ABCA1 deficiency‐induced cholesterol accumulation aggravates colonic epithelial injury in knockout hamsters

Abstract

Background

ATP‐binding cassette transporter A1 (ABCA1) mediates cellular cholesterol efflux, yet its role in inflammatory bowel disease (IBD) remains incompletely understood. This study investigated the consequences of intestinal epithelial cholesterol accumulation—primarily driven by ABCA1 deficiency—on colonic inflammation and the underlying molecular mechanisms.

Methods

ABCA1 knockout (KO) and wild‐type (WT) Syrian golden hamsters were fed a high‐cholesterol high‐fat diet for 12 weeks. Colon specimens were analyzed for pathological changes, inflammatory markers, cholesterol accumulation, and gene expression profiles. Furthermore, in vitro experiments were performed using the human colon epithelial cell line Caco‐2.

Results

ABCA1 deficiency, in the context of a high‐cholesterol high‐fat diet, triggered significant colonic inflammation in KO hamsters, characterized by epithelial damage, inflammatory cell infiltration, and elevated pro‐inflammatory cytokine levels. These pathological changes were associated with impaired mucosal barrier integrity, a four‐fold increase in epithelial cholesterol accumulation, and reduced E‐cadherin, ZO‐1 and occludin expression. Furthermore, ABCA1 deficiency promoted oxidative stress and induced marked remodeling of the gut microbiota.

Conclusions

Our findings establish a pathogenic role for ABCA1 deficiency in driving colonic epithelial injury and inflammation, suggesting that ABCA1 may play an important role in the pathogenesis of IBD.

We established ABCA1‐knockout Syrian golden hamsters as a novel and physiologically relevant model for colitis research. When challenged with a high‐fat diet, ABCA1‐deficient hamsters developed exacerbated colonic inflammation and impaired mucosal barrier integrity, driven by intracellular cholesterol accumulation. The underlying mechanisms are associated with cholesterol‐induced oxidative damage in colonic epithelial cells, which leads to cell death, as well as marked remodeling of the gut microbiota.

1. INTRODUCTION

Inflammatory bowel disease (IBD) is a growing global health challenge affecting more than 7 million people worldwide. ¹ Emerging evidence highlights that metabolic reprogramming, particularly disturbances in lipid metabolism, plays a critical role in shaping intestinal inflammation. ² Among these, cholesterol metabolism has recently been recognized as a key modulator of immune responses and epithelial homeostasis in the gut. ³ , ⁴ , ⁵

ATP‐binding cassette transporter A1 (ABCA1) is a crucial membrane protein that plays a central role in cellular cholesterol homeostasis. The primary function of ABCA1 is to mediate the efflux of cellular cholesterol and phospholipids to specific extracellular acceptors. ⁶ , ⁷ It has been reported that abnormalities in ABCA1 expression can cause many disorders such as dyslipidemia, atherosclerosis, neurodegenerative disease, and so on. A line of evidence also suggests that ABCA1 expression in the intestinal epithelial cells is closely related to the pathogenesis of IBD. ⁸ This contention is initially supported by the fact that cytokines such as IL‐23 and TNF‐α, decrease ABCA1 expression and attenuate cholesterol efflux in the intestinal cells thereby affecting the progression of IBD. ⁴ , ⁹ In addition, interactions of the ABC transporters expressed at the apical surface of the epithelial cells of the intestine and gut microbes may also involve IBD, although this issue has not been investigated. ¹⁰ In humans, mutations in the ABCA1 gene cause Tangier disease (TD), a rare genetic disorder characterized by a near‐total absence of high‐density lipoproteins (HDLs) in the blood, accumulation of cholesterol in tissues (leading to enlarged orange‐toned tonsils), ¹¹ , ¹² , ¹³ and an increased risk of atherosclerotic cardiovascular disease. ¹⁴ , ¹⁵ TD is exceedingly rare, with only 200 patients reported worldwide exhibiting the trait. Despite this, a case report showed that one TD patient suffered from severe chronic diarrhea, ¹⁶ which strongly indicates that disturbed cholesterol metabolism in the intestine caused by ABCA1 deficiency directly leads to IBD.

To examine whether ABCA1 deficiency plays a causal role in IBD, we analyzed the colon pathology of ABCA1 knockout hamsters, an alternative animal model for studying lipid metabolism and atherosclerosis. ¹⁷ , ¹⁸ Compared with mice, hamsters have lipoprotein metabolism features more like humans such as plasma cholesteryl ester transfer protein and intestinal apolipoprotein B editing, ¹⁹ both of which are important in lipoprotein metabolism. In addition, the immune system of hamsters more closely resembles that of humans. For instance, certain human cytokines, such as interleukin‐12, do not function in mouse models but can cross‐react with hamster cells. ²⁰ , ²¹ Moreover, owing to the similarity in immune responses, hamsters have been widely used as animal models for infectious diseases, ²² including coronavirus infections. ²³ Our current study showed that ABCA1 deficiency in ABCA1 KO hamsters fed a high‐cholesterol high‐fat (HCHF) diet disrupts the colonic mucosal barrier and causes colitis associated with altered gut microbiota, highlighting a critical role for ABCA1 in mediating cholesterol metabolism–inflammation crosstalk during IBD pathogenesis.

