Abstract

   Severe acute pancreatitis (SAP) is a critical condition characterized
   by an imbalance between systemic inflammatory response syndrome (SIRS)
   and compensatory anti‐inflammatory response syndrome (CARS).
   Disruptions in intestinal microbiota exacerbate this imbalance;
   however, the mechanisms by which commensal bacteria regulate
   SAP‐induced SIRS and/or CARS remain unclear. This study reports that
   the abundance of the gut commensal Akkermansia is significantly reduced
   in fecal samples from patients with acute pancreatitis (AP) and is
   inversely associated with the severity of systemic inflammatory
   responses. Using Foxp3‐DTR and IL‐10‐KO mice, it is found that
   Akkermansia muciniphila and its derived protein Amuc_1409 can suppress
   SAP‐induced pancreatic and systemic inflammation by increasing
   peripheral regulatory T cells (Tregs) and enhancing the IL‐10
   expression. Mechanistically, Amuc_1409 interacts with Ube2k to reduce
   the ubiquitination degradation of Foxp3, and it promotes the
   differentiation of Tregs and the production of IL‐10, effectively
   alleviating inflammation induced by SAP. These findings highlight A.
   muciniphila and its derivative Amuc_1409 as promising probiotic and
   biomolecule for managing acute pancreatitis and potentially other
   inflammatory conditions.

   Keywords: akkermansia muciniphila, Amuc_1409, regulatory T cells,
   severe acute pancreatitis, Ube2k
     __________________________________________________________________

   Akkermansia depletion correlates with acute pancreatitis (AP) severity.
   A. muciniphila and its protein Amuc_1409 mitigate pancreatic/systemic
   inflammation by expanding Tregs and IL‐10. Mechanistically, Amuc_1409
   binds Ube2k, inhibiting Foxp3 ubiquitination to stabilize Treg
   differentiation and IL‐10 production, countering SAP‐induced SIRS/CARS
   imbalance. These findings highlight  A. muciniphila and Amuc_1409 as
   therapeutic candidates for AP and inflammatory disorders.

   graphic file with name ADVS-12-e04214-g006.jpg

1. Introduction

   AP is one of the most common gastrointestinal diseases, and its
   incidence rate has been increasing.^[ [42]^1 ^] AP can be classified
   into mild acute pancreatitis (MAP), mild severe acute pancreatitis
   (MSAP), and SAP based on the extent of inflammation. MAP is primarily
   characterized by interstitial pancreatic edema and rarely results in
   death. However, 15–25% of patients with MAP may progress to SAP, which
   is marked by severe systemic inflammatory responses and persistent
   multiple organ failure (PMOF). Due to the complex etiology, the lack of
   specific clinical drugs, and the absence of effective intervention, SAP
   has a poor prognosis, with a mortality rate as high as 36–50%.^[ [43]^2
   ^]

   Cumulative evidence strongly supports that premature activation of
   trypsinogen is a critical pathological mechanism leading to acinar cell
   necrosis.^[ [44]^3 ^] Damaged acinar cells release chemokines and
   cytokines, which recruit and mediate the infiltration of monocytes and
   neutrophils to the injury site.^[ [45]^4 ^] Subsequently, monocytes
   were activated by damage‐associated molecular patterns (DAMPs) released
   from necrotic acinar cells, leading to an amplified production of
   pro‐inflammatory cytokines. Macrophages in distant organs are also
   activated, exacerbating systemic inflammation and causing distant organ
   injury, which contributes to the development of SIRS and PMOF.^[ [46]^5
   ^] This phase of hyperinflammation is often followed by a CARS mediated
   by an increase in Tregs, aimed at resolving the deadly cytokine storm,
   allowing most patients to restore a balance between hyperinflammatory
   response and immunosuppression.^[ [47]^6 ^] However, if this
   homeostasis is disrupted, the patient may succumb to either an early,
   overwhelming inflammatory response or to late‐stage
   immunosuppression.^[ [48]^7 ^] Currently, there remains no effective
   strategy to control the excessive inflammatory response or to reduce
   mortality in SAP.

   SAP is often accompanied by increased intestinal permeability and
   mucosal barrier damage, resulting in bacterial translocation and even
   abdominal infection.^[ [49]^8 ^] Our previous study indicated that
   perturbation of the intestinal microbiota was a critical factor to
   exacerbate pancreatic and systemic inflammatory responses.^[ [50]^9 ^]
   Furthermore, there is growing evidence that metabolites and proteins of
   commensal bacteria attenuate SAP‐induce SIRS by suppressing the
   inflammatory mediated signaling of macrophages and neutrophils.^[
   [51]^9 , [52]^10 ^] Unfortunately, our understanding of how commensal
   bacteria regulate SAP‐induced CARS is limited. Additionally, the
   upstream mechanisms and downstream targets through which commensal
   bacteria influence the function of Tregs in AP remain poorly
   understood.

   Here, we screened the intestinal microbiome of patients with AP and
   identified that the abundance of Akkermansia genus was significantly
   reduced in patients with AP than in healthy control. Subsequently, we
   discovered that Akkermansia muciniphila and its derivative Amuc_1409
   could interact with Ube2k to induce Foxp3 expression in Tregs, which
   produces anti‐inflammatory cytokine, IL‐10, to ameliorate SAP severity
   by suppressing pancreatic and systemic inflammatory responses.

2. Results

2.1. Patients with AP Exhibit a Decreased Fecal Abundance of Akkermansia

   Previous studies have identified patients with AP exhibit intestinal
   flora dysregulation, but the role of intestinal flora in the
   progression of AP is unknown. Here, we collected fecal samples from 86
   hospitalized patients (38 with MAP, 16 with MSAP, and 32 with SAP) and
   46 healthy control volunteers (HC). There was no significant difference
   in age, sex, lifestyle, etiology or antibiotics use of AP between AP
   groups (Table [53] 1 ) as determined by logistic regression. Patients
   with MAP (31.6%) and MSAP (37.5%) had a higher recurrence rate,
   compared with those with SAP (6.3%) (Table [54]1). Then, we analyzed
   the difference of their fecal microbiota composition by 16S rRNA
   sequencing. Principal coordinate analysis (PCoA) of all samples
   revealed a different microbial signature of patients with AP compared
   with HC, with the most prominent changes showed between patients with
   SAP and HC (Figure [55] 1A). There were no significant differences in
   reads between the groups, but a remarked reduced richness and Chao1 was
   observed in patients with MAP and SAP compared with HC
   (Figure [56]1B,C). The heatmap exhibited a reduced abundance of
   Akkermansia, Lachnoclostridium, and Ruminococcus, and an increasing
   abundance of Enterococcus, Pseudomonas, and Escherichia‐Shigella in the
   patients with AP compared with the HC (Figure [57]1D). To validate the
   critical role of gut microbiota in mediating the progression of AP, we
   performed fecal microbiota transplantation (FMT) experiment.
   Specifically, fecal samples from SAP patients and healthy individuals
   were orally administered to antibiotic‐treated (Abx) mice. After 48 h
   of microbial colonization, a murine SAP model was established using
   caerulein. Compared to mice receiving FMT from healthy donors, the SAP
   patient FMT group exhibited significantly elevated serum amylase
   levels, increased expression of inflammatory factors (Il6, Il1b) in
   both pancreatic tissue and serum, and markedly aggravated pancreatic
   pathological damage, including acinar cell edema, necrosis, and
   inflammatory infiltration (Figure [58]1E—H; Figure [59]S1A, Supporting
   Information). These findings suggest that gut microbiota alterations in
   AP patients are a key factor mediating disease severity.

Table 1.

   Clinical characteristics of patient with acute pancreatitis.
   Property Level of pancreatitis p‐value[60] ^* ^)
   Mild (n = 38) Moderately severe (n = 16) Severe (n = 32)
   Age, y, mean (range) 47±14.9 (20‐78) 50±12.6 (25‐82) 50±16.9 (20‐88)
   0.65
   Sex, f/m 9/29 4/12 11/21 0.6023
   Etiology, n (%)
   Gallstone 18 (47.4) 5 (31.3) 21 (65.6) 0.0734
   Alcohol 1 (2.6) 1 (6.3) 1 (3.1) 0.7802
   hyperlipemia 17 (44.7) 9 (56.3) 11 (34.4) 0.3602
   Others[61] ^# ^) 2 (5.3) 1 (6.3) 0 (0) 0.4059
   Lifestyle, n (%)
   Smoking 12 (31.6) 3 (18.8) 3 (9.4) 0.0760
   Drinking 14 (36.8) 5 (31.3) 7 (21.9) 0.3557
   IV antibiotics on admission 2 (5.3) 1 (6.3) 1 (3.1) 0.9999
   Previous pancreatitis 12 (31.6) 6 (37.5) 2 (6.3) 0.0097
   [62]Open in a new tab
   ^*)

   p‐value calculated using Fisher's exact probability test between mild,
   moderately severe, and severe groups.
   ^^#)

   Other = Idiopathic, traumatic pancreatitis.

Figure 1.

   Figure 1
   [63]Open in a new tab

   Patients with AP exhibit a decreased fecal abundance of Akkermansia
   that was inversely correlated with the inflammation of AP. The 16S rRNA
   sequencing analysis of fecal samples from patients with AP or Healthy
   control. A) PCoA, B) total reads, C) Alpha diversity indicators of
   Richness (left) and Chao (right) among the four groups (HC n = 46, MAP
   n = 38, MSAP n = 16, SAP n = 32). D) The heatmap of relative abundance
   of bacteria at the genus level. E) The concentration of serum amylase
   (n = 5), F) serum IL‐6 level (n = 7), G) pathology scores, and H)
   pancreatic histopathology from Abx‐treated mice received FMT from HC or
   patients with SAP donors at 12 h post caerulein‐induced SAP modeling.
   I) Venn diagram analysis between the gut bacteria reduced in AP mouse
   models and those in AP patients. J) Genomic copies of A. muciniphila (n
   = 15). K) The correlations between the concentrations of IL‐6 (left)
   and TNF‐α (right) and the relative abundance of Akkermansia. The
   two‐sided p‐values were examined by one‐way ANOVA with Dunnett's
   multiple comparisons test (C) or Student's t‐test (E–G, J) and data
   were presented as mean ± sem. R^2 and exact two‐sided p‐values
   calculated by Pearson's test are shown (K).

