Distinct Single-Cell Expression Pattern of the Soluble Epoxide Hydrolase and Cytochrome p450 Epoxygenases in Human and Mouse Small Intestinal Epithelia

Abstract

Lipid signaling molecules are essential for maintaining intestinal mucosal barrier homeostasis. Over 80% of polyunsaturated fatty acids (PUFAs) are metabolized through the cytochrome P450 epoxygenase (CYP)-soluble epoxide hydrolase (sEH) axis, generating bioactive epoxy and diol fatty acids. However, the expression patterns and functional roles of these enzymes in the intestinal epithelium remain poorly defined. To address this, we used RNAscope in situ hybridization and publicly available single-cell RNA sequencing (scRNAseq) datasets to map the cell-type-specific expression of PUFA-metabolizing enzymes in the small intestine. In humans, we identified three major epithelial expression patterns: (1) stem-cell-dominant (CYP2E1), (2) enterocyte-dominant (CYP1A1), and (3) widespread expression across epithelial subsets for key CYP epoxygenases (CYP2C8, 2C9, 2C18, 2C19, 2J2, 2S1), sEH, and additional CYPs (3A4, 4A11, 4F2, 4F3, 4F8, 4F12). In mice, five distinct expression patterns were found: (1) stem cell and transit-amplifying cells (CYP2E1), (2) transit-amplifying dominant (CYP4F18), (3) transit-amplifying and enterocytes (CYP2C44), (4) enterocyte-dominant (CYP1A1), and (5) broadly expressed across all epithelial clusters (sEH, CYP2B10, 2S1, 2C55, 4F13, 4F16). Furthermore, in a radiation-induced small-bowel injury-regeneration model, we observed dynamic changes in CYP expression patterns. These findings provide the first high-resolution single-cell atlas of CYP–sEH axis enzymes in the intestinal epithelium, offering key insights into their potential roles in epithelial differentiation, injury response, and mucosal barrier integrity. This foundational work enables future studies to define the biological functions and therapeutic relevance of PUFA-derived lipid mediators in intestinal health and disease:

Graphical Abstract.

This is a visual representation of the abstract.

NEW & NOTEWORTHY: This study is the first to map, at single-cell resolution, the expression patterns of cytochrome P450 epoxygenases and soluble epoxide hydrolase (sEH) across human and mouse small intestinal epithelial lineages. By combining RNAscope in situ hybridization with scRNAseq data analysis, we identify distinct epithelial cell-type-specific expression patterns of PUFA-metabolizing enzymes and demonstrate their dynamic regulation in a radiation-induced injury model. These findings provide foundational insight into the CYP-sEH lipid signaling axis in intestinal mucosal barrier biology and regeneration.

Introduction

The intestinal mucosal barrier plays a critical role in defense, healing, and maintaining homeostasis. ¹ It comprises three key components: (1) Physical barrier—epithelial cells and lamina propria vascular/immune cells; (2) Chemical barrier—intestinal fluid, mucin, and digestive enzymes; and (3) Endogenous regulatory factors. Cyclooxygenase (COX)-mediated eicosanoids, particularly PGE2, are well-established endogenous regulators of intestinal barrier homeostasis. ²

In general, there are three dominant metabolic pathways of polyunsaturated fatty acids (PUFAs), and each of these pathways plays a crucial role in PUFA metabolism, contributing to physiological and pathological processes. ³ COX pathway converts PUFAs into prostaglandins (PGs), thromboxanes (TXs), and prostacyclins (PGIs), which plays a key role in inflammation, vasodilation, and platelet aggregation. Lipoxygenase (LOX) pathway converts PUFAs into leukotrienes (LTs), hydroxyeicosatetraenoic acids (HETEs), and lipoxins, which are involved in immune response, inflammation, and resolution of inflammation. Cytochrome P450 (CYPs) mediate PUFA metabolism through two distinct enzymatic activities: (1) epoxygenase activity, which converts PUFAs into epoxy fatty acids (EpFAs), and (2) hydroxylase activity, which generates hydroxy fatty acids (hydroxy-FAs). ⁴ The final product depends on the specific CYP enzyme and the type of Omega-3 or Omega-6 PUFA substrate. CYP epoxygenase pathway converts PUFAs into EpFAs (e.g., EETs, EEQs, EDPs), which have anti-inflammatory, vasodilatory, and protective effects.^{4 –6} However, under physiologic condition, these EpFAs are rapidly hydrolyzed into less-active diol fatty acids (DFAs) by soluble epoxy hydrolase (sEH), which serves as the sole enzyme for EpFA metabolism in vivo,^5,7 as summarized in Fig. 1.

Figure 1.

Dominant epoxy-diol metabolic pathway of polyunsaturate fatty acids–arachidonic acid in addition to prostaglandin H2 synthase–cyclooxygenase and lipoxygenase pathways.

In general, CYP epoxygenases convert arachidonic acid (AA) into four regio-isomers of epoxyeicosatrienoic acids (EETs): 5,6-EET, 8,9-EET, 11,12-EET, and 14,15-EET. Some CYPs display regioselectivity and stereoselectivity, meaning the same CYP isoform can act differently, favoring either epoxygenase or hydroxylase activity, depending on the substrate. These regioselective and stereoselective differences could ultimately influence intestinal epithelial physiology, as numerous studies suggest that EpFAs exert biological activity in a regioselective and stereoselective manner. Therefore, identification of expression pattern of multiple CYPs in the intestinal epithelium warrants further investigation to elucidate their biological active metabolites and their functional roles in intestinal epithelial lineage development, homeostasis, and barrier defense.

Among the multiple enzymatic pathways involved in PUFA metabolism, the CYP–soluble epoxide hydrolase (sEH) axis represents a predominant route, accounting for more than 80% of PUFA metabolic conversion in mammalian tissues.^6,8,9 This pathway generates a diverse spectrum of bioactive epoxy and DFAs that exert potent effects on epithelial cell proliferation, differentiation, and immune modulation. Despite its quantitative dominance and biological relevance, the cell-type-specific expression patterns and functional significance of CYP–sEH enzymes in the intestinal epithelium remain largely undefined. Understanding this pathway is therefore crucial for elucidating lipid signaling mechanisms that underlie epithelial integrity and the response to injury.^8,9 Since CYP-sEH enzymes play a key role in the metabolism of PUFAs, leading to the production of lipid mediators, the important question also remains: what are the single-cell expression patterns and biological roles of CYP epoxygenases and sEH in the intestinal mucosal barrier?