2. METHODS

2.1. Animals

Generation of ABCA1 knockout Syrian golden hamsters was described previously using the CRISPR/Cas9 gene editing system. ²⁴ The hamsters were housed under a temperature‐controlled environment with a 12/12 h light/dark cycle and were fed a normal chow diet containing 20% protein and 4% fat (Beijing Keao Xieli Feed Co., Ltd., Beijing, China) and water ad libitum. To investigate the impact of ABCA1 deficiency on colons in the context of lipid metabolism disorders, 8‐week‐old male hamsters were fed either a chow diet for 20 weeks or a high‐cholesterol high‐fat (HCHF) diet for 12 weeks. The HCHF diet was formulated by supplementing a standard powdered chow diet with 0.5% cholesterol and 15% lard (Chengdu Hualanxu Biotechnology, Chengdu, China). All animal experiments were conducted in accordance with protocols approved by the laboratory animal ethics committee of Peking University (LA2015012).

2.2. Pathological analysis

The colon specimens were collected and measured for length. After fixed in 4% paraformaldehyde solution, the colon was cut into three parts designated as the ascending, transverse, and descending colons, each part was sectioned into 4–6 segments. Paraffin sections of the colon were stained with hematoxylin and eosin (HE) or immunohistochemically stained with monoclonal CD68 antibody against macrophages (BA3638, Boster, Wuhan, China). Furthermore, colonic mucosal barrier changes were assessed by immunofluorescence staining for adherens junction proteins. The sections of colon tissues were incubated with PBS containing 0.1% (v/v) Triton X‐100 for 15 min. After blocking in PBS with 3% (w/v) BSA, the sections were incubated with E‐cadherin (2255385, Invitrogen, Carlsbad, CA) or ZO‐1 (P50518‐1R2F, Abmart, Shanghai, China) antibody in the blocking buffer overnight at 4°C. Afterwards, the sections were washed with PBS containing 0.1% (v/v) Tween‐20 and then incubated with Alexa Fluor‐conjugated secondary antibodies (ab150115, Abcam, Cambridge, UK). DAPI (C1005, Beyotime, Shanghai, China) was used to display the cell nuclei and the fluorescence signals were observed under a fluorescence microscope.

2.3. RNA sequencing analysis

RNA sequencing (RNA‐seq) analysis was performed as previously described. ²⁵ In brief, total RNA was extracted from colon tissue samples. Then, 1 μg of total RNA was used for library construction. The quality of the resulting libraries was assessed using an Agilent 4200 system. Libraries passing quality control were sequenced on the NovaSeq X Plus/DNBSEQ‐T7 platforms (Chi Biotech Co. Ltd., Shenzhen, China). The reference genome index was built using STAR (v2.7.10b), and the paired‐end clean reads were aligned to the reference genome. FeatureCounts (v2.0.3) was used to quantify the reads mapped to each gene. Subsequent differential expression and functional enrichment analyses were performed in R Studio using the DESeq2, clusterProfiler, and CBNplot packages.

2.4. Western blot

Colon tissue or cellular samples were loaded onto SDS‐PAGE gel for electrophoresis. Then, proteins were transferred to a polyvinylidene difluoride membrane and hybridized with the following antibodies: ABCA1 (26564‐1‐AP, Proteintech, Wuhan, China), ABCA1 (96292, Cell Signaling Technology, Danvers, MA), ZO‐1 (A0659, ABclonal, Wuhan, China), E‐cadherin (sc‐8426, Santa cruz Biotechnology, Santa Cruz, CA), Occludin (66378‐1‐Ig, Proteintech, Wuhan, China), p‐p38 (8796‐1‐AP, Proteintech, Wuhan, China) or β‐actin (66009‐1‐Ig, Proteintech, Wuhan, China). Target protein bands were visualized by reaction with horseradish peroxidase (HRP)‐conjugated secondary antibodies followed by enhanced chemiluminescence (ECL) detection (E411‐05, Vazyme, Nanjing, China). Band intensities were quantified with the ImageJ software (National Institutes of Health, USA).

2.5. RNA isolation and quantitative real‐time PCR

To examine the expression levels of inflammatory cytokines in colonic tissues, total RNA was isolated from colonic tissues of hamsters using the FastPure Cell/Tissue Total RNA Isolation Kit (RC112‐01, Vazyme, Nanjing, China) according to the manufacturer's instructions. Quantitative real‐time PCR was carried out with gene‐specific primers detailed in Table SI.

2.6. Analysis of plasma lipoproteins and tissue lipid contents

Following a 12‐h fasting period, blood was collected from the retro‐orbital sinus of hamsters using heparinized capillary tubes and the plasma was separated. Total cholesterol (TC) and triglyceride (TG) concentrations of plasma were measured using commercial enzymatic assay kits (Biosino Bio‐technology and Science, Beijing, China). To analyze the lipoprotein distribution, fast protein liquid chromatography (FPLC) of plasma was performed as described previously. ²⁶ To estimate the lipid contents of colon tissues, 25 mg tissue samples were homogenized in 1 mL cold PBS and then 4 mL of chloroform/methanol (v:v = 2:1) was added. The mixture was vortexed for 2 min and then allowed to stand at room temperature for 30 min. After centrifugation at 3000 rpm for 30 min, the chloroform layer was transferred to a new glass tube using a glass syringe and dried under a nitrogen stream. Lipids were dissolved with 500 μL of 5% Triton X‐100, and the contents of cholesterol and triglyceride were measured according to the method described above.

