Abstract
The large intestine has a dense milieu of indigenous bacteria, generating a complex ecosystem with crosstalk between individual bacteria and host cells. In vitro host cell modeling and bacterial interactions at the anaerobic interphase have elucidated the crosstalk molecular basis. Although classical cell lines derived from patients with colorectal cancer including Caco-2 are used, whether they adequately mimic normal colonic epithelial physiology is unclear. To address this, we performed transcriptome profiling of Caco-2 and Monolayer-cultured epithelial cells derived from healthy Human Colonic Organoids (MHCO) cultured hemi-anaerobically. Coculture with the anaerobic gut bacteria, Bifidobacterium longum subsp. longum differentiated the probiotic effects of test cells from those of physiologically normal intestinal and colorectal cancer cells. We cataloged non- or overlapping gene signatures where gene profiles of Caco-2 represented absorptive cells in the small intestinal epithelium, and MHCO showed complete colonic epithelium signature, including stem/progenitor, goblet, and enteroendocrine cells colonocytes. Characteristic gene expression changes related to lipid metabolism, inflammation, and cell-cell adhesion were observed in cocultured live Bifidobacterium longum and Caco-2 or MHCO. B. longum-stimulated MHCO exhibited barrier-enhancing characteristics, as demonstrated in clinical trials. Our data represent a valuable resource for understanding gut microbe and host cell communication.
Similar content being viewed by others
Introduction
The large intestine, also known as the colon, is a crucial component of the human digestive system with a wide range of physiological functions vital for maintaining overall health1. One of the primary physiological functions of the colon is the absorption of water and electrolytes from residual undigested food materials entering the colon from the small intestine2. The content of the partially digested chyme that passes through the colon is concentrated through water absorption, an essential process that maintains body fluid balance and prevents dehydration.
Additionally, the colon hosts a diverse community of beneficial bacteria, known as the gut microbiota, which contribute significantly to human health3. These microorganisms aid in fermenting undigested carbohydrates and breaking down certain substances to produce metabolites, gasses, and vitamins4. The gut microbiota plays a crucial role in bolstering the immune system, protecting against harmful pathogens and maintaining the intestinal lining integrity5. The role of microorganisms in maintaining homeostasis and influencing various aspects of human physiology has spurred research into probiotics development.
Probiotics are live microorganisms that confer health benefits when consumed in adequate quantities, and they have emerged as a promising strategy for improving human health and well-being6. To promote research and development of probiotics, the physiological functions of microorganisms and their effects on the host need to be investigated7. However, details of the fundamental molecular mechanisms underlying host cell-bacteria interactions remain largely unknown. Recent improvements in next-generation sequencing and mass spectrometry technologies have increased interest in the trans-omics analysis of gut microbiota. However, the results only provide a correlation between changes in microorganisms and their traits and do not allow direct causal inference. Therefore, in vitro models are required to elucidate the detailed mechanisms of microbe-host cell interactions. However, existing models do not assess the response of normal human intestinal epithelial cells because of the difference in oxygen demands between gut bacteria (most bacteria living in the colon are anaerobes) and host cells (propagate under normoxia).
To overcome this shortcoming, researchers including our group, have developed a state-of-the-art coculture system of human colonic epithelium with highly oxygen-sensitive gut bacteria8,9,10,11,12. Colonic epithelial cells were cultured using a two-dimensional (2D) design on the Boyden chamber to establish the coculture system. Immortalized cell lines derived from human colorectal cancer tissue, such as Caco-2 cells, are commonly used to analyze gut bacteria and their metabolites13. Although the molecular basis is still unclear, the morphological and biochemical properties of Caco-2 are known to spontaneously transform, converting them into a small intestine-like epithelium following prolonged culture under confluent 2D layer conditions14. Consequently, Caco-2 monolayers have been extensively used as in vitro models of the small intestine to characterize certain aspects of nutrient uptake15.
Most probiotic bacteria, such as Akkermansia and Bifidobacterium species, reside and function in the colon rather than the small intestine16. Thus, colonic epithelium must be used in analyzing the role of microbes inhabiting the colon. Recently, normal colonic epithelial cells were stably cultured for extended periods using three-dimensional (3D) mouse and human organoid culture methods17,18. Self-renewing 2D Monolayer-cultured epithelial cells derived from healthy Human Colonic Organoids (MHCO) are also a physiologically relevant model10,19. Moreover, primary cultured intestinal epithelial cells derived from organoids are increasingly used to study intestinal microbes.
In this study, we investigated the feasibility of using organoid-based normal human epithelium instead of Caco-2 in an intestinal hemi-anaerobic coculture system (iHACS)10 containing the probiotic bacteria, Bifidobacterium longum. This system was established to address the limitation of currently available in vitro coculture systems of representative human and microbial cells. We conducted a detailed molecular analysis of the differential effects on cell physiology induced by coculture experiments with Caco-2 and normal colonic epithelium. By stimulating a colorectal cancer-derived cell line or normal colonic epithelium with bacteria, we aimed to generate an expression profile for the stratification of probiotic responses.
Results
Characterization of Caco-2 and colonic organoid-derived monolayer
We performed RNA-sequencing (RNA-seq) analysis to decipher the complete genetic information of the intestinal epithelium cultured using iHACS and compared two in vitro cell culture models of Caco-2 and 2D cultured normal colonic epithelium derived from MHCO. To collect RNA samples for transcriptome analysis, we first aligned the culture conditions of the Caco-2 and MHCO (Fig. 1A). Caco-2 was found to be able to grow in the organoid-conditioned medium, which was devoid of serum completely, so after 17 days of culture under sufficiently confluent conditions in fetal bovine serum (FBS)-based DMEM Cell Culture Medium20,21. The medium was changed to the organoid-conditioned medium, and RNA was recovered from cells after a further 3 days culture (Fig. 1A). Contrastingly, MHCO were cultured for 3 days after reaching confluence under similar conditions to those used for the Caco-2 (Fig. 1A). Importantly, we comparatively analyzed the gene expression in MHCO and Caco-2 following growth under hemi-anaerobic conditions.
Comparison of gene expression profiles between Monolayer cells derived from healthy Human Colonic Organoid (MHCO) and Caco-2. (A) A schematic representation of the experiment. (B) PCA of the transcriptomes of MHCO and Caco-2. Transcriptomes from three replicates were plotted. Ellipses indicate the 95% confidence interval. Filled and unfilled circles indicate two independent experiments (each three biological replicates). (C) Volcano plot depicting the DEGs of the transcriptomes of MHCO and Caco-2. The x- and the y-axis show the log2 transformed fold change and the -log10 transformed adjusted p-value. Red and blue dots indicate significantly upregulated genes in MHCO and Caco-2 according to adjusted p-value < 0.01 and log2 fold change > 1.5. (D) Heatmap depicting the top-1000 DEGs of the transcriptomes of MHCO and Caco-2 when sorted by adjusted p-value. (E) TPM values of marker genes for each gut epithelial cell type. Data are presented as the mean ± Standard derivatives (SD). Statistical significance was evaluated by Deseq2.