   Subsequently, we conducted a joint analysis of 16S rRNA data from
   previous AP mouse models.^[ [64]^9 ^] By performing Venn diagram
   analysis between the gut bacteria reduced in AP mouse models and those
   in AP patients, we identified three co‐altered bacterial species:
   Akkermansia, Bifidobacterium, and Erysipelatoclostridium
   (Figure [65]1I). Notably, the genomic copies and relative abundance of
   Akkermansia muciniphila (A. muciniphila) was significantly decreased in
   the SAP group compared with HC (Figure [66]1J; Figure [67]S1B,
   Supporting Information). Then, we determined the expression of
   inflammatory cytokines, IL‐6 and TNF‐α in the serum of patients with
   AP, and correlation analysis further revealed that the relative
   abundance of A. muciniphila in fecal samples showed negative
   correlations with pro‐inflammatory cytokines (IL‐6 and TNF‐α)
   (Figure [68]1K). Furthermore, the serum concentrations of C‐reactive
   protein (CRP) and amylase inversely correlated with the abundance of
   Akkermansia in patients with AP (Figure [69]S1C,D, Supporting
   Information). Taken together, these data suggested that Akkermansia was
   inversely correlated with the inflammation of AP.

2.2. Akkermansia Muciniphila Protects Against SAP by Attenuating Pancreatic
and Systemic Inflammation

   Our previous work also has identified Akkermansia could be a potential
   probiotic that alleviates AP in mice.^[ [70]^9 ^] To test the potential
   probiotic role of Akkermansia during the progression of AP, we used
   highly concentrated broad‐spectrum antibiotic (Abx) by oral gavage to
   deplete commensal bacteria and constructed AP mouse model by caerulein
   injection (Figure [71]S2A, Supporting Information). As previously
   reported,^[ [72]^9 ^] Abx‐treated mice displayed a distinct microbial
   profile, as revealed by PCoA, with significantly diminished α‐diversity
   indices (Richness and Chao1) when compared to PBS‐treated controls
   (Figure [73]S2E,F, Supporting Information). Notably, Abx‐treated mice
   demonstrated substantial alterations in microbial composition,
   characterized by decreased relative abundances of beneficial taxa
   including Akkermansia, Lactobacillus, and Ruminococcaceae, concomitant
   with an enrichment of potentially pathogenic genera such as
   Pseudomonas, Sphingomonas, and Escherichia‐Shigella (Figure [74]S2G,H,
   Supporting Information). Furthermore, Abx‐treated mice exhibited a
   significantly higher content of serum amylase and more severe
   pancreatic inflammatory lesions than PBS‐treated controls in SAP
   models. FMT from PBS‐treated mice into Abx‐treated mice reversed the
   amylase content and improved pancreatic pathological lesions
   (Figure [75] 2A–C), suggesting that intestinal microbiota alteration
   influenced the disease progression. Then, we colonized A. muciniphila
   or Enterococcus faecalis (E.F) to Abx‐treated mice and found that A.
   muciniphila colonization, not E.F colonization, markedly reduced serum
   amylase concentration and alleviated pancreatic edema and inflammatory
   cells infiltration (Figure [76]2A–C). Due to the complex pathogenic
   factors of AP, multiple animal model validations are required. We then
   used L‐arginine to construct another SAP mouse model, as described
   previously^[ [77]^11 ^] (Figure [78]S2B, Supporting Information). A.
   muciniphila colonization significantly reduced serum amylase level and
   improved pancreatic pathological lesion (Figure [79]2D,E). The two
   different mouse models suggested the preventive role of A. muciniphila
   on SAP.

Figure 2.

   Figure 2
   [80]Open in a new tab

   Akkermansia muciniphila protects the host from MAP and SAP by
   attenuating pancreatic and systemic inflammation. A) The concentration
   of serum amylase, B) pancreatic histopathology, and C) pathology scores
   from mice treated with PBS, Abx, or Abx‐treated mice received FMT or
   colonized with Akkermansia muciniphila (AKK) or E.F at 12 h post
   caerulein‐induced SAP modeling (n = 6–10). D) The serum amylase, E)
   pancreatic histopathology from mice treated with PBS, Abx, or
   Abx‐treated mice colonized with AKK at 3 days post L‐arginine‐induced
   SAP modeling (n = 6). F) Serum amylase level (n = 5), G) pancreatic Il6
   mRNA level (n = 4) and serum IL‐6 level (n = 5), H) pancreatic
   histopathology, and I) pathology scores from mice treated with Abx or
   Abx‐mice treated with AKK, pasteurized A. muciniphila (p‐AKK),
   supernatant A. muciniphila (s‐AKK) at 12 h post caerulein‐induced SAP
   modeling. J) Serum amylase level (n = 9), K) serum IL‐6 level (n = 10),
   L) pancreatic histopathology, and pathology scores from
   L‐arginine‐induced SAP mice treated with AKK, p‐AKK, s‐AKK at 3 days
   post modeling. The two‐sided p‐values were examined by one‐way ANOVA
   with Dunnett's multiple comparisons test and data were presented as
   mean ± sem (A, C,D, F,G, I–L).

   To determine the protective component of A. muciniphila, we filtered
   the supernatant of A. muciniphila (s‐AKK, comprised by metabolites) and
   pasteurized component of A. muciniphila (p‐AKK, comprised by proteins),
   and gavaged to Abx‐treated mice (Figure [81]S2A, Supporting
   Information). Intriguingly, the treatment of p‐AKK significantly
   alleviated SAP‐induced serum amylase and downregulated IL‐6 expression
   in the local pancreas and systemic circulation (Figure [82]2F,G).
   Notably, p‐AKK administration significantly attenuated pancreatic edema
   and inflammatory cell infiltration. In contrast, while oral s‐AKK
   treatment demonstrated a modest reduction in SAP‐induced inflammatory
   response and pancreatic tissue damage, these protective effects failed
   to achieve statistical significance (Figure [83]2H,I). This
   differential efficacy strongly suggests that the observed benefits were
   primarily mediated by microbial proteins rather than bacterial
   metabolites. To further explore the therapeutic effect of A.
   muciniphila, we constructed SAP mouse model using L‐arginine, and
   gavaged with A. muciniphila, p‐AKK or s‐AKK daily until the end of the
   experiments (Figure [84]S2B, Supporting Information). As previously
   observed, A. muciniphila and p‐AKK exhibited a significant reduction of
   serum amylase and IL‐6 expression (Figure [85]2J,K; Figure [86]S2I,
   Supporting Information), and an observable improvement of pancreatic
   edema, necrosis, and inflammatory cell infiltration (Figure [87]2L).
   Together, these concordant results in two different SAP mouse models
   indicate a protective and therapeutic role of A. muciniphila‐driven
   proteins in alleviating SAP‐induced inflammatory responses and
   pathological lesion in the local pancreas and systemic circulation.

2.3. Akkermansia Muciniphila Upregulates Peripheral Regulatory T Cells and
Promotes Intestinal Barrier Integrity

   To explore the regulation of A. muciniphila on immune cells, we
   performed single‐cell RNA sequencing of peripheral blood mononuclear
   cells (PBMCs) from L‐arginine‐induced SAP mice, A. muciniphila‐treated
   SAP mice and non‐treated control to identify the key cells regulated by
   A. muciniphila. After strict quality control and filtering
   (Figure [88]S3A, Supporting Information), we obtained 26 176 single
   cells from these samples. Five major cell clusters were identified,
   including B cells (Ms4a1 ^+), CD4^+ T cells (Cd4^+ ), CD8^+ T cells
   (Cd8a ^+), myeloid cells (Cd14 ^+), and natural killer (NK) cells (Nkg7
   ^+) (Figure [89]S3B, Supporting Information). Then, we analyzed the
   immune cell composition in each of the individuals and noted that the
   percentage of myeloid cells were significantly increased in SAP mice
   treated with or without A. muciniphila, compared with negative control
   (NC) mice (Figure [90]S3C, Supporting Information). We subdivided
   myeloid cells into mononuclear macrophage (Mono‐mac), neutrophil, mast
   cells and proliferative cells (Figure [91]S3D,E, Supporting
   Information), and noted that the percentage of mononuclear macrophages
   was decreased in the A. muciniphila‐treated SAP mice, compared with SAP
   mice (Figure [92]S3F, Supporting Information). Previous studies have
   demonstrated that activated macrophages and neutrophils play a crucial
   role as key immune cells in initiating and amplifying the inflammatory
   response induced by SAP.^[ [93]^5 , [94]^9 , [95]^10 ^] To determine
   whether A. muciniphila alleviates the inflammatory response by
   inhibiting macrophages and neutrophils, we examined the changes of
   peripheral macrophages and neutrophils using flow cytometry.
   Expectedly, SAP significantly upregulated frequency of peripheral
   macrophages and neutrophils, but A. muciniphila group did not exhibit
   lower frequency of peripheral macrophages and neutrophils, compared
   with SAP group (Figure [96]S4A,B, Supporting Information), suggesting
   that A. muciniphila may alleviate acute pancreatitis through the
   involvement of other immune cells.

   It's worth noting that the percentage of CD4^+ T cells and CD8^+ T
   cells was decreased in A. muciniphila‐treated SAP mice, compared with
   SAP mice (Figure [97]S3C, Supporting Information). Subsequently, we
   subdivided T cells into CD4 naïve cells, CD8 naïve cells, CD4 exhausted
   cells, CD4 Tem cells, CD8 T effector cells, and Tregs (Figure [98] 3A;
   Figure [99]S3G, Supporting Information). Then, we analyzed the changes
   of T cell subtypes in each of groups and noted that the frequency of
   CD4 exhausted cells and CD8 T effector cells was enriched and the
   percentage of CD4 Tem cells were decreased in SAP mice, compared with
   NC mice. Interestingly, we found that the percentage of Tregs was
   decreased in the SAP mice and more enriched in the A.
   muciniphila‐treated SAP mice, compared with NC mice (Figure [100]3B,C),
   and confirmed this phenomenon by flow cytometry (Figure [101]3D),
   implying A. muciniphila may alleviate SAP‐induced inflammation via
   upregulating the percentage of peripheral Tregs. We further determined
   the immunomodulatory effects of A. muciniphila in another
   caerulein‐induced SAP model, and detected the changes of macrophages,
   neutrophils, and Tregs in PBMC and splenic cells by flow cytometry. As
   expected, the occurrence of SAP led to a substantial increase in the
   percentage of neutrophils and macrophages in PBMC (Figure [102]S4C,D,
   Supporting Information) and splenic cells (Figure [103]S5A,B,
   Supporting Information), and neither A. muciniphila nor p‐AKK reversed
   these changes. Notably, A. muciniphila and p‐AKK could restore the
   decreased Treg levels in PBMCs and splenic cells of SAP mice to normal
   levels (Figure [104]3E,F; Figure [105]S5C, Supporting Information).

Figure 3.