2.7. siRNA interference

Human intestinal Caco‐2 cells were purchased from Solarbio (Beijing, China) and were maintained in MEM medium (C11095500BT, Gibco, Gaithersburg, MD) supplemented with 10% fetal bovine serum and 1% antibiotic‐antimycotic solution. Caco‐2 cells were cultured in 6‐well plates and transfected with negative control siNC (5′‐UUCUCCGAACGUGUCACGUTT‐3′) or siABCA1 (5′‐GGCUACCAGUUACAUUUGATT‐3′) purchased from Genepharma (Suzhou, China), using lipomaster 3000 (TL301, Vazyme, Nanjing, China) as the transfection reagent. The cells were then incubated for 24 or 48 h.

2.8. Cholesterol efflux assay

Caco‐2 cells transfected with siNC or siABCA1 were treated with 1 μg/mL NBD‐cholesterol (N1148, Invitrogen, Carlsbad, CA) for 8 h. After cholesterol loading, the cells were washed and equilibrated for 2 h with 0.2% BSA (ST023, Beyotime, Shanghai, China), and then incubated with or without 10 μg/mL ApoAI (10686‐H02H, Sino Biological, Beijing, China) as cholesterol acceptor in MEM medium containing no phenol red (51200038, Gibco, Gaithersburg, MD) with 0.2% BSA for 4 h. Then, the cells were washed with PBS twice and fixed with 4% paraformaldehyde for 10 min at room temperature. After washing with PBS, the cell nuclei were stained with Hoechst 33342, and images were captured using a fluorescence microscope (IX71, Olympus, Tokyo, Japan).

2.9. Maleimide‐labeling assay (‐SH groups detection)

Following the designated treatments, the protein samples from colon tissues, plasma or Caco‐2 cellular lysates were incubated with Alexa Fluor 680C2 maleimide (129585, Sigma‐Aldrich, St Louis, MO) at a final concentration of 10 μmol/L for 2 h at 4°C. After the reaction, the proteins were separated by SDS‐PAGE. The signal of fluorescent maleimide in the gel was captured with Amersham ImageQuant™ 800 and quantified with ImageJ software. Coomassie Brilliant Blue (P0017F, Beyotime, Shanghai, China) staining was performed to confirm the equal loading of proteins.

2.10. Detection of ROS production

Cells grown in 24‐well plates were incubated with the indicated concentration of cholesterol which was dissolved in anhydrous ethanol for 8 h and then washed twice with phosphate‐buffered saline to remove residual culture medium. Subsequently, cells were incubated with 10 μmol/L 2′,7′‐dichlorodihydrofluorescein diacetate (DCFH‐DA, 4091‐99‐0, Solarbio, Beijing, China) in the dark at 37°C for 30 min to facilitate probe uptake and intracellular oxidation by reactive oxygen species (ROS). After that, the cells were fixed with 4% paraformaldehyde solution. After washing with PBS, the cell nuclei were stained with Hoechst 33342, and fluorescence signals were captured under a fluorescence microscope.

2.11. Calcein‐AM and propidium iodide (PI) staining

To assess the viability of Caco‐2 cells under cholesterol accumulation conditions, we performed live/dead staining using Calcein‐AM and PI with a commercial assay kit (C542, Dojindo, Kumamoto, Japan). Briefly, cells were allowed to incubate with a mixture of Calcein‐AM (green) and PI (red) solution for 10–20 min and observed under a fluorescent microscope. Calcein‐AM positive green cells were considered alive, while PI‐positive red cells were considered dead.

2.12. Microbiota analysis

Microbiota analysis was performed as described previously. ²⁵ Briefly, fresh feces of hamsters from indicated groups were collected. Total genomic DNA was extracted using the CTAB/SDS method and verified by 1% agarose gel electrophoresis. The 16S rRNA gene regions were amplified with barcoded primers, purified (Qiagen Gel Extraction Kit), and used to prepare libraries (NEBNext® Ultra™ II Kit). Libraries were assessed (Qubit® 2.0, Agilent Bioanalyzer 2100) and sequenced on an Illumina NovaSeq platform. Raw reads were demultiplexed, merged (VSEARCH v2.28.1), and processed into high‐quality tags. Operational Taxonomic Units (OTUs) were clustered at 97% similarity using UPARSE in VSEARCH, while Amplicon Sequence Variants (ASVs) were generated with DADA2. Taxonomy was assigned in QIIME2 (v2024.2) against SILVA 138.2. Alpha and beta diversity were analyzed in QIIME2 and R (vegan), and differentially abundant taxa identified by LEfSe (LDA > 4) and Metastats (p < 0.05).

2.13. Statistical analysis

All data are expressed as the mean ± SD. Student's t‐test (for parametric data) and Mann–Whitney U test (for non‐parametric data) were used for statistical analyses using the GraphPad Prism 7.0 software (GraphPad Software, San Diego, CA). In all cases, statistical significance was set at p < 0.05.

