Abstract

   The precise mechanism underlying the therapeutic effects of
   dihydroartemisinin (DHA) in alleviating colitis remains incompletely
   understood. A strong correlation existed between the elevation of glial
   fibrillary acidic protein (GFAP)^+/S100 calcium binding protein B
   (S100β)^+ enteric glial cells (EGCs) in inflamed colonic tissues and
   the disruption of the intestinal epithelial barrier (IEB) and gut
   vascular barrier (GVB) observed in chronic colitis. DHA demonstrated
   efficacy in restoring the functionality of the dual gut barrier while
   concurrently attenuating intestinal inflammation. Mechanistically, DHA
   inhibited the transformation of GFAP^+ EGCs into GFAP^+/S100β^+ EGCs
   while promoting the differentiation of GFAP^+/S100β^+ EGCs back into
   GFAP^+ EGCs. Furthermore, DHA induced apoptosis in GFAP^+/S100β^+ EGCs
   by inducing cell cycle arrest at the G0/G1 phase. The initial
   mechanism is further validated that DHA regulates EGC heterogeneity by
   improving dysbiosis in colitis. These findings underscore the
   multifaceted therapeutic potential of DHA in ameliorating colitis by
   improving dysbiosis, modulating EGC heterogeneity, and preserving gut
   barrier integrity, thus offering promising avenues for novel
   therapeutic strategies for inflammatory bowel diseases.

   Keywords: colitis, dihydroartemisinin, dysbiosis, enteric glial cells,
   gut barrier, phenotypic transformation
     __________________________________________________________________

   In chronic colitis, microbial products or metabolites, along with
   inflammatory stimuli, can lead to concurrent damage to both the
   intestinal epithelial barrier (IEB) and the gut vascular barrier (GVB),
   accompanied by an increase in GFAP^+/S100β^+ enteric glial cells (EGCs)
   within the inflamed gut microenvironment. Dihydroartemisinin (DHA)
   regulates the heterogeneity of EGCs by improving dysbiosis, thereby
   effectively alleviating colitis and promoting the restoration of dual
   gut barrier function.

   graphic file with name ADVS-11-2403461-g001.jpg

1. Introduction

   Ulcerative colitis (UC) has traditionally been defined as diffuse,
   superficial inflammation that is restricted to the mucosa.^[ [46]^1 ,
   [47]^2 ^] UC and Crohn's disease (CD) constitute the two primary types
   of inflammatory bowel disease (IBD), characterized as a group of
   chronic, relapsing, and inflammatory disorders.^[ [48]^3 ^] The
   impairment of gut barrier function and the presence of chronic
   inflammation serve as pivotal drivers of clinical complications in IBD.
   Consequently, the restoration of gut barrier integrity and mitigation
   of intestinal inflammation represent critical therapeutic objectives in
   the management of IBD.^[ [49]^4 ^] In 2015, Spadoni et al. put forward
   the concept of the gut vascular barrier (GVB), suggesting that the GVB
   and intestinal epithelial barrier (IEB) are distinct entities.^[ [50]^5
   ^] Their proposition suggests that the GVB dismantling plays a crucial
   role in facilitating the entry of harmful substances, including toxins,
   into the systemic circulation, thereby leading to liver damage. The GVB
   controls the passage of antigens and commensal gut microbiota from the
   intestine into the bloodstream, preventing bacteria that have
   successfully crossed the IEB from entering the bloodstream.^[ [51]^6 ^]
   This insight underscores the potential benefits of enhancing
   dual‐barrier function as an underutilized strategy for treating IBD.
   While current research primarily focuses on the damage to the IEB
   during IBD, attention to the GVB offers a promising avenue for
   therapeutic intervention.

   The enteric nervous system (ENS) is a “brain‐like” neural network that
   integrates several neurons and glia to give local control over vital
   gastrointestinal activities.^[ [52]^7 ^] Enteric glial cells (EGCs) are
   abundant peripheral glial cells that perform homeostatic functions in
   the gastrointestinal tract. Loss or enhancement of glial function can
   lead to alterations in the homeostasis of the ENS and gastrointestinal
   pathophysiology.^[ [53]^7 ^] In response to intestinal inflammation,
   enteric glia undergo dynamic changes both in vivo and in vitro,
   involving modifications in glial fibrillary acidic protein (GFAP)
   expression, morphological alterations, cytokine release, and shifts in
   gene expression profiles, as suggested by several studies.^[ [54]^8 ,
   [55]^9 , [56]^10 ^] Despite these observations, the precise mechanisms
   underlying these responses remain elusive. Functioning as a central
   hub, EGCs regulate various processes such as homeostasis, pathogen
   defense, and tissue regeneration through direct and indirect
   interactions with different cell types within the gut.^[ [57]^7 ^] As
   proposed by Kabouridis et al. EGCs serve as primary targets for the gut
   microbiota, highlighting the regulatory role of microbes and their
   metabolites in the development and maturation of the EGC network.^[
   [58]^11 ^] EGCs exhibit prominent heterogeneity and phenotypic
   plasticity, enabling them to adjust their functions based on triggers
   in the surrounding environment to preserve local homeostasis and
   facilitate effective resolution of intestinal inflammation. However,
   our understanding of the regional or local heterogeneity of EGCs and
   how their phenotypic changes affect gut barrier function and
   inflammation progression during enteritis remains limited. The
   potential relationship between microbiota diversity and EGC
   heterogeneity in intestinal inflammation remains poorly understood. In
   this study, we aimed to elucidate the relationship between alterations
   in EGC heterogeneity induced by chronic intestinal inflammation and
   changes in microbiota diversity. Furthermore, we sought to uncover the
   specific mechanisms by which these factors collectively regulate dual
   gut barrier function.

   Artemisinin, derived from the herb Artemisia annua, is a sesquiterpene
   lactone peroxide. Its active metabolite, dihydroartemisinin (DHA), is
   extensively utilized for malaria treatment.^[ [59]^12 ^] Beyond its
   well‐known anti‐parasitic effects, recent studies have unveiled its
   diverse properties, including antiviral, antifungal, anticancer, and
   anti‐inflammatory activities.^[ [60]^13 ^] Notably, a study reported
   that DHA prominently inhibits microglial activation and
   neuroinflammation by modulating the miR‐16‐mediated Toll‐like receptor
   4 (TLR4)/nuclear factor kappa B (NF‐κB) signaling pathway.^[ [61]^7 ^]
   Moreover, Gao et al. discovered that DHA ameliorates lipopolysaccharide
   (LPS)‐induced neuroinflammation by inhibiting the phosphatidylinositol
   3‐kinase (PI3K)/protein kinase B (AKT) pathway, suggesting the
   potential of DHA in mitigating neuroinflammatory diseases.^[ [62]^7 ^]
   While the effects of DHA on the central nervous system (CNS) have been
   extensively investigated, its impact on the ENS remains less
   understood.^[ [63]^14 ^]

   We aimed to explore the pathogenesis and therapeutic strategies of
   colitis. Drawing insights from the GVB and the ENS, we delve into the
   essential drivers of gut barrier function and the role of EGCs in this
   process. Moreover, we focus on DHA as a potential treatment for UC,
   aiming to clarify its capacity to restore the dual gut barrier function
   and alleviate chronic colitis and its relationship with EGC
   heterogeneity and gut microbiota diversity.

2. Results

2.1. Dual Gut Barrier Impairment in Colitis

   To investigate the extent of gut barrier damage in inflamed colon
   tissues of colitis, considering that the damage primarily occurs in the
   mucosal and submucosal layers,^[ [64]^15 , [65]^16 ^] we obtained colon
   tissues from mice with dextran sulfate sodium (DSS)‐induced chronic
   colitis using a stripping method, which removed the muscular and
   serosal layers. Subsequently, we analyzed the expression of barrier
   proteins in the tissues (Figure [66] 1A,B). Our findings revealed that
   mice with chronic colitis exhibited impairment not only in the IEB
   (with downregulation of key tight junction (TJ) proteins zonula
   occludens‐1 (ZO‐1) and Occludin) but also in the GVB (downregulation of
   adherens junction (AJ) protein β‐catenin,^[ [67]^5 ^] upregulation of
   the plasmalemma vesicle‐associated protein 1 (PV1),^[ [68]^17 ^] a
   marker of vascular endothelial cell permeability^[ [69]^18 , [70]^19
   ^]). Furthermore, the damage to the GVB was evidenced by significantly
   elevated levels of LPS in the peripheral blood of colitic mice compared
   to the control group (Figure [71]S1A, Supporting Information). The
   release of 70 kDa dextran from the bloodstream was observed
   (Figure [72]1C), as the shifting of molecules ≈70 kDa is known to be
   resisted by intact GVB.^[ [73]^5 ^] Simultaneously, the levels of
   inflammatory cytokines interleukin 1 beta (IL‐1β) and tumor necrosis
   factor (TNF) were significantly elevated in inflamed colon tissues
   compared to the control group (Figure [74]S1B,C, Supporting
   Information), while the anti‐inflammatory cytokine interleukin 10
   (IL‐10) showed a noticeable decrease in concentration (Figure [75]S1D,
   Supporting Information). The electron microscopy results revealed a
   significant depletion of goblet cells in DSS‐induced colitis tissues
   (Figure [76]1D), along with an increase in fenestration of the
   endothelial cells at the tips of the villus in the colitis group
   (Figure [77]1E). Consistently, increased angiogenesis was observed in
   the mucosa, submucosa, and lamina propria of the colonic tissues of
   mice with colitis, accompanied by a significant upregulation of PV1
   expression (Figure [78]1F–I). The histological examination of inflamed
   mucosal biopsies obtained from active UC patients and non‐IBD controls
   revealed extensive architectural tissue damage, characterized by
   disrupted epithelial barrier, significant loss of mucosal villi,
   disorganized glandular cell arrangement, prominent glandular loss,
   extensive infiltration of inflammatory cells, and reduced goblet cells
   (Figure [79]S1E,F, Supporting Information). The histological scores for
   active UC patients were markedly elevated compared to the non‐IBD
   controls (Figure [80]S1F, Supporting Information). Similar to findings
   in mice with colitis, inflamed mucosa from patients with UC
   demonstrated increased vascularization marked by platelet and
   endothelial cell adhesion molecule 1 (CD31)(Figure [81]1J,K), as well
   as a notable upregulation of PV1 expression (Figure [82]1J,L). It is
   noteworthy that the majority of newly formed blood vessels expressed
   high levels of PV1 (Figure [83]1M), indicating severe GVB damage in the
   inflamed biopsy specimens of patients with active UC. The above results
   indicate the presence of dual gut barrier damage, including the IEB and
   GVB, in the inflamed colon tissues of chronic colitic mice.

