Yes-Associated Protein Drives Helicobacter pylori–Induced Metaplastic Changes in Gastric Epithelium Authors: Byeong Min Yu, So Dam Lee, Bo Ram Hwang, Haengdueng Jeong, Jiseon Kim, Seokyoung Hwang, Minsoo Noh, Ki Taek Nam, Yong Chan Lee Published: 2026 ABSTRACT Background & AimsHelicobactor pylori infection triggers a spectrum of metaplastic changes in the gastric epithelium; however, the upstream regulatory mechanisms remain unclear. In human gastric epithelial models, H pylori cytotoxin-associated gene A–dependent signaling stabilized and activated yes-associated protein, thereby promoting metaplastic reprogramming of the gastric epithelium.MethodsHuman gastric tissue specimens were obtained from patients at Severance Hospital, Yonsei University College of Medicine. Human gastric epithelial cell lines (AGS, MKN74, N87, KATO III) and 8- to 10-week-old C57BL/6J mice were used. Transcriptomic analysis, immunoprecipitation, histologic and immunofluorescence staining, and molecular assays (Western blotting and quantitative reverse-transcription polymerase chain reaction) were performed to assess yes-associated protein signaling and metaplastic marker expression.ResultsGenetic and pharmacologic modulation of yes-associated protein and OTU deubiquitinase, ubiquitin aldehyde binding 2 demonstrated that these metaplastic marker alterations are yes-associated protein–dependent. In vivo, comparative infection with H pylori with or without cytotoxin-associated gene A revealed significantly increased spasmolytic polypeptide–expressing metaplasia markers from chief cells (Griffonia simplicifolia II lectin/gastric intrinsic factor, CD44 variant 9), parietal cell loss, and an expanded proliferative region in the presence of cytotoxin-associated gene A. Helicobacter felis infection was used to broaden the dynamic range of in vivo pathologic outcomes. Manipulating yes-associated protein specifically in chief cells with a muscle, intestine, stomach expression 1 driver verified both the induction and suppression of these lesions, and the administration of recombinant OTU deubiquitinase, ubiquitin aldehyde binding 2 and treatment with selective inhibitors enhanced and reduced metaplastic signaling, respectively.ConclusionsYes-associated protein acts as a central regulator of H pylori–induced metaplastic transformation by reprogramming gastric epithelial and chief cell lineages. OTU deubiquitinase, ubiquitin aldehyde binding 2 serves as an upstream modulator that reinforces yes-associated protein stability and signaling, highlighting yes-associated protein as a potential therapeutic target for preventing metaplastic progression. SummaryHelicobacter pylori promotes gastric metaplastic transformation through CagA-dependent activation of YAP signaling. Genetic and pharmacologic studies identify OTUB2 as an upstream regulator of YAP and a potential target for preventing metaplastic progression. What You Need to KnowBackgroundHelicobacter pylori infection induces gastric metaplasia, but upstream regulators remain unclear. Yes-associated protein signaling has emerged as a key pathway linking bacterial virulence factors to epithelial reprogramming.ImpactThis study identifies OTU deubiquitinase, ubiquitin aldehyde binding 2–mediated yes-associated protein stabilization as a central driver of H pylori–induced gastric metaplasia, linking cytotoxin-associated gene A signaling to chief cell reprogramming and disease progression.Future DirectionsFuture studies should validate yes-associated protein activity and downstream targets in human gastric metaplastic and neoplastic tissues and explore pharmacologic inhibition to prevent metaplasia progression. Summary Helicobacter pylori promotes gastric metaplastic transformation through CagA-dependent activation of YAP signaling. Genetic and pharmacologic studies identify OTUB2 as an upstream regulator of YAP and a potential target for preventing metaplastic progression. What You Need to Know Background Helicobacter pylori