   Figure 3
   [106]Open in a new tab

   Akkermansia muciniphila upregulated peripheral regulatory T cells and
   promotes intestinal barrier integrity. L‐arginine‐induced SAP mice
   treated with AKK daily for 3 days. PBMCs were isolated at 3 days post
   modeling and single‐cell transcriptomics sequencing was performed. A)
   UMAP plots of T cells, including six major clusters. B) UMAP plots of T
   cells, C) Ro/e of T cells, D) the percentage of
   CD4^+/CD25^+/Foxp3^+‐Tregs in PBMC from NC mice, SAP mice and
   AKK‐treated SAP mice (n = 3). Mice treated with Abx or Abx‐treated mice
   colonized with AKK or p‐AKK at 12 h post caerulein‐induced SAP
   modeling. E) Representative flow cytometry plots, F) the percentage of
   CD4^+/CD25^+/Foxp3^+‐Tregs (n = 3). L‐arginine‐induced SAP mice treated
   with AKK, p‐AKK or s‐AKK daily for 3 days. G) Colony forming units
   (CFU) were counted from pancreas homogenates and exceeded the cut‐off
   level of 1 CFU/mg tissue weight (n = 2–4). H) Representative indirect
   fluorescent assay (IFA) images of colon. Zo‐1, Occludin, and Muc2 as
   makers for tight junction proteins. I) Representative Western blot
   images of Occludin, Muc2 and Actin in colon. J) Quantification of the
   proteins in Figure [107]3I. The protein levels of Muc2 (left) and Ocln
   (right) were normalized to those of actin. The amount of each protein
   in the NC group was defined as 1 (n = 3). The two‐sided p‐values were
   examined by one‐way ANOVA with Dunnett's multiple comparisons test and
   data were presented as mean ± sem (D, F, J).

   In addition to hyper‐inflammatory responses, SAP often accompanies
   intestinal dysfunction, which in severe cases can lead to intestinal
   failure and bacterial translocation.^[ [108]^8 ^] Hence, we determined
   the bacterial infection of pancreas by colony growth assays of the
   tissue homogenate on agar plates and detected a marked bacterial load
   in the pancreas of SAP mice, exceeding 1 colony‐forming unit (CFU) per
   milligram of pancreas. A. muciniphila and p‐AKK, not s‐AKK, were
   reduced bacterial loads in the pancreas (Figure [109]3G), suggesting
   that A. muciniphila and p‐AKK effectively reduce bacterial
   translocation. To further evaluate the effect of A. muciniphila on the
   intestinal barrier, we tested pathogenic lesion by H&E staining, and
   analyzed the expression of structural protein components of the
   intestinal barrier like tight junctions by immunofluorescence assay
   (IFA) and Western blot. The pathogenic lesion of duodenum and colon was
   observed in SAP mice and s‐AKK‐treated SAP mice, while it was
   alleviated in A. muciniphila‐ and p‐AKK‐treated SAP mice
   (Figure [110]S5D, Supporting Information). Furthermore, we observed a
   significant reduction of mRNAs encoding for mucins 2 (Muc2), Occluding
   (Ocln) in the colon of SAP mice, whereas those mRNA levels were
   markedly upregulated after treating with A. muciniphila and p‐AKK
   (Figure [111]S5E, Supporting Information). The protein expression of
   those tight junction proteins, Muc2 and Ocln, not Zo‐1, were also
   downregulated in the duodenum and colon of SAP mice and s‐AKK‐treated
   SAP mice, and restored in A. muciniphila‐ and p‐AKK‐treated SAP mice
   (Figure [112]3H—J; Figure [113]S5F, Supporting Information), implying
   that A. muciniphila and its protein components promote intestinal
   barrier integrity. Together, those results indicated a role for A.
   muciniphila‐driven proteins in suppressing inflammatory responses via
   upregulating Tregs and preventing bacterial translocation via promoting
   intestinal barrier integrity.

2.4. Akkermansia Muciniphila‐Driven Protein Amuc_1409 Protects the Host from
SAP by Dampening Pancreatic and Systemic Inflammation

   Eun‐Jung Kang et al. performed comprehensive proteomic profiling of A.
   muciniphila secretory proteins using nanoflow liquid
   chromatography‐tandem mass spectrometry (nLC‐MS/MS). Their quantitative
   analysis revealed Amuc_1409 as the most abundant secreted protein,
   based on both MS/MS spectral counts and intensity‐based absolute
   quantification (iBAQ) values.^[ [114]^12 ^] Notably, Amuc_1100— a
   previously characterized multifunctional outer membrane protein of A.
   muciniphila—has been demonstrated to attenuate AP by reducing
   macrophage and neutrophil infiltration.^[ [115]^10 , [116]^13 ^] To
   test whether A. muciniphila has a protective effect on SAP mediated by
   Amuc_1409, we treated Abx mice with purified His‐tagged Amuc_1409 and
   Amuc_1100 produced in Escherichia coli (E. coli) and constructed SAP
   models using caerulein, in which Amuc_1100 treatment as positive
   control (Figure [117]S2A, Supporting Information). Treated with
   Amuc_1409 or Amuc_1100 significantly reduced serum amylase
   concentration (Figure [118] 4A) and downregulated IL‐6 expression in
   the local pancreas and systemic circulation (Figure [119]4B,C).
   Moreover, Amuc_1409 or Amuc_1100 treatment remarkedly alleviated
   pancreatic edema and inflammatory cell infiltration (Figure [120]4D,E).
   Amuc_1409 also demonstrated significant protective effects in both
   caerulein‐induced MAP and L‐arginine‐induced SAP models
   (Figure [121]S6A–G, Supporting Information). Treatment with Amuc_1409
   or Amuc_1100 promoted peripheral Tregs proportion in caerulein‐induced
   (Figure [122]4F,G) and L‐arginine‐induced SAP model (Figure [123]S6H,I,
   Supporting Information), and Amuc_1409 exhibited a stronger promotion
   on Tregs than Amuc_1100.

Figure 4.

   Figure 4
   [124]Open in a new tab

   A. muciniphila‐driven protein Amuc_1409 protects the host from SAP by
   dampening pancreatic and systemic inflammation. Mice treated with Abx
   or Abx‐treated mice gavaged with Amuc_1100 or Amuc_1409 at 12 h post
   caerulein‐induced SAP modeling. A) Serum amylase level (n = 3–5). B)
   Pancreatic Il6, Il1b mRNA level (n = 5) and C) serum IL‐6 level (n =
   5). D) Pathology scores (n = 3) and E) pancreatic histopathology. F)
   Representative flow cytometry plots, G) the percentage of
   CD4^+/CD25^+/Foxp3^+‐Tregs in PBMC (n = 3). H) The percentage of
   CD4^+/CD25^+/Foxp3^+‐Tregs in MLN in L‐arginine‐induced SAP model (n =
   3). I) Representative Western blot images of Occludin, Muc2, and Actin
   in colon. J) Quantification of the proteins in Figure [125]4I. The
   protein levels of Muc2 (left) and Ocln (right) were normalized to those
   of Actin. The amount of each protein in the NC group was defined as 1
   (n = 3). K) Representative IFA images of colon. Occludin, Muc2 as
   makers for tight junction proteins, His‐tagged for Amuc_1100 and
   Amuc_1409. The two‐sided p‐values were examined by one‐way ANOVA with
   Dunnett's multiple comparisons test and data were presented as mean ±
   sem (A–D, G,H, J).

   Mesenteric lymph nodes (mLNs) are highly efficient sites for the
   peripherally induced Tregs.^[ [126]^14 ^] Furthermore, commensal
   bacteria‐derived antigens can rapidly and efficiently activate T cell
   in mLNs, promoting the generation of peripheral Tregs.^[ [127]^15 ^]
   Based on this, we hypothesized that A. muciniphila and its derived
   protein Amuc_1409 promote Tregs differentiation in mLNs and increase
   the proportion of peripheral Tregs. To test this hypothesis, we
   measured the percentage of CD4^+CD25^+Foxp3^+‐Tregs in mLNs and found
   that treatment with A. muciniphila, Amuc_1409 or Amuc_1100 increased
   the proportion of Tregs in mLNs. Notably, A. muciniphila and Amuc_1409
   induced a higher proportion of Tregs compared to Amuc_1100
   (Figure [128]4H; Figure [129]S6J, Supporting Information). To assess
   Treg infiltration in pancreatic tissue, we performed indirect
   immunofluorescence staining to analyze CD3 and Foxp3 expression in
   PBS‐treated and Amuc_1409‐treated SAP mice. Quantitative analysis
   revealed a marked increase in CD3⁺Foxp3⁺ double‐positive cells and
   distinct localization of Amuc_1409 in Amuc_1409‐treated mice compared
   to SAP controls (Figure [130]S7A,B, Supporting Information),
   demonstrating that Amuc_1409 significantly promotes Treg recruitment to
   pancreatic tissue.

   Then, we examined the effects of Amuc_1409 or Amuc_1100 on intestinal
   barrier and found that the gene expression of markers for tight
   junction, Muc2, Ocln, and Zo‐1, was significantly upregulated in colon
   of Amuc_1409 treated SAP mice (Figure [131]S7C, Supporting
   Information). Consistent with the gene expression changes, the protein
   expression of Ocln and Muc2 was also increased in colon
   (Figure [132]4I—K; Figure [133]S7D, Supporting Information) and
   duodenum (Figure [134]S7E, Supporting Information) of Amuc_1409 treated
   SAP mice. Further, His‐tagged proteins were observably expressed in
   colon and duodenum of Amuc_1100 or Amuc_1409 treated mice. Previous
   studies have demonstrated that the Th17/Treg imbalance in SAP
   correlates with disease severity and poor prognosis,^[ [135]^16 ^] and
   Th17/Treg ratio is important for the intestinal barrier. To investigate
   this relationship, we isolated colon tissues from SAP model mice and
   analyzed single‐cell suspensions for Th17 and Treg proportions. Results
   demonstrated that compared to NC mice, SAP mice exhibited a significant
   reduction in colonic Treg populations and elevated Th17 cells
   populations. Amuc_1409 treatment normalized this imbalance, increasing
   Treg proportions, and reducing Th17 cells populations. The Th17/Treg
   ratio in Amuc_1409‐treated mice approached that of NC, contrasting
   sharply with untreated SAP mice (Figure [136]S7F,G, Supporting
   Information)). These results suggest that Amuc_1409 maybe ameliorate
   intestinal barrier dysfunction by resolving Th17/Treg imbalance.

   Collectively, these results suggest that Amuc_1409 protects the host
   from SAP by dampening pancreatic and systemic inflammation and
   promoting intestinal barrier integrity.

