Multiomics analysis reveals the biological effects of live Roseburia intestinalis as a high‐butyrate‐producing bacterium in human intestinal epithelial cells

Butyrate‐producing bacteria play a key role in human health, and recent studies have triggered interest in their development as next‐generation probiotics. However, there remains limited knowledge not only on the identification of high‐butyrate‐producing bacteria in the human gut but also in the metabolic capacities for prebiotic carbohydrates and their interaction with the host. Herein, it was discovered that Roseburia intestinalis produces higher levels of butyrate and digests a wider variety of prebiotic polysaccharide structures compared with other human major butyrate‐producing bacteria (Eubacterium rectale, Faecalibacterium prausnitzii, and Roseburia hominis). Moreover, R. intestinalis extracts upregulated the mRNA expression of tight junction proteins (TJP1, OCLN, and CLDN3) in human intestinal epithelial cells more than other butyrate‐producing bacteria. R. intestinalis was cultured with human intestinal epithelial cells in the mimetic intestinal host–microbe interaction coculture system to explore the health‐promoting effects using multiomics approaches. Consequently, it was discovered that live R. intestinalis only enhances purine metabolism and the oxidative pathway, increasing adenosine triphosphate levels in human intestinal epithelial cells, but that heat‐killed bacteria had no effect. Therefore, this study proposes that R. intestinalis has potentially high value as a next‐generation probiotic to promote host intestinal health.


INTRODUCTION
The intestinal tract of humans harbors complex microbial ecosystems that contain several hundred bacterial species, whose genetic contents are >150 times more diverse than those of their human hosts. [1]e genetic material present in the gut microbiome encodes various proteins that catabolize indigestible nutrients and produce diverse metabolites, [2] and several metabolites, such as short-chain fatty acids, can improve host health. [3]Therefore, there is a strong research emphasis on discovering intestinal bacteria that produce beneficial metabolites to help enhance gut health and resolve host diseases. [4]tyrate, which is mainly fermented by the gut clostridial cluster, is well known to maintain gut health and alleviate inflammatory bowel disease (IBD), diabetes, and colon cancer. [5]Butyrate is mainly produced in the human large intestine via bacterial fermentation as humans do not encode metabolic enzymes for producing butyrate. [6]Butyrate is rapidly absorbed by intestinal epithelial cells and is therefore used as a primary energy source for intestinal epithelial cells to maintain colonic mucosal health. [7]Particularly, butyrate enhances the integrity of the intestinal epithelial barrier by inducing tight junction assembly of zonula occludens 1 (ZO-1) and occludin via activation of adenosine monophosphate (AMP)-activated protein kinase (AMPK)-dependent pathway in intestinal epithelial cells. [8]ditionally, butyrate can potentially prevent colon cancer or improve cancer outcomes as a histone deacetylase inhibitor. [9]Butyrate also plays a role in maintaining immune homeostasis. [10]Based on these advantages, the discovery of high-butyrate-producing bacteria has the potential to develop next-generation probiotics to improve human gut health. [11]cent studies have shown that the colonization of butyrateproducing bacteria in the gut improves host health and reduces pathological risks, such as IBD. [12]Particularly, Eubacterium rectale, Faecalibacterium prausnitzii, Roseburia hominis, and Roseburia intestinalis have been reported as dominant butyrate-producing bacteria in the human gut, and are effective in mitigating certain diseases. [13]r example, the abundance of F. prausnitzii in the gut is negatively correlated with the development of IBD or colon cancer. [14,15]This bacterium regulates immune homeostasis by inhibiting the production of inflammatory cytokines, such as interleukin (IL)-6 and IL-12. [16]reover, supplementation of R. intestinalis enhances the production of antimicrobial peptides, gut barrier function, and upregulation of Toll-like receptor signaling to prevent colitis. [17,18]A recent study also proposed that the composition of R. intestinalis in the gut revealed an inverse correlation with the risk of alcoholic fatty liver disease. [19]e presence of R. intestinalis improves gut barrier function in mice and increases the expression of IL-22 and REG3G in the murine gut. [20]However, despite the health-promoting effects of butyrateproducing bacteria, exploration of their metabolic capacities and comparison of their effects on intestinal epithelial cells has remained limited.
Herein, we compared the butyrate production and digestibility of prebiotic carbohydrates by major human butyrate-producing bacteria (E.rectale, F. prausnitzii, R. hominis, and R. intestinalis) to discover potential probiotic candidates.We found that R. intestinalis can produce a high amount of butyrate digested from various prebiotic carbohydrates and that the bacterial extract improves the expression of tight junction protein mRNA in intestinal epithelial cells.Moreover, we evaluated the host-microbe interaction of live R. intestinalis with human intestinal epithelial cells using multiomics approaches.In this regard, we utilized the mimetic intestinal host-microbe interaction coculture system (MIMICS), which we recently developed to investigate the biological effects of live anaerobic bacteria in human cells. [21]Through this, we found that live R. intestinalis enhances the oxidative pathway to increase adenosine triphosphate (ATP) levels in intestinal epithelial cells, whereas dead bacteria did not play a role.These findings support the use of live R. intestinalis as a butyrate-producing probiotic bacterium in the future, which has high potential for improving human gut health.