infection induces gastric metaplasia, but upstream regulators remain unclear. Yes-associated protein signaling has emerged as a key pathway linking bacterial virulence factors to epithelial reprogramming. Impact This study identifies OTU deubiquitinase, ubiquitin aldehyde binding 2–mediated yes-associated protein stabilization as a central driver of H pylori–induced gastric metaplasia, linking cytotoxin-associated gene A signaling to chief cell reprogramming and disease progression. Future Directions Future studies should validate yes-associated protein activity and downstream targets in human gastric metaplastic and neoplastic tissues and explore pharmacologic inhibition to prevent metaplasia progression. Helicobacter pylori infection is a canonical driver of multistep gastric pathology, progressing from chronic gastritis through metaplasia and dysplasia to gastric carcinoma.1, 2, 3, 4, 5 The major H pylori virulence factor cytotoxin-associated gene A (CagA) perturbs host signaling pathways, disrupts epithelial junctional integrity, and promotes a tumor-permissive inflammatory microenvironment.6, 7, 8, 9, 10, 11 Although these early epithelial alterations are widely recognized as initiating events in gastric carcinogenesis, the molecular mechanism through which bacterial virulence signals are translated into epithelial reprogramming remain incompletely defined.12,13 Yes-associated protein (YAP) is a transcriptional coactivator that integrates cues from cell adhesion, polarity, and mechanical signaling to engage TEA domain transcription factor (TEAD)-dependent transcriptional programs.14, 15, 16, 17 In the stomach, YAP activation and metaplastic remodeling are frequently observed in gastric mucosa from patients with chronic H pylori infection and are considered early events preceding neoplastic transformation.18, 19, 20, 21, 22 Emerging studies have positioned YAP as a central regulator of epithelial plasticity and injury-induced cell fate reprogramming.22,23 However, in gastric epithelial cells, the upstream mechanisms through which CagA-dependent signals converge on YAP activation have not been clearly established.12,13,18,20 OTU deubiquitinase, ubiquitin aldehyde binding 2 (OTUB2) is a deubiquitinase capable of stabilizing YAP through the modulation of ubiquitin signaling, raising the possibility that it acts as an upstream mediator linking microbial cues to YAP activation.24,25 Nevertheless, how OTUB2 activity is regulated during H pylori infection and whether post-translational mechanisms contribute to YAP stabilization in this context remain largely unexplored. In particular, whether small ubiquitin-like modifier (SUMO)ylation, a post-translational modification known to regulate protein function and localization, is engaged by bacterial virulence signals to modulate YAP stability has not been determined.26,27 Gastric metaplasia represents a spectrum of epithelial reprogramming states rather than a single pathologic entity.28 Human epithelial models enable precise interrogation of epithelial-intrinsic transcriptional changes, including intestinal-type programs typified by caudal type homeobox (CDX) 2 induction.29, 30, 31, 32 In contrast, in vivo systems preserve the composite tissue microenvironment and are required to capture tissue-level remodeling, such as spasmolytic polypeptide–expressing metaplasia (SPEM) arising from chief cells.33, 34, 35, 36, 37 Given species-specific differences in gastric lineage markers, integrated analyses across multiple markers and at single-cell resolution are critical for accurate interpretation of metaplastic changes.38,39 Therefore, understanding how YAP signaling integrates microbial stimuli with epithelial plasticity across these interconnected metaplastic programs is central to elucidating the early mechanisms of gastric disease progression.18,19,21,22 In this study, we investigated whether H pylori CagA activates YAP signaling through OTUB2 and whether this pathway drives epithelial