bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

1 Research Article 2 3 Microbiota induces aging-related leaky gut and inflammation by dampening mucin 4 barriers and butyrate-FFAR2/3 signaling 5 6 7 8 Sidharth P Mishra1,2,3,4, Bo Wang5, Shaohua Wang1,2,3,4, Ravinder Nagpal4, Brandi Miller1,2,3,4, 9 Shalini Jain1,2,3, Jea Young Lee2,3, Cesar Borlongan2,3, Subhash Taraphdar6, Sushil G. Rane7, 10 Hariom Yadav1,2,3,4,8* 11 12 13 1USF Center for Microbiome Research, 2Department of Neurosurgery and Brain Repair and 14 3Center for Excellence of Aging and Brain Repair, University of South Florida Morsani College of 15 Medicine, Tampa, FL, USA 16 17 4Department of Nutrition and Integrative Physiology, Florida State University, Tallahassee, FL, 18 USA 19 20 5Department of Chemistry, North Carolina A&T State University, Greensboro, NC, USA 21 22 6Department of Animal Genetics and Breeding, West Bengal University of Animal & Fishery 23 Sciences, Kolkata, West Bengal, India 24 25 7Diabetes, Endocrinology and Obesity Branch, National Institute of Digestive and Diabetes and 26 Kidney Diseases, National Institutes of Health, Bethesda, MD, USA 27 28 29 8Department of Internal Medicine- Digestive Diseases and Nutrition, University of South Florida 30 Morsani College of Medicine, Tampa, FL, USA 31 32 33 34 *Corresponding author 35 Hariom Yadav, PhD, Director of USF Center for Microbiome Research, Associate Professor of 36 Neurosurgery and Brain Repair, MDT 1753, Center for Excellence of Aging and Brain Repair, 37 University of South Florida Morsani College of Medicine, Tampa, FL 33612, USA. 38 Email: [email protected] 39 40

- 1 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

41 ABSTRACT 42 43 Increased chronic inflammation is one of the key risk factors of aging-related disorders although 44 its precise etiology remains elusive. Here, we demonstrate that aged, but not young, microbiota 45 triggers inflammation by promoting gut permeability (leaky gut) via disruption of mucus barriers. 46 Levels of the beneficial short-chain fatty acid, butyrate, are suppressed in the aged gut. 47 Consistent with feedback regulation, the expression of butyrate-sensing receptors, free fatty 48 acid receptor 2/3 (FFAR2/3), are also reduced in aged gut. Butyrate treatment of aged mice 49 revereses the reduced mucin production, increased gut permeability and inflammation 50 associated with low butyrate levels. In agreement, intestine-specific FFAR2/3 knockout mice 51 manifest a compromised gut phenotype typically seen in aged mice,, such as increased gut 52 permeability and inflammation with reduced mucin production. Taken together, our results 53 demonstrate that an aged gut microbiota causally instigates inflammation by increasing gut 54 permeability due to reduced butyrate levels, FFAR2/3 expression, and mucin barriers. Thus, 55 butyrate-FFAR2/3 agonism could ameliorate the deleterious effects seen in aged gut and their 56 implications on metabolic health. 57 58 Keywords: Microbiome, aging, gut, barrier, mucus, permeability, inflammation 59 60 INTRODUCTION 61 62 With the increasing aging and veteran population, the prevalence of aging-related diseases, 63 such as diabetes, obesity, cardiovascular dysfunctions, Alzheimer’s disease, and cancer, is on 64 the rise (Disease et al., 2018). Geroscience is an emerging field focused on understanding the 65 basic biology of aging, which may lead to improve the clinical interventions, as well as included 66 inflammation as one of the pillars of Geroscience and may lead to the generation of clinically 67 meaningful outcomes to improve health of aged adults (Ferrucci & Fabbri, 2018; Kennedy et al., 68 2014). Persistent low-grade chronic inflammation is often observed in older adults and is 69 considered one of the major risk factors of aging-related diseases (Ferrucci & Fabbri, 2018). 70 The persistent decline in adaptive immunity and chronic increase in the levels of cytokines like 71 interleukin (IL)-1β, IL-6, tumor necrosis factor-alpha (TNF-α) in the systemic circulation are 72 hallmarks of increased inflammation in older adults (Koelman et al., 2019). However, the precise 73 origins of chronic inflammation in elderly are obscure. We and others have shown that gut 74 permeability (referred to as ‘leaky gut’) is often increased in older adults and is an understudied 75 source of inflammation in aging (Ahmadi, Wang, et al., 2020; Wilms et al., 2020). Increased gut 76 permeability allows the non-specific diffusion of microbes, microbial ingredients, metabolites, 77 and antigens from the gut lumen to the blood which, in turn, can promote the migration of pro- 78 inflammatory molecules and systemic inflammation. For example, lipopolysaccharide (LPS) - 79 one of the cell wall components of gram-negative bacteria - leaks from gut causing systemic 80 inflammation resulting in endotoxemia (Cani et al., 2008). Indeed, levels LPS-binding protein 81 (LBP) and soluble CD14 (sCD14) – markers of leaky gut - are significantly higher in the older 82 adults and are linked with age-associated phenotypes such as decreased physical activity and 83 increased risk of heart-failure (Awoyemi et al., 2018; Stehle et al., 2012). However, the 84 mechanisms underlying compromised gut barrier functions in the aged gut are unknown. 85 86 Gut barrier functions are under the purview of two major structural components: mucin 87 and tight-junction proteins (Chelakkot et al., 2018). Mucin forms a physical barrier by sealing 88 gaps between the intestinal epithelial cells, while tight junction proteins tightly adhere epithelial 89 cells. Intestinal goblet cells produce mucin as part of a thick gel like viscous mucus layer on the 90 intestinal epithelia, thus facilitating the selective diffusion of small molecules from the gut lumen 91 to intestinal epithelial cells (Pelaseyed et al., 2014). Although this mucus layer restricts any

- 2 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

92 direct bacterial interactions with intestinal epithelial cells, it allows the colonization of bacteria on 93 the mucuosal surface (Li et al., 2015). Mucin also serves as a food source for numerous gut 94 bacteria such as the mucin-degrading bacteria Akkermansia (Liu et al., 2021). Mucin depletion, 95 however, permits easy interaction of microbes with intestinal epithelial cells, thus initiating 96 degradation of tight junctions proteins, increasing gut permeability and inflammation in the gut 97 environment (Okumura & Takeda, 2018). Mucin 2 (Muc2) and mucin 6 (Muc6) are the major 98 mucin isoforms in the intestine (McGuckin et al., 2011). In agreement, Muc2-deficient mice 99 develop severe colitis and gut inflammation, thereby highlighting the physiological importance of 100 mucin in maintaining gut barrier functions and gut permeability while staving off inflammation 101 (Van der Sluis et al., 2006). Indeed, aging rodent and primate models exhibit thin mucin lining 102 linking increased gut inflammation with age-associated diseases (Elderman et al., 2017; Sovran 103 et al., 2019; Vemuri et al., 2020). Taken together, maintaining a healthy mucus layer is critical 104 for ideal gut barrier function with advancing age. However, the mechanisms and molecular 105 drivers of reduced mucin levels in the aging gut are poorly understood. 106 107 The gut microbiota in older adults is distinct compared to young adults (Langille et al., 2014). 108 Specifcally, aged gut microbiota signature is often characterized by skewing of the ratio in favor 109 of increased abundance of harmful bacteria and reduced proportions of beneficial bacteria. 110 Termed dysbiosis these perturbations to the gut microflora dysregulate the production of 111 adverse and beneficial metabolites (Ahmadi, Wang, et al., 2020). However, evidence that 112 fluctuations in the microbiota population instigate gut permeability and promote gut inflammation 113 in elderly adults is lacking. It is also unknown if the gut microbiota impact goblet cells and mucin 114 barriers in the gut of elderly adults. Furthermore, the mechanisms by which microbiota impact 115 the mucus and tight junction barrier functions are unclear. Using un-biased, non-supervised and 116 comprehensive analyses, we address these open questions. We unravel causal microbiota- 117 dependent mechanisms with translational potential to reverse gut permeability and inflammation 118 in aged adults and help mitigate the consequential adverse health outcomes. 119 120 RESULTS 121 122 1. Aged gut microbiota propogates gut permeability and inflammation. Microbiome 123 diversity is an important indicator of microbiota balance and health in the gut (Lozupone et al., 124 2012). The β-diversity indicates similarity between the two microbiome communities (a higher 125 value indicates more differences between the microbiomes), while the α-diversity indicates the 126 microbial diversity within the samples (a higher α-diversity indicates more diversity) (Wagner et 127 al., 2018). Principal co-ordinate analysis (PCA) of β-diversity showed that the gut of aged mice 128 displayed significantly (p<0.05) distinct microbiota signatures than that from younger mice 129 (Figure 1a). The α-diversity indices like Shannon index and number of operational taxonomic 130 units (OTUs) were lower in the aged gut compared to the younger gut (Figure 1b, 131 Supplementary Figure S1a), suggesting that the aged microbiome skewed towards unhealthy 132 state. Furthermore, compared to the younger gut the abundance of major phyla such as 133 Bacteroidetes, Verrucomicrobia, Proteobacteria, Deferribateres and Actinobacteria were lower 134 in the aged gut while the Firmicutes abundance was high (Figure 1c, Supplementary Figure 135 S1b). Similar differences in the old and young gut were observed in other taxonomic ranks, 136 including class, orders and family (summarized in Supplementary Figure S1c-e). Furthermore, 137 the abundances of major genera like Barnesiella, Dysgonomas, Akkermansia, Lactobacillus, 138 Parabacteroides, Alistipes, Turicibacter and Bacteroides were significantly lower in the aged 139 gut, while the abundances of Clostridium_XIVa, Acetobacteroides and Lachnospiracea genera 140 were higher (Figure 1d). Similar results were obtained at the genera level by both heatmap and 141 linear discriminatory analysis (LDA) effect size (LEfSe) analysis (Supplementary Figure S1f,g). 142 At the species level, we also found that the abundance of Muribaculum intestinale, Akkermansia