3. RESULTS

3.1. ABCA1 knockout hamsters showed reduced body weight and shortened colon length when fed a high‐cholesterol high‐fat (HCHF) diet

For uncovering the role and underlying mechanisms of ABCA1‐regulated cholesterol metabolism in colitis, we generated ABCA1 knockout (KO) hamsters, as reported previously. ²⁴ As shown in Figure SIA, compared with wild type (WT) littermates, Western blot analysis of colon tissues showed that ABCA1 was undetectable in homozygous KO hamsters. To investigate the impact of ABCA1 deficiency in the pathogenesis of colitis, different dietary interventions, including a normal standard diet or a high‐cholesterol high‐fat (HCHF) diet, were performed using KO hamsters and age‐matched WT littermates (Figure 1A). On a normal chow diet, histological analysis of KO hamsters did not reveal any obvious abnormalities in the colonic tissues compared with WT hamsters (Figure SIB). Given that dysregulated lipid metabolism is a key contributor to colitis, we further investigated the impact of ABCA1 deficiency on colitis under HCHF diet conditions. On a HCHF diet, ABCA1 KO hamsters showed consistently elevated plasma triglyceride (TG) levels. Total cholesterol (TC) levels, however, were increased only at the end of the 12‐week period compared with controls (Figure SIIA). The plasma lipoprotein profile analyzed by FPLC revealed an excessive accumulation of very‐low‐density lipoproteins and an almost complete absence of high‐density lipoproteins (HDLs) in ABCA1 KO hamsters on a HCHF diet (Figure SIIB). The changes in lipid and lipoprotein profiles associated with ABCA1 deficiency closely resemble those observed in humans with ABCA1 mutations, further confirming that ABCA1 KO hamsters are a suitable model for studying metabolic disorders caused by ABCA1 dysfunction and dysregulated lipid metabolism.

FIGURE 1.

ABCA1 knockout hamsters showed reduced body weight and shortened colon length when fed a high‐cholesterol high‐fat (HCHF) diet. (A) Schematic illustration of experimental design and procedure. (B) Body weight changes in hamsters fed a HCHF diet for 12 weeks. Data are expressed as mean ± SD (n = 9 or 10; *p < 0.05, ***p < 0.001 vs. ABCA1^+/+ by Student's t‐test). (C) The changes in colon length. The colon was taken at necropsy on the last day of the 12 week experimental period and photographed (left panel). The colon length was measured in centimeters in the bar graph (mean ± SD, n = 8–9; ***p < 0.001 vs. ABCA1^+/+ by Student's t‐test).

Compared with WT hamsters, HCHF‐fed KO hamsters ceased to gain body weight (Figure 1B). Gross examination revealed that the total colon length of ABCA1 KO hamsters was significantly shorter than that of WT controls (Figure 1C). Furthermore, histological examination revealed that the colon wall of ABCA1 KO animals became thinner due to reductions in both the mucosal and muscle layers (Figure 2A). The mucosal epithelial cells showed a reduction in goblet cells, loss of microvilli, and crypt distortion. The mucosal surface was often detached or eroded and was covered by inflammatory cells and necrotic tissue (Figure 2B). In the lamina propria, inflammatory cells, including macrophages and plasma cells, were prominent, and many macrophages exhibited a foamy appearance (Figures 2B and SIIIA). Inflammatory infiltrates were also present in the submucosa. Taken together, these histological features suggest that colitis is indeed present in ABCA1 KO hamsters challenged with HCHF diet. This notion is further supported by observations of marked upregulation in the mRNA expression of inflammatory cytokines such as tumor necrosis factor‐α (TNF‐α), interleukin‐1β (IL‐1β), and interleukin‐6 (IL‐6) in the colons of ABCA1 KO hamsters (Figure SIIIB).

FIGURE 2.

ABCA1^−/− hamsters exhibit exacerbated colitis on a HCHF diet. Representative images of HE staining of colon tissues of ABCA1^+/+ and ABCA1^−/− hamsters on a HCHF diet. (A) Evaluation of colonic wall integrity: (a) ABCA1^+/+ hamsters show a relatively normal colonic architecture with a thick mucosa; (b) higher magnification of panel a; (c) ABCA1^−/− hamsters exhibit destruction of the mucosal and muscle layers, accompanied by marked thinning of the colonic wall; (d) higher magnification of panel c. (B) Pathological alterations in ABCA1^−/− hamsters fed a HCHF diet, including: (a, b) detachment or erosion of the mucosal surface, reduced goblet cell numbers, loss of microvilli, and crypt distortion; (c) mucosal surface lesions covered by inflammatory exudate and necrotic tissue; (d) macrophages and plasma cells infiltration in the lamina propria, with many macrophages exhibited a foamy appearance.