junctional disruption and metaplastic reprogramming. Using complementary human epithelial models and in vivo genetic approaches, we identified a CagA–OTUB2–YAP signaling axis that links microbial infection to epithelial remodeling and established YAP as a central regulator of metaplastic transformation in gastric epithelium during infection. Results Yes-Activated Protein Activation in Human Specimens, Single-Cell Data, and Cell Lines Immunofluorescence (IF) analysis of human gastric specimens representing chronic superficial gastritis, intestinal metaplasia (IM), intestinal-type gastric cancer, and diffuse-type gastric cancer revealed a stepwise increase in YAP signaling with disease progression (Figure 1A). In contrast, the expression of the tight-junction protein zonula occludens-1 (ZO-1) progressively decreased across disease stages (Figure 1A), indicating that YAP activation accompanies the weakening of epithelial junctional integrity during the progression of gastric lesions. Consistently, the analysis of a public cohort using gene expression profiling interactive analysis demonstrated significantly higher YAP messenger RNA (mRNA) expression in tumor tissues, compared with that in matched adjacent normal tissues (Figure 1B).Figure 1Expression of YAP in human gastric lesions and cancer cell lines. (A) Representative IF staining of YAP (red) and ZO-1 (green) in human gastric lesions: CSG, IM, IGC, and DGC. Scale bars, 250 and 100 μm. (B) GEPIA comparing YAP mRNA in tumors (n = 408) vs matched adjacent normal tissues (n = 211). Mann–Whitney U test; ∗P < .05. (C) UMAP visualization of 48,670 single cells derived from paired nonmalignant and malignant gastric tissues after quality control and batch correction. Ten major clusters identified via unsupervised clustering were annotated on the basis of the expression of canonical markers. (D) UMAP distribution of single cells stratified by lesion stage. Cells are colored according to lesion stage to visualize stage-specific distributions across CSG, IM, IGC, and DGC. (E) Dot plot showing the expression of canonical marker genes across annotated cell types. Dot size indicates the percentage of cells expressing each gene, and color intensity indicates the average expression level. (F) UMAP feature plot showing YAP expression across cell populations. (G) Dot plot showing YAP expression across annotated cell types, stratified by lesion stage. Color indicates average expression level, and dot size indicates the fraction of cells expressing YAP. (H) Subclustering of chief cells based on SPEM-associated markers. Chief cells were classified into SPEM marker–negative and SPEM marker–positive populations according to the expression of AQP5 and WFDC2. Red indicates SPEM marker–positive chief cells, and gray indicates SPEM marker–negative chief cells. (I) Dot plot showing differential expression of canonical chief cell markers (PGA4 and PGA5) and SPEM-associated markers (AQP5 and WFDC2), along with YAP expression, between SPEM marker–negative and SPEM marker–positive chief cell populations. (J) UMAP feature plots showing the expression of SPEM-associated markers AQP5 and WFDC2. The inset highlights their spatial enrichment within a subset of chief cells, along with YAP expression. (K) Quantitative analysis of gene expression in SPEM marker–negative and SPEM marker–positive chief cells, confirming increased YAP, AQP5, and WFDC2 expression and decreased PGA4 and PGA5 expression in the SPEM marker-positive population. Data are presented as log (TP10k + 1); mean ± SEM. (L) Representative WB showing basal protein levels of Hippo pathway components (YAP, TAZ, and LATS2) in gastric cancer cell lines (AGS, N87, KATO III, MKN45, and MKN74), a colorectal line (Caco-2), and a renal epithelial line (293T) (n = 3). (M) WB analysis of YAP, TAZ, LATS2, and CagA expression after incubating cells with control (uninfected), H pylori 60,190 (CagA+), or ΔCagA strains (n = 3) for 5 hours. (N) Densitometry of YAP bands normalized