Butyrate measurement using liquid chromatography-mass spectrometry (LC-MS)
The level of butyrate was determined using chemical derivatization, as previously described. [22,23]After culturing bacterial cells for 48 h, culture supernatants were obtained after centrifugation, and 50 μL of the supernatants was added to 450 μL distilled water.

Bacterial cell count quantification using quantitative polymerase chain reaction (qPCR)
To measure bacterial cell counts in mixed cultures using qPCR, we used primer sets that were designed to target specific 16S rDNA regions for four species of butyrate-producing bacteria following a previous study. [24]We extracted genomic DNA (gDNA) from each bacterial cell culture at the log phase and quantified their colony forming unit (CFU) values using the same bacterial cell culture.The extracted gDNA was serially diluted, and we used these samples as standards for bacterial cell counting in each qPCR run.After incubation of the mixed bacterial cultures, whole bacterial cells were extracted using a bacterial DNA isolation kit (QIAGEN, Hilden, Germany).Bacterial DNA samples were analyzed by qPCR, and the assay was performed using a CFX Connect (Bio-Rad, CA) with SsoAdvanced Universal SYBR Green Supermix (Bio-Rad, CA).The thermocycling conditions in this assay were denaturation at 95 • C for 10 min, followed by 40 cycles of denaturation at 95 • C for 20 s, annealing at 60 • C for 20 s, and extension at 72 • C for 20 s.The results were processed using the CFX Maestro software (Bio-Rad, CA).

Caco-2 cell culture in multiwell plates and treatment of bacterial cell culture extracts
The Caco-2 cell line was obtained from the Korean Cell Line Bank (KCLB no.30037.1,Seoul, Korea).Caco-2 cells were maintained in Dulbecco's modified Eagle medium (DMEM, pH 7.4, Gibco, NY) supplemented with 10% fetal bovine serum (FBS, Biowest, MA) and 1% penicillin/streptomycin (Biowest, MA).Caco-2 cells were cultured for 21 days in 24-well plates at 37 • C in a humidified incubator with 5% CO 2 before treatment with bacterial cell culture extracts.To extract the organic compounds of bacterial cells with cell culture media, we used the single bacterial cell cultures of four species of butyrateproducing bacteria.First, we collected each bacterial culture at the same bacterial cell counts in the exponential phase.The bacterial cultures were mixed with ethyl acetate in a 1:1 ratio.The mixtures were rigorously vortexed for 10 min and centrifuged at 3000 × g for 10 min at 4 • C.After collecting the organic phase, the bacterial medium extract was concentrated using a centrifugal vacuum concentrator (Vision Scientific, Daejeon, South Korea) for 5 h.The concentrated component was dissolved in DMEM and filtered with a syringe filter (polyvinylidene fluoride, pore size 0.45 μm, Millipore, MA).Extracts of 500,000 CFU of each bacterium were added to each culture well and cultured for 24 h before the subsequent assays.