3.2. ABCA1 deficiency leads to dysregulation of lipid metabolism and cholesterol accumulation

Based on the well‐established role of ABCA1 in lipid metabolism, we hypothesized that the exacerbated colitis phenotype observed in KO hamsters on a HCHF diet condition was primarily attributable to a disruption in cholesterol homeostasis. Indeed, quantitative analysis of colon lipid content revealed that cholesterol content was increased 4‐fold, whereas triglyceride content was reduced 5‐fold compared with controls (Figure SIV). To investigate the role of ABCA1 in colonic lipid metabolism, we performed transcriptome sequencing of colonic tissues from ABCA1 KO and WT hamsters. Comparative analysis revealed substantial differences in global gene expression (Figure 3A). Gene Set Enrichment Analysis (GSEA) further indicated significant enrichment of pathways related to cholesterol storage and cellular stress responses in ABCA1 KO hamsters compared with WT controls (Figure 3B). To examine the possible mechanism responsible for an imbalance in cholesterol homeostasis observed in ABCA1 KO hamsters, we performed cholesterol efflux assay in vitro using human intestinal Caco‐2 cells, which express ABCA1 protein. siRNA‐mediated downregulation of ABCA1 expression, with successful knockdown verified by Western blot, led to the accumulation of nitrobenzoxadiazole (NBD)‐labeled cholesterol in Caco‐2 cells (Figure 3C,D). These findings indicate that ABCA1 deficiency not only disrupts systemic lipid metabolism—evidenced by altered plasma lipid profiles—but also exacerbates local cholesterol accumulation in the colon, which may underlie the further aggravation of colitis observed under HCHF dietary conditions.

FIGURE 3.

ABCA1 deficiency leads to dysregulation of lipid metabolism and cholesterol accumulation. (A) RNA sequencing result of colon tissues reveals substantial differences in overall gene expression in ABCA1^−/− hamsters compared with wild‐type controls. (B) Gene Set Enrichment Analysis (GSEA) of lipid metabolism pathways between ABCA1^+/+ and ABCA1^−/− hamsters. (C) ApoAI‐mediated cholesterol efflux assay of colon epithelial cell line Caco‐2 cells with ABCA1 silencing. Caco‐2 cells were transfected with control siRNA (siNC) or ABCA1‐targeting siRNA (siABCA1) for 24 h, then preloaded with 1 μg/mL NBD‐cholesterol (green) for 8 h, and then incubated with or without ApoAI for 4 h. Caco‐2 cell nuclei were stained with Hoechst (blue). (D) Quantification of fluorescence intensity of NBD‐cholesterol in Caco‐2 cells. Data are expressed as mean ± SD (n = 4; ***p < 0.001 vs. siNC without ApoAI by Student's t‐test). Western blot showing ABCA1 expression in Caco‐2 cells transfected with siNC or siABCA1 are shown at top right.

3.3. ABCA1 deficiency disrupts the junctional integrity of the colonic mucosal barrier

The colonic mucosal epithelial barrier serves as the first line of defense against luminal environmental factors. When this barrier is compromised, it facilitates the translocation of luminal antigens—including pathogenic microbes and their toxins—across the mucosal layer into the systemic circulation, thereby disrupting the delicate balance of gut homeostasis and significantly increasing the susceptibility to systemic inflammation. We evaluated whether the absence of ABCA1 affects the integrity of the colonic barrier. Immunofluorescence staining showed that the abundance of E‐cadherin (Figure 4A) and ZO‐1 (Figure SVA), molecular markers of intestinal barrier integrity, were significantly reduced in the KO hamsters. The regeneration of colonic epithelial cells plays a vital role in restoring the integrity of the intestinal barrier during the initiation of colitis. However, our in vitro studies demonstrated that ABCA1 silencing in Caco‐2 cells under the condition of cholesterol overload results in substantial cell death (Figures 4B,C and SVB), accompanied by decreased levels of the tight junction proteins E‐cadherin, ZO‐1, and occludin (Figure 4D). These findings suggest that the accumulation of cholesterol, caused by the absence of ABCA1, which in turn affects the survival and proliferation of epithelial cells, may represent a potential underlying mechanism contributing to the disruption of the intestinal barrier.

FIGURE 4.

ABCA1 deficiency disrupts the junctional integrity of the colonic mucosal barrier. (A) Representative images of immunofluorescence staining of the zippering junction with E‐cadherin. (B, C) Effects of ABCA1 silencing on cholesterol crystals (CHO)‐induced cell injury. Caco‐2 cells were transfected with siNC or siABCA1 for 24 h, and then exposed to 1 mmol/L CHO or not for 24 h. The cell viability was determined using Calcein‐AM/PI staining (B) and CCK8 assay (C). The data in panel C are expressed as mean ± SD (n = 4; ***p < 0.001 vs. siNC without CHO by Student's t‐test). (D) ABCA1 silencing aggravates CHO‐mediated barrier dysfunction in Caco‐2 cells. Caco‐2 cells were transfected with siNC or siABCA1 for 24 h, and then exposed to 1 mmol/L CHO for 6 h. Cellular lysates were assayed for E‐cadherin, Occludin and ZO‐1 expression. The respective densitometric quantitations of bands are presented in the bar graphs to the right of the blots. Data are expressed as mean ± SD (n = 3; *p < 0.05, **p < 0.01, ***p < 0.001 vs. siNC without CHO by Student's t‐test).