to GAPDH (for L and M). Statistics: Data are presented as mean ± SEM. Multi-group comparisons used 1-way or 2-way ANOVA with Tukey’s post hoc test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. APQ5, aquaporin-5; CSG, chronic superficial gastritis; DGC, diffuse-type gastric cancer; GEPIA, gene expression profiling interactive analysis; IGC, intestinal-type gastric cancer; SEM, standard error of the mean; UMAP, Uniform Manifold Approximation and Projection. Expression of YAP in human gastric lesions and cancer cell lines. (A) Representative IF staining of YAP (red) and ZO-1 (green) in human gastric lesions: CSG, IM, IGC, and DGC. Scale bars, 250 and 100 μm. (B) GEPIA comparing YAP mRNA in tumors (n = 408) vs matched adjacent normal tissues (n = 211). Mann–Whitney U test; ∗P < .05. (C) UMAP visualization of 48,670 single cells derived from paired nonmalignant and malignant gastric tissues after quality control and batch correction. Ten major clusters identified via unsupervised clustering were annotated on the basis of the expression of canonical markers. (D) UMAP distribution of single cells stratified by lesion stage. Cells are colored according to lesion stage to visualize stage-specific distributions across CSG, IM, IGC, and DGC. (E) Dot plot showing the expression of canonical marker genes across annotated cell types. Dot size indicates the percentage of cells expressing each gene, and color intensity indicates the average expression level. (F) UMAP feature plot showing YAP expression across cell populations. (G) Dot plot showing YAP expression across annotated cell types, stratified by lesion stage. Color indicates average expression level, and dot size indicates the fraction of cells expressing YAP. (H) Subclustering of chief cells based on SPEM-associated markers. Chief cells were classified into SPEM marker–negative and SPEM marker–positive populations according to the expression of AQP5 and WFDC2. Red indicates SPEM marker–positive chief cells, and gray indicates SPEM marker–negative chief cells. (I) Dot plot showing differential expression of canonical chief cell markers (PGA4 and PGA5) and SPEM-associated markers (AQP5 and WFDC2), along with YAP expression, between SPEM marker–negative and SPEM marker–positive chief cell populations. (J) UMAP feature plots showing the expression of SPEM-associated markers AQP5 and WFDC2. The inset highlights their spatial enrichment within a subset of chief cells, along with YAP expression. (K) Quantitative analysis of gene expression in SPEM marker–negative and SPEM marker–positive chief cells, confirming increased YAP, AQP5, and WFDC2 expression and decreased PGA4 and PGA5 expression in the SPEM marker-positive population. Data are presented as log (TP10k + 1); mean ± SEM. (L) Representative WB showing basal protein levels of Hippo pathway components (YAP, TAZ, and LATS2) in gastric cancer cell lines (AGS, N87, KATO III, MKN45, and MKN74), a colorectal line (Caco-2), and a renal epithelial line (293T) (n = 3). (M) WB analysis of YAP, TAZ, LATS2, and CagA expression after incubating cells with control (uninfected), H pylori 60,190 (CagA+), or ΔCagA strains (n = 3) for 5 hours. (N) Densitometry of YAP bands normalized to GAPDH (for L and M). Statistics: Data are presented as mean ± SEM. Multi-group comparisons used 1-way or 2-way ANOVA with Tukey’s post hoc test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. APQ5, aquaporin-5; CSG, chronic superficial gastritis; DGC, diffuse-type gastric cancer; GEPIA, gene expression profiling interactive analysis; IGC, intestinal-type gastric cancer; SEM, standard error of the mean; UMAP, Uniform Manifold Approximation and Projection. To further define YAP expression at single-cell resolution, we integrated and reanalyzed 48,670 single-cell transcriptomes derived from nonmalignant and malignant human gastric tissues. After quality control and batch correction, cell populations were visualized using Uniform Manifold Approximation and