Eicosapentaenoic acid enhances intestinal stem cell-mediated colonic epithelial regeneration by activating the LSD1-WNT signaling pathway

Graphical abstract

Highlights

• •EPA mitigates DSS-induced colitis and promotes colonic epithelium regeneration.
• •EPA enhances intestinal organoid growth and ISC proliferation and differentiation.
• •EPA upregulates LSD1 expression to induce ISC proliferation and differentiation.
• •LSD1 in ISCs is necessary for EPA-mediated prevention of DSS-induced colitis.
• •EPA ameliorates DSS-induced colitis through activating the WNT signaling pathway.

Abstract

Introduction

Inflammatory bowel disease (IBD) is often associated with impaired proliferation and differentiation of intestinal stem cells (ISCs). Eicosapentaenoic acid (EPA), which is predominantly found in fish oil, has been recognized for its intestinal health benefits, although the potential mechanisms are not well understood.

Objectives

This study aimed to investigate the regulatory role and mechanism of EPA in colonic epithelial regeneration, specifically from the perspective of ISCs.

Methods

Wild-type mice whose diet was supplemented with 5% EPA-enriched fish oil were subjected to dextran sulfate sodium (DSS) to induce colitis. We utilized intestinal organoids, ISC-specific lysine-specific demethylase 1 (LSD1) knockout mice, and WNT inhibitor-treated mice to explore how EPA influences ISC proliferation and differentiation. ISC proliferation, differentiation and apoptosis were assessed using tdTomato and propidium iodide tracer testing, histological analyses, and immunofluorescence staining.

Results

EPA treatment significantly mitigated the symptoms of DSS-induced acute colitis, as evidenced by lower body weight loss and decreased disease activity index, histological scores and proinflammatory cytokine levels. Additionally, EPA increased the numbers of proliferative cells, absorptive cells, goblet cells, and enteroendocrine cells, which enhanced the regeneration of intestinal epithelium. Pretreatment with EPA increased ISC proliferation and differentiation, and protected against TNF-α-induced cell death in intestinal organoids. Mechanistically, EPA upregulated G protein-coupled receptor 120 (GPR120) to induce LSD1 expression, which facilitated ISC proliferation and differentiation in organoids. ISC-specific ablation of LSD1 negated the protective effect of EPA on DSS-induced colitis in mice. Moreover, EPA administration activated the WNT signaling pathway downstream of LSD1 in ISCs, while inhibiting WNT signaling abolished the beneficial effects of EPA.

Conclusions

These findings demonstrate that EPA promotes ISC proliferation and differentiation, thereby enhancing colonic epithelial regeneration through the activation of LSD1-WNT signaling. Consequently, dietary supplementation with EPA represents a promising alternative therapeutic strategy for managing IBD.

Introduction

Dysregulation of intestinal homeostasis heightens the susceptibility to inflammatory bowel diseases (IBDs), which have become major public health concerns [1]. The maintenance of intestinal homeostasis is dependent on the regulation of intestinal stem cells (ISCs) and involves fine-tuning of the balance between ISC proliferation and differentiation [2]. Once the intestine is damaged, ISCs immediately trigger the efficient regeneration of the epithelium and continuously generate transit amplifying (TA) progenitors. These progenitors migrate toward the villi and differentiate into various epithelial cell lineages [3]. Recent evidence has suggested that individuals afflicted with IBD exhibit a marked reduction in the population of Lgr5-positive ISCs and Paneth cells [4], [5], which emphasizes the pivotal role of ISC abnormalities in the etiology of intestinal diseases. Consequently, the precise regulation of ISC proliferation and differentiation holds significant potential for the development of innovative therapeutic strategies to combat IBDs.

Dietary cues and nutrients are pivotal in determining the fate of ISCs, which in turn are essential for maintaining intestinal homeostasis [6], [7]. Specifically, fatty acids have recently attracted substantial attention as vital modulators of intestinal health and disease [8]. Wang et al. [9] demonstrated that arachidonic acid facilitates small intestine epithelial regeneration following injury by augmenting the proliferation of + 4 intestinal stem/progenitor cells. In contrast, butyrate, a gut microbial metabolite, supports enterocyte growth and concurrently inhibits the proliferation of colonic Lgr5-positive ISCs and precursor cells in mice [10]. These divergent impacts of different fatty acids on ISC regulation underscore the complexity of nutrient-ISC interactions. Eicosapentaenoic acid (EPA), which is predominantly found in fish oil, has beneficial effects on intestinal health, as evidenced in both animal models and clinical trials. Epidemiological studies have reported that patients with ulcerative colitis have a significant decrease in the availability of EPA [11]. Although previous research has identified the role of EPA in mitigating intestinal injury and mucosal inflammation [12], [13], the potential mechanisms remain unclear. Given the dominant role of ISCs in intestinal damage, these studies motivated us to investigate whether EPA regulates ISC function to maintain intestinal homeostasis.

Lysine specific demethylase 1 (LSD1), a histone demethylase, is known to regulate the epigenetic landscape of stem cells, influencing their proliferation and differentiation [14]. Previous studies have highlighted the critical function of LSD1 in intestinal epithelial regeneration, including its roles in the maturation of Paneth and goblet cells [15], [16]. Recent research showed that demethylation by LSD1 prevented β-catenin degradation thereby maintaining its nuclear levels [17]. Meanwhile, WNT/β-catenin signaling pathway is essential for ISC activity, governing key processes such as crypt regeneration and epithelial repair. These studies suggest that the LSD1-WNT axis may play a critical role in maintaining intestinal homeostasis and promoting epithelial regeneration. Here, we used dextran sulfate sodium (DSS)-induced colitis and conditional knockout models to investigate the role and underlying mechanisms of EPA in colonic epithelium regeneration following injury in both mouse and organoid models. Our findings revealed that EPA significantly enhanced the regeneration of colonic epithelium by promoting ISC proliferation and differentiation. Intriguingly, we observed that EPA upregulated G protein-coupled receptor 120 (GPR120) to induce the expression of LSD1, an epigenetic factor that activates the WNT signaling pathway, which in turn enhanced the expansion of ISCs. This cascade of molecular events induced by EPA underscores its potential as a therapeutic agent. The elucidation of the role of EPA in ISCs will provide potential therapeutic strategies for IBD treatment.