2.5. Akkermansia Muciniphila and Amuc_1409 Protect the Host from SAP in a
Treg‐and IL‐10‐Dependent Manner

   To verify whether A. muciniphila and Amuc_1409 protect the host from
   SAP through Tregs, we depleted Tregs in Foxp3‐DTR mice by
   intraperitoneal injection of diphtheria toxin (DT), and confirmed a
   significant reduction of CD4^+CD25^+Foxp3^+‐Tregs in DT‐treated mice
   (Figure [137] 5A). Then, we gavaged A. muciniphila or Amuc_1409 to
   Tregs‐depletion mice and induced SAP using caerulein, and found that A.
   muciniphila or Amuc_1409 significantly reduced serum amylase content,
   pancreatic inflammatory cytokine Il6, Il1b mRNA level and serum IL‐6
   expression, and upregulated pancreatic anti‐inflammatory cytokine Il10
   mRNA level in PBS‐treated Foxp3‐DTR mice, whereas A. muciniphila or
   Amuc_1409 did not alleviate serum amylase and pancreatic and systemic
   inflammatory responses in DT‐treated Foxp3‐DTR mice (Figure [138]5B–D).
   Moreover, A. muciniphila or Amuc_1409 treatment remarkedly alleviated
   pancreatic edema, necrosis, and inflammatory cells infiltration in
   PBS‐treated Foxp3‐DTR mice, but A. muciniphila or Amuc_1409 did not
   alleviate SAP induced pancreatic pathological lesion in DT‐treated
   Foxp3‐DTR mice (Figure [139]5E,F). Notably, Tregs depletion
   significantly aggravated pancreatic necrosis and inflammatory cells
   infiltration (Figure [140]5E,F). We further analyzed the pancreatic
   IL‐6 and IL‐10 expression through IFA assay, and found that A.
   muciniphila or Amuc_1409 significantly downregulated pancreatic IL‐6
   expression and upregulated IL‐10 expression in PBS‐treated Foxp3‐DTR
   mice, whereas A. muciniphila or Amuc_1409 did not regulate the
   expression of IL‐6 and IL‐10 in DT‐treated Foxp3‐DTR mice
   (Figure [141]5E,G,H). Furthermore, we found that in PBS‐treated
   Foxp3‐DTR mice, A. muciniphila or Amuc_1409 administration
   significantly ameliorated colonic injury, with preserved epithelial
   integrity. However, in DT‐treated Foxp3‐DTR mice, the therapeutic
   effects of A. muciniphila or Amuc1409 were abolished (Figure [142]S8A,
   Supporting Information), indicating a Treg‐dependent mechanism
   underlying their protective action on intestinal repair.

Figure 5.

   Figure 5
   [143]Open in a new tab

   A. muciniphila and Amuc_1409 protect the host from SAP in a
   Treg‐dependent manner. Foxp3‐DTR mice were intraperitoneally injected
   with DT daily for 5 days, and splenic CD4^+/CD25^+/Foxp3^+‐Tregs were
   detected by flow cytometry. A) Representative flow cytometry plots
   (left), and the percentage of CD4^+/CD25^+/Foxp3^+‐Tregs (right) (n =
   3). AKK or Amuc_1409 were gavaged to Tregs‐depletion mice and induced
   SAP using caerulein. B) Serum amylase level (n = 5), C) pancreatic Il6,
   Il1b, Il10 mRNA level (n = 5), and D) serum IL‐6 level (n = 5), E)
   representative pancreatic histopathology (upper), IFA images of
   pancreatic IL‐6 (middle) and IL‐10 (down), F) pathology scores (n = 3),
   G,H) mean fluorescence intensity of (G) IL‐6 and (H) IL‐10. The
   two‐sided p‐values were examined by one‐way ANOVA with Dunnett's
   multiple comparisons test and data were presented as mean ± sem (A–D,
   F–H).

   As a pivotal yet complex anti‐inflammatory cytokine, IL‐10 exhibits
   paradoxical effects in SAP patients–while some studies report
   significant elevation of IL‐10 levels,^[ [144]^17 ^] others demonstrate
   marked reduction.^[ [145]^18 ^] We hypothesize that this phenomenon may
   reflect dynamic equilibrium between pro‐inflammatory and
   anti‐inflammatory mediators during disease progression. To further
   elucidate IL‐10′s role in pancreatitis pathogenesis, we established SAP
   models using IL‐10 knockout (KO) mice. The IL‐10‐KO mice manifested
   exacerbated inflammatory responses and heightened pancreatic tissue
   damage compared to wild‐type controls (Figure [146]S8B—E, Supporting
   Information), suggesting that IL‐10 deficiency may serve as a critical
   factor contributing to aggravated SAP pathology. To verify whether
   Amuc_1409 protect the host from SAP through IL‐10, we gavaged Amuc_1409
   to IL‐10‐KO mice and induced SAP using caerulein, and found that
   Amuc_1409 significantly reduced serum amylase content, pancreatic
   inflammatory cytokine Il6, Il1b mRNA level and serum IL‐6 expression,
   and alleviated pancreatic edema, necrosis and inflammatory cells
   infiltration in WT mice, whereas Amuc_1409 did not alleviate serum
   amylase, pancreatic and systemic inflammatory responses and pancreatic
   damage in IL‐10‐KO mice (Figure [147]S8B–E, Supporting Information).

   Overall, these results suggest that the ameliorating effect of A.
   muciniphila and Amuc_1409 on SAP severity is mainly due to the increase
   of peripheral Tregs and IL‐10. Depletion of Tregs and IL‐10 discourage
   the protection of A. muciniphila or Amuc_1409 on SAP.

2.6. Amuc_1409 Promoted Tregs Differentiation and IL‐10 Production through
Interaction with Ube2k

   To investigate whether Amuc_1409 modulates macrophage‐mediated
   inflammatory responses, we conducted in vitro experiments using bone
   marrow‐derived macrophages (BMDMs) from C57BL/6 mice. Notably,
   Amuc_1409 treatment failed to attenuate LPS‐induced inflammatory
   activation in BMDMs within our experimental system (Figure [148]S9A,
   Supporting Information). To examine the effect of Amuc_1409 on Tregs
   differentiation in vitro, we isolated mouse naïve CD4^+ T cells, and
   differentiated them to Tregs via treating with or without Amuc_1409
   under IL‐2 and TGF‐β condition. There was a significantly increase in
   the proportion of CD4^+/Foxp3^+‐Tregs under Amuc_1409 treatment, and in
   a dose‐dependent manner (Figure [149] 6A). We next assessed whether
   dendritic cells (DCs) were required for Amuc_1409‐mediated potentiation
   of Tregs differentiation. Murine bone marrow‐derived DCs (BMDCs) were
   induced by M‐CSF and IL‐4 in vitro, stimulated with Amuc_1409 or LPS
   (Positive control), and cocultured with Naïve CD4^+ T cells to assess
   Tregs differentiation. Flow cytometry results revealed LPS, not
   Amuc_1409 stimulation significantly upregulated the proportion of
   CD11C^+/F4/80^+‐DCs (Figure [150]S9B, Supporting Information), and
   Amuc_1409‐stimulated BMDC did not significantly enhanced Tregs
   differentiation, compared with vehicle‐treated BMDC (Figure [151]S9C,
   Supporting Information), suggesting that Amuc_1409 promotes Tregs
   differentiation in a DCs‐independent manner.

Figure 6.

   Figure 6
   [152]Open in a new tab

   Amuc_1409 promoted Tregs differentiation and IL‐10 production through
   interaction with Ube2k. A) Mice naïve CD4^+ T cells were isolated from
   spleen, and differentiated to Tregs via treating with or without
   different does of Amuc_1409 under IL‐2 and TGF‐β condition.
   Representative flow cytometry plots (left), and the percentage of
   CD4^+/Foxp3^+‐Tregs (right) (n = 3). B) Differentiated Tregs were
   treated with or without different does of Amuc_1409 under IL‐2 and
   TGF‐β condition. Representative flow cytometry plots (left), and the
   percentage of CD4^+/CD25^+/IL‐10^+‐cells (right) (n = 3). C)
   Representative SDS‐PAGE analysis of binding complexes in Tregs
   extracts, incubated with or without His‐tagged Amuc_1409. D) HEK293T
   cells were co‐transfected with pcDNA3.1‐Amuc_1409‐His and
   pcDNA3.1‐Ube2k‐Flag. Co‐Immunoprecipitation assay was performed at 48
   ho post transfections. Representative Western blot. E) Stimulation of
   the Amuc_1409 (Green)‐Ube2k (Blue) interaction. F) Differentiated Tregs
   were transfected with siNC or siUbe2k, and treated with or without
   200 µg/mL of Amuc_1409 under IL‐2 and TGF‐β condition, the percentage
   of CD4^+/Foxp3^+‐Tregs (n = 3). G,H) Transcriptome analysis of
   differentiated Tregs transfected with siNC or siUbe2k. G) Volcano plot,
   H) Bubble plots of GO enrichment in upregulated genes in siUbe2k vs
   siNC transfected Tregs. I) HEK293T cells were co‐transfected with
   pcDNA3.1‐Foxp3‐HA and pcDNA3.1‐Ube2k‐Flag. Co‐Immunoprecipitation assay
   was performed at 72 h post transfections. Representative Western blot.
   J) HEK293T cells were co‐transfected with the indicated combinations of
   expression constructs encoding Foxp3‐HA, Ube2k‐Flag, and Myc‐tagged
   ubiquitin molecules—either wild‐type (Ub‐Myc) or a variant with a
   K‐to‐R mutation at either lysine residue 48 (K48R) or 63 (K63R)
   responsible for preventing K48‐ and K63‐type polyubiquitination,
   respectively. Co‐Immunoprecipitation assay was performed at 72 h post
   transfections. Representative Western blot. The two‐sided p‐values were
   examined by Student's t‐test (B) or one‐way ANOVA with Dunnett's
   multiple comparisons test (A, F) and data were presented as mean ± sem.

   To further test the role of Amuc_1409 on Tregs proliferation in vitro,
   we differentiated naïve CD4^+ T cells to Tregs, and then treated with
   or without Amuc_1409. Flow cytometry results revealed Amuc_1409
   remarkedly upregulated the proportion of CD4^+/Foxp3^+‐Tregs in a
   dose‐dependent manner (Figure [153]S9D, Supporting Information). It
   suggests that Amuc_1409 promotes Tregs proliferation in vitro. In
   addition, we analyzed anti‐inflammatory cytokine IL‐10 expression of
   Tregs using flow cytometry. Amuc_1409 treatment significantly increased
   the IL‐10 expression of Tregs (Figure [154]6B). We further determined
   mRNA levels of inflammatory cytokines and anti‐inflammatory cytokines
   in Tregs using qPCR. Treatment with Amuc_1409 observably decreased the
   mRNA levels of Il1b, Il6, whereas increased the expression of Foxp3,
   Il10 (Figure [155]S9E, Supporting Information), implying that Amuc_1409
   enhances anti‐inflammatory properties of Tregs.