Projection (Figure 1C). Mapping of cells by lesion stage revealed distinct shifts in cellular composition between non-neoplastic and tumor tissues (Figure 1D). Unsupervised clustering of the single-cell RNA sequencing dataset resolved 10 major cell clusters. Cluster identities were then inferred by analyzing cluster-enriched differentially expressed genes (DEGs) in conjunction with established canonical marker genes. On this basis, the clusters were annotated as endothelial cells (VWF and CD34), enteroendocrine cells (CHGA and CHGB), fibroblasts (LUM and PDGFRA), gland mucous cells (mucin [MUC] 6 and trefoil factor [TFF] 2), IM-associated epithelial cells (TFF3, CDX1, and CDX2), chief cells (pepsinogen [PGA] 4 and PGA5), myofibroblasts (ACTA2 and TAGLN), pit mucous cells (MUC5AC, TFF1, gastrokine [GKN] 1, and GKN2), parietal cells (ATPase transporting subunit H+/K+ β [ATP4B] and CBLIF), and cancer cells (CEACAM5 and CEACAM6). These annotations were further supported by dot plot visualization of representative marker genes across clusters (Figure 1E), allowing integrated assessment of marker expression intensity and detection frequencies. Because gland mucous and metaplastic cell programs are known to partially overlap, cell identities were assigned on the basis of combined canonical marker expression and overall transcriptional features rather than by any single marker alone.38 Accordingly, the gland mucous cell cluster may include cells with pyloric metaplasia–like features.36 Visualization of YAP expression across cell types showed that YAP was broadly expressed across multiple epithelial lineages of the gastric epithelium (Figure 1F). Notably, chief cells derived from cancer tissues exhibited significantly higher YAP expression, compared with chief cells from noncancerous tissues (Figure 1G). Next, we performed unsupervised subclustering of chief cells and identified SPEM marker–negative and SPEM marker–positive subpopulations based on the expression of established SPEM-associated markers (Figure 1H). The SPEM marker–positive population was characterized by elevated expression of SPEM-associated markers (APQ5 and WAP 4-disulfide core domain 2 [WFDC2]), reduced expression of canonical chief cell markers (PGA4 and PGA5), and higher YAP expression (Figure 1I and K). Consistently, Uniform Manifold Approximation and Projection visualization revealed that aquaporin-5– and WFDC2-expressing cells were enriched along a metaplastic continuum within a subset of chief cells (Figure 1J), supporting the existence of heterogeneity within the chief cell compartment. Collectively, these findings indicate a gradual expansion of metaplastic programs during gastric lesion development and suggest that YAP expression is associated with the emergence of SPEM-like features within the chief cell lineage. Additionally, we assessed the basal expression of core Hippo pathway components, including YAP, transcriptional coactivator with PDZ-binding motif (TAZ), and large tumor suppressor kinase 2 (LATS2), across a panel of human gastric cancer cell lines (Figure 1L). Cells exposed to H pylori strain 60,190 (CagA+) for 5 hours showed a marked increase in YAP level, compared with that in uninfected controls and cells infected with the ΔCagA strain (Figure 1M and N), indicating CagA-dependent activation of YAP signaling. Based on these findings, subsequent in vitro experiments were performed using AGS cells infected with H pylori 60,190 (CagA+) to elucidate the mechanism. Positive H pylori Induces Alterations in Hippo Signaling, K63-Linked Deubiquitination, and Inflammatory Gene Expression in Gastric Cancer Cells RNA-seq analysis revealed marked changes in Hippo signaling–associated genes (Gene Ontology [GO]:0035329) in AGS cells infected with H pylori 60,190 (CagA+) (Figure 2A). Notably, the transcript levels of YAP, WWTR1 (TAZ), LATS2, SAV1, STK4, and TEAD1 were higher in the H pylori 60,190 than in uninfected controls and ΔCagA-infected cells. Gene set enrichment