NAD+‐Dependent Enzyme SIRT3 Limits Intestinal Epithelial Cell Functions Through NAD+ Synthesis Pathway in Colorectal Cancer

TL;DR

This study shows how the enzyme SIRT3 in intestinal cells affects immune responses in colorectal cancer and colitis through NAD+ synthesis.

Contribution

The novel finding is that microbiota-derived quinolinic acid replenishes NAD+ in intestinal cells under SIRT3 deficiency, influencing immune function.

Findings

SIRT3 deficiency in intestinal epithelial cells promotes TH1 and CTL differentiation through IL-1β-IL-1R1 signaling.

Microbiota-derived 3-hydroxyaminobenzoic acid is a source of quinolinic acid for NAD+ synthesis in intestinal cells.

Targeting intestinal epithelial cells via NAD+ regulation may offer new treatment approaches for immune-associated diseases.

Abstract

Intestinal epithelial cells (IEC) are crucial for regulating intestinal local immunity to potentiate mucosal barrier function, but the mechanism remains unclear. In this study, we showed that the nicotinamide adenine dinucleotide (NAD+)‐dependent enzyme SIRT3 in IECs is required for local T cell differentiation in colorectal cancer and colitis. Modest IEC SIRT3 overexpression reduces the secretion of proinflammatory cytokine IL‐1β and inhibits IFNγ‐producing CD4+T cells (TH1) and cytotoxic T lymphocytes (CTLs) differentiation. IEC SIRT3 deficiency enhances the production of IL‐1β and promotes local TH1 and CTL differentiation in limiting colorectal cancer growth and aggravating colitis. Mechanistically, SIRT3 deficiency promotes IEC functions through quinolinic acid (QA)‐mediated NAD+ synthesis for limiting tumor growth. Microbiota‐derived 3‐hydroxyaminobenzoic acid is the source of intracellular QA in IECs. IL‐1β‐IL‐1R1 signaling is required for IEC SIRT3 deficiency‐induced TH1 and CTL differentiation in cancer. Thus, our findings showed that microbiota‐derived QA is used as an alternative source of replenishing the intracellular NAD+ pool induced by SIRT3 deficiency to regulate IEC and T cell function, which has implications for targeting IECs as an approach to the treatment of immune‐associated diseases, including colorectal cancer and colitis.

Microbiota‐derived quinolinic acid is used as an alternative source of replenishing the intracellular NAD+ pool induced by SIRT3 deficiency to regulate intestinal epithelial cell and T cell function, which has implications for targeting intestinal epithelial cells as an approach to the treatment of immune‐associated diseases, including colorectal cancer and colitis.

Figures

SIRT3 signaling from IECs is associated with TH1 and CTL function in colorectal cancer. (A,B) ScRNA‐seq datasets were retrieved from GSE261388. (A) Gene expression analysis of IEC clusters in colitis and colorectal cancer using the “Find All Markers” function. (B) Violin plots of IEC SIRT3 expression in colitis and colorectal cancer. (C) Survival analysis of human colon adenocarcinoma associated with IEC SIRT3 expression based on samples from the TCGA database. Analysis was performed using GEPIA 2 (http://gepia2.cancer‐pku.cn) online. (D–G) C57BL/6 mice were treated with 1.5 % dextran sulfate sodium salt (DSS) for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer, and Sirt3 expression was examined by qPCR in sorted Epcam1+ IEC cells from colon, and the percent of IFNγ+ TH1 and CTL cells was examined by flow cytometry at different time points. (D) The mRNA expression levels of SIRT3 in mouse IECs under different physiological conditions. Correlations between the percent of TH1 (E) and CTL cells (F) and Sirt3 mRNA levels. Correlations between tumor size and the Sirt3 mRNA level (G). The data are representative of three independent experiments (C, n = 3‐5; D‐H; n = 10 mice per group). * p<0.05 and * p<0.001, compared with the indicated groups.

IEC SIRT3 deficiency promotes TH1 and CTL cell function in inhibiting colorectal cancer growth. Wild‐type (WT) and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer. (A) A typical image of colorectal cancer (left) and a summary of tumor volume (right). (B) HE staining of colorectal cancer samples. Scale bars, 50 µm. Original magnification, 50X. (C) Expressions of MHCII and CD54 among Epcam1+ IECs in the tumor. (D) Percent of IL‐1α, IL‐1β, IL‐12, and IL‐23 among Epcam1+ IECs in the tumor. (E,F) Percent of CD44highCD62L− cells among CD4+T cells (E; left) and CD8+T cells (E; right) in tumor and draining lymph node (dLN) in WT and Sirt3 ∆Epi mice. A representative dot plot is shown (E), and data are summarized (F). (G) Expression of IL‐4, IL‐17A, and Foxp3 among CD4+T cells in the tumor in mice. A representative dot plot is shown on the left, and data are summarized on the right. (H) Expression of IFNγ among CD4+T cells and CD8+T cells in tumor and dLN in mice. A representative dot plot is shown on the left, and the data are summarized on the right. Data are representative of three individual experiments (n = 4‐10). * p<0.001, compared with the indicated groups. n.s., not significant.

IEC SIRT3 deficiency promotes TH1 and CTL cell differentiation and function in aggravating colitis. Wild‐type (WT) and Sirt3 ∆Epi mice were treated with 1.0 % DSS for 14 days. (A) A typical image of colon (upper panel) and percent of mouse weight loss (lower panel). (B) Percent of Epcam1+ IEC and CD45+ lymphocyte cells in the colon, and a representative dot plot is shown. (C) Expressions of MHCII and CD54 among Epcam1+ IECs in the colon. A representative dot plot is shown on the left, and a data summary on the right. (D) Expressions of IL‐1α, IL‐1β, IL‐12, and IL‐23 among Epcam1+ IECs in the colon are summarized. (E,F) Percent of CD44highCD62L− cells among CD4+T cells (E) and CD8+T cells (F) in colon, dLN, and MLN in WT and Sirt3 ∆Epi mice. A representative dot plot is shown on the left, and data are summarized on the right. (G,H) Expression of IFNγ among CD4+T cells (G) and CD8+T cells (H) in colon, dLN, and MLN in mice. A representative dot plot is shown on the left, and the data are summarized on the right. Data are representative of three individual experiments (n = 4‐10). * p<0.05, p<0.01 and * p<0.001, compared with the indicated groups. n.s., not significant.

SIRT3 inhibits IEC function and TH1 and CTL differentiation. (A–F) IECs isolated from wild‐type (WT) mouse colon and transduced with control AAV9 (Ctrl) or SIRT3‐expressing AAV9 (Sirt3 OE). (A) Western blot of SIRT3 in the eGFP+ populations from the IECs transduced with Ctrl or Sirt3 OE. Representative data are shown on the left, and data are summarized on the right. NAD+ level (B) and IL‐1β expressions (C) in IECs transduced with Ctrl or Sirt3 OE and stimulated for 6 h with LPS (10 ng/mL, Sigma). (D–F) eGFP+ populations from the IECs transduced with Ctrl or Sirt3 OE were cocultured with T cells isolated from colon lamina propria (D). (E) Mean fluorescence intensity (MFI) of IL‐1R1 among CD4+T and CD8+T cells. (F) Percent of IFNγ+ cells among CD4+T cells and CD8+T cells. A representative dot plot is shown on the left, and data are summarized on the right. (G–I) SIRT3 expressions (G), NAD+ level (H), IL‐1β expressions (I) in IECs isolated from mouse colon and stimulated for 6 h with the LPS (10 ng/mL, Sigma) in the presence of Gardenia Yellow (1–10 nM, Selleck). (J,K) Epithelial cells isolated from the colon of WT mice. IECs were cocultured with T cells isolated from colon lamina propria in the presence of Gardenia Yellow (1–10 mM, Selleck) for 3 days. (J) MFI of IL‐1R1 among CD4+T and CD8+T cells. (K) Percent of IFNγ+ cells among CD4+T cells and CD8+T cells. A representative dot plot is shown on the left, and data are summarized on the right. Data are representative of three individual experiments (n = 4). p<0.01 and * p<0.001, compared with the control groups (without treatment with Gardenia Yellow).

Quinolinic acid constitutes a NAD+ salvage pathway involved in regulating IEC function and T cell differentiation induced by SIRT3 deficiency in colorectal cancer. (A–E) Wild‐type (WT) and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer. (A) Quinolinic acid (QA) and the NAD+ salvage pathway. (B) Changes in the NAD+ level in IECs isolated from the mouse colon. (C) Mass spectrometry analysis of the metabolite abundance of QA, nicotinic acid (NA), NAM, and NMN in IEC in the colon. (D) Changes in the NAD+ level in IECs isolated from the mouse colon and stimulated for 6 h with the LPS (10 ng/mL, Sigma) in the presence of FK866 (5 nM, Selleck), QA (1–10 mM, Selleck), or NA (5–10 mM, Selleck). (E) Intracellular IL‐1β+ cells among Epcam1+ IEC cells in the colon of WT and Sirt3 ∆Epi mice were stimulated for 6 h with LPS (10 ng/mL) in the presence of FK866 (5 nM, Selleck), QA (1–10 mM, Selleck), or NA (5–10 mM, Selleck). (F,G) Epithelial cells isolated from the colon of WT and Sirt3 ∆Epi mice. IECs were cocultured with CD8+T cells isolated from colon lamina propria in the presence of FK866 (5 nM, Selleck), QA (1–10 mM, Selleck), or NA (5–10 mM, Selleck) for 3 days (F). (G) Percent of IFNγ+ cells among CD8+T cells. Data are representative of three individual experiments (n = 3‐4). p<0.01 and * p<0.001, compared with the indicated groups. n.s., not significant.

Gut microbiota helps IECs use QA for NAD+ synthesis induced by SIRT3 deficiency in colorectal cancer. (A–C) Wild‐type (WT) and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer. (A) 3‐HA and QA in the NAD+ synthesis pathway. (B) mRNA expression of QPRT and 3‐HAO in IECs isolated from the colon. (C) A single‐cell suspension was prepared from the colon, and feces were collected. Mass spectrometry analysis of the metabolite abundance of 3‐HA in cells from colon and gut microbes. (D–G) WT and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer, and an antibiotic injection was administered once every two days. (D) A cartoon picture. (E) NAD+ relative level of IECs isolated from the colon. (F) Summary of tumor volume. (G) Percent of IL‐1β among IECs in the colon. (H–J) WT and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer. Gut microbes were collected and cultured to get the supernatant. IECs were isolated from mice and stimulated with LPS (10 ng/mL) in the presence of vehicle (PBS), microbe culture supernatant (microbe sup.), and QA (5 mM, Selleck) for 5 days. (H) A cartoon picture. Ratio of the NAD+ level in Sirt3 ∆Epi and WT mice (I) and the ratio of IL‐1β+ cells in IECs from Sirt3 ∆Epi and WT mice (J). Data are representative of three individual experiments (n = 4‐10). p<0.01 and * p<0.001, compared with the indicated groups.

IEC SIRT3 deficiency regulates T cell differentiation and function through IL‐1R1 signaling pathway in colorectal cancer. (A–C) Epithelial cells isolated from the colon of wild‐type (WT) and Sirt3 ∆Epi mice were cocultured with CD4+T cells or CD8+T cells isolated from colon lamina propria for 3 days in the presence of IL‐1R1i (anakinra, 500 ng/mL, MCE) or vehicle (Veh). (A) Experimental mode diagram. (B) IL‐1R1 expression among T cells in the colon. (C) Expressions of IFNγ among CD4+T cells (left) or CD8+T cells (right). (D–G) WT and Sirt3 ∆Epi mice were induced with orthotopic rectal cancer, as described above. IL‐1R inhibitor (anakinra, 15 µg in 15 µL) or vehicle was locally injected on Days 0 and 7. (D) Experimental mode diagram. (E) Tumor volume. (F) Immunofluorescence analysis of colorectal cancer samples. Scale bars, 50 µm. Original magnification, 630X. (G) Expressions of IFNγ among CD8+T cells in the tumor. A representative dot plot is shown on the left, and the data are summarized on the right. Data are representative of three individual experiments (n = 3–5). * p<0.001, compared with the indicated groups. n.s.

Full text

Introduction

1

Intestinal epithelial cells (IECs) are crucial mediators of intestinal homeostasis that enable the establishment of an immunological environment that allows commensal bacterial colonization and prohibits the growth of pathogens and tumors [1, 2, 3]. In addition to the immune regulatory role of circulating immune cells, local intestinal immune cells also play a crucial role in immune regulation under physiological and pathological conditions [4]. Following priming by intestine‐derived antigen‐presenting cells in secondary lymphoid tissues, conventional effector T cells, including CD4^+^T cells and CD8^+^T cells, recirculate through the body before settling in the intestine, where they exert their tolerogenic or inflammatory effect on the local environment [5]. Mature T cells are subject to the direct influence of IECs for their functional maintenance and survival in the lamina propria and in the intestines [6]. Recent studies have advanced the understanding of the development of these cells and the functions that they have at the intestinal barrier. IEC‐derived thymic stromal lymphopoietin (TSLP), transforming growth factor‐beta (TGFβ), and retinoic acid, interleukin‐1 (IL‐1) β, IL‐6, IL‐18, IL‐23, and IL‐33 produced in response to foreign bacterial or other antigen signals, promote the development of dendritic cells (DCs), neutrophils, and T cell differentiation and function [7, 8, 9, 10]. Although IEC and IEC‐derived signals have played important roles in promoting the differentiation and functional regulation of immune cells, including T cells, the regulatory mechanism is still unclear.

Sirtuin 3 (SIRT3), a nicotinamide adenine dinucleotide (NAD^+^)‐dependent deacetylase, plays a crucial role in regulating cell aging, cell proliferation and differentiation, tissue metabolism, and organ function [11]. SIRT3 is widely expressed in various cells, including epithelia cells or immune cells, allowing it to participate in regulating metabolic activities in the liver, skeletal muscle, adipose tissue, and intestine, especially in response to energy metabolism alterations caused by dietary changes [12]. Research has shown that changes in SIRT3 expression are associated with the progression of diseases, including colorectal cancer [13]. It is believed that SIRT3 regulates mitochondrial energy metabolism, thereby maintaining metabolic homeostasis and regulating the proliferation and differentiation of cancer cells [13]. But the cell‐specific regulatory role of SIRT3 in inflammation and cancer remains unclear. Recently, our results showed that SIRT3 suppresses the functions of T follicular helper cells and promotes tumor growth [14]. However, it remains unclear whether SIRT3 signaling from IECs regulates the differentiation and function of local T cells in colitis and colorectal cancer.

Herein, we showed that SIRT3 is a crucial negative determinant of local IFNγ‐producing CD4^+^T cells (T_H_1) and cytotoxic T lymphocytes (CTLs) differentiation and functions in colorectal cancer and colitis. IEC SIRT3 deficiency promotes the production of proinflammatory cytokine IL‐1β and the local T_H_1 and CTL differentiation and function in limiting colorectal cancer growth and aggravating colitis. Importantly, NAD^+^‐dependent enzyme SIRT3 deficiency promotes the functions of IECs through quinolinic acid (QA)‐mediated NAD^+^ synthesis in limiting tumor growth. Microbiota‐derived 3‐hydroxyaminobenzoic acid (3‐HA), which is a source of intracellular QA in IECs. Thus, these findings identify a previously unrecognized QPRT‐mediated switch in NAD^+^ metabolism by exploiting microbiota‐derived QA as an alternative source of replenishing intracellular NAD^+^ pools induced by SIRT3 deficiency to regulate IEC and T cell function in colorectal cancer and colitis. This provides a new approach for targeting IECs and gut microbiota for NAD^+^ metabolic remodeling intervention in immune therapy for colorectal cancer and colitis.

Results

2

IEC SIRT3 Expression is Related to Local TH1 and CTL Function in Colorectal Cancer and Colitis

2.1

IEC metabolism modulation is crucial for the occurrence and development of colitis and colorectal cancer [15]. Focusing on the changes in carbon metabolism of IECs, we first analyzed the different gene expression of IECs in colitis and colorectal cancer with a ScRNA‐seq dataset (GSE261388). As expected, we found that the well‐known metabolism regulators, including glycolytic metabolic enzymes, crucial metabolism signaling molecules, and oncogenes, are in line with their crucial immune regulatory functions [16, 17]. Notably, SIRT3, an NAD^+^‐dependent metabolic regulatory molecule located on mitochondria, expression in IEC was higher in colitis and colorectal cancer samples than in normal health samples, suggesting its important roles in human tumor progression (Figure 1A,B). Consistently, inflammatory stimuli enhanced SIRT3 expression in IECs in mice in vitro with dose dependent manner (Figure S1A). Moreover, we set up dextran sulfate sodium (DSS)‐induced colitis and colorectal orthotopic tumor mouse model, as described previously [16, 18]. IECs showed higher SIRT3 expression in colitis and colorectal cancer in mice than in normal control mice (Figure S1B). Additionally, data from the TCGA database also showed that IEC SIRT3 expression is negatively associated with the survival of human colon adenocarcinoma (Figure 1C). And, IEC SIRT3 expressions are related to the T_H_1 and CTL infiltration, but not T_reg_ infiltration, in human colon adenocarcinoma patients (Figure S2A), which suggests that IEC SIRT3 expressions are critically involved in regulating T_H_1 and CTL functions in human colitis and colorectal cancer.

SIRT3 signaling from IECs is associated with TH1 and CTL function in colorectal cancer. (A,B) ScRNA‐seq datasets were retrieved from GSE261388. (A) Gene expression analysis of IEC clusters in colitis and colorectal cancer using the “Find All Markers” function. (B) Violin plots of IEC SIRT3 expression in colitis and colorectal cancer. (C) Survival analysis of human colon adenocarcinoma associated with IEC SIRT3 expression based on samples from the TCGA database. Analysis was performed using GEPIA 2 (http://gepia2.cancer‐pku.cn) online. (D–G) C57BL/6 mice were treated with 1.5 % dextran sulfate sodium salt (DSS) for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer, and Sirt3 expression was examined by qPCR in sorted Epcam1+ IEC cells from colon, and the percent of IFNγ+ TH1 and CTL cells was examined by flow cytometry at different time points. (D) The mRNA expression levels of SIRT3 in mouse IECs under different physiological conditions. Correlations between the percent of TH1 (E) and CTL cells (F) and Sirt3 mRNA levels. Correlations between tumor size and the Sirt3 mRNA level (G). The data are representative of three independent experiments (C, n = 3‐5; D‐H; n = 10 mice per group). * p<0.05 and *** p<0.001, compared with the indicated groups.

To determine whether SIRT3 is related to T cell differentiation and tumor growth, we used DSS‐induced colitis and colorectal orthotopic tumor mouse model and found that SIRT3 is gradually upregulated in the IEC cells isolated from colitis and colorectal cancer (Figure 1D). IEC does not relate to the percent of CD11b^+^Ly6G^+^ neutrophils and CD11c^+^ DC infiltration in the tumor (Figure S2B,C). Specifically, IEC SIRT3 mRNA expression is negatively related to T_H_1 and CTL infiltration percentage and positively related to tumor volume (Figure 1E–G). Collectively, these data suggest that IEC SIRT3 is likely involved in regulating T_H_1 and CTL differentiation and function in colorectal cancer and colitis.

IEC SIRT3 Deficiency Does not Affect Local T Cell Differentiation and Function Under Physiological Conditions

2.2

To investigate the effect of SIRT3 in IECs on T cell function, we crossed Sirt3 ^flox/flox^ and Villin‐Cre mice to conditionally delete Sirt3 gene expression in IECs (called Sirt3 ^∆Epi^ mice hereafter). IEC SIRT3 deficiency does not affect the immune cell infiltration in the colon (Figure S3A). In Sirt3 ^∆Epi^ mice, there were no significant effects on the percent of Epcam1^+^ IECs and CD45^+^ immune cells, expressions of MHCII and CD54, and productions of pro‐inflammatory cytokine IL‐1β and IL‐23 in IECs (Figure S3B,D). NAD^+^ level is important for the activity of the NAD^+^‐dependent enzyme SIRT3 [19]. But Sirt3 ^∆Epi^ does not alter the level of NAD^+^ in IECs (Figure S3E). Also, Sirt3 ^∆Epi^ does not alter the percent of CD44^high^CD62L^−^ T cells (Figure S4A,B) and the production of IFNγ and TNFα in the CD4^+^T and CD8^+^T cells (Figure S5A–D). These data suggest that IEC SIRT3 deficiency has no significant effects on immune cell infiltration and function under physiological conditions.

IEC SIRT3 Deficiency Promotes Local TH1 and CTL Differentiation and Function in Limiting Colorectal Cancer Growth

2.3

To directly determine the role of SIRT3 in IECs in colorectal cancer, we established an orthotopic mouse colon cancer model via in situ transplantation of colon cancer cells, as described previously [20]. Sirt3 ^∆Epi^ presented smaller tumors (Figure 2A) and more immune cell infiltration into tumors (Figure 2B). Specifically, Sirt3 ^∆Epi^ promoted the expressions of MHCII and CD54 and productions of IL‐1β, but not IL‐1α, IL‐12, IL‐23 in IECs (Figure 2C,D). Importantly, Sirt3 ^∆Epi^ enhanced the percent of CD44^high^CD62L^−^ T cells and production of IFNγ of CD4^+^T (T_H_1) and CD8^+^T cells (CTL), but not IL‐4 expression (T_H_2), IL‐17A expression (T_H_17), and Foxp3 expression (T_reg_) in CD4^+^T cells (Figure 2E–H). These suggest that IEC SIRT3 negatively regulates T_H_1 and CTL functions in colorectal cancer. Furthermore, functions of tumor‐specific CTL were studied, as described previously [4]. MC38‐OVA colon cancer cells were injected into the rectal mucosa to induce rectal cancer, and tumor‐specific CTL were stained by OVA‐specific TCR tetramer. Sirt3 ^∆Epi^ promoted the IFNγ and TNFα production in OVA‐tetramer^+^CD8^+^T cells (Figure S6A,B). These suggest that IEC SIRT3 is critical for the functions of tumor‐specific CTL.

IEC SIRT3 deficiency promotes TH1 and CTL cell function in inhibiting colorectal cancer growth. Wild‐type (WT) and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer. (A) A typical image of colorectal cancer (left) and a summary of tumor volume (right). (B) HE staining of colorectal cancer samples. Scale bars, 50 µm. Original magnification, 50X. (C) Expressions of MHCII and CD54 among Epcam1+ IECs in the tumor. (D) Percent of IL‐1α, IL‐1β, IL‐12, and IL‐23 among Epcam1+ IECs in the tumor. (E,F) Percent of CD44highCD62L− cells among CD4+T cells (E; left) and CD8+T cells (E; right) in tumor and draining lymph node (dLN) in WT and Sirt3 ∆Epi mice. A representative dot plot is shown (E), and data are summarized (F). (G) Expression of IL‐4, IL‐17A, and Foxp3 among CD4+T cells in the tumor in mice. A representative dot plot is shown on the left, and data are summarized on the right. (H) Expression of IFNγ among CD4+T cells and CD8+T cells in tumor and dLN in mice. A representative dot plot is shown on the left, and the data are summarized on the right. Data are representative of three individual experiments (n = 4‐10). *** p<0.001, compared with the indicated groups. n.s., not significant.

IEC SIRT3 Deficiency Promotes Local TH1 and CTL Differentiation and Function in Aggravating Colitis

2.4

To directly determine the role of SIRT3 in IECs in intestinal inflammation, we established a DSS‐induced mouse colitis model, as described previously [16]. Sirt3 ^∆Epi^ presented shorter colon length and more weight loss (Figure 3A). Although there is no significant alteration in the percent of Epcam1^+^ IEC cells and CD45^+^ lymphocyte cells, Sirt3 ^∆Epi^ promoted the expressions of MHCII and CD54 and productions of IL‐1β, but not IL‐1α, IL‐12, and IL‐23 in IECs (Figure 3B–D). Importantly, Sirt3 ^∆Epi^ enhanced the percent of CD44^high^CD62L^−^ T cells and production of T_H_1 and CTL, but not IL‐4 expression (T_H_2), IL‐17A expression (T_H_17), and Foxp3 expression (T_reg_) in CD4^+^T cells in colon, dLN, and MLN (Figure 3E–H; Figure S7A–C). These suggest that IEC SIRT3 negatively regulates T_H_1 and CTL functions in colitis.

IEC SIRT3 deficiency promotes TH1 and CTL cell differentiation and function in aggravating colitis. Wild‐type (WT) and Sirt3 ∆Epi mice were treated with 1.0 % DSS for 14 days. (A) A typical image of colon (upper panel) and percent of mouse weight loss (lower panel). (B) Percent of Epcam1+ IEC and CD45+ lymphocyte cells in the colon, and a representative dot plot is shown. (C) Expressions of MHCII and CD54 among Epcam1+ IECs in the colon. A representative dot plot is shown on the left, and a data summary on the right. (D) Expressions of IL‐1α, IL‐1β, IL‐12, and IL‐23 among Epcam1+ IECs in the colon are summarized. (E,F) Percent of CD44highCD62L− cells among CD4+T cells (E) and CD8+T cells (F) in colon, dLN, and MLN in WT and Sirt3 ∆Epi mice. A representative dot plot is shown on the left, and data are summarized on the right. (G,H) Expression of IFNγ among CD4+T cells (G) and CD8+T cells (H) in colon, dLN, and MLN in mice. A representative dot plot is shown on the left, and the data are summarized on the right. Data are representative of three individual experiments (n = 4‐10). * p<0.05, ** p<0.01 and *** p<0.001, compared with the indicated groups. n.s., not significant.

Furthermore, the role of SIRT3 in IECs was investigated in a bacterial infection model. We set up Citrobacter rodentium infected intestinal inflammation mouse model, as described previously [21]. Consistently, Sirt3 ^∆Epi^ presented shorter colon length, more weight changes, and less bacterial survival (Figure S8A–C). These suggested that SIRT3 deficiency in IECs limited the bacterial survival and showed anti‐bacterial infection effects in mice. As expected, Sirt3 ^∆Epi^ promoted the expressions of MHCII and CD54 and productions of IL‐1β, but not IL‐23 in IECs (Figure S8D–F). Importantly, Sirt3 ^∆Epi^ enhanced the functions of T_H_1 and CTL, but not IL‐4 expression (T_H_2), IL‐17A expression (T_H_17), and Foxp3 expression (T_reg_) in CD4^+^T cells in ileum, peyer's patch, and colon (Figure S9A–E). These suggest that IEC SIRT3 negatively regulates T_H_1 and CTL functions in bacterial infectious colitis. Additionally, functions of infection‐specific CTL were studied, as described previously [4]. Bacterial‐specific CTL were stained by OVA‐specific TCR tetramer in CR‐OVA‐infected mice. Sirt3 ^∆Epi^ promoted the IFNγ production in OVA‐tetramer^+^CD8^+^T cells (Figure S9F). These suggest that IEC SIRT3 is required for the functions of bacterial infection‐specific CTL cells in colitis.

SIRT3 Inhibits IEC Function and TH1 and CTL Differentiation

2.5

To directly clarify the role of SIRT3 in IECs and T cell function, we ectopically expressed SIRT3 with an adeno‐associated virus (AAV) 9 and sorted the positive green fluorescent protein (GFP)^+^ IECs. Modest SIRT3 overexpression (Sirt3 ^OE^) resulted in a decrease in NAD^+^ level and IL‐1β secretion in IECs (Figure 4A–C). Additionally, IECs with Sirt3 ^OE^ cocultured with T cells, which reduced the expression of IL‐1R1 in T cells and inhibited the functions of T_H_1 and CTL cells (Figure 4D–F).

SIRT3 inhibits IEC function and TH1 and CTL differentiation. (A–F) IECs isolated from wild‐type (WT) mouse colon and transduced with control AAV9 (Ctrl) or SIRT3‐expressing AAV9 (Sirt3 OE). (A) Western blot of SIRT3 in the eGFP+ populations from the IECs transduced with Ctrl or Sirt3 OE. Representative data are shown on the left, and data are summarized on the right. NAD+ level (B) and IL‐1β expressions (C) in IECs transduced with Ctrl or Sirt3 OE and stimulated for 6 h with LPS (10 ng/mL, Sigma). (D–F) eGFP+ populations from the IECs transduced with Ctrl or Sirt3 OE were cocultured with T cells isolated from colon lamina propria (D). (E) Mean fluorescence intensity (MFI) of IL‐1R1 among CD4+T and CD8+T cells. (F) Percent of IFNγ+ cells among CD4+T cells and CD8+T cells. A representative dot plot is shown on the left, and data are summarized on the right. (G–I) SIRT3 expressions (G), NAD+ level (H), IL‐1β expressions (I) in IECs isolated from mouse colon and stimulated for 6 h with the LPS (10 ng/mL, Sigma) in the presence of Gardenia Yellow (1–10 nM, Selleck). (J,K) Epithelial cells isolated from the colon of WT mice. IECs were cocultured with T cells isolated from colon lamina propria in the presence of Gardenia Yellow (1–10 mM, Selleck) for 3 days. (J) MFI of IL‐1R1 among CD4+T and CD8+T cells. (K) Percent of IFNγ+ cells among CD4+T cells and CD8+T cells. A representative dot plot is shown on the left, and data are summarized on the right. Data are representative of three individual experiments (n = 4). ** p<0.01 and *** p<0.001, compared with the control groups (without treatment with Gardenia Yellow).

Next, we tested the application of a pharmacological approach to target SIRT3 in mouse IEC cells and determined whether our findings concerning the genetic target SIRT3 could be recapitulated. The pharmacological SIRT3 agonist gardenia yellow treatment enhanced the SIRT3 expression and reduced the NAD^+^ level and IL‐1β secretion in a dose‐dependent manner (Figure 4G–I). Importantly, IECs treated with SIRT3 agonist, and cocultured with T cells, reduced the expression of IL‐1R1 in T cells and suppressed the functions of T_H_1 and CTL cells (Figure 4J–K). Altogether, these data indicate SIRT3 is sufficient for regulating IECs function and T cell differentiation.

QA Constitutes a NAD+ Salvage Pathway Involved in Regulating IEC Function Induced by Sirt3

∆Epi in Colorectal Cancer

2.6

NAD^+^ level and the NAD^+^ synthesis pathway are important for the activity of the NAD^+^‐dependent enzyme SIRT3 [13, 22, 23]. As shown, Modest SIRT3 overexpression or SIRT3 agonist treatment enhanced SIRT3 expression and reduced NAD^+^ level (Figure 4A,B,G,H). We further investigated the NAD^+^ level in the IECs in colorectal cancer in mice. SIRT3 deficiency in IECs enhanced NAD^+^ levels (Figure 5A,B). Specifically, SIRT3 deficiency led to less metabolite QA and more nicotinamide mononucleotide (NMN) but not nicotinic acid (NA) or nicotinamide (NAM), in IECs (Figure 5C). As reported [23], NAD^+^ promotes IEC proinflammatory factor secretion in ulcerative colitis. Consistently, as a NAD^+^ precursor, NMN treatment enhanced NAD^+^ levels and promoted IL‐1β production in IECs (Figure S10A,B). These data reveal that the NAD^+^ level and the NAD^+^ salvage synthesis pathway are likely involved in the regulation of IECs function induced by Sirt3 ^∆Epi^ in colorectal cancer.