Figure 1.

   Figure 1
   [84]Open in a new tab

   Dual Gut Barrier Impairment in Colitis. A,B) Western blot analysis of
   the level of IEB barrier proteins (ZO‐1 and Occludin) and GVB barrier
   proteins (β‐catenin and PV1) in the mucosa and lamina propria of
   control and chronic DSS‐induced colitis mice. C) Serum FITC‐dextran
   70 kDa (FD70) concentration in control and chronic DSS‐induced colitis
   group. Representative images of goblet cells D) and tight junctions E)
   in colonic tissue from control and chronic DSS‐induced colitis mice by
   transmission electron microscopy (TEM). Red arrows indicated sites of
   tight junctions. F) Representative immunofluorescence images of PV1
   (red) expression on colonic vessels stained with CD31 (green) in
   control and chronic DSS‐induced colitis mice. DAPI (blue) stained cell
   nuclei. Quantification of CD31^+ cells G), PV1^+ cells H), and
   CD31^+/PV1^+ cells I) in colon tissue from control and chronic
   DSS‐induced colitis mice. J) Representative immunofluorescence images
   of PV1 (red) expression on colonic vessels stained with CD31 (green) in
   healthy individuals and UC patients. DAPI (blue) stained cell nuclei.
   Quantification of CD31^+ cells K), PV1^+ cells L), and CD31^+/PV1^+
   cells M) in colon tissue from healthy individuals and UC patients.
   About 50 cells were surveyed in each sample, and five fields of vision
   were used for quantification. IEB, intestinal epithelial barrier; GVB,
   gut vascular barrier; DSS, dextran sulfate sodium; DAPI,
   4′,6′‐diamidino‐2‐phenylindole; UC, ulcerative colitis. Scale bars are
   shown in the figures. n = 5 for each group. Data are mean ± SEM. **p
   <0.01, ***p <0.001, ****p <0.0001.

2.2. Increased Presence of GFAP^+/S100 Calcium Binding Protein B (S100β)^+
EGCs in Inflamed Colonic Tissues

   As depicted by the ultrastructure of DSS‐induced colitis mouse colon
   tissues, gut microvascular endothelial cells are surrounded by EGCs and
   pericytes, collectively forming the intact gut‐vascular unit^[ [85]^5
   ^] (Figure [86] 2A). The IEB is made up of a monolayer of intestinal
   epithelial cells structured into invaginations (crypts) and finger‐like
   projections (villi) in the small intestine, as well as a series of
   crypts alternating with a flat epithelial surface in the colon.^[
   [87]^20 ^] EGCs anatomically reside within the IEB/GVB (Figure [88]2A).
   Given the pivotal anatomical location of EGCs, involved in both the
   formation of GVB and their association with the IEB, we further
   assessed the expression of EGCs in the inflamed colon full‐layer
   sections of mice with DSS‐induced colitis. Regarding DSS‐induced
   inflammation, S100β^+ cells constituted the great majority of GFAP^+
   cells, representing our primary identification of the EGCs subtype
   (Figure [89]S2A, Supporting Information). Previous evidence indicates
   continuous dynamic regulation of gene and protein expression in
   subtypes of enteric glia, alongside potential variations in
   responsiveness to intercellular mediators both within and across these
   subtypes.^[ [90]^21 , [91]^22 ^] Moreover, it is evident that
   GFAP^+/S100β^+ EGCs in colitic tissues predominantly localize within
   the mucosa and submucosa (i.e., the areas affected by inflammation).
   The identification of elevated GFAP and S100β protein expression in the
   inflamed colonic tissues after the removal of the serosa and myenteron
   provided further evidence in support of this finding (Figure [92]S2B,C,
   Supporting Information). To reinforce this finding, we performed
   submucosal whole‐mount IF staining (Figure [93]2B–E). Our analysis
   revealed a modest reduction in GFAP^+ EGCs in the submucosa of colitic
   groups compared to controls, while no notable disparity in S100β^+ EGCs
   was observed between the cohorts. Additionally, the GFAP^+/S100β^+ EGCs
   exhibited a significant upregulation in the submucosa of mice with
   DSS‐induced colitis.

Figure 2.

   Figure 2
   [94]Open in a new tab

   Increased abundance of GFAP^+/S100β^+ EGCs in inflamed colonic tissues.
   A) EGCs participate in gut vascular barrier (GVB) formation and are
   closely associated with intestinal epithelial barrier (IEB).
   Representative whole‐mount staining B) and quantification C–E) of GFAP
   (red) and S100β (green) in the submucosal plexus of control and chronic
   DSS‐induced colitis mice. DAPI (blue) stained the cell nuclei. The
   percentage of GFAP^+ EGCs F,G) S100β^+ EGCs H,I) and GFAP^+/S100β^+
   EGCs J,K) were assayed in the colonic mucosa and lamina propria of
   control and DSS‐induced colitis mice. L) Representative
   immunofluorescence images of EGCs labeled with GFAP (red) and S100β
   (green) in full‐layer colonic sections of healthy individuals and UC
   patients. DAPI (blue) stained the cell nuclei. Quantification of GFAP^+
   cells M), S100β^+ cells N), and GFAP^+/S100β^+ cells O) in colon tissue
   from healthy individuals and UC patients. P) Unsupervised subclustering
   of hEGC cluster 6 in healthy individuals (green dots) and those with UC
   (red dots). Q) Proportion of EGCs expressing GFAP or S100β in each
   subpopulation. Single cells possessing greater than 3 and non‐zero
   values are defined as GFAP^+ and S100β^+, respectively (n = 5 Drop‐seq
   replicates). Unsupervised analysis of EGC clusters shows that EGC
   subpopulation gene signatures correlate with S100β R), GFAP S), and
   S100β+GFAP T) enrichment in EGC#1, EGC#2, and EGC#0, respectively. U)
   Schematic representation of the mucosal and submucosal microenvironment
   in the gut under healthy and inflamed conditions: In the healthy state,
   GFAP^+ EGCs predominate, while in the inflamed state, there is an
   increase in GFAP^+/S100β^+ EGCs. About 50 cells were surveyed in each
   sample, and five fields of vision were used for quantification. DSS,
   dextran sulfate sodium; EGCs, enteric glial cells; DAPI,
   4′,6′‐diamidino‐2‐phenylindole; UC, ulcerative colitis. Scale bars are
   shown in the figures. n = 5 for each group. Data are mean ± SEM. **p
   <0.01, ***p <0.001.

   To furnish direct evidence of increased expression of GFAP^+/S100β^+
   EGCs in the inflamed colonic mucosa and submucosa and to gain further
   insights into downstream signaling events, we isolated primary EGCs
   from the inflamed colonic mucosa and submucosa of DSS‐induced colitis
   mice. Subsequently, we utilized flow cytometry to assess the expression
   of marker proteins (Figure [95]2F–K). Consistent with the observations
   of submucosal immunofluorescence, although no noticeable distinction
   was observed in S100β^+ EGCs across the groups, there was a reduction
   in GFAP^+ EGCs. Conversely, there was a notable increase in
   GFAP^+/S100β^+ EGCs within the submucosa of the colitic groups compared
   to the control groups. Crucially, in the inflamed biopsy tissues of
   patients with active UC, GFAP^+/S100β^+ EGCs were predominantly found
   among areas with inflammatory injury (Figure [96]2L,O). The findings
   observed in patients with active UC are consistent with those observed
   in mice with DSS‐induced colitis, where GFAP^+ EGCs were slightly
   decreased in inflamed intestinal tissues compared to normal controls
   (Figure [97]2L,M), and the expression of S100β^+ EGCs showed no
   significant difference between the two groups (Figure [98]2L,N). This
   further supports the notion of increased expression of GFAP^+/S100β^+
   EGCs in inflamed colonic tissues.

   To explore the heterogeneity of EGC in colitis tissues, we analyzed
   single‐cell RNA sequencing (scRNA‐seq) data from colonic mesenchyme
   cell populations of healthy controls and patients with UC. After doing
   an unsupervised clustering analysis, we identified 12 disparate stromal
   cell populations (Figure [99]S2H, Supporting Information). Utilizing
   glial cell markers S100β and GFAP, we designated cluster 6 as the EGC
   population. Subsequently, we defined three subpopulations inside the
   EGC cluster 6 using unsupervised sub‐clustering analysis, denoted as
   EGC#0–EGC#2 (Figure [100]S2I, Supporting Information). As shown in
   Figure [101]2P, we discovered that whereas EGC#0 and EGC#2 are more
   common in patients with UC, cluster EGC#1 is more prevalent in healthy
   material. Upon comparing the expression of S100β and GFAP, within these
   subclusters, we discovered that while S100β was expressed to variable
   degrees within most subpopulations, EGC#1 exhibited consistently high
   S100β expression, with 48.5% of S100β^+ cells (Figure [102]S2K,
   Supporting Information; Figure [103]2Q,R), in contrast to 36.4% and 15%
   in EGC#0 and EGC#2, respectively. On the other hand, EGC#2 had 77% of
   GFAP^+ cells, whereas EGC#0 and EGC#1 had 7.7% and 15.3%, respectively
   (Figure [104]S2J, Supporting Information; Figure [105]2Q,S).
   Furthermore, GFAP expression was higher in EGC#2, whereas S100β was
   substantially expressed in EGC#1. Most notably, we observed that
   GFAP^+/S100β^+ EGCs were predominantly enriched in cluster 0, comprised
   almost entirely of EGCs sourced from UC patients (Figure [106]2Q,T,
   highlighted in yellow). Delineating GFAP^+/S100β^+, S100β^+, and GFAP^+
   labels for EGC#0, EGC#1, and EGC#2, respectively, this scRNA‐seq study
   highlights the variability of EGC populations. The evidence above
   substantiates the assertion that multiple phenotypes of EGCs exist
   within colonic tissues, with a conspicuous increase in GFAP^+/S100β^+
   EGCs observed in the inflamed colon (Figure [107]2U).