To ensure the reliability of our transcriptomic analysis, we conducted two independent RNA-seq experiments (shown in filled and open circles dots in Figures, respectively), each using three biological replicates (n = 3) of separately prepared organoid-derived monolayers and Caco-2. Principal component analysis (PCA) of RNA-seq data from Caco-2 and MHCO cultured using iHACS revealed substantial differences and a clear separation of transcriptional patterns between both samples (Fig. 1B). The transcriptomes of Caco-2 and MHCO differed significantly, and among genes with an expression difference of > 1.5 fold change (adjusted P-value [Padj] < 0.01), we identified 5,766 and 5,141 genes that were enriched explicitly in Caco-2 and MHCO, respectively (Fig. 1C). Furthermore, we visualized two non-overlapping classes of cell type-specific transcripts using hierarchical clustering, Caco-2 specific and MHCO signatures that were expressed in a mutually exclusive manner (Fig. 1D). Among the top 1,000 differentially expressed genes (DEGs), we observed several marker genes of functionally differentiated cells constituting the colonic epithelium: Keratin 19 (KRT19, colonocytes), Mucin 2 (MUC2), Trefoil factor 3 (TFF3, goblet cells), and Chromogranin A/B (CHGA and CHGB, enteroendocrine cells) in the MHCO RNA-seq data (Fig. 1D, E). The stem cell marker gene, Leucine rich repeat containing G protein-coupled receptor 5 (LGR5) was highly expressed in the cancer-derived cell line, whereas other markers of differentiated cells constituting the colonic epithelial tissue were highly expressed in MHCO (Fig. 1E). This finding was further supported by data showing that several known marker genes of the small intestine, such as Alkaline Phosphatase, intestinal (ALPI)22,23, Alanyl Aminopeptidase, membrane (ANPEP)24, and Apolipoprotein B (APOB)25, were more highly expressed in Caco-2 than they were in MHCO (Fig. 1D, E).
To address the concern regarding whether the long-term culture of MHCO might affect its property, the normal colonic epithelium can acquire small intestine-like characteristics observed in Caco-2, we performed RNA-seq analysis on MHCO and Caco-2 after 5 days (short period) or 21 days (long period) culture, respectively. PCA of those RNA-seq data demonstrated that MHCO and Caco-2 exhibited consistently distinct global gene expression patterns, irrespective of whether they were culture for a short or long period (SFig. 1 A). Notably, the effect of culture duration on gene expression was more pronounced in MHCO than in Caco-2. One possible explanation for this observation is the downregulation of cell cycle-related genes observed specifically in MHCO (SFig. 1 A and B). The expression of intestinal stem cell marker LGR5 was markedly reduced in MHCO after prolonged culture (SFig. 1 C). In contrast, the expression levels of differentiated cell markers for colonocyte, goblet cells, and enteroendocrine cells remained consistently high, indicating the preservation of colonic epithelial identity during extended culture (SFig. 1 C). Moreover, genes associated with small intestinal functions, such as enterocyte and a lipid transport-related genes, remained low in MHCO after long-term culture compared to Caco-2 (SFig. 1D). Thus, the overall gene expression profile remained characteristic of small intestine like-epithelium in Caco-2, and of colonic epithelium in MHCO, regardless of the number of days in culture.
Next, the DEGs that differed significantly between Caco-2 and MHCO (Fig. 1D) were examined using gene ontology (GO) enrichment analysis. GO analysis of the genes upregulated in cultured Caco-2 showed enrichment of GO terms associated with the Wnt signaling pathway (GO:0016055) and protein-lipid complex (GO: 0034368, Fig. 2A). These findings are consistent with those previously reported on the identity of Caco-2 derived from a patient with colorectal cancer who had a mutation in the APC gene. APC is a negative regulator of Wnt signaling26 that induces the expression of the target genes, Yes-associated protein 1 (YAP1)27 and Midkine (MDK)28,29. Additionally, GO analysis also revealed a significant association with the transporters of lipids gene, Microsomal triglyceride transfer protein (MTTP)30 because Caco-2 are an extensively used in vitro assay for chylomicron secretion studies15 (Fig. 2B).
GO enrichment analysis of DEGs of the transcriptomes of MHCO and Caco-2. (A) Dot plot showing GO terms associated with DEGs for Caco-2 (left panel) and MHCO (right panel) identified in Fig. 1D. (B, C) TPM values of representative genes for each GO term enriched in Caco-2 (B) or MHCO (C). Data are presented as the mean ± Standard derivatives (SD). Statistical significance was evaluated by Deseq2.
In contrast, genes upregulated in MHCO showed a pronounced enrichment of GO terms related to tissue homeostasis (GO:0001894) involved in goblet cell marker genes (MUC2 and TFF3) and different glycoprotein biosynthetic process (GO: 0009101) involved in the glycosylation of mucin proteins (ST6 N-acetylgalactosaminide alpha-2,6-sialyltransferase 1 [ST6GALNAC1], beta-1,3-galactosyltransferase 5 [B3GALT5], and Fucosyltransferase 2 [FUT2]; Figs. 1E and 2A and C).
Furthermore, GO analysis indicated the enrichment of tissue homeostasis genes (GO:0001894) exclusive to MHCO and involved in the mucosal immune system (Fig. 2A). Immune responses to gut bacteria, Toll-like receptor 4 (TLR4)31, and Aldehyde dehydrogenase 1A1 (ALDH1A1)32, were detected in MHCO (Fig. 2C). Collectively, the presence of these genes and enrichment of relevant GO confirmed that MHCO are more pertinent to observe the physiological response of the colonic epithelium than Caco-2 when those cells were cultured under semi-anaerobic conditions.
Response of Caco-2 and MHCO to live anaerobes in the colon
Caco-2 have been used to analyze interactions between host cells and gut bacteria33 and, therefore, we determined whether the responses of those cells to gut bacteria differed from those of MHCO. To test the physiological effects of commensal anaerobes in the colon, we cocultured Caco-2 or MHCO with a probiotic strain of B. longum using iHACS. Because Bifidobacterium is generally localized in the colon, we used healthy MHCOs in this study34.
