Society: AGA

Background & Aim:
Phosphatase and tensin homolog (PTEN) is a lipid phosphatase, while its protein phosphatase activity is being suggested. Since the classical role of PTEN as a tumor suppressor is unclear in the gut, we hypothesize PTEN’s alternative function in regulating gut homeostasis.

Methods:
We examined human colonic mucosal biopsies of pediatric Ulcerative Colitis (pUC) patients and control subjects to evaluate PTEN mRNA levels. We generated intestinal epithelial cell (IEC)-Pten-knockout (KO) (Pten-dIEC/dIEC) mice using the Villin-Cre system. We tested in vivo gut permeability with FITC-Dextran, intestinal epithelial cell junction, and autophagy using transmission electron microscopy and fluorescence confocal microscopy. We knocked down (KD) endogenous PTEN gene in human colonic epithelial HT-29 and DLD-1 cells to perform immunoblotting and immunoprecipitation. We used fluorescence in situ hybridization and 16S rRNA gene sequencing. We generated Pten-dIEC/dIEC;Myd88-KO, Pten-dIEC/dIEC;Tlr5-KO, and Pten-dIEC/dIEC;Tlr4-KO mice to study the pathological link of IEC-Pten-KO.

Results:
PTEN mRNA level is 70% reduced (P<0.01, Mann-Whitney U test) in colonic mucosa tissues of pUC (n=9) compared to controls (n=19). IEC-Pten-KO mice exhibit disrupted intestinal epithelial cell junctions and increased in vivo gut permeability. These mice have reduced levels of cell-junction proteins (ZO-1, JAM-C, Claudin-1) in IECs. PTEN-KD decreases cell-junction protein levels, but increases SNAIL/SLUG transcription repressor proteins in HT-29 and DLD-1 cells. PTEN-KD promotes the phosphorylation of Ubiquitin protein and increases the ubiquitination of SNAIL/SLUG in IECs. Loss of PTEN initiates the autophagy pathway in IECs. But, it lowers YKT6 levels to inhibit the fusion between autophagosome and lysosome in IECs. Thereby, PTEN deficiency inhibits the protein degradation machinery of autophagy. We further observed that IEC-Pten-KO mice have increased Bacteroides, but decreased Akkermansia muciniphila in the feces. IEC-Pten-KO mice have increased gut bacterial dissemination to the colon submucosa. Accordingly, Pten-dIEC/dIEC;Myd88-KO, Pten-dIEC/dIEC;Tlr5-KO mice develop massive inflammation in extra-intestinal organs, including the liver and kidney. Pten-dIEC/dIEC;Tlr4-KO mice exhibit mild inflammation in the liver where the abundance of Bacteroides is markedly increased.

Conclusion:
PTEN deficiency promotes the ubiquitination of SNAIL/SLUG transcription repressor protein, while blocking the autophagy flux by inhibiting the autolysosome formation in IECs. Thus, ubiquitinated SNAIL/SLUG cannot be degraded in PTEN-deficient IECs, resulting in augmented SNAIL/SLUG that suppresses the expression of cell-junction proteins and then disrupts the intestinal epithelium. Consequently, PTEN deficiency in IECs increases the gut permeability and risk of gut microbe-induced disorders.

Background and Aims: The Aryl Hydrocarbon Receptor (AhR) plays important functions in intestinal stem cell differentiation, intestinal homeostasis, and immune regulation in the gut, however, its role in regulating intestinal immune tolerance remains poorly understood. In this study, we assessed the role of AhR knockout on the intestinal paracellular tight junction (TJ) barrier, and how it affects the factors influencing intestinal anergy and immune tolerance.
Methods: Ahr^fl/fl, and AhR^Vil-Cre mice were maintained in a specific pathogen-free area until injected with tamoxifen to delete the AhR gene, and then moved to the normal housing. The AhR deletion was maintained by weekly tamoxifen injections for 3 weeks. We performed physiological assessments of the gut epithelial barrier and used flow cytometry, and qRT-PCR analysis to determine the effect of epithelia-specific AhR^-/- on the gut immune system.
Results: Conditional gut epithelium-specific knockout of AhR in AhR^Vil-Cre mice significantly increased the colonic transepithelial resistance (TER) and reduced the paracellular flux of small molecule, urea, and macromolecule inulin, compared to AhR^fl/fl mice. Assessment of transcript levels in the mouse colonic tissues, after 3 weeks of AhR deletion, showed that AhR knockout did not alter the baseline levels of TNF-α, IFN-γ, IL-4, and TGF-β, however, the transcript levels of IL-6, IL-1β, and Il-17A showed marked elevation. Inflammatory stimuli in the form of intraperitoneal LPS administration further amplified the increase in IL-6, IL-1β, and IL-17A mRNA levels in AhR^Vil-Cre mice. Knocking out AhR in the gut epithelial cells also reduced the colonic transcript levels of IL-10, a key anti-inflammatory cytokine, and Programmed death-ligand 1 (PD-L1) and cytotoxic T-lymphocyte-associated protein 4 (CTLA4), the two key proteins, involved in limiting the pro-inflammatory responses in the gut mucosa. Flow cytometry analysis using control Ahr^fl/fl and AhR^Vil-Cre mouse colons subjected to the same tamoxifen administration cycles and LPS challenge showed an increased abundance of IL-6^hi inflammatory M1 macrophages and IL-17A^hi RoRγt⁺CD4+ Th17 cells in the AhR^Vil-Cre colonic lamina propria under baseline conditions which were further increased by the administration of LPS.
Conclusion: Our data show that deletion of the AhR gene in the gut epithelial cells reduces the epithelial paracellular TJ permeability. This causes the down-regulation of immunological-anergy-promoting proteins and increases the presence of circulating M1 polarized macrophages and Th17 cells in the mice colon in the event of an inflammatory challenge. Our study also highlights the role of paracellular TJ permeability in intestinal immune homeostasis.