Material and methods

Animals

Lgr5-EGFP-IRES-creERT2 knock-in mice (No.008875) and LSD1 ^flox/flox mice (No.023969) were purchased from the Jackson Laboratory. Rosa26-CAG-LSL-tdTomato transgenic mice (No. C001181) were obtained from Cyagen Biosciences. To generate ISC-specific LSD1 knockout mice, LSD1 ^flox/flox mice were crossbred with Lgr5-EGFP-IRES-creERT2 mice and/or Rosa26-CAG-LSL-tdTomato mice. All mouse strains were housed in a specific-pathogen free (SPF) facility with a 12-hour light/dark cycle at 22 ± 2 °C and humidity 55 ± 15 %. Mice had continuous access to a standard diet, in addition to ad libitum access to water. All studies were conducted on adult animals aged at least 8 weeks at the beginning of the experiment.

Ethics statement

All experiments involving animals were conducted according to the ethical policies and procedures approved by the ethics committee of Wuhan Polytechnic University, China (Approval No. EM20230512001).

In vivo experiments

For EPA experiments, 8-week-old male C57 BL/6J mice were fed the standard diet plus 5 % w/w EPA-enriched fish oil for a period of 28 days. The fish oil is obtained from Fuzhou Gaolong Industrial Co., Ltd and contains 50 % EPA. The dosages of EPA were determined according to the literature [18]. On day 14, mice were administered with 2.5 % DSS (MP Biomedicals, 36 000–50 000 Da) in water for 7 days, and then replaced with fresh water for 7 days. On day 29, mice were euthanized and then sacrificed. For LSD1 deletion induction, mice were intraperitoneally injected with 100 mg/kg body weight of tamoxifen (Sigma) every day for 3 consecutive days. For lineage tracing experiments, LSD1^flox/flox mice were bred with Lgr5-EGFP-IRES-CreERT2 mice and Rosa26-CAG-LSL-tdTomato mice to generate Lgr5 lineage tracer mice. The offspring were induced by same dose of tamoxifen. Corn oil was used as control. For WNT inhibitor treatments, EPA or vehicle‐treated mice were subjected to intraperitoneal injection of LGK974 (3 mg/kg) or vehicle (0.1 % DMSO) every day for 5 days, followed by drinking water for 2 days. The distal colon (10 mm from middle) was isolated and used for analysis.

Pathologic scoring

For histologic scoring, the colon was fixed with 4 % paraformaldehyde and then subjected to histological examination using hematoxylin and eosin (H&E) staining. A pathologist scored colonic sections based on the sum of crypt damage, inflammation and ulceration scores. The severity of inflammation was scored as 0 (absent), 1 (slight), 2 (moderate), and 3 (severe). The crypt damage was scored as 0 (absent), 1 (one-third damaged), 2 (two-thirds damaged), 3 (only epithelium intact) and 4 (lost of entire crypt and epithelium). The ulceration was scored as 0 (absent), 1 (mucosa), 2 (mucosa and submucosa), and 3 (transmural and epithelium lost).

Disease activity index (DAI)

The DAI score includes the sum of body weight, diarrhea, and bleeding scores, which is used to evaluate the phenotype of colitis. Each component of DAI was assigned a score ranging from 0 to 4. Body weight changes were scored as follows: no weight loss (0), 1–5 % loss (1), 6–10 % loss (2), 11–20 % loss (3), and over 20 % loss (4). The gross bleeding was scored as 0 (absent), 1 (stool color changed), 2 (small area bleeding), 3 (large area bleeding) and 4 (visible blood on feces). The fecal consistency was scored as 0 (firm stools), 1 (soft but retrievable stools), 2 (unshapely but non-retrievable stools), 3 (shapeless and non-retrievable stools), and 4 (watery consistency).

Inflammatory cytokine assay

The concentrations of IL-6, TNF-α, MCP-1, and IFN-γ in serum were measured by a cytometric bead array (CBA) mouse inflammation kit (BD Biosciences) in accordance with manufacturer’s instructions. Briefly, samples or standard samples with known concentrations (ranging from 0 to 5000 pg/mL) are mixed with capture antibody beads and phycoerythrin (PE)-conjugated detection antibody, and then incubated in the dark for 2 h before washing. Data were collected and analyzed using a Beckman SRT flow cytometer.

Immunofluorescence staining

For immunofluorescence staining, the frozen sections were first blocked with 1 % normal goat serum for 30 min and washed 3 times. Then, they were incubated overnight at 4 ℃ with the following primary antibodies: Ki67 (No. ab16667, Abcam), Muc2 (No.sc-515032, Santa Cruz), and ChgA (No. sc-393941, Santa Cruz). The sections were incubated with the corresponding secondary antibody (Invitrogen) for 2 h, followed by staining with DAPI for 10 min and washing with water 5 times. Fluorescence images were acquired using a confocal microscope (FV10i, Olympus) for both qualitative and quantitative evaluation.

Crypt isolation and organoid culture

Intestinal crypt isolation, purification, and organoid culture were performed according to the experimental procedure described by Wang et al. [9]. Briefly, the proximal colon was isolated, opened longitudinally, and washed with cold 0.1 % BSA. After washing, we cut the colon into approximately 1–2 mm pieces. Next, the pieces were incubated in Gentle cell dissociation reagent (STEMCELL, No. 07174) for 15 min at room temperature on a horizontal shaker. After blowing the pieces, the supernatant was filtered through a 70 μm cell strainer, and we then observed the purity of the crypts under a microscope, separated and purified the crypts by centrifugation at 250 g for 5 min. 200 isolated crypts were mixed with 50 μL of growth factor-reduced Matrigel (Corning, No.356237), were plated in 24-well plates in IntestiCult Organoid Growth Medium (STEMCELL, No.06005). Five non-overlapping fields were randomly captured using an inverted microscope for each well, and the surface area and budding number of organoids were analyzed using ImageJ software.