In humans, principal CYP epoxygenases include CYP2C8, CYP2C9, CYP2C18, CYP2C19, and CYP2J2, which primarily metabolize PUFAs into epoxylipids known as epoxyeicosanoids (EpFAs). ¹⁰ While CYP2S1 has been implicated in some lipid-metabolism pathways, its role as a primary epoxygenase is less established than the others. ¹¹ Examples include AA is directly converted to EETs, eicosapentaenoic acid (EPA) to epoxyeicosatetraenoic acids (EEQs/EpETEs), and docosahexaenoic acid (DHA) to epoxydocosapentaenoic acids (EDPs). In addition, the following CYP enzymes exhibit both hydroxylase and epoxygenase activity: CYP1 (1A1, 1A2), CYP2 (2E1), CYP3 (3A4), and CYP4 (4A11, 4F2, 4F3A, 4F3B, 4F8, 4F12, 4F22). ¹⁰ These enzymes metabolize PUFAs into both hydroxy fatty acids (HETEs) and EpFAs, contributing to diverse physiological and pathophysiological processes.

In mice, CYP epoxygenases are not as clearly classified. However, multiple CYP enzymes, including CYP1a1, 1a2, 3a4, 4a10, 4a11, 4a12a, 4a12b, 46a1, 2b10, 2b23, 2c8, 2c9, 2c18, 2c19, 2c29, 2c38, 2c39, 2c44, 2c50, 2c55, 26c1, 2e1, 4f2, 4f3a, 4f3b, 4f8, 4f12, 4f13, 4f16, 4f18, 4f22, 2j2, and 2s1, exhibit epoxygenase activity. ¹² However, whether they also possess hydroxylase activity remains unclear.

The small-bowel epithelial barrier basically follows a crypt–villus differentiation axis: intestinal stem cells (ISCs) are at the crypt base, transit-amplifying (TA) cells in the proliferative compartment, and five differentiated epithelial cell types—enterocytes, goblet cells, Paneth cells, enteroendocrine cells (EECs), and tuft cells. Previous studies using bulk analysis methods (e.g., western blot, RT-PCR, immunohistochemistry) suggest that CYP epoxygenases and sEH are highly expressed in the small bowel. ¹³ However, their single-cell expression patterns remain undefined. In this study, we utilized RNAscope mRNA in situ hybridization and publicly available single-cell RNA sequencing (scRNAseq) datasets to comprehensively map the expression patterns of CYP epoxygenases and sEH in human and mouse small-bowel epithelium along the crypt–villus axis.^14,15 In addition, using publicly available scRNAseq dataset of a radiation-induced small-bowel injury-regeneration model (3 days post-12 Gy whole-body radiation), we further analyzed changes in the single-cell expression patterns of CYP epoxygenases and sEH. ¹⁶ These findings provide a foundation for future studies on the biological roles of these enzymes in intestinal barrier function and repair.

Materials and Methods

RNAscope mRNA In Situ Hybridization Assay

The manual RNAscope multiplex fluorescent mRNA in situ assay was performed using the ACD RNAscope Multiplex Fluorescent V2 Assay (Cat. No. 323270, ACD, Newark, CA) following the manufacturer’s protocol. Proper RNAscope 3-plex negative control probes (Cat. No. 320871) and mouse-specific RNAscope 3-plex positive control probes (Cat. No. 320881) were included. The following target probes were also obtained from ACD: CYP1a1 (Cat. No. 464611), CYP2e1 (Cat. No. 402781), CYP2b10 (Cat. No. 500821), EPHX2 (Cat. No. 558701), LGR5 (Cat. No. 312171), and Ki-67 (Cat. No. 416771).

The experimental procedure was briefly as follows: small-bowel tissue sections were deparaffinized and treated with Pretreat 1 (pretreatment kit, ACD, CA) for 10 minutes at room temperature (RT). Target retrieval was performed for 15 minutes at 100–104°C, followed by protease treatment for 15 minutes at 40°C. Probes were hybridized for 2 hours at 40°C, followed by RNAscope amplification and TSA Vivid Dyes detection. Stained slides were analyzed using a fluorescent microscope and scored according to the RNAscope manufacturer’s scoring system.

RNAscope CI-KIT, positive control CI-PPIB, and negative control dapB probe signal results (staining intensity) were measured using Image pro-plus software (visible at 40× magnification). The scoring of mRNA expression was independently performed by three pathologists (GYY, JL, and LYS), with a concordance rate of 87%.

For quantification, images were analyzed using Image pro-plus software (visible at 40× magnification). Regions of interest (ROIs) were manually defined based on the histologic architecture (e.g., crypt base, TA zone, villus epithelium). Individual fluorescent puncta representing single mRNA molecules were quantified either automatically using the QuPath spot-detection algorithm or semi-quantitatively scored by trained observers. Expression levels were reported as average intensity per cell or percentage of positive cells per ROI. For comparisons between groups or cell types, at least three biological replicates per condition and five representative fields per sample were analyzed. Statistical analyses were performed using GraphPad Prism v10, and data are presented as mean ± standard error of the mean.

scRNAseq Dataset and Analysis

ScRNAseq Datasets

Three publicly available scRNAseq datasets were used in our study—(1) Mouse small intestinal cells (Haber et al., Nature 2017): ¹⁴ The dataset was generated by Aviv Regev’s lab and accessed through the GEO repository (GSE92332) https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE92332. ¹⁴ The matrix files titled “Atlas xx,” corresponding to total small intestinal CD45⁻ cell data, were used for analysis, as shown in Fig. 1. (2) Human small intestinal cells (Burclaff et al., Cell Mol Gastroenterol Hepatol 2022): ¹⁵ The dataset was generated by Scott T Magness’s lab and accessed through the GEO repository (GSE185224) https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE185224. ¹⁵ The matrix file titled “GSE185224_clustered_annotated_adata_k10_lr0.92_v1.7.h5ad” was used for analysis. (3) Radiation-induced mouse small-bowel epithelial damage-regeneration model (Ayyaz et al., Nature 2019): ¹⁶ The dataset was generated by Alex Gregorieff’s lab and accessed through the GEO repository (GSE123516) https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE123516. ¹⁶ The matrix file titled “GSE123516_RAW” was used for analysis.

ScRNAseq Data Analysis

All downstream analyses were performed using the Seurat R package (v4.3.0.1) in R version 4.3.1. Matrix files generated from CellRanger were used as inputs into Seurat.