Background: Microvillus inclusion disease (MVID) is a congenital disorder characterized by severe diarrhea that presents within the first few days of life. MVID is caused by inactivating mutations in myosin Vb (MYO5B), a motor protein. Mice lacking MYO5B in their intestinal epithelium (VilCre^ERT2; MYO5B^fl/fl) have demonstrated the importance of MYO5B in enterocyte brush border formation and epithelial cell differentiation. Additionally, the proliferative zone of the intestinal crypt of MYO5B-deficient mice is elongated compared to controls. This study aims to better understand the differentiation deficits and crypt-specific effects caused by MYO5B loss utilizing the progenitor cell-specific Cre driver under Lrig1.

Methods: Tamoxifen-inducible Lrig1-Cre^ERT2;R26R-YFP and Myo5b^fl/fl mice were crossbred to generate Lrig1-Cre^ERT2;R26R-YFP;MYO5B^fl/fl(Lrig1^ΔMYO5B) mice. Eight to 10-week-old Lrig1^ΔMYO5B mice and control littermates received 2 mg tamoxifen intraperitoneally at day 0. Cells derived from Cre-induced cells also express YFP (Fig. 1A). GI tissues were collected 2.5, 3, 4, and 5 days after tamoxifen administration for histological characterization.

Results: Lrig1^ΔMYO5B mice began experiencing significant weight loss 4 days after tamoxifen injection and lost 20% of their starting body weight on day 5 (Fig. 1B). At day 5, the Lrig1^ΔMYO5B mouse colon was devoid of solid feces and their small intestine contained clear luminal contents, together suggesting a watery diarrheal phenotype (Fig. 1C). Enterocytes of Lrig1^ΔMYO5B mice on day 5 contained abnormal cytoplasmic accumulation of PAS staining (Fig. 1D). Like the differentiation deficits seen in VilCre^ERT2; MYO5B^fl/fl mice, fewer goblet cells were apparent in intestinal sections from day 3 and 5 Lrig1^ΔMYO5B mice compared to controls. Crypts were elongated starting at day 3. Similarly, the PCNA-expressing epithelial region was expanded in Lrig1^ΔMYO5B mice, even reaching into the villi at day 5 (Fig. 2A). The time course of the PCNA⁺ region expansion corresponded to the wave of MYO5B-loss beginning at the base of crypts, indicating that MYO5B loss directly altered progenitor cell proliferation and differentiation. MYO5B expression was diminished in the crypts at day 3 and nearly absent from the epithelium by day 5. Sodium/glucose cotransporter 1 (SGLT1), which is present along the apical surface of mature enterocytes, was progressively diminished and mis-localized away from the apical surface after tamoxifen administration (Fig. 2B).
Conclusion: We established a novel mouse model that demonstrates progressive MYO5B loss in the intestinal epithelium, starting at the base of the crypt and reaching villi tips over the course of 5 days following Cre induction. This Lrig1^ΔMYO5B strain will allow for lineage tracing studies to better understand the effect of MYO5B loss on intestinal epithelial differentiation.

<b>Figure 1.</b> Characterizing the phenotype Lrig1<sup>ΔMYO5B</sup> mice. (A) Schematic illustrating the timeline of tamoxifen (TMX) administration and tissue collection. Day 3 represents a loss of MYO5B in the crypt, corresponding to YFP+ cells restricted to the crypt region. Day 5 represents a loss of MYO5B throughout the crypt-villus axis, corresponding to YFP<sup>+</sup> cells throughout the intestinal epithelium. (B) Body weight changes over the course of 5 days following tamoxifen administration in control and Lrig1<sup>ΔMYO5B</sup> mice. Data points were compared using a 2-way ANOVA with Tukey’s multiple comparisons test, **** p < 0001. (C) GI tract from control and Lrig1<sup>ΔMYO5B</sup> mice on day 5. Scale bar = 1 cm. (D) Alcian blue and Periodic-acid Schiff (PAS) staining of the duodenum from control and Lrig1<sup>ΔMYO5B</sup> mice at day 3 and 5. Scale bar = 50 µm.