Gene expression analysis using quantitative real-time PCR (qRT-PCR)
Total RNA was extracted from Caco-2 cells using an RNeasy Mini Kit (QIAGEN, Hilden, Germany) following the manufacturer's instructions.
Concentration and purity of the extracted RNA were measured using an ultraviolet spectrophotometer.Complementary DNA (cDNA) was synthesized from 1 μg of extracted RNA using an iScript cDNA Synthesis Kit (Bio-Rad, CA).Validated sequences of the primers for TJP1, OCLN, CLDN3, and GAPDH (glyceraldehyde 3-phosphate dehydrogenase) were designed using PrimerBank [25] ; GAPDH was used as an internal control for qRT-PCR.qRT-PCR was performed using a CFX Connect with SsoAdvanced Universal SYBR Green Supermix as in bacterial cell counting.The thermocycling conditions in this assay were denaturation at 95 • C for 10 min, followed by 40 cycles of denaturation at 95 • C for 20 s, annealing at 60 • C for 20 s, and extension at 72 • C for 20 s.The results were processed using the CFX Maestro software.

Cell viability assays
Culture media were aspirated after treatment with organic extracts.Each culture well was washed three times with sterilized phosphate-buffered saline (PBS) solution.

Coculture of R. intestinalis with Caco-2 cells in the MIMICS
R. intestinalis was cocultured with Caco-2 cells in the MIMICS as previously described with modifications. [21]Briefly, 2 × 10 6 Caco-2 cells were transferred to the MIMICS before coculturing with R. intestinalis and incubated for 21 days at 37 • C 5% CO 2 .Then, the MIMICS was washed two times with PBS and filled with DMEM supplemented with 10% FBS, but without antibiotics.To prepare the cultures of R. intestinalis for coculturing in the MIMICS, R. intestinalis was cultured in YCFA medium supplemented with 0.5 g L −1 glucose until exponential growth (optical density (OD) = 1).Then, the bacterial cells were washed twice with PBS and resuspended with DMEM supplemented with 10% FBS.The MIMICS culture media was replaced with 3 mL DMEM supplemented with 10% FBS with or without R. intestinalis, which was preconditioned under anaerobic conditions.After coculturing for 24 h, the cell culture membrane supports were removed from the MIMICS and washed three times with PBS before the following steps for preparation.Butyrate was measured in DMEM supplemented with 10% FBS and in culture media from coculturing compartments with or without R. intestinalis.

CFU assay for R. intestinalis
To measure changes in the viability of R. intestinalis after coculturing in the MIMICS, serial dilutions of R. intestinalis (before coculture and after coculture for 24 h under anaerobic conditions) were plated onto YCFA agar plates supplemented with 0.5 g L −1 glucose in an anaerobic chamber.Culture plates of R. intestinalis were cultured for 3 days, and the colonies on the plates were counted.

Inactivation of R. intestinalis
R. intestinalis was cultured in YCFA medium supplemented with 0.5 g L −1 glucose until OD = 1.Collected bacterial cells were washed twice with PBS.To inactivate live cells, the bacterial cells were heated at 80 • C for 30 min.After heat-inactivation, the bacterial cells were resuspended with DMEM supplemented with 10% FBS.Culture medium in the coculturing compartments was replaced with 3 mL of media containing heat-killed R. intestinalis.Butyrate was then quantified in DMEM supplemented with 10% FBS and in the culture media from coculturing compartments with heat-killed or live R. intestinalis.

Proteomic analysis using LC-MS/MS
Cellular proteome was prepared as described previously. [26]Briefly, Caco-2 cells on the cell culture membrane supports were lysed using Proteomic analysis was performed using a Q Exactive mass spectrometer (Thermo Fisher Scientific, MA) equipped with a nano-electrospray ionization source coupled to an Ultimate 3000 RSLC nano-LC system (Thermo Fisher Scientific, MA) as described previously. [27]Peptide samples were injected and trapped in an Acclaim PepMap 100 trap column (100 mm × 2 cm, nanoViper C18, Protein identification and quantification were performed using the MaxQuant software with the human UniProt database. [28]Up to two missed cleavages were allowed for peptide identification.
Carbamidomethylation of cysteines was set as a static modification, whereas oxidation of methionine, N-terminal acetylation, and Nterminal methionine excision were set as variable modifications.Mass tolerance was set at 20 ppm.Quantified protein data were processed with the Perseus software. [29]Gene ontology enrichment analysis was performed using the PANTHER classification system. [30]Multiomics analysis was performed using the MetScape 2 bioinformatics tool. [31]