analysis (GSEA) further showed significant positive enrichment of YAP signaling-related signatures in H pylori 60,190-infected cells, compared with both uninfected controls and ΔCagA-infected cells (Figure 2B).Figure 2CagA-positive Helicobacter pylori alters Hippo signaling, K63-linked deubiquitination, and inflammatory gene expression in gastric cancer cells. (A) Heatmap of genes in the Hippo pathway GO term (GO: 0035329) in AGS cells infected with H pylori 60,190 (CagA+) or ΔCagA for 5 hours. (B) GSEA plot for YAP signaling signatures comparing Control vs H pylori 60,190 (CagA+) and H pylori 60,190 (CagA+) vs ΔCagA-infected AGS cells. (C) GSEA plot for K63-linked deubiquitination comparing control vs H pylori 60,190 (CagA+) and H pylori 60,190 (CagA+) vs ΔCagA-infected AGS cells. (D) GSEA plot showing positive enrichment of the CXCR chemokine receptor binding gene set comparing control vs H pylori 60,190 (CagA+) and H pylori 60,190 (CagA+) vs ΔCagA-infected AGS cells. (E) ELISA of IL-8 production in AGS cells infected with 60,190 or ΔCagA (n = 4 per group). (F) Representative WB showing time-course of YAP and TAZ after infection with 60,190 or ΔCagA (n = 3). (G) Representative WB showing the induction of YAP/TAZ and OTUB2 after the overexpression of CagA WT or the EPIYA-PR mutant (n = 3). Statistics: Data are presented as mean ± SEM. Group comparisons used 2-way ANOVA with Tukey’s post hoc test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. CXCR, C-X-C chemokine receptor; PR, phosphorylation-resistant; SEM, standard error of the mean. CagA-positive Helicobacter pylori alters Hippo signaling, K63-linked deubiquitination, and inflammatory gene expression in gastric cancer cells. (A) Heatmap of genes in the Hippo pathway GO term (GO: 0035329) in AGS cells infected with H pylori 60,190 (CagA+) or ΔCagA for 5 hours. (B) GSEA plot for YAP signaling signatures comparing Control vs H pylori 60,190 (CagA+) and H pylori 60,190 (CagA+) vs ΔCagA-infected AGS cells. (C) GSEA plot for K63-linked deubiquitination comparing control vs H pylori 60,190 (CagA+) and H pylori 60,190 (CagA+) vs ΔCagA-infected AGS cells. (D) GSEA plot showing positive enrichment of the CXCR chemokine receptor binding gene set comparing control vs H pylori 60,190 (CagA+) and H pylori 60,190 (CagA+) vs ΔCagA-infected AGS cells. (E) ELISA of IL-8 production in AGS cells infected with 60,190 or ΔCagA (n = 4 per group). (F) Representative WB showing time-course of YAP and TAZ after infection with 60,190 or ΔCagA (n = 3). (G) Representative WB showing the induction of YAP/TAZ and OTUB2 after the overexpression of CagA WT or the EPIYA-PR mutant (n = 3). Statistics: Data are presented as mean ± SEM. Group comparisons used 2-way ANOVA with Tukey’s post hoc test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. CXCR, C-X-C chemokine receptor; PR, phosphorylation-resistant; SEM, standard error of the mean. The K63-linked deubiquitination gene set displayed a higher enrichment score in H pylori 60,190–infected cells in both comparisons (control vs 60,190 and 60,190 vs ΔCagA), and OTUB2 was concomitantly upregulated at the transcript level under the same condition (Figure 2C). C-X-C chemokine receptor chemokine receptor binding gene sets were positively enriched after H pylori 60,190 infection, relative to both uninfected controls and ΔCagA-infected cells, and chemokine C-X-C motif ligand 8 (CXCL8; interleukin-8 [IL-8]) expression increased (Figure 2D). Consistent with the transcriptomic data, enzyme-linked immunosorbent assay (ELISA) at 8 hours postinfection showed significantly higher IL-8 secretion in H pylori 60,190–infected cells than in uninfected or ΔCagA-infected cells (Figure 2E). Time-course immunoblotting revealed the induction of YAP and TAZ proteins, with YAP peaking at approximately 5 hours postinfection (Figure 2F). Moreover, plasmid-mediated CagA overexpression (wild-type [WT] and phosphorylation-resistant mutant) increased the levels of YAP, TAZ, and the