical difference in HI between the two groups (Table 5). Thus, absence of bacteria suppresses the tissue damage even in the presence of inflammatory substance. Further, we transferred the larvae that have been fed on unsterilized DSS diet from the neonate stage (for 10 days) to an antibiotics-supplemented diet for 5–6 days. Histopatholoical examination of these insects revealed that administration of antibiotics repressed the tissue damage caused by DSS. The microphotograph obtained for such larvae exhibits recovery and shows minimal degree of tissue damage when compared to larvae that kept on feeding DSS supplemented without antibiotics treatment diet from neonate stage (Fig. 5C and D).

Figure 5.

Histopathological structural changes in the midgut section of axenic S. litura larvae. The tissue sections were stained with haematoxylin and eosin stain. (A) Cross section of midgut from axenic S. litura larvae fed on sterile diet without DSS (×100; HI = 0). The arrow (black) denotes intact basement membrane, arrow (red) points towards smooth PM, double-arrow indicates monolayer of midgut epithelium; (B) Cross section of midgut from axenic S. litura larvae fed on sterile diet supplemented with 5% DSS (×100; HI = 0). The morphology of the midgut structure remains intact except slight proliferation at certain regions. (C) Cross section of midgut from S. litura larvae reared on 5% DSS supplemented for 10 days and then transferred to an antibiotic-supplemented diet for 5–6 days (×100; HI = 1). The microphotograph depicts recovery from tissue damage after being administered antibiotics. (D) Cross section of midgut from S. litura larvae fed on 5% DSS supplemented diet from neonate stage up to early fifth instar stage (×100; HI = 4). The microphotograph clearly depicts damaged basement membrane, complete loss of PM and highly corroded midgut epithelium. Microphotographs are obtained after conducting a minimum of three independent experiments with 10 larvae each. Section thickness = 5μm, scale reference bar = 100μm.

Functional role of Pseudomonas, Enterobacter and Acinetobacter in Spodoptera litura midgut tissue inflammation under stress condition

Investigations were done on axenic insect monoassociated with Pseudomonas, Enterobacter or Acinetobacter individually. Such larvae were offered DSS-treated diet with controls reared on DSS-free diet after re-colonization with bacterium. A 16S rRNA gene sequence analysis of fecal matter obtained from the control larvae after bacterial colonization exhibits successful establishment of bacteria in the insect. Pseudomonas (Figs 6A and B) and Enterobacter (Fig. 6C) monoassociated insect exhibited inflammatory changes like midgut epithelium damage and hyperproliferation on feeding DSS for 8 days as compared to their control. On the contrary, those re-colonized with Acinetobacter exhibited tolerance towards DSS and their midgut section appeared to be healthy (Fig. 6D). The statistical data were in agreement with histological alterations. The HI was significantly higher (P < 0.0001) in treated larvae colonized with Pseudomonas sp. and Enterobacter sp. as compared to their control (Table 5). However, there was no significant statistical difference between treated larvae colonized with Acinetobacter and its control (Table 5). Further, the HI exhibits that conventionally reared larvae or larvae colonized with either Enterobacter or Pseudomonas were more prone to gut tissue damage on feeding DSS-treated diet when compared to larvae colonized with Acinetobacter (Table 5).

Figure 6.

Histopathological structural changes in the midgut section of axenic S. litura larvae re-colonized with different bacterial isolate. The tissue sections were stained with haematoxylin and eosin stain. (A) Cross section of midgut from Pseudomonas monoassociated S. litura larvae fed on DSS-supplemented diet (×200; HI = 2). The arrow (black) denotes degrading PM, arrow (red) points towards the mature cells falling off into the lumen; (B) Cross section of midgut from Pseudomonas monoassociated S. litura larvae fed on DSS-supplemented diet (×100; HI = 2.5). The arrow shows degrading PM whereas doublearrow denotes proliferating gut epithelium; (C) Cross section of midgut from Enterobacter mono associated S. litura larvae reared on DSS-supplemented diet (×200; HI = 1.5). The double arrow denotes proliferating gut epithelium and partially damaged PM; (D) Cross section of midgut from Acinetobacter monoassociated S. litura larvae fed on DSS-supplemented diet (×200; HI = 0). The arrows (red) depict intact PM, arrows (black) denote smooth basement membrane, double arrow denotes midgut epithelium and a black line displays clear midgut lumen. Microphotographs are obtained after conducting a minimum of three independent experiments with 10 larvae each. Section thickness = 5μm, scale reference bar = 100μm.