Quality Control and Filtering, Data Standardization and Integration Across scRNAseq Datasets

To minimize bias and ensure comparability across datasets, all scRNAseq data were reprocessed using a uniform computational workflow. Raw or count matrix files were imported into Seurat (v5.0) in R, and gene expression values were log-normalized using the same scaling factor. To ensure high-quality data and to ensure consistent filtering criteria, the following filtering criteria were applied: (1) Cells with low-quality metrics (e.g., <500 unique molecular identifiers [UMIs]) were removed. (2) Cells with fewer than 250 unique genes detected were excluded. (3) Cells with fewer than seven genes per UMI were filtered out. And (4) cells with more than 20% mitochondrial gene expression were excluded (except for data from Haber et al., 2017, which lacked mitochondrial gene annotations).

For the dataset from Haber et al. (2017), all available libraries were combined into a single dataset containing 3623 cells for downstream analysis.

The top 2000 highly variable genes were identified within each dataset, followed by data scaling and centering to standardize gene expression distributions. To correct for batch effects and platform-specific variability, datasets were integrated using the Seurat integration pipeline (FindIntegrationAnchors/IntegrateData) based on canonical correlation analysis (CCA). Dimensionality reduction and clustering were then performed on the integrated dataset using the same parameters (e.g., PCA with 30 components, shared nearest neighbor [SNN] clustering at resolution 0.6).

For cell-type annotation, we used canonical intestinal epithelial marker genes (e.g., Lgr5, Alpi, Muc2, Chga, Lyz1) applied uniformly across datasets to ensure consistent identification of corresponding epithelial cell populations. Annotation accuracy was further validated by cross-referencing with published cluster identities from the original studies. This standardized approach allowed for robust cross-species and cross-condition comparison while minimizing dataset-specific biases in clustering and annotation.

Clustering and Cell-Type Identification

To identify distinct cell populations, clustering was performed as follows: (1) for Normalization and Scaling, the dataset was log-normalized and scaled using the ScaleData () function. (2) Principal component analysis (PCA) was performed using known marker genes for different small intestinal cell populations. The first 10 principal components (PCs) were selected for downstream clustering and visualization. (3) For Clustering and Dimensionality Reduction, (a) the FindNeighbors() function was applied using the top 10 PCs, (b) the FindClusters() function was used with a resolution of 0.8, and (c) Uniform Manifold Approximation and Projection (UMAP) was used for dimensionality reduction and visualization.

Visualization and Functional Analysis

Various Seurat functions were used for downstream analyses and data visualization, including (1) DoHeatMap (), for cluster-specific gene expression heatmaps; (2) FeaturePlot (), for visualizing gene expression in UMAP space; (3) VlnPlot (), for violin plots of gene expression; (4) DotPlot (), for comparative expression across clusters; (5) RidgePlot (), for density distribution of gene expression; and (6) AddModuleScore (), for calculating module scores on a single-cell basis, including proliferation and senescence scores.

This approach allowed for the robust identification and characterization of small intestinal epithelial cell (IEC) populations under various conditions, providing insight into cell-type-specific gene expression patterns.

Ethics

This study was approved by our institutional review board (IRB#: STU00209176).

Results

Single-Cell Expression Patterns of sEH (EPHX2) and CYP Epoxygenases in Human and Mouse Small-Bowel Epithelium

Mouse Small-Bowel Epithelium

A total of 13 cell clusters were identified for the mouse small-bowel epithelia, based on their biomarker gene expression. These cell clusters include two stem-cell clusters (Lgr5+ stem cells and +4 stem cells/Hopx+ cells), four TA cell clusters (TA-1, TA-2, TA-3, and Ki-67+ TA), and seven differentiated cell clusters (enterocyte villi-top-distal [FGR15-high cell population], enterocyte villi-top-proximal, enterocyte villi-bottom, Paneth cells, Goblet cells, endocrine cells, and Tuft cells), as illustrated in the UMAP plot (Fig. 2A). The detailed classification of these specific cell clusters is described in the Supplemental material.

Figure 2.

scRNAseq UMAP and expression distribution of soluble epoxide hydrolase and cytochrome P450 epoxygenases in mouse small intestinal epithelium. (A) UMAP visualization of single-cell RNA sequencing data from mouse small intestinal epithelial cells, showing 13 cell clusters. Cells are color-coded by cluster identity. (B)–(I) Violin plots showing normalized expression levels of Ephx2 (encoding soluble epoxide hydrolase) and a panel of cytochrome P450 epoxygenases across epithelial cell types. Each violin represents the distribution and density of gene expression within the indicated cluster. Abbreviation: UMAP, Uniform Manifold Approximation and Projection.

sEH is encoded by EPHX2 gene. The transcriptional single-cell expression pattern of the EPHX2 gene was identified in the small-bowel epithelium, as shown in Fig. 2B. Key observations include: (1) A high level of expression and a high percentage of positive cells were observed in Clusters 5 and 12, which correspond to villus-top differentiated enterocytes. Thus, this gene also serves as a biomarker that distinguishes this cell type in both the proximal and distal small bowel from other epithelial cell types. (2) Compared to the villi-top enterocyte clusters in both proximal and distal small bowel, a significantly lower level of expression was observed in the villi-bottom enterocytes (cell Cluster 3, p<0.0001). (3) Except TA cell Cluster 6, all other TA cell clusters (Clusters 1, 2, and 8) showed moderate level of EPHX2 expression. And (4) the remaining cell clusters including stem-cell clusters, Goblet cells, Tuft cells, and Paneth cells displayed significantly low level of expression compared to the differentiated enterocytes (p<0.0001).

Ten of the 47 mouse CYP epoxygenases (CYP1a1, 1a2, 3a4, 3a13, 3a57, 4a10, 4a11, 4a12a, 4a12b, 46a1, 2b10, 2b19, 2b23, 2c29, 2c38, 2c39, 2c40, 2c44, 2c50, 2c54, 2c55, 2c65, 2c66, 2c67, 26c1, 2d22, 2e1, 4f2, 4F3, 4F8, 4F12, 4f13, 4f14, 4f15, 4f16, 4f17, 4f18, 4f37, 2j5, 2j6, 2j8, 2j9, 2j11, 2j12, 2j13, 2s1, 2u1) were found to be expressed in the mouse small-bowel epithelium, including CYP1a1, CYP2b10, CYP2c29, CYP2c55, CYP2c44, CYP2e1, CYP4f13, CYP4f16, CYP4f18, and CYP2s1. ¹² Based on their expression patterns across epithelial cell clusters, these CYP epoxygenases were classified into five unique groups: Group 1 was a stem cell and TA cell expression pattern—CYP2e1 exhibited low-level expression in these cell clusters (Fig. 2C). As depicted in Fig. 2C, CYP2e1 was expressed in TA-1, Ki-67+ TA, Lgr5+ stem cells, +4 cells/Hopx+ stem cells, TA-3, and TA-2 clusters. To demonstrate the nature of CYP2e1 expression in IECs, we determine its co-expression pattern with stem cell and proliferative cell markers. In a total of 52 CYP2e1-positive mouse IECs, (1) for stem-cell markers, it was observed that there were 15 cells co-expressed with Lgr5 (29%), and 49 cells with Nmp1 and Mif1, indicating CYP2e1 mainly expressed in the cell cluster of +4 cells/Hopx+ stem cells. For cell proliferation marker, there were only 17 cells expressed with MKi-67 (33%, 17/52).