We confirmed that B. longum growth increased similarly in Caco-2 and MHCO cultured with iHACS (Fig. 3A, B). After 24 h, an inoculum of 1 × 107 colony forming units (CFU)/mL of B. longum increased to a median of 6.98 × 108 and 8.93 × 108 CFU/mL in coculture with MHCO and Caco-2, respectively (P = 0.26, Fig. 3B). RNA was extracted from Caco-2 and MHCO to assess their general responses to the probiotic bacteria. In the PCA of the characterized compound classes in Caco-2 and MHCO in the presence or absence of B. longum exposure (Fig. 3C), the first two principal components explained 32.1% (PC1) and 9.8% (PC2) of the total variation. PC1 indicated that MHCO were more distant from the clusters of Caco-2, whereas PC2 showed the response of Caco-2 cocultured with B. longum. Samples of MHCO alone and cocultured with B. longum showed a low variation in each data set (Fig. 3C). Compared to the bacteria-free axenic culture, Caco-2 and MHCO incubated with B. longum showed distinct transcriptomic responses, as shown in the volcano plot of replicates exposed to identical conditions (Fig. 3D). After filtering for adjP < 0.01, the results showed that stimulation with B. longum affected the expression of 4,806 genes in Caco-2 by 1.5-fold, consisting of 2,274 and 2,532 upregulated and downregulated genes, respectively (Fig. 3D). Similarly, 2,600 genes were differentially expressed in MHCO, consisting of 1,172 and 1,428 upregulated and downregulated genes, respectively (Fig. 3D). Additionally, the overlap of upregulated DEGs between MHCO and Caco-2 was minimal, with only 204 commonly expressed genes (Fig. 3E). Similarly, 334 DEGs exhibited decreased expression (Fig. 3E).
Differences in gene expression change among MHCO and Caco-2 after coculturing with B. longum. (A) A schematic representation of the experiment. (B) Colony-forming units (CFU) counts per milliliter of B. longum inoculated after 24-hour incubation with MHCO and Caco-2 using iHACS. (C) PCA of the transcriptomes of MHCO and Caco-2 cocultured with or without B. longum. Filled and unfilled circles indicate two independent experiments (each three replicates). Transcriptomes from three replicates were plotted. Ellipses indicate the 95% confidence interval. (D) Volcano plot depicting the upregulated and downregulated genes of Caco-2 (left panel) and MHCO (right panel) after coculturing with B. longum. The x- and the y-axis show the log2 transformed fold change and the -log10 transformed adjusted p-value. Red and blue dots indicate significantly upregulated and downregulated genes after coculturing with B. longum according to adjusted p-value < 0.05 and log2 fold change > 2. (E) Venn diagram comparing the significantly upregulated (left panel) and downregulated genes (right panel) in MHCO or Caco-2 after coculturing with B. longum.
The results of the GO analysis of genes that exhibited variations in expression in each cell are shown in Fig. 4. We identified GO terms derived from Caco-2 or MHCO-specific gene expression variations following stimulation with B. longum. The DEGs that showed increased expression exclusively in MHCO were predominantly enriched in extracellular matrix organization (GO:0030198) and glycerolipid metabolic process (GO:0046486) (Fig. 4A). Genes downregulated in MHCO were associated with cell proliferation pathways such as nuclear division (GO:0000280) and chromosome segregation (GO:0007059, Fig. 4B). In Caco-2, the most enriched and least downregulated GO terms were cytoplasmic translation (GO:0002181) and the amide metabolic process (GO:0043603), respectively (Fig. 4C, D). Conversely, most enriched terms in the commonly upregulated genes were associated with lipopolysaccharide (GO: 0032496), whereas downregulated genes were related to double-strand break repair (GO:0006302) (Fig. 4E, F).
GO enrichment analysis of upregulated and downregulated genes in MHCO and Caco-2 after coculturing with B. longum. (A–F) Dot plot showing GO terms associated with upregulated (A, C, E) and downregulated genes (B, D, F) in MHCO (A, B), Caco-2 (C, D), and both (E, F) identified in Fig. 3E.
B. longum regulates the expression of transcription factors and genes involved in host metabolic pathway.
Our analysis of the differentially induced gene expression in Caco-2 and MHCO following bacterial treatment showed a significant increase in the expression of genes involved in peroxisome proliferator-activated receptor (PPAR) signaling in MHCO (Fig. 5A). Furthermore, our transcriptome data showed that the expression of PPARG and PPARA were upregulated in MHCO, while those tended to decrease in Caco-2 following stimulation with B. longum (Fig. 5A, B). Genes of signaling pathways associated with PPARs, fatty acid transporter/oxidation, adipocyte differentiation, and gluconeogenesis were upregulated in MHCO but not in Caco-2 (Fig. 5A).
B. longum regulates the PPAR signaling pathway. (A) Heatmap of relative expression of PPAR signaling-related genes by co-cultivation of B. longum in Caco-2 and MHCO. Statistical significance was evaluated by a Deseq2 (*adjusted p-value < 0.05, ** adjusted p-value < 0.01, *** adjusted p-value < 0.001). (B) TPM values of representative PPAR receptor and its target genes. Filled and unfilled circles indicate two independent experiments (each three replicates). Data are presented as the mean ± Standard derivatives (SD). Statistical significance was evaluated using a Deseq2. (C) Enrichment plot of genes in Ppar Signaling Pathway (KEGG) for MHCO (left panel) and Caco-2 (right panel). Each vertical bar represents a gene, and genes enriched in upregulated and downregulated after coculturing with B. longum are at the left and right part of the graph, respectively. The normalized enrichment score (NES), the p-value, and the false discovery rate (Adjusted p-value) were indicated in the insert.
To validate the upregulation of PPAR signaling in MHCO cocultured with B. longum, we focused on the expression of the PPARA and PPARG target genes. In MHCO coculturing with B. longum, a fatty acid oxidation gene acyl-CoA oxidase 1 (ACOX1), a direct target gene of PPARA in the intestine35, was elevated (Fig. 5B). Angiopoietin like 4 (ANGPTL4), another identical target gene of PPARG in colonocytes36, was also upregulated in MHCO (Fig. 5B). As expected, gene set enrichment analysis (GSEA) indicated that MHCO, but not Caco-2 cocultured with B. longum were enriched in genes related to PPAR signaling (normalized enrichment score [NES] = 1.91, adjP = 0.04 in MHCO, NES = -0.86, adjP = 0.96, Fig. 5C). These data indicated that live B. longum has a potential to activate PAPR signaling on normal physiological conditions of colonic epithelium.
Differential inflammatory response to MHCO and Caco-2 to B. longum stimulation
Next, we characterized the reaction of each of the cell lines to microbial stimulation by determining their immune responses following stimulation and further investigated gene expression profiles in the production of cytokine and chemokine molecules and their regulators37. GSEA revealed that Caco-2 and MHCO stimulated by B. longum were commonly enriched for genes associated the NF-κB signaling pathway (NES = 1.75, adjP = 0.05 in MHCO; NES = 2.1, adjP = 0.01 in Caco-2; Fig. 6A). To assess the responses of the Caco-2 and MHCO comparatively, we calculated the vehicle control and B. longum-induced log2 fold change (FC) for each gene. As expected, the responses induced by five chemokine genes (C-C motif chemokine ligand 20 [CCL20], C-X-C motif chemokine ligand 1 [CXCL1], CXCL2, CXCL3, and CXCL8) implicated in bacterial sensing38 showed more pronounced FC values in Caco-2 than in MHCO (Fig. 6B, C).