Targeted metabolomic analysis using LC-MS/MS
To extract cellular metabolites, Caco-2 cells on the cell culture membrane supports were lysed with 5 mL ice-cold 80% v/v methanol and incubated for 15 min at −70 • C.After centrifugation, the supernatant was dried using a centrifugal vacuum concentrator.Dried pellets were dissolved in 100 μL 50% v/v methanol.Then, the cellular metabo- The capillary temperature was 300 • C, and the electrospray ionization spray voltage was 4000 V. Peak areas of the metabolites were collected using the Agilent MassHunter Qualitative Analysis software [32] .

Comparison of butyrate production and digestibility of prebiotic polysaccharides in human butyrate-producing bacteria
To investigate the carbohydrate preferences for human butyrateproducing bacteria, we measured the growth of four species of butyrate-producing bacteria (E.rectale, F. prausnitzii, R. hominis, and R. intestinalis) after 48-h culture in YCFA media supplemented with monosaccharides (arabinose, fructose, galactose, glucose, and xylose) or prebiotic polysaccharides (amylopectin, arabinoxylan, galactan, inulin, pullulan, and xylan) (Figure 1A).In addition, we analyzed the butyrate concentration in each culture medium containing either glucose or prebiotic polysaccharides (Figure 1B).R. intestinalis exhibited the widest range of polysaccharide digestion by growing in media containing amylopectin, arabinoxylan, and pullulan and could also ferment all five monosaccharides.R. intestinalis produced the highest butyrate level in culture media with prebiotic carbohydrates.A previous study found that R. intestinalis encodes more glycoside hydrolase (GH) enzymes, such as GH2, GH3, GH13, and GH43, than other human butyrate-producing bacteria. [33]The major activities reported for these GHs are α-amylases, α-L-arabinofuranosidases, and β-Dxylosidases, which are needed to digest the various dietary fibers that humans ingest. [34]Particularly, we determined that R. intestinalis grew and fermented high levels of butyrate even in media containing arabinoxylan that comprises a backbone of β-(1,4)-linked xylose residues, which are substituted with arabinose residues on the C(O)-2 and/or C(O)-3 position. [35]The complex structures of dietary fibers such as arabinoxylan have received more focus as prebiotics [36] as their complex structures allow these carbohydrates to progress further into the large intestine compared with other common dietary fibers. [37]erefore, we found that R. intestinalis has high metabolic capacity for butyrate production and can even ferment complex prebiotic polysaccharide structures.Furthermore, we investigated the competitiveness of bacterial growth in mixed cultures of butyrate-producing bacteria.To do this, we inoculated 10 7 CFU mL −1 of each of the four species of butyrateproducing bacteria into the same culture flask with media containing prebiotic polysaccharides and measured the butyrate production after 48 h of culture (Figure 1C).The butyrate levels were higher in the media containing amylopectin, arabinoxylan, and pullulan compared with that containing inulin, which R. intestinalis cannot ferment well.
Using qPCR for bacterial cell counting, we determined that R. intestinalis was the dominant bacterium in media containing amylopectin, arabinoxylan, and pullulan but not for inulin (Figure 1D).Competition of probiotic bacterium with other bacteria is an important factor to dominate nutrient use in complex microbial ecosystems. [38]As probiotic bacteria are often weakly competitive with other bacteria, they can be easily washed out and have problems colonizing the gut. [39]Accordingly, our results suggest that R. intestinalis can become the dominant butyrate-producing bacterium in the gut via nutrient competition with other butyrate-producing bacteria.