- 3 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

143 muciniphila, Turicibacter sanguinus, Acetatifactor muris, Mucispirillum schaedleri, Lactobacillus 144 intestinalis, Barnesialla intestinihominis and Faecalibaculum rodentium species were 145 significantly lower, while Bacteroidetes bacterium, Eubacterium cellulosolvens and Flintibacter 146 butyricus were higher in old gut compared to young (Figure 1e). LEfSe analyses also revealed 147 similar results at species level (Supplementary Figure S1h) with the aged gut harboring a 148 significantly distinct microbiome signature on different taxonomic levels like unique phyla, class, 149 order, family and genera (Figure 1f). In addition, we also find that the aged mice have 150 significantly higher (p<0.01) gut permeability (measured by increased transfusion of FITC 151 [Fluorescein isothiocyanate]-dextran from the intestine to systemic blood circulation) and 152 increased mRNA expression of inflammatory markers (IL-1β, IL-6 and TNF-α mRNA 153 expression) (Figure 1g,h). Moreover, the concentration of systemic inflammatory markers like 154 IL-6 and TNF-α concentrations were significantly higher (p<0.01) in the aged mice as compared 155 to the younger mice (Figure 1i,j). These results indicate that aged gut harbors a microbiota 156 distinct from that seen in the younger gut, with the former linked with increased gut permeability 157 and systemic inflammation. We then inquired whether the microbiota differences are causal in 158 the acquisition of aging-related increases in gut permeability and inflammation. 159 160 Fecal microbiota transplantation (FMT) from aged mice significantly increased FITC-dextran 161 diffusion and elevated levels of systemic leaky gut markers like LBP and sCD14 in young 162 recipient mice (Figure 1k-m), suggesting that aged microbiota triggers leaky gut in recipient 163 young mice. In addition, recipient mice with aged FMTs also showed higher levels of systemic 164 inflammation markers like Il-1β, Il-6 and Tnf-α mRNA expression in the intestine and the serum 165 (IL-6 and TNF-α) compared to the young FMT controls (Figure 1n-p). These findings suggest 166 that the aged microbiota can initiate gut permeability and inflammation. Furthermore, we 167 showed that, compared to FCM of younger mice, the treatment of human intestinal epithelial 168 HT-29 cell monolayers with fecal conditioned media (FCM) of aged mice significantly decreased 169 the trans-epithelial electrical resistance (TEER) and increased the diffusion of FITC-dextran 170 (Figure. 1q,r). These observations also suggest that the aged microbiota instigates leakiness in 171 the gut epithelia, which in turn stimulates inflammation. Further, these results demonstrate that 172 FCM treatments recapitulate the effects of FMTs using a cell culture system and, thereby, are 173 alternatives to FMTs to study the impact of the microbiota on epithelial barrier function. Overall, 174 these results indicate that the gut of aged mice harbors an abnormal microbiome with reduced 175 microbial diversity, which elevates gut permeability and inflammation. 176 177 2. Aged microbiota induces gut permeability and inflammation by disrupting mucus 178 barriers. FMT of aged mice significantly reduced mRNA expression of mucin (Mucin 2 [Muc2] 179 and mucin [Muc6]), tight junction genes (tight junction protein 1 [Tjp1] and occludin [Ocln]) 180 expression in the gut of older FMT recipient mice compared to their young FMT recipient 181 controls (Figure 2a,b). Similarly, expression of gut barrier markers was reduced in intestinal 182 organoids and intestinal goblet cells (CMT93 cells) upon incubation with aged mice FCM (Figure 183 2c-e). To understand the mechanism of action, we next performed unsupervised correlation 184 analyses (heat-map), random forest analyses (RFA) and differential expression (Volcano graph) 185 via gene expression and ELISA from mice, organoids and CMT-93 cells. Muc2 expression 186 exhibited the highest negative correlation with markers of leaky gut, inflammation and 187 comprised intestinal barrier (Figure 2f-h). This suggests that mucin may be the primary target of 188 aged microbiota to induce gut permeability. Histological analyses also showed that the number 189 of goblet cells were significantly decreased in mice receiving aged mice FMT, compared to 190 those receiving young mice FMT (Figure 2i-j). These data indicate that the aged microbiota 191 reduced the abundance of goblet cells, accompanied by lower Muc2 expression and poor 192 intestinal barrier functions. Further, the gut of aged donor mice displayed significantly less 193 expression of Muc2, Muc6, Tjp1 and Ocln, along with reduced goblet cell numbers and lower

- 4 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

194 mucin levels in their feces compared to young mice (Figure 2k-o). Thus, the aged mice manifest 195 compromised gut barrier functions due to a reduced number of goblet cells, which is linked to 196 decreased mucin production. Taken together, these results indicate that aged microbiota 197 promotes gut permeability by disrupting intestinal barrier functions via reduction in goblet cell 198 numbers and decrease in mucin production. 199 200 3. Reduced short-chain fatty acids production in aged gut is linked to reduced mucosal 201 barrier and increased gut permeability and inflammation. Muc2 expression shows the 202 highest negative correlation with gut permeability and inflammatory markers (Figure 3a). This 203 further suggests that reduced mucin production is a primary manifestation in aged donor mice, 204 instigated by an abnormal microbiota. Considering that metabolites produced by gut microbiota 205 are key mediators that impact host cells (Hariom Yadav, 2016), we next performed 206 metabolomics analysis to determine the mechanism by which gut microbiota impacts Muc2 207 expression. PCA revealed that the feces of aged mice had significantly different metabolic 208 signatures compared to the feces of young mice (Figure 3b). Further, differential analyses 209 showed that the abundance of beneficial metabolites such as short-chain fatty acids (SCFAs 210 like acetate, propionate and butyrate) and taurine were significantly reduced in the aged gut 211 while abundance of anserine, total bile acids, cholesterol and cholate were significantly 212 increased (Figure 3c). Un-biased RFAs showed that the SCFAs (like butyrate and propionate) 213 were most significantly reduced in the aged gut compared to the younger gut (Figure 3d). 214 Among the metabolites, levels of butyrate were the most significantly reduced in the the aged 215 gut (Figure 3e, and Supplementary Figure 2a). Correlation analyses also showed that the 216 reduced butyrate levels were linked to the reduced Muc2 expression and markers of increased 217 gut permeability and inflammation (Figure 3f-h, Supplementary Figure 2b). Together, these 218 findings are consistent with the notion that the aged microbiota reduces butyrate production, 219 thereby decreasing mucin production and increasing gut permeability and inflammation. 220 221 4. Butyrate treatment reduces gut permeability and inflammation by increasing mucus 222 barriers. Feeding butyrate to mice significantly reduced gut permeability and lowered the 223 expression of inflammatory markers (including Il-6, Tnf-α, monocytes chemoattractant protein-1 224 [Mcp-1] and plasminigen activator inhibitor-1[Pai-1]) (Figure. 4a-b). Further, butyrate treatment 225 significantly increased goblet cell numbers and Muc2 expression in the intestine (ileum) of mice 226 (Figure 4c,d). Butyrate feeding also increased the mucin content in the feces of treated mice 227 compared to non-treated controls (Figure. 4e). These results suggest that butyrate treatment 228 reduced gut permeability and inflammation by increasing goblet cell numbers and mucin 229 production. To determine if butyrate can ameliorate the detrimental effects of aged microbiota 230 (FCM) on mucin barriers, we incubated CMT-93 goblet cells and intestinal organoids with aged 231 gut FCM supplemented with butyrate. Butyrate-incubated CMT-93 cells showed increased 232 mucin production (as suggested by increased PAS staining in Figure 4f) and higher expression 233 of Muc2 and Muc6 compared to controls (Figure 4g). Similarly, the expression of Muc2 and 234 Muc6 was significantly higher and similar to young FCM-treated controls in butyrate-incubated 235 intestinal organoids that were exposed to aged FCM (Figure 4h). These observations showed 236 that butyrate treatment protected the goblet cells from the detrimental effects of the aged 237 microbiota by preserving mucin production. Overall, these results indicate that butyrate reduces 238 aged microbiota-induced disruptions in the mucin barriers mitigating gut permeability and 239 inflammation. 240 241 5. Deficiency of SCFAs receptors FFAR2 and FFAR3 increases gut permeability and 242 inflammation via reducing mucin barriers thus resulting in early aging gut phenotype. . 243 Although it is known that the SCFAs, like butyrate, act on host cells by activating G-coupled 244 protein receptor signaling (such as FFAR2 [Gpr43] and FFAR3 [Gpr41] (Mishra et al., 2020),,the

- 5 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

245 role of FFAR2/3 in preserving mucin barriers is poorly understood. Expression levels of Ffar2 246 and Ffar3 were significantly lower in the gut of aged mice compared to younger mice (Figure 247 5a). Furthermore, expression of Ffar2/3 was significantly decreased in the gut of old FMT 248 recipients as well as in the intestinal organoids, CMT93 and HT29 cells treated with FCM from 249 aged mice compared to young FCM treated controls (Figure 5b-e). These results indicated that 250 the aged microbiota plays a causal role in suppressing the of FFAR2 and FFAR3 expression. 251 Furthermore, the ability of butyrate to induce mucin genes (Muc2, Muc6 and Muc13) was 252 significantly diminished when CMT93 cells and intestinal organoids were treated with Ffar2 253 inhibitor (CATPB; 1 µM) and Ffar3 inhibitor (Pertussis Toxin; 1 µM) (Figure 5f,g). PAS staining 254 also demonstrated that that butyrate induced mucin production was significantly diminished in 255 CMT93 cells upon treatment with FFAR2 and FFAR3 inhibitors (Figure 5h). Together, these 256 results suggest that stimulation of mucin expression by butyrate is dependent on the activation 257 of FFAR2 and FFAR3 receptors. Further, mice lacking FFAR2 and FFAR3 in their intestine 258 (induced by Villin-Cre) exhibited higher gut permeability and inflammation with reduced 259 expression of mucin and tight junction proteins (Figure 5i-p). These data suggest that deficiency 260 of FFAR2 and FFAR3 results in early age onset gut abnormalities manifested as increased gut 261 permeability and inflammation accompanied with reduced mucin expression. Thus, the aged 262 microbiota suppresses FFAR2/3 signaling, which compromises mucin barrier functions and 263 increases gut permeability. Notably, the suppression of FFAR2/3 signaling attenuates the 264 beneficial effects of butyrate supplementation. In young adult mice, deficiency of FFAR2/3 265 signaling in the intestine promotes a pathology reminiscent of aged gut, suggesting that 266 suppression of SCFA signaling contributes to the induction of early aging gut phenotype. 267 268 DISCUSSION 269 270 Low grade chronic inflammation is one of the major risk factors of aging-related disorders and, 271 as proposed by the Geroscience hypothesis, stands as one of the key targets to reduce aging 272 related disorders (Kennedy et al., 2014). However, the etiology of inflammation during aging is 273 largely unknown. We and others have reported that aged mice and humans display increased 274 gut permeability (also referred to as leaky gut) that is linked with increased inflammation and 275 poor health (Ferrucci & Fabbri, 2018). Therefore, we reasoned that gut permeability may be an 276 understudied source of inflammation in the aged gut. However, the mechanisms contributing to 277 leakiness in the aged gut remain unknown. Here, we demonstrated that aged gut harbors an 278 abnormal microbiome which causally promotes gut permeability and inflammation by disrupting 279 mucus barriers. We also show that the aged microbiota produces lower amounts of beneficial 280 metabolites like butyrate and suppresses FFAR2/3 signaling resulting in disruption of mucin 281 barriers that promote gut permeability and inflammation. 282 283 Inflammation is an immune response that can be induced by microbes. In particular, the non- 284 specific diffusion of microbial antigens from the gut to mucosal and systemic immune systems 285 can stimulate inflammation. Emerging evidence indicates that the gut of aged individuals 286 harbors an abnormal gut microbiome (Tachon et al., 2013). Herein, we demonstrated that the 287 gut microbiota of aged mice was significantly different in comparison to their younger 288 counterparts. However, whether abnormalities in the gut microbiota are merely associated with 289 inflammation or if they play a causal role in the process was unknown. To investigate this, we 290 performed FMT studies in mice and found that abnormal microbiota from the aged gut plays a 291 causal role in gut permeability and inflammation. Specifically, we find that the microbiota from 292 the aged gut directly contributes to the induction of inflammation and gut permeability. We used 293 monolayers of human intestinal cells (HT29) treated by fecal conditioned media (FCM) to 294 confirm these findings and demonstrated that the aged microbiota FCM significantly induced gut 295 permeability. We find that the effects of the aged microbiota on intestinal epithelia are