Figure 1. Characterizing the phenotype Lrig1^ΔMYO5B mice. (A) Schematic illustrating the timeline of tamoxifen (TMX) administration and tissue collection. Day 3 represents a loss of MYO5B in the crypt, corresponding to YFP+ cells restricted to the crypt region. Day 5 represents a loss of MYO5B throughout the crypt-villus axis, corresponding to YFP⁺ cells throughout the intestinal epithelium. (B) Body weight changes over the course of 5 days following tamoxifen administration in control and Lrig1^ΔMYO5B mice. Data points were compared using a 2-way ANOVA with Tukey’s multiple comparisons test, **** p < 0001. (C) GI tract from control and Lrig1^ΔMYO5B mice on day 5. Scale bar = 1 cm. (D) Alcian blue and Periodic-acid Schiff (PAS) staining of the duodenum from control and Lrig1^ΔMYO5B mice at day 3 and 5. Scale bar = 50 µm.

<b>Figure 2.</b> Progressive elongation of the crypt and loss of apical nutrient transporters after tamoxifen treatment in Lrig1<sup>ΔMYO5B</sup> mice. (A) Immunofluorescence images of duodenum from control and Lrig1<sup>ΔMYO5B</sup> mice at days 3 and 5 stained for MYO5B (red), PCNA (magenta), ACTG1 (white), and YFP (green). (B) Immunofluorescence images for SGLT1 (magenta), YFP (green), and ACTG1 (white) in duodenum from control and Lrig1<sup>ΔMYO5B</sup> mice at 2.5 through 5 days after tamoxifen administration. Sections stained. Scale bar = 50 µm

Figure 2. Progressive elongation of the crypt and loss of apical nutrient transporters after tamoxifen treatment in Lrig1^ΔMYO5B mice. (A) Immunofluorescence images of duodenum from control and Lrig1^ΔMYO5B mice at days 3 and 5 stained for MYO5B (red), PCNA (magenta), ACTG1 (white), and YFP (green). (B) Immunofluorescence images for SGLT1 (magenta), YFP (green), and ACTG1 (white) in duodenum from control and Lrig1^ΔMYO5B mice at 2.5 through 5 days after tamoxifen administration. Sections stained. Scale bar = 50 µm

Background:. TGF-β induces multiple transitions in hepatic stellate cells (HSC) including mitochondrial biogenesis, and transcriptional changes. However, the connection between these mitochondrial signals and fibrogenic activation of HSC are not clearly established. In the current study we have demonstrated a novel pathway involving TGF-β induced HSC activation via mitochondrial DNA release and activation of the cGAS-STING-IRF3 pathway. Methods: Hepatic stellate cells were isolated from C57BL6 mice and LX2 cells (human HSC line) and were subjected to 5ng/ml TGF-β for 16h and were assayed for transdiffertation. Inhibitors such as VBIT4 10μM (VDAC inhibitor), STING inhibitor (IFM-4490 at 20μM) were used. A mouse model of high fat/fibrogenic diet (Choline deficient – L-amino acid sufficient) for 6 weeks with/without STING antagonist 5mg/kg. Results: TGF-β stimulus in HSC resulted in phospho-STING (S366) localization to Golgi (p<0.001) and phospho-IRF3 nuclear translocation and expression of α-smooth muscle actin. Knockdown of STING in LX2 cells significantly suppressed the above changes. Live cell imaging of LX2 cells expressing TFAM-eGFP and TOM20-mPlum with super resolution microscopy showed DNA release following 8h of stimulation with TGF-β. We have termed the cGAS binding to DNA on TOM20 structures as “mtDNA cap”. Similar mitochondrial membrane protrusions were also observed in electron micrographs of TGFβ stimulated LX2. Depletion of mtDNA in LX2 abrogated TGF-β induced HSC activation. Blocking Ca²⁺ flux in LX2 mitochondria resulted in suppression of TGFβ induced activation. Immunofluorescence and proximity ligation assay confirmed that mtDNA is released via VDAC3 upon TGF-β stimulus. Furthermore, VBIT4 significantly suppressed the ECM gene signature upon TGFβ stimulus further providing evidence for VDAC3 mediated mtDNA release. Pretreatment of LX2 with a STING inhibitor significantly suppressed TGF-β induced IRF3 nuclear translocation as well as ACTA2 transcription. Moreover, RNA sequence analysis of LX2 cells subjected to TGF-β stimulus with or without STING inhibitor shows significant suppression of gene signatures involved in ECM production and HSC differentiation. Mice treated with STING antagonist and subjected to high fat – CDAA diet were significantly protected from steatosis induced hepatic fibrosis (n=8; P<0.001). Conclusion: TGF-β signaling in HSC results in mtDNA release via VDCA3 onto the outer mitochondrial membrane. Cytosolic cGAS binds this mtDNA to form a “mtDNA-cap” structure which activates the STING-IRF3 axis. IRF3 further mediates transactivation of a panel of genes leading to HSC differentiation and fibrosis progression. These findings provide novel insights into fibrogenic activation of HSCs and potential therapeutic targets for hepatic fibrosis.