Comparison of the improvement of expression of tight junction barrier protein mRNA by butyrate-producing bacteria
To explore which butyrate-producing bacterium was more effective for improving the integrity of the intestinal epithelial barrier, we compared the gene expression levels of the tight junction proteins (ZO-1, occludin, claudin 3) of Caco-2 cells after treatment with the extracts from both cells and culture media of butyrate-producing bacteria.
First, we extracted organic components from E. rectale, F. prausnitzii, R. hominis, R. intestinalis, and basal culture media (YCFA medium).Upon measuring the butyrate level in the extracts, the medium extract from R. intestinalis culture showed the highest butyrate level (Figure S1).
The cell viability measurements showed that not all bacterial extracts affected Caco-2 cell viability as well as butyrate treatment did (Figure S2).Interestingly, the whole cell culture extracts of R. intestinalis highly increased the gene expression of ZO-1, occludin, and claudin 3 in Caco-2 cells compared with the extracts of other butyrate-producing bacteria (Figure 2A-C).ZO-1, occludin, and claudin 3 are pivotal proteins for maintaining tight junction assembly in intestinal epithelial cells. [40]Butyrate plays an important role in enhancing the integrity of the intestinal epithelial barrier via the upregulation of tight junction proteins. [8,41,42]Hence, butyrate-producing bacteria help defend against translocation of pathogenic toxins, such as lipopolysaccharides, into systemic circulation and improve inflammatory diseases via the production of butyrate. [43]Based on these results, we finally selected R. intestinalis as a probiotic bacterium with high potential for improving host gut health.

Coculture of live R. intestinalis with human intestinal epithelial cells using the MIMICS
In our previous study, we developed the MIMICS to investigate hostmicrobe interaction by coculturing live anaerobic bacteria with human intestinal epithelial cells. [21]Using this state-of-the-art device, we evaluated the effects of live R. intestinalis on human intestinal epithelial cells (Figure 3A).First, we cultured R. intestinalis with Caco-2 cells for 24 h and evaluated whether the bacterium affected host cell viability (Figure 3B).Additionally, we confirmed that R. intestinalis maintained viability for 24 h and produced butyrate, meaning that this anaerobic bacterium was metabolizing normally in the MIMICS (Figure 3C,D).
To investigate host-microbe interaction accurately, it is required to maintain the metabolism of anaerobic bacteria as well as their viability during coculturing with host cells. [21]We observed that live R. intestinalis does not change the viability of host cells, which revealed that the bacterium is unlikely to damage intestinal cells.In vivo evidence that R. intestinalis does not damage the host has been previously provided. [44]Therefore, we explored the biological changes in intestinal epithelial cells treated with live R. intestinalis using the In vivo model for host-microbe interaction.

Multiomics analysis to uncover host-R. intestinalis interaction
We performed metabolomic and proteomic analyses of Caco-2 cells after coculturing with live R. intestinalis to investigate the biological effects on intestinal epithelial cells.Consequently, we measured 73 metabolite changes; the levels of 43 metabolites considerably increased with treatment of live R. intestinalis, while that of nine metabolites decreased (Figure 4A).Butyrate is easily absorbed into intestinal epithelial cells and used as a primary source for the tricarboxylic acid cycle. [7]Our results show that treatment with live R. intestinalis increased the level of tricarboxylic acid cycle intermediates in intestinal epithelial cells (Table S1).Notably, metabolites related to nucleic acid metabolism, such as ATP, AMP, and guanosine diphosphate, were highly overproduced in comparison with other metabolites in the R. intestinalis coculture (Table S1).Consequently, we found that live R. intestinalis changed the primary metabolism in intestinal epithelial cells.Proteomic analysis showed that 2370 proteins were detected, of which 578 proteins were upregulated in expression in the coculture with live R. intestinalis, while 106 proteins were downregulated in expression (Figure 4B, Table S2).To check whether our proteomic analysis was consistent with previous In vivo results, we first observed the proteomic changes in tight junction assembly.Our results show that expression was upregulated in 14 of the 17 proteins involved in tight junctions (Figure 4C, Table S3).Particularly, expressions of the genes for ZO-1 (TJP1), claudin 3 (CLDN3), and 5-AMP-activated protein kinase catalytic subunit (PRKAA1) were upregulated by 2.14-, 1.74-, and 1.50-fold, respectively.Previous studies have shown that oral administration of R. intestinalis increased levels of protein and mRNA of tight junction proteins such as ZO-1 in a mouse model, which was consistent with our proteomic results. [19,45]Furthermore, we found that expression of PRKAA1 increased following treatment with live R. intestinalis, which is expected as butyrate is known to enhance tight junction assembly via the activation of AMPK in intesti-nal epithelial cells. [8]Interestingly, the coculture with R. intestinalis reduced the levels of Ras-related protein Rab-8A (RAB8A).As RAB8A inhibits the recycling of claudin and occludin, our observation also suggest another mechanism by which R. intestinalis can improve the integrity of the intestinal epithelial barrier. [46]Based on these results, we believe that the proteomic analysis is consistent with previous In vivo studies.