B. longum regulates inflammatory response differently in MHCO and Caco-2. (A) Enrichment plot of genes in Nf-Kappa B Signaling Pathway (KEGG) for MHCO (left panel) and Caco-2 (right panel). (B) Heatmap of relative expression of inflammatory response-related genes by co-cultivation of B. longum in Caco-2 and MHCO. Statistical significance was evaluated by a Deseq2 (*adjusted p-value < 0.05, ** adjusted p-value < 0.01, *** adjusted p-value < 0.001). (C) TPM values of representative chemokine, NF-κB related, and inflammatory cytokine genes. Filled and unfilled circles indicate two independent experiments (each three replicates). Data are presented as the mean ± Standard derivatives (SD). Statistical significance was evaluated by Deseq2.
Moreover, bacterial inoculation of Caco-2 and MHCO enhanced the expression of NFKB signaling target genes (NFKB inhibitor epsilon [NFKBIE], NFKB inhibitor zeta [NFKBIZ])39 (Fig. 6B, C). In addition, the expression level of other NF-κB pathway-related genes, TNFAIP3, NFKB inhibitor alpha (NFKBIA), and RELB proto-oncogene, NF-κB subunit (RELB) were more prominently increased in Caco-2, indicating a comparatively stronger inflammatory response in Caco-2 following bacterial stimulation (Fig. 6B, C). On the other hand, Interleukin 1 A (IL-1 A) and IL-1B, inflammatory cytokine genes crucial for inflammation and infection defense, were significantly upregulated following coculturing of B. longum and MHCO specifically40 (Fig. 6B, C). Those results suggest that a probiotic microbe, B. longum, raises the level of immune surveillance.
B. longum upregulates intestinal barrier function in MHCO
We investigated the effect of Bifidobacteria on the intestinal barrier function, by examining the effects on expression patterns of relevant genes. Our RNA-seq results revealed substantial differences in expression patterns of genes related to the mucosal barrier function between Caco-2 and MHCO in response to B. longum. The tight junction-related genes, Claudin 3 (CLDN3), CLDN4, and Occludin (OCLN), were upregulated in both Caco-2 and MHCO following exposure to B. longum. In contrast, other related genes, CLDN7 and Tight junction protein 3 (TJP3), were only upregulated in MHCO, whereas CLDN2 was upregulated in Caco-2 (Fig. 7A, B). Other components of the adherens junction (Cadherin 1 [CDH1], CDH17, Epithelial cell adhesion molecule [EPCAM]) or desmosomes (Desmoglein 2 [DSG2]) were linked to a set of genes that were upregulated in B. longum-stimulated MHCO and conversely downregulated in Caco-2 (Fig. 7A, C, D). The polarity of intestinal epithelial cells plays a central role in establishing the barrier function for symbiotic relationships with microbiota41. The B. longum-induced expression pattern of Integrin subunit α6 (ITGA6), a significant factor in the epithelium and basement membrane interaction, also differed between MHCO and Caco-2 (Fig. 7A, E).
B. longum regulates intestinal barrier function. (A) Heatmap of relative expression of cell junction-related genes by co-cultivation of B. longum in Caco-2 and MHCO. Statistical significance was evaluated by a DESeq2 (*adjusted p-value < 0.05, ** adjusted p-value < 0.01, *** adjusted p-value < 0.001). (B–E) TPM values of representative tight junction (B), adherence junction (C), desmosome (D), and hemidesmosome (E) related genes. Filled and unfilled circles indicate two independent experiments (each three replicates). Data are presented as the mean ± Standard derivatives (SD). Statistical significance was evaluated using a Deseq2. (F, G) Relative TEER value adjusted by the initial value (F) and permeability of Lucifer Yellow (LY) from apical to basolateral compartment (G) in MHCO, and relative TEER value adjusted by the initial value in Caco-2 (H) cocultured with B. longum versus vehicle control. Data are expressed as means ± SD. Statistical significance was evaluated using Welch’s t-test.
To investigate whether increased gene expression induced by B. longum was associated with improved barrier function, we measured the transepithelial electrical resistance (TEER) in MHCO. Compared to the vehicle control, TEER was significantly increased in cells cocultured with B. longum (Fig. 7F). We also evaluated the paracellular permeability of the colonic epithelium using the Lucifer yellow (LY) flux assay with the fluorescent molecule LY, which does not interact with cell components42. Consistent with the changes in TEER, LY flux was downregulated when MHCO were cultured with B. longum (Fig. 7G). Transcriptomics data indicated that some tight junction-associated genes, including CLDN3, CLDN4, and OCLN, were upregulated in both cell types upon stimulation of bacteria. However, TEER analysis revealed a notable difference in barrier function, and no comparable enhancement was observed in Caco-2 (Fig. 7H). Overall, these results suggest that probiotic Bifidobacteria enhance intestinal barrier integrity through the PPARG/STAT3 pathway.
Discussion
In this study, we compared gene expression data between Caco-2 and MHCO cultured under hemi-anaerobic conditions to provide valuable datasets for investigating host cells and anaerobic bacteria in the colon. We demonstrated the presence of all cell types that form the colonic epithelium, including stem, goblet, and enteroendocrine cells and colonocytes among the MHCO cultured under hemi-anaerobic conditions. In contrast, the gene expression pattern of Caco-2 was similar to that of epithelial cells of the small intestine, as shown previously43. Therefore, MHCO may be a physiologically appropriate model for studying probiotic activities, including of Bifidobacterium and Akkermansia, in the colon34,44.
The Caco-2, a cell line generated from a patient with colorectal cancer is the most commonly used in studying the physiological functions of the human intestinal epithelium. These cells spontaneously transform to differentiate and form monolayers of polarized cells with functions similar to those of enterocytes in the small intestine when grown under confluent conditions in Boyden chambers. Consequently, Caco-2 have been widely used to investigate the metabolism and absorption of drugs and nutrients15,45.
Gnotobiotic mice have been the primary animal model used in previous studies of the effect of microbes on hosts. Although studies using in vivo mouse models provide important insights into the physiology of host-microbiome interactions, they are unable to account for significant differences between mice and humans because bacteria exhibit strong species specificity46. Conventionally, cell lines derived from patients with cancer have been used to study microbial responses. In this study, healthy primary colonic epithelial cells cultured as organoids and a conventional colorectal cancer-derived cell line were used to examine the physiological response of the host epithelium to microbiota. The gene expression patterns in the DEG analysis in our study indicated that healthy human colonic epithelium derived from organoids was more robust against gut bacteria than the cancer-derived Caco-2 cell line was. These results demonstrated the observed low incidence of overlapping upregulated DEGs between MHCO and Caco-2 and decreased expression of DEGs, indicating that the impact of coculturing with B. longum on gene expression was not conserved across these two cell types.