Group 2 was a TA cell–dominant expression pattern—CYP4f18 was identified to be expressed in TA-1, Ki-67+ TA, and enterocyte villi-bottom clusters (Fig. 2D). Group 3 was a TA cell and enterocyte expression pattern—CYP2c44 had low-level expression in TA-1, TA-3, Ki-67+ TA, and enterocyte villi-bottom/top-proximal clusters (Fig. 2E). Group 4 was an enterocyte expression pattern—CYP1a1 and CYP2c29 were primarily expressed in enterocyte villi-top-proximal (Fig. 2F and G). CYP1a1 showed a minimal expression in Goblet cells, and CYP2c29 displayed a minimal expression in Goblet cells and TA-2 cell cluster. Group 5 was a dominant enterocyte expression pattern with minimal expression in other cell clusters—CYP2b10,2c55 (Fig. 2H and I). Group 6 was a ubiquitous expression pattern across multiple clusters—4f13, 4f16, and 2s1 belonged to this category (Fig. 2J–L).

In addition, the low number of cells and low level of expression of CYP2e1 (Fig. 2C), CYP4f18 (Fig. 2D), CYP2c44 (Fig. 2E), CYP1a1 (Fig. 2F), and CYP2c29 (Fig. 2G) were identified. First, as our method, we used various Seurat functions for downstream analyses and data visualization, including violin plots of gene expression, comparative expression across cell clusters, density distribution of gene expression, and calculating module scores on a single-cell basis. Second, with Fisher’s enrichment test, statistical significance was further confirmed (p<0.0001). Last, RNAscope validation was performed for selected CYPs (e.g., CYP2E1, CYP1A1) to confirm focal expression (the result in the Figure 7).

Figure 7.

In situ expression patterns of EPHX2 and selected CYP epoxygenases in mice and human small intestinal mucosa using RNAscope mRNA in situ hybridization: (A)–(E) Small intestinal mucosa in mice including (A) EPHX2 expression was predominantly localized to villous enterocytes in the duodenum and proximal jejunum, with strong staining at the villus tips and a gradual decrease toward the crypts. (The upper portion of the dashed line indicates the villous enterocytes, whereas the lower portion corresponds to the crypt.) In the distal small bowel (ileum), expression was markedly reduced in both villous and crypt epithelium. (B) CYP2J2, a principal human CYP epoxygenase, was exclusively expressed in human small-bowel epithelium, primarily in villous enterocytes of the duodenum and ileum, with minimal crypt expression. (C) CYP2S1 exhibited a similar pattern, with strong villous enterocyte expression and significantly lower expression in crypt cells. (D) CYP1A1 expression was confined to villous enterocytes in the duodenum and proximal jejunum, with no detectable expression in the ileum. (E) CYP2E1 displayed a scattered expression pattern in both villous and crypt enterocytes without clear regional restriction, differing from the stem-cell-specific crypt pattern observed in mice. Bottom panel of each figure: quantitative analysis confirmed a statistically significant reduction in expression in the ileum compared to proximal segments (* or **: p value and statistical significance). (F–J) Small intestinal mucosa in human including: (F) Ephx2 expression was highest in the villous enterocytes of the duodenum and proximal jejunum, with a gradual decrease toward the crypts. Both villous and crypt expression were significantly reduced in the distal ileum. (G) Cyp2s1 exhibited strong and diffuse expression across both villous and crypt epithelium in the duodenum and ileum, showing less regional variation than humans. (H) Cyp1a1 was expressed in villous enterocytes of the duodenum and proximal jejunum but was absent in crypt epithelium and distal ileum. (I) Cyp2b10, detected only in mice, showed a pattern similar to Ephx2: strong expression in villous enterocytes throughout the small bowel, minimal crypt expression, and higher intensity in proximal compared to distal segments. (J) Cyp2e1 exhibited a unique mouse-specific pattern, with intense staining confined to epithelial cells at the crypt base in the duodenum and jejunum, and no expression in the ileum. This crypt-localized expression aligns with scRNAseq findings identifying Cyp2e1 in stem-cell and transit-amplifying clusters. Bottom panel of each photo: quantitative analysis confirmed a statistically significant difference between proximal and distal segments (* or **: p value and statistical significance). Scale bar: 100 μm.

Human Small-Bowel Epithelium

For the human small-bowel epithelia, based on biomarker gene expression, a total of 15 cell clusters were identified, as seen in UMAP in Fig. 3A, including (1) three intestinal stem-cell clusters (Cluster 8 [villi-bottom stem cells], Cluster 10 [crypt base stem cells], and Cluster 11 [Best4+]); (2) one TA cell cluster (Cluster 13); (3) seven differentiated cell clusters of enterocytes/intestinal absorptive cells (Clusters 1, 2, 4, 5, 6, 7, and 12); (4) Paneth cells (Cluster 9); (5) Goblet cells (Cluster 15), (6) EECs (Cluster 14); and (7) tuft cells (Cluster 3). Based on representative gene expression patterns (Supplemental material), the seven enterocyte clusters were categorized as (1) proximal small-bowel enterocytes (Clusters 4, 5, 6, 7, and 12); (2) distal small-bowel enterocytes (Clusters 1 and 2); (c) villous top enterocytes (Clusters 5, 7, and 12); and (4) villous bottom enterocytes (Clusters 4 and 6).

Figure 3.

scRNAseq UMAP and expression distribution of soluble epoxide hydrolase and cytochrome P450 epoxygenases in human small intestinal epithelium. (A) UMAP visualization of single-cell RNA sequencing data from human small intestinal epithelial cells, showing 15 cell clusters. Cells are color-coded by cluster identity. (B)–(I) Violin plots showing normalized expression levels of Ephx2 (encoding soluble epoxide hydrolase) and a panel of cytochrome P450 epoxygenases across epithelial cell types. Each violin represents the distribution and density of gene expression within the indicated cluster. Abbreviation: UMAP, Uniform Manifold Approximation and Projection.