Image Acquisition and analysis of organoid cell death

For the propidium iodide (PI)-traced organoid cell death assay, organoids were stained with 1 μg/mL PI for 10 min, then removed and replaced with fresh culture medium. The images were taken every 4 h by real-time tracking and photography. The percentage of the PI-positive area per organoid was calculated by ImageJ software.

RNA extraction and quantitative RT-PCR

RNA was extracted from isolated colon crypts and organoids. The extraction utilized the TRIzol reagent (Invitrogen), and the subsequent cDNA synthesis was performed using M−MLV (H-) Reverse Transcriptase and oligo (dT) primer (Vazyme, No. R021-01). The cDNA served as the template for quantitative real-time PCR, which was conducted using PowerUp™ SYBR™ Green (Applied Biosystems, No. A25741) on an CFX96 Real-Time PCR (Bio-Rad). The primer sequences were shown in Supplementary Table S1.

Western Blotting

The proteins from colon crypts or organoids were extracted using standard procedures. In short, the samples were lysed in RIPA buffer (KeyGEN, No. KGP704) with 1 mmol/L PMSF and subsequently centrifuged at 12000 g for 15 min. Total protein concentration was quantified using a commercially protein detection kit (KeyGEN, No. KGP902). The proteins were then separated by 10 % SDS-PAGE electrophoresis, and transferred onto PVDF membranes. The membranes were blocked with 5 % non-fat dry milk for 3 h, followed by incubation with primary antibodies: Muc2 (No. sc-515032, Santa Cruz, 1:1000), ChgA (No. sc-393941, Santa Cruz, 1:1000), Villin (No. sc-58897, Santa Cruz, 1:1000), Lgr5 (No. A12327, ABclonal, 1:1000), Non-phospho (Active) β-catenin (No. A22180, ABclonal, 1:1000), β-catenin (No. 610153, BD Biosciences, 1:1000) and β-actin (No. SAB4300645, Sigma, 1:1000) for normalization. The grayscale values of bands were quantified using Image J Software and normalized to β-actin.

Proliferation assay

The ability of proliferation was assessed by Cell-Light EdU DNA cell proliferation kit (No.C103102, RiboBio), according to the manufacturer’s instructions. In short, we treated intestinal organoids with 20 mmol/L EdU for 2 h, and then fixed them with 4 % paraformaldehyde for 24 h. After infiltration with 0.5 % Triton X-100, intestinal organoids were reacted with 1 × Apollo reaction mixture for 30 min. Subsequently, the organoids were stained with DAPI for 10 min and visualized under a confocal microscopy (FV10i, Olympus).

Statistical analysis

All results were presented as the mean ± SEM and were performed with GraphPad Prism software (Version 8.0, USA) and ImageJ software (Version 2.1.0). The data were analyzed by one-way ANOVA followed by LSD multiple comparisons for experiments with three or more groups. Significance was assumed with *and ^#p < 0.05; ** and ^##p < 0.01; *** and ^###p < 0.001.

Results

EPA supplementation ameliorates DSS-induced colitis in mice

To evaluate whether EPA ameliorates colonic epithelium damage and colitis, eight-week-old mice were fed standard diet supplemented with 5 % EPA-enriched fish oil for 28 days (Fig. 1A). Throughout the 14 days of feeding before DSS treatment, body weight change and food intake of mice was not affected by EPA treatment (Supplementary Fig. S1A-B). We then administered 2.5 % DSS to the mice via the drinking water for 7 days to establish a colitis model, followed by normal water for 7 days. After 5 days of DSS treatment, mice began to lose weight, while EPA greatly reduced body weight loss on day 8 (Fig. 1B). The disease activity index was significantly decreased in the EPA group after DSS treatment (Fig. 1C). EPA treatment also significantly reduced the DSS-induced increase in the serum contents of the inflammatory cytokines including IL-6, MCP-1 and TNF-α (Fig. 1D), but did not affect IFN-γ level in serum (Supplementary Fig. S1C). Histological examination revealed that DSS treatment led to significant loss of epithelial cells, extensive crypt shedding, and massive infiltration of immune cells in the colon (Fig. 1E). In contrast, EPA supplementation significantly decreased colonic damage and inflammatory cell infiltration in DSS-treated mice (Fig. 1E-F). Moreover, histological changes in morphology, villus height and crypt depth in the jejunum were not affected by EPA treatment (Supplementary Fig. S1D-F). Furthermore, EPA supplementation dramatically reduced DSS-induced shortening of colon length and crypt depth (Fig. 1G-I). In addition, the expression of pro-inflammatory cytokines TNF-α, IL-6, and IL-1β were increased in the colon of DSS-treated mice, which was significantly restrained in EPA-treated mice (Fig. 1J). Taken together, these results imply that EPA supplementation effectively ameliorates DSS-induced colonic inflammation and injury in mice.

Fig. 1.

EPA supplementation ameliorates DSS-induced colitis in mice. Eight-week-old mice were treated with 5 % EPA-enriched fish oil in standard diet for 28 days. On day 14, mice were treated with 2.5 % DSS for 7 days, and then replaced with fresh water for 7 days before being euthanized. (A) Schematic of experimental design. (B) Body weight. (C) Disease activity index at day 7 after DSS treatment. (D) Level of serum IL-6, MCP-1 and TNF-α. (E-F) Histological assessment of colon tissues. Scale bar, 200 μm; A indicates discontinuity or rupture of the epithelial layer; B indicates immune cell infiltration in the lamina propria; C indicates crypt structure disorganization. (G) The representative images of the colon. (H-I) Colon length and crypt depth. (J) Relative mRNA expression of TNF-α, IL-6, and IL-1β in the colon. Data are presented as mean ± SEM, n = 9. DEPA represents DSS + EPA group; ND represents not detected. *p < 0.05, **p < 0.01, ***p < 0.001 vs controls; ^#p < 0.05 vs DSS group. Statistical significance was assessed by using one-way ANOVA with multiple comparisons test.