- 6 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

296 conserved in mice and humans and that these effects can be recapitulated in in-vitro and/or ex- 297 vivo settings. 298 299 Gut permeability occurs as a result of structural disruptions in the gut barrier. The mucus layer 300 and tight junction proteins serve as the main components of this barrier (Chelakkot et al., 2018). 301 FMT studies revealed that the aged gut microbiota significantly reduced the expression of 302 mucin and tight junction proteins supporting the notion that the aged microbiota induces gut 303 permeability. Further, unbiased and non-supervised correlational analyses of mice, organoids 304 and human intestinal cells (HT-29) revealed that the Muc2 gene expression was significantly 305 impacted by the aged microbiota. Mucin is an important biophysical barrier that forms a 306 semipermeable gel like structure on the intestinal epithelial layer to seal and prevent the 307 leakage of gut lumen antigens (Paone & Cani, 2020). However, deterioration of the mucus layer 308 can allow direct contact between the bacterial cells and other antigens in the gut lumen. Such 309 adverse effect on intestinal epithelial cells stimulates the breakdown of tight junction proteins 310 leading to leakage in the gut epithelium. 311 312 A common event in mammalian cell interactions with microbial cells (including involving 313 beneficial bacteria) entails an inflammatory response (Belkaid & Hand, 2014). Increased 314 inflammation instigates further breakdown of tight junctions in the intestinal epithelia (Luissint et 315 al., 2016; Zuo et al., 2020). In addition, increased inflammation also reduces production of 316 mucin in the gut (Ahmadi, Razazan, et al., 2020). In aged adults, the mucin layer is thin with this 317 phenotype co-occuring with an abnormal microbiota and increased gut permeability and 318 inflammation. However, if the gut microbiota or inflammation contributes to reducing gut mucin 319 was unclear. Our observations demonstrating that the FMTs of aged microbiota reduced mucin 320 expression in the gut epithelia suggest that microbiota initiatie the disruption of mucin barriers in 321 the gut. However, the reason behind decreased mucin barriers instigated by aged FMTs 322 remained elusive. We here show that old FMTs in mice and treatment of intestinal organoids 323 with old FCMs reduced the number of goblet cells in the intestinal epithelia suggesting that aged 324 microbiota reduces the number of goblet cells. Goblet cells are primarily responsible for mucin 325 production in the gut and further detailed studies are needed to elucidate the mechanism by 326 which the aged microbiota reduces goblet cell numbers in the gut. 327 328 To better understand the mechanism of action, we tested the hypothesis that the gut 329 microbiota produces metabolites that impact intestinal epithelial cells (Visconti et al., 2019). To 330 that end, we found that the feces of aged mice exhibited a significantly distinct metabolite 331 signature compared to younger mice, eventhough both groups of mice were fed an identical 332 diet. Furthermore un-biased and non-targeted metabolomic analyses further revealed that the 333 abundance of beneficial metabolites, including SCFAs (such as butyrate), were significantly 334 reduced in the gut of aged mice. Butyrate is a well-studied gut microbiome derived metabolite 335 and considered to be among the most important energy sources for intestinal epithelial cells 336 (Donohoe et al., 2011). Its decrease in the aged gut may result in energy deprivation in 337 intestinal epithelial cells thereby inducing cellular stress. Butyrate levels were positively 338 correlated with mucin expression and butyrate feeding significantly reduced the gut permeability 339 and inflammation via increased mucin production and elevated goblet cell mass. We and others 340 have demonstrated that the feeding of HFD induces gut permeability and inflammation (Ahmadi, 341 Razazan, et al., 2020; Nagpal, Mishra, et al., 2020) by disrupting gut barrier functions (Ahmad et 342 al., 2017). Ahmadi et al also showed that HFD feeding reduces mucin production in aged mice 343 (Ahmadi, Razazan, et al., 2020; Ahmadi, Wang, et al., 2020). However, it is unclear why 344 butyrate levels were decreased in the gut of aged mice. Butyrate levels are often reduced in 345 pathophysiological conditions due to a reduced abundance of butyrate producing bacteria 346 (Jakobsdottir et al., 2013; Lu et al., 2016). A similar mechanism may be involved in aged gut

- 7 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

347 (Lee et al., 2020). On the other hand, it is possible that butyrate uptake is elevated in aged 348 intestinal cells resulting in depletion of butyrate levels. 349 350 Although butyrate is one of the most widely studied microbiome metabolites and has important 351 roles in several biological and physiological functions (Liu et al., 2018), its mechanisms of action 352 on mucin biology and gut permeability are poorly understood. Butyrate stimulates G-protein 353 coupled receptors FFAR2 and 3 (also known as Gpr43 and Gpr41) (Mishra et al., 2020). These 354 receptors in turn initiate signaling cascades that target genes involved in gut epithelial cell 355 integrity and inflammation. However, the precise roles of FFAR2 and 3 signaling and its 356 interaction with aged microbiome especially as they pertain to mucin biology and gut 357 permeability are unknown. We showed that expression of FFAR2 and FFAR3 genes were 358 significantly reduced in the aged gut. Further, aged gut microbiota transfer via FMTs or FCM 359 significantly reduced the expression of these genes. Taken together, these findings are 360 consistent with the notion that the aged microbiota suppresses FFAR2/3 expression in the gut. 361 Furthermore, inhibition of FFAR2/3 signaling dampened the beneficial effects of butyrate on 362 mucin expression and production suggesting that FFAR2/3 signaling contributes to butyrate- 363 dependent mucin production. However, SCFAs and butyrate are also known to be taken up and 364 used as an energy source by intestinal cells through infusion into the mitochondrial energy 365 metabolism (NEED ORIGINAL REF OR DELETE THIS) and can also inhibit the activities of 366 histone-deacetylase (NEED ORIGINAL REF OR DELETE THIS). However, the role of these 367 mechanisms in mucin production is not known. Furthermore, intestine specific FFAR2 and 368 FFAR3 knockout mice exhibit significantly low mucin expression along with elevated gut 369 permeability and inflammation at a young age. These findings suggest that compromised 370 FFAR2/3 signaling can recapitulate phenotypes typically seen in the aged gut underscoring the 371 potential importance of this signaling in the early aging phenotype. 372 373 Overall, we demonstrate that the aged microbiome promotes gut permeability and inflammation, 374 by disrupting gut barrier functions that are facilitated by a reduction in mucin and goblet cells 375 numbers. The aged gut microbiota reduces butyrate synthesis and FFAR2/3 signaling which, in 376 turn, decreases mucin formation thus weakening gut barrier functions to promote gut 377 permeability and inflammation. Importantly, butyrate administration reverses the phenotype via 378 activation of FFAR2/3 signaling, suggesting that the butyrate supplementation and/or FFAR2/3 379 agonism represent an effective approach for reversing microbiome abnormalities, gut 380 permeability and inflammation typically seen in aging. 381 382

- 8 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

383 Material and Methods 384 Chemical, reagents, and instruments. The specific information about the chemicals, reagents 385 and instruments including catalog numbers and vendors is provided in Supplementary Table S5. 386 Mice studies. Twelve-week (young donor, n=10) and 78 week (aged donor, n=10)C57BL/6J 387 (B6) male mice were purchased from Jackson Laboratory (Bar Harbor, ME, USA) and were 388 acclimatized for 2 weeks in the animal vivarium under a 12h light/-dark cycle. After 389 acclimatization, fresh fecal samples were collected, snap frozen and stored in the -80°C freezer, 390 and were used for 16S metagenomics, metabolomics, and FMT and FCM studies, as described 391 below. Eight week old B6 mice received FMTs from young and old donors, after 7 days of FMT, 392 gut permeability assay was performed and mice were euthanized, and tissues were collected for 393 further analysis. Because high fat diet (HFD) rapidly progresses leaky gut in aging models, 394 some B6 mice were given butyrate (2.5% solution) in their normal drinking water (n=6) while 395 controls were given normal drinking water (n=6), while both groups were fed with high fat diet 396 (60% kcal fat). Homozygous Ffar2flox/flox (fl/fl) and Ffar3fl/fl mice were breed with VilCre+ 397 heterozygotes mice to develop the intestinal tissue specific -Ffar2 knockout mice (intFFAR2 KO 398 [Ffar2fl/flVilCre+]), -Ffar3 knock-out mice (intFFAR3 KO [FFAR3fl/fl/VilCre+]) and compared with their 399 corresponding wild type counterparts like intFFAR2 WT (FFAR2fl/fl-VilCre-) and Ffar3 400 (Ffar3fl/flVilCre-) at the 10 weeks of age for gut permeability, inflammation and mucus barriers 401 while fed with normal chow. All the animal experiments and procedures were approved by the 402 IACUC of the Wake Forest School of Medicine and University of South Florida. 403 Fecal microbiome transplantation (FMT). We have used gut cleansed mice instead of germ- 404 free mice in our studies, because germ free mice have several abnormalities in the gut which 405 may confound the results of our studies. To perform gutcleansing in the recipient mice, we gave 406 them an antibiotic cocktail (consisting of 1 g/l Ampicillin, 1 g/l Metronidazole, 1 g/l Neomycin, 407 and 0.5 g/l Vancomycin) for 4 days in the drinking water. On the 4th day, all the mice were fasted 408 for 4 hours and 4 doses of polyethylene glycol (PEG) were given at 20 minutes interval using 409 oral gavage. PEG physically cleanses the gut and reduces upto the 95% of microbes. After 410 cleansing, the fecal slurry (200 uL) was prepared from the donor groups/for the recipients 411 groups. We have transplanted 9 mice with the young microbiota and 9 mice with old microbiota 412 (Ahmadi, Razazan, et al., 2020; Wang et al., 2020; Wrzosek et al., 2018). We continued giving 413 one dose of fecal slurry every day for an additional 4 days to stabilize the microbiome and 414 maintain it closer to the donor microbiome. The fecal slurry was prepared using snap frozen 415 fecal samples (500 mg) from donor mice, which were dissolved in 5 ml of reduced PBS 416 (phosphate buffer supplemented with 0.1% Resazurin (w/v) and 0.05% L-cysteine-HCl in 417 anaerobic chamber (Kang et al., 2017). 418 Gut permeability assay. Mice were pre-fasted for 4 hours and administered with 4-kD FITC 419 (Fluorescein isothiocyanate)-dextran at a concentration of 1g/kg body weight through oral 420 gavage. After, an additional 4 hours of fasting, blood serum was collected from the tail vein to 421 measure the appearance of FITC fluorescence at 485 nm excitation and at 530 nm emission 422 wavelength using a fluorescence 96-well plate. The FITC concentration was determined using a 423 standard FITC curve as previously described in our earlier publications (Ahmadi et al., 2019; 424 Ahmadi, Razazan, et al., 2020; Ahmadi, Wang, et al., 2020; Nagpal, Newman, et al., 2018; 425 Wang et al., 2020). 426 Markers of systemic leaky gut and inflammation assays. The concentrations of LBP and 427 sCD14 were measured in mice serum to determine the systemic leaky gut; while the 428 concentration of IL-6 and TNF-α were measured in mice serum to determine systemic 429 inflammation using the ELISA kits.

- 9 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

430 Fecal mucin content. Fecal mucin content was quantified using an ELISA kit following the 431 manufacturer`s protocol. Mucin content was measured in triplicate from each mice feces. 432 Histochemistry. The intestinal tissues were collected, washed with PBS, and fixed in 10% 433 formalin. Tissues were sliced into 8-µm thick sections, which were stained with hematoxylin and 434 eosin (H&E). Goblet cells were counted as unstained vacuoles in the H&E 435 immunohistochemistry slides. Mucus staining in the goblet cell line- CMT93 cells was performed 436 in a 12-well cell culture plate by using the Alcian Blue/PAS kit following the manufacturer 437 instructions. All the images were captured at 20X magnification using AmScope microscope. 438 Gut microbiota analyses. Our well established 16S rRNA sequencing with the MiSeq Illumina 439 platform and bioinformatics were used to analyze the gut microbiota composition (Ahmadi et al., 440 2019; Nagpal, Mishra, et al., 2020; Nagpal et al., 2019; Nagpal, Neth, et al., 2020; Nagpal, 441 Shively, et al., 2018; Nagpal, Wang, et al., 2018). In brief, genomic DNA was extracted from 442 ~100 mg of mice feces using the Qiagen QiaAmp PowerFecal Pro kit. Primers 515 F (barcoded) 443 and 806 R were used to amplify the V4 region of bacterial 16S rDNA (Caporaso et al., 2010). 444 After being purified and quantified with AMPure® magnetic purification beads (Agencourt) and 445 Qubit-3 fluorimeter, respectively. Equal amounts (8 pM) of the amplicons were used for 446 sequencing on an Illumina MiSeq sequencer (using Miseq reagent kit v3). The sequences were 447 de-multiplexed, quality filtered, clustered, and analyzed using the Quantitative Insights into 448 Microbial Ecology (QIIME) and R-based analytical tools (Navas-Molina et al., 2013). 449 Metabolomics analysis. Fecal samples were extracted using a previously described method 450 (Gratton et al., 2016) with slight modifications. The extracted water samples were mixed with 451 phosphate buffer and the final solutions contain 10% of D2O with 0.1 M phosphate buffer (pH = 452 7.4) and 0.5 mM trimethylsilylpropanoic acid (TSP). NMR experiments were conducted on a 453 Bruker Ascend 400 MHz high-resolution NMR using a 1D first increment of a NOESY 454 (noesygppr1d) with water suppression and a 4-s recycle delay. All NMR spectra were phased 455 and referenced to TSP in TopSpin 4.06 (Bruker BioSpin, Ettlingen, Germany). The NMR peaks 456 were extracted using Amix 4.0 (Bruker Biospin, Ettlingen, Germany) with a previously reported 457 automatic bucketing method to minimize peak splitting. The metabolite identification was carried 458 out using Chenomx 8.6 (Chenomx Inc, Edmonton, Alberta Canada). Total intensity 459 normalization was applied before further data analysis. 460 Preparation of Fecal Condition Media (FCM). The snap frozen feces were ground in liquid 461 nitrogen using a mortar and pestle. The fine powdered feces were suspended in cold 462 Dulbecco`s modified eagle`s medium (DMEM)media at a concentration of 100 mg fecal 463 powdered content in 100 ml of media. The suspended fecal media was shaken at a speed of 464 200 rpm for 1 hour at 4°C to mix properly. The suspended media was filtered twice through a 465 0.45 µm pore nylon membrane sterile filter and twice through a 0.22 µm nylon membrane 466 sterile syringe filter in sterile condition. This prepared FCM with 1:40 dilution was used to 467 challenge the organoid and cells at a concentration of 2.5 mg/ml. 468 Intestinal organoid development. The intestinal organoids were prepared as described earlier 469 by us (Nagpal, Newman, et al., 2018). In brief, the entire small intestine of 6-8 week old B6 mice 470 was washed with pre-cooled Dulbecco`s phosphate buffered saline (DPBS), and fragmented 471 into the 2 mm small pieces after being cut opened lengthwise. These tissue fragments were 472 washed gently with pre-cooled PBS by pipetting up and down for next 15-20 times until the 473 supernatant become clear.Then, the fragments were suspended in 25 ml of pre-warmed trypsin 474 and shaken for 15 minutes at room temperature. Trypsin was removed by adding four times of 475 fresh 10 ml pre-cold PBS with 0.1% bovine serum albumin (BSA). Digested fragments were 476 filtered through a 70 μm cell strainer and washed by centrifuging at 290x g for 5 min at 4°C. This 477 process was repeated two more times. The intestinal crypt-containing pellets were