2.3. The Protective Role of DHA in Colitis – Restoring Dual Gut Barrier
Function

   To explore the anti‐inflammatory effects of DHA in chronic colitis, we
   administered intraperitoneal injections of 20 mg kg^−1 (body weight) of
   DHA daily to DSS‐treated C57BL/6 mice (Figure [108] 3A). After one week
   of continuous administration, the treatment was paused for one week
   before resuming, repeating four rounds to conclude the modeling. During
   DSS administration, the DHA‐treated group showed significantly reduced
   body weight loss compared to the colitic mice (Figure [109]S3A,
   Supporting Information), manifested as body weight loss appearing
   earlier and being more severe at any time point in DSS‐induced mice not
   treated with DHA. At the conclusion of the experiment, the improvement
   in colon length shortening (Figure [110]S3B, Supporting Information)
   observed in the DHA‐treated group confirmed a significant difference in
   disease activity scores (Figure [111]S3C, Supporting Information) and
   spleen index (Figure [112]S3D, Supporting Information) between the
   groups. Histological examination of colon cross‐sections revealed that
   mice in the DHA‐treated group exhibited smaller areas of colonic
   inflammation, improved tissue edema, and less disrupted crypt
   architecture compared to the DSS group, with contractible immune cell
   infiltration and lower colonic histological scores (Figure [113]S3E,F,
   Supporting Information). In addition, our real‐time quantitative
   polymerase chain reaction (qPCR) results (Figure [114]S3G, Supporting
   Information) revealed that compared to colitis mice untreated with DHA,
   those in the DHA‐treated group exhibited significantly reduced
   expression of inflammatory cytokines, such as interleukin 6 (IL6),
   IL1B, interferon gamma (IFNG), and TNF. The expression of the
   anti‐inflammatory cytokine IL10 was augmented following DHA
   administration. Mice colon cross‐sections in steady state (no DSS)
   showed no signs of pathology. Furthermore, DHA treatment markedly
   decreased the levels of inflammatory cytokines IL‐1β (Figure [115]S3H,
   Supporting Information) and TNF (Figure [116]S3I, Supporting
   Information) in inflamed colonic tissues, while elevating the content
   of the anti‐inflammatory cytokine IL‐10 (Figure [117]S3J, Supporting
   Information). Additionally, DHA supplementation also attenuated the
   elevation of LPS levels in the peripheral blood of DSS‐induced colitis
   mice (Figure [118]S3K, Supporting Information), providing more
   extensive evidence for the protective role of DHA in the context of
   colitis.

Figure 3.

   Figure 3
   [119]Open in a new tab

   DHA protects dual gut barrier function in colitis. A) Flowchart of
   animal experiment modeling process. B) Representative images of tight
   junctions in colonic tissue from each group of mice by transmission
   electron microscopy (TEM). Red arrows indicated sites of tight
   junctions. C) Representative immunofluorescence images of PV1 (red)
   expression on colonic vessels stained with CD31 (green) in each group.
   DAPI (blue) stained cell nuclei. Quantification of CD31^+ cells D),
   PV1^+ cells E), and CD31^+/PV1^+ cells F) in colon tissue from each
   group. About 50 cells were surveyed in each sample, and five fields of
   vision were used for quantification. G) The integrity of the GVB was
   assessed by measuring the fluorescence intensity of FD70 (green),
   referring to the emergence of FD70 in the liver and spleen after gut
   lumen exposure to FD70 in different groups. Quantification of
   fluorescence intensity of FD70 in the spleen H) and liver I) from each
   group. J) Experimental illustration of DHA's impact on IEB/GVB via EGCs
   modulation under inflammatory conditions. K‐L) Western blot analysis of
   the level of IEB barrier proteins (ZO‐1, Occludin, Claudin‐2, and
   Claudin‐1) in HT29 under EGCs conditioned medium treatment with or
   without DHA therapy under inflammatory conditions. M‐N) Western blot
   analysis of the level of GVB barrier proteins (ZO‐1, β‐catenin,
   Occludin, PV1, Claudin‐2, and Claudin‐1) in HUVECs under EGCs
   conditioned medium treatment with or without DHA therapy under
   inflammatory conditions. DHA, dihydroartemisinin; DSS, dextran sulfate
   sodium; DAPI, 4′,6′‐diamidino‐2‐phenylindole; FD70, FITC‐labeled
   fluorescent dextran 70 KDa; IEB, intestinal epithelial barrier; GVB,
   gut vascular barrier; EG‐CM, enteric glial‐conditioned medium; EGCs,
   enteric glial cells. Scale bars are shown in the figures. n = 5 for
   each group (B, C, and G). Data are mean ± SEM. *p < 0.05, **p < 0.01,
   ***p < 0.001, ****p < 0.0001.

   To determine whether DHA can improve the dual gut barrier function in
   DSS‐induced colitis, we performed the western blot analysis on colonic
   tissues from mice with mucosal and muscular layers removed
   (Figure [120]S3L,M, Supporting Information). As predicted, our results
   indicated that DHA administration significantly restored the
   downregulation of markers related to IEB integrity induced by DSS, such
   as the TJ proteins (ZO‐1 and Occludin). More importantly, the AJ
   protein β‐catenin associated with GVB was elevated following DHA
   treatment, while the expression of GVB permeability‐related molecule
   PV1 was reduced by DHA administration. To further investigate the
   reparative effect of DHA on GVB in colitis, the ultrastructure of GVB
   in the colonic tissues of each group of mice was examined using
   transmission electron microscopy (TEM), as shown in Figure [121]3B.
   Moreover, we observed through immunofluorescence staining that the
   colonic whole‐layer sections of mice treated with DHA (Figure [122]3C)
   exhibited a more pronounced reduction in CD31^+/PV^+ cells compared to
   colitis mice with no DHA treated (Figure [123]3F). The increase in
   CD31^+ cells (Figure [124]3D) and PV1^+ cells (Figure [125]3E) in the
   inflamed colons of DSS‐induced mice was also blocked by DHA treatment,
   indicating that DHA can inhibit neovascularization in colitis mice,
   especially the generation of highly permeable endothelial cells. The
   results revealed a significant reduction in the fenestration of
   vascular endothelial cells in the colonic tissues of mice treated with
   DHA compared to those without DHA treatment.

   We further evaluated the effect of DHA treatment on gut barrier
   function in colitis by conducting fluorescein isothiocyanate
   (FITC)‐dextran intestinal permeability assay in each group of mice.
   This experiment serves as the most direct assessment of GVB
   permeability.^[ [126]^23 ^] In line with TEM observations, DHA
   administration significantly reduced the levels of FITC‐dextran in the
   serum of colitis mice (Figure [127]S3N, Supporting Information).
   Indeed, the frozen sections of the liver and spleen from mice treated
   with DHA exhibited significantly attenuated green fluorescence
   intensity under the microscope as compared with colitis mice not
   treated with DHA (Figure [128]3G–I), providing evidence for the
   extensive role of DHA in mediating dual gut barrier protection in
   colitis. Given that EGCs are positioned at the critical anatomical site
   associated with IEB/GVB (Figure [129]2A), and the occurrence of the
   abnormally increased proportion of GFAP^+/S100β^+ EGCs in inflamed
   colonic tissues correlates positively with the presence of dual gut
   barrier damage, more importantly, both are significantly improved after
   DHA treatment. These findings suggest a close association between DHA's
   reparative effect on dual gut barrier damage during enteritis and EGCs.
   To further elucidate the impact of DHA on dual gut barrier function
   through its action on EGCs in the context of inflammation, we incubated
   EGCs with DHA in vitro and obtained the conditioned medium.
   Subsequently, we treated intestinal epithelial cells (HT29) and
   vascular endothelial cells (HUVECs) with the conditioned medium
   separately and assessed the expression of relevant barrier proteins
   (Figure [130]3J). Regarding the impact of DHA‐treated TNF‐activated
   EGCs on IEB, compared to the untreated inflammation group, western blot
   analysis revealed an upregulation in the expression of TJ proteins
   ZO‐1, Occludin, and Claudin‐1, along with a downregulation in Claudin‐2
   (a modulator of intestinal inflammation‐related leaky gut barrier)^[
   [131]^24 ^] (Figure [132]3K,L). These findings demonstrated the
   reparative effect of DHA‐treated TNF‐activated EGCs on IEB.
   Consistently, the impact of DHA‐treated TNF‐activated EGCs on GVB was
   also favorable. Compared to the inflammation group without DHA
   treatment, similar trends were observed in the expression of TJ
   proteins within the GVB as observed in the IEB (Figure [133]3M,N).
   Additionally, the expression of AJ‐associated protein β‐catenin in GVB
   was upregulated, along with a decrease in the fenestration protein PV1
   in the DHA‐treated inflammation group. These results indicated that DHA
   administration not only restores the integrity of IEB but also repairs
   the GVB.

2.4. Modulation of EGC Phenotypes by DHA in Colitis

   To further dissect the protective role of DHA in experimental colitis
   in vivo, the western blot analysis of colonic tissues, depleted of the
   serosa and myenteron, indicated a dramatic decrease in the expression
   of GFAP and S100β protein in the DHA‐treated group compared to the
   untreated group of colitis mice (Figure [134]S4A,B, Supporting
   Information), reinforcing the beneficial role of DHA in DSS‐induced
   colitis. Immunofluorescence staining of colonic cross‐sections
   (Figure [135]S4C, Supporting Information) also exhibited a similar
   trend, wherein the fluorescence intensity of GFAP and S100β in the
   colitis mice treated with DHA was weaker compared to the untreated
   group. Furthermore, we observed a more obvious augment in the number of
   GFAP^+ cells in the DHA‐treated group compared to those untreated with
   DHA but with colitis (Figure [136]4B), as indicated by
   immunofluorescence analysis of submucosal whole‐mount IF staining
   (Figure [137] 4A). Interestingly, the treatment with DHA showed no
   significant effect on the abundance of S100β^+ cells (Figure [138]4C).
   Of particular note, DHA treatment significantly suppressed the content
   of GFAP^+/S100β^+ cells in DSS‐induced colitis (Figure [139]4D),
   providing further substantiation for the broader pivotal role of DHA in
   managing gut inflammation. Given the crucial role of GFAP^+ EGCs in the
   repair of IBD,^[ [140]^25 ^] we thus hypothesize that in the
   inflammatory context, DHA not only inhibits the differentiation of
   GFAP^+ EGCs into GFAP^+/S100β^+ EGCs but also promotes the
   transformation of GFAP^+/S100β^+ EGCs to GFAP^+ EGCs.