Caco-2 have been used to assess the efficacy of probiotics and intestinal bacteria in evaluating mucosal barriers and inflammatory responses. However, two independent reports demonstrated the immunoreactivity of Faecalibacterium prausnitzii (recently reclassified as Faecalibacterium duncaniae) in host cells using Caco-2 or primary colonic epithelium (organoid)12,47. To the best of our knowledge, this study is the first to evaluate the effect of gut bacteria on Caco-2 and colonic organoids incubated under matched culture and hemi-anaerobic conditions.
We performed RNA-seq to compare gene profiles between Caco-2 and MHCO cultured in the same condition. Our GO enrichment analysis observed that genes upregulated in MHCO were significant with the GO term “tissue homeostasis” (GO: 0001894). Notably, this category includes key goblet cell marker genes such as MUC2 and TFF3. These genes are well known for protecting and stabilizing the intestinal mucosal barrier, thereby contributing to epithelial homeostasis. Through the secretion of mucins and other protective factors, Goblet cells play a crucial role in maintaining gut integrity and defense against pathogens and foreign substances. This functional link has been substantiated by previous studies48,49. Thus, the observed enrichment of the “tissue homeostasis” GO term in MHCO likely reflects the activation contribution of goblet cell-related genes to the maintenance of colonic epithelial function.
In this study, the expression of various inflammation-related genes was similarly induced by B. longum in Caco-2 and organoid-derived healthy epithelial cells. However, the fold-change ratio was higher in Caco-2 than it was in the organoid cells. This probably occurred because Caco-2 lack goblet cells and poorly express genes such as MUC2, which renders the barrier function of the intestinal epithelium vulnerable. Moreover, the bacteria may have more readily caused inflammation in the Caco-2 than in MHCO. Therefore, we focused on immune responses to characterize the reaction of Caco-2 and MHCO to microbial stimulation because genes commonly upregulated in these cell lines show pronounced enrichment of GO terms associated with inflammation. The intestinal epithelium interfaces directly with the external environment and acts as a constant frontline defense for the immune response50. Epithelial cells play an essential role in preventing bacterial invasion by producing cytokines and chemokines that determine the nature of the immune response by controlling immune cells. These facts prompted our investigation of gene expression profiles in the production of immune-related molecules and their regulators, including NFKB inhibitors and TNFAIP37. The results suggest that the probiotic microbe, B. longum, raised levels of immune surveillance.
In addition, our GSEA results showing the enrichment of genes related to PPAR signaling in bacterial-stimulated MHCO but not Caco-2 led us to conclude that live B. longum has the potential to activate PAPR signaling under normal physiological conditions in the colonic epithelium. These observations further confirmed that MHCO are more pertinent for observing the physiological response of the colonic epithelium than Caco-2 are when cultured under semi-anaerobic conditions.
Moreover, the results of the TEER analysis corroborated the relationship between B. longum-induced increase in gene expression and improved barrier function in MHCO. These observations further confirmed the findings of previous studies in MHCO cocultured with live B. longum under hemi-anaerobic culture conditions, which showed that probiotic Bifidobacteria enhanced the intestinal barrier integrity through the PPARG/signal transducer and activator of transcription 3 (STAT3) pathway. Previous animal and clinical studies have revealed that Bifidobacterium, a probiotic, positively regulates the barrier function of mucosal epithelial cells51.
Epithelial cells are held together by strong anchoring junctions that include tight junctions, adherens junctions, and desmosomes, which are formed by transmembrane adhesion proteins52. The expression of CLDN4, a component of tight junctions, has been shown to be upregulated in the colonic epithelial cells of mice administered Bifidobacterium53. Similarly, our investigation using human colonic organoids cocultured with B. longum also showed the upregulation of CLDN4. Contrastingly, several molecules related to epithelial cell junctions, such as CLDN2, showed conflicting expression patterns in Caco-2 and MHCO stimulated with B. longum. CLDN2 is a pore-forming CLDN that permeates water and ions and has been reported to be elevated in patients with inflammatory bowel disease or neonatal necrotizing enterocolitis54,55,56,57. Furthermore, we found that CLDN7, a tight junction-related gene, was upregulated only in MHCO cocultured with B. longum. Increased CLDN2 and decreased CLDN7 levels have previously been observed in the gastrointestinal tracts of patients with ulcerative colitis58, and CLDN7 knockout mice develop colitis spontaneously59. In this study, to the best of our knowledge, we provide the first proof that Bifidobacterium modulates CLDN7 expression in the physiologically normal human colonic epithelium.
The polarity of intestinal epithelial cells plays a central role in establishing a barrier function for symbiotic relationships with microbiota41. Hemidesmosomes, which are composed of ITGA6 and are a significant factor in the interaction between the epithelium and basement membrane, are expressed at specific junctions in the basal portion of intestinal epithelial cells60. In addition to other factors, improvement of the mucus barrier function is a known molecular mechanism in coculture systems. A novel finding of this study is that more genes were altered by bacterial stimulation in Caco-2 than they were in the organoids. While both organoids and Caco-2 share certain variable gene sets, each cell line exhibited a unique variation in the expression of genes linked to inflammation and the mucosal barrier, which are well-known roles of Bifidobacterium. However, the potential usefulness and applicability of Bifidobacterium needs to be further investigated using appropriate in vitro culture systems.