- 10 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

478 resuspended in 10 mL of cold DMEM: Nutrient Mixture F-12 (DMEM/F12) and centrifuged at 479 500x g for 10 minutes at 4 °C before resuspension into 150 μL IntestiCult Organoid Growth 480 Medium containing with 50 μg/mL gentamicin. The Matrigel Matrix was added to the suspension 481 and 50 μL of the mixture was pipetted slowly at the center of a prewarmed 24-well culture plate 482 to form a dome. The plate was immediately incubated at 37°C and 5% CO2 for 30 minutes to 483 allow the Matrigel to set, and 500 μL of IntestiCult Organoid Growth Medium was added into 484 each well. To maintain the cultures, the IntestiCult Organoid Growth Medium was changed three 485 times per week. After 48 hours, the developed organoids were challenged with Ffar2 inhibitor (1 486 µM CATBP) and Ffar3 inhibitor (1 µM Pertussis Toxin) and on day 10, the organoid were 487 challenged with young and old FCM (2.5 mg/ml) and butyrate (6 µM). The organoids were 488 harvested on 12th day for gene expression analyses. Each experiment was conducted 3 times 489 with 3 replica each. 490 Cell Culture. Human HT-29 cells and mouse CMT-93 were purchased from American Type 491 Culture Collection (ATCC). The HT-29 cells (passages 10-15) and CMT93 cells (passages 6-14) 492 were maintained in 4.5 g/L D-Glucose with L-Glutamine DMEM supplemented with 10% fetal 493 bovine serum (FBS) and 100 U/mL penicillin, 100 U/mL streptomycin. The CMT93 cells were 494 grown for 48 hours then treated with young and old FCM (2.5 mg/ml), butyrate (6 µM), Ffar2 495 inhibitor (1 µM CATBP) and Ffar3 inhibitor (1 µM Pertussis Toxin) for 12 hours. Fully 496 differentiated 21-days HT-29 cells were used for young and old FCM treatment (2.5 mg/ml) for a 497 time duration of 8 hours to measure the TEER and FITC diffusion assay. The cells were washed 498 properly with DPBS before harvesting. 499 Measurements of transepithelial electrical resistance (TEER) in HT-29 Cells. Caco-2 cells 500 were seeded on apical chamber made of polyester membrane filters with 0.4 μm pore size of 501 12-well transwell plates at a density of 3 × 105/well. The culture media from both the apical and 502 basolateral compartments was changed every two days. The cells were allowed to fully 503 differentiate for the next 21 days. On the 21st day, the fully differentiated cells were challenged 504 with FCM for 8 hours with continuous measuring of TEER values using an EVOM2 Epithelial 505 Voltohmmeter according to the manufacturer`s instruction. The blank inert resistance value (the 506 insert with only culture media) was subtracted from the measured resistance value of each 507 sample and the final resistance in (reported in ohm×cm2) was calculated by multiplying the 508 sample resistance by the area of the membrane(Ahmadi, Wang, et al., 2020). 509 FITC-dextran permeability assay in HT-29 cells. The HT-29 cell transwell plates were 510 prepared using the same method as in the TEER experiments. On the 21st day, the HT-29 cells 511 monolayer was treated with FCM and 4-kD-FITC-dextran solution (1 mg/ml) was added on the 512 apical (upper) side of the monolayers.The 4-kD-FITC level of the basolateral side was 513 determined using a fluorescent 96-well [late at a 485 nm excitation and at 525 nm emission 514 wavelength using fluorescence 96-well plate reader. Values were normalized using a standard 515 FITC curve as described in our paper(Wang et al., 2020).The assay was conducted three times 516 in triplicate. 517 Gene expression using real-time PCR. Total RNA was extracted using RNeasy Mini Kit 518 following the manufacturing protocol from frozen tissues, intestinal organoids, and cells. cDNA 519 was synthesized using High-Capacity cDNA reverse transcription kit. The gene expression was 520 quantified using ABI 7500 real time PCR machine using powerUp SYBR Green master mix 521 using specific primers (listed in Supplementary Table S3 and S4). Relative gene expression was 522 analyzed using ΔΔCT method by normalizing with 18S as an internal housekeeping control 523 (Ahmadi, Razazan, et al., 2020; Ahmadi, Wang, et al., 2020; Yadav et al., 2017; Yadav et al., 524 2013).

- 11 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

525 Statistical Analyses. Different datasets were analyzed using the student`s t-test. All the 526 statistically analyzed figures are presented in the form of mean ± standard error of mean (SEM). 527 Alpha-diversity indices and the bacterial abundance between the old versus young was 528 compared using the unpaired two-tailed Student’s t-test. Differences in beta-diversity were 529 tested by permutational multivariate analysis of variance (PERMANOVA), a permutation-based 530 multivariate analysis of variance to a matrix of pairwise distance to partition the inter-group and 531 intra-group distance. Random forest analysis (RFA) and principal component analysis (PCA) 532 were analysed in R programming (version 3.6.0 https://www.r-project.org/) using packages 533 “randomForest”, “ggplot2”, “caret”, “psych”, “ggbiplots”, “nnet” and “devtools”. Hierarchical 534 clustering and heatmaps of top 50 genera based on abundance was constructed by using the 535 pheatmap and “ggplots” packages of R v6.0. LEfSE (Linear discriminatory analysis [LDA] Effect 536 Size) was used to identify unique bacterial taxa that drive differences in old vs young mice fecal 537 sample from Galaxy server (https://huttenhower.sph.harvard.edu/galaxy/) (Segata et al., 2011). 538 The alpha parameter significance threshold for the Kruskal–Wallis as well as the Wilcoxon 539 signed-rank test, which were implemented among classes was set to 0.01, the logarithmic LDA 540 score cut-off was set to 3, and the strategy for multi-class analysis was set to “all-against-all”. 541 Welch`s t-test was applied for statistical significance analysis for metabolites using Amix 3.9 542 (Bruker Biospin) and a false discovery rate (FDR) were applied to control the family wised error. 543 Hierarchical clustering and heat-maps were constructed based on the average linkage on 544 Euclidean distance and depict the patterns of abundance and log values of metabolites and the 545 correlation of these metabolites with gut permeability, inflammatory and tight junction marker 546 were constructedin R v6.0 using the pheatmap and “ggplots” packages. The Pearson`s 547 correlation between gut permeability, inflammatory and tight junction markers were calculated 548 between respective groups in IBM SPSS (statistical package for social science) and clustering 549 correlation heatmaps were constructed using the pheatmap package, R v6.0. P<0.05 was 550 considered statistically significant for all the analyses unless otherwise specified. 551 552 Acknowledgement 553 All the authors are grateful for all the fellow staff and laboratory members of Yadav’s lab for their 554 cooperation and help during the study. 555 556 Funding 557 We are thankful for the support provided by National Institutes of Health (NIH) grants 558 R56AG064075, R56AG069676, R21AG072379, and the Department of Defense funding 559 W81XWH-19-1-0236 (HY), as well funds and services provided from the USF Center for 560 Microbiome Research and Center for Excellence in Aging and Brain Repair. 561 562 Declaration of conflict of interest 563 All authors declare no conflict of interest for this work. 564 565 References 566 567 Ahmad, R., Rah, B., Bastola, D., Dhawan, P., & Singh, A. B. (2017, Jul 11). Obesity-induces Organ 568 and Tissue Specific Tight Junction Restructuring and Barrier Deregulation by Claudin 569 Switching. Sci Rep, 7(1), 5125. https://doi.org/10.1038/s41598-017-04989-8 570 571 Ahmadi, S., Nagpal, R., Wang, S., Gagliano, J., Kitzman, D. W., Soleimanian-Zad, S., Sheikh- 572 Zeinoddin, M., Read, R., & Yadav, H. (2019, May). Prebiotics from acorn and sago

- 12 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

573 prevent high-fat-diet-induced insulin resistance via microbiome-gut-brain axis 574 modulation. J Nutr Biochem, 67, 1-13. https://doi.org/10.1016/j.jnutbio.2019.01.011 575 576 Ahmadi, S., Razazan, A., Nagpal, R., Jain, S., Wang, B., Mishra, S. P., Wang, S., Justice, J., Ding, J., 577 McClain, D. A., Kritchevsky, S. B., Kitzman, D., & Yadav, H. (2020, Jun 18). Metformin 578 Reduces Aging-Related Leaky Gut and Improves Cognitive Function by Beneficially 579 Modulating Gut Microbiome/Goblet Cell/Mucin Axis. J Gerontol A Biol Sci Med Sci, 75(7), 580 e9-e21. https://doi.org/10.1093/gerona/glaa056 581 582 Ahmadi, S., Wang, S., Nagpal, R., Wang, B., Jain, S., Razazan, A., Mishra, S. P., Zhu, X., Wang, Z., 583 Kavanagh, K., & Yadav, H. (2020, May 7). A human-origin probiotic cocktail ameliorates 584 aging-related leaky gut and inflammation via modulating the microbiota/taurine/tight 585 junction axis. JCI Insight, 5(9). https://doi.org/10.1172/jci.insight.132055 586 587 Awoyemi, A., Troseid, M., Arnesen, H., Solheim, S., & Seljeflot, I. (2018, Jul 21). Markers of 588 metabolic endotoxemia as related to metabolic syndrome in an elderly male population 589 at high cardiovascular risk: a cross-sectional study. Diabetology & Metabolic Syndrome, 590 10. https://doi.org/ARTN 59 591 10.1186/s13098-018-0360-3 592 593 Belkaid, Y., & Hand, T. W. (2014, Mar 27). Role of the microbiota in immunity and inflammation. 594 Cell, 157(1), 121-141. https://doi.org/10.1016/j.cell.2014.03.011 595 596 Cani, P. D., Bibiloni, R., Knauf, C., Waget, A., Neyrinck, A. M., Delzenne, N. M., & Burcelin, R. 597 (2008, Jun). Changes in gut microbiota control metabolic endotoxemia-induced 598 inflammation in high-fat diet-induced obesity and diabetes in mice. Diabetes, 57(6), 599 1470-1481. https://doi.org/10.2337/db07-1403 600 601 Chelakkot, C., Ghim, J., & Ryu, S. H. (2018, Aug 16). Mechanisms regulating intestinal barrier 602 integrity and its pathological implications. Exp Mol Med, 50(8), 1-9. 603 https://doi.org/10.1038/s12276-018-0126-x 604 605 Disease, G. B. D., Injury, I., & Prevalence, C. (2018, Nov 10). Global, regional, and national 606 incidence, prevalence, and years lived with disability for 354 diseases and injuries for 607 195 countries and territories, 1990-2017: a systematic analysis for the Global Burden of 608 Disease Study 2017. Lancet, 392(10159), 1789-1858. https://doi.org/10.1016/S0140- 609 6736(18)32279-7 610 611 Donohoe, D. R., Garge, N., Zhang, X., Sun, W., O'Connell, T. M., Bunger, M. K., & Bultman, S. J. 612 (2011, May 4). The microbiome and butyrate regulate energy metabolism and 613 autophagy in the mammalian colon. Cell Metab, 13(5), 517-526. 614 https://doi.org/10.1016/j.cmet.2011.02.018 615