3.4. Cholesterol accumulation caused by ABCA1 deficiency induces oxidative stress in colons

Cholesterol accumulation not only induces inflammation by activating inflammasomes, but also disrupts cellular redox homeostasis, ultimately leading to organ damage. To explore the underlying causes of epithelial cell death and barrier disruption induced by cholesterol accumulation due to ABCA1 deficiency, we quantified the levels of free sulfhydryl (‐SH) groups and hydrogen peroxide (H₂O₂) in the colonic tissues. Notably, the ‐SH content, a critical antioxidant defense, was significantly reduced both in the colons (Figure 5A,B) and plasma (Figure SVI) of KO hamsters, while H₂O₂ levels were markedly elevated (Figure 5C). The results obtained by the DCFH method provide more direct evidence that, upon siRNA‐mediated downregulation of ABCA1 expression, the level of reactive oxygen species (ROS) in Caco‐2 cells stimulated with cholesterol crystals was dramatically increased (Figure 5D,E). Moreover, impaired ABCA1 expression aggravated cholesterol crystal‐induced sulfenic acid formation (shown as the reduction of ‐SH groups) in cellular proteins (Figure 5F,G) and activates the p38 MAPK pathway (Figure 5H,I), with phosphorylated p38 (p‐p38) serving as a well‐established marker of oxidative stress‐induced cellular damage. Collectively, these results suggest that ABCA1‐deficient hamsters suffered more severe oxidative stress under HCHF diet conditions compared to WT counterparts.

FIGURE 5.

Cholesterol accumulation caused by ABCA1 deficiency induces oxidative stress in colons. (A) Detection of sulfhydryl (‐SH) groups in colon tissues. Proteins extracted from colon tissues of ABCA1^+/+ and ABCA1^−/− hamsters on a HCHF diet were allowed to react with fluorescent maleimide and the level of ‐SH was detected using Western blot analysis. Coomassie blue staining of the gels showed equal amounts of samples were loaded. (B) Quantitative analysis of relative expression level of ‐SH groups. The densitometric values are presented as fold‐changes from ABCA1^+/+. The data are expressed as mean ± SD (n = 4; ***p < 0.001 via Student's t‐test). (C) Hydrogen peroxide (H₂O₂) detection of the colon tissues of ABCA1^+/+ and ABCA1^−/− hamsters. The data are expressed as mean ± SD (n = 4; *p < 0.05 vs. ABCA1^+/+ by Student's t‐test). (D, E) Effects of ABCA1 silencing on cholesterol (CHO)‐induced ROS production. Caco‐2 cells were transfected with siNC or siABCA1 for 24 h, and then exposed to 1 mmol/L CHO for 24 h. The levels of ROS were detected by DCFH‐DA (green). Nuclei were counterstained with Hoechst (blue). Representative images are shown in panel D. Note the markedly increased fluorescence intensity in the siABCA1 plus CHO group compared to the siNC plus CHO group. The densitometric intensity of the cellular fluorescence was measured, and the data are presented in panel E (mean ± SD, n = 4; ***p < 0.001 vs. siNC without CHO by Student's t‐test). (F) Changes of ‐SH groups in Caco‐2 cells transfected with siNC or siABCA1 were examined, followed by treatment without or with CHO. An equal amount of proteins from cells was subjected to Western blot analysis for ‐SH groups (left image). Coomassie blue staining of the gel was used as a loading control (right image). (G) Quantitative analysis of relative expression level of ‐SH groups in Caco‐2 cells. Data are presented as mean ± SD (n = 4; **p < 0.01, ***p < 0.001 vs. siNC without CHO via Student's t‐test) (H, I) Effect of ABCA1 silencing on CHO‐induced activation of p38. Cells were transfected with siNC or siABCA1 and then treated without or with CHO. The cellular proteins were extracted and subjected to Western blot analysis for p‐p38, and β‐actin served as a loading control (H). The densitometric analysis of the blot in panel H is shown in panel I. The results are expressed as the percentage of change relative to siNC without CHO control. Data are presented as mean ± SD from three separate experiments (n = 3; *p < 0.05, **p < 0.01, ***p < 0.001 via Student's t‐test).

3.5. ABCA1 knockout hamsters display more pronounced gut microbiota dysbiosis

Considering the close relationship between gut microbiota homeostasis and IBD, it is essential to investigate how ABCA1 deficiency‐induced cholesterol metabolic dysregulation affects the gut microbiota, particularly in terms of microbial diversity and the balance between beneficial and pathogenic bacterial populations. Thus, we performed a microbiota analysis and found that the KO group exhibited elevated abundances of potential pro‐inflammatory or opportunistic pathogenic taxa, including Escherichia‐Shigella (Proteobacteria) and Ruminococcaceae_UCG‐014 (Actinobacteria) (Figure 6A–C). Differential species analysis revealed that Escherichia‐Shigella was enriched in the KO group, whereas Bacteroides vulgatus was more abundant in the WT group (Figure 6D). Accordingly, the KO group displayed an “inflammation‐associated” microbial profile, while the WT group was characterized by a “metabolically beneficial” community structure (Figure 6E).

FIGURE 6.