EPA promotes colonic epithelium regeneration after DSS-induced injury in mice

To investigate the role of EPA in colonic epithelium regeneration, we first measured the expression of cell proliferation and differentiation markers. The mRNA levels of Ki67 and PCNA (cell proliferation markers), Villin (absorptive cell marker), Muc2 (goblet cell marker), ChgA (enteroendocrine cell marker), and Lgr5 and Ascl2 (ISC markers) were significantly reduced in colon tissues of DSS-treated mice (Fig. 2A-G). However, EPA supplementation significantly blocked the DSS-induced decrease in the expression of these markers (Fig. 2A-G). Notably, EPA supplementation also increased the protein expression of Villin, Muc2, ChgA and Lgr5 regardless of whether DSS was administered (Fig. 2H-I). Furthermore, immunofluorescence staining revealed that a fewer Ki67-positive cells was observed in the DSS group than in the control group (Fig. 2J). EPA treatment significantly increased the number of Ki67-positive cells under normal conditions and after injury (Fig. 2J). Compared with control mice, DSS-treated mice had fewer Muc2-positive and ChgA-positive cells in the colon (Fig. 2 K-L). Supplementation with EPA increased the number of Muc2-positive and ChgA-positive cells regardless of whether DSS was administered (Fig. 2 K-L), which is consistent with what was observed at the translational level (Fig. 2D-E).

Fig. 2.

EPA promotes colonic epithelium regeneration after DSS-induced injury. (A-G) Relative mRNA expression of Ki67 (A), Pcna (B), Villin (C), Muc2 (D), ChgA (E), Lgr5 (F) and Ascl2 (G) in the colon. (H-I) Representative images (H) and quantification (I) of Villin, ChgA, Muc2 and Lgr5 protein expression in the colon. (J-M) Representative images and quantification of Ki67-positive regenerative crypts (J), Muc2-positive cells (K), ChgA-positive cells (L) and tdTomato-positive crypts (M) in the colon. Scale bar, 100 μm or 50 μm. Data are presented as mean ± SEM, n = 9 (A-G), n = 4 (H-M). *p < 0.05, **p < 0.01, ***p < 0.001 vs controls; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 vs DSS group. Statistical significance was assessed by using one-way ANOVA with multiple comparisons test.

To further confirm the effect of EPA on ISC differentiation, Lgr5-IRES-CreERT2 mice were crossed with Rosa26-LSL-tdTomato mice to generate Lgr5 lineage tracer mice, which enabled tdTomato labeling of Lgr5-positive ISCs and their offspring upon tamoxifen treatment. Compared with control mice, DSS-treated mice generated 3 times fewer tdTomato-positive crypts with fewer labeled progeny extending up the crypt-villous units (Fig. 2 M). EPA supplementation markedly enhanced the ability of ISCs to generate tdTomato-positive progeny toward the villi in mice compared with control mice or DSS-treated mice (Fig. 2 M). These data suggest that EPA promotes ISC proliferation and differentiation and colonic epithelium regeneration.

EPA enhances intestinal organoid growth and ISC proliferation and differentiation

Next, we utilized a 3D intestinal organoid model to verify the function of EPA in intestinal regeneration after injury ex vivo. We cultured primary organoids in media containing 0, 1, 5 or 20 μmol/L EPA (Supplementary Fig. 2A). After 48 h of treatment, treatment with 5 and 20 μmol/L EPA significantly increased the budding efficiency of organoids (Supplementary Fig. S2B). We then used TNF-α, which mimics the role of DSS, to establish an organoid injury model. PI staining revealed that treatment with 50 ng/mL TNF-α significantly induced abundant cell death in normal organoids but not in EPA-treated organoids after 24 h (Fig. 3A-B and Supplementary Video 1A-D). In addition, supplementation with 20 μmol/L EPA significantly blocked the TNF-α-induced increase in the necrosis rate of the organoids (Fig. 3C). The number of buds per organoid was reduced, but the surface area of the organoids was not affected in the TNF-α group (Fig. 3D-E). The organoids treated with EPA contained a greater number of buds and larger surface area regardless of whether they were treated with TNF-α (Fig. 3D-E). The mRNA expression of Ki67, PCNA, and Villin was lower in the TNF-α group than in the control group (Fig. 3F-G and Supplementary Fig. S2C). EPA treatment not only alleviated TNF-α-induced decrease of Ki67, PCNA, and Villin mRNA expression (Fig. 3F-G and Supplementary Fig. S2C), but also increased the mRNA expression of ChgA, Muc2, Lgr5 and the quiescent ISC marker Hopx in the organoids (Fig. 3H-J and Supplementary Fig. S2D). Interestingly, EPA treatment increased the mRNA expression of the quiescent ISC markers Bmil1 and m-Tert in normal organoids but not in TNF-α-treated organoids (Supplementary Fig. S2E-F). Immunofluorescence staining revealed that the number of EdU-positive cells of organoids was decreased in TNF-α group compared to the control group (Fig. 3 K). However, EPA significantly increased the number of EdU-positive cells (Fig. 3 K), which explains the increased surface area and number of budding organoids. We also observed a significantly greater percentage of Muc2-positive cells and ChgA-positive cells in the organoids after EPA treatment regardless of whether they were treated with TNF-α (Fig. 3 K). These data indicate that EPA increases ISC proliferation and differentiation, which promotes the growth of intestinal organoids.

Fig. 3.

EPA enhances intestinal organoid growth and ISC proliferation and differentiation. Intestinal organoids were treated with EPA (20 μmol/L) for 48 h and then treated with TNF-α (50 ng/mL) for 24 h. (A) Time-lapse confocal microscopic imaging of propidium iodide (PI)-stained organoids; (B) The relative fluorescence intensity of PI per organoid (n = 20 organoids); (C-E) Quantification of necrosis rate (C), budding number (D), and surface area (E) in intestinal organoids; (F-J) Relative mRNA expression of Ki67 (H), Villin (I), ChgA (H), Muc2 (I), and Lgr5 (J) in intestinal organoids; (K) Representative images and quantification of EDU-positive cells, Muc2-positive cells and ChgA-positive cells in intestinal organoids. Scale bars, 100 mm. Data are presented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 vs controls; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 vs DSS group. Statistical significance was assessed by using one-way ANOVA with multiple comparisons test.