- 13 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

616 Elderman, M., Sovran, B., Hugenholtz, F., Graversen, K., Huijskes, M., Houtsma, E., Belzer, C., 617 Boekschoten, M., de Vos, P., Dekker, J., Wells, J., & Faas, M. (2017). The effect of age on 618 the intestinal mucus thickness, microbiota composition and immunity in relation to sex 619 in mice. PLoS One, 12(9), e0184274. https://doi.org/10.1371/journal.pone.0184274 620 621 Ferrucci, L., & Fabbri, E. (2018, Sep). Inflammageing: chronic inflammation in ageing, 622 cardiovascular disease, and frailty. Nat Rev Cardiol, 15(9), 505-522. 623 https://doi.org/10.1038/s41569-018-0064-2 624 625 Gratton, J., Phetcharaburanin, J., Mullish, B. H., Williams, H. R., Thursz, M., Nicholson, J. K., 626 Holmes, E., Marchesi, J. R., & Li, J. V. (2016, May 3). Optimized Sample Handling Strategy 627 for Metabolic Profiling of Human Feces. Anal Chem, 88(9), 4661-4668. 628 https://doi.org/10.1021/acs.analchem.5b04159 629 630 [Record #54 is using a reference type undefined in this output style.] 631 632 Jakobsdottir, G., Xu, J., Molin, G., Ahrne, S., & Nyman, M. (2013, Nov 13). High-Fat Diet Reduces 633 the Formation of Butyrate, but Increases Succinate, Inflammation, Liver Fat and 634 Cholesterol in Rats, while Dietary Fibre Counteracts These Effects. PLoS One, 8(11). 635 https://doi.org/ARTN e80476 636 10.1371/journal.pone.0080476 637 638 Kang, D. W., Adams, J. B., Gregory, A. C., Borody, T., Chittick, L., Fasano, A., Khoruts, A., Geis, E., 639 Maldonado, J., McDonough-Means, S., Pollard, E. L., Roux, S., Sadowsky, M. J., Lipson, K. 640 S., Sullivan, M. B., Caporaso, J. G., & Krajmalnik-Brown, R. (2017, Jan 23). Microbiota 641 Transfer Therapy alters gut ecosystem and improves gastrointestinal and autism 642 symptoms: an open-label study. Microbiome, 5(1), 10. https://doi.org/10.1186/s40168- 643 016-0225-7 644 645 Kennedy, B. K., Berger, S. L., Brunet, A., Campisi, J., Cuervo, A. M., Epel, E. S., Franceschi, C., 646 Lithgow, G. J., Morimoto, R. I., Pessin, J. E., Rando, T. A., Richardson, A., Schadt, E. E., 647 Wyss-Coray, T., & Sierra, F. (2014, Nov 6). Geroscience: linking aging to chronic disease. 648 Cell, 159(4), 709-713. https://doi.org/10.1016/j.cell.2014.10.039 649 650 Koelman, L., Pivovarova-Ramich, O., Pfeiffer, A. F. H., Grune, T., & Aleksandrova, K. (2019). 651 Cytokines for evaluation of chronic inflammatory status in ageing research: reliability 652 and phenotypic characterisation. Immun Ageing, 16, 11. 653 https://doi.org/10.1186/s12979-019-0151-1 654 655 Langille, M. G. I., Meehan, C. J., Koenig, J. E., Dhanani, A. S., Rose, R. A., Howlett, S. E., & Beiko, 656 R. G. (2014). Microbial shifts in the aging mouse gut. Microbiome, 2(1). 657 https://doi.org/10.1186/s40168-014-0050-9 658

- 14 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

659 Lee, J., d'Aigle, J., Atadja, L., Quaicoe, V., Honarpisheh, P., Ganesh, B. P., Hassan, A., Graf, J., 660 Petrosino, J., Putluri, N., Zhu, L., Durgan, D. J., Bryan, R. M., Jr., McCullough, L. D., & 661 Venna, V. R. (2020, Jul 31). Gut Microbiota-Derived Short-Chain Fatty Acids Promote 662 Poststroke Recovery in Aged Mice. Circ Res, 127(4), 453-465. 663 https://doi.org/10.1161/CIRCRESAHA.119.316448 664 665 Li, H., Limenitakis, J. P., Fuhrer, T., Geuking, M. B., Lawson, M. A., Wyss, M., Brugiroux, S., Keller, 666 I., Macpherson, J. A., Rupp, S., Stolp, B., Stein, J. V., Stecher, B., Sauer, U., McCoy, K. D., 667 & Macpherson, A. J. (2015, Sep 22). The outer mucus layer hosts a distinct intestinal 668 microbial niche. Nat Commun, 6, 8292. https://doi.org/10.1038/ncomms9292 669 670 Liu, H., Wang, J., He, T., Becker, S., Zhang, G., Li, D., & Ma, X. (2018, Jan 1). Butyrate: A Double- 671 Edged Sword for Health? Adv Nutr, 9(1), 21-29. 672 https://doi.org/10.1093/advances/nmx009 673 674 Liu, X., Zhao, F., Liu, H., Xie, Y., Zhao, D., & Li, C. (2021, Apr 27). Transcriptomics and 675 metabolomics reveal the adaption of Akkermansia muciniphila to high mucin by 676 regulating energy homeostasis. Sci Rep, 11(1), 9073. https://doi.org/10.1038/s41598- 677 021-88397-z 678 679 Lozupone, C. A., Stombaugh, J. I., Gordon, J. I., Jansson, J. K., & Knight, R. (2012, Sep 13). 680 Diversity, stability and resilience of the human gut microbiota. Nature, 489(7415), 220- 681 230. https://doi.org/10.1038/nature11550 682 683 Lu, Y., Fan, C., Li, P., Lu, Y., Chang, X., & Qi, K. (2016, Nov 28). Short Chain Fatty Acids Prevent 684 High-fat-diet-induced Obesity in Mice by Regulating G Protein-coupled Receptors and 685 Gut Microbiota. Sci Rep, 6, 37589. https://doi.org/10.1038/srep37589 686 687 Luissint, A. C., Parkos, C. A., & Nusrat, A. (2016, Oct). Inflammation and the Intestinal Barrier: 688 Leukocyte-Epithelial Cell Interactions, Cell Junction Remodeling, and Mucosal Repair. 689 Gastroenterology, 151(4), 616-632. https://doi.org/10.1053/j.gastro.2016.07.008 690 691 McGuckin, M. A., Linden, S. K., Sutton, P., & Florin, T. H. (2011, Apr). Mucin dynamics and 692 enteric pathogens. Nat Rev Microbiol, 9(4), 265-278. 693 https://doi.org/10.1038/nrmicro2538 694 695 Mishra, S. P., Karunakar, P., Taraphder, S., & Yadav, H. (2020, Jun 8). Free Fatty Acid Receptors 2 696 and 3 as Microbial Metabolite Sensors to Shape Host Health: Pharmacophysiological 697 View. Biomedicines, 8(6). https://doi.org/10.3390/biomedicines8060154 698 699 Nagpal, R., Mishra, S. P., & Yadav, H. (2020, May 13). Unique Gut Microbiome Signatures Depict 700 Diet-Versus Genetically Induced Obesity in Mice. Int J Mol Sci, 21(10). 701 https://doi.org/10.3390/ijms21103434 702

- 15 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

703 Nagpal, R., Neth, B. J., Wang, S., Craft, S., & Yadav, H. (2019, Sep). Modified Mediterranean- 704 ketogenic diet modulates gut microbiome and short-chain fatty acids in association with 705 Alzheimer's disease markers in subjects with mild cognitive impairment. EBioMedicine, 706 47, 529-542. https://doi.org/10.1016/j.ebiom.2019.08.032 707 708 Nagpal, R., Neth, B. J., Wang, S., Mishra, S. P., Craft, S., & Yadav, H. (2020, Sep). Gut mycobiome 709 and its interaction with diet, gut bacteria and alzheimer's disease markers in subjects 710 with mild cognitive impairment: A pilot study. EBioMedicine, 59, 102950. 711 https://doi.org/10.1016/j.ebiom.2020.102950 712 713 Nagpal, R., Newman, T. M., Wang, S., Jain, S., Lovato, J. F., & Yadav, H. (2018). Obesity-Linked 714 Gut Microbiome Dysbiosis Associated with Derangements in Gut Permeability and 715 Intestinal Cellular Homeostasis Independent of Diet. J Diabetes Res, 2018, 3462092. 716 https://doi.org/10.1155/2018/3462092 717 718 Nagpal, R., Shively, C. A., Appt, S. A., Register, T. C., Michalson, K. T., Vitolins, M. Z., & Yadav, H. 719 (2018). Gut Microbiome Composition in Non-human Primates Consuming a Western or 720 Mediterranean Diet. Front Nutr, 5, 28. https://doi.org/10.3389/fnut.2018.00028 721 722 Nagpal, R., Wang, S., Ahmadi, S., Hayes, J., Gagliano, J., Subashchandrabose, S., Kitzman, D. W., 723 Becton, T., Read, R., & Yadav, H. (2018, Aug 23). Human-origin probiotic cocktail 724 increases short-chain fatty acid production via modulation of mice and human gut 725 microbiome. Sci Rep, 8(1), 12649. https://doi.org/10.1038/s41598-018-30114-4 726 727 Navas-Molina, J. A., Peralta-Sanchez, J. M., Gonzalez, A., McMurdie, P. J., Vazquez-Baeza, Y., Xu, 728 Z. J., Ursell, L. K., Lauber, C., Zhou, H. W., Song, S. J., Huntley, J., Ackermann, G. L., Berg- 729 Lyons, D., Holmes, S., Caporaso, J. G., & Knight, R. (2013). Advancing Our Understanding 730 of the Human Microbiome Using QIIME. Microbial Metagenomics, Metatranscriptomics, 731 and Metaproteomics, 531, 371-444. https://doi.org/10.1016/B978-0-12-407863- 732 5.00019-8 733 734 Okumura, R., & Takeda, K. (2018). Maintenance of intestinal homeostasis by mucosal barriers. 735 Inflamm Regen, 38, 5. https://doi.org/10.1186/s41232-018-0063-z 736 737 Paone, P., & Cani, P. D. (2020, Dec). Mucus barrier, mucins and gut microbiota: the expected 738 slimy partners? Gut, 69(12), 2232-2243. https://doi.org/10.1136/gutjnl-2020-322260 739 740 Pelaseyed, T., Bergstrom, J. H., Gustafsson, J. K., Ermund, A., Birchenough, G. M., Schutte, A., 741 van der Post, S., Svensson, F., Rodriguez-Pineiro, A. M., Nystrom, E. E., Wising, C., 742 Johansson, M. E., & Hansson, G. C. (2014, Jul). The mucus and mucins of the goblet cells 743 and enterocytes provide the first defense line of the gastrointestinal tract and interact 744 with the immune system. Immunol Rev, 260(1), 8-20. 745 https://doi.org/10.1111/imr.12182 746