3.5
Activation of nucleic acid metabolism in the coculture with R. intestinalis Interestingly, our metabolomic data showed that 20 of 27 metabolites involved in nucleic acid metabolism were considerably overproduced in the coculture with R. intestinalis, although only three metabolites, including xanthine, experienced drastically reduced production (Figure 5A).Moreover, we found that the expression of 16 proteins was F I G U R E 5 (A) Changes in nucleic acid metabolism of Caco-2 cells.(B) Multiomics view of purine metabolism.Upregulation refers to the overexpression of proteins in the coculture with R. intestinalis compared with that of the control, while downregulation refers to the opposite.considerably upregulated among the 59 proteins involved in nucleic acid metabolism, whereas there was no considerable downregulation of the expression of any of these proteins (Table S4).Multiomics analysis based on MetScape showed that purine metabolism was highly enriched and that the levels of ATP and associated reactions, in particular, were highly upregulated in the coculture with R. intestinalis (Figure 5B).Nucleic acid metabolism plays a pivotal role in the metabolism of DNA and RNA, storage of chemical energy, and formation of carriers of activated metabolites for biosynthesis in intestinal cells. [47]ATP is the source of energy for use and storage at the cellu-lar level, and is the final product of the oxidative pathway and nucleic acid metabolism. [48]Therefore, improvement of purine metabolism is essential for the accumulation of cellular energy. [49]In limited cases, such as uric acid accumulation, purine metabolism can be pathological to developing intestinal diseases. [50]However, intracellular production of xanthine, which is a precursor of uric acid, was significantly reduced in the coculture with R. intestinalis; thus, live R. intestinalis may only enhance the synthetic pathway of purine metabolism.Consequently, we found that R. intestinalis increases purine metabolism to accumulate cellular energy.

Activation of the oxidative pathway of carbohydrate and fatty acid metabolism
We focused on the increase in metabolite levels involved in the oxidative pathway of carboxylic acid metabolism following treatment with live R. intestinalis.Levels of α-ketoglutarate, fumarate, hexose phosphate, phosphoglycerate, lactate, and glucose increased in the coculture with R. intestinalis, and more interestingly, there was a 55.2fold overproduction of ATP in Caco-2 cells compared with that in the control (Figure 6A,B).The proteomic results were consistent with those for the improvement of the oxidative pathway of carboxylic acid metabolism (Table S5).Gene ontology enrichment analysis using the PANTHER database showed that glycolysis, TCA metabolism, fatty acid oxidation, and respiratory electron transport chain were highly enriched with treatment of R. intestinalis (Figure 6C).All 130 proteins involved in these pathways were identified, and the expression of 32 proteins was considerably upregulated in the coculture with R. intestinalis, while that of only five proteins was downregulated (Figure 6D, Table S5).Oxidative metabolism of carboxylic acids and fatty acids in intestinal cells plays a pivotal role in regulating intestinal epithelial cell homeostasis. [51]Decrease in oxidative metabolism impairs intestinal epithelial cells via lipotoxicity and can even cause intestinal diseases through gut dysbiosis. [52]Moreover, the increase in ATP level via the activation of the oxidative pathway induces actin polymerization in intestinal epithelial cells, followed by the enhancement of intestinal epithelial barrier integrity. [53]Therefore, we found that live R. intestinalis can improve gut health via activation of the oxidative metabolism of carboxylic acids and fatty acids at the molecular level.Based on previous studies, we hypothesized that the improvement of oxidative metabolism in intestinal epithelial cells was most probably caused by the metabolic products of live R. intestinalis, particularly butyrate. [7]Butyrate acts as cellular fuel in enterocytes and would therefore be able to improve oxidative metabolism.Thus, we added heat-killed R. intestinalis into the MIMICS for 24 h and analyzed metabolic changes in the Caco-2 cells.First, treatment with heat-killed R. intestinalis did not produce butyrate in the culture medium (Figure 7A).In addition, the intracellular metabolomics of Caco-2 cells showed that heat-killed R. intestinalis treatment reduced ATP levels and did not increase production of metabolites involved in oxidative metabolism (Figure 7B,C).The administration of dead probiotic cells has been used as a biological response modifier, that is, via immunomodulation, using probiotic strains. [54]For instance, heat-killed strains of Lactobacillus, such as L. acidophilus, were able to modulate immune responses by stimulating the proliferation of murine splenocytes as this does not depend on the cells being alive. [55]We found that only live R. intestinalis activated the oxidative pathway of intestinal epithelial cells, whereas there were no effects after treatment with heat-killed R. intestinalis.Thus, our results suggest that it is necessary to administrate live cells for the health-promoting effects of R. intestinalis.