Figure 4.

   Figure 4
   [141]Open in a new tab

   DHA regulation of EGC heterogeneity. A) Representative whole‐mount
   staining and quantification B–D) of GFAP (red) and S100β (green) in the
   submucosal plexus of different groups of mice. DAPI (blue) stained the
   cell nuclei. Quantification of GFAP^+ EGCs (B), S100β^+ EGCs (C), and
   GFAP^+/S100β^+ EGCs (D) in the submucosal plexus of different groups of
   mice. E) Schematic diagram of mouse colonic primary EGCs extraction for
   flow cytometry analysis. The percentage of GFAP^+ EGCs F) and
   GFAP^+/S100β^+ EGCs G) were assayed in the colonic mucosa and lamina
   propria of colitis mice treated with or without DHA and corresponding
   control mice. H‐I) Western blot analysis of the expression of EGCs
   surface markers (GFAP and S100β) under inflammatory conditions with or
   without DHA treatment. The intracellular levels of GDNF J), IL‐1β K),
   and CXCL10 L) in EGCs of each group determined by ELISA. M) The
   extracellular IL‐1β levels in EGCs of each group determined by ELISA.
   N) Graphical abstract of the modulation of EGC heterogeneity by DHA
   under inflammatory conditions: DHA inhibits the transformation of
   GFAP^+ EGCs into GFAP^+/S100β^+ EGCs and promotes the conversion of
   GFAP^+/S100β^+ EGCs into GFAP^+ EGCs in the context of inflammation.
   DHA, dihydroartemisinin; DSS, dextran sulfate sodium; LMMP,
   longitudinal muscle layer, and the longitudinal muscle/myenteric
   plexus; EGCs, enteric glial cells; DAPI,
   4′,6′‐diamidino‐2‐phenylindole. Scale bars are shown in the figures. n
   = 5 for each group (A, F, and G). Data are mean ± SEM. **p < 0.01, ***p
   < 0.001, ****p < 0.0001.

   To investigate the exact relationship between DHA inhibition of
   GFAP^+/S100β^+ EGCs phenotypic transformation in inflamed colonic
   tissues and colitis remission, we extracted primary EGCs from colonic
   tissues of each group of mice according to the protocol provided by
   Wang et al.^[ [142]^26 ^] We then conducted flow cytometry to assess
   their phenotypic changes (Figure [143]4E). In DSS‐induced colitis, flow
   cytometry analysis revealed that 1.24% of EGCs were GFAP^+, while in
   the colitis group treated with DHA, the proportion of GFAP^+ EGCs
   increased to 23.5% (Figure [144]4F). Consistent with previous
   observations, DSS did not induce an increase in the proportion of
   S100β^+ cells in colitis mice. However, treatment with DHA induced this
   property, especially in the colitis group, as the ratio of S100β^+
   cells in mice treated with DHA was more than doubled compared to the
   untreated group (Figure [145]S4D, Supporting Information). It is worth
   noting that the significant increase in GFAP^+/S100β^+ EGCs observed in
   the DSS‐induced colitis group was markedly reduced after DHA
   administration (Figure [146]4G), reinforcing the hypothesis that DHA
   can induce the transition of GFAP^+/S100β^+ EGCs to GFAP^+ EGCs in the
   inflamed colon. Next, we sought to determine whether the EGC‐specific
   changes observed in vivo could also be replicated in vitro. We
   pre‐incubated EGC line CRL‐2690 with TNF for 24 h to simulate the
   inflammatory damage caused by colitis (Figure [147]4H). Subsequently,
   Western blot analysis revealed that DHA treatment reduced the elevated
   protein levels of GFAP and S100β following TNF pre‐incubation
   (Figure [148]4I), providing decisive evidence for the fact that DHA can
   convert the increased GFAP^+/S100β^+ EGCs observed in DSS‐induced
   colitis to GFAP^+ EGCs.

   Reactive EGC transcriptome profiling in humans and mice indicates that
   EGCs enhance the generation of neurotrophin (like glial cell‐derived
   neurotrophic factor (GDNF)), inflammatory cytokines (like IL‐1β), and
   chemokines (like C‐X‐C motif chemokine ligand 10 (CXCL10)).^[ [149]^8 ,
   [150]^9 ^] Furthermore, we found that regardless of inflammation, DHA
   treatment significantly increased the intracellular levels of the
   barrier‐enhancing factor GDNF in EGCs (Figure [151]4J), as reported
   previously,^[ [152]^27 ^] without affecting the secretion of GDNF by
   EGCs (Figure [153]S4E, Supporting Information). This suggests that the
   barrier‐protective effect exerted by EGC through GDNF is not mediated
   via secretion. Inflammation led to a significant increase in the
   intracellular levels of IL‐1β (Figure [154]4K) and CXCL10
   (Figure [155]4L) in EGCs. However, treatment with DHA markedly reduced
   their intracellular levels in EGCs. Even more importantly, DHA
   alleviates the increase in IL‐1β secretion induced by inflammation in
   EGCs (Figure [156]4M). Pre‐treatment with TNF significantly reduces the
   ability of EGCs to secrete CXCL10, indicating that inflammation impairs
   the chemotactic ability of EGCs toward immune cells (Figure [157]S4F,
   Supporting Information). However, DHA treatment does not increase the
   secretion of CXCL10 by EGCs. Building on the previous findings
   (Figure [158]3), this also underscores the close association between
   the functional damage of IEB/GVB and GFAP^+/S100β^+ EGCs. These studies
   demonstrate that DHA possesses the ability to promote the phenotypic
   transition of increased GFAP^+/S100β^+ EGCs to GFAP^+ EGCs during
   intestinal inflammation while simultaneously inhibiting the
   differentiation of GFAP^+ EGCs into GFAP^+/S100β^+ EGCs
   (Figure [159]4N).

2.5. DHA‐Induced Cell Cycle Arrest and Apoptosis of GFAP^+/S100β^+ EGCs in
the Inflammatory State

   To investigate the molecular mechanisms underlying the effect of DHA on
   intestinal inflammation, we conducted RNA‐sequencing (RNA‐Seq)
   transcriptomic analysis on EGCs treated with and without DHA following
   TNF pre‐incubation. Volcano plots comparing DHA treated with untreated
   inflammatory samples showed 378 up‐regulated genes and 226
   down‐regulated genes (Figure [160] 5A). We subsequently performed Gene
   Ontology (GO) analysis on these differentially expressed genes (DEGs)
   to elucidate the biological processes (BP), cellular components (CC),
   and molecular functions (MF) they are engaged in, aiming to further
   investigate the impact of DHA treatment on TNF pre‐incubated EGCs
   (Figure [161]S5A, Supporting Information). It revealed that compared to
   the untreated inflammatory EGCs, the DHA‐treated inflammatory EGCs
   showed enrichment in BP related to chromosome segregation, CC
   associated with the chromosomal region, and MF related to adenosine
   triphosphate (ATP) and DNA‐related activities. This indicates that DHA
   treatment mainly affects the cell division and proliferation of
   inflammatory EGCs. The Kyoto Encyclopedia of Genes and Genomes (KEGG)
   analysis of DEGs revealed that the enriched pathways primarily involve
   cell cycle, DNA replication, cellular senescence, homologous
   recombination, base excision repair, and mismatch repair‐related KEGG
   pathways (Figure [162]5B). Moreover, these pathways mainly belong to
   the KEGG categories of cellular growth and death (hsa04110 and
   hsa04218), replication and repair (hsa03030, hsa03440, hsa03460,
   hsa03410, and hsa03430). The heatmap in Figure [163]5C illustrates the
   DEGs associated with cell cycle, DNA replication, cellular senescence,
   and homologous recombination. The RNA‐Seq data revealed that DHA
   treatment had the most significant impact on the cell cycle of EGCs
   pre‐incubated with TNF. Consequently, we assessed the cell cycle
   changes in EGCs from each group (Figure [164]5D). We observed a
   dramatic increase in the proportion of cells in the G0/G1 phase and a
   corresponding decrease in cells in the S and G2/M phases in EGCs after
   DHA treatment, suggesting that DHA can arrest the cell cycle of EGCs at
   the G0/G1 phase (Figure [165]5E), further indicating that the effect of
   DHA on inflammatory EGCs involves proliferation inhibition.

Figure 5.

   Figure 5
   [166]Open in a new tab

   In the inflammatory condition, DHA causes GFAP^+/S100β^+ EGCs to enter
   a cell cycle arrest and undergo apoptosis. A) Volcano plot of actively
   transcribed genes in inflamed EGCs after DHA treatment, with
   significantly differentially expressed genes (DEGs) (P‐value < 0.05,> ±
   twofold) marked red (upregulated) and blue (downregulated). B) KEGG
   pathway enrichment analysis of DEGs in inflamed EGCs after DHA
   treatment. C) Heatmap of partial DEGs for KEGG enrichment analysis. D)
   Cell cycle analysis of inflammatory EGCs with or without DHA treatment.
   E) Quantitative analysis of cell cycle. F) Cell apoptosis analysis of
   inflammatory EGCs with or without DHA treatment. G) Quantitative
   analysis of cell apoptosis. H‐I) Western blot analysis of apoptotic
   protein (Cleaved Caspase‐3, Bcl‐2, and Bax) expression in EGCs treated
   with or without DHA under inflammatory conditions. J) Representative
   immunofluorescence images of Cleaved Caspase‐3 (red) expression on EGCs
   stained with GFAP (magenta) and S100β (green) in full‐layer colonic
   sections of colitis mice treated with or without DHA and corresponding
   control mice. DAPI (blue) stained cell nuclei. Quantification of
   GFAP^+/Cleaved Caspase‐3^+ cells (K), S100β^+/Cleaved Caspase‐3^+ cells
   L), and GFAP^+/S100β^+/Cleaved Caspase‐3^+ cells M) in colon tissue
   from each group. About 50 cells were surveyed in each sample, and five
   fields of vision were used for quantification. DHA, dihydroartemisinin;
   DSS, dextran sulfate sodium; EGCs, enteric glial cells; DAPI,
   4′,6′‐diamidino‐2‐phenylindole; KEGG, Kyoto Encyclopedia of Genes and
   Genomes. Scale bars are shown in the figures. n = 5 for each group J).
   Data are mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p <
   0.0001.