- 16 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

747 Segata, N., Izard, J., Waldron, L., Gevers, D., Miropolsky, L., Garrett, W. S., & Huttenhower, C. 748 (2011, Jun 24). Metagenomic biomarker discovery and explanation. Genome Biol, 12(6), 749 R60. https://doi.org/10.1186/gb-2011-12-6-r60 750 751 Sovran, B., Hugenholtz, F., Elderman, M., Van Beek, A. A., Graversen, K., Huijskes, M., 752 Boekschoten, M. V., Savelkoul, H. F. J., De Vos, P., Dekker, J., & Wells, J. M. (2019, Feb 753 5). Age-associated Impairment of the Mucus Barrier Function is Associated with 754 Profound Changes in Microbiota and Immunity. Scientific Reports, 9. 755 https://doi.org/ARTN 1437 756 10.1038/s41598-018-35228-3 757 758 Stehle, J. R., Leng, X. Y., Kitzman, D. W., Nicklas, B. J., Kritchevsky, S. B., & High, K. P. (2012, 759 Nov). Lipopolysaccharide-Binding Protein, a Surrogate Marker of Microbial 760 Translocation, Is Associated With Physical Function in Healthy Aged Adults. Journals of 761 Gerontology Series a-Biological Sciences and Medical Sciences, 67(11), 1212-1218. 762 https://doi.org/10.1093/gerona/gls178 763 764 Tachon, S., Zhou, J. N., Keenan, M., Martin, R., & Marco, M. L. (2013, Feb). The intestinal 765 microbiota in aged mice is modulated by dietary resistant starch and correlated with 766 improvements in host responses. Fems Microbiology Ecology, 83(2), 299-309. 767 https://doi.org/10.1111/j.1574-6941.2012.01475.x 768 769 Van der Sluis, M., De Koning, B. A. E., De Bruijn, A. C. J. M., Velcich, A., Meijerink, J. P. P., Van 770 Goudoever, J. B., Buller, H. A., Dekker, J., Van Seuningen, I., Renes, I. B., & Einerhand, A. 771 W. C. (2006, Jul). Muc2-deficient mice spontaneously develop colitis, indicating that 772 Muc2 is critical for colonic protection. Gastroenterology, 131(1), 117-129. 773 https://doi.org/10.1053/j.gastro.2006.04.020 774 775 Vemuri, R., Sherrill, C., Davis, M. A., & Kavanagh, K. (2020, Oct 6). Age-Related Colonic Mucosal 776 Microbiome Community Shifts in Monkeys. J Gerontol A Biol Sci Med Sci. 777 https://doi.org/10.1093/gerona/glaa256 778 779 Visconti, A., Le Roy, C. I., Rosa, F., Rossi, N., Martin, T. C., Mohney, R. P., Li, W. Z., de Rinaldis, E., 780 Bell, J. T., Venter, J. C., Nelson, K. E., Spector, T. D., & Falchi, M. (2019, Oct 3). Interplay 781 between the human gut microbiome and host metabolism. Nature Communications, 10. 782 https://doi.org/ARTN 4505 783 10.1038/s41467-019-12476-z 784 785 Wagner, B. D., Grunwald, G. K., Zerbe, G. O., Mikulich-Gilbertson, S. K., Robertson, C. E., 786 Zemanick, E. T., & Harris, J. K. (2018, May 22). On the Use of Diversity Measures in 787 Longitudinal Sequencing Studies of Microbial Communities. Frontiers in Microbiology, 9. 788 https://doi.org/10.3389/fmicb.2018.01037 789

- 17 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

790 Wang, S., Ahmadi, S., Nagpal, R., Jain, S., Mishra, S. P., Kavanagh, K., Zhu, X., Wang, Z., McClain, 791 D. A., Kritchevsky, S. B., Kitzman, D. W., & Yadav, H. (2020, Feb). Lipoteichoic acid from 792 the cell wall of a heat killed Lactobacillus paracasei D3-5 ameliorates aging-related leaky 793 gut, inflammation and improves physical and cognitive functions: from C. elegans to 794 mice. Geroscience, 42(1), 333-352. https://doi.org/10.1007/s11357-019-00137-4 795 796 Wilms, E., Troost, F. J., Elizalde, M., Winkens, B., de Vos, P., Mujagic, Z., Jonkers, D. M. A. E., & 797 Masclee, A. A. M. (2020, Jan 16). Intestinal barrier function is maintained with aging - a 798 comprehensive study in healthy subjects and irritable bowel syndrome patients. 799 Scientific Reports, 10(1). https://doi.org/ARTN 475 800 10.1038/s41598-019-57106-2 801 802 Wrzosek, L., Ciocan, D., Borentain, P., Spatz, M., Puchois, V., Hogot, C., Ferrere, G., Mayeur, C., 803 Perlemuter, G., & Cassard, A. M. (2018, May 1). Transplantation of human microbiota 804 into conventional mice durably reshapes the gut microbiota. Scientific Reports, 8. 805 https://doi.org/ARTN 6854 806 10.1038/s41598-018-25300-3 807 808 Yadav, H., Devalaraja, S., Chung, S. T., & Rane, S. G. (2017, Feb 24). TGF-beta1/Smad3 Pathway 809 Targets PP2A-AMPK-FoxO1 Signaling to Regulate Hepatic Gluconeogenesis. J Biol Chem, 810 292(8), 3420-3432. https://doi.org/10.1074/jbc.M116.764910 811 812 Yadav, H., Lee, J. H., Lloyd, J., Walter, P., & Rane, S. G. (2013, Aug 30). Beneficial metabolic 813 effects of a probiotic via butyrate-induced GLP-1 hormone secretion. J Biol Chem, 814 288(35), 25088-25097. https://doi.org/10.1074/jbc.M113.452516 815 816 Zuo, L., Kuo, W. T., & Turner, J. R. (2020). Tight Junctions as Targets and Effectors of Mucosal 817 Immune Homeostasis. Cellular and Molecular Gastroenterology and Hepatology, 10(2), 818 327-340. https://doi.org/10.1016/j.jcmgh.2020.04.001 819 820

- 18 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

821 FIGURES AND FIGURE LEGENDS 822 823 Figure 1. An aged gut harbors significantly distinct microbiota than a younger gut, which 824 is causal to induce leaky gut and inflammation. a) Principal component analysis (PCA) as a 825 b-diversity indices revealed that aged gut microbiome are significantly distinct from a younger 826 gut. b) The microbiome diversity a-diversity indices indicated by the Shannon index was 827 significantly lower in feces from an aged gut as compared to a young gut. c-f) The abundance of 828 major phyla, genera and species presented in bar-graphs (c-d) and Linear discriminant analysis 829 (LDA) effect size LEfSe cladogram is significantly distinct in an aged microbiome compared to a 830 younger microbiome. g-j) Aged mice tend to have an increased amount of leakiness from the 831 gut as represented by FITC-dextran leakage from gut to blood, and increased expression of IL- 832 1β, IL-6 and TNF-α) in the gut along with higher circulating IL-6 and TNF-a as compared to a 833 younger gut. k-p) FMT of aged feces significantly increased leaky gut (FITC-dextran leakage) 834 and its systemic markers like LBP and sCD14 in serum along with increase in inflammatory 835 markers (IL-1β, IL-6 and TNF-α) in gut and IL-6 and TNF-a in serum of recipient mice compared 836 to young FMT recipients. q-r) Treatment of FCM made from aged gut feces significantly reduced 837 transepithelial electrical resistance (TEER) and increased FITC-dextran permeability in the 838 monolayers of HT-29 cells compared to FCM of younger gut feces. All the values presented are 839 mean of 5-9 animals in each group and error bars represent standard error of means. P-values 840 with **p<0.01, ***p<0.001 are statistically significant. 841 842 Figure 2. Aged gut microbiota disrupts gut mucin barriers, which in turn induces leaky 843 gut and inflammation. a-e) The expression of mucin isomers like Muc2 and Muc6 (a,c,e), and 844 tight junction protein like Tjp1/zonulin and Ocln (b,d) genes was significantly reduced in the gut 845 of old FMT recipient mice (a,b) and old FCM treated organoids (c,d) and CMT93 goblet cell line 846 (e) in comparison to their corresponding controls. f-h) Pearson`s correlation heatmap (f), 847 random forest analysis (g) and volcano plot (h) representing data of mucin, inflammation and 848 gut permeability markers in the gut of aged FMT recipient mice and FCM treated organoids and 849 CMT93 cells shows Muc2 was the highest influenced gene from an old microbiome. i-j) 850 Similarly, goblet cell numbers were significantly lower in the intestine of old FMT recipients than 851 young FMT recipient controls. k-o) the expression of mucin (k), tight junction proteins (l), goblet 852 cell numbers (m,n) and mucin in feces (o) were significantly lower in aged mice as compared to 853 younger mice. All the values presented are mean of 5-9 animals in each group and error bars 854 represent standard error of means. P-values with *p<0.05, **p<0.01, ***p<0.001 are statistically 855 significant. 856 857 Figure 3. The abundance of short chain fatty acids (SCFAs) like butyrate was 858 significantly reduced in an aged gut as compared to a younger gut, due to reduced 859 mucin barriers. a) Pearson correlation between the mucin, tight junction, leaky gut and 860 inflammatory markers in the gut of aged mice further show Muc2 was dominantly reduced in the 861 aged gut. b,e) Principal component analysis (PCA) (b) and hierarchical differential clustering (e) 862 analyses revealed that an aged gut has significantly different metabolites than a young gut. c,d) 863 Differential abundance analyses of metabolites in volcano plot (c) and random forest analysis 864 show that butyrate was the most significantly reduced metabolite in the aged gut compared to a 865 younger gut. f-h) Further, correlation analyses of Muc2 expression with acetate, propionate and 866 butyrate revealed butyrate was more significantly correlated with Muc2 than acetate and 867 propionate. All the values presented are mean of 5-9 samples in each group. 868 869 Figure 4. Butyrate treatment decreases leaky gut and inflammation and enhances mucus 870 barriers. a-e) Butyrate feeding in drinking water significantly reduced leakiness (FITC-dextran

- 19 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

871 leakage from gut to blood) (a) and inflammation (IL-6, TNF-a, MCP-1 and PAI-1) (b) and 872 increased goblet cell numbers (c) and expression of Muc2 in the gut along with (d) and mucin in 873 feces of mice fed with a high fat diet as compared to non-butyrate treated controls. e) In 874 addition, butyrate treatment significantly increased mucin production (revealed by Periodic acid– 875 Schiff (PAS) staining) in CMT-93 cells compared to non-treated controls. f-g) Pre-butyrate 876 treatment abrogated detrimental effects of old FCM on the expression of mucin genes in both 877 CMT93 cells (g) and intestinal organoids (h). All the values presented are mean of 5-8 animals, 878 and 3-4 independent replicates in cells and organoid cultures in each group and error bars 879 represent standard error of means. P-values with *p<0.05, **p<0.01, ***p<0.001 are statistically 880 significant. *p<0.05, **p<0.01, ***p<0.001 881 882 Figure 5. Microbiota reduces FFAR2/3 expression in an aged gut which further depletes 883 butyrate action; Depletion of FFAR2/3 develops early aging in gut. a) The Ffar2 and Ffar3 884 gene expression was significantly reduced in the gut of old donor mice as compared to young 885 donor mice. b-e) Old FMT and FCM also reduced FFAR2/3 expression in the intestine of the 886 recipient mice gut (b) and treated organoids (c), CMT93 cells (d) and HT29 cells (e) compared 887 to their controls. f-g) Inhibition of FFAR2/3 dampened the effects of butyrate to induce 888 expression of Muc2, Muc6 and Muc13) in CMT93 (f) and intestinal organoids (g), as well as 889 mucin production indicated by PAS staining in CMT93 cells (h). i-p) Intestine specific FFAR2 890 (intFFAR2 KO) and FFAR3 (intFFAR3 KO) knockout mice show significantly leakier gut (FITC- 891 dextran leakage) (i,m) and expression of inflammatory genes (Il-1β, Il-6 and Tnf-α) (j,n) along 892 with reduced mucin genes (Muc2 and Muc6) (k,o) and tight junction genes (Tjp1 and Ocln) (l,p) 893 in their gut compared to wild type (intFFAR2-WT and intFFAR3-WT) controls. All the values 894 presented are mean of 5-8 animals, and 3-4 independent replicates in cells and organoid 895 cultures in each group and error bars represent standard error of means. P-values with *p<0.05, 896 **p<0.01, ***p<0.001 are statistically significant. *p<0.05, **p<0.01, ***p<0.001