CONCLUSION
Despite spotlights on the health-promoting effects of human gut butyrate-producing bacteria, their metabolic capabilities for butyrate production, carbohydrate digestion, and biological influence on human epithelial cells are not well understood.Herein, we discovered that R. intestinalis produces higher amounts of butyrate and digests more complex prebiotic polysaccharide structures than other human butyrate-producing bacteria (Figure 8).In addition, cellular extracts of R. intestinalis increased the gene expression of intestinal epithelial 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) was dissolved in DMEM, and 500 μL MTT solution was added to each culture well and incubated at 37 • C for 4 h.After removing the MTT solution, the purple formazan crystals were dissolved in acid-isopropanol (0.04 N HCl in isopropanol).The cell viability was measured spectrophotometrically at 550 nm.

5
mL of radioimmunoprecipitation assay buffer (Thermo Fisher Scientific, MA).Lysed Caco-2 cells were then sonicated with a probe sonicator on ice.The lysed cell solution was centrifuged at 4000 × g for 10 min at 4 • C, and the supernatant was removed into a new tube.Trichloroacetic acid (1 mL) was added, and the mixture was incubated for an hour at 4 • C to precipitate proteins.After centrifugation, the protein pellet was dissolved in 8 M urea after washing with ice-cold acetone.The concentration of the extracted protein was quantified using the bicinchoninic acid assay (Pierce BCA Protein Assay Kit, Thermo Fisher Scientific, MA).Proteins (100 μg) were reduced with dithiothreitol (dithiothreitol:protein = 1:50, w/w) and alkylated using iodoacetamide (iodoacetamide:protein = 1:10, w/w).After adding 3.5 mL of 25 mM ammonium bicarbonate buffer (pH 8), the proteins were digested by proteomics-grade trypsin for 18 h at 37 • C. Digested peptides were then purified with a C18 cartridge (Waters, MA) and fully dried using a centrifugal vacuum concentrator (Vision Scientific, Daejeon, South Korea).Dried peptides were dissolved in a 0.1% v/v formic acid solution.

5
mm, 100 Å; Thermo Fisher Scientific, MA).An Acclaim PepMap 100 capillary column (75 mm × 15 cm, nanoViper C18, 3 mm, 100 Å; Thermo Fisher Scientific, MA) was used for LC separation.The flow rate was 350 nL min −1 ; solvent A comprised water containing 0.1% v/v formic acid, while solvent B was ACN containing 0.1% v/v formic acid.The LC gradient program was set as follows: t = 0 min, 2% B; 30 min, 35% B; 40 min, 90% B; 45 min, 90% B; 60 min, 5% B. The mass spectrometer was operated in the positive ion mode; the ion spray voltage was 2100 eV.MS data were obtained using the Xcaliber software (Thermo Fisher Scientific, MA).An MS analyzer scanned precursor ions with a mass range of m/z 350-1800 and resolution of 70,000.The 15 most abundant precursor ions were selected for collision-induced dissociation.The normalized collision energy was 32.
lites were analyzed via multiple reaction monitoring methods equipped with an Agilent 6420 Triple Quadrupole LC/MS (Agilent Technologies, CA, United States) coupled with an Agilent 1260 Infinity Binary LC (Agilent Technologies, CA).Dissolved solution (10 μL) was injected into the LC-MS/MS system, and the metabolites were separated in a Luna NH 2 column (250 × 2 mm, 5 mm particle size, Phenomenex, CA).The flow rate was 0.5 mL min −1 ; solvent A comprised 5% v/v ACN solution with 20 mM ammonium acetate and 20 mM ammonium hydroxide (pH 9.45), while solvent B was 100% ACN.The LC gradient method was set as follows: t = 0 min, 85% B; 15 min, 0% B; 30 min, 85% B; 40 min, 85% B.
measurements of four species of human butyrate-producing bacteria in media containing various monosaccharides and prebiotic polysaccharides after 48-h incubation (n = 3).Normalized growth means the relative average value of the maximum OD reached by each bacterium.(B) The butyrate production measured after culturing each bacterium in YCFA media supplemented with glucose and polysaccharides for 48 h (n = 3).(C) Comparison in the butyrate concentrations in the media after 48-hour mixed cultures of four species of butyrate-producing bacteria in media supplemented with glucose and polysaccharides (n = 3).(D) Comparison in the bacterial abundance for the media containing glucose or polysaccharides after a 48-h mixed culture of four species of butyrate-producing bacteria (n = 6).All bar graphs represent the mean values, and error bars indicate the standard deviation.(*p < 0.05, ****p < 0.001; p values were derived from two-way ANOVA).