A ubiquitous single-cell expression pattern was identified for the human EPHX2 gene/sEH in small-bowel epithelia, displaying more extensive and high expression across all epithelial cell clusters, from villous enterocyte to TA and stem-cell clusters and from proximal to distal small bowel (Fig. 3B), which are markedly different from mouse small-bowel epithelia (Fig. 2B).

Fifteen of the 47 CYP epoxygenases were found to be expressed in human small-bowel epithelia, including CYP1A1, 3A4, 4A11, 46A1, 2C8, 2C9, 2C18, 2C19, 2E1, 4F2, 4F3, 4F8, 4F12, 2J2, and 2S1. The expression pattern of these CYPs was further grouped into four groups. Group 1 was that (1) all six “principal CYP epoxygenases” (CYP2C8, 2C9, 2C18, 2C19, 2J2, and 2S1) showed a ubiquitous single-cell expression pattern in small-bowel epithelia (Fig. 3C–I), and (2) their single-cell expression patterns were nearly identical, showing significantly moderate to high level of expression in enterocyte clusters, while their expression was relatively low in other cell clusters, including ISCs and TA cells (p<0.0001).

Among the “non-principal CYP epoxygenases,” group 2 showed an identical single-cell expression pattern as seen in “principal CYP epoxygenases,” including CYP3A4, 4A11, 46A1, 4F2, 4F3, 4F8, and 4F12, exhibiting significantly high expression in enterocyte clusters and relatively low expression in intestinal stem and TA cells (Fig. 3J–P). Group 3 was CYP1A1 that showed a dominant pattern in the proximal enterocyte clusters (Clusters 4, 7, and 12) with low numbers of positively expressed cells (Fig. 3Q). Group 4 was CYP2E1 that showed low level of expression in all four clusters of stems cells, as well as in enterocytes, tuft cells, and EECs (Fig. 3R).

Co-Expression Pattern of EPHX2 and CYP Epoxygenases in Mouse and Human Small-Bowel Epithelium

Co-Expression in Mice

In our single-cell RNA-seq analysis, co-expression of EPHX2 and CYP epoxygenases was defined based on normalized expression values (log-normalized counts) greater than 0.5 in the same cell. The threshold of 0.5 was chosen empirically to minimize background noise while capturing biologically meaningful expression, consistent with previously published single-cell RNA-seq studies using similar normalization methods. This cutoff effectively distinguishes low-level background detection from genuine transcript expression in sparse epithelial populations. As seen in Figs. 4 and 5, the color thresholding used in the visualization of co-expression patterns was applied to enhance interpretability of expression intensity across individual cells. Specifically, co-expression values were mapped to a color scale ranging from green (EPHX2) and red (CYP) expression to yellow (merged co-expression), with the threshold set to exclude background or near-zero expression values. The color cutoff was empirically determined based on the distribution of normalized expression values for each gene to ensure that only biologically meaningful signals were visualized. This approach prevents misleading interpretation from low-intensity background noise while maintaining consistency across datasets. Cells meeting this criterion for both EPHX2 and a given CYP gene were considered co-expressing.

Figure 4.

Co-expression pattern of EPHX2 and each cytochrome P450 epoxygenase (including CYP in mouse small intestinal epithelium revealed by scRNAseq. Dot plot showing the co-expression of Ephx2/sEH and individual cytochrome P450 epoxygenase across epithelial cell clusters in the mouse small bowel. Each dot represents the proportion of cells expressing the gene, and color intensity indicates the average normalized expression level and overlapping within each cluster.

Figure 5.

Co-expression pattern of EPHX2 and each cytochrome P450 epoxygenase (including CYP in human small intestinal epithelium revealed by scRNAseq. Dot plot showing the co-expression of Ephx2/sEH and individual cytochrome P450 epoxygenase across epithelial cell clusters in the human small bowel. Each dot represents the proportion of cells expressing the gene, and color intensity indicates the average normalized expression level and overlapping within each cluster.

In a total of 3048 cells analyzed by scRNAseq, 19.7% of cells expressed the EPHX2 gene (601/3048). In Fig. 4, the percentage of each CYP co-expression with EPHX2 in the total EPHX2-positive cells (601 cells) was identified, in which, there were 3.2% (19/601) CYP1A1, 24.6% (148/601) CYP2b10, 3.8% (23/601) CYP2c19, 1.7% (10/601) CYP2c44, 14.3% (86/601) CYP2c55, 1.3% (8/601) CYP2e1, 16.8% (101/601) CYP4f13, 20.8% (125/601) CYP4f16, and 13.3% (80/601) CYP2s1. Among a total of 3048 cells analyzed, there were four CYP4f18-expressed cells, but none of these four cells was positive for EPHX2. We further analyzed the overall co-expression of CYP epoxygenases with EPHX2, defining co-expression as the presence of at least one CYP epoxygenase in EPHX2-positive cells. Among the total 601 EPHX2-positive cells, 57% (343/601) exhibited co-expression.

On the other hand, we also analyzed the EPHX2 co-expression with each epoxygenase. Within a total number of positive cells for each CYP, our results showed that the percentage of EPHX2-positive cells was 90.5% (19/21) for CYP1a1; 60.2% (148/246) for CYP1b10, 74.2% (23/31) for CYP2c29, 76.9% (10/13) for CYP2c44, 54.1% (86/159) for CYP2c55, 15.7% (8/51) for CYP2e1, 26.5% (101/381) for CYP4f13, 27.2% (125/460) for CYP4f16, 0% (0/4) for CYP4f18, and 24.1% (80/332) for CYP2s1. Among these CYPs, CYP1a1, 2b10, 2c29, 2c44, and 2c55 showed high co-expression (≥50% co-expressed cells). In contrast, CYP2e1, 4f13, 4f16, and 2s1 exhibited low co-expression (<30% co-expressed cells). CYP4f18 had no co-expression with EPHX2.