- 20 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

897 SUPPLEMENTARY FIGURES AND FIGURE LEGENDS 898 899 Supplementary Figure S1. Microbiota in aged donor mice is significantly distinct from 900 young donor mice. a) Alpha diversity indices show the number of observed is significantly 901 lower in the feces of aged mice compared to younger mice. b) The ratio of 902 Firmicutes/Bacteroidetes was significantly higher in aged mice feces compared to younger 903 counterparts. c-e) The abundance of major class (c), order (d) and family (e) of bacteria that are 904 distinct in aged versus younger gut. f) Hierarchical heatmap of top 50 genus based on 905 abundance in old vs young mice feces. g-h) Linear discriminant analysis (LDA) effect size LEfSe 906 of old microbiome at genera and species level as compared to younger microbiome. All the 907 values presented are mean of 5-9 animals in each group and error bars represent standard 908 error of means. P-values with ***p<0.001 are statistically significant. 909 910 Supplementary Figure S2. The abundance of metabolites was significant different in aged 911 versus younger gut and are distinctly correlated with mucin, tight junctions and leaky 912 and inflammatory markers. a) The abundance of SCFAs (acetate, propionate and butyrate) 913 was significantly lower in the feces of aged mice compared to younger mice. b) Heatmap of 914 Pearson correlation values among microbiome metabolites and mucin, tight junction protein, 915 leaky gut and inflammatory markers revealed significantly unique clustering, with the highest 916 correction in butyrate and Muc2 gene expression. 917

- 21 - bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.Mishra It is et made al Figure 1 available under aCC-BY 4.0 International license. Alpha Diversity Major Phyla Major Genera a b c d 100% 100% Other 2.5 90% 90% Bacteroides 10 80% Other 80% 2 *** Lachnospiracea 70% Tenericutes 70% Turicibacter 1.5 60% Actinobacteria 60% Alistipes 0 50% Deferribacteres Parabacteroides 1 50% 40% Proteobacteria Lactobacillus PC2 (20.5%) PC2 40% 30% Acetobacteroides 0.5 Verrucomicrobia 30% Akkermansia -10 20% Firmicutes 20% Clostridium_XlVa 0 10% Bacteroidetes Shannon Species Diversity Dysgonomonas 0% 10% Old Barnesiella Young 0% 0 Old -10 10 Old Young PC1 (43.7%) Young e f Major Species 100% Other 90% Flintibacter_butyricus 80% Eubacterium_cellulosolvens 70% Faecalibaculum_rodentium 60% Barnesiella_intestinihominis Lactobacillus_intestinalis 50% Mu cispirillum_schaedleri 40% Acetatifactor_muris 30% Turicibacter_sanguinis 20% Bacteroidetes_bacterium 10% Akkermansia_muciniphila Mu rib aculum_intestinale 0% Old Young Donor Mice Recipient FMT Mice

100 g 6 h 180 i 350 j 60 k 60 l 35 m *** *** ** *** *** *** 30 80 5 150 300 *** 50 50 *** 250 25 4 120 40 40 60 *** (pg/ml)] 20 α 200 3 90 - 30 30 6 (pg/ml)]

40 - 150 15 2 60 20 20 100 10 20 1 10 30 50 Serum LBP (ng/ml)] 10 5 Serum sCD14 (ng/ml)] sCD14 Serum Serum IL Serum TNF Young Old Old Young Young Old 0 0 0 0 0 0 Old 0 Gut permeability FITC(µg/ml) permeability Gut

IL-1β IL-6 Tnf-α FITC(µg/ml) permeability Gut Old Old Old Old Old Old Young Young Young Recipient FMT Mice HT-29 Cells 120 250 1400 5 n o p q 50 r *** *** *** 100 *** 1200 *** 200 40 ****** 4 1000 *** 80 *** (pg/ml)] *** 150 800 30 3 α

*** - 6 (pg/ml)] *** - 60 *** *** 600 20 *** 20 *** 2 100 0 *** 40 400 *** 10

TEER (ohm.cm2) TEER -10 Young Young Young *** FITC (µg/ml) FITC change/18S) -20 *** 10 0

1 Serum IL 20 50 200 *** -30 *** *** Serum TNF 0 Old 0 Gene expression (Fold (Fold expression Gene Old ung Old Old Old 0 0 un 0 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 Yo IL-1β IL-6 Tnf-α Yo g Hour Hour bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.Mishra It is et made al Figure 2 available under aCC-BY 4.0 International license. FMT Mice Organoid 1.2 a 1.2 b 1.5 c 1.4 d 1 1 1.2 0.8 0.8 1 1 ** 0.8 0.6 0.6 *** * *** *** 0.6 0.4 0.4 0.5 *** *** 0.4 *** change/18S) 0.2 0.2 0.2 Gene expression expression Gene (Fold change/18S) (Fold Young Young Young Old Old Young Old Old Young Old Old Young Young Old Old Young Young Young 0 0 Old Old 0 (Fold expression Gene 0 Muc2 Muc6 Muc13 Tjp1 Ocln Muc2 Muc6 Muc13 Tjp1 Ocln Correlation value

-.611 1 e CMT93 f 1 .450 .547 .243 -.722 -.583 -.507 -.147 .634 .492 .387 .362 FITC .061 1 .425 .424 -.734 -.597-.643 -.692 -.134 .763 .725 .455 .486 IL1b 1.2 .019 .079 1 .737 .572 -.634 -.672-.271 -.216 .832 .697 .772 .548 IL6 1 .332 .080 4e-4 1 -.345 -.591-.523 -.235-.048 .686 .639 .608 .508 TNFa 0.5 0.8 .001 .001 .013 .160 1 .628 .653 .500 .321 -.723 -.696 -.539 -.491 ** Tjp1 0.6 .011 .009 .005 .010 .005 1 .846 .691 .206 -.846 -.750 -.772-.677 Ocln .026 0.4 *** .007 .004 .002 .003 1e-5 1 .822 .448 -.866 -.849 -.783 -.716 Muc 2 0 .032 .001 .277 .347 .034 .002 2e-5 1 .466 -.629 -.577 -.434 -.408 Muc 6 change/18S) 0.2 .561 .596 .389 .849 .194 .411 .062 .051 1 -.180-.189 -.188-.064 Muc 13 Old Old Young Old Young 0 Young Gene expression (Fold (Fold expression Gene .005 2e-4 .000 .002 .001 1e-5 3e-6 .005 .476 1 .895 .772 .703 LBP_ELISA Muc2 Muc6 Muc13 -0.5 .001 .001 .004 .001 8e-6 .012 g .038 3e-4 .452 1e-6 1 .750 .751 CD14_ELISA .112 .058 1e-4 .007 .021 1e-4 1e-4 .072 .455 1e-4 3e-4 1 .847 IL6_ELISA Mean decrease gini .140 .041 .019 .031 .039 .002 .001 .093 .800 .001 3e-4 9e-6 1 TNFa_ELISA Correlationsignificance level FITC Il1b Il6 Tnfa Tjp1 Ocln Muc2 Muc6 Muc13 LBP_ELISA CD14_ELISA IL6_ELISA TNFa_ELISA Muc2

Il1b 732021_Mucin_Tjps_ELISA_FITC_VP_FMTh Old Mice FITC 6 Tnfa Muc2 Il6_elisa

CD14_elisa 5 Ocln Il6 FMT Recipient Mice Il1b j Il6 4 Tjp1 i 120 Value) - 100 Ocln Tnfa 3 Tjp1 80 log10pvalue log10(p 2 - 60 Muc6 villi *** LBP_elisa 40 per 1 TNFa_elisa 20 IL6_elisa 0

-2 -1 0 1 2 percentage cell Goblet Muc13 Young Old log2(foldlog2fc change) Old Young

0.0 0.2 0.4 0.6 Donor Mice

m 0.5 1.4 k 1.2 l 120 n o 1.2 1 100 0.4 1 0.8 80 0.8 0.3 0.6 60 villi 0.6 0.2 *** 0.4 *** *** 40 0.4 per

change/18S) *** *** 0.1 0.2 *** 0.2 20 Mucin (mg/g of feces) of (mg/g Mucin

Gene expression (Fold (Fold expression Gene 0 0 0 0 Mu c2 Mu c6 Mu c13 Tjp1 Ocln percentage cell Goblet Young Old Old Old Young Old Young Young bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.Mishra It is et made al Figure 3 available under aCC-BY 4.0 International license. a Correlation value b 10 1 1 .748 .450 .455 -.645 -.788 -.766 -.686 -.012 .822 .717 FITC 5 3e-4 1 .425 .424 -.734 -.597 -.728 -.694 -.173 .749 .608 Il1b 6e-2 7e-2 1 .737 -.572 -.634 -.716 -.666 -.221 .817 .614 Il6 0.5 0 5e-2 7e-2 4e-4 1 -.345 -.591 -.658 -.666 -.082 -.671 .346 Tnf a

3e-3 5e-4 1e-2 1e-1 1 .628 .739 .716 .183 -.745 -.707 Tjp1 PC2(16.8%) -5 1e-4 8e-3 4e-3 9e-3 5e-3 1 .785 .727 .004 -.823 -.748 Ocln 0

2e-4 6e-4 8e-4 1e-4 .973 .111 -.885 -.704 2e-3 4e-4 1 Muc 2 -10 1e-3 1e-3 2e-3 2e-3 8e-4 6e-4 1e-11 1 -.017 -.83 -.665 Muc 6 -8 -4 0 4 8 9e-1 4e-1 .3e-1 7e-1 4e-1 9e-1 6e-1 9e-1 1 -.107 .167 Muc 13 -0.5 PC1(47.3%) 2e-5 3e-4 3e-5 2e-3 3e-4 2e-5 1e-6 2e-5 6e-1 1 .803 IL6_ELISA 8e-4 7e-3 6e-3 1e-1 1e-3 3e-4 1e-3 2e-3 5e-1 6e-5 1 TNFa_ELISA Correlationsignificance level Anserine FITC Il1b Il6 Tnfa Tjp1 Ocln Muc2 Muc6 Muc13 IL6_ELISA TNFa_ELISA e 2 Hypoxanthine N-A cetylcysteine Volcano Plot of Young vs Old Metabolites Ethanolamine Glycine Mean decrease accuracy 1 c d Formate Butyrate Histidine Methanol Butyrate 3.53.5 Anserine 5-Aminopentanoate Taurine 0 Fumarate Nicotinate 3 3.0 Propionate Fructose Acetate Total bile acid & Cholate Tartrate 2.52.5 cholesterol

Value) Acetate Total bile acid and cholesterol -1 -

2 2.0 Taurine Anserine Arabinose

log10pvalue Ethanol 1.5 Methanol 1.5 Cholate Threonine -2

log10(p Valine - 3-Methyl-2-oxovalerate

1 1.0 3-Phenylpropionate Serine Methionine

0.50.5 Glycine 3-Hydroxyisobutyrate Methy lamine Galactose --22 --11 0 1 Anserine 1 Total bile acid & cholesterol Citrulline 2 Hypoxanthine log2(foldlog2fc change) 3-Methyl-2-oxovalerate Leucine N-A cetylcysteine Xylose Xylose Ethanolamine f 2.5 Glutamine 1 Choline Formate Alanine Isoleucine Histidine 2 Lactate 4-Hydroxyphenylacetate 5-Aminopentanoate R² = 0.3173 Lysine Tartrate Nicotinate 0 1.5 Fructose Ornithine Arabinose UDP-Glucose Tartrate