   Further expanding our current RNA‐Seq study and complementing it with
   flow cytometry analysis, we can confirm no statistically significant
   differences in the early apoptosis rates among the cell groups
   (Figure [167]5F,G). However, DHA treatment significantly increased the
   late apoptosis rate of EGCs, regardless of whether the EGCs were
   pre‐treated with TNF, suggesting a lethal role of DHA in EGCs.
   Moreover, we detected the expression of apoptosis‐related proteins in
   EGCs pre‐treated with TNF and then treated with DHA. As anticipated,
   DHA treatment significantly upregulated the expression of Bax and
   Cleaved Caspase‐3 in inflamed EGCs and downregulating the expression of
   Bcl‐2, supporting the notion that DHA can induce apoptosis in
   GFAP^+/S100β^+ EGCs (Figure [168]5H,I). To further investigate the
   apoptosis of GFAP^+/S100β^+ EGCs in vivo, we detect the expression of
   the well‐known apoptosis protein Cleaved Caspase‐3^[ [169]^28 ^] by
   immunofluorescence in healthy control and UC patients’ colon biopsy
   (Figure [170]S5B–E, Supporting Information). Consistent with the
   observations in the in vitro inflammatory‐induced EGCs model, we found
   no significant difference in the apoptosis rates of GFAP^+ EGCs
   (Figure [171]S5C, Supporting Information) and S100β^+ EGCs
   (Figure [172]S5D, Supporting Information) in the inflamed colon of UC
   patients compared to healthy controls. However, the UC group indicated
   an overall increased number of apoptotic GFAP^+/S100β^+ EGCs
   (Figure [173]S5E, Supporting Information). In another set of
   experiments, we aimed to investigate the effect of DHA on inducing
   apoptosis of GFAP^+/S100β^+ EGCs in vivo (Figure [174]5J–M). There was
   no statistically significant apoptosis observed in GFAP^+ EGCs among
   the groups (Figure [175]5K). We found that both S100β^+ EGCs
   (Figure [176]5L) and GFAP^+/S100β^+ EGCs (Figure [177]5M) exhibited
   increased apoptosis in DSS‐induced colitis mice. However, only
   GFAP^+/S100β^+ EGCs showed a significant decrease in apoptosis rate
   after DHA treatment. Taken together, these findings demonstrate that
   DHA induces cell cycle arrest and apoptosis in EGCs, particularly in
   GFAP^+/S100β^+ EGCs.

2.6. DHA Modulated EGC Heterogeneity by Regulating Gut Dysbiosis in
DSS‐Induced Colitis

   Previous studies,^[ [178]^21 , [179]^22 , [180]^29 , [181]^30 ^] in
   conjunction with our aforementioned findings, reveal dynamic changes in
   the expression levels of major glial markers in EGCs, influenced by
   environmental factors and stimuli encountered by EGCs. Moreover, the
   gut microbiota profoundly impacts the abundance of EGCs within the
   gut.^[ [182]^11 , [183]^31 ^] Kabouridis et al. definitively
   demonstrated that EGCs are the primary target cells of the gut
   microbiota, and in‐situ gut microbiota regulates the homeostasis of
   EGCs within the intestinal mucosa.^[ [184]^11 ^] Additionally, the
   expression of GFAP, which is observable, varies depending on the
   neuroglial state,^[ [185]^8 ^] subtype,^[ [186]^21 ^] and genetic
   targeting strategies.^[ [187]^32 ^] Spica et al. suggested that
   glial‐derived S100β is involved in mediating the regulation of gut
   microbiota diversity.^[ [188]^33 ^] Thus, to investigate whether DHA
   treatment alters the gut microbiota of colitis mice and consequently
   affects EGC heterogeneity, we evaluated the intestinal microbiota
   diversity in mouse feces.

   Results from 16S ribosomal RNA (rRNA) gene sequencing of all four
   groups (Control, DHA, DSS, and DSS+DHA) of mice showed that the
   richness and diversity of microbial communities, as indicated by the
   Chao index (Figure [189] 6A), were significantly increased after DHA
   treatment, regardless of the occurrence of colitis. Furthermore,
   compared to normal control mice, at the phylum level, there was a
   significant decrease in the proportions of Actinobacteriota and
   Campylobacterota but a significant increase in Proteobacteria in the
   feces of colitis mice (Figure [190]6B). At the class level,
   Actinobacteria was significantly reduced in colitis mice, while
   Gammaproteobacteria showed a significant increase. Treatment with DHA
   helped improve this situation (Figure [191]6C). At the order level,
   Bifidobacteriales showed an observable decrease in the feces of colitis
   mice, but DHA administration did not improve this condition
   (Figure [192]6D). At the family level, there was a notable increase in
   the abundance of Bacteroidaceae in the colitis group compared to the
   control group. However, in the colitis group treated with DHA, the
   proportion of Bacteroidaceae decreased (Figure [193]6E). Similarly,
   Bacteroides at the genus level showed a significant upregulation in the
   DSS‐induced colitis group, which was markedly attenuated by DHA
   treatment (Figure [194]6F). Furthermore, at the species level, the
   significantly elevated proportion of Bacteroides_stercorirosoris in the
   colitis group decreased after DHA administration (Figure [195]6G).
   These findings collectively suggest that the alterations in Bacteroides
   diversity are crucial in the mitigation of DSS‐induced colitis by DHA.
   The principal co‐ordinate analysis (PCoA) employed to quantify beta
   diversity^[ [196]^34 ^] demonstrated that DHA therapy dramatically
   affected the gut microbiota composition in colitis animals
   (Figure [197]6H). To further identify specific taxonomic units with
   differential abundance between DHA‐treated and untreated colitis mice,
   we employed Linear discriminant analysis Effect Size (LEfSe) analysis
   to compare bacterial abundances at each taxonomic level.^[ [198]^35 ^]
   The Linear Discriminant Analysis (LDA) scores revealed the dominating
   groupings and their influence at various taxonomic gradations, ranging
   from phylum to species. There was a significant difference in the
   abundance of 18 bacterial genera between DHA‐treated and untreated
   colitis mice, as shown by an LDA score[log10]>3.6 and a P‐value <0.05
   (Figure [199]6I,J). At the genus level, Bacteroides exhibited a
   significant increase in the fecal samples of colitis mice, and
   Muribaculaceae showed a notable increase in the fecal samples of
   colitis mice treated with DHA.

Figure 6.

   Figure 6
   [200]Open in a new tab

   16S ribosomal RNA (rRNA) sequencing analysis of gut microbiota of
   colitis mice treated with or without DHA and corresponding control
   mice. A) Chao1 index of each group showed the 𝛼‐diversity of the
   microbial community. The community histogram showed the microbial
   compositional profiling at the phylum B), class C), and order D)
   levels. Each row represents each mouse and n = 3 for each group.
   Heatmap exhibited the relative abundance of microbial compositional
   profiling at the family E), genus F), and species G) levels. Each
   column represents each mouse and n = 3 for each group. H) A principal
   co‐ordinates analysis revealed the gut microbiome's diversity. Each
   point represents each mouse and n = 3 for each group. Clustering's
   significance was assessed using analysis of similarities (ANOSIM). I)
   Linear discriminant analysis (LDA) determines the most abundant genus
   in various groupings. Taxa that meeting an LDA significant threshold of
   3.6 are shown. J) A cladogram revealed the disparity in richness, as
   well as the group with the greatest abundance. The brightness of each
   dot positively relates to its effect. K‐L) Western blot analysis of the
   expression of GFAP, S100β, and bacterial product‐related signaling
   molecules (c‐Fos, TLR4, NF‐κB, and iNOS) of EGCs after pre‐incubation
   with LPS with or without DHA treatment. DHA, dihydroartemisinin; DSS,
   dextran sulfate sodium; EGCs, enteric glial cells; LPS,
   lipopolysaccharide. Data are mean ± SEM. *p < 0.05, **p < 0.01, ***p <
   0.001, ****p < 0.0001.

   To further investigate the regulatory mechanism of DHA on EGC
   heterogeneity mediated by microbial products, we pre‐incubated EGCs
   with LPS (5 µg mL^−1) for 2 h before DHA treatment to observe the
   stimulatory effect of microbial products on EGCs (Figure [201]6K).
   Consistent with the results of 16S rRNA sequencing, we observed
   significant activation of LPS‐pre‐treated EGCs, manifested by markedly
   increased expression of the EGC activation marker c‐Fos^[ [202]^36 ^]
   (Figure [203]6K,L). TLR4, which mediates EGC homeostasis^[ [204]^37 ^]
   and is a pivotal receptor for interactions between microbes and their
   products with EGCs,^[ [205]^7 , [206]^38 , [207]^39 ^] is also
   significantly upregulated in LPS‐pre‐treated EGCs. LPS stimulation
   further enhances the upregulation of downstream molecules NF‐κB and
   inducible nitric oxide synthase (iNOS) through TLR4, which is
   significantly correlated with the upregulation of S100β and GFAP
   proteins.^[ [208]^33 , [209]^40 , [210]^41 ^] This is consistent with
   previous research findings that human EGCs respond to microbial
   dysbiosis through differential expression of Toll‐like receptors (TLRs)
   and production of nitric oxide (NO), with pathogens (rather than
   probiotics) significantly inducing overexpression of the S100β protein,
   and EGCs releasing NO.^[ [211]^40 ^] It is noteworthy that treatment
   with DHA significantly reversed the aforementioned conditions, further
   indicating that the regulation of EGC heterogeneity by DHA under
   inflammatory conditions is achieved through the modulation of microbial
   products.

3. Discussion

   Although there is a substantial correlation between the pathophysiology
   of chronic enteritis and the phenotypic plasticity of EGCs,^[ [212]^7
   ^] the underlying processes are still mostly unknown. In this study, we
   elucidated the pivotal role of EGC phenotypic alterations induced by
   changes in microbiota diversity during enteritis in regulating the
   integrity of the dual gut barrier and the severity of gut inflammation.
   We confirmed that DHA modulates the heterogeneity of EGCs by reshaping
   the gut microbiota homeostasis in colitis, thereby alleviating colitis
   progression. Specifically, DHA induces the transformation of
   GFAP^+/S100β^+ EGCs into GFAP^+ EGCs or prevents the transition of
   GFAP^+ EGCs into GFAP^+/S100β^+ EGCs while promoting apoptosis of
   GFAP^+/S100β^+ EGCs (Figure [213] 7 ).