EPA directly upregulates LSD1 expression to induce ISC proliferation and differentiation in organoids

The mechanism by which EPA regulates ISC proliferation and differentiation remains unclear. Recent studies have demonstrated that LSD1 modulates the transcription of genes critical for ISC proliferation and differentiation [15], [16]. We observed a reduction in LSD1 expression in both DSS-treated colon tissues and TNF-α-treated organoids (Fig. 4A −D). Our results demonstrated that EPA significantly enhanced both the protein and mRNA levels of LSD1 in the colon and organoids under both normal and injury conditions (Fig. 4A-D). To explore the role of LSD1 in the EPA-mediated regulation of ISC proliferation and differentiation, we treated intestinal organoids with the specific LSD1 inhibitor GSK-LSD1. PI tracer experiments revealed that GSK-LSD1 treatment did not influence cell death, the necrosis rate, the number of buds per organoid, and the surface area of the organoids (Fig. 4E-I). However, the inhibition of LSD1 by GSK-LSD1 significantly blocked the EPA-mediated alleviation of TNF-α-induced cell death in the organoids (Fig. 4E-F and Supplementary Video 2A-D). Additionally, the beneficial effects of EPA on reducing the necrosis rate and increasing the budding number and surface area of TNF-α-treated organoids were completely negated by GSK-LSD1 treatment (Fig. 4G-I). EPA treatment significantly increased LSD1 expression and the mRNA expression of Ki67, PCNA, Villin, ChgA and Lgr5 in TNF-α-treated organoids (Fig. 4 J and Supplementary Fig. S3A). In contrast, the expression of the ISC marker Ascl2 expression remained unchanged (Supplementary Fig. S3B). Importantly, inhibition of LSD1 completely blocked the effects of EPA on LSD1, Ki67, PCNA, Villin, Muc2 and Lgr5 expression in the organoids (Fig. 4 J). Immunofluorescence staining indicated a significant increase in the number of EdU-positive and Muc2-positive cells in the organoids treated with EPA compared with the control group (Fig. 4 K). Nonetheless, GSK-LSD1 treatment completely abolished the effects of EPA (Fig. 4 K). These findings suggest that LSD1 mediates the EPA-induced proliferation and differentiation of ISCs.

Fig. 4.

EPA upregulates LSD1 expression to induce ISC proliferation and differentiation in organoids. Intestinal organoids were treated with EPA (20 μmol/L) or GSK-LSD1 (50 μmol/L) for 48 h and then treated with TNF-α (50 ng/mL) for 24 h. (A-B) Protein expression of LSD1 in the colon of mice treated with EPA or DSS; (C-D) The mRNA expression of LSD1 in the colon and intestinal organoids; (E) The images of PI-stained organoids; (F) The relative fluorescence intensity of PI per organoid; (G-I) Quantification of necrosis rate (G), budding number (H), and surface area (I) in intestinal organoids; (J) Relative mRNA expression of LSD1, Ki67, Villin, Muc2, and Lgr5 in intestinal organoids; (K) Representative images and quantification of EDU-positive cells and Muc2-positive cells in intestinal organoids. Scale bars, 100 mm. Values are means ± SEM, n = 4. LSD1i represents LSD1 inhibiter (GSK-LSD1) group. *p < 0.05, **p < 0.01, ***p < 0.001 vs controls; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 vs DSS group. Statistical significance was assessed by using one-way ANOVA with multiple comparisons test.

To explore how EPA influences LSD1, we initially assessed the colonic expression of long-chain fatty acid receptors, including G protein-coupled receptor 120 (GPR120) and GPR40. Our results showed that there was a significant upregulation of GPR120 mRNA expression, but not GPR40, in the colon of mice administered EPA (Supplementary Fig. S4A-B). In vitro experiments demonstrated that EPA dose-dependently increased GPR120 mRNA level in intestinal organoids (Supplementary Fig. S4C). To determine the role of GPR120 in EPA-mediated regulation of LSD1 and ISC proliferation and differentiation, we treated organoids with AH-7614, a selective GPR120 inhibitor. Administration of AH-7614 reduced GPR120 expression and inhibited EPA-stimulated LSD1 expression in organoids (Supplementary Fig. S4D-E). Moreover, GPR120 inhibition by AH-7614 significantly blocked the positive impact of EPA on enhancing the surface area and budding efficiency of intestinal organoids (Supplementary Fig. S4F-H). The EPA-induced increases in mRNA expression levels of Ki67, ChgA, Muc2, and Lgr5 were also inhibited (Supplementary Fig. S4I-L), indicating that EPA promoted LSD1 expression and ISC proliferation and differentiation in a GPR120-dependent manner.

LSD1 in ISCs is crucial for EPA-mediated prevention of DSS-induced colitis in mice

Given the critical role of LSD1 in ISC function, we further investigated the role of LSD1 in ISCs in the EPA-mediated prevention of DSS-induced colitis in mice. We generated ISC-specific LSD1 knockout (LSD1 IKO) mice by crossing LSD1-floxed mice with Lgr5-EGFP-IRES-CreERT2 knock-in mice (Supplementary Fig. S5A). As anticipated, tamoxifen-induced LSD1 IKO mice exhibited LSD1 deletion in crypts (Supplementary Fig. S5B). Both control (LSD1 ^flox/flox) mice and LSD1 IKO mice were fed a standard diet supplemented with EPA for 3 weeks, followed by treatment with 2.5 % DSS for 6 days to induce colitis (Fig. 5A). Control mice whose diet was supplemented with EPA showed less body weight loss and a reduced DAI after DSS treatment, whereas this effect was not observed in LSD1 IKO mice (Fig. 5B-C). The serum levels of inflammatory cytokines, including IL-6, MCP-1 and TNF-α, were significantly decreased in EPA-treated control mice, but this decrease was attenuated by LSD1 deficiency (Fig. 5D). The histopathological scores of EPA-treated control mice were significantly decreased in mice with DSS-induced colitis, whereas no effect was observed in LSD1 IKO mice (Fig. 5E-F). Importantly, LSD1 IKO mice exhibited more severe colitis, as indicated by higher histological scores, and greater TNF-α and IL-1β expression in the colon, compared with control mice (Fig. 5E-F and 5H), despite no significant difference in body weight change (Fig. 5B-C). EPA treatment did not affect colon length in LSD1 IKO mice (Fig. 5G). Furthermore, the mRNA expression of TNF-α, IL-6, and IL-1β in the colon was significantly reduced in EPA-treated control mice exposed to DSS (Fig. 5H), but LSD1 deficiency abrogated the beneficial effects of EPA (Fig. 5H). EPA treatment significantly increased the mRNA expression of LSD1, Ki67, PCNA, ChgA and Lgr5 in the colons of control mice but not in those of LSD1 IKO mice (Fig. 5I). The tracer experiment revealed a greater number of tdTomato-positive crypts in the colon of EPA-treated control mice, while LSD1 deficiency completely abolished the effects of EPA (Fig. 5 J). Compared with control mice, LSD1 IKO mice produced fewer labeled progeny with tdTomato-positive crypts (Fig. 5 J). These data support an essential role for LSD1 in the EPA-mediated alleviation of DSS-induced colitis.