   Then, we investigated the underlying mechanism by which Amuc_1409
   promotes Tregs differentiation and proliferation to induce IL‐10
   expression. We first observed the subcellular localization of Amuc_1409
   in Tregs via IFA, and found that Amuc_1409 mainly localized in the
   cytoplasm (Figure [156]S10A, Supporting Information), suggesting that
   Amuc_1409 may enter cells and promote Tregs differentiation by
   interacting with certain proteins. Subsequently, by performing an in
   vitro binding assay to identify the proteins interacted with Amuc_1409
   in Tregs via label‐free quantitative proteomics, we found that
   Amuc_1409 interacted with a range of proteins, such as, proteasome
   subunit beta type‐2 (Psmb2), ubiquitin‐conjugating enzyme E2 K (Ube2k),
   ATP‐dependent RNA helicase DHX8 (Dhx8), nuclear factor related to
   kappa‐B‐binding protein (Nfrkb), annexin A4 (Anxa4), protein
   phosphatase 1 regulatory subunit 12A (Ppp1r12a), and so on
   (Figure [157]6C). Among them, Psmb2 and Ube2k exhibited the highest
   protein abundance. Hence, we examined the interaction of these two
   proteins to Amuc_1409 through Co‐immunoprecipitation (Co‐IP) assay, and
   found that Amuc_1409 interacted with Ube2k but not Psmb2
   (Figure [158]6D; Figure [159]S10B, Supporting Information).
   Furthermore, we stimulated the Amuc_1409‐Ube2k interaction at the
   molecular level using AlphFold3, and found that Amuc_1409 interacted
   with six major amino acids (Asn^69, Asp^94, Gln^99, Trp^100, Ala^102,
   and Ala^103) from UBC core domain of Ube2k (Figure [160]6E). To further
   investigate whether Amuc_1409 promotes Tregs differentiation through
   interacting with Ube2k, we knocked down Ube2k expression of Tregs
   transfecting small interfering RNA (siRNA) (Figure [161]S10C,
   Supporting Information), and then treated with or without Amuc_1409 to
   determine the differentiation of Tregs using Flow cytometry. We found
   that Amuc_1409 enhanced the differentiation of Tregs transfecting with
   siNC, whereas Amuc_1409 lose the ability to promote Tregs
   differentiation when transfected with siUbe2k. In addition, knocking
   down Ube2k expression promoted differentiation of Tregs with or without
   Amuc_1409 treatment (Figure [162]6F; Figure [163]S10D, Supporting
   Information). Taken together, these results suggests that Amuc_1409
   enhances Tregs differentiation and IL‐10 production through interaction
   with Ube2k.

   To further investigate the contribution of Ube2k on regulation of Tregs
   differentiation, we performed RNA‐seq transcriptome profiling for Tregs
   transfected with siNC or siUbe2k. Differential expression analysis
   between siUbe2k and siNC‐transfected Tregs revealed 542 upregulated
   genes (log[2] fold change > 0.585, adjusted p‐value < 0.05) and 354
   downregulated genes (log[2] fold change < −0.585, adjusted p‐value <
   0.05) (Figure [164]6G). Gene Ontology (GO) enrichment analysis
   indicated that upregulated differentially expressed genes were mainly
   involved in terms such as regulation of regulatory T cell
   differentiation and regulation T cell differentiation (Figure [165]6H),
   and downregulated genes were involved in negative regulation of growth
   (Figure [166]S10E, Supporting Information). Kyoto Encyclopedia of Genes
   and Genomes (KEGG) pathway enrichment analysis revealed upregulated
   genes in Tregs‐related pathways like Th17 cell differentiation
   (Figure [167]S10F, Supporting Information), and downregulated genes
   were mainly enriched in JAK‐STAT signaling pathway (Figure [168]S10G,
   Supporting Information). Taken together, knockdown Ube2k of Tregs
   exhibited a higher enrichment of Tregs and T cells differentiation and
   cells growth related signaling.

   As a ubiquitin‐conjugating enzyme, Ube2k is capable of catalyzing the
   formation of ubiquitin chains with diverse linkage types, including
   K48‐linked and K63‐linked.^[ [169]^19 ^] Stable expression of Foxp3 is
   essential for both the generation and suppressive function of Tregs.^[
   [170]^20 ^] However, Foxp3 is subjected to K48‐linked
   polyubiquitination, which mediates its degradation via the
   proteasome.^[ [171]^21 ^] Therefore, we hypothesized that Ube2k may
   mediate Foxp3 degradation via ubiquitination modification. To test
   this, we performed Co‐IP assays with the lysates of HEK293T cells
   transfected with expression vectors encoding HA‐tagged Foxp3 and
   Flag‐tagged Ube2k. Upon immunoprecipitation of Foxp3, we detected Ube2k
   protein in the material recovered from lysate (Figure [172]6I),
   indicating that Ube2k interacts with the Treg transcription factor
   Foxp3. Next, Co‐expression of Foxp3‐HA and Ube2k‐Flag constructs, along
   with distinct Myc‐tagged ubiquitin variants, allowed us to confirm the
   nature of the Foxp3 modification arising from this interaction.
   Pull‐down of Foxp3‐HA protein from cell lysates by anti‐HA antibody
   enabled us to examine the various ubiquitinated forms of Foxp3 by
   immunoblot analysis. We observed substantial ubiquitination of Foxp3 in
   cells triple‐transfected with expression constructs encoding Foxp3,
   Ube2k, and wild‐type ubiquitin (Figure [173]6J). Interestingly, the
   expression of ubiquitin molecules with a mutated lysine residue at K48
   (K48R), which cannot support K48 polyubiquitination, led to a robust
   Foxp3 ubiquitination in this system. In contrast, expression of a K63R
   ubiquitin mutation resulted in a significant reduction of Foxp3
   ubiquitination (Figure [174]6J). These results suggest that Ube2k
   facilitates a K63‐type ubiquitination of Foxp3.

2.7. Patients with SAP Exhibit a Decreased Percentage of Tregs that was
Inversely Correlated with Fecal Genomic Copies of A. Muciniphila

   Finally, we isolated the PBMCs of healthy individuals and patients with
   SAP, and analyzed the population of peripheral Tregs by flow cytometry.
   The population of peripheral Tregs were significantly reduced in PBMC
   of patients with SAP, compared with healthy control (Figure [175] 7A).
   Subsequently, we quantified the fecal genomic copies of A. muciniphila
   and E. faecalis, and analyzed their correlations with the populations
   of peripheral Tregs, neutrophils, and monocytes. Our results revealed a
   positive correlation between the fecal genomic copies of A. muciniphila
   and the population of peripheral Tregs (Figure [176]7B). However, no
   such correlation was observed with the populations of peripheral
   neutrophils (Figure [177]7C) or monocytes (Figure [178]7D).
   Additionally, the population of peripheral Tregs showed no correlation
   with the fecal genomic copies of E. faecalis (Figure [179]7E), but the
   serum concentrations of IL‐10 showed a positive correlation with the
   fecal genomic copies of A. muciniphila (Figure [180]7F). Collectively,
   these correlation analysis data imply that A. muciniphila might
   mitigate inflammatory responses in patients with SAP by augmenting the
   population of peripheral Tregs.

Figure 7.

   Figure 7
   [181]Open in a new tab

   Amuc_1409 in regulating the population of Tregs in PBMCs from patients
   with SAP. A) The PBMCs of healthy individuals and patients with SAP
   were isolated, and the population of peripheral
   CD3^+/CD4^+/Foxp3^+‐Tregs analyzed by flow cytometry. Representative
   flow cytometry plots (left), and the percentage of
   CD3^+/CD4^+/Foxp3^+‐Tregs (right) (n = 4–15). B–F) The percentages of
   Tregs, Neutrophils, and Monocytes were determined in PBMCs of patients
   with SAP, and the genomic copies of A. muciniphila and Enterococcus
   fecalis were determined in feces of patients with SAP. B–D) Dot plot
   showing the correlation of genomic copies of A. muciniphila and
   percentages of Tregs (B), Neutrophils (C), or Monocytes (D). E) Dot
   plot showing the correlation of percentages of Tregs and genomic copies
   of E. fecalis, and F) the correlation of concentration of IL‐10 and
   genomic copies of A. muciniphila. G) The PBMCs of healthy individuals
   and patients with SAP were isolated, then treated with or without
   200 µg mL^−1 Amuc_1409 and analyzed the population of peripheral
   CD3^+/CD4^+/Foxp3^+‐Tregs by flow cytometry. Representative flow
   cytometry plots (left), and the percentage of CD3^+/CD4^+/Foxp3^+‐Tregs
   (right) (n = 8). H) Schematic illustration of A. muciniphila alleviates
   severe acute pancreatitis via Amuc1409‐ube2k‐foxp3 axis in Tregs. The
   two‐sided p‐values were examined by Student's t‐test and data were
   presented as mean ± sem (A,G). R^2 and exact two‐sided p‐values
   calculated by Pearson's test are shown (B–F).

   To explore the therapeutic efficacy of Amuc_1409 in patients with SAP,
   PBMCs were isolated from these patients and then treated either with or
   without Amuc_1409. Subsequently, the population of Tregs was analyzed
   via flow cytometry. Notably, treatment with Amuc_1409 led to a
   significant increase in the population of Tregs within the isolated
   PBMCs from SAP patients (Figure [182]7G). These findings validate the
   function of Amuc_1409 in modulating the Treg population in PBMCs
   obtained from SAP patients. Moreover, they suggest that in vivo
   administration of Amuc_1409 could potentially exert therapeutic
   benefits in patients with SAP by upregulating the Treg population,
   thereby suppressing inflammation.

3. Discussion

   In this study, we elucidated an A. muciniphila–Amuc_1409–Ube2k axis
   that enhances the anti‐inflammatory responses mediated by peripheral
   Tregs, thereby mitigating host SIRS associated with SAP and potentially
   other systemic inflammatory diseases (Figure [183]7H). The patients
   with AP exhibited a significantly lower abundance of fecal Akkermansia,
   which was also reduced with the severity of AP, suggesting a potential
   role of Akkermansia as a microbial biomarker for outcome of AP. One
   limitation of this study was the inability to collect stool samples
   from patients with AP before they show clinical symptoms. Consequently,
   we cannot determine whether the decrease in fecal Akkermansia abundance
   was a result of AP itself, or if inherent differences in fecal
   Akkermansia levels among patients contributed to varied AP outcomes.
   This raises the question of whether patients with lower baseline fecal
   Akkermansia abundance might be more susceptible to severe conditions.
   However, in our previous mouse model, we observed a significant
   reduction in fecal Akkermansia abundance in MAP mice compared to PBS
   controls,^[ [184]^9 ^] suggesting that AP may indeed contribute to a
   decrease in Akkermansia abundance. In another disease involving severe
   fever with thrombocytopenia syndrome virus (SFTSV) infection‐induced
   systemic inflammatory response, we also observed a similar phenomenon,
   where changes in the abundance of Akkermansia caused by the infection
   affected the disease outcome.^[ [185]^22 ^]

   Recent studies have highlighted the role of A. muciniphila in reducing
   systemic inflammatory responses. Research shows that A. muciniphila
   mitigates systemic inflammation through its metabolite harmaline, which
   acts via the bile acid‐G‐protein‐coupled receptor‐5 (TGR5)‐NF‐kB
   signaling.^[ [186]^22 ^] Furthermore, A. muciniphila secreted
   threonyl‐tRNA synthetase (Am TARS) targets macrophages and activates
   the anti‐inflammatory TLR‐CREB axis, leading to increase IL‐10
   production and attenuates colitis in mice.^[ [187]^23 ^] In this study,
   we observed that the protein components of A. muciniphila demonstrate
   superior efficacy compared to its metabolites in attenuating both
   pancreatic and systemic inflammatory responses. Therefore, our study
   primarily focuses on the regulatory effects of A. muciniphila‐derived
   proteins, rather than metabolites, on AP. However, this focus does not
   preclude the potential regulatory role of A. muciniphila‐derived
   metabolites such as Am TARS in AP pathogenesis. Notably, Amuc_1100, an
   outer membrane protein of A. muciniphila, plays a crucial role in
   immune regulation and intestinal barrier function by activating TLR2
   and TLR4, subsequently increasing IL‐10 production.^[ [188]^13 ,
   [189]^24 ^] Recent literature also highlights the role of Amuc_1100 in
   acute pancreatitis, with findings indicating that Amuc_1100
   pretreatment alleviates AP by reducing macrophage and neutrophil
   infiltration and modulating gut microbiota composition and tryptophan
   metabolism.^[ [190]^10 ^] Consistent with these findings, our study
   demonstrates that Amuc_1100 effectively mitigates SAP‐induced
   inflammatory responses and pancreatic immunopathological injury. This
   study revealed that another key protein, Amuc_1409, significantly
   reduced pancreatic and systemic inflammatory responses, thereby
   alleviated pancreatic immunopathological injury. Amuc_1409 is
   recognized as one of the most abundant secreted proteins, capable of
   localizing within the periplasmic space and outer membrane before
   export to the extracellular space.^[ [191]^25 ^] Additionally,
   Amuc_1409 exhibits high thermal stability, similar to that of
   Amuc_1100.^[ [192]^12 , [193]^13 ^] Future studies are required to
   develop a mutant A. muciniphila strain with silenced Amuc_1100 and/or
   Amuc_1409 genes to determine whether A. muciniphila's protective
   effects against SAP are mediated through Amuc_1100 and/or Amuc_1409.
   Currently, there have been no reported side effects associated with
   Amuc proteins. However, it is conceivable that prolonged or high‐dose
   administration of proteins such as Amuc_1100 may excessively activate
   Toll‐like receptors (TLRs) in vivo, leading to systemic immune
   overactivation, cytokine storms, or immune dysregulation. Consequently,
   further investigations are required to clarify the appropriate dosage,
   frequency, and potential adverse effects of these immunomodulatory
   proteins.

   Using Foxp3‐DTR mice, we verified that the protective effects of A.
   muciniphila and its secreted protein Amuc_1409 against SAP in mice
   depend on the presence of Tregs, underscoring the critical role of
   Tregs in SAP. Recent studies have shown that while preventive depletion
   of Tregs enhances inflammatory responses, it concurrently stabilizes
   the immune barrier functions of Th17 cells and CD8^+γδTCR^+ intestinal
   epithelial cells, significantly reducing the translocation of
   intestinal bacteria to the pancreas.^[ [194]^26 ^] However, our results
   indicate that the increase in Tregs in PBMCs helps alleviate
   inflammation induced by SAP, seemingly presenting contradictory
   conclusions. This discrepancy arises from the distinct roles Tregs play
   in different organs during AP. In the intestine, Tregs disrupt the
   immune barrier functions of Th17 cells and CD8^+γδTCR^+ intestinal
   epithelial cells, increasing the likelihood of bacterial translocation
   from the gut to the pancreas. The referenced study used a bile duct
   ligation‐induced acute pancreatitis model, which elicited a relatively
   mild inflammatory response in mice. Thus, Tregs with immunosuppressive
   functions did not inhibit the body's inflammatory response but rather
   weakened the intestinal immune barrier, exacerbating bacterial
   translocation. In contrast, our SAP model, induced by cerulein combined
   with LPS or arginine combined with LPS, generated a robust initial
   inflammatory response. As a result, the Tregs induced by A. muciniphila
   and Amuc_1409 could rapidly suppress excessive inflammation, thereby
   alleviating SAP in mice. This dual role of Treg function presents a
   critical therapeutic challenge: the precise timing and dosage of A.
   muciniphila or its driven protein Amuc_1409 administration must be
   carefully optimized to balance their beneficial anti‐inflammatory
   effects in early AP against potential risks of immunosuppression in
   later stages. Addressing this challenge will be a major focus of our
   future research.

   During the initial phase of SAP or sepsis, a massive release of
   pro‐inflammatory cytokines (e.g., IL‐6, TNF‐α) establishes a critical
   immunoregulatory imbalance. Elevated IL‐6 levels suppress Treg
   differentiation and survival through STAT3 signaling pathway
   activation, while concurrently driving Th17 cell expansion.^[ [195]^27
   ^] This imbalance leads to a transient depletion of Tregs as the immune
   system prioritizes pro‐inflammatory responses, potentially through Treg
   functional inhibition or apoptosis. Our experimental design
   specifically targets this early inflammatory window, with sample
   collection timed to capture the cytokine storm phase, aiming to develop
   strategies that enhance Treg‐mediated immunosuppression to rapidly
   terminate pathological hyperinflammation. Notably, excessive Treg
   activation may paradoxically induce the CARS, predisposing patients to
   infectious pancreatic necrosis during later disease stages. This
   biphasic immunopathology underscores the therapeutic challenge of
   precisely regulating Treg induction timing and activation intensity.
   Optimal intervention requires: stage‐specific modulation of Treg
   plasticity, threshold‐controlled immunomodulatory signals, and
   microenvironment‐adapted functional programming. These parameters
   constitute critical foci for our ongoing investigations into SAP
   pathophysiology and targeted immunotherapy development.

   Through in vitro pull‐down and Co‐IP experiments, we confirmed
   Amuc_1409 interacts with the ubiquitin‐conjugating enzyme Ube2k,
   promoting Treg differentiation, proliferation, and anti‐inflammatory
   IL‐10 expression in a Ube2k‐dependent manner. Ube2k, an E2
   ubiquitin‐conjugating enzyme, plays an important role in gene
   regulation and cellular function. Recent studies have shown that Ube2k
   binds to histone H3, induces its ubiquitination, and regulates neural
   stem cell differentiation and proliferation through proteasome‐mediated
   degradation of histone H3.^[ [196]^28 ^] Additionally, Ube2k promotes
   the proliferation and migration of hepatocellular carcinoma cells by
   upregulating c‐Myc, while knockdown of Ube2k significantly inhibits the
   malignancy of liver cancer.^[ [197]^29 ^] These findings highlight
   Ube2k's critical regulatory role in cell differentiation and
   proliferation. Our results demonstrate that Ube2k directly interacts
   with Foxp3 and promotes its K63‐linked polyubiquitination, which may
   lead to protein destabilization of this critical transcriptional
   regulator of Treg cell differentiation and IL‐10 production. Given
   Foxp3's pivotal role in maintaining Treg populations and
   anti‐inflammatory responses, Ube2k‐mediated ubiquitination could
   consequently reduce Treg numbers and diminish the transcription of key
   anti‐inflammatory factors. Importantly, we propose that Amuc_1409 may
   exert its protective effects by binding to Ube2k and competitively
   inhibiting Foxp3 ubiquitination, thereby preserving Foxp3 stability and
   promoting Treg differentiation along with IL‐10 expression. Future
   studies employing genetic manipulation approaches (e.g.,
   CRISPR‐Cas9‐mediated knockout or overexpression of Ube2k in Tregs)
   would provide mechanistic validation of its role in Amuc_1409‐induced
   Treg cell differentiation.

   In conclusion, this study demonstrates that Amuc_1409, derived from A.
   muciniphila, plays a crucial role in suppressing SAP‐induced
   inflammation through its direct interaction with the Ube2k protein in
   Tregs. This work advances the understanding of microbiota‐host
   interactions mediated by bioactive molecules. More importantly, it
   opens up potential avenues for developing a new probiotic product aimed
   at improving acute pancreatitis and other inflammatory disorders.

4. Experimental Section

Ethics Statement

   The experimental protocols were approved by the Institutional Animal
   Care and Use Committee of Sir Run Run Shaw Hospital (No. 202 402 078).
   All methods were carried out in accordance with relevant guidelines and
   regulations. This study was conducted as recommended by the ARRIVE
   guidelines.

Patients and Sample Collection

   A total of 86 hospitalized patients (38 with MAP, 16 with MSAP, and 32
   with SAP) and 46 healthy control volunteers (HC), aged between 20 and
   88 years old, were included in the study. These individuals, who
   exhibited symptoms within 7 days of onset, were recruited from the
   Intensive Care Unit at Sir Run Run Shaw Hospital, affiliated with
   Zhejiang University Medical College. All participants met the criteria
   for SAP as defined by the revised Atlanta classification. Faecal and
   serum specimens were collected from patients on admission. Specimens
   were stored in sterile EP tubes, and kept at −80 °C. Every participant
   provided a written informed consent, adhering to protocols approved by
   the ethical committees of Sir Run Run Shaw Hospital (No. 2025‐2053‐01).

16S rRNA Amplicon Sequencing and Data Analyses

   ≈200 mg of fecal samples were resuspended in Qiagen's ASL buffer and
   homogenized for 2 min. Total fecal DNA was extracted from the
   supernatant using the QIAamp DNA Stool Mini Kit (Qiagen). The
   concentration and purity of the extracted DNA were assessed using the
   Qubit fluorometer (Thermo Fisher Scientific).

   For amplification, the stool DNA was subjected to PCR using the Phusion
   High‐Fidelity PCR Master Mix (New England Biolabs), targeting the V3–V4
   variable regions of the 16S rRNA gene. The forward primer used was
   ACTCCTACGGGAGGCAGCA, and the reverse primer was GGACTACHVGGGTWTCTAAT.
   Amplicons, tagged with sample‐specific barcodes, were then sequenced
   using an Illumina NovaSeq platform.

   Paired‐end reads from the sequencing were merged into longer sequences
   using FLASH v.1.2.7,^[ [198]^30 ^] a tool optimized for merging
   paired‐end reads with overlapping regions. The merged sequences were
   subsequently analyzed with the QIIME v.1.9.1 software package.^[
   [199]^31 ^]

Animals, Bacteria, and Cell Culture

   C57BL/6J mice were obtained from the Model Animal Research Center of
   Nanjing University (Nanjing, China). B6‐Il10‐KO mice (Strain NO.
   T005959) were purchased from GemPharmatech (Nanjing, China). All
   experiments were conducted using 6‐week‐old mice with an average weight
   of 20 ± 5 g. The mice were housed in a sterile environment under
   controlled conditions, with an average temperature of 22 °C and a
   standard 12 h light/dark cycle (7:00 am to 7:00 pm) at the animal
   experimentation center of Zhejiang University.

   Akkermansia muciniphila (catalog no. BAA835) was obtained from the
   American Type Culture Collection (ATCC) and cultured in Brain Heart
   Infusion (BHI) medium (Oxoid) at 37 °C under anaerobic conditions.
   Enterococcus faecalis (catalog no. AB 2 018 154) was purchased from the
   CCTCC and cultured in lysogeny broth (LB) medium (Oxoid) at 37 °C. The
   concentration of each bacterial species was quantified by measuring the
   optical density at 600 nm (OD600).

   HEK293T cells were purchased from ATCC and cultured in DMEM with 10%
   fetal bovine serum (FBS, Gibco). Human PBMCs were extracted by
   density‐gradient centrifugation using human PBMC isolation kits (TBD)
   following the manufacturer's protocol. After isolation, PBMCs were
   cultured in RPMI‐1640 medium supplemented with 10% FBS at 37 °C.

   Naïve CD4^+ T cells were isolated using EasySep Mouse Naïve CD4^+ T
   Cell Isolation Kit (STEMCELL) following the manufacturer's protocol.
   After isolation, Naïve CD4^+ T cells were plated to 1 µg mL^−1 CD3/CD28
   pretreated cell plates and cultured in RPMI‐1640 medium supplemented
   with 10% FBS, 1 ng mL^−1 TGF‐β, 5 ng mL^−1 IL‐2 at 37 °C. Tregs were
   harvested at 3 days post‐differentiation.

   Isolation and in vitro induction of murine bone marrow‐derived
   dendritic cells (BMDCs) and murine bone marrow‐derived macrophages
   (BMDMs) were carried out as previously described.^[ [200]^32 ^]
   Briefly, the femurs and tibias were surgically excised and thoroughly
   washed with sterile phosphate‐buffered saline (PBS) to procure a cell
   suspension. Subsequently, red blood cells were lysed, and bone marrow
   cells (BM cells) were harvested by centrifugation. For induction of
   BMDCs, the BM cells were then plated at a density of 1 × 10^6 cells
   mL^−1 in 2 mL of cell culture medium supplemented with 20 ng mL^−1
   granulocyte‐macrophage colony‐stimulating factor (GM‐CSF) and 20 ng
   mL^−1 interleukin‐4 (IL‐4) in the central area of 6‐well plates. For
   induction of BMDMs, the BM cells were then plated at a density of
   1 × 10^6 cells mL^−1 in 2 mL of cell culture medium supplemented with
   20 ng mL^−1 GM‐CSF in the central area of 6‐well plates. On the third
   and fifth days of culture, half of the supernatant was carefully
   aspirated, and fresh cytokines were added at an equal volume. On Day 5,
   Amuc_1409 (at a concentration of 100 µg mL^−1) or lipopolysaccharide
   (LPS, at a concentration of 1 µg mL^−1) was incorporated into the
   culture media.

   Foxp3‐DTR mice received intraperitoneal (i.p.) injections of 1 µg DT
   for 5 days to deplete Tregs. The efficiency of depletion was determined
   by Flow cytometry.

Antibiotics (Abx) Treatment, Faecal Microbial Transfer, and Bacterial
Colonization

   Mice were administered an antibiotic cocktail consisting of ampicillin,
   neomycin, metronidazole, and vancomycin via oral gavage for 5 days,
   following previously described protocols.^[ [201]^33 ^] Antibiotics
   were also added to their drinking water, and the animals were
   maintained on either Abx‐containing or PBS‐containing water for the
   duration of the experiments.

   For FMT experiments, fecal sample were collected from none‐treated
   mice, HC volunteers or patients with SAP and processed as described
   previously.^[ [202]^34 ^] In brief, fecal samples were weighted and
   homogenized with 1 mL sterile PBS, filtered through a 100 µm strainer,
   and centrifuged at 6000 g for 15 min. The pellets were resuspended in
   PBS with 10% (v/v) glycerol and stored at −80°C. Fecal sample were
   centrifuged and resuspended in sterile PBS, and gavaged to Abx‐treated
   mice with 20 mg fecal in 200 µL PBS at 6 d after Abx treatment.

   For the bacterial colonization and FMT experiments, Abx mice were
   gavaged with 10^10 CFU of A. muciniphila, E. faecalis or 20 mg FMT in
   200 µL of PBS at 6 d after Abx treatment. After 48 h of colonization,
   SAP models were constructed.

Bacterial Supernatant and Protein Components Preparation

   A. muciniphila cell were centrifuged at 6000 g for 15 min, and the
   collected supernatant was centrifuged at 12 000 g for 15 min at 4 °C.
   Subsequently, the supernatant was passed through polyether‐sulfone
   filters (0.22 µm, Merck Millipore) to remove the residual bacterial
   cells. The supernatants were passed through 100, 50 and 10 kDa filters
   (Merck Millipore) at 3200 g for 10 min. Each intercepting fluid
   (including secretory proteins) was mixed with A. muciniphila cell, and
   then the mixture was inactivated by pasteurization for 30 min at 70 °C
   (Named p‐AKK). The filtrate passed 10 kDa filter were collected and
   named s‐AKK.

   The mice gavaged with 200 µL s‐AKK or 10^9 CFU p‐AKK were initiated in
   parallel with Abx treatment until the end of experiments.

Expression and Purification of His‐Tagged Amuc_1409 and His‐Tagged Amuc_1100
Protein

   The Amuc_1409 and Amuc_1100 genes were synthesized and subcloned into
   pET30a vectors using E. coli DH5α. The resulting plasmids, pET30a‐1409
   and pET30a‐1100, were verified by DNA sequence analysis. The purified
   plasmids containing the recombinant Amuc_1409 and Amuc_1100 genes were
   transformed into E. coli BL21 (DE3). The transformed E. coli strains
   were grown in LB broth supplemented with kanamycin (50 µg mL^−1) at 30
   °C while shaking at 120 rpm. In the mid‐exponential phase, 0.5 mm IPTG
   was added to induce protein expression. After centrifugation at 6000 g
   for 20 min at 4 °C, the cells were harvested. Next, the cell pellets
   were resuspended in an ice‐cold buffer consisting of 300 mm NaCl, 50 mm
   Tris‐HCl (pH 8.0), and 10 mm imidazole. The cell suspensions were
   sonicated to obtain crude cell extracts. The crude cell extracts were
   further processed by centrifugation at 12 000 g for 30 min at 4 °C. The
   resulting cell lysates containing the 6xHis‐Amuc1409 and 6xHis‐Amuc1100
   fused proteins were applied to Ni Sepharose 6 Fast Flow columns.
   Endotoxin was removed from the recombinant proteins using high‐capacity
   endotoxin removal spin columns according to the manufacturer's
   instructions. After buffer exchange using 7 MWCO zeba spin desalting
   columns (Thermo Fisher Scientific), the protein contents were
   determined using the BCA assay (Thermo Fisher Scientific). The purified
   Amuc_1409 and Amuc_1100 proteins were then stored at −80 °C.

   The mice gavaged with Amuc_1409 or Amuc_1100 proteins (5 µg per mouse)
   were initiated in parallel with Abx treatment until the end of
   experiments.

Induction of Severe Acute Pancreatitis Mouse Models

   For the cerulein‐induced SAP model, a group of mice received seven
   intraperitoneal (i.p.) injections of cerulein (50 µg kg^−1, MCE) at 1 h
   intervals, followed by a single i.p. injection of LPS (10 mg kg^−1,
   Sigma) after the final cerulein injection. Samples were collected 12 h
   after the first cerulein injection.

   For the cerulein‐induced MAP model, a group of mice received seven
   intraperitoneal (i.p.) injections of cerulein (50 µg kg^−1, MCE) at 1 h
   intervals, after the final cerulein injection. Samples were collected
   12 h after the first cerulein injection.

   For the L‐arginine‐induced SAP model, mice were given two hourly i.p.
   injections of 10% L‐arginine (4 g kg^−1, pH 7). Three days later, the
   mice received an i.p. injection of LPS (10 mg kg^−1), and samples were
   collected 4 h after the initial LPS injection.

Serum Amylase and Cytokine Expression Analysis

   Serum amylase concentration was determined by the Alpha‐amylase
   Determination Kit (BIOSINO, Beijing, China) following the
   manufacturer's protocol.

   Total RNA from bead‐homogenized tissue samples or cell culture was
   extracted using TRIzol reagent (Invitrogen) following the
   manufacturer's protocol. The cytokine level was determined by
   quantitative PCR with reverse transcription using the HiScript II One
   Step qRT–PCR SYBR Green Kit (Vazyme).

   The following primers were used, GAPDH, forward,
   5′‐GCCTTCCGTGTTCCTACCC‐3′; reverse, 5′‐CCCTCAGATGCCTGCTTCAC‐3′. IL‐6,
   forward, 5′‐AGGATACCACTCCCAACAGA‐3′; reverse,
   5′‐ACTCCAGGTAGCTATGGTACTC‐3′. IL‐1β, forward, 5′‐
   AAGCCTCGTGCTGTCGGACC‐3′; reverse, 5′‐ TGAGGCCCAAGGCCACAGGT‐3′. TNF‐α,
   forward, 5′‐CATCTTCTCAAAATTCGAGTGACAA‐3′; reverse, 5′‐
   CCAGCTGCTCCTCCACTTG‐3′. UBE2K, forward, 5′‐CAGCGAATCAAGCGGGAGTT‐3′;
   reverse, 5′‐AGGTCCTGCTATTTCTCCTCTT‐3′.

   IL‐6, TNF‐α, IL‐10 protein levels in serum samples were measured by the
   corresponding enzyme‐linked immunosorbent assay (ELISA) kits
   (MultiSciences) following the manufacturer's protocol.

Fecal Bacteria Quantification

   Fecal bacterial quantification was determined through qPCR. The fecal
   bacterial DNA was isolated with a TIANamp Stool DNA Kit (TIANGEN). SYBR
   Green Real‐time PCR Master Mix (TOYOBO) was used for qPCR.

   The following primers were used, Akkermansia muciniphila, forward,
   5′‐CAGCACGTGAAGGTGGGGAC‐3′; reverse, 5′‐CCTTGCGGTTGGCTTCAGAT‐3′.
   Enterococcus fecalis, forward, 5′‐CCGAGTGCTTGCACTCAATTGG‐3′; reverse,
   5′‐CTCTTATGCCATGCGGCATAAAC‐3′.

Tissue Histology and Staining

   Pancreas, colon, and duodenum tissues were fixed in 4%
   paraformaldehyde, dehydrated, embedded in paraffin, and sectioned into
   4 µm thick slices. These sections were then stained with hematoxylin
   and eosin (H&E) following standard protocols.