Figure 7.

   Figure 7
   [214]Open in a new tab

   Graphical abstract: DHA regulates EGC phenotypic changes by promoting
   microbiota homeostasis to exert protective effects in colitis. In
   chronic colitis, microbial products or metabolites, along with
   inflammatory stimuli, can lead to concurrent damage to both the
   intestinal epithelial barrier (IEB) and the gut vascular barrier (GVB),
   accompanied by an increase in GFAP^+/S100β^+ enteric glial cells (EGCs)
   within the inflamed gut microenvironment. Treatment with
   dihydroartemisinin (DHA) effectively alleviates colitis and promotes
   the restoration of gut barrier function primarily through several
   mechanisms: DHA induces the phenotype conversion of GFAP^+/S100β^+ EGCs
   to GFAP^+ EGCs or induces apoptosis of GFAP^+/S100β^+ EGCs while
   inhibiting the conversion of GFAP^+ EGCs to GFAP^+/S100β^+ EGCs. The
   regulatory effect of DHA on EGC heterogeneity is mediated through the
   microbial products or metabolites/TLR4/NF‐κB/iNOS signaling axis,
   accompanied by the activation of c‐Fos. EGCs, enteric glial cells; DHA,
   dihydroartemisinin; LPS, lipopolysaccharide; GDNF, glial cell‐derived
   neurotrophic factor.

   Previous studies have reported the protective effects of DHA against
   IBD, primarily attributing this protective mechanism to DHA's
   regulatory effects on immune cells.^[ [215]^42 , [216]^43 ^] Yan et al.
   demonstrated that DHA protects mice from oxazolone (OXA) and
   2,4,6‐trinitrobenzene sulfonic acid (TNBS)‐induced colitis by
   triggering apoptosis of activated CD4^+ T cells, therefore regulating
   the balance of T helper (Th) and regulatory T (Treg) cells.^[ [217]^42
   ^] Another artemisinin derivative, SM934, has been shown to exert
   protective effects against colitis by inhibiting neutrophils,
   macrophages, and the NF‐κB signaling pathway.^[ [218]^43 ^] The
   anti‐inflammatory effects of DHA are also evident in its alleviation of
   LPS‐induced neuroinflammation by inhibiting the PI3K/AKT pathway.^[
   [219]^44 ^] In human and mouse models, Schneider et al. discovered a
   unique pathogenic P2X2‐dependent pathway that drives ATP‐induced
   enteric gliosis, inflammation, and disruptions in motility during
   intestinal inflammation.^[ [220]^45 ^] This prompts us to focus our
   attention on the ENS, specifically on EGCs.

   Traditionally, EGCs have been considered supportive cells for enteric
   neurons. However, as our understanding of the diverse roles EGCs play
   in regulating gastrointestinal functions expands, this conventional
   perception of EGCs as passive cells is evolving. Due to their
   distribution across the gut wall's layers, EGCs can modify their
   activities in response to changes in their microenvironment. When they
   sense stimuli from their microenvironment, they can adjust the
   expression of their surface markers.^[ [221]^25 ^] As evidenced by this
   study, there was a significant increase in GFAP^+/S100β^+ EGCs during
   colitis, suggesting that local inflammatory changes in the
   microenvironment stimulate EGCs. As observed in Figure [222]4K,M,
   inflammation triggers an increase in the synthesis and secretion of
   IL‐1β by EGCs. Additionally, inflammation enhances the synthesis of
   CXCL10 in EGCs (Figure [223]4L) but impairs its secretion
   (Figure [224]S4F, Supporting Information). Baidoo et al. found that
   aging in the human descending colon decreases the density of S100β^+
   EGCs in the myenteron, leading to colonic dysfunction.^[ [225]^46 ^] In
   addition to supporting neurons and the immune system, EGCs are reported
   to interact with epithelial cells to regulate gut permeability.^[
   [226]^25 ^] Several studies have shown that EGCs can produce GDNF and
   S‐nitrosoglutathion (GSNO) to promote the expression of barrier‐related
   proteins, thereby enhancing the integrity of the IEB.^[ [227]^47 ,
   [228]^48 ^] In our study, DHA treatment significantly enhances the
   ability of EGCs to synthesize GDNF, which may also be a crucial factor
   in the protective effect of DHA against colitis. Soret et al. suggested
   that supplementing exogenous GDNF helps alleviate dysbiosis caused by
   colitis, but this protective effect is ineffective against
   Bacteroides.^[ [229]^49 ^] In our study, DHA treatment did not increase
   EGC secretion of GDNF while alleviating the abnormally elevated
   proportion of Bacteroides in the feces of DSS‐induced colitis mice,
   indicating that the action of DHA on Bacteroides does not necessarily
   require mediation by GDNF. Undeniably, ablating EGCs in vivo leads to a
   sharp decrease in intestinal permeability.^[ [230]^25 ^] We believe
   that the effect of EGC ablation is not solely attributed to IEB damage
   but also extends to damage in the GVB. As observed in our study,
   colitis results in impairment of the dual intestinal barriers,
   concurrent with an increase in the proportion of GFAP^+/S100β^+ EGCs.
   This observation raises suspicions regarding the potential role of
   GFAP^+/S100β^+ EGCs in promoting the progression of intestinal
   inflammation.^[ [231]^50 ^]

   In Figures [232]4F and Figure [233]S4D (Supporting Information), we
   observed an increase in the proportion of GFAP^+ and S100β^+ primary
   EGCs derived from mice that did not develop colitis but received DHA
   treatment. However, compared to the control group, there was no
   difference in the proportion of primary S100β^+ EGCs derived from mice
   with colitis. Therefore, investigating the increase in the proportion
   of S100β^+ EGCs induced by DHA administration would be meaningless
   based on this premise. Thus, we focus more on the significant increase
   in GFAP^+/S100β^+ EGCs and the remarkable decrease in GFAP^+ EGCs
   observed in colitis mice. Besides, western blotting in vitro
   experiments are more indicative of phenotype changes in GFAP^+/S100β^+
   EGCs (Figure [234]4H), as the EGCs treated in vitro are not influenced
   by other interfering factors present in the complex microenvironment in
   vivo while simultaneously exhibiting dual expression of GFAP and S100β.
   This consistency is demonstrated by the trend observed in both
   Figure [235]4G,H. When investigating the specific mechanism of DHA
   action on EGCs during colitis, RNA‐Seq analysis unveiled a noteworthy
   downregulation of genes in the DHA‐treated group compared to the
   untreated inflammatory group. These downregulated genes were notably
   enriched in two major KEGG pathways: cell growth and death, and
   replication and repair (Figure [236]5B). Subsequent mice and cell
   experiments provided further confirmation of these findings
   (Figure [237]5). To distinguish between normal and pathological
   signaling, glia only express more iNOS when there is inflammation.^[
   [238]^7 ^] In the face of various stimuli within the gut
   microenvironment, EGCs often react rapidly, but they also tend to
   sacrifice themselves in the process. During clostridioides difficile
   infection, both clostridium difficile toxin B alone and in conjunction
   with pro‐inflammatory cytokines induce apoptosis in EGCs through the
   activation of three distinct signaling pathways mediated by caspases,
   calpains, and cathepsin B.^[ [239]^51 ^] TLRs, especially TLR4, as
   pattern recognition receptors (PRRs), are crucial in maintaining
   symbiosis between the gut microbiota and the host.^[ [240]^52 ^] Their
   expression in the enteric glial suggests the potential of the enteric
   glial lineage that directly responds to stimuli from the microbiota. As
   shown in Figure [241]6K, stimulation with LPS induces an increase in
   TLR4 expression in EGCs. At this point, EGCs are in an activated state,
   often referred to as reactive EGCs. “Reactive” describes the state in
   which EGCs respond to any degree of pathological or physiological
   disturbance. Reactive enteric gliosis is a highly complex and dynamic
   continuum characterized by environmental dependency and disease
   specificity, which can be considered as an attempt by EGCs to maintain
   internal balance.^[ [242]^7 ^]

   Research has shown that germ‐free mice or those treated with
   broad‐spectrum antibiotics, which lack microbiota, exhibit a
   significant reduction in EGCs in both the submucosal and myenteric
   plexuses.^[ [243]^11 ^] In mice with colitis, the more severe the
   extent of inflammation was associated with increased ratios of
   GFAP^+/S100β^+ EGCs, while concurrently, the proportion of Bacteroides
   increased in the gut microbiota, suggesting a possible correlation of
   GFAP^+/S100β^+ EGCs with Bacteroides in enteritis. Remarkably,
   treatment with DHA significantly reverses this phenomenon. S100β is
   known to play a role in regulating microbial diversity in the mouse
   intestine.^[ [244]^33 , [245]^53 ^] Our research suggests that GFAP,
   along with S100β, participates in this complex regulatory process
   simultaneously. Lei et al. investigated the effects of DHA on the gut
   microbiota of mice with acute colitis.^[ [246]^54 ^] In contrast to the
   reduced proportion of Bacteroides observed in the feces of acute
   colitis mice, we observed a significant increase in Bacteroides
   abundance in mice with chronic colitis (Figure [247]6E–G), suggesting
   that during chronic inflammation, Bacteroides may act as opportunistic
   pathogens contributing to the progression of colitis.

   Overall, our study demonstrates that the damage to IEB/GVB occurring in
   chronic colitis is closely associated with the increase of
   GFAP^+/S100β^+ EGCs in inflamed colonic tissues. By suppressing the
   phenotypic change of GFAP^+ EGCs to GFAP^+/S100β^+ EGCs, increasing the
   transition of GFAP^+/S100β^+ EGCs into GFAP^+ EGCs, and encouraging the
   apoptosis of GFAP^+/S100β^+ EGCs, DHA slows the evolution of gut
   inflammation and enhances the function of the dual gut barrier. We
   confirmed that the regulatory effect of DHA on EGC heterogeneity is
   achieved through the modulation of intestinal microbiota homeostasis.
   DHA supplementation increased the abundance of gut microbiota in
   colitis mice, leading to an elevated proportion of beneficial bacteria
   (such as Muribaculaceae) and normalization of the proportion of
   opportunistic pathogens (such as Bacteroides), thereby reducing the
   stimulating effects of harmful microbes and their metabolites on EGCs.