Methods
Colonic organoid culture
Clinical samples for healthy human colonic organoid establishment and coculturing microbes were obtained from patients at Gunma University Hospital with informed consent after study approval by the ethical committees (HS2022-054). The informed consent was obtained from donors as required, and all methods were carried out in under the Declaration of Helsinki. The human colonic organoid was cultured as in the previous report with minor modifications10. Briefly, three-dimensional colonic organoids were maintained with Modified human colonic organoid (MHCO) medium, consisting of advanced Dulbecco’s modified Eagle’s medium (DMEM)/F12 supplemented with penicillin/streptomycin, 10mM HEPES, 2 mM Glutamax, 1×B-27 Supplement (Thermo Fisher Scientific), 1 mM N-acetylcysteine (Sigma-Aldrich), 50 ng/mL recombinant mouse epidermal growth factor (Thermo Fisher Scientific), 100 ng/mL mouse recombinant noggin (Peprotech), 1 µg/mL human recombinant R-spondin1 (R&D), 500 nM A83-01 (Tocris), 100 ng/mL recombinant human insulin-like growth factor-1 (BioLegend), 50 ng/mL recombinant human fibroblast growth factor-basic (FGF-2) (PeproTech) and conditioned medium containing Wnt3A. The human colonic organoid was derived from a nonpathological biopsy and confirmed by genomic sequencing analysis that it has no driver mutations related to colorectal cancer, as shown in the previous paper61. Organoids were passed approximately every 5–7 days by physical dissociation using fire-polished Pasteur pipettes. To generate MHCO, ThinCert culture inserts (24-well insert, 0.4 μm pore polyester membrane; Greiner bio-one) were coated with 4% Matrigel diluted with advanced DMEM/F12 medium and incubated at 25ºC for 30 min, then Matrigel solution was removed. The membrane was dried in a tissue-culture hood for 15 min. Human colonic organoids were cultured for 3 to 5 days before being used to plate into monolayer culture in MHCO medium. Three-dimensional cultured organoids were treated with TrypLE Express (Thermo Fisher Scientific) to dissociate into single cells. The cells were resuspended to 1 × 106 cells/mL in MHCO medium containing 10 µM Y-27,632 (FUJIFILM Wako Pure Chemical Corporation), and 200 uL of cell suspension was added into the transwell inserts. After 1 day of monolayer culture, MHCO without Y-27,632 was used, and the medium was changed every 2 days. When MHCO was cultured in the inserts, MHCO without penicillin/streptomycin was used for their maintenance.
Caco-2 culture
The cell line Caco-2 maintained in our group was confirmed by the Cell Line Authentication Service (ATCC). Caco-2 was maintained with the medium consisting of DMEM (high glucose) with L-Glutamine and Phenol Red (FUJIFILM Wako Pure Chemical Corporation) supplemented with 10% heat-treated FBS (Hyclone), MEM Non-Essential Amino Acids Solution (Nacalai tescue), and penicillin/streptomycin. Caco-2 cells were passaged approximately 3–5 days before reaching 80% confluency using 0.25% Trypsin-EDTA (Gibco) and seeded at 2–4 × 104 cells/cm2. 105 cells were seeded in transwell culture insert coated Matrigel solution described above. Cells were cultured with the 10% FBS-containing DMEM medium for 17 days. Before the coculture experiment, the culture medium was changed to MHCO medium without antibiotics and cultured for 2 days.
Bacterial culture
B. longum JCM1217T (B. longum) was obtained from the Riken BioResource Research Center (JCM, Tokyo, Japan). This strain was cultured in modified GAM broth (Nissui Pharmaceutical, Gifu, Japan) for 4 h at 37ºC under anaerobic conditions using a BACTRON300 anaerobic chamber (Sheldon Manufacturing, Inc., Cornelius, USA). The culture medium was centrifuged (8,000 x g, 1 min, 25ºC) and suspended to approximately 5.0 × 107 CFU/mL in a deoxygenated WENRAIF medium supplemented with 100mM HEPES in an anaerobic chamber. Then, 200 uL bacterial suspension was added at the apical side of the transwell in an anaerobic chamber and capped in butyl gum caps. Then, the basolateral side’s medium was cultured in a CO2 incubator for 24 h. The number of colony-forming units (CFU) on TOS propionate agar (Yakult Pharmaceutical Industry) was used to calculate the number of live B. longum.
RNA isolation and sequencing
Total RNA was extracted using the RNeasy Mini Kit (Qiagen) with an RNase-Free DNase Set (Qiagen). Library preparation was performed using Illumina Stranded mRNA Prep and IDT® for Illumina® RNA UD Indexes Set A-B Ligation (Illumina). The concentration and quality of the extracted RNA and adapter-tagged sequence library were calculated using Agilent RNA 6000 Nano and Agilent High Sensitivity DNA Kits (Agilent Technologies), respectively. Sequences were obtained using the NextSeq 1,000 system with the Illumina NextSeq 1000/2000 P2 Reagent kit (100 cycles) (Illumina). Read files were trimmed and mapped to the human reference genome (GRCh38.p.13) using Trim Galore! (ver.0.6.4, https://github.com/FelixKrueger/TrimGalore) and HISAT2 (ver.2.2.1) with default settings62. Transcript assembly, GTF document merging, and transcript quantification were performed using Stringtie (ver.2.2.0)63. A matrix of read counts mapped to genes was extracted using prepDE.py (http://ccb.jhu.edu/software/stringtie/dl/prepDE.py3).
Data analysis and visualization
Statistical analysis and data visualization were performed using R (ver. 4.3.2). Normalization from read counts and identifying differentially expressed genes (DEGs) were performed using DESeq2 packages (ver. 1.46.0). Batch correlation was performed by using limma (ver. 3.36.2). Adjusted p-value < 0.01 and log2 fold change > 1.5 were defined as DEGs. The principal component analysis (PCA) was performed using the ‘prcomp’ function with rlog transformed counts matrix in the stats package (ver. 4.6.3, https://www.rdocumentation.org/packages/stats/versions/4.6.3) with scaling. The hierarchical clustering and plotting of the heatmap were performed by the ‘Heatmap’ function in the ComplexHeatmap package (ver. 2.22.0, https://github.com/jokergoo/ComplexHeatmap). Functional enrichment analysis of DEGs was performed using the ‘enrichGO’ function in the clusterProfiler package (ver. 4.14.6, https://github.com/YuLab-SMU/clusterProfiler) and GO terms with p-value < 0.05 and adjusted p-value < 0.05 were considered significant. KEGG pathway enrichment analysis was performed using the ‘gseKEGG’ function in the clusterProfiler package, and the significantly enriched pathways were defined by nominal adjusted p-value < 0.05.
Measurement of the integrity of the cell monolayer
The integrity of the cell monolayer was assessed with relative TEER value and LY assay. TEER was measured by Millicell ERS-2 Voltohmmeter (Merck Millipore) following the manufacturer’s instruction. Relative TEER was calculated by TEER value after Hemi-Anerobic cultivation for 24 h relative to their initial value (0 h). HBSS supplemented with 10mM HEPES, D-PBS(+) Preparation Reagent (Ca, Mg Solution) (Nacalai tesque) were used as transport buffers. After washing the monolayer with transport buffer, 300uL of Lucifer yellow solution (300 μm prepared in transport buffer) and 1000uL of transport buffer were added to the apical and basolateral sides of the monolayers, respectively. The monolayers were incubated for 60 min in a CO2 incubator. After incubation, 200uL of apical and basolateral solution was collected in the 96-well black assay microplate, and the fluorescence was measured at an emission wavelength of 538 nm and excitation at 485 nm using Perkin Elmer Enspire 2300 Multi-mode Microplate Reader with EnSpire Workstation (version 4.13.3005.1482) n = 4 biological replicates.
Data availability
The raw sequence data for RNA-seq analysis have been deposited at Japanese Genotype-phenotype Archive (JGA) under accession number JGAS000740.
References
Turner, J. R. Intestinal mucosal barrier function in health and disease. Nat. Rev. Immunol. 9, 799–809 (2009).
Kiela, P. R. & Ghishan, F. K. Physiology of intestinal absorption and secretion. Best Pract. Res. Clin. Gastroenterol. 30, 145–159 (2016).
Lynch, S. V. & Pedersen, O. The human intestinal Microbiome in health and disease. N Engl. J. Med. 375, 2369–2379 (2016).
Rowland, I. et al. Gut microbiota functions: metabolism of nutrients and other food components. Eur. J. Nutr. 57, 1–24 (2018).
Zheng, D., Liwinski, T. & Elinav, E. Interaction between microbiota and immunity in health and disease. Cell. Res. 30, 492–506 (2020).
Sanders, M. E., Merenstein, D. J., Reid, G., Gibson, G. R. & Rastall, R. A. Probiotics and prebiotics in intestinal health and disease: from biology to the clinic. Nat. Rev. Gastroenterol. Hepatol. 16, 605–616 (2019).
Plaza-Diaz, J., Ruiz-Ojeda, F. J., Gil-Campos, M. & Gil, A. Mechanisms of Action of Probiotics. Adv. Nutr., 10 S49–S66. (2019).
Jalili-Firoozinezhad, S. et al. A complex human gut Microbiome cultured in an anaerobic intestine-on-a-chip. Nat. Biomed. Eng. 3, 520–531 (2019).
Kim, R. et al. An in vitro intestinal platform with a self-sustaining oxygen gradient to study the human gut/microbiome interface. Biofabrication 12, 015006 (2019).
Sasaki, N. et al. Development of a scalable coculture system for gut anaerobes and human colon epithelium. Gastroenterology 159, 388–390 e385. (2020).
Shah, P. et al. A microfluidics-based in vitro model of the Gastrointestinal human-microbe interface. Nat. Commun. 7, 11535 (2016).
Zhang, J. et al. Primary human colonic mucosal barrier crosstalk with super oxygen-sensitive Faecalibacterium prausnitzii in continuous culture. Med 2, 74–98e79 (2021).
Qi, Y., Yu, L., Tian, F., Zhao, J. & Zhai, Q. Vitro models to study human gut-microbiota interactions: applications, advances, and limitations. Microbiol. Res. 270, 127336 (2023).
Sambuy, Y. et al. The Caco-2 cell line as a model of the intestinal barrier: influence of cell and culture-related factors on Caco-2 cell functional characteristics. Cell. Biol. Toxicol. 21, 1–26 (2005).
Ding, X. et al. Differentiated Caco-2 cell models in food-intestine interaction study: current applications and future trends. Trends Food Sci. Technol. 107, 455–465 (2021).
Donaldson, G. P., Lee, S. M. & Mazmanian, S. K. Gut biogeography of the bacterial microbiota. Nat. Rev. Microbiol. 14, 20–32 (2016).
Sato, T. et al. Long-term expansion of epithelial organoids from human colon, adenoma, adenocarcinoma, and Barrett’s epithelium. Gastroenterology 141, 1762–1772 (2011).
Sato, T. et al. Single Lgr5 stem cells build crypt-villus structures in vitro without a mesenchymal niche. Nature 459, 262–265 (2009).
Wang, Y. et al. Self-renewing monolayer of primary colonic or rectal epithelial cells. Cell. Mol. Gastroenterol. Hepatol. 4, 165–182. e167 (2017).
Fujii, M. et al. Human intestinal organoids maintain Self-Renewal capacity and cellular diversity in Niche-Inspired culture condition. Cell. Stem Cell. 23, 787–793e786 (2018).
Jumarie, C. & Malo, C. Caco-2 cells cultured in serum-free medium as a model for the study of enterocytic differentiation in vitro. J. Cell. Physiol. 149, 24–33 (1991).
Tetteh, P. W. et al. Clevers, replacement of lost Lgr5-Positive stem cells through plasticity of their Enterocyte-Lineage daughters. Cell. Stem Cell. 18, 203–213 (2016).
Grant, C. N. et al. Human and mouse tissue-engineered small intestine both demonstrate digestive and absorptive function. Am. J. Physiol. Gastrointest. Liver Physiol. 308, G664–677 (2015).
Zweibaum, A. et al. Immunohistological evidence, obtained with monoclonal antibodies, of small intestinal brush border hydrolases in human colon cancers and foetal colons. Int. J. Cancer. 34, 591–598 (1984).
Wang, Y. et al. Single-cell transcriptome analysis reveals differential nutrient absorption functions in human intestine. J. Exp. Med. 217 (2020).
Mouradov, D. et al. Colorectal cancer cell lines are representative models of the main molecular subtypes of primary cancer. Cancer Res. 74, 3238–3247 (2014).
Konsavage, W. M. Jr., Kyler, S. L., Rennoll, S. A., Jin, G. & Yochum, G. S. Wnt/beta-catenin signaling regulates Yes-associated protein (YAP) gene expression in colorectal carcinoma cells. J. Biol. Chem. 287, 11730–11739 (2012).
Tang, S. L., Gao, Y. L. & Chen, X. B. Wnt/beta-catenin up-regulates midkine expression in glioma cells. Int. J. Clin. Exp. Med. 8, 12644–12649 (2015).
Hashimoto, M. et al. Spatial and single-cell colocalisation analysis reveals MDK-mediated immunosuppressive environment with regulatory T cells in colorectal carcinogenesis. EBioMedicine 103, 105102 (2024).
Hickey, J. W. et al. Organization of the human intestine at single-cell resolution. Nature 619, 572–584 (2023).
Burgueno, J. F. & Abreu, M. T. Epithelial Toll-like receptors and their role in gut homeostasis and disease. Nat. Rev. Gastroenterol. Hepatol. 17, 263–278 (2020).
Hall, J. A., Grainger, J. R., Spencer, S. P. & Belkaid, Y. The role of retinoic acid in tolerance and immunity. Immunity 35, 13–22 (2011).
McCoy, R. et al. In vitro models for investigating intestinal Host-Pathogen interactions. Adv. Sci. (Weinh). 11, e2306727 (2024).
Derrien, M., Turroni, F., Ventura, M. & van Sinderen, D. Insights into endogenous Bifidobacterium species in the human gut microbiota during adulthood. Trends Microbiol. 30, 940–947 (2022).
Luo, Y. et al. Intestinal PPARalpha protects against colon carcinogenesis via regulation of methyltransferases DNMT1 and PRMT6. Gastroenterology 157, 744–759 e744. (2019).
Kersten, S. et al. Characterization of the fasting-induced adipose factor FIAF, a novel peroxisome proliferator-activated receptor target gene. J. Biol. Chem. 275, 28488–28493 (2000).
Rahman, M. M. & McFadden, G. Modulation of NF-kappaB signalling by microbial pathogens. Nat. Rev. Microbiol. 9, 291–306 (2011).
Friedrich, M., Pohin, M. & Powrie, F. Cytokine networks in the pathophysiology of inflammatory bowel disease. Immunity 50, 992–1006 (2019).
Bren, G. D. et al. Transcription of the RelB gene is regulated by NF-kappaB. Oncogene 20, 7722–7733 (2001).
Mantovani, A., Dinarello, C. A., Molgora, M. & Garlanda, C. Interleukin-1 and related cytokines in the regulation of inflammation and immunity. Immunity 50, 778–795 (2019).
McGuckin, M. A., Eri, R., Simms, L. A. & Florin, T. H. Radford-Smith, intestinal barrier dysfunction in inflammatory bowel diseases. Inflamm. Bowel Dis. 15, 100–113 (2009).
Marr, E. E. et al. A platform to reproducibly evaluate human colon permeability and damage. Sci. Rep. 13, 8922 (2023).
Kulthong, K. et al. Transcriptome comparisons of in vitro intestinal epithelia grown under static and microfluidic gut-on-chip conditions with in vivo human epithelia. Sci. Rep. 11, 3234 (2021).
Cheng, D. & Xie, M. Z. A review of a potential and promising probiotic candidate-Akkermansia muciniphila. J. Appl. Microbiol. 130, 1813–1822 (2021).
Panse, N. & Gerk, P. M. The Caco-2 model: modifications and enhancements to improve efficiency and predictive performance. Int. J. Pharm. 624, 122004 (2022).
Mallott, E. K. & Amato, K. R. Host specificity of the gut Microbiome. Nat. Rev. Microbiol. 19, 639–653 (2021).
Sadaghian Sadabad, M. et al. A simple coculture system shows mutualism between anaerobic faecalibacteria and epithelial Caco-2 cells. Sci. Rep. 5, 17906 (2015).
Gustafsson, J. K. & Johansson, M. E. V. The role of goblet cells and mucus in intestinal homeostasis. Nat. Rev. Gastroenterol. Hepatol. 19, 785–803 (2022).
McCauley, H. A. & Guasch, G. Three cheers for the goblet cell: maintaining homeostasis in mucosal epithelia. Trends Mol. Med. 21, 492–503 (2015).
Allaire, J. M. et al. The intestinal epithelium: central coordinator of mucosal immunity. Trends Immunol. 39, 677–696 (2018).
Abdulqadir, R., Engers, J. & Al-Sadi, R. Role of Bifidobacterium in modulating the intestinal epithelial tight junction barrier: current knowledge and perspectives. Curr. Dev. Nutr. 7, 102026 (2023).
Garcia, M. A., Nelson, W. J. & Chavez, N. Cell-Cell junctions organize structural and signaling networks. Cold Spring Harb Perspect. Biol. 10 (2018).
Shu, X. et al. Bifidobacterium lactis TY-S01 protects against alcoholic liver injury in mice by regulating intestinal barrier function and gut microbiota. Heliyon 9, e17878 (2023).
Bergmann, K. R. et al. Bifidobacteria stabilize claudins at tight junctions and prevent intestinal barrier dysfunction in mouse necrotizing Enterocolitis. Am. J. Pathol. 182, 1595–1606 (2013).
Horowitz, A., Chanez-Paredes, S. D., Haest, X. & Turner, J. R. Paracellular permeability and tight junction regulation in gut health and disease. Nat. Rev. Gastroenterol. Hepatol. 20, 417–432 (2023).
Prasad, S. et al. Inflammatory processes have differential effects on claudins 2, 3 and 4 in colonic epithelial cells. Lab. Invest. 85, 1139–1162 (2005).
Zeissig, S. et al. Changes in expression and distribution of Claudin 2, 5 and 8 lead to discontinuous tight junctions and barrier dysfunction in active crohn’s disease. Gut 56, 61–72 (2007).
Oshima, T., Miwa, H. & Joh, T. Changes in the expression of claudins in active ulcerative colitis. J. Gastroenterol. Hepatol. 23 (Suppl 2), S146–150 (2008).
Ding, L. et al. Inflammation and disruption of the mucosal architecture in claudin-7-deficient mice. Gastroenterology 142, 305–315 (2012).
Stutzmann, J., Bellissent-Waydelich, A., Fontao, L., Launay, J. F. & Simon-Assmann, P. Adhesion complexes implicated in intestinal epithelial cell-matrix interactions. Microsc Res. Tech. 51, 179–190 (2000).
Fujii, M. et al. A colorectal tumor organoid library demonstrates progressive loss of niche factor requirements during tumorigenesis. Cell. Stem Cell. 18, 827–838 (2016).
Kim, D., Paggi, J. M., Park, C., Bennett, C. & Salzberg, S. L. Graph-based genome alignment and genotyping with HISAT2 and HISAT-genotype. Nat. Biotechnol. 37, 907–915 (2019).
Pertea, M. et al. StringTie enables improved reconstruction of a transcriptome from RNA-seq reads. Nat. Biotechnol. 33, 290–295 (2015).
Acknowledgements
The authors thank the members of Gastroenterological Surgery, School of Medical Faculty of Medicine at Gunma University for human sample collection and all members of the laboratory for Mucosal Ecosystem Design for their assistance in experiments and discussion. A Wnt3a-producing cell line was a kind gift from Hans Clevers (Hubrecht Institute).
Funding
This work was supported in part by the Japan Agency for Medical Research and Development (AMED) grant (JP 23ae0121046), JST FOREST program (JPMJFR2161), Grants-in-Aid for the Japanese Society for the Promotion of Science (JSPS) KAKENHI (19H03455, 23H02713), a Grant-in-Aid for Challenging Research (Pioneering, 23K17415), the Astellas Foundation for Research on Metabolic Disorders, and the LOTTE Foundation.
Author information
Authors and Affiliations
Contributions
A.S., A.I., F.S., and T. Odamaki performed all the empirical experiments and analyzed the data. A.S., A. I., K. Y., E. M., and T. Oda. analyzed the RNA-seq data. S.U., T. Okada, T.Y., and H.S. provided tissue specimens. N.S. conceived and conducted the project. A.S. and N.S. wrote the manuscript. All authors discussed the results and edited the manuscript.
Corresponding author
Ethics declarations
Competing interests
A. S. and T. O. are employees of Morinaga Milk Industry Co. Ltd. The remaining authors disclose no conflicts.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Sen, A., Imai, A., Miyauchi, E. et al. Comprehensive gene expression analysis of organoid-derived healthy human colonic epithelium and cancer cell line stimulated with live probiotic bacteria. Sci Rep 15, 22325 (2025). https://doi.org/10.1038/s41598-025-07391-x
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-025-07391-x