Co-Expression in Human

In a total of 6507 cells analyzed for scRNAseq expression, 73.8% (4799/6507) exhibited EPHX2 expression. As seen in Fig. 5, the percentage of each CYP co-expression with EPHX2 in the total EPHX2-positive cells was 0.3% (12/4799) for CYP1a1, 58.7% (2816/4799) for CYP 3a4, 1.8% (84/4799) for CYP4a11, 0.4% (19/4799) for CYP46a1, 2.1% (101/4799) for CYP2c8, 38.8% (1861/4799) for CYP2c9, 46.9% (2250/4799) for CYP2c18, 33.3% (1597/4799) for CYP1c19, 0.1% (6/4799) for CYP2e1, 35.5% (1706/4799) for CYP4f2, 40.9% (1965/4799) for CYP4f3, 0.2% (9/4799) for CYP4f8, 55.0% (2640/4799) for CYP4f12, 45.3% (2174/4799) for CYP2j2, and 27.9% (1337/4799) for CYP2s1. We further analyzed the overall co-expression of CYP epoxygenases with EPHX2. Among the 4799 EPHX2-positive cells, 4339 (90.4%) showed co-expression with at least one CYP epoxygenase.

On the other hand, in total CYP-positive cells for each epoxygenase, the co-expression percentages with EPHX2 were as follows: 92.3% (12/13) for CYP1A1, 84.0% (2816/3352) for CYP3A4, 88.4% (84/95) for CYP4A11, 86.4% (19/22) for CYP46A1, 90.1% (101/112) for CYP2C8, 82.6% (1861/2254) for CYP2C9, 88.7% (2250/2538) for CYP2C18, 90.4% (1597/1767) for CYP2C19, 50.0% (6/12) for CYP2E1, 89.8% (1706/1900) for CYP4F2, 84.4% (1965/2329) for CYP4F3, 81.8% (9/11) for CYP4F8, 84.0% (2640/3143) for CYP4F12, 85.0% (2174/2558) for CYP2J2, and 84.6% (1337/1580) for CYP2S1. All “principal CYP epoxygenases” (CYP2C8, 2C9, 2C18, 2C19, 2J2, and 2S1) exhibited extremely high co-expression (≥80% co-expressed cells). In addition, other CYP epoxygenases, including CYP1A1, 3A4, 4A11, 46A1, 2E1, 4F2, 4F3, 4F8, and 4F12, followed a similar pattern, with more than 80% of cells co-expressing EPHX2. The only exception was CYP2E1, which had a low number of expressing cells and lower co-expression levels (50%).

Single-Cell Expression Pattern of CYP–sEH Axis Metabolic Enzymes in Radiation-Induced Mouse Small-Bowel Epithelium Damage-Regeneration Process

To determine whether there are any changes in single-cell expression patterns in mice following radiation-induced small-bowel epithelial damage-regeneration, we analyzed publicly available scRNAseq data. ¹⁶ Based on biomarker gene expression, a total of 12 cell clusters were identified, including one stem-cell cluster, two TA cell clusters (TA, and Mki-67+ TA cells), eight differentiated cell clusters (including: enterocytes [villi-top-distal or FGR15-high cell population, villi-top-proximal, villi-bottom], three enterocyte clusters [villi-top proximal 1, 2, and 3, villi-top distal], Paneth cells, goblet cells, endocrine cells, and tuft cells), as well as one intraepithelial lymphocyte cluster, as visualized in UMAP (Fig. 6). The detailed classification of these specific cell clusters is described in the Supplemental material.

Figure 6.

scRNAseq UMAP and expression distribution of soluble epoxide hydrolase and cytochrome P450 epoxygenases in mouse small intestinal epithelium treated with radiation. (A) UMAP visualization of single-cell RNA sequencing data from radiated mouse small intestinal epithelial cells, showing 12 cell clusters. Cells are color-coded by cluster identity. (B)–(I) Violin plots showing normalized expression levels of Ephx2 (encoding soluble epoxide hydrolase) and a panel of cytochrome P450 epoxygenases across epithelial cell types. Each violin represents the distribution and density of gene expression within the indicated cluster either in radiated mice or non-radiation control mice.

The scRNAseq expression patterns of EPHX2 in these small-bowel epithelial cell clusters after radiation showed (1) loss of EPHX2 expression in the Paneth cell cluster and (2) a significant decrease (overall p<0.001) in EPHX2 expression in cell clusters of enterocytes (villi-top proximal 3, Cluster 1), goblet cells (Cluster 5), enterocytes (villi-tope proximal 3, Cluster 6), Lgr5+ stem cells (Cluster 7), and Mki-67+ TA cells (Cluster 8) (Fig. 6).

For CYP epoxygenases, only CYP4F18 showed a significant increase in expression in proximal and distal villous top enterocytes (Clusters 1 and 3, p<0.001), Goblet cells (Cluster 5, p<0.0001), and TA cells (Cluster 4, p<0.0001) after radiation (Fig. 6), although overall expression level of this enzyme is low in non-treated mice (Fig. 6). Loss of CYP1A1 and CYP2C38 was observed in small-bowel epithelial cells following radiation (Fig. 6). In addition, a significant decrease in CYP2B10, CYP2C55, and CYP2S1 expression was identified in almost all small-bowel epithelial cell clusters (Fig. 6, p<0.0001).

In Situ Villi–Crypt Axis Expression Pattern and Distribution of EPHX2 and CYP Epoxygenases in Human and Mouse Small-Bowel Epithelium

Using RNAscope mRNA in situ hybridization assay, the in situ expression pattern of EPHX2-coded sEH was analyzed and exhibited an identical expression pattern in the small-bowel epithelium of both mouse and human small bowel. Top villous enterocytes in the duodenum and jejunum exhibited more extensive intense positive staining and further displayed a gradual reduction in expression levels from the top of the villi down to the crypts (Fig. 7A and F). Compared to the proximal small bowel including duodenum and proximal jejunum, staining intensity was significantly lower in both villous and crypt epithelium in the distal small bowel (ileum) (Fig. 7A and F). Quantitative analysis confirmed the statistically significant difference (p<0.001, Fig. 7A and F, bottom histogram).

For the representative CYPs with both hydroxylase and epoxygenase activity, CYP1a1 exhibited an identical expression pattern in both mice and human, showing strong positive staining in the villous enterocytes, but not in the crypt cells of the duodenum and proximal jejunum (Fig. 7D and G). No expression of CYP1a1 was identified in the distal ileum (Fig. 7D and G). Since CYP2B10 only expressed in mouse small epithelium per our scRNAseq results, we analyzed its in situ expression pattern. CYP2B10 in mice displayed an expression pattern similar to sEH/EPHX2 and showed (1) extensive expression in the villous enterocytes in both proximal and distal small bowel, (2) much less expressed in the crypts, and (3) more intense expression in the enterocytes in proximal small bowel than in distal small bowel (Fig. 7C).