   To sum up, our work sheds light on the specific mechanisms underlying
   the progression of chronic colitis from the perspective of the ENS,
   aiding in the identification of innovative therapeutic targets for
   intervention in IBD and providing new insights for future strategies
   using DHA for IBD therapy. Ultimately, our findings aim to enhance the
   quality of life for IBD patients.

4. Experimental Section

Animals

   All mice utilized in this investigation shared a C57BL/6 background.
   Wild‐type male C57BL/6 mice, aged 6–8 weeks, were acquired from Beijing
   Vital River Laboratory Animal Technology Company in China. Ethical
   standards and animal welfare were upheld through approved protocols
   following the guidelines of the Animal Ethics Committee of Zhongnan
   Hospital of Wuhan University, China (Approval No. WP20230028). All
   experimental procedures adhered to the National Law on the Use of
   Experimental Animals in China.

DSS‐Induced Colitis

   Male wild‐type C57BL/6 mice were randomly assigned to control and
   colitis groups. All mice were housed in a 12‐hour light‐dark cycle
   specific pathogen‐free (SPF) environment, fed a standard diet, and
   acclimatized for one week before commencing the modeling procedure.
   There were five mice in each group. The colitis group received drinking
   water containing 2% DSS (molecular mass: 36000–50000 Da; MP
   Biomedicals, California, USA) for 7 days, followed by 7 days of
   distilled water, constituting one cycle. This cycle was repeated four
   times to establish a DSS‐induced chronic colitis model, effectively
   simulating the clinical process of remission and exacerbation in UC.
   The control group received distilled water throughout the experiment.

Drug Administration

   For the DHA treatment group, DHA (Aladdin Biochemical Technology Co.,
   Ltd., China) was administered via intraperitoneal injection, consistent
   with the methodology outlined in prior studies by Zeng et al.^[
   [248]^55 ^] The concentration of DHA was 20 mg kg^−1 (body weight),
   diluted in corn oil (HY‐Y1888, MedChemExpress, USA). DHA administration
   occurred during the interlude period of the modeling procedure,
   precisely during the 7‐day interval of distilled water intake following
   the 7‐day DSS exposure. Refer to Figure [249]3A for the detailed
   modeling process.^[ [250]^56 , [251]^57 ^] Three separate observers
   calculated the disease activity index (DAI) of mice by daily monitoring
   of body weight, stool consistency, and occult blood.^[ [252]^58 ^] Body
   weight loss was quantified on a 0 to 4 scale (0, no change; 1, <5%; 2,
   5–10%; 3, 11–20%; 4, >20%). Stool consistency was scored on a 0 to 4
   scale (0, normal; 1, soft but formed; 2, extremely soft; 3, half
   diarrhea; 4, diarrhea). Fecal bleeding severity was assessed on a 0 to
   4 scale (0‐2, none; 3, blood tracked in visible stool; 4, entirely
   rectal bleeding).^[ [253]^59 ^] Each mouse's spleen index was
   determined using the following formula: (Spleen Weight (g)/Body Weight
   (g)) × 100%.

   On the 57th day, euthanasia was performed on the mice, followed by the
   removal and measurement of the length of the entire large intestine
   (cecum, colon, and rectum). The spleen was collected and weighed, and
   the spleen index (SI) was calculated. Visceral organs such as the
   heart, liver, spleen, lungs, and kidneys were collected, fixed
   overnight in 4% paraformaldehyde, embedded in paraffin, and sectioned
   into 4µm‐thick slices for subsequent staining with hematoxylin and
   eosin (H&E), as well as for further immunohistochemical or
   immunofluorescent staining. The leftover tissues were immediately
   liquid nitrogen flash‐frozen and kept at −80 °C.

H&E Staining, Immunohistochemistry (IHC), and Immunofluorescence (IF) Assay

   The conventional protocol for staining tissue slides with H&E was
   followed. The slides underwent two consecutive 15‐min xylene washes,
   and then five minutes of ethanol washes at progressively lower
   concentrations (100%, 100%, 95%, and 70%). Lastly, the slides were
   incubated for five minutes in deionized water. Hematoxylin and eosin
   were used to stain the prepared slides. Using a double‐blind
   methodology, two gastrointestinal pathologists assessed and scored
   changes in the intestinal tissues’ histopathology based on the
   following criteria: inflammation (severity and extent), epithelial
   alterations (erosion/ulceration), submucosal edema, crypt alterations
   (crypt loss, cryptitis/crypt abscess, regeneration/hyperplasia, and
   goblet cell loss), and percentage of pathology‐affected tissue area
   compared to the total tissue area on the slide. They employed a 0–4
   scoring system: 0, inside standard deviations; 1, minimum; 2, gentle;
   3, moderate; 4, severe.

   For IHC, the dewaxed sections underwent high‐temperature (95–99°C)
   antigen retrieval in a 10 mmol L^−1 citrate buffer (pH 6.0) for 10–15
   min and were incubated with 3% hydrogen peroxide (H[2]O[2]) for 30 min
   to quench endogenous peroxidase activity. Then, block the slides with
   5% goat serum for 30 min. Following this, the slides were incubated
   overnight at 4 °C in the specified concentration of the primary
   antibody. The next day, the slides were incubated for 1 h at 37 °C with
   the horseradish peroxidase (HRP)‐conjugated secondary antibody
   (Proteintech, PR30009). Finally, the detection was performed using the
   DAB (3,3′‐Diaminobenzidine) substrate kit (ZSGB‐BIO, ZLI‐9018, Beijing,
   China). The slides were protected and mounted with neutral resins after
   nuclei were counterstained with hematoxylin (Beyotime, Cat#C0105,
   China). We used an Olympus microscope to take light pictures of the
   stained slices.

   For IF assay, the sections, having undergone antigen retrieval, were
   incubated in 5% goat serum at 37 °C for 1 h. Tissues were then
   incubated overnight at 4 °C with the designated primary antibody at an
   appropriate concentration. On the following day, tissue slices were
   incubated with the corresponding fluorescent secondary antibodies
   (1:400, Goat anti‐mouse, DyLight 549, A23310; and 1:300, Goat
   anti‐rabbit, DyLight 488, A23220; Abbkine, Wuhan, China) (1:300, Goat
   anti‐mouse, Alexa Fluor 488, 33106ES60; and 1:300, Goat anti‐rabbit,
   Alexa Fluor 647, 33113ES60, Yeasen Biotechnology, Shanghai, China) at
   37 °C for 1 h. Following incubation with 4′,6′‐diamidino‐2‐phenylindole
   (DAPI) for 5–10 minutes, mounting was performed. Intervals between each
   step were washed with phosphate buffered saline (PBS). Finally, we used
   a Leica SP8 confocal fluorescence microscope to analyze the slides.
   Images were processed with Fiji imagej software
   ([254]https://imagej.net).

RNA Extraction and RT–qPCR Analysis

   Total RNA was extracted from mouse colonic tissues using the RNA
   preparation kit (TRIpure Reagent, Aidlab, RN01, Beijing, China). The
   cDNA synthesis was performed with the ReverTra Ace qPCR RT Kit (Toyobo,
   FSQ‐101, Osaka, Japan). Quantitative real‐time polymerase chain
   reaction (qRT‐PCR) was conducted in a 10 µL reaction system using
   UltraSYBR Mixture (CWBIO, Jiangsu, China) on the Roche Light Cycler 96
   (Roche, Basel, Switzerland). 2^(−ΔΔCt) was used to compute the relative
   mRNA levels, with GAPDH as the internal reference. Table [255]S1
   (Supporting Information) lists the precise primers used in the RT‐PCR.

TEM Analysis

   We employed TEM to study changes in colon ultrastructure in each group.
   Colon tissues were cut into cubic millimeters and immediately fixed in
   2.5% glutaraldehyde (volume fraction) for 4 h at −4 °C. The samples
   were washed three times in PBS and fixed in 1% (volume fraction) of
   sodium tetroxide for 2 h. After drying in an ethanol gradient and
   embedding in epoxy resin, we used a diamond knife on an ultramicrotome
   (Leica, EM UC7, Germany) to cut ultrathin sections (60‐70 nm). The
   tissue was treated with 2% (volume fraction) uranyl acetate for 20 min.
   Finally, the materials were examined and photographed using a
   transmission electron microscope (HITACHI, HT7800, Japan).

GVB Permeability Assessment (Gut Transit Assay)

   24 hours before anesthesia with 2, 2, 2‐Tribromoethanol (Aibei
   Biotechnology, M2920, Nanjing, China), mice were subjected to a fasting
   period with free access to water. Following midline laparotomy, a 4 cm
   intestinal loop was exteriorized and ligated. Subsequently, 2 mg of
   70 kDa FITC‐Dextran (FD70, Sigma‐Aldrich) was dissolved in 2 ml of
   sterile saline solution (0.9% NaCl) and intraluminally injected through
   the ligated intestinal segment.^[ [256]^5 ^] 1 h later, blood was
   collected from the heart, and the concentration of FITC‐dextran in
   plasma was measured using a fluorescence microplate reader (BioTek
   Epoch, USA). To achieve accurate quantification, the fluorescence
   intensity was adjusted with FITC‐glucan standards, as per the
   manufacturer's instructions. Concurrently, liver and spleen tissues
   from the mice were fixed with 4% paraformaldehyde for subsequent
   preparation of frozen sections (20 µm). The fluorescence density of
   FD70 in the liver and spleen was quantified using the Leica SP8
   confocal fluorescence microscope.

Cell Culture and Treatment

   The primary umbilical vein endothelial cells (HUVEC) (PCS‐100‐013,
   ATCC) and rat enteroglial cell line EGC (CRL‐2690, ATCC) were grown in
   Dulbecco's Modified Eagle's Medium (DMEM, Hyclone), which included 10%
   heat‐inactivated fetal bovine serum (FBS, HyClone) and
   penicillin‐streptomycin (PS, HyClone). We pre‐incubated CRL‐2690 cells
   with Recombinant Rat TNF (PeproTech, 400–14, USA) for 24 h or LPS
   (Sigma‐Aldrich, USA) for 2 h to activate EGCs for subsequent
   experiments. The HT‐29 (HTB‐38, ATCC) human colonic epithelial cell
   line was grown in Roswell Park Memorial Institute (RPMI, HyClone) 1640
   medium supplemented with 10% heat‐inactivated FBS (HyClone) and PS
   (HyClone). All cells were cultivated in a monolayer at 37 °C in
   humidified air containing 5% CO2.