F I G U R E 2
Gene expression changes of tight junction proteins after treatment with butyrate or the whole cell culture extracts of butyrate-producing bacteria: (A) ZO-1, (B) occludin, (C) claudin 3 (n = 3).All bar graphs represent the mean values, and error bars indicate the standard deviation (*p < 0.05, ***p < 0.001; p values were derived from one-way ANOVA).F I G U R E 3 (A) Schematic diagram for the investigation of host-R.intestinalis interaction using multiomics approaches.(B) Changes in the viability of Caco-2 cells after coculturing with R. intestinalis for 24 h (n = 3).(C) Changes in the bacterial cell count after coculturing with R. intestinalis for 24 h (n = 3).(D) Comparison of butyrate levels in the culture media of Caco-2 cells during 24-h monoculture or co-culture with R. intestinalis (n = 3).All bar graphs represent the mean values, and error bars indicate the standard deviation.(****p < 0.0001; p values were derived from Student's t-test and one-way ANOVA).

F
I G U R E 4 (A) Heatmap visualization of all metabolite changes in Caco-2 cells (n = 4).(B) Volcano plot of the proteomic results.Fold-change refers to the protein expression in the coculture with R. intestinalis divided by the control (n = 3).(C) Changes in protein expression involved in tight junction assemblies.

F
I G U R E 6 (A) Changes in carboxylic acid metabolism of Caco-2 cells (n = 4).(B) Comparison of ATP levels (n = 4).(C) Gene ontology enrichment analysis in the proteomic results.(D) Proteomic changes in oxidative pathways.Fold-change refers to the protein expression of treatment with R. intestinalis divided by the control.All bar graphs represent the mean values, and error bars indicate the standard deviation.(*p < 0.05, **p < 0.01, ***p < 0.001; p values were derived from Student's t-test).

F
I G U R E 7 (A) Comparison of butyrate concentration in culture media after co-culturing with heat-killed or viable R. intestinalis (n = 3).(B) Changes in ATP levels after treatment with heat-killed R. intestinalis (n = 3).(C) Changes in the metabolites involved in carboxylic acid metabolism after treatment with heat-killed R. intestinalis (n = 3).All bar graphs represent the mean values, and error bars indicate the standard deviation.(*p < 0.05; p values were derived from Student's t-test.).

F I G U R E 8
Summary of the discovery of high-butyrate-producing bacterium and the associated health-promoting effects on intestinal epithelial cells.barrier proteins.To investigate the comprehensive biological effects of R. intestinalis in the human gut, we cocultured the bacterium with human intestinal epithelial cells in the MIMICS, which is an In vivo coculture system to explore host-anaerobe interaction.Consequently, we demonstrated that live R. intestinalis enhanced the level of cellular ATP in human intestinal epithelial cells by increasing oxidative pathway and purine metabolism.However, heat-killed R. intestinalis did not induce an increase in cellular ATP.Therefore, this study suggests that the use of live R. intestinalis as a probiotic bacterium is expected to improve gut health and diseases in the future.