Fig. 5.

LSD1 is crucial for EPA-mediated prevention of DSS-induced colitis in ISCs. LSD1 IKO mice and LSD1 ^flox/flox mice were administered tamoxifen five times, and then were treated with 5 % EPA-enriched fish oils in standard diet for 23 days. On day 19, mice received 2.5 % DSS for 6 days to induce colitis, followed by three days on fresh water. (A) Schematic of experimental design; (B) Body weight changes, shown as percentages of initial weight, from the onset of DSS treatment; (C) Disease activity index post-DSS treatment; (D) Serum levels of pro-inflammatory cytokines; (E-F) Histological assessment of colon tissues by H&E staining and histologic score analysis. Scale bars, 200 μm; A indicates discontinuity or rupture of the epithelial layer; B indicates immune cell infiltration in the lamina propria; C indicates crypt structure disorganization. (G) The length of colon; (H) Relative mRNA levels of TNF-α, IL-6, and IL-1β in the colon. (I) Relative mRNA expression of LSD1, Villin, Ki67, Muc2, ChgA and Lgr5 in the colon. (J) Quantitative analysis and representative images of tdTomato-positive crypts by IF staining. Scale bars, 100 μm. Values are means ± SEM, n = 4. *p < 0.05, **p < 0.01, ***p < 0.001 vs controls. Statistical significance was assessed by using one-way ANOVA with multiple comparisons test.

EPA ameliorates DSS-induced colitis through activating the WNT signaling pathway

Recent studies have shown that LSD1 can regulate the WNT signaling pathway via stem cells in damaged adult tissue [17]. We hypothesized that EPA could upregulate WNT signaling, which is the downstream of LSD1 to promote colonic epithelial regeneration. Our findings showed that EPA treatment significantly upregulated the mRNA expression of key WNT signaling genes, including Wnt3a, Axin2 and Cyclin D1, in the colon and intestinal organoids (Supplementary Fig. S6A-B). The protein expression of total β-catenin and active β-catenin in the colon were increased by EPA treatment both in control mice and DSS-treated mice (Fig. 6A). Additionally, LSD1 knockout in ISCs blocked the EPA-induced elevation of Wnt3a, Axin2 and Cyclin D1, as well as total β-catenin and active β-catenin protein expression (Fig. 6B and Supplementary Fig. S6C), indicating that LSD1 is necessary for EPA to activate the WNT signaling pathway. To further validate the crucial role of WNT signaling in the protective effect mediated by EPA against DSS-induced colitis, we used the specific PORCN inhibitor LGK974 to inhibit WNT pathway activation by EPA in mice (Fig. 6C). We observed that LGK974 treatment markedly reduced the EPA-induced increase in Wnt3a and Axin2 expression in the colon (Supplementary Fig. S5D). Additionally, LGK974 significantly negated the ability of EPA to mitigate DSS-induced body weight loss and to increase the DAI (Fig. 6D-E). Histological assessments revealed that inhibition of WNT signaling by LGK974 led to severe inflammation, edema, and ulceration in the colons of EPA-treated mice (Fig. 6F-G). Moreover, colon length was significantly reduced following LGK974 administration (Fig. 6H). In EPA-treated mice, serum inflammatory cytokines IL-6, MCP-1, and TNF-α were substantially lower than that in control mice post-DSS treatment, and this effect was successfully reversed by LGK974 treatment (Fig. 6I). In addition, LGK974 injection significantly alleviated the EPA-induced decrease in TNF-α, IL-6 and IL-1β mRNA expression in DSS-treated mice (Fig. 6J). LGK974 treatment inhibited the EPA-induced increases in PCNA, Ki67, Muc2, ChgA, Lgr5 and Ascl2 expression in the colon (Fig. 6 K). Ex vivo experiments confirmed that inhibition of WNT signaling by LGK974 significantly blocked the protective effect of EPA against TNF-α-induced cell death in the organoids (Fig. 6L-M and Supplementary Video 3A-C). These results demonstrate that WNT signaling is necessary for EPA to exert its protective effects of EPA on DSS-induced colitis in mice.

Fig. 6.

EPA ameliorates DSS-induced colitis through activating the WNT signaling pathway. (A-B) Representative images and quantification of β-catenin and active β-catenin protein expression in the colon of EPA-treated wild-type mice (A) or LSD1 IKO mice (B). Wild-type mice were treated with 5 % EPA-enriched fish oils in standard diet for 22 days. On day 14, mice were injected with 3 mg/kg LGK974 every 2 days and received 2.5 % DSS for 6 days, followed by 3 days in fresh water, n = 6. (C) Schematic of experimental design. (D-E) Body weight changes (D) and disease activity index (E) were monitored daily starting from DSS treatment. (F-G) Histological assessment of colon tissues by H&E staining and histologic score analysis. Scale bars, 100 μm. (G) The length of colon. (I) Serum levels of pro-inflammatory cytokines. (J) Relative mRNA expression of TNF-α, IL-6, and IL-1β in the colon. (K) Relative mRNA expression of Pcna, Ki67, Muc2, ChgA, Villin, Ascl2 and Lgr5 in the colon. Intestinal organoids were treated with EPA (20 μmol/L), and then treated with/without 500 nmol/L LGK974 in the presence TNF-α (50 ng/mL) for 24 h, n = 4. (L) Time-lapse confocal microscopic imaging of PI-stained organoids. (M) The relative fluorescence intensity of PI per organoid; Data are presented as mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001 vs controls; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001 vs DSS + EPA group. Statistical significance was assessed by using one-way ANOVA with multiple comparisons test.