Flow Cytometry

   EGC taken from the mouse colon's submucosa and lamina propria were
   centrifuged at 400 ×g for 5 minutes and washed once with PBS. The cells
   were then resuspended in Fixation/Permeabilization solution (BD
   Pharmingen, 554722, USA) and incubated at 4°C for 30 min. Subsequently,
   the cells were washed once with Perm/Wash™ buffer (BD Pharmingen,
   554723, USA) and incubated in Perm/Wash™ buffer containing the
   appropriate concentration of fluorescent antibodies at 4°C in the dark
   for 1 h. The cells were washed and resuspended in the staining buffer
   (BD Pharmingen, 1029962, USA) for further flow cytometry analysis.

Isolation of EGCs from the Colonic Submucosa and Lamina Propria of the Mice

   Following the experimental procedure outlined by Wang et al.,^[
   [257]^26 ^] EGCs were extracted from mouse colonic tissues. The mouse
   was euthanized and dissected, with the whole colon separated and the
   mesentery removed. The colon was then rinsed of residual fecal contents
   in pre‐cooled Dulbecco's phosphate buffered saline (DPBS). The colon
   was segmented into 3 cm sections, and, following the method described
   by Sundaresan et al.,^[ [258]^60 ^] the longitudinal muscle layer and
   the longitudinal muscle/myenteric plexus (LMMP) were sequentially
   removed. Subsequently, the colon segments were minced and incubated in
   the pre‐cooled sterile EDTA/HEPES/DPBS solution, followed by grinding
   to eliminate the mucosa. The suspension was purified through a 100 µm
   nylon cell strainer. The tissue was incubated in the commercially
   available cell recovery solution (Corning, 354253, USA) for 30 min,
   followed by a 10‐min grinding step to separate EGC from the lamina
   propria. The resulting mixture was filtered through a 40 µm strainer,
   centrifuged at 2000 × g for 5 min, and the sediment was collected to
   collect EGC, and analysis was performed using FlowJo analysis software
   (Treestar Inc., Ashland, OR, USA). EGCs were stained with the following
   antibodies: PE mouse anti‐GFAP, BD Pharmingen, 561483; S100β Alexa
   Fluor® 488, Santa Cruz, sc‐393919AF488.

Enzyme‐Linked Immunosorbent Assay (ELISA)

   Following the manufacturer's instructions (R&D Systems, Minneapolis,
   Minnesota), the levels of IL‐1β, TNF, and IL‐10 in mouse colonic
   tissues were measured using ELISA kits. Equal quantities of tissue
   samples from each group were homogenized in the PBS supplemented with a
   protease inhibitor mixture, followed by homogenization on ice using
   ultrasonication. The samples were centrifuged at 1000 × g for 15 min,
   and the supernatant was collected to measure the absorbance at 450 nm
   using a microplate reader (BioTek Epoch, USA). The concentrations of
   the corresponding cytokines in the samples were calculated based on a
   standard curve. LPS concentrations in mice's peripheral blood were
   measured using a kit (R&D Systems, Minneapolis, Minnesota) according to
   the manufacturer's instructions. According to the manufacturer's
   instructions, levels of GDNF (Elabscience, Wuhan, China), CXCL10
   (Elabscience, Wuhan, China), and IL‐1β (R&D Systems, Minneapolis,
   Minnesota) in both intracellular and supernatant of EGCs were measured
   using ELISA kits.

Western Blot Analysis

   Total protein was extracted from the colonic mucosal layer using RIPA
   buffer containing a phosphatase and protease inhibitor mixture. Protein
   quantification was performed using a bicinchoninic acid (BCA) protein
   assay kit (Beyotime Biotech, P0009, China). Equal amounts of total
   protein were dissolved in 1× sodium dodecyl sulfate (SDS) sample buffer
   and denatured. Subsequently, proteins were separated by 7%−12%
   SDS‐polyacrylamide gel electrophoresis (PAGE) and transferred onto PVDF
   membranes (Millipore, IPVH00010, USA) on ice. After blocking with 5%
   bovine serum albumin (BSA), the membranes were incubated overnight at 4
   °C with the corresponding primary antibodies: rabbit anti‐GAPDH
   (Proteintech, 10494‐1‐AP), mouse anti‐GFAP (Cell signaling, 3670),
   rabbit anti‐S100β (Cell signaling, 90393), rabbit anti‐Occludin
   (Proteintech, 27260‐1‐AP), rabbit anti‐Claudin‐1 (Proteintech,
   28674‐1‐AP), rabbit anti‐Claudin‐2 (Proteintech, 26912‐1‐AP), rabbit
   anti‐PV1 (Cell signaling, 38238), rabbit anti‐ZO‐1 (Absin, abs122482,
   Shanghai, China), rabbit anti‐β‐Catenin (Proteintech, 51067‐2‐AP),
   rabbit anti‐Cleaved Caspase‐3 (Cell signaling, 9661), mouse anti‐Bcl‐2
   (Proteintech, 68103‐1‐Ig), rabbit anti‐Bax (Proteintech, 50599‐2‐Ig),
   mouse anti‐c‐Fos (Abcam, ab208942), rabbit anti‐iNOS (Proteintech,
   22226‐1‐AP), rabbit anti‐ NF‐κB p65 (Cell signaling, 8242), rabbit
   anti‐Phospho‐NF‐κB p65 (Cell signaling, 3033), mouse anti‐TLR4
   (Proteintech, 66350‐1‐Ig). Following three washes in 1×TBST, the
   membrane was incubated with HRP‐conjugated secondary antibodies
   (1/4000, Proteintech), and visualized using chemiluminescence by a
   detection kit (Affinity, Shanghai). Membranes were imaged using a
   GeneGnome XRQ system, and band intensity was calculated using Fiji
   software.

Human Patient Tissues with IBD

   Human colon biopsies were obtained from the Department of
   Gastroenterology, Zhongnan Hospital of Wuhan University, China. IBD was
   diagnosed based on endoscopic and histological examination of the
   biopsy tissue. All human biopsies were collected with informed
   permission, and the Ethics Committee at Zhongnan Hospital of Wuhan
   University (Approval No. 2024062K) authorized the study. Colon biopsies
   with no abnormalities from healthy individuals of the same age and
   gender, such as those having screening colonoscopies, were utilized as
   controls.

RNA‐Seq and Analysis

   Total RNA was purified for mRNA using poly‐T oligo‐attached magnetic
   beads, followed by cDNA synthesis, blunt‐end conversion, adaptor
   ligation, and PCR. Sequencing on an Illumina Novaseq platform generated
   150 bp paired‐end reads. Raw fastq data underwent fastp processing for
   high‐quality clean reads. A Hisat2 v2.0.5 index facilitated alignment,
   and StringTie (v1.3.3b) assembled mapped reads. featureCounts v1.5.0‐p3
   quantified gene reads, and FPKM (fragments per kilobase of transcript
   per million mapped reads) values were calculated. DESeq2 analyzed
   conditions with biological replicates, while edgeR assessed conditions
   without replicates, setting a threshold of corrected P‐value 0.05 and
   an absolute fold change of 2.

   Gene enrichment and functional annotation analyses of DEGs were carried
   out using either DAVID (the database for annotation, visualization and
   integrated discovery) Bioinformatics Resources or the GOtats R package,
   which was adjusted for gene‐length bias. Significantly enriched GO
   terms by DEGs were defined as those with an adjusted p‐value of less
   than 0.05. To examine the statistical enrichment of differential
   expression of genes in KEGG pathways, we utilized the R package
   ClusterProfiler. Heatmaps were created using the pheatmap package.

Analysis of scRNA‐Seq Data

   The pre‐aligned single‐cell transcriptomes of humans are sourced from
   the Gene Expression Omnibus dataset [259]GSE114374, comprising the
   following samples: Human_HC_1: [260]GSM3140593; Human_HC_2:
   [261]GSM3140594; Human_UC_1: [262]GSM3140595; and Human_UC_2:
   [263]GSM3140596. The human mesenchyme was analyzed using an in‐house
   pipeline.^[ [264]^61 ^] To put it succinctly, cells expressing 200 or
   more genes, genes with nonzero values, and genes linked to at least 5
   cells were chosen, and the library size and number of genes were
   submitted to strict quality control filters. Both datasets were
   subjected to a relaxed mitochondrial filter, which is two times the
   median absolute deviation of the fraction of mitochondrial genes as the
   median of the entire dataset. The Seurat R package^[ [265]^62 ^] was
   then used to read the filtered and normalized data for all upcoming
   analyses. Following data scaling and principal component analysis based
   on all genes, a Shared nearest neighbor (SNN) graph was created. This
   graph was then utilized in conjunction with an SNN modularity optimized
   clustering algorithm to determine cell groupings. Uniform Manifold
   Approximation and Projection (UMAP)^[ [266]^63 ^] graphs were used to
   depict the cells. To establish clusters, up to 1000 top genes (P < 0.05
   and log fold change > 0.25) were chosen from among the DEGs in each
   cluster, which were found using the Wilcox test.

Statistical Analysis

   GraphPad Prism 9.5.1 (Graph Pad Software, San Diego, CA) was used for
   all statistical analysis and charting of results. When data fell under
   a normal distribution, the student's t‐test and analysis of variance
   were used appropriately. A value of P < 0.05 was considered
   statistically significant. These error bars represent the standard
   error of the mean (SEM) as measured.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   P.Q., Y.C., and Q.Z. were associated with concept study and design.
   P.Q., L.L., X.C., H.N, C.X., and Y.H. performed experiments and data
   analysis. P.Q., Y.C., and H.W. wrote the final manuscript. P.Q. and
   S.W. performed visualization. Q.Z., M.Z., and Y.C. provided financial
   or technical support.

Supporting information

   Supporting Information
   [267]ADVS-11-2403461-s001.docx^ (13.6MB, docx)

Acknowledgements