Quinolinic acid constitutes a NAD+ salvage pathway involved in regulating IEC function and T cell differentiation induced by SIRT3 deficiency in colorectal cancer. (A–E) Wild‐type (WT) and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer. (A) Quinolinic acid (QA) and the NAD+ salvage pathway. (B) Changes in the NAD+ level in IECs isolated from the mouse colon. (C) Mass spectrometry analysis of the metabolite abundance of QA, nicotinic acid (NA), NAM, and NMN in IEC in the colon. (D) Changes in the NAD+ level in IECs isolated from the mouse colon and stimulated for 6 h with the LPS (10 ng/mL, Sigma) in the presence of FK866 (5 nM, Selleck), QA (1–10 mM, Selleck), or NA (5–10 mM, Selleck). (E) Intracellular IL‐1β+ cells among Epcam1+ IEC cells in the colon of WT and Sirt3 ∆Epi mice were stimulated for 6 h with LPS (10 ng/mL) in the presence of FK866 (5 nM, Selleck), QA (1–10 mM, Selleck), or NA (5–10 mM, Selleck). (F,G) Epithelial cells isolated from the colon of WT and Sirt3 ∆Epi mice. IECs were cocultured with CD8+T cells isolated from colon lamina propria in the presence of FK866 (5 nM, Selleck), QA (1–10 mM, Selleck), or NA (5–10 mM, Selleck) for 3 days (F). (G) Percent of IFNγ+ cells among CD8+T cells. Data are representative of three individual experiments (n = 3‐4). ** p<0.01 and *** p<0.001, compared with the indicated groups. n.s., not significant.

As QA may serve as a precursor for NAD^+^, which in turn may direct a portion of tryptophan (Trp) catabolism toward replenishing NAD^+^ levels in response to inflammation [24]. We next assessed the functional consequences of QA for NAD^+^‐mediated immune responses in IECs induced by Sirt3 ^∆Epi^. Blocking de novo NAD^+^ synthesis with the nicotinamide phosphoribosyl transferase (NAMPT) inhibitor FK866 reduced NAD^+^ levels in IEC and IL‐1β production in IECs induced by Sirt3 ^∆Epi^ (Figure 5D,E), which indicated that NAMPT‐mediated NAD^+^ synthesis is required for IEC functions induced by Sirt3 ^∆Epi^. Interestingly, QA, but not NA, restored NAD^+^ levels and the production of IL‐1β via NAMPT inhibition in IECs (Figure 5D,E). These suggest that QA‐fed, but not the NA‐fed salvage pathway, is required for the NAD^+^ synthesis and IEC function induced by Sirt3 ^∆Epi^. Furthermore, the effects of IECs on T cells have also been studied. We set up a co‐culture system between IECs and CD8^+^T cells isolated from the colon in colorectal cancer, as shown in Figure 5F. Although Sirt3 ^∆Epi^ IEC induced more IFNγ expression in CD8^+^T cells, blocking de novo NAD^+^ synthesis with FK866 reduced the expression of IFNγ in CD8^+^T cells. Consistently, QA, but not NA, restored the expression of IFNγ in CD8^+^T cells (Figure 5G). Taken together, these data suggest that IECs use QA to replenish NAD^+^ stocks and regulate the IEC and T cell function induced by Sirt3 ^∆Epi^. Thus, these data reveal that the QA‐fed but not the NA‐fed NAD^+^ salvage pathway is activated in Sirt3 ^∆Epi^ IECs in colorectal cancer. A switch from the NA salvage pathway to the QA salvage pathway to synthesize NAD^+^ is crucial for regulating the functions of IECs and the differentiation of T cells induced by Sirt3 ^∆Epi^ in colorectal cancer.

3‐HA are Produced in Gut Microbes to Help IECs Use QA for NAD+ Synthesis Induced by Sirt3

∆Epi in Colorectal Cancer

2.7

3‐HAO and QPRT are key enzyme that catalyzes the synthesis of QA from 3‐HA [25]. Sirt3 ^∆Epi^ led to more expressions of QPRT and 3‐HAO in colorectal cancer (Figure 6A,B). These suggest that 3‐HAO and QPRT are crucial for NAD^+^ synthesis induced by Sirt3 ^∆Epi^ in colorectal cancer. As we all know, QA is generated through QPRT and 3‐HAO utilizing 3‐HA [24]. To determine whether IECs can produce 3‐HA, we isolated cells from the colon in colorectal cancer, as well as gut microbes from feces, and detected changes in the metabolite 3‐HA. Importantly, Sirt3 ^∆Epi^ led to less 3‐HA in cells and more 3‐HA in gut microbes in mice with colorectal cancer (Figure 6C). These data suggest that 3‐HA produced in gut microbes probably helps IECs use QA for NAD^+^ synthesis induced by Sirt3 ^∆Epi^ in colorectal cancer.

Gut microbiota helps IECs use QA for NAD+ synthesis induced by SIRT3 deficiency in colorectal cancer. (A–C) Wild‐type (WT) and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer. (A) 3‐HA and QA in the NAD+ synthesis pathway. (B) mRNA expression of QPRT and 3‐HAO in IECs isolated from the colon. (C) A single‐cell suspension was prepared from the colon, and feces were collected. Mass spectrometry analysis of the metabolite abundance of 3‐HA in cells from colon and gut microbes. (D–G) WT and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer, and an antibiotic injection was administered once every two days. (D) A cartoon picture. (E) NAD+ relative level of IECs isolated from the colon. (F) Summary of tumor volume. (G) Percent of IL‐1β among IECs in the colon. (H–J) WT and Sirt3 ∆Epi mice were treated with 1.5 % DSS for 5 days beginning on Day 0, and then MC38 colon cancer cells were injected on Day 8 under the rectal mucosa to induce orthotopic rectal cancer. Gut microbes were collected and cultured to get the supernatant. IECs were isolated from mice and stimulated with LPS (10 ng/mL) in the presence of vehicle (PBS), microbe culture supernatant (microbe sup.), and QA (5 mM, Selleck) for 5 days. (H) A cartoon picture. Ratio of the NAD+ level in Sirt3 ∆Epi and WT mice (I) and the ratio of IL‐1β+ cells in IECs from Sirt3 ∆Epi and WT mice (J). Data are representative of three individual experiments (n = 4‐10). ** p<0.01 and *** p<0.001, compared with the indicated groups.

To test this hypothesis, we treated WT and Sirt3 ^∆Epi^ mice daily with antibiotics during the induction of colorectal cancer, and the gut microbiota of mice could be effectively eliminated (Figure 6D). We found that although Sirt3 ^∆Epi^ led to higher NAD^+^ levels, smaller tumor volume, and higher IL‐1β production, elimination of gut microbiota almost restored these alterations to control levels (Figure 6E–G). These data suggest that gut microbiota‐derived metabolites, probably 3‐HA, are responsible for the functions of IECs induced by Sirt3 ^∆Epi^ in colorectal cancer.

To confirm this point, we isolated IECs from WT and Sirt3 ^∆Epi^ colorectal cancer mice and stimulated these IECs with LPS in the presence of microbe culture supernatant (microbe sup.) or QA in an in vitro culture system, as shown in Figure 6H. Sirt3 ^∆Epi^ enhanced the NAD^+^ level, but the increased values gradually decreased. Interestingly, IECs treated with microbe sup. or QA almost restored these alterations to the control level (Figure 6I,J). These data suggested that 3‐HA may be one of the components of microbiota‐derived metabolites that are required for NAD^+^ synthesis and IEC functions induced by Sirt3 ^∆Epi^ in colorectal cancer. Taken together, these data reveal that 3‐HA produced by microbiota helps IECs use QA to replenish the NAD^+^ stock and regulate IEC functions induced by Sirt3 ^∆Epi^ in colorectal cancer.

IL‐1β Is Required for Sirt3

∆Epi IEC‐Directed TH1 and CTL Function in Colitis and Colorectal Cancer

2.8

SIRT3 deficiency enhanced IL‐1β production, but not IL‐1α, IL‐12, and IL‐23 in colitis and colorectal cancer, which prompted us to investigate the role of IL‐1β in IECs‐directed T cell functions. We observed these effects in MC38 colorectal orthotopic tumor mice. Tumors were locally injected with an anti‐IL‐1β or IgG control antibody, as shown in Figure S11A. Although Sirt3 ^∆Epi^ delayed tumor growth, enhanced IL‐1β production, IL‐1R1 expressions, and T cell function and infiltration into the tumor, blocking IL‐1β with anti‐IL‐1β generally reversed these alterations to the control level (Figure S11B–F). Consistently, we observed these effects in colitis mice. Mice were treated with anti‐IL‐1β or IgG control (Figure S12A). Although Sirt3 ^∆Epi^ led to more weight loss, higher IL‐1β production in IECs, and more T_H_1 and CTL differentiation, blocking IL‐1β with anti‐IL‐1β generally reversed these alterations to the control level (Figure S12B–F). Collectively, these data suggest that IL‐1β signaling is required for T cell function and differentiation in colitis and colorectal cancer induced by Sirt3 ^∆Epi^.

IEC SIRT3 Promotes Local TH1 and CTL Function Through IL‐1R1 Signaling Pathway in Colorectal Cancer

2.9

T‐cell polarizing cytokines often induce the expression of their corresponding cytokine receptor on T cells, resulting in robust programming of cell fate determination [26, 27]. The IL‐1β‐IL‐1 receptor (IL‐1R) 1 signaling pathway is critical for regulating IEC function in inflammation [28, 29, 30, 31, 32]. IL‐1β upregulated the IL‐1R1 expression in IECs and IFNγ production in T cells (Figure S13A,B). IEC Sirt3 ^OE^ or SIRT3 agonist gardenia yellow treatment reduced the IL‐1R1 expression in T cells (Figure 4E,J). Additionally, Sirt3 ^∆Epi^ led to more expressions of IL‐1R1 in CD4^+^T cells and CD8^+^T cells in colorectal cancer in mice (Figure S14A,B). These further suggest IL‐1R1 signaling is probably involved in regulating T cell functions induced by Sirt3 ^∆Epi^ IECs. To check the role of IL‐1R1 signaling in regulating T cell functions induced by IECs. We set up a coculture system between IECs and T cells in vitro, as shown in Figure 7A. CD4^+^T or CD8^+^T cells isolated from the colon lamina propria of mice and pretreated with or without IL‐1R1 inhibitor (anakinra, 500 ng/mL), as described previously [33, 34]. IECs isolated from WT and Sirt3 ^∆Epi^ mice and cocultured with T cells. Although Sirt3 ^∆Epi^ enhanced the IFNγ production in CD4^+^T cells and CD8^+^T cells, IL‐1R1 inhibitor treatment almost completely recovered these alterations to the WT control level (Figure 7B,C). Furthermore, we observed these effects in MC38 colorectal orthotopic tumor mice. Tumors were locally injected with an IL‐1R1 inhibitor (Figure 7D), as described previously [35]. Although Sirt3 ^∆Epi^ delayed tumor growth and promoted T cell function and infiltration into the tumor, blocking IL‐1R1 with an inhibitor generally reversed these alterations to the control level (Figure 7E–G). Collectively, these data suggest that Sirt3 ^∆Epi^ delays tumor growth and potentiates T cell function through the IL‐1β‐IL‐1R1 signaling pathway.

IEC SIRT3 deficiency regulates T cell differentiation and function through IL‐1R1 signaling pathway in colorectal cancer. (A–C) Epithelial cells isolated from the colon of wild‐type (WT) and Sirt3 ∆Epi mice were cocultured with CD4+T cells or CD8+T cells isolated from colon lamina propria for 3 days in the presence of IL‐1R1i (anakinra, 500 ng/mL, MCE) or vehicle (Veh). (A) Experimental mode diagram. (B) IL‐1R1 expression among T cells in the colon. (C) Expressions of IFNγ among CD4+T cells (left) or CD8+T cells (right). (D–G) WT and Sirt3 ∆Epi mice were induced with orthotopic rectal cancer, as described above. IL‐1R inhibitor (anakinra, 15 µg in 15 µL) or vehicle was locally injected on Days 0 and 7. (D) Experimental mode diagram. (E) Tumor volume. (F) Immunofluorescence analysis of colorectal cancer samples. Scale bars, 50 µm. Original magnification, 630X. (G) Expressions of IFNγ among CD8+T cells in the tumor. A representative dot plot is shown on the left, and the data are summarized on the right. Data are representative of three individual experiments (n = 3–5). *** p<0.001, compared with the indicated groups. n.s.

Discussion

3

IECs have the capacity to function as immune rheostats by employing sensory mechanisms that induce an immunomodulatory output in infection and cancer [21, 36]. In the present study, we found that NAD^+^ is required for the IEC function and antigen‐specific differentiation of T_H_1 and CTL cells induced by Sirt3 ^∆Epi^ in colitis and colorectal cancer. Moreover, QPRT‐mediated switch in NAD^+^ metabolism by exploiting microbiota‐derived QA as an alternative source of replenishing intracellular NAD^+^ pools induced by Sirt3 ^∆Epi^ IECs. Thus, our results define the importance of the SIRT3‐NAD^+^ synthesis pathways for inducing IL‐1β production in IECs and IL‐1R1 expression in local T_H_1 and CTL cells. These results have implications for the targeting of IECs and gut microbiota as an approach to the treatment of immune system disorders and immune‐associated diseases, including colorectal cancer and colitis (Figure S15).

IEC can reshape the immune microenvironment and enhance mucosal barrier capacity, affecting the outcomes of intestinal cancer and inflammation [6]. IL‐1 is considered a “master regulator” of inflammation. IL‐1α and IL‐1β bind the IL‐1R1 to activate the downstream target including NF‐κB pathways [28, 29, 30, 31, 32]. Epithelial NF‐kB regulates important barrier functions required for establishing tolerance and immunity, including maintenance of the mucus layer and antimicrobial defense [37]. IEC‐derived IL‐1 family member including IL‐18 and IL‐33, produced in response to foreign bacterial or antigen signals, promote the development of DCs, neutrophils, and T cell differentiation and function [7, 8, 9, 10, 38]. IL‐18 is constitutively expressed by IECs in the steady state that mediates type 1 immune responses. Constitutive secretion of IL‐18 from IECs inhibits T_H_17 differentiation and promotes T_reg_ cell function for limiting local inflammation in the colon [39]. Intestinal T_reg_s are further regulated by the epithelial alarmin IL‐33. Epithelial production of IL‐33 in response to inflammatory tissue damage promotes the function of colonic T_reg_ cells, which preferentially express the IL‐33 receptor subunit ST2, by enhancing T_reg_ cell maintenance and accumulation in the inflamed tissue [38]. In this study, our data showed that Sirt3 ^∆Epi^ promotes T_H_1 and CTL differentiation and function through the IL‐1β‐IL‐1R1 signaling pathway in colorectal cancer and colitis. It is worth noting that the DSS‐induced colitis and Citrobacter rodentium infected intestinal inflammation, as well as orthotopic mouse colon cancer, exhibited mainly T_H_1 and CTL changes, without affecting the differentiation and function of other T cell subsets, showing very similar immune microenvironment changes. These also provide similar immunological regulatory mechanisms for the SIRT3 signal from IECs to regulate T cell differentiation and function, and alter the outcomes of colitis and colorectal cancer diseases.

SIRT3 belongs NAD^+^‐dependent enzyme family, which regulates signaling pathways involved in cellular proliferation and differentiation, metabolism, response to stress, and cancer [11]. SIRT3, as the major member located in mitochondria, plays a crucial role in modulating mitochondrial metabolism [12]. Under stress conditions, SIRT3 translocates to the mitochondrial matrix, where it is activated by a protease, which yields the 28‐kDa active form of SIRT3 [22]. Recent studies show a possible nuclear localization of SIRT3 in the nucleus, where it may regulate gene expression in response to stress [40]. However, SIRT3 exhibits opposite effects in different types of tumors. Acts as a tumor promoter in breast cancer [41], colon cancer [42], gastric cancer [43], esophageal cancer [44], oral squamous cell carcinoma [45], melanoma [46], and renal cancer [47] or acts as a tumor suppressor in breast cancer [48], hepatocellular carcinoma [49], B cell malignancies and leukemia [50] and metastatic ovarian cancer [51]. These suggest the complex regulatory roles of SIRT3 in cancer, which may be related to the different expression of SIRT3 in specific types of cells in the tumor microenvironment. Under physiological conditions, the expression of SIRT3 in IECs is low, and inflammatory stimuli or inflammatory microenvironment significantly upregulated SIRT3 expression in colitis and colorectal cancer (Figure S1A,B). These indicate that changes in SIRT3 expression are the main cause of functional changes in IECs in inflammatory (colitis) and tumor environments (colorectal cancer). Recently, our data showed that SIRT3 suppresses the functions of T_FH_ cell differentiation and promotes tumor growth [14]. This study further demonstrates that Sirt3 ^∆Epi^ promotes IEC function and T_H_1 and CTL differentiation in colorectal cancer and inflammation.

The Trp‐to‐QA pathway has been shown to be critical in several psychiatric conditions and inflammation [52, 53]. However, the roles of the tryptophan metabolite QA and the QA‐NAD^+^ pathway in colitis and colorectal cancer have not been studied. SIRT3 is a crucial regulator of T cell‐mediated adaptive immunity, and NAD^+^ level sensitively regulate the expression of SIRT3 in T cells [14]. The present study further showed that NAD^+^‐dependent enzyme Sirt3 ^∆Epi^ promotes IEC functions through the NAD^+^ synthesis pathway. Sirt3 ^∆Epi^ regulates IEC function by altering the NAD^+^ salvage synthesis pathway for reshaping T_H_1 and CTL differentiation and function in colitis and colorectal cancer. Therefore, NAD^+^‐SIRT3‐NAD^+^ develops a negative feedback loop in regulating IEC function and T cell differentiation in colorectal cancer and colitis. More importantly, these data confirm that gut microbiota‐derived QA is critical for NAD^+^ synthesis of IECs. Thus, these data reveal the role of gut microbiota and IEC coordinated regulation in reshaping the immune microenvironment of colorectal cancer and colitis induced by Sirt3 ^∆Epi^. FK866, as a NAMPT inhibitor, increases tumor cell apoptosis by inhibiting the new NAD^+^ synthesis, and is currently being evaluated in phase II clinical trials for cancer patients [54]. Although the results of therapeutic NAD^+^ depletion are promising, the inhibitory effect of NAMPT may be influenced by the activity of QA‐ and NA‐dependent NAD^+^ synthesis pathways. QA and NA can both serve as alternative substrates for NAD^+^ synthesis. The present study data show that SIRT3 deficiency enhances the expression of 3‐HAO and QPRT, and utilizes microbiota‐derived QA as an alternative source of intracellular NAD^+^ pool supplementation to regulate IEC function in colitis and colorectal cancer. Additionally, In Sirt3 ^∆Epi^ mice, although the microbiota‐derived 3‐HA was abundant, the IECs showed a significant decrease in QA, indicating efficient utilization of QA within the IECs. These also suggest that exogenous supplementation of QA may be a more direct intervention approach in mice.

In summary, targeting NAD^+^‐dependent enzyme SIRT3, integrating microbiota‐derived QA to replenish NAD^+^ synthesis in IECs controls the production of IL‐1β, thereby IL‐1R1 in T cells contributes to the antigen‐specific differentiation of T_H_1 and CTL cells in colorectal cancer and colitis (Figure S15). Thus, our results define the importance of the SIRT3‐NAD^+^ synthesis pathway in IECs for inducing the differentiation of T_H_1 and CTL cells through the IL‐1β‐IL‐1R1 pathway. These results have implications for the targeting of IECs as an approach to the treatment of immune system disorders and immune‐associated diseases, including colorectal cancer and colitis.

Materials and Methods

4

Mice

4.1

All animal experiments were performed with the approval of the Animal Ethics Committee of Beijing Institute of Microbiology and Epidemiology and Beijing Normal University (IACUC‐DWZX‐2021‐003 and CLS‐EAW‐2021‐002). C57BL/6 Sirt3 ^fl/flx^ mice have been described previously [14, 55]. Villin‐Cre was obtained from GemPharmatech (Nanjing, China). Sirt3 ^fl/fl^ Villin‐Cre mice were mated in our laboratory. All the mice were bred and maintained in specific pathogen‐free conditions. Sex‐matched littermate mice that were 6–12 weeks of age were used for the experiments.

Colorectal Orthotopic Cancer

4.2

The murine colon cancer cell line MC38 (RRID: CVCL_B288) was obtained from the Cell Resource Center, Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences (CAMS/PUMC, Beijing, China), and preserved in liquid nitrogen in our laboratory. Cells were cultured in RPMI 1640 medium supplemented with 10 % fetal bovine serum (FBS) and 1 % penicillin‐streptomycin (P/S) at 37°C in a humidified incubator with 5 % CO_2_. The ovalbumin (OVA)‐expressing MC38 cell line (MC38‐OVA) was kindly provided by Professor Liufu Deng (Shanghai Jiao Tong University, Shanghai, China) and preserved in liquid nitrogen in our laboratory. Culture conditions were the same as those of the common MC38 cell line.

To establish orthotopic MC38 tumors, drinking water supplemented with 1.5 % DSS (MP Biomedicals, USA) was available ad libitum to the mice starting from day −9 for 5 days; the water was replaced with fresh drinking water and was available to the mice for 4 days. After 12 h of fasting, the mice were administered 2 × 10^6^ MC38 tumor cells suspended in 15 µL of 1 × PBS via injection into the rectal submucosa using a 50 µL Hamilton Microliter Syringe (Day 0). The mice were sacrificed 2 weeks (Day 14) after tumor inoculation, as previously described [18]. Anakinra (15 µg per mouse each time, MCE, USA) was locally injected into the tumor in mice. To delete the mouse gut microbiota, broad‐spectrum antibiotics were added to water, including 10 g/L ampicillin, 10 g/L neomycin sulfate, 10 g/L metronidazole, and 5 g/L vancomycin. The mixed antibiotic water was given by gavage in 0.5 mL once every two days.

Intestinal Inflammation

4.3

Age‐matched mice with initial body weights over 20 g were administered 1.2 % DSS in drinking water with ad libitum access to food. DSS treatment was continued for 9 consecutive days, followed by replacement with regular drinking water for an additional 5 days before euthanasia. Throughout the colitis induction period, mice were monitored daily for body weight changes and fecal bleeding as clinical indicators of disease, as described previously [16]. Age‐matched mice with initial body weights over 20 g were administered Citrobacter rodentium (CR; 2 × 10^10^‐3 × 10^10^ CFU) in 200 µL volume by gavage after fasting for 4 h. CR used for gavage was cultured overnight and amplified for an additional 8–9 h until OD_600_ reached 1.0. Throughout the enterocolitis induction period, mice were monitored daily for body weight changes for 8 days before euthanasia. CR‐OVA infection was the same as wild‐type CR except for the induction of expression before gavage.

Cell Isolation and Culture

4.4

For isolation of immune cells from the colon lamina propria, colon tissue was collected and washed with 1 x HBSS (calcium‐ and magnesium‐free [CMF]) to remove the intestinal contents. The colons were cut into 0.5 cm pieces and shaken twice in 20 mL CMF solution containing 25 mM NaHCO_3_, 2 % FBS, and 5 mM EDTA at 37°C for 30 min. Subsequently, the colon sections were harvested and incubated in 2 mL RPMI 1640 medium containing 5 % FBS for 5 min. The colon sections were then digested in 2 mL RPMI 1640 medium containing 5 % FBS and 12 mg/mL collagenase type IV (Worthington) at 37°C for 45 min. The cells released from the colon sections were collected by filtration through 70 µm nylon mesh, as described previously [21]. For isolation of IECs from the colon, cells were collected from supernatant (about 16 mL CMF solution according to requirements) of the second incubation mentioned above by centrifugation. Collected epithelial cells were washed with RPMI 1640 medium (10 % FBS) by centrifugation at 220 rpm for 5 min at 37°C and filtered to get a single‐cell suspension.

For the coculture between IECs and T cells. IECs isolated from the colon were stained and purified by flow sorting (7AAD^−^CD45^−^EpCam1⁺). CD4^+^T cells and CD8^+^T cells isolated from the colon lamina propria with flow cytometry. Cells were set in a 48‐well plate in RPMI 1640 medium with 10 % FBS and antibiotics (100 U/mL penicillin, 0.1 mg/mL streptomycin, 0.25 µg/mL amphotericin B) in each well. Anti‐CD3 (2 µg/mL) was pre‐coated at the bottom of the plate, and anti‐CD28 (2 µg/mL) and IL‐2 (5 ng/mL) were added in the medium for T cell survival. For IL‐1β signaling pathway blocking, coculture between IECs and T cells for 3–5 days was performed in the presence of anakinra (500 ng/mL, MCE, USA) or vehicle, after which the T cells were collected and analyzed.

Adeno‐Associated Virus Transfection

4.5

IECs were sorted from single‐cell suspensions of colon from mice, as described above. A total of 5 × 10^5^ cells was plated in 500 µL in a 48‐well flat‐bottom plate and incubated for 4 h. The Sirt3 overexpression system was established using adeno‐associated virus (AAV)9. The AAV9 carrying Sirt3 cDNA (NM_001177804.1) alongside a promoter (CAG‐MCS‐BGH PA, VB251021‐2002tpu). The Sirt3 overexpressing AAV9 (Sirt3 ^OE^) and the negative control vector (Ctrl) were constructed by VectorBuilder (Guangzhou, China). Sirt3 ^OE^ cocultured with intestinal epithelial cells at 37°C for 48 h, while the control (Ctrl) group received AAV9‐empty vector (MOI = 10^4^). Cells were sorted based on the expression of GFP by flow cytometry, and further analysis was performed.

Targeted Metabolomics Analysis

4.6

IEC cells were washed twice with pre‐cooled PBS and extracted using pre‐chilled 80 % methanol (Solarbio, Beijing, China). Targeted metabolomics was performed on a liquid chromatography with tandem mass spectrometry (LC‐MS/MS 8060, Shimadzu Corporation, Kyoto, Japan) equipped with an electrospray ionization source, as described previously [14]. Briefly, 5 mL of cell extract was injected onto a Waters ACQUITY BEH Amide column. Separation was performed at 40°C with a flow rate of 0.4 mL/min using 10 mM ammonium acetate in water (mobile phase A) and acetonitrile (mobile phase B). The generated data was processed using Shimadzu LabSolutions Postrun Analysis (version 5.89, Shimadzu, Kyoto, Japan). The integrated peak areas of targeted metabolites were normalized to the internal standard for semi‐quantification.

Immunofluorescence and Histological Analyses

4.7

Tumor tissues were collected and fixed in 4 % paraformaldehyde (PFA). Paraffin‐embedded tissues were sliced into 4 µm thick sections and stained with hematoxylin and eosin (HE). The sections were stained using standard protocols for xylene and an alcohol gradient for deparaffinization. After antigen retrieval and unmarking procedures, DAPI (Sigma, USA) was used to indicate the nucleus, anti‐CD4 PE, and anti‐CD8 FITC (both from Biolegend, USA) were incubated overnight at 4°C in a humidified chamber to indicate immune cell infiltration in tumors. After rinsing in PBS‐T (0.1 % Tween‐20) for 5 min of 3 times. Immunohistochemistry (IHC) slides were scanned with a Pannoramic Digital Slide Scanner (SDHISTECH, Budapest, Hungary), and images were cropped from virtual slides in Pannoramic Viewer.

Flow Cytometry

4.8

Single‐cell suspensions were prepared from colon or tumor tissues and from cell culture systems in vitro. For analysis of cell surface markers, the cells were stained with the indicated antibodies in 1 x PBS containing 2 % (w/v) BSA and 0.1 % NaN_3_ for 20 min on ice, as previously described [55, 56]. Anti‐mouse antibodies from Bioledgend (clones are indicated in brackets): CD4 (RM4‐5), CD8 (53‐6.7), CD54 (YN1/1.7.4), CD62L (MEL‐14), B220 (RA3‐6B2), CD25 (PC61), CD45 (30‐F11). Antibodies from eBioscience: CD45 (30‐F11), EpCam1 (G8.8). Antibodies from BD: 7‐AAD, MHC II (I‐A/I‐E, M5/114.15.2), CD44 (IM7), IL‐4 (11B11). Antibody IL‐1R1 (Leu20‐Lys338, R&D Systems) and OVA‐tetramers (Creative Biosciences).

For the intracellular staining of cytokines, the cells were treated with LPS (10 ng/mL, Sigma–Aldrich) or PMA (50 ng/mL, Sigma–Aldrich)/ionomycin (1 µg/mL, PeproTech) for 5 h. Next, 1 % GolgiStop (BD Biosciences) was added to the cell culture medium simultaneously. The cells were fixed and permeabilized by using a BD Cytofix/Cytoperm Kit according to the manufacturer's instructions (BD Biosciences). Anti‐IL‐23 (fc23cpg) antibody was obtained from eBioscience; anti‐IFNγ (XMG1.2), anti‐IL‐4 (11B11), anti‐IL‐17A (eBio17B7), and anti‐TNFα (MP6‐XT22) antibodies were obtained from Biolegend; and an anti‐IL‐1β (NJTEN3) antibody was obtained from eBioscience. For intracellular staining of transcription factors, the cells were permeabilized and fixed with reagents from a Foxp3 Staining Buffer Set. Anti‐Foxp3 (FJK‐16s) from eBioscience. Flow cytometry data were acquired on an ACEA NovoCyte (ACEA Biosciences) and analyzed with NovoExpress (Agilent Technologies) or FlowJo (TreeStar). Cell sorting was carried out on a FACS Aria III (BD Biosciences).

NAD+ Level Assay

4.9

NAD^+^ levels were measured using an NAD^+^/NADH assay kit (N6035, UElandy, Jiangsu, China).

ScRNA‐seq Analysis

4.10

ScRNA‐seq datasets were retrieved from GSE261388. Analysis of IEC clusters in intestinal tissue using the “Find All Markers” function to show the relationship between SIRT3 and intestinal inflammation and intestinal cancer. Then, performing GO functional enrichment and KEGG signaling pathway enrichment on the differentially expressed genes. The expression levels of SIRT3 in IECs from the intestine were presented by violin plots. Data were imported in R software (version 4.4.1). The “Seurat” (version 5.2.1) and “ggplot2” (version 3.5.2) packages were used for statistical analysis and visualization, respectively, as described previously [57].

Statistical Analysis

4.11

All the data are presented as the means ± SDs. Student's unpaired t‐test for parametric data or the Mann–Whitney test for nonparametric data was used when two samples were compared, and one‐way ANOVA with Dunnett's post hoc test for parametric data or the Kruskal–Wallis test for nonparametric data was used when more than two samples were compared. Differences between groups were considered statistically significant when the p‐value (alpha value) was less than 0.05.

Resource Availability

4.12

Lead Contact

4.12.1

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Yujing Bi (byj7801@sina.com) and Guangwei Liu (liugw@bnu.edu.cn).

Materials Availability

4.12.2

This study did not generate new, unique reagents.

Data and Code Availability

4.12.3

All data generated or analyzed during this study are included in this published article and its Supporting Information files.

Author Contributions

R.N., Y.D., J.T., and J.X. designed and conducted the experiment with cells and mice, analyzed data. L.Z., Z.G., Y.L., J.Z., X.L., M.F., and X.J. participated in discussions. Y.B. contributes to analyzing data and writing the manuscript. G.L. developed the concept, designed and conducted the experiments, analyzed data, wrote the manuscript, and provided overall direction.

Funding

The authors’ research is supported by grants from the National Natural Science Foundation for Key Programs of China (31730024, G.L.) and National Natural Science Foundation for General Programs of China (32370924 and 32170911).

Ethics Statement

All animal experiments were approved by the Animal Ethics Committee of Beijing Institute of Microbiology and Epidemiology and Beijing Normal University (IACUC‐DWZX‐2017‐003 and CLS‐EAW‐2017‐002).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File: advs73603‐sup‐0001‐SuppMat.pdf.