ABCA1 knockout hamsters display more pronounced gut microbiota dysbiosis. (A) Phylogenetic cladogram showing the taxonomic distribution of gut microbiota in ABCA1^+/+ and ABCA1^−/− hamsters, with LEfSe analysis highlighting differentially abundant clades. (B) Principal coordinate analysis (PCoA). (C) LEfSe analysis identifying microbial biomarkers significantly enriched in ABCA1^+/+ (blue) or ABCA1^−/− (red) groups at the genus level. (D) Bar plot displaying the differential abundance and group means of bacterial genera between ABCA1^+/+ and ABCA1^−/− hamsters. (E) Comparative analysis of predicted functional pathways (based on PICRUSt2) in ABCA1^+/+ and ABCA1^−/− groups, showing relative abundance and intergroup differences.

4. DISCUSSION

As a pivotal regulator in the reverse cholesterol transport pathway, ABCA1 functions as the gatekeeper for the removal of excess cholesterol from peripheral tissues. ²⁷ Although cholesterol homeostasis has been extensively investigated in various pathological conditions, ²⁸ the role of ABCA1 and ABCA1‐mediated cholesterol efflux in colitis remains poorly understood. Although mouse models have been the primary animals used to study human diseases, ABCA1‐genetically modified mice differ from humans in both their lipid profiles and lipoprotein metabolism pathways. For instance, ABCA1 knockout mice exhibit a dramatic ~70% reduction in plasma TC levels, ²⁹ whereas TD patients with ABCA1 deficiency show only a moderate decrease in TC. Moreover, plasma TG levels remain largely unchanged in ABCA1 KO mice, ³⁰ in stark contrast to TD patients, who display markedly elevated TG levels. Given the limitations of mouse models for ABCA1‐related lipid metabolism studies, and based on our prior experience with gene‐edited hamster models and lipid metabolism research, we opted to use the hamster model for this study.

In the current study, we demonstrated for the first time that ABCA1 deficiency confers susceptibility to, and exacerbates, HCHF diet‐induced colitis. Several mechanisms may underlie the pathological role of ABCA1 in colitis. In the absence of ABCA1, cholesterol homeostasis in colonic epithelial cells is disrupted, leading to abnormal cholesterol accumulation and consequent cellular lipotoxicity, ultimately causing epithelial cell injury and death. This event may also compromise the barrier integrity of epithelial cells, likely facilitating bacterial translocation across the impaired barrier and triggering the recruitment of inflammatory cells into the mucosa. This hypothesis is supported not only by the pathological changes observed in the colon (reduced E‐cadherin and ZO‐1 expression) but also by in vitro studies using cultured intestinal epithelial cells, in which ABCA1 silencing led to reduced cell viability and decreased expression of E‐cadherin, ZO‐1, and occludin.

It should be noted that in the colon of ABCA1 KO hamsters, the levels of free sulfhydryl (‐SH) groups and hydrogen peroxide were altered compared with controls, suggesting that oxidative stress may be involved in the inflammatory cascade of colitis. This hypothesis was further substantiated by experimental evidence showing a marked reduction of ‐SH groups in both plasma and colonic tissues of KO hamsters, which was concomitant with a significant elevation in H₂O₂ levels within the colonic tissue. Free ‐SH groups neutralize oxidants to less toxic forms and thus regulate cell function and fate. ³¹ In cultured colonic epithelial cells lacking ABCA1 and stimulated with cholesterol, reactive oxygen species levels were significantly elevated, while ‐SH group levels were markedly reduced, indicating that cholesterol‐induced oxidative stress is further exacerbated by ABCA1 deficiency. Consistent with our previous study, ³² the current results suggest that identifying an appropriate antioxidant therapy may serve as a promising approach for the management or mitigation of colitis. In addition, the restriction of our mechanistic investigation to the ROS/p38 pathway constitutes a notable limitation of this work. Future research will be required to elucidate whether additional pathways participate in the colonic epithelial injury elicited by cholesterol overload, and to evaluate whether targeted blockade of ROS/p38 signaling is effective in ameliorating the diet‐induced colonic lipotoxic injury observed in the setting of ABCA1 deficiency.

Furthermore, ABCA1 knockout hamsters exhibited more pronounced gut microbiota dysbiosis under high‐fat diet conditions. Although it remains to be determined whether microbiota dysbiosis is a consequence or a cause of colitis, it is likely that this dysbiosis enhances the vicious cycle of the inflammatory process.

Beyond the impact of intracellular cholesterol accumulation, the pronounced depletion of high‐density lipoprotein cholesterol (HDL‐C) resulting from impaired RCT in KO hamsters may represent an equally significant contributor to the dysregulation of the colonic microenvironment. Indeed, strategies aimed at specifically elevating HDL levels have emerged as a promising therapeutic avenue for the treatment of colitis. ³³ , ³⁴ Moreover, the present study did not fully assess the role of systemic hyperlipidemia in colonic epithelial injury. Future research is needed to clarify how ABCA1 deficiency‐induced increases in plasma TC and TG affect colonic tissue. In conclusion, our results demonstrate that the interplay between ABCA1 deficiency and dietary cholesterol overload disrupts colonic homeostasis and promotes colon inflammation. This highlights that ABCA1‐mediated cholesterol efflux is particularly critical for maintaining gut integrity under hyperlipidemic conditions. Therefore, evaluating ABCA1 expression in colitis patients and developing therapies that enhance ABCA1 function may offer novel therapeutic opportunities for inflammatory bowel disease (IBD).