CYP2E1 displayed species-specific differences. Interestingly, in mice, CYP2E1 showed a unique expression pattern, with intense positive staining in the epithelial cells at the bottom of the crypts in the duodenum and jejunum, but not in the ileum (Fig. 7E). As aforementioned, the scRNAseq expression pattern of CYP2e1 demonstrated its presence in stem-cell and TA cell clusters, corroborating findings from RNAscope mRNA in situ hybridization, which identified CYP2e1+ epithelial cells in crypts (Fig. 7E). RNAscope mRNA in situ hybridization further demonstrated that both CYP2E1 and Lgr5 were expressed in the epithelial cells at the crypt base, whereas MKi-67 labeled proliferating epithelial cells located higher up in the crypts compared to CYP2E1- and Lgr5-positive cells (Fig. 8).

Figure 8.

In situ expression patterns of CYP2e1, Lgr5, and Mki-67 in small-bowel mucosa in mice: (A) CYP2e1: positively stained cells located at the base of the crypts (arrow). (B) Lgr5: positively stained cells at the base of the crypts (arrow). (C) Mki-67: positive cells diffusely distributed throughout the crypts. Scale bar: 100 μm.

For the selected human CYP epoxygenases, the well-recognized and typical “principal CYP epoxygenases” CYP2J2 and CYP2S1 were analyzed for their in situ expression pattern. CYP2J2 was exclusively expressed in human small-bowel epithelium (Fig. 7H) and predominantly expressed in the villous enterocytes of both the duodenum and ileum, with minimal expression in the crypts (Fig. 7H). Since CYP2J2 was absent in mouse small-bowel epithelium (per our scRNAseq analysis), we did not perform RNAscope mRNA in situ hybridization assay for mouse. CYP2S1, on the other hand, was expressed in both mouse and human small-bowel epithelium with a similar pattern (Fig. 7D and I). In mice, CYP2S1 was diffusely and highly expressed in both villous and crypt epithelium in both the duodenum and ileum (Fig. 7D). In humans, CYP2S1 was highly expressed in the villous enterocytes of both the duodenum and ileum, with a significantly lower expression in the crypt epithelium than in the villi (Fig. 7I). In humans, CYP2E1 did not show a stem-cell-specific expression pattern as seen in mice and displayed a scattered positive staining throughout both villous and crypt enterocytes (Fig. 7J).

Discussion

A single layer of polarized IECs is a key component of the intestinal barrier. ¹⁷ Stem cells in the intestinal crypt form a rapidly proliferating cell population that completely renews the epithelium with newly generated cells every 4–5 days. Maintaining structural integrity, stem-cell–enterocyte axis maturation, and barrier function in the gut requires tight regulation of epithelial cell proliferation and cell death rates. It is well known that eicosanoid metabolites, such as prostaglandin E2 (PGE2), play a critical role in sustaining intestinal mucosal barrier homeostasis. ¹⁸ However, there has been limited research on the importance of the CYP–sEH axis as a key PUFA/eicosanoid metabolic pathway in maintaining intestinal mucosal barrier homeostasis. In the present study, we are the first to comprehensively analyze and identify the unique single-cell expression patterns and distribution of CYP–sEH axis enzymes using the RNAscope in situ hybridization approach, combined with publicly available scRNAseq datasets. Our findings provide a foundation for exploring the biological function of this metabolic pathway in the small-bowel mucosal barrier.

Overall information and findings using these three publicly available scRNAseq datasets have been summarized in Table 1. First, to minimize bias and ensure comparability across datasets, all scRNAseq data were reprocessed using a uniform computational workflow. For quality control and consistency across the three datasets, raw or count matrix files were processed using Seurat (v5.0) in R. Gene expression values were log-normalized and filtered to remove low-quality cells (<250 detected genes or >20% mitochondrial gene expression). The top 2000 highly variable genes were selected from each dataset, followed by data scaling and centering. To minimize batch effects and technical variability, datasets were integrated using Seurat’s CCA-based integration pipeline. Dimensionality reduction and clustering were performed with uniform parameters, and cell types were annotated based on canonical intestinal epithelial markers (e.g., Lgr5, Alpi, Muc2, Chga, and Lyz1). Annotation accuracy was further confirmed by comparison, with cluster identities reported in the original studies. This standardized workflow ensured consistent data processing, reliable integration, and robust cross-dataset comparisons.

Table 1.

Summary of scRNAseq Datasets and Analysis of Single-Cell Expression Pattern of CYP and sEH in Small-Bowel Epithelial Cells.

Species	Source/Dataset	Platform	Tissue/Region	Cell Types Analyzed	Analysis and Findings in Our Study	Reference/Accession
Human	Burclaff et al., Cell Mol Gastroenterol Hepatol 2022	10x Genomics	Small intestine	All epithelial subsets	Single-cell expression pattern of CYP and sEH in intestinal epithelial cell-type specificity	GSE185224
Mouse	Haber et al., Nature 2017	10x Genomics	Small intestine	All epithelial subsets	Single-cell expression pattern of CYP and sEH in intestinal epithelial cell-type specificity	GSE92332
Mouse (radiation model)	Ayyaz et al., Nature 2019	10x Genomics	Small intestine (irradiated vs. control)	All epithelial subsets	Assess dynamic CYP/sEH changes during radiation-injury and regeneration	GSE123516

sEH is recognized as the sole enzyme responsible for metabolizing endogenous epoxyeicosanoids (EpFAs) in vivo, as summarized in Fig. 1. ⁶ Thus, sEH functions as a critical inactivating pathway that limits the biological activity of EpFAs. ⁶ First, we identified a ubiquitous expression pattern of the mouse and human EPHX2 gene, which encodes sEH, with high-level expression in enterocytes and low-level expression in all other small-bowel epithelial cell clusters in both mice and humans. Second, we comprehensively analyzed a large panel of CYP epoxygenases. In mice, CYP2b10, 2s1, 2c55, 4f13, and 4f16 exhibited an expression pattern similar to EPHX2. In humans, both “principal” CYP epoxygenases (CYP2c9, 2c18, 2c19, 2j2, and 2s1) and “non-principal” CYP epoxygenases (CYP3a4, 4a11, 4f2, 4f3, 4f8, and 4f12) displayed a similar expression pattern to EPHX2.

As our results, among the EPHX2-positive cells, 90.4% (4339/4799) in human and 57% (343/601) in mice showed co-expression with at least one CYP epoxygenase. The high co-expression patterns of CYP epoxygenases and EPHX2 enzymes in the small bowel suggest a potentially important biological role of the CYP–sEH metabolic axis in sustaining small-bowel epithelial mucosal barrier homeostasis.