Muc2 Pheny lalanine 1 Ethanolamine Total bile acid and cholesterol -1 Tyrosine 2-Hydroxyisobutyrate Arabinose Cholate 0.5 Leucine Ethanol Valine Pyruvate Threonine -2 0 3-Methyl-2-oxovalerateFucose Glutamine Is oleuc ine 0 1 2 Serine Hypoxanthine Methionine Uracil Acetate Allantoin g Formate Methy lamine Succinate 1.8 Dimethylamine Galactose Aspartate Citrulline 1.5 Fumarate Adenine Leucine Ethanol Glucose 1.2 Alanine Xylose Glutamine 3-Phenylpropionate 0.9 R² = 0.1017 Alanine Glutamate Py r uv ate Muc2 Lactate h 0 2 4 Dimethylamine 0.6 Lysine Sarcosine 2 Ornithine 0.3 Total bile acid and cholesterol_1 R² = 0.7759 UDP-Glucose Fumarate Pheny lalanine 0 1.5 Methanol Tyrosine 0 1 2 Cholate 3-Hydroxyisobutyrate Propionate 1 Valine Acetate Taurine Fucose Butyrate Muc2 Is oleuc ine 0.5 Glycine Uracil 4-Hydroxyphenylacetate Allantoin Pr opionate 0 Succinate 2-Hydroxyisobutyrate maybe Aspartate 0 1 2 Choline

Young AdenineOld Butyrate Glucose 3-Phenylpropionate Glutamate Py r uv ate Dimethylamine Sarcosine Total bile acid and cholesterol_1 Fumarate Methanol 3-Hydroxyisobutyrate Acetate Taurine Butyrate Glycine 4-Hydroxyphenylacetate Pr opionate 2-Hydroxyisobutyrate maybe Choline Young Old bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY 4.0 International license.

Mishra et al Figure 4 a b c d 50 e

- 1.5 200 3 0.4 40 Control Butyrate * * * 150 *** 0.3 30 1 2 * * 100 0.2 20 ** 0.5 *** 1

change/18S) 50 0.1 dextran (ng/mL) dextran 10 Serum 4.4kDa FITC 4.4kDa Serum Mucin (mg/ g feces) g (mg/ Mucin Butyrate Butyrate Control Control Butyrate Butyrate Control Control mRNA expression (fold (fold mRNAexpression 0 0 0 (%) numbers cell Goblet 0 0 a -6 - 1 -1 change/18S) mRNA fold Muc2 Il - Tnf Pai Mcp f

Control Butyrate PAS Staining in CMT93 Cells

g h Young+Butyrate Old CMT-93 Cells Organoid 1.4 1.4 Old+Butyrate 1.2 1.2 1 1 * 0.8 * 0.8 * 0.6 0.6 *** 0.4 0.4 *** change/18S) 0.2 0.2 0 0 Gene expression (Fold (Fold expression Gene Mu c2 Mu c6 Mu c13 Muc2 Muc6 Muc13 bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.Mishra It is et made al Figure 5 available under aCC-BY 4.0 International license. Donor Mice FMT Recipient Mice Organoid CMT93 Cells HT29 Cells a b c d e 1.2 1.2 1.2 1.4 1.2 Young 1.2 Old 1 1 1 1 1 0.8 0.8 0.8 0.8 ** 0.8 ** *** 0.6 0.6 0.6 ** 0.6 *** *** 0.6 *** *** 0.4 *** 0.4 0.4 0.4 0.4 *** 0.2 0.2 0.2 0.2 0.2 Gene expression expression Gene (Fold change/18S) (Fold 0 0 0 0 0 Ffar2 Ffar3 Ffar2 Ffar3 Ffar2 Ffar3 Ffar2 Ffar3 Ffar2 Ffar3 f g h CMT93 Cells Muc 2 1.5 Intestinal Organoid Muc 2 1.5 Muc 6 Muc 6 DMSO Ffar2 Inhibitor Ffar3 Inhibitor Muc 2 1.5 Muc 2 1.5 Muc 13 1 Muc 13 1 Muc 6 Muc 6 Spdef Spdef Muc 13 1 0.51 0.5 Muc 13 Control Butyrate Inhibitor Ffar2 DMSO Inhibitor Ffar3 Inhibitor Ffar2 Inhibitor Ffar3 Control DMSO Inhibitor Ffar2 Butyrate

Ffar3 Inhibitor Ffar3 Gf i

Gf i Inhibitor Ffar3 Spdef Inhibitor Ffar2

Spdef No treatment Elf 3 ELF3 0.5 Gf i 0.5 0 0 Gf i Cells Math1 Math1 Elf 3 ELF3 0 -0.50 Tf f 3 -0.5 Tf f 3 Control DMSO Butyrate Ffar2.KO Ffar3.KO Ffar2.KO_Butyrate Ffar3.KO_Butyrate Control DMSO Butyrate Ffar2.KO Ffar3.KO Ffar2.KO_Butyrate Ffar3.KO_Butyrate Math1 Math1 - - - - Butyrate Butyrate Butyrate Butyrate Butyrate Tf f 3 -0.5 Tf f 3 -1-0.5 PAS Staining in CMT93 Control DMSO Butyrate Ffar2.KO Ffar3.KO Ffar2.KO_Butyrate Ffar3.KO_Butyrate -1 Control DMSO Butyrate Ffar2.KO Ffar3.KO Ffar2.KO_Butyrate Ffar3.KO_Butyrate

-1 -1.5-1

intFFAR2VilCre Mice -1.5

50 4 1.4 l i j *** k 1.2 ** 3.5 40 1.2 1 3 1 0.8 30 2.5 *** ** 0.8 *** 2 *** *** 0.6 20 0.6 ** 1.5 0.4 change/18S) 0.4 1 10 0.2 0.2

Gene expression (Fold (Fold expression Gene 0.5 0 0 0 Gut permeability FITC(µg/ml) permeability Gut 0 FITC Il-1b Il-6 Tnf-a Mu c2 Mu c6 Mu c13 Tjp-1 Ocln-1

intFFAR2_WT intFFAR2_KO

intFFAR3VilCre Mice

50 3 1.4 1.2 m ** n ** ** o p 1.2 40 2.5 ** 1 1 2 0.8 30 0.8 *** 1.5 0.6 *** 20 0.6 *** 1 0.4 change/18S) 0.4 *** 10 0.5 0.2 0.2 Gene expression (Fold (Fold expression Gene

Gut permeability FITC(µg/ml) permeability Gut 0 0 0 0 FITC Il-1b Il-6 Tnf-a Muc2 Muc6 Muc13 Tjp1 Ocln1 intFFAR3_WT intFFAR3_KO bioRxiv preprint doi: https://doi.org/10.1101/2021.08.18.456856; this version posted August 19, 2021. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to displayMishra the preprintet alSupplementary in perpetuity. It is made Figure 1 c available under aCC-BY 4.0 International license. a Major Class d Major Order e Major Family 100% 100% 100% 90% Other Other Other 500 90% 90% Actinobacteria Campylobacterales Clostridiaceae_1 80% 80% 400 Negativicutes Selenomonadales 80% Desulfovibrionaceae 70% 70% Deferribacteres Deferribacterales 70% Bacteroidaceae 300 60% Rikenellaceae *** Deltaproteobacteria 60% Desulfovibrionales 60% Erysipelotrichaceae 50% Erysipelotrichia 50% 200 Erysipelotrichales 50% Verrucomicrobiae Lactobacillaceae 40% 40% Lactobacillales Bacilli 40% Verrucomicrobiaceae 100 30% 30% Verrucomicrobiales Ruminococcaceae Clostridia 30% 20% Clostridiales Lachnospiraceae 0 Bacteroidia 20% 20% Porphyromonadaceae Number of Species Identified Bacteroidales 10% 10% Old 10% Young 0% 0% 0% Old Young Old Old Young Young Top 50 genus based on Abundance b f h Species LDA > 3.0 Sporobacter Guggenheimella 2 Anaerotruncus Sporobacter 1.8 *** Guggenheimella Intes tinimonasRatio 4 Acetobacteroides1.6 Anaerotruncus Alkaliphilus Intes tinimonas 4 Oscillibacter 1.4 Acetobacteroides Pr oteinibor us 1.2 2 Alkaliphilus Clos tridium_XlV b Oscillibacter 1 Pr oteinibor us 2 Lactobacillus Clos tridium_XlV b Tannerella 0.8 Akkermansia 0 Lactobacillus Bacteriodetes 0.6 Tannerella Mic r obac ter Akkermansia 0 Johnsonella 0.4 Mic r obac ter Por phy r omonas -2 Pr ev otella 0.2 Johnsonella Por phy r omonas -2 Syntrophococcus0 Pr ev otella Mar v inbr y antia Syntrophococcus Par as utter ellaFirmicutes/ -4 Old Mar v inbr y antia Ros eburia -4 Clos tridium_XlV a Young Par as utter ella Ruminoc oc c us 2 Ros eburia Clos tridium_IV Clos tridium_XlV a Lachnospiracea_incertae_sedis Ruminoc oc c us 2 Brassicibacter Clos tridium_IV Blautia Lachnospiracea_incertae_sedis Falsiporphyromonas Brassicibacter Bifidobacteriumg Blautia Odoribacter Falsiporphyromonas Helic obac ter Bifidobacterium Faecalitalea Odoribacter Muc is pir illum Genus LDA > 3.5 Helic obac ter Des ulf ov ibrio Faecalitalea Coprobac ter Muc is pir illum Dy s gonomonas Des ulf ov ibrio Alistipes Coprobac ter Bacteroides Dy s gonomonas Clostridium_sensu_stricto Alistipes Butyricicoccus Bacteroides Clos tridium_III Clostridium_sensu_stricto Par abac ter oides Butyricicoccus Coprobac illus Clos tridium_III Ruminoc oc c us Par abac ter oides Eis enber giella Coprobac illus Barnesiella Ruminoc oc c us Eubac ter ium Eis enber giella Acetatifactor Barnesiella Anaerostipes Eubac ter ium Natranaerov irga Acetatifactor Turicibacter Anaerostipes Young Old Natranaerov irga Turicibacter Young Old 0.5 0 -0.5 Glycine Ornithine Glutamine 3-Phenylpropionate 2_Hydroxyisobutyrate Methanol Choline Serine Citrulline Galactose N_Acetylcysteine Lysine Propionate Alanine Fructose Fucose Lactate Leucine Valine Cholate Anserine Isoleucine cholesterol and acid bile Total 3_Methyl2oxovalerate Formate Ethanolamine 5_Aminopentanoate Allantoin Uracil Threonine Sarcosine Adenine Dimethylamine Methylamine Arabinose Histidine Tartrate Ethanol Aspartate cholesterol_1 and acid bile Total Methionine 3_Hydroxyisobutyrate Nicotinate 4_Hydroxyphenylacetate Butyrate Acetate Taurine Fumarate Pyruvate Glutamate Phenylalanine Tyrosine Glucose Succinate Hypoxanthine UDP_Glucose Xylose Tjp1 The copyright holder for this preprint this for holder copyright The Muc2 Muc6 Mishra et alSupplementary Figure 2 alSupplementary et Mishra Ocln1 . Muc13 IL6_ELISA FITC TNFa_ELISA Tnfa IL6 IL1b this version posted August 19, 2021. 2021. 19, August posted version this ; Old Donor mice metabolite Correlation Heatmap Correlation metabolite mice Donor Old CC-BY 4.0 International license International 4.0 CC-BY b available under a under available Young Old

*** https://doi.org/10.1101/2021.08.18.456856 Butyrate * doi: doi:

**

Propionate Acetate 0 1

0.2 0.6 0.4 0.8 1.4 1.2 bioRxiv preprint preprint bioRxiv (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made made It is in perpetuity. preprint the to display a license bioRxiv granted has who the author/funder, is review) peer by certified not was (which Change Fold a