AUTHOR CONTRIBUTIONS

Chang Liu: Project administration; writing – original draft. Jingjie Zheng: Data curation; validation. Sijing Shi: Formal analysis; visualization. Jianwen Chen: Methodology. Geroge Liu: Methodology. Wei Huang: Supervision. Xunde Xian: Resources. Yuhui Wang: Methodology; resources. Jianglin Fan: Conceptualization; supervision; writing – review and editing. Ping Zhu: Conceptualization; funding acquisition; supervision. Xiawen Yang: Conceptualization; funding acquisition; supervision; writing – review and editing.

FUNDING INFORMATION

This work was supported by the Guangdong Major Project of Basic and Applied Basic Research (2023B0303000005) and Guangdong Basic and Applied Basic Research Foundation (2025A1515012787).

CONFLICT OF INTEREST STATEMENT

Xunde Xian is an editorial board member of Animal Models and Experimental Medicine (AMEM) and one of the authors of this article. To minimize bias, he was excluded from all editorial decision making related to the acceptance of this article for publication.

ETHICS STATEMENT

All experiments were performed under the principle of experimental animal care (NIH publication No. 85Y23, revised 1996) and were approved by the laboratory animal ethics committee of Peking University (LA2015012).

Supporting information

Figure SI. Characterization of colon tissues of ABCA1 knockout hamsters on a normal chow diet. (A) Western blotting analysis of ABCA1 expression in colon tissues of ABCA1^+/+ and ABCA1^−/− hamsters. (B) Pathological changes of colon tissue of hamsters fed a normal chow diet for 20 weeks. Representative images of hematoxylin–eosin (HE) staining of colon tissues are shown.

Figure SII. Analysis of plasma lipids and lipoprotein profiles of hamsters on a HCHF diet. (A) Plasma total cholesterol (TC) and triglyceride (TG) levels were determined from ABCA1^+/+ and ABCA1^−/− hamsters at the indicated time points during HCHF diet feeding. Data are presented as mean ± SD (n = 9–10; ***p < 0.001 vs. ABCA1^+/+ by Student's t‐test). (B) The distribution of TC and TG contents in lipoprotein fractions of pooled plasma samples from ABCA1^+/+ and ABCA1^−/− hamsters after 12 weeks of HCHF diet feeding were investigated by fast protein liquid chromatography (FPLC).

Figure SIII. Detection of MΦ infiltration and inflammatory cytokine expression in colon tissues of KO hamsters on a HCHF diet. (A) Representative images of immunohistochemical staining of macrophages (MΦ) in colonic tissue sections from ABCA1^+/+ and ABCA1^−/− hamsters on a HCHF diet. (B) Inflammatory cytokine expression levels of tumor necrosis factor‐α (TNF‐α) and interleukin‐1β (IL‐1β) and interleukin‐6 (IL‐6) in colon tissues of hamsters on a HCHF diet were detected by quantitative real‐time PCR. Data for TNF α and IL 1β are presented as mean ± SD (n = 6; *p < 0.05 vs. ABCA1^+/+ by Student's t‐test). For IL 6, data are presented as mean ± SD and analyzed by Mann–Whitney U test (n = 6; *p < 0.05 vs. ABCA1^+/+).

Figure SIV. Analysis of lipid contents in colon tissues of hamsters on a HCHF diet. TC and TG levels in colon tissues of ABCA1^+/+ and ABCA1^−/− hamsters after 12 weeks of HCHF diet feeding were determined. Data for colon TC are presented as mean ± SD (n = 5–7; ***p < 0.001 vs. ABCA1^+/+ by Student's t‐test). Data for colon TG are presented as mean ± SD (n = 5–7; **p < 0.01 vs. ABCA1^+/+ by Mann–Whitney U test).

Figure SV. Analysis of intestinal barrier integrity and colonic epithelial cells survival with cholesterol stimulation. (A) Representative images of immunofluorescence staining of the tight junction protein ZO‐1 in colon tissues of ABCA1^+/+ and ABCA1^−/− hamsters on a HCHF diet.

(B) Representative images show the morphological changes in Caco‐2 cells transfected with siNC or siABCA1 after treated with different concentration of CHO (0–1000 μM). And quantitative analysis of total protein levels of Caco‐2 cells transfected with siNC or siABCA1 after treated with different concentration of CHO (mean ± SD, n = 3; ***p < 0.001 vs. siNC plus CHO 0 μM by Student's t‐test).

Figure SVI. Plasma ‐SH groups examination of hamsters on a HCHF diet. (A) Changes in sulfhydryl (‐SH) groups in the plasma of ABCA1^+/+ and ABCA1^−/− hamsters on a HCHF diet. (B) Quantitative analysis of relative expression level of ‐SH groups. The densitometric values are presented as fold of ABCA1^+/+. The data are mean ± SD (n = 4; ***p < 0.001 via Student's t‐test).

Table SI. Primers used for quantitative real‐time PCR.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.