Discussion

Recent studies have extensively documented the beneficial effects of n-3 polyunsaturated fatty acids (PUFAs) on intestinal health [19]. However, the mechanisms underlying their impact on ISC proliferation and differentiation have not yet been defined. In this study, we utilized DSS-treated mice and TNF-α-treated intestinal organoids to elucidate whether EPA facilitates colonic epithelial regeneration both in vivo and ex vivo. Our observations revealed that EPA supplementation promoted ISC proliferation and differentiation, which enhanced colonic epithelial regeneration and mitigated DSS-induced colitis. We discovered that LSD1, a critical epigenetic regulator, is essential for the EPA-driven activation of the WNT signaling pathway, which in turn protects against colitis. These findings reveal novel insights into the role of EPA in enhancing colonic epithelial regeneration following damage and emphasize its role in ISC-mediated tissue repair.

Recent evidence has shown that EPA and docosahexaenoic acid (DHA), which are key components of n-3 PUFAs, have divergent effects on UC. Studies have revealed significantly reduced serum EPA levels but increased DHA levels in UC patients compared with controls [20]. Further, ethanol-induced intestinal damage is correlated with reduced EPA but not DHA levels [21]. Similarly, the severity of colitis was shown to be negatively correlated with the EPA level and positively associated with the DHA level in the colonic mucosa [22]. Notably, EPA effectively mitigates DSS-induced colitis in mice, outperforming DHA [23]. These observations suggest a strong relationship between EPA levels and IBDs, but any causal relationship is unclear. Previous research has indicated that EPA ameliorates small intestinal damage induced by deoxynivalenol and boosts intestinal barrier function [13], yet its specific impact on the colon has not been explored. In this study, the administration of 5 % EPA-enriched fish oil reduced DSS-induced colonic inflammation and body weight loss in mice, which highlights its regulatory potential in colitis. Moreover, clinical studies have confirmed the efficacy of EPA in reducing mucosal inflammation and preventing symptomatic relapses in UC patients [24]. Collectively, these findings support dietary EPA supplementation as a promising nutritional intervention for IBD management.

Previous studies have revealed that IBDs are linked to a disruption in the balance of ISC proliferation and differentiation [25]. Overactive ISC proliferation can expand the TA cell pool, which increases the risk of tumorigenesis. Conversely, excessive ISC differentiation can diminish the ISC pool and hinder intestinal epithelial regeneration [26]. Numerous reports suggest that this balance is influenced by a high-fat diet, which emphasizes the crucial role of fatty acids in ISC dynamics [27], [28], [29]. Research has predominantly focused on the impact of long-chain fatty acids, apart from n-3 PUFAs, on ISC regulation [30]. For instance, the treatment of mouse organoids with palmitate or oleic acid increases the number of Lgr5-positive ISCs and promotes secondary organoid formation [31]. Additionally, arachidonic acid not only promotes ISC proliferation but also enhances small intestinal regeneration by activating WNT signaling [9]. Our research demonstrated that dietary EPA supplementation significantly increased the number of differentiated cells such as enterocytes, goblet cells, and enteroendocrine cells. These cells primarily arise from ISC differentiation, which shows the influence of EPA on ISC function. Using Lgr5-EGFP-IRES-CreERT and Rosa26-tdTomato-postive double transgenic mice, which serve as robust models for tracing the ISC lineage due to their ability to permanently label Lgr5-positive cells and their progeny with tdTomato [32], we observed that EPA notably mitigated the DSS-induced reduction in ISC progeny, thus playing a critical role in ISC proliferation and differentiation.

Intestinal organoids, also known as “mini-guts”, are structures developed from a single ISC that form crypt-villus architectures and contain all types of intestinal epithelial cells. Due to their high structural and physiological similarity to the native intestine, intestinal organoids have been widely used in studies involving nutrient absorption, ISC proliferation and differentiation, as well as the interaction between gut microbiota and hosts [33]. Our findings indicated that pretreatment with EPA at concentrations of 1, 5 and 20 μmol/L significantly enhanced the budding efficiency of the organoids in a dose-dependent manner. Notably, EPA not only promoted organoid growth but also prevented TNFα-induced cell death. These beneficial effects of EPA are primarily attributed to improvements in ISC proliferation and differentiation. Supporting evidence has suggested that the organoids derived from the crypts of n-3 PUFA desaturase-1 (fat-1) transgenic mice are significantly larger than those derived from controls due to increased endogenous n-3 PUFA levels [21]. Additionally, the expression of ISC proliferation markers, along with mucus production, is significantly increased in intestinal organoids derived from fat-1 mice [21]. These results indirectly substantiate the role of EPA in promoting organoid growth. EPA is metabolized through enzymatic reactions involving cyclooxygenases and specific prostaglandin synthases to produce prostaglandin E3 (PGE3). Moreover, 10 μmol/L PGE3 supplementation increased the percentage of proliferative cells and goblet cells in organoid culture [34]. Thus, the involvement of EPA in regulating ISC function may be mediated by its conversion to PGE3.