   For immunohistochemistry, deparaffinized pancreas sections were blocked
   with 10% normal goat serum for 30 min and incubated overnight at 4 °C
   with an anti‐IL‐6 polyclonal antibody or an anti‐IL‐10 polyclonal
   antibody (Proteintech). The pancreas, spleen, colon, and duodenum were
   blocked with 10% normal goat serum for 30 min and incubated overnight
   at 4 °C with the anti‐ZO‐1 polyclonal antibody, anti‐OCLN polyclonal
   antibody, anti‐MUC‐2 polyclonal antibody, anti‐CD3 polyclonal antibody,
   anti‐Foxp3 polyclonal antibody, and anti‐His polyclonal antibody
   (Proteintech). The sections were subsequently incubated with a
   FITC/Cy3/Cy5‐conjugated anti‐rabbit secondary antibody for 55 min at
   room temperature. Finally, the sections were incubated with
   4′,6‐diamidino‐2‐phenylindole (DAPI) solution for 10 min at room
   temperature. Fluorescence microscopy was performed, and images were
   captured accordingly.

Isolation of Lymphocytes from Colon

   Mice were euthanized by cervical dislocation, and the colon was
   aseptically excised. After removing Peyer's patches and adipose tissue,
   the intestine was longitudinally opened, washed with PBS, and cut into
   fragments. Tissue digestion was performed using a solution containing
   1 mg mL^−1 collagenase D and 1 mg mL^−1 DNase I in RPMI‐1640 medium (37
   °C, 40–60 min). The digested suspension was filtered through a 70 µm
   cell strainer, and single‐cell suspensions were enriched via Percoll
   gradient centrifugation (30 and 70% layers, 2000 rpm, 20 min). Cells at
   the 70–30% interface were collected, washed, and resuspended in
   staining buffer.

Flow Cytometry Analysis

   Flow cytometry analysis was performed on PBMCs, colonic lymphocytes,
   and splenocytes. The cells were separated through a 70 µm cell
   strainer. After washing and lysis of erythrocytes using red blood cell
   lysis buffer (solarbio), cells were preincubated with FcR blocking
   reagent (Biolegend) and stained for extracellular surface marker (CD4,
   CD25; CD11b, F4/80, Ly6G) at 4 °C for 20 min. For staining of Treg
   cells, the True‐Nuclear Transcription Factor Buffer Set (Biolegend) was
   used for fixation and permeabilization. Then, the cells were stained
   for transcription factors (Foxp3) at 4 °C for 60 min. For staining of
   Th17 cells, the Fixation Buffer and Intracellular Staining Perm Wash
   Buffer (Biolegend) were used for fixation and permeabilization. Then,
   the cells were stained for IL17A at 4 °C for 60 min. Data evaluation
   was performed by flow cytometry (CytoFLEX LX, BECKMAN).

Single‐Cell Sequencing

   scRNA‐seq was performed by Cosmos Wisdom Biotech Co., Ltd. In brief,
   the scRNA‐seq libraries were generated using the 10X Genomics Chromium
   Controller Instrument and Chromium Single Cell 3′ V3.1 Reagent Kits
   (10X Genomics, Pleasanton, CA). Briefly, cells were concentrated to
   1000 cells uL^−1 and ≈10 000 cells were loaded into each channel to
   generate single‐cell Gel Bead‐In‐EMulsions (GEMs), which resulted into
   expected mRNA barcoding of 5000 single‐cells for each sample. After the
   RT step, GEMs were broken and barcoded‐cDNA was purified and amplified.
   The amplified barcoded cDNA was fragmented, A‐tailed, ligated with
   adaptors and index PCR amplified. The final libraries were quantified
   using the Qubit High Sensitivity DNA assay (Thermo Fisher Scientific)
   and the size distribution of the libraries were determined using a High
   Sensitivity DNA chip on a Qsep100 (Bioptic). All libraries were
   sequenced by NovaSeq 6000 (Illumina, San Diego, CA) on a 150 bp
   paired‐end run.

Single‐Cell RNA Statistical Analysis

   Fastp^[ [203]^35 ^] was applied with default parameter filtering the
   adaptor sequence and removed the low quality reads to achieve the clean
   data. Then, the feature‐barcode matrices were obtained by aligning
   reads to the mouse genome using CellRanger v7.1.0 (10X Genomics). All
   scRNA‐seq reads were aligned to the mouse reference genome (mm10) and
   the cell‐by‐gene count matrices were produced by the CellRanger
   pipeline with default parameters. This process performed sequence
   alignment, filtering, label calculation, and unique molecular
   identifier (UMI) calculation. Finally, the gene file, tag file, and
   gene‐tag matrix file corresponding to each sample were generated.

   The result files generated by CellRanger were processed using R package
   Seurat (v4.0.4) to individually integrate each sample and generate
   Seurat objects. The filtration criteria for low quality cells were as
   follows: 1) The abnormally total number of genes. 2) The abnormally
   total number of transcripts. 3) The high expression percentage of
   mitochondrial genes, ribosomal genes, heat‐shock genes, and dissociated
   genes in each cell. The quickPerCellQC function in the R package scater
   (v1.18.6) was used to filter low quality cells. The nmad parameter in
   the function quickPerCellQC represented the median absolute deviation
   (MAD, in this study nmad = 3). Cells that meet one of the above quality
   control criteria were considered to be of low quality and were filtered
   out.

   After quality control of each sample, the Seurat objects of all samples
   were integrated, and the Seurat package was used to perform standard
   process on the merged objects for subsequent analysis. Specifically,
   variable genes were found using the FindVariableFeatures function with
   parameters: selection.method = “vst”, nfeatures = 500, and the
   ScaleData function was used to regress out variation due to differences
   in total UMIs per cell. Principal component analysis was used on the
   scaled expression profiles for the variable genes, and the top 20
   principal components (PCs) were conducted for cell clustering. Cells
   were clustered in PCA space using Seurat's FindNeighbors on the top 30
   PCs, followed by shared nearest neighbor modularity optimization‐based
   clustering algorithm. All cells were visualized by uniform manifold
   approximation and projection (UMAP). Cell subtypes were classified and
   annotated by examining differentially expressed genes in each cell
   subtype and observing the expression of specific marker genes of known
   cell types collected from publications. The FindAllMarkers function was
   used to calculate the differentially expressed genes of each subtype
   with log[2] fold change = 0.5. A large number of classical marker genes
   specifically expressed for different cell subtypes were collected.
   After annotating the main cell subtypes, the cell subtypes were
   classified and annotated in more detail in the same way, such as T
   cells can be annotated as CD4^+ cells, CD8^+ cells, etc.

   To quantify the enrichment of cell subtypes in different tissues, the
   method reported in the previous study was adopted to calculate the
   ratio of the observed and expected number of subtype cells in the
   tissues.^[ [204]^36 ^] The formula was as follows:
   [MATH: <mrow><mrow><msub><mi mathvariant="normal">R</mi><mrow><mi
   mathvariant="normal">o</mi><mo>/</mo><mi
   mathvariant="normal">e</mi></mrow></msub><mo
   linebreak="badbreak">=</mo><mfrac><mi>Observed</mi><mi>Expected</mi></m
   frac></mrow></mrow> :MATH]
   (1)

   The observed value was the actual number of subtype cells in specific
   tissue, and the expected value was obtained by counting the number of
   subtype cells and the number of specific tissues using Chi‐square test.
   When the R [o/e ]> 1, this subtype was considered to be enriched in
   this tissue.

Bacteriological Analysis of Murine Pancreatic Tissue

   The pancreas of mice was removed under sterile conditions, weighed, and
   immediately homogenized in sterile PBS. Homogenates were plated on LB
   Agar plates and cultured for 24 h at 37 °C. Colonies were counted and
   CFU/mg pancreatic tissue were calculated.

RNA Interference

   siRNAs were transfected into isolated Tregs with Lipofectamine 3000
   reagent (Invitrogen) following the manufacturer's instructions. Mouse
   Ube2k‐specific siRNA was designed and synthesized by Hanheng
   Biotechnology. The efficiency of interference was determined by qPCR.

Immunoprecipitation, Pull‐Dwn and Immunoblotting Assays

   HEK293T cells were co‐transfected with plasmid as indicated for 48 h
   and lysed with cell lysis buffer for Western blot and
   immunoprecipitation (Beyotime). Whole‐cell lysates (WCL) were used as
   input and were subjected to IP with 2 µg antibody overnight. 50 µL
   protein G magnetic beads (Bio‐Rad) prewashed with 0.1% PBST were added
   and incubated for another 4 h at 4 °C. Protein complex‐containing beads
   were washed four times with 0.1% PBST. Proteins were eluted and boiled
   in 1% SDS loading buffer.

   For pull down assays, Dynabeads His‐tag magnetic beads (Invitrogen)
   were preincubated with either vehicle or his‐tagged Amuc_1409 protein
   in PBS for 2 h at room temperature, then washed with 0.01% PBST, and
   incubated with Tregs cell lysates overnight at 4 °C. The beads were
   washed four times with 0.01% PBST and analyzed by immunoblot.

   Tissues and cells treated as indicated were lysed with RIPA lysis
   buffer (Beyotime). The lysates were subjected to 12% SDS–polyacrylamide
   gel electrophoresis and then transferred to polyvinylidene difluoride
   membranes (Millipore). Proteins were further incubated with the
   indicated primary antibodies and then horseradish peroxidase
   (HRP)‐conjugated secondary antibodies. Protein bands were probed using
   an enhanced chemiluminescence kit (Vazyme) with a ChemiDoc Touch Gel
   Imaging System (Bio‐Rad).

Plasmid Constructs

   Full‐length cDNAs of mice Pbmb2, Ube2k, Foxp3, ubiquitin,
   ubiquitin‐K48R, ubiquitin‐K63R fragments were synthesized, then
   subcloned into pcDNA3.1 vectors (Clontech). All constructs were
   confirmed by DNA sequencing.

Statistics

   Statistical analyses were performed with Prism GraphPad
   softwarev.10.1.1. Error bars represented standard errors of the means
   in all figures, and the two‐sided p‐values were examined by one‐way
   ANOVA with Dunnett's multiple comparisons test or Student's t‐test. R^2
   and exact two‐sided p‐values calculated by Pearson's test. A two‐sided
   p‐value < 0.05 was considered statistically significant.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   J.X., L.D., and Y.L. contributed equally to this study. H.Y., F.G., and
   J.X. designed the experiments. J.X., L.D., and X. Z. performed the
   experiments. J.X. and Y.L. conducted bioinformatics analysis. X.G.,
   Y.T., B.S., and X.Y. commented on and revised drafts of the manuscript.
   J.X. wrote the paper. H.Y., F.G., X.Y., and J.X. supervised research,
   coordination and strategy.

Supporting information

   Supporting Information
   [205]ADVS-12-e04214-s002.docx^ (9.4MB, docx)

   Supporting Information
   [206]ADVS-12-e04214-s001.pdf^ (6MB, pdf)

Acknowledgements