Further, flow cytometry analysis depicts that axenic insect fed on DSS diet did not show any significant change in the number of stem cells as compared to its control i.e. the larvae fed on axenic diet without DSS (Fig. 7A). On the contrary, in conventionally raised larvae, DSS treatment leads to an increase in the number of stem cells as seen above (Fig. 4F). This suggests that indigenous gut microbiota is essential for inflicting an inflammatory response in host. Further, insect re-colonized with Pseudomonas (Fig. 7B) and Enterobacter (Fig. 7C) exhibited a significant increase in stem cell number following DSS feeding. On the contrary, larvae fed on DSS diet after being colonized with Acinetobacter (Fig. 7D) did not show significant change in stem cell number.

Figure 7.

Monitoring stem cell proliferation in S. litura larvae re-colonized with different bacterial phylotypes. (A) Axenic S. litura larvae did not exhibit increase in the number of stem cells on feeding 5% DSS-supplemented diet; (B) Axenic S. litura larvae re-colonized with Pseudomonas exhibits increase in the number of stem cells on feeding 5% DSS-supplemented diet; (C) Axenic S. litura larvae re-colonized with Enterobacter exhibits increase in the number of stem cells on feeding 5% DSS-supplemented diet; (D) Axenic S. litura larvae re-colonized with Acinetobacter did not exhibit increase in the number of stem cells on feeding 5% DSS-supplemented diet. 1 × 10⁵ cells were counted in a flow cytometer each day for three consecutive days. The numbers above the bar represent mean percentage of stem cells from three independent experiments. Error bars equate to standard error of mean. N_S = axenic larvae fed on the sterile diet, T_S = axenic larvae fed on DSS-treated sterile diet; N_P = larvae monoassociated with Pseudomonas fed on the sterile diet, T_P = larvae monoassociated with Pseudomonas fed on DSS-treated sterile diet; N_E = larvae monoassociated with Enterobacter fed on the sterile diet, T_E = larvae monoassociated with Enterobacter fed on DSS-treated sterile diet; N_A = larvae monoassociated with Acinetobacter fed on the sterile diet, T_A = larvae monoassociated with Acinetobacter fed on DSS-treated sterile diet; ns ₌ non-significant, asterisk sign presents the significant P-value of each day—P*** < 0.001, P** < 0.01 and P* < 0.05.

DISCUSSION

In the present work, we have examined the potential role of native midgut microbiota of Spodoptera litura larvae in commencement or suppression of tissue inflammation in the presence of dextran sulfate sodium salt (40 kD). DSS, a sulfated polysaccharide, has been successfully used to induce colitis in small experimental animals like Wistar rats (Chiba 1993), mice (Okayasu et al.1990), guinea pigs (Iwanaga et al.1994) and hamsters (Ohkusa 1985). The spread of colitis in these mammals involves complex interaction of lymphocytes, defective immune response and the gut microbiota. However, Lepidopterans do not have gut-associated immune cells (Jiang, Vilcinskas and Kanost 2010), thus, histological and flow cytometry examination of S. litura midgut tissue revealed extensive epithelial cell loss and stem cell proliferation as the only major intestinal inflammatory alteration upon exposure to DSS. As stem cells are the only cells capable of self-renewal (Hakim et al.2009), stress conditions like epithelium damage probably stimulate them to restore the deteriorating gut epithelium. Further, as seen in our results (Fig. 4F), during flow cytometry analysis the number of stem cells decreased from day one to day three after third moult i.e. during fourth instar stage. Stem cells greatly increase in number at the beginning of a molt, differentiate and intercalate between existing mature epithelial cells in such a way as to perpetuate the larval midgut cell pattern (Baldwin and Hakim 1991). In this way, the larval gut is enlarged at the moult to accommodate growth in the next instar. Thus, our results demonstrate that a fine balance exists between cellular damage induced by chemical stress and intestinal stem cell proliferation to achieve effective intestinal homeostasis. In this regard, intestinal epithelium renewal can be considered as an essential host defense mechanism under stress conditions. Recently, it has been shown that low doses of Cry1Aa toxin protein, DSS or bacterial infection were not lethal to insect midgut integrity; instead, they stimulate stem cell proliferation and differentiation that contributes to midgut epithelium renewal (Amcheslavsky, Jiang and Tony 2009; Apidianakis et al.2009; Tanaka, Yoshizawa and Sato 2012). Further, as discussed above, midgut tissue injury is often correlated with hyperproliferation of intestinal regenerative cells. However, due to the absence of reliable mammalian stem cell surface markers (Barker et al.2007; Demidov et al.2007; Scoville et al.2008), it was difficult to determine whether the proliferated cells were intestinal stem cells. In S. litura, using flow cytometry, we could monitor the surge in the turnover of midgut stem cells on tissue injury inflicted by DSS. To date, only a single published report has successfully described the establishment of flow cytometry gate for discriminating midgut cell types of lepidopteran insects using Heliothis virescens as a model organism (Castagnola, Eda and Jurat-Fuentes 2011). Thus, in agreement with previous report, our results confirmed the validity of flow cytometric gates for discriminating midgut cells in S. litura and present a simple model system to study crucial physiological phenomenon such as stem cell mediated tissue remodeling using flow cytometry.

Further, we performed a bacterial profiling via pyrosequencing to determine the bacterial phylotypes actively participating during midgut inflammation. Strikingly, bacteria profiling depicted microbial dysbiosis wherein a loss of balance between protective and harmful microbes was observed between DSS treated and untreated insect. DSS-treated insect midgut displayed complete absence of bacterial phylotypes belonging to genera Clostridium. In insects, its role has been well documented in nutrition and defense against pathogens (Engel and Moran 2013). Contrarily, 16S rRNA sequencing revealed a temendous increase in bacterial phylotypes Escherichia, Enterobacter and Pseudomonas belonging to subclass Gammaproteobacteria. E. coli strains routinely colonize the human intestine and are maintained at very low level in the gut of healthy human being. However, E. coli strains when overrepresented, could contribute significantly to either acute or chronic gastrointestinal disorders (Mukhopadhya et al.2012). Highly adherent, invasive and pathogenic strains of E. coli known as AIEC (adherent-invasive Escherichia coli) were usually isolated from intestinal biopsies and mucosa of patients suffering from gastrointestinal pathology (Martin et al.2004; Sasaki et al.2007; Martinez-Medina et al.2009). However, the direct support for the role of intestinal flora in midgut inflammation came from the experiments conducted in axenic insects. Both histological and flow cytometry examination of midgut tissue did not exhibit epithelial perturbation and stem cell proliferation respectively in axenic S. litura larvae reared on DSS free or DSS supplemented diet. Further, administration of antibiotics after DSS treatment caused the larvae to recover from the tissue damage. This suggests that indigenous bacteria are essential for host tissue susceptibility to inflammatory agent. Recently, it has been shown that in the absence of Enterobacter, an indigenous gut bacterium of various Lepidopterans, the pathogenicity of the B. thuringiensis toxin gets reduced against the host (Broderick et al.2009).

Further, axenic S. litura larvae when re-colonized with Enterobacter and Pseudomonas individually displayed ruptured PM, mature cell loss and hyperproliferation of midgut epithelium upon exposure to DSS. In Lepidopterans, Enterobacter has been established to aid in the toxicity of Bacillus thuringiensis (Broderick et al.2009). This suggests that high abundance of this bacterium might have contributed to the overall microbiota load during intestinal inflammation in S. litura. Further, in insects, Pseudomonas has been documented to increase the susceptibility of midgut epithelium to toxin (Engel and Moran 2013). In humans, Pseudomonas fluorescens has been reported to activate or triggers Crohn's disease (Sutton et al.2000; Wei et al.2002; Nagalingam and Lynch 2012). Thus, we hypothesize that Pseudomonas might play a significant role in the initiation of midgut inflammation in insects; however, other enteric bacteria like Enterobacter and Escherichia begin to rise and perpetuate the gut inflammatory response. In most of the studies, it's the Gammaproteobacteria that has been found to be consistently associated with the inflamed gut. The studies are marked by an increase in the members of the genus belonging to bacterial division Gammaproteobacteria during inflammation. However, none of the study conducted to date has been able to identify a single bacterial genus that is responsible for inflammation of the gut (Mukhopadhya et al.2012). Further, it can be said that severe tissue injury inflicted by DSS alters the gut environment for microbes. In such injured midgut tissue, host enteric bacteria like Escherichia coli, Pseudomonas and Enterobacter rapidly adapt to new gut environment and perpetuate the inflammatory reaction by deriving their nutrition from sloughed off mature cells to gain a rapid growth. Hence, dysbiosis due to environmental stress could result in susceptibility of host to intestinal inflammation.

On the contrary, axenic larvae reconstituted with Acinetobacter schindleri exhibited resistance against developing severe gut epithelium perturbations on feeding DSS. A detailed bioinformatics analysis of the 16S rRNA sequences of the Acinetobacter sp. (KT991636) isolated from this study using UVA FASTA server exhibits 99.9% sequence similarity to Acinetobacter sp. (AJXD00000000) whose whole genome has been sequenced. The annotation results have sown the presence of protein genes coding for esterase enzymes in this particular strain. Further, sequence analysis revealed its close similarity to some of the xenobiotic degrading organisms like it showed 96% 16S rRNA nucleotide sequence similarity to phenol degrader Acinetobacter sp. AF392991. Further, several other strains of this bacterium were reported to be actively involved in the metabolism and biodegradation of xenobiotic pollutants like benzene, phenol and styrene (Abdel-El-Haleem 2003; Achaleke et al.2009). Additionally, it has been shown that A. schindleri secretes esterase at alkaline condition (pH 10.5) that is the normal physiological condition in insect gut and degrades quinalphos readily when compared to other native gut bacterium (Priya 2012: unpublished study). Recently, several bacteria have been demonstrated to activate aryl hydrocarbon receptor pathway that involves recognition and metabolism of xenobiotics by inducing cyp1a gene (Takamura et al.2011). In insects, it was demonstrated that certain P450 enzymes that belong to the subfamily CYP1 were induced on exposure to xenobiotics (Nelson et al.1996). Thus, it will be interesting to elucidate the mechanism underlying the inflammatory and protective function of gut microbiota against DSS-induced intestinal inflammation. To summarize, our observations in S. litura suggest that indigenous gut bacteria that remain benign in optimum environment can exert pathogenic or beneficial consequence under stress conditions. Such bacterium–epithelium interactions with protective functions might be exceptional in insects and hence need to be addressed. In view of the comparative cellular framework and physiology between mammals and lepidopteran insects, we anticipate that understanding of host–microbiota interaction using S. litura as a model might contribute to understand unanswered questions of human intestinal pathology and physiology. Although complex manifestation of the disorder cannot be entirely replicated in this model yet it can provide sincere insight into the pathophysiology of disease without involving complications of the tedious mammalian gut framework. Hence, we foresee that better understanding of host epithelial response to stress conditions will encourage development of new promising ways for managing gastrointestinal diseases in future.

SUPPLEMENTARY DATA

Supplementary data are available at FEMSEC online.

Acknowledgements

We thank Mr Prateek Arora for his technical assistance in Fluorescence-activated cell sorting (FACS) analysis. We also thank Naseer Sangwan for his suggestions that improved the manuscript. We also thank Mr Ashok Pal for laboratory assistance.

FUNDING

The National Fund for Basic Strategic and Frontier Application Research in Agriculture, of the Indian Council for Agricultural Research, Govt of India, and Faculty Research Grants of University of Delhi funded this work. NP was provided research fellowship by Council of Scientific and Industrial Research, Government of India.

Conflict of interest. None declared.