Obviously, 9.6% EPHX2-positive cells in humans and 43% in mice exhibit no co-expression with CYP epoxygenase. There are several possibilities for this difference in co-expression: First, species-specific differences in transcriptional regulation or metabolic demand may contribute to the higher proportion of co-expressing cells in humans compared with mice. Second, cellular heterogeneity within epithelial populations could explain the variability, as EPHX2 may be expressed in broader epithelial subsets, including those with minimal CYP epoxygenase activity. Third, temporal and functional regulation may play a role, since CYP epoxygenases are often inducible by physiological or environmental stimuli, whereas EPHX2 expression tends to be more constitutive. Fourth, differences in tissue microenvironment or metabolic state, such as exposure to dietary lipids, oxidative stress, or inflammatory cues, may influence the co-expression pattern. Finally, technical factors, including variation in mRNA stability, probe sensitivity, or detection thresholds, could partially contribute to the apparent discrepancy in co-expression frequencies between species.

Overall, the co-expression of EPHX2 with CYP epoxygenases in IECs suggests a tightly coordinated metabolic system that regulates the balance between the production and inactivation of bioactive lipid mediators derived from PUFAs. Functionally, this CYP–sEH co-expression may serve to fine-tune the local levels of EETs and related epoxy metabolites, ensuring appropriate responses to physiological stimuli such as oxidative stress, inflammation, or injury. High co-expression levels in humans, compared to mice, may indicate a greater reliance on this pathway for maintaining epithelial barrier integrity and homeostatic renewal in the human intestine. Conversely, EPHX2-positive cells without CYP co-expression could reflect cell states (such as cell cycle–related expression) or lineages that are primarily involved in metabolite clearance rather than synthesis, or that depend on alternate lipid metabolic routes.

The crypt–villus axis forms a histologic architecture in the intestinal epithelial barrier, where ISCs in the crypts give rise to mature enterocytes. ¹⁷ The unique expression patterns of CYP–sEH axis enzymes in the crypt–villus axis have been identified through the scRNAseq analysis and the RNAscope approach. We found that the key or “principal” CYP epoxygenase CYP2s1 in mice and CYP2j2 and CYP2s1 in humans are extensively or ubiquitously expressed in IECs from crypts to mature enterocytes, particularly in absorptive enterocytes, throughout both the proximal and distal parts of the small bowel. These expression patterns are similar to those observed for EPHX2 gene expression, indicating the potential importance of the CYP–sEH axis in maintaining intestinal epithelial barrier homeostasis.

In addition, CYP2b10, 2c55, 4f13, and 4f16 in mice and CYP3a4, 4a11, 46a1, 4f2, 4f3, 4f8, and 4f12 in humans exhibit a similar ubiquitous expression pattern as that seen in the principal CYP epoxygenases CYP2j2 and CYP2s1 in IECs. Notably, these CYPs possess both hydroxylase and epoxygenase activities, enabling them to metabolize PUFAs into biologically active epoxy- and hydroxyl-metabolites. This underscores the need for further biological studies to explore their potential roles in intestinal mucosal barrier homeostasis and defense.

The limited number of cells and low level of expression of CYP2e1 (Fig. 2C), CYP4f18 (Fig. 2D), CYP2c44 (Fig. 2E), CYP1a1 (Fig. 2F), and CYP2c29 (Fig. 2G) in mice were noted. Similarly, in human small-bowel epithelia, CYP2c8, CYP4a11, CYP46a1, CYP4f8, CYP1a1, and CYP2e1 were also identified with a low expression level. Obviously, the low expression and limited positive cell numbers for CYP epoxygenases in the small intestine raise questions regarding their biological relevance and detection sensitivity. Our RNAscope analysis provides strong evidence to support their unique single-cell expression patterns of CYPs in IECs with low-level expression and potential biological significance. For example, Our RNAscope analysis demonstrates that (1) CYP1a1 is predominantly expressed in absorptive enterocytes in both mice and humans, and (2) CYP2e1 is primarily expressed in ISCs in mice, and co-expression with biomarkers of +4 cells/Hopx+ stem cells and Lgr5. As seen in an scRNAseq assay, CYP4f18 exhibits a TA cell-dominant expression pattern in mice, and CYP2c29 is dominantly expressed in absorptive enterocytes in mice but not in humans. These unique expression patterns in specific cell lineages with low-level expression suggest their potential biological functions in cell lineage development.

The differential expression of CYP epoxygenases among intestinal epithelial lineages likely reflects lineage-specific metabolic requirements and developmental programs along the crypt–villus axis. For instance, CYPs enriched in stem or TA cells (e.g., CYP2e1, CYP4f18) may participate in regulating proliferation, differentiation, or response to stress signals. In contrast, the prominent expression of CYP1a1, CYP2s1, and CYP2j2 in mature absorptive enterocytes may support their role in mucosal defense, detoxification, and maintenance of epithelial integrity. These lineage-specific patterns imply that the CYP–sEH axis is integrated into epithelial cell differentiation programs, enabling context-dependent lipid signaling that supports both barrier function and epithelial renewal.

Radiation-induced changes in CYP and EPHX2 expression further highlight the dynamic regulation of this metabolic network under conditions of epithelial stress and injury. Radiation disrupts intestinal epithelial integrity by inducing oxidative and inflammatory stress, and altered CYP–sEH expression may represent a compensatory or protective response aimed at restoring redox balance, modulating inflammation, or promoting epithelial regeneration. Therefore, the radiation model provides a valuable platform to study how environmental or therapeutic insults reshape PUFA metabolism and to elucidate the protective roles of the CYP–sEH pathway in intestinal injury and repair.

Beyond PUFA metabolism, our findings of the high-level expression of CYPs in IECs suggest additional roles in (1) drug and xenobiotic detoxification, (2) cellular metabolism and homeostasis, and (3) intestinal epithelial barrier defense—all of which warrant further detailed investigation.

Overall, our integrated scRNAseq and RNAscope analyses reveal distinct, conserved, and inducible expression patterns of CYP epoxygenases and sEH in intestinal epithelial lineages. These findings underscore the functional versatility and physiological relevance of the CYP–sEH axis in epithelial homeostasis, lineage differentiation, and stress adaptation, setting the stage for future mechanistic studies of this pathway in intestinal health and disease. Further research is needed to explore the functional impact of CYP-mediated PUFA metabolism in intestinal epithelial physiology.