Accumulating evidence highlights the critical role of LSD1 in ISC-mediated epithelial regeneration following intestinal injury [15]. Our previous research demonstrated that fatty acids, particularly propionate, can modulate intestinal LSD1 expression [35]. We hypothesized that LSD1 supports the enhancement of ISC activity by EPA. In this study, inhibition of LSD1 by GSK-LSD1 significantly reversed the beneficial effects of EPA on the TNF-α-induced decrease in ISC proliferation and differentiation of organoids. Notably, GSK-LSD1 did not impact organoid growth or ISC activity. In contrast, Zwiggelaar et al. [15] reported that GSK-LSD1 treatment inhibited the differentiation of Paneth cells and goblet cells and increased the number of Lgr5-positive ISCs in human intestinal organoids. This discrepancy might be attributed to species-specific differences. Given that GSK-LSD1 cannot selectively target LSD1 in ISCs, we employed a conditional knockout model to further explore the role of LSD1 in mediating the effects of EPA on intestinal regeneration. Knockout of LSD1 in ISCs abrogated the protective effects of EPA against DSS-induced colitis, leading to exacerbated inflammation and extensive immune cell infiltration in the colon following DSS treatment. Interestingly, body weight changes and serum inflammatory cytokine levels in LSD1-deficient mice were comparable to those in LSD1^fl/fl mice. It is possible that, in the absence of LSD1, compensatory mechanisms may be activated to maintain colonic homeostasis. Another consideration is the possibility that the body weigh changes associated with the inflammatory cytokine increase may take longer to manifest. Similarly, Parmar et al. [16] reported that mice with intestinal epithelial-specific LSD1 deficiency were more susceptible to Citrobacter rodentium infection than control mice. However, these mice displayed greater regenerative capacity after radiation damage. Furthermore, Sun et al. [36] reported that LSD1 inhibition increased Cdx2 expression via H3K4me3 methylation, which promoted differentiation into mouse intestinal epithelial cells. These findings suggest that the influence of LSD1 on intestinal regeneration is complex and warrants further investigation.

GPR120, a G-protein-coupled receptor, is highly expressed in human lungs and colon and can be activated by omega-3 fatty acids such as EPA and DHA [37]. Recent research indicates that activating GPR120 causes AMPK phosphorylation through a PLC/Ca²⁺/CaMKK-dependent mechanism in THP-1 macrophage [38]. Our previous research identified LSD1 as a downstream target of AMPK, playing a role in intestinal lipolysis [35], suggesting a potential connection between GPR120 and LSD1 in the gut. Notably, we found that EPA significantly upregulated GPR120 expression in a dose-dependent manner. Moreover, inhibition of GPR120 effectively blocked EPA-induced LSD1 expression and its subsequent impact on ISC proliferation and differentiation. These findings indicate that EPA regulates LSD1 through a GPR120-dependent pathway, highlighting the interaction between GPR120 and epigenetic factors in preserving ISC homeostasis. However, the exact mechanism by which LSD1 regulates ISC proliferation and differentiation remain obscure. Recent studies have highlighted that LSD1 influences the differentiation of intestinal epithelial secretory lineages independently of canonical bone morphogenetic protein (BMP) signaling and the microbiota [39]. In addition to its role in intestinal epithelium, recent research has indicated that muscle stem cells lacking LSD1 retain the regenerative and self-renewal capacities postinjury via the WNT pathway [17]. Thus, we hypothesize that LSD1 may regulate genes involved in the WNT signaling pathways, which is crucial for intestinal epithelial regeneration and ISC fate determination [40]. Wnt ligands bind to the transmembrane receptors Frizzleds and LRP5/6, which inhibits β-catenin phosphorylation. Subsequently, β-catenin passes through the nuclear membrane and binds with TCF/LEF response elements to regulate the transcription of downstream target genes such as c-Myc and Cyclin D1 [41]. We noted that EPA treatment upregulated Wnt3a and the Wnt target genes Axin2 and Cyclin D1 in both the colon and organoids, consistent with the findings of Zhang et al. [23]. Additionally, treatment with the PORCN inhibitor LGK974 abrogated the positive effects of EPA on DSS-induced colitis, which highlights the pivotal role of WNT signaling in mediating the beneficial effects of EPA. LGK974 also effectively suppressed EPA-induced intestinal organoid growth, which indicates that EPA accelerated ISCs-mediated intestinal epithelial development via WNT signaling. Our study indicates that LSD1 plays a crucial role in regulating the Wnt/β-catenin signaling pathway in ISCs. Specifically, we observed that knocking out LSD1 in ISCs blocked the activation of active β-catenin by EPA, suggesting that LSD1 is involved in the regulation of WNT signaling. This finding aligns with recent research demonstrating that LSD1 demethylates β-catenin, preventing its degradation and thereby maintaining its nuclear levels, which is essential for activating the transcription of target genes in the Wnt pathway [17], [42]. Although our study focused on the LSD1-Wnt axis in ISCs due to its established role in epithelial regeneration. Due to the complexity of the pathogenesis of IBD, the crosstalk between LSD1-Wnt signaling and other pathways, such as the Notch and Hippo pathways, both of which may be associated with epithelial regeneration during IBD.

Conclusions

This study identified a novel mechanism by which EPA enhances ISC-mediated colonic epithelial regeneration following injury. Specifically, EPA activates GPR120 expression, which in turn stimulates the LSD1-WNT signaling axis, promoting ISC proliferation and differentiation. However, our study still has limitations. EPA-enriched fish oil used in this study may also contain other beneficial components, such as DHA and other omega-3 fatty acids, which could influence colonic epithelial regeneration. Future research should dissect the individual and combined roles of these components to fully understand EPA's specific contributions to colonic epithelial health. Additionally, IBD is multifactorial disease influenced by genetic, environmental, immune, and microbial factors. While the DSS-induced colitis model is widely used, it may not fully capture the complexity and heterogeneity of human IBD. Further studies are needed to elucidate EPA's role in preventing colitis in human populations. Despite these limitations, our findings enhance our understanding of the role of EPA in maintaining colonic epithelial homeostasis and offer promising therapeutic avenues for treating intestinal diseases such as IBDs.

Ethical statement

The current study did not include human and clinical trials. All experiments involving animals were conducted according to the ethical policies and procedures approved by the ethics committee of Wuhan Polytechnic University, China (Approval No. EM20230512001).

CRediT authorship contribution statement

Dan Wang: Writing – original draft, Data curation, Investigation, Formal analysis, Visualization, Funding acquisition. Nianbang Wu: Investigation, Validation, Visualization, Writing – original draft. Pei Li: Investigation, Validation, Visualization. Xiaojuan Zhang: Investigation, Validation. Wenshuai Xie: Investigation. Shunkang Li: Investigation, Visualization. Ding Wang: Validation, Data curation. Yanling Kuang: Validation, Data curation. Shaokui Chen: Investigation. Yulan Liu: Conceptualization, Resources, Supervision, Funding acquisition, Project administration, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: