Synbiotic (Lactiplantibacillus pentosus GSSK2 and isomalto-oligosaccharides) supplementation modulates pathophysiology and gut dysbiosis in experimental metabolic syndrome

Metabolic syndrome a lifestyle disease, where diet and gut microbiota play a prodigious role in its initiation and progression. Prophylactic bio-interventions employing probiotics and prebiotics offer an alternate nutritional approach towards attenuating its progression. The present study aimed to evaluate the protective efficacy of a novel synbiotic (Lactiplantibacillus pentosus GSSK2 + isomalto-oligosaccharides) in comparison to orlistat in an experimental model of metabolic syndrome. It was observed that supplementation of synbiotic for 12 weeks to Sprague Dawley rats fed with high fat diet (HFD), ameliorated the morphometric parameters i.e. weight gain, abdominal circumference, Lee’s index, BMI and visceral fat deposition along with significantly increased fecal Bacteroidetes to Firmicutes ratio, elevated population of Lactobacillus spp., Akkermansia spp., Faecalibacterium spp., Roseburia spp. and decreased Enterobacteriaceae compared with HFD animals. Additionally, synbiotic administration to HFD animals exhibited improved glucose clearance, lipid biomarkers, alleviated oxidative stress, prevented leaky gut phenotype, reduced serum lipopolysaccharides and modulated the inflammatory, lipid and glucose metabolism genes along with restored histomorphology of adipose tissue, colon and liver compared with HFD animals. Taken together, the study highlights the protective potential of synbiotic in comparison with its individual components in ameliorating HFD-induced metabolic complications.

Improved glucose tolerance and reduced insulin. It was observed that administration of either probiotic L. pentosus GSSK2 (Group IV), synbiotic (Group VIII) or orlistat (Group IX) to HFD animals significantly (p < 0.05) lowered the fasting blood glucose level and enhanced glucose tolerance compared with HFD (Group II) animals that had increased fasting blood glucose level and impaired glucose clearance from circulation as depicted by an increase in the area under the concentration-time curve (AUC) during OGTT (Fig. 2a-c). However, animals belonging to IMOs + HFD (Group VI) did not show any significant improvement in glucose tolerance while animals supplemented either with probiotic (Group III), prebiotic (Group V) or synbiotic (Group VII) had blood glucose parameters comparable to control (Group I, Fig. 2a-c).Interestingly, serum insulin levels were significantly (p < 0.05) reduced in animals belonging to L. pentosus GSSK2 + HFD (Group IV), IMOs + HFD (Group VI) synbiotic + HFD (Group VIII) and orlistat + HFD (Group IX) animals compared with HFD (Group II) animals (Fig. 2d).

Increased lactic acid bacteria (LAB) count and lipids in feces.
The LAB count in feces, an indicator of healthy gut, increased significantly (p < 0.05) in animals belonging to synbiotic, probiotic and prebiotic inspite of HFD feeding compared with control (Group I) and HFD (Group II) animals but maximum increase in LAB was in synbiotic (Group VII) animals (Fig. 2e). Further, supplementation of synbiotic led to increased fecal lipid excretion inspite of HFD, compared with counter controls (Group II, IV, VI, IX) (Fig. 2f). Improved gut bacteria composition. Supplementation of synbiotic to HFD animals (Group VIII) led to maximum increase in the Bacteroidetes to Firmicutes ratio followed by L. pentosus GSSK2 + HFD (Group IV), IMOs + HFD (Group VI) and orlistat + HFD (Group IX) respectively, compared with HFD (Group II) animals ( Fig. 3a-c). Further, animals belonging to synbiotic + HFD (Group VIII) had significantly (p < 0.05) elevated population of Lactobacillus spp., Akkermansia spp., Faecalibacterium spp., and Roseburia spp. while L. pentosus GSSK2 + HFD (Group IV) had significantly (p < 0.05) high number of Lactobacillus spp. and Roseburia spp. while IMOs + HFD animals (Group VI) had increased population of Lactobacillus spp. and Faecalibacterium spp whereas, orlistat + HFD (Group IX) animals did not show any significant change in any of the bacterial genera compared with HFD animals (Fig. 3d-j). Moreover, Enterobacteriaceae population decreased significantly (p < 0.05) in synbiotic + HFD (Group VIII), L. pentosus GSSK2 + HFD (Group IV), IMOs + HFD (Group VI) and orlistat + HFD (Group IX) animals compared with HFD animals (Fig. 3k). Further, animals administered with probiotic (Group III) had increased abundance of Lactobacillus spp. and Roseburia spp., prebiotic (Group V) animals had increased Lactobacillus spp., Akkermansia spp, Faecalibacterium spp. and Ruminococcus spp. while  Improved serum biomarkers. It was found that animals belonging to synbiotic + HFD (Group VIII) and orlistat + HFD (Group IX) had significantly (p < 0.05) reduced obesity associated serum biomarkers i.e. total cholesterol, triglycerides and LDL cholesterol with increased HDL cholesterol while L. pentosus GSSK2 + HFD (Group IV) animals had significantly (p < 0.05) reduced levels of triglycerides and LDL-cholesterol and IMOs + HFD (Group VI) had reduced levels of triglycerides compared with HFD animals (Group II, Table 1). Further, L. pentosus GSSK2 + HFD (Group IV), synbiotic (Group VIII) and orlistat (Group IX) supplementation to HFD animals significantly (p < 0.05) reduced bilirubin, AST and ALT levels while IMOs + HFD (Group VI) animals showed significant reduction in bilirubin compared with HFD (Group II) animals (Table 1).
It was also observed that synbiotic + HFD (Group VIII) and orlistat (Group IX) animals had significantly (p < 0.05) reduced levels of serum LPS, TNF-α and IL-6 followed by L. pentosus GSSK2 + HFD (Group IV) and IMOs + HFD (Group VI) animals compared with their elevated levels in serum of HFD (Group II) animals (Table 1).

Enhanced antioxidant and suppressed oxidant level.
It was interesting to observe that supplementation of synbiotic to HFD animals (Group VIII) led to significant (p < 0.05) reduction in oxidant MDA level both in adipose tissue and colon whereas orlistat + HFD (Group IX) animals had maximum reduction of MDA in colon followed by L. pentosus GSSK2 + HFD (Group IV) and IMOs + HFD (Group VI) respectively compared with HFD animals (Table 2). Further, antioxidant GSH and SOD level increased in adipose tissue, colon as well as liver of synbiotic + HFD (Group VIII), followed by L. pentosus GSSK2 + HFD (Group IV) and orlistat + HFD (Group IX) while IMOs + HFD (Group VI) animals did not show significant (p < 0.05) change in GSH and SOD level compared with HFD animals ( Table 2).
The expression of gut integrity gene, claudin in colon was significantly (p < 0.05) upregulated in synbiotic + HFD (Group VIII) and orlistat + HFD (Group IX) while Muc-2 expression was upregulated in L. pentosus GSSK2 + HFD (Group IV) and synbiotic + HFD (Group VIII) compared with HFD (Group II) animals. Further, synbiotic + HFD animals (Group VIII) had maximum downregulation of TLR-4 expression followed by L. pentosus GSSK2 + HFD (Group IV), orlistat + HFD (Group IX) and IMOs + HFD (Group VI) animals respectively while there was no significant change in CDX-2 expression in animals belonging to various groups compared with HFD (Group II) animals (Fig. 4c).
Histological modulation. Histological analysis of adipose tissue of HFD (Group II) animals showed hypertrophied adipocytes indicated by increased mean adipocyte size compared with normal histoarchitecture of adipocytes of control (Group I) animals ( Fig. 5a,b,j). Interestingly, supplementation of probiotic, prebiotic, synbiotic and orlistat to HFD animals for 12 weeks led to reduced adipocyte hypertrophy. Maximum reduction in mean adipocyte size was observed in synbiotic + HFD (Group VIII) followed by orlistat + HFD (Group IX), L. pentosus GSSK2 + HFD (Group VI) and IMOs + HFD (Group VI) animals respectively ( Fig. 5d,f,h-j). However, adipose tissue of animals belonging to either probiotic (Group III), prebiotic (Group V), or synbiotic (Group VII) had normal histoarchitecture of adipocytes ( Fig. 5c,e,g).
Morphological examination of liver of animals belonging to HFD (Group II) showed ballooning degeneration of hepatocytes and hypertrophy compared with normal hepatocyte of probiotic (Group III), prebiotic (Group V), synbiotic (Group VII) and control (Group I) animals ( Fig. 6a,b). However, supplementation of either probiotic (Group IV), synbiotic (Group VIII) or orlistat (Group IX) to HFD animals led to reduced hepatic steatosis along with minimal fat deposition in hepatocytes compared with HFD (Group II) animals whereas, IMOs + HFD (Group VI) had increased hepatic steatosis with ballooned hepatocytes (Fig. 6c-i).
Colon segments of HFD (Group II) animals showed disrupted crypts, focal colitis in form of excess lymphocytes between glands and hyperplasia compared with intact mucosal epithelium of probiotic (Group III), prebiotic (Group V), synbiotic (Group VII) and control (Group I) animals ( Fig. 7a-c,e,g). Interestingly, the colon of L. pentosus GSSK2 + HFD (Group IV), synbiotic + HFD (Group VIII) and orlistat + HFD (Group IX) had intact epithelium lining and closely packed mucus glands with minimal inflammatory cells compared with accumulation of lymphocytes in colonic mucosa of IMOs + HFD (Group VI) animals ( Fig. 7d,f,h,i).

Discussion
Trialogue between diet, gut microbiota and host immune response has revealed novel prophylactic interventions employing probiotics for metabolic diseases. In our earlier studies, we have observed that indigenous probiotic L. pentosus GSSK2 showed potent anti-inflammatory activity in LPS induced RAW 264.7 cells, metabolized prebiotic IMOs and ameliorated adiposity parameters in HFD fed SD rats 14,15 . In present study, an attempt was made to investigate the prophylactic potential of a novel synbiotic intervention (L. pentosus GSSK2 + IMOs) in experimental metabolic syndrome compared with orlistat, commonly used weight loss drug.
Increased body weight is an important hallmark of metabolic syndrome and is accompanied by fat mass deposition. Interestingly, synbiotic + HFD animals had improved adiposity parameters with no change in average feed intake which might be attributed to lipoprotein lipase inhibition and release of appetite-reducing hormones  www.nature.com/scientificreports/ glucagon-like peptide-1 and peptide YY either by probiotic or its metabolites resulting into increased energy expenditure, reduced visceral adipose tissue deposits and decreased mean adipocyte size [17][18][19] . Esposito et al. 20 have also observed that supplementation of VSL#3 to HFD rats reduced fat mass with equal food intake. Hyperglycemia, a key outcome of metabolic syndrome, was ameliorated in synbiotic + HFD animals along with reduced insulin level suggesting improved glucose metabolism which might be due to alleviated β cell dysfunction, increased expression of GLUT-4 leading to increased glucose uptake by adipose tissue and muscles [21][22][23] . Lim et al. have also demonstrated that probiotic Latilactobacillus sakei OK67 reduced the blood glucose levels in HFD-fed mice due to decrease in LPS producing Gram negative bacteria and preventing LPS induced inflammation 24 . Accumulating evidences have demonstrated the role of gut dysbiosis in the etiology and progression of diet induced disorders 25 . The decreased ratio of Bacteroidetes to Firmicutes and decreased abundance of beneficial   [31][32][33][34] . It was observed that, synbiotic supplementation to HFD animals positively modulated the serum lipid profile paralleled with reduced hepatic steatosis in histological analysis and increased fecal lipid excretion thereby  www.nature.com/scientificreports/ On molecular basis, synbiotic supplementation to HFD animals downregulated the expression of lipid metabolism regulators, FASN and HSL leading to suppressed lipid synthesis, increase β-oxidation, improved hepatosteatosis and corroborates with earlier studies where selenium enriched probiotics reduced the expression of FASN while Singh et al. reported decreased HSL expression in cobiotic (IMOs + lycopene) supplemented HFD animals 39,40 . Further, the liver GLUT-4 mRNA level was enhanced while glucokinase was reduced in synbiotic + HFD animals that might have attributed to improved insulin resistance and maintained glucose homeostasis 41 .
Interestingly, the expression of PPARγ and C/EBPα, the main transcription factors of adipocyte differentiation, lipid storage and adipokine signalling was reduced in synbiotic + HFD animals, suggesting decreased adipogenesis via limiting the conversion of preadipocytes to mature adipocytes 42 . Park et al. have also reported that L. plantarum Q180 inhibited 3T3-L1 adipocyte differentiation by downregulation of C/EBPα and PPARγ and reduction of adipocyte size in diet-induced obese mice 43 . Obesity is often accompanied by resistance to leptin, a hormone secreted by adipocytes, leading to increased hunger and reduced energy expenditure occurring due to hyperleptinemia 44 . Notably, synbiotic supplementation to HFD animals led to decreased leptin mRNA expression and is in agreement with previous study where L. plantarum A29 supplementation reduced fat mass and downregulated the expression of leptin in adipocytes, resulting in reduced bodyweight of HFD mice 45 . It was also noted that synbiotic + HFD animals had decreased expression of proinflammatory markers, TNF-α and IL-6, upregulated expression of anti-inflammatory adipokine i.e. adiponectin in adipose tissue along with reduced serum level of inflammatory markers (LPS, TNF-α and IL-6) which corroborated with previous study where administration of L. pentosus S-PT84 to LPS and HFD fed mice exerted anti-inflammatory effect by restoring adiponectin production and decreased pro-inflammatory mediators 46,47 . Alleviation of systemic endotoxemia by synbiotic intervention could be due to decreased LPS producing Enterobacteriaceae in the gut and downregulation of key signaling pathways by probiotic or its metabolites as observed in our earlier in vitro study, where L. pentosus GSSK2 attenuated LPS-induced inflammation by downregulating MAPK pathway and COX-2 14,48,49 .
In the present study, mucin gene Muc-2 and tight junction protein claudin were upregulated with reduced expression of TLR-4 in colon suggesting that synbiotic supplementation to HFD animals attenuated mucosal damage and regulated gut barrier function which is further supported by reduced infilteration of immune cells in colon of synbiotic + HFD animals. Similarly, Mennigen et al. reported that probiotic mixture VSL#3 protected the epithelial barrier function by maintaining tight junction expression in murine colitis model 50 .
Based on the present study, synbiotic biointervention (L. pentosus GSSK2 + IMOs) was found to be the most effective and comparable to antiobesity drug orlistat in terms of improved anthropometric parameters, biochemical markers, gene expression and histoarchitecture of metabolic tissues. Thus, the results clearly highlight that the protective potential of the synbiotic biointervention is multifarious and is mediated via its modulation of adipocytes, liver, colon and immune cells as all these play critical role in regulating metabolic homeostasis. Therefore, the proposed molecular mechanism of modulation of HFD-induced metabolic alterations by synbiotic may be attributed to remodulation of gut microbiota by probiotic as well as prebiotic, that may have led to altered adiposity by remodeling energy metabolism, activation of nutrient sensing pathways, mobilizing fats by regulating the expression of glucose and lipid metabolism genes, increased fatty acid oxidation, cholesterol binding and assimilation by probiotic. Moreover, increased short chain fatty acid production due to biofermentation of prebiotic and reduced circulating LPS levels due to decreased pathobionts may have led to mitigation of chronic inflammation, metabolic endotoxemia and restoration of intestinal barrier function that in turn regulated the glucose homeostasis and prevented fat accumulation in liver and adipose tissue thereby attenuating the progression of metabolic syndrome.
However, fecal microbiota analysis by next generation sequencing and short chain fatty acid analysis would have provided a better understanding of the impact of synbiotic intervention on gut microbiota and metabolites produced while insulin tolerance test would have given better insight about alleviation of insulin resistance. Moreover, due to species and strain specific response of probiotics and entirely different gut microbiota of rodent, detailed translational safety and efficacy studies for substantiating the functional benefits of the synbiotic, standardizing optimum dosage and monitoring the variability in response to these intervention is required in human subjects. Taken together, it is proposed that such novel synbiotic intervention may be employed for development of functional foods to combat the growing incidence of metabolic syndrome that could be considered as a promising live bacteriotherapy for maintaining the metabolic homeostasis.

Methods
Animals. Male Sprague-Dawley (SD) rats (150-180 g) were used in the study as they are more vulnerable than the females to the impacts of HFD on metabolic alterations and were procured from inbred population of the Central Animal House, Panjab University, Chandigarh, India 51 . Rats were housed in polypropylene cages with a hygienic bed of husk in room with 12 h light/dark cycle, acclimatized for 7-10 days and given standard pellet diet and water ad libitum. Preparation of HFD. Standard pellet diet (SPD) (6% calories from fat) was procured from Ashirwad Industries, Chandigarh, India and HFD (60% calories from fat) was prepared in-house as described previously 15  www.nature.com/scientificreports/ Preparation of dose.
Experimental design. Animals were divided into nine groups, each comprising of 6 animals and treated as follows.
• Group I (Control): Animals were fed with SPD for 12 weeks.
• Group II (HFD): Animals were fed with HFD for 12 weeks.
• Group III (L. pentosus GSSK2): Animals were fed with a single dose of probiotic (1 × 10 9 lactobacilli/0.1 mL) daily via orogastric gavage and were given SPD for 12 weeks. Follow up of animals. Body weight and LAB count were monitored once a week, throughout the experiment. A day before sacrificing the animals, fasting blood glucose level was monitored, oral glucose tolerance test (OGTT) was performed and feces of animals were collected, for estimation of fecal lipids and analysis of gut bacterial abundance. Animals were sacrificed after 12 weeks of respective treatments by injecting ketamine hydrochloride (80 mg/kg) intraperitoneally followed by cervical dislocation. Blood was drawn through retroorbital bleeding for estimation of serum biochemical parameters. Liver, adipose tissue (epididymal and retroperitoneal) and colon were collected for analysis of oxidants and antioxidants, histopathological alterations and molecular markers ( Supplementary Fig. S2).

Evaluation of morphometric parameters and adiposity markers. Body weight of animals was
recorded weekly on ordinary balance (SD-300, S.D fine chemicals Ltd, Chandigarh, India) while abdominal circumference was measured at the beginning and end of the study using ordinary measuring tape 15 . Lee's Index was calculated as cube root of body weight (g)/ naso-anal length (cm), Body mass index (BMI) was monitored as body weight (g)/length 2 (cm 2 ) at the end of experiment 13 . Feed intake of animals was recorded twice a month and was calculated by subtracting the amount of residual food in each cage from the weighed amount of food provided on previous day (g/day) and represented as average feed intake (g/day/ rat) by dividing the feed intake by total number of animals per cage 15 . Post sacrifice, liver and adipose tissue were weighed using ordinary balance.
Blood glucose and OGTT . Fasting blood glucose levels of animals were recorded weekly via tail snip method using glucometer (Freestyle Optium Glucometer, Abbott Ltd., Oxon, UK). For OGTT, animals were fasted for 6 h and blood glucose level was measured before and after oral administration of D-glucose (2 g/kg) at an interval of 15, 30, 60, 90, and 120 min respectively 15 . AUC was calculated using GraphPad PRISM 5 software. Serum insulin level was determined using commercially available ELISA kit as per manufacturer's instructions (Ray Biotech, Norcross, GA, USA).

Fecal LAB count and fecal lipids.
To assess the effect of synbiotic supplementation on LAB in the colon, freshly voided fecal material (0.5 g/animal) was collected weekly from each group, homogenized in normal saline, serially diluted and plated on MRS agar. The plates were incubated at 37 °C for 48 h and colony forming units (CFU) were recorded 11 .
Fecal lipids were extracted using phase separation based method by Folch et al. 52 followed by estimation of total lipids by method of Fringes and Dunn 53 . Briefly, 200 mg dry feces was taken in a centrifuge tube and 3 mL of chloroform-methanol mixture (2:1, v/v) was added, vortexed for 1 min and centrifuged at 3,000 g for 10 min. The chloroform phase containing lipid fraction was collected in a fresh tube and completely dried followed by estimation of total lipids. Selected gut bacterial abundance. Bacterial  www.nature.com/scientificreports/ was performed using Infinite® M200 Pro NanoQuant (Tecan). DNA extracted as above was subjected to qPCR to quantify the abundance of Lactobacillus spp., Bifidobacterium spp., Roseburia spp., Akkermansia spp., Faecalibacterium spp., Ruminococcus spp., Prevotella spp., using genus-specific primers and that of Enterobacteriaceae, Bacteroidetes and Firmicutes using phylum-specific primers taking total bacteria as an internal control. q-PCR conditions and primer details are given in supplementary material (Supplementary Table 1). Data was analyzed using the ΔΔCt method and values expressed as fold change relative to the control group 13 .

Assessment of oxidant and antioxidant level.
Tissue homogenates of colon, liver and adipose tissue samples were prepared in 0.15 M PBS (pH 7.2) using potter Elvehjem homogenizer. Post mitochondrial supernatant (PMS) was prepared by cold centrifuging tissues homogenates at 16,000 × g for 10 min and supernatant was labeled as PMS. Protein concentration in tissue homogenate and PMS was measured using standard method of Lowry et al. 54 .
The amount of malondialdehyde (MDA), a measure of lipid peroxidation, was assayed in homogenates as per Wills 55 and results were expressed as nanomoles of MDA per milligram of protein. Superoxide dismutase (SOD) activity was assayed in PMS of tissue homogenates according to the method of Kono 56 and expressed as units of SOD per milligram of protein, where 1 U activity is defined as the amount of SOD required to inhibit the rate of Nitroblue tetrazolium reduction by 50%. Reduced glutathione (GSH) levels were estimated in tissue homogenates as per Ellman 57 , absorbance was measured at 412 nm and results were expressed as µmole of GSH/mg of protein.
Gene expression analysis. q-PCR based gene expression analysis was done for fatty acid synthase (FASN), hormone sensitive lipase (HSL), glucokinase, GLUT-4, TNF-α and IL-6 in liver; CCAAT/ enhancer-binding protein alpha (C/EBPα), peroxisome proliferators-activated receptor gamma (PPAR-γ), leptin, adiponectin, TNF-α and IL-6 in adipose tissue; claudin, CDX-2, Muc2 and TLR-4 in colon. Total RNA was extracted using Trizol (Sigma Aldrich, USA). 1 µg of RNA sample was used for c-DNA synthesis using commercially available kit (Biorad iscript kit 1,708,891) as per the kit's instructions. Relative expression of different genes was determined by qPCR using SYBR® based dye (Biorad C1000 Touch Real-Time PCR machine). q-PCR conditions and primer details are given in supplementary material (Supplementary Table 2). Data was analyzed using the ΔΔCt method and values were expressed as fold change relative to control group. GAPDH was used as internal reference gene to normalize the expression of target genes.

Histological analysis.
A part of distal colon, liver and adipose tissue were fixed immediately in 10% buffered formalin, processed, stained with hematoxylin and eosin, and examined for histological alterations using light microscope. The mean adipocyte sizes in adipose tissue sections (minimum 2 animals per group) were estimated in 10-12 images (40X objective), using Image J software 58 .
Statistical analysis. Results were expressed as mean ± standard deviation (SD). The inter group variation was assessed by one-way analysis of variance (ANOVA) followed by Tukey's post Hoc Test using PRISM software (5.0). The statistical significance was defined as p and calculated at p < 0.05. www.nature.com/scientificreports/