<b>Figure1.</b> <b>A.</b> Super resolution microscopy of LX2 cells stimulated with TGF-β 5ng/ml, showing cGAS (green) bound to DNA (yellow) on the surface of mitochondria (TOM20 – red). The associated graph shows significant increase in cGAS-DNA-TOM20 colocalization in TGF-β stimulated LX2 cells as compared to control. <b>B. </b>Graphical representation of overall events on the hepatic stellate cells upon TGF-β signaling. In summary, TGF-β triggers DNA release via VDAC channels. cGAS is recruited to DNA release sites and forms a mtDNA CAP which leads to cGAS activation and subsequent STING-TBK1-IRF3 activation. Activated IRF3 leads to transactivation of HSC to myofibroblasts with extensive production of ECM proteins.

Figure1. A. Super resolution microscopy of LX2 cells stimulated with TGF-β 5ng/ml, showing cGAS (green) bound to DNA (yellow) on the surface of mitochondria (TOM20 – red). The associated graph shows significant increase in cGAS-DNA-TOM20 colocalization in TGF-β stimulated LX2 cells as compared to control. B. Graphical representation of overall events on the hepatic stellate cells upon TGF-β signaling. In summary, TGF-β triggers DNA release via VDAC channels. cGAS is recruited to DNA release sites and forms a mtDNA CAP which leads to cGAS activation and subsequent STING-TBK1-IRF3 activation. Activated IRF3 leads to transactivation of HSC to myofibroblasts with extensive production of ECM proteins.

Background and aims: Intestinal tuft cells initiate type II immune responses to parasitic helminth infections. Low tuft cell numbers are associated with the pathology of ulcerative colitis and Crohn’s disease. Understanding the mechanisms regulating human tuft cell differentiation could provide insights into mechanisms underlying tuft cell hyperplasia during worm clearance and inflammatory bowel disease etiology. In neurons multiple Sry-box (SOX) transcription factors are expressed in a temporally coordinated manner to direct lineage specification. Previous work by our group suggests that tuft and enteroendocrine cell subsets (EECs) arise from a common progenitor that expresses high levels of SOX4. In mice, mature tuft cells and rare EECs in the villi express high levels of SOX9 but low or unappreciable levels of SOX4. We hypothesize that human tuft cell differentiation is coordinated by a SOX factor relay, where SOX4 is transiently upregulated to specify tuft/EEC progenitors then SOX9 is upregulated to predominantly specify the tuft lineage.
Methods: SOX4 and SOX9 protein and mRNA levels were assessed in human intestinal epithelium via immunofluorescence and single-cell RNA-sequencing (scRNA-seq) respectively. A doxycycline inducible SOX4-overexpression cell line was generated in two dimensional (2D) monolayers of primary human intestinal stem cells (hISCs). qPCR was performed on control and SOX4 overexpressing hISCs to assess the impact of SOX4 on tuft and EEC differentiation.
Results: scRNA-seq revealed that human secretory progenitors express high levels of SOX4 (SOX4^hi) and low levels of SOX9 (SOX9^lo). The opposite was observed for tuft cells (SOX4^lo/SOX9^hi). Cells expressing high SOX4 and high SOX9 are mutually exclusive. qPCR revealed that SOX4 overexpression significantly inhibits SOX9. SOX4 overexpression significantly inhibits the master regulator of tuft cell differentiation POU2F3. SOX4 overexpression significantly upregulates EEC markers CHGA and CHGB but surprisingly not the terminal EEC specifier NEUROD1.
Conclusions: Inhibition of tuft and EEC transcription factors by high SOX4 is consistent with a model where SOX4 primes tuft and EEC secretory progenitors but is not sufficient to specify the tuft or EEC lineage. Our prior work demonstrates CHGA marks cycling EEC progenitors. Thus, our finding that SOX4 overexpression significantly upregulates CHGA and CHGB, could be due to an increase in the number of EEC progenitors rather than EEC maturation.

Ismael Gomez Martinez

Appearances