deubiquitinase OTUB2 (Figure 2G). Collectively, these findings indicate that CagA-dependent signaling converges on the Hippo–YAP signaling, ubiquitin regulatory circuits, and inflammatory responses, motivating mechanistic interrogation of the OTUB2-regulated YAP signaling as an upstream node. Linking Yes-Associated Protein Nuclear Translocation to Tight-Junction Loss and Intestinal-Type Metaplastic Change To determine whether H pylori–induced YAP activation involves nuclear translocation, we performed nuclear–cytoplasmic fractionation in AGS cells. Nuclear YAP levels were markedly increased at 5 hours postinfection with H pylori 60,190 (CagA+), compared with control and ΔCagA strain (Figure 3A). Consistent with this finding, quantitative fluorescence imaging revealed a prominent accumulation of YAP within the nuclei following 5 hours of H pylori 60,190 infection (Figure 3B).Figure 3Tight-junction disruption and IM are associated with nuclear translocation of YAP. (A) Representative WB of nuclear and cytoplasmic fractions showing YAP and phospho-YAP in AGS cells infected with H pylori 60,190 or ΔCagA for 5 hours (n = 3). (B) IF staining of YAP and CagA in AGS cells fixed 5 hours after H pylori 60,190 infection. Nuclear translocation of YAP was confirmed by fluorescence intensity analysis. Scale bar, 50 μm. (C) Representative WB of ZO-1 and CDX2 in AGS cells infected with H pylori 60,190 or ΔCagA (n = 5). (D) Representative IF images of ZO-1 (green) and CDX2 (red) in AGS cells incubated with H pylori 60,190 or ΔCagA for 5 hours; scale bar, 50 μm. (E) RT-PCR analysis of ZO-1 and CDX2 mRNA after infection with 60,190 or ΔCagA (n = 4). (F) TEAD luciferase reporter activity with H pylori infection and with YAP knockdown (n = 3). (G) IL-8 production in siYAP-treated AGS cells with or without infection (n = 5). (H) Representative WB of ZO-1 and CDX2 in YAP-silenced AGS cells (n = 3). (I) Representative IF images of ZO-1 (red), CDX2 (red), and YAP (green) in YAP-silenced AGS cells incubated with H pylori 60,190 for 5 hours. Scale bar, 100 μm. (J) In vivo xenograft tumor growth of AGS cells following YAP depletion. Top left: Longitudinal measurement of tumor volume in nude mice injected with pLKO.1 control, shYAP#1, or shYAP#2 AGS cells over a 60-day period. Top right: Representative images of excised xenograft tumors at the experimental endpoint. Bottom left: Representative bioluminescence images of tumor-bearing mice injected with pLKO.1 control or shYAP#1- or shYAP#2-expressing AGS cells. Bottom right: Quantification of bioluminescence signal shown as average radiance (photons/sec/cm2/sr). n = 5 mice per group. Scale bar, 10 mm. (K) Representative images of normal and gastric cancer organoids after siYAP or YAP overexpression through electroporation; scale bars, 200 μm (top), 100 μm (bottom). The bar graph shows the proportion of single-layered and multilumen organoids (n = 5 independent organoid lines). (L) Transmission electron microscopy/scanning electron microscopy and IF analysis of ALI-cultured 2D organoids infected with 60,190 for 48 hours (ZO-1, H pylori, YAP, CagA); scale bars, 1000 nm (transmission electron microscopy), 10 μm (scanning electron microscopy), 50 μm (IF). Statistics: Data are presented as mean ± SEM. Multi-group/factorial data were analyzed by 1-way or 2-way ANOVA with Tukey’s post hoc test; 2-group comparisons used 2-tailed unpaired Student’s t tests. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. ALI, air–liquid interface. SEM, standard error of the mean. Tight-junction disruption and IM are associated with nuclear translocation of YAP. (A) Representative WB of nuclear and cytoplasmic fractions showing YAP and phospho-YAP in AGS cells infected with H pylori 60,190 or ΔCagA for 5 hours (n = 3). (B) IF staining of YAP and CagA in AGS cells fixed 5 hours after H pylori 60,190 infection. Nuclear translocation of YAP was confirmed by fluorescence intensity analysis. Scale bar, 50 μm. (C) Representative WB of ZO-1 and CDX2 in AGS cells infected with H pylori 60,190 or ΔCagA (n = 5). (D) Representative IF images of ZO-1 (green) and CDX2 (red) in AGS cells incubated with H pylori 60,190 or ΔCagA for 5 hours; scale bar, 50 μm. (E) RT-PCR analysis of ZO-1 and CDX2 mRNA after infection with 60,190 or ΔCagA (n = 4). (F) TEAD luciferase reporter activity with H pylori infection and with YAP knockdown (n = 3). (G) IL-8 production in siYAP-treated AGS cells with or without infection (n = 5). (H) Representative WB of ZO-1 and CDX2 in YAP-silenced AGS cells (n = 3). (I) Representative IF images of ZO-1 (red), CDX2 (red), and YAP (green) in YAP-silenced AGS cells incubated with H pylori 60,190 for 5 hours. Scale bar, 100 μm. (J) In vivo xenograft tumor growth of AGS cells following YAP depletion. Top left: Longitudinal measurement of tumor volume in nude mice injected with pLKO.1 control, shYAP#1, or shYAP#2 AGS cells over a 60-day period. Top right: Representative images of excised xenograft tumors at the experimental endpoint. Bottom left: Representative bioluminescence images of tumor-bearing mice injected with pLKO.1 control or shYAP#1- or shYAP#2-expressing AGS cells. Bottom right: Quantification of bioluminescence signal shown as average radiance (photons/sec/cm2/sr). n = 5 mice per group. Scale bar, 10 mm. (K) Representative images of normal and gastric cancer organoids after siYAP or YAP overexpression through electroporation; scale bars, 200 μm (top), 100 μm (bottom). The bar graph shows the proportion of single-layered and multilumen organoids (n = 5 independent organoid lines). (L) Transmission electron microscopy/scanning electron microscopy and IF analysis of ALI-cultured 2D organoids infected with 60,190 for 48 hours (ZO-1, H pylori, YAP, CagA); scale bars, 1000 nm (transmission electron microscopy), 10 μm (scanning electron microscopy), 50 μm (IF). Statistics: Data are presented as mean ± SEM. Multi-group/factorial data were analyzed by 1-way or 2-way ANOVA with Tukey’s post hoc test; 2-group comparisons used 2-tailed unpaired Student’s t tests. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. ALI, air–liquid interface. SEM, standard error of the mean. To assess whether nuclear YAP accumulation correlates with epithelial barrier integrity and intestinal differentiation, we evaluated the expression of ZO-1 and CDX2. Immunoblotting and IF analyses revealed a decrease in ZO-1, a tight-junction protein, and a concomitant increase in CDX2, a representative marker of intestinal metaplasia (Figure 3C and D). These changes were consistently observed at the transcript level, as reverse-transcription polymerase chain reaction (RT-PCR) showed reduced ZO-1 mRNA and elevated CDX2 mRNA, following H pylori 60,190 infection (Figure 3E). Next, we examined whether YAP activation results in functional transcriptional output. TEAD luciferase reporter activity increased following H pylori 60,190 infection. However, this induction was abolished in YAP-depleted cells (shYAP#1/#2), indicating that transcriptional activation is YAP-dependent (Figure 3F). Similarly, siYAP-treated cells exhibited reduced IL-8 production (Figure 3G), and YAP depletion blunted the infection-induced decrease in ZO-1 expression and increase in CDX2 expression (Figure 3H and I). These findings support a causal role for YAP in regulating junctional integrity and metaplastic gene programs. Based on these in vitro observations, we evaluated whether YAP contributes to proliferative potential in vivo. In an AGS xenograft model, tumors derived from YAP-depleted cells (shYAP#1 or shYAP#2-mediated) exhibited reduced tumor growth. Bioluminescence imaging revealed significantly lower radiant efficiency in tumors formed by YAP-depleted cells, compared with pLKO.1 control tumors, indicating that YAP depletion suppresses tumor growth in vivo (Figure 3J). To investigate the structural role of YAP in epithelial morphogenesis, YAP expression was modulated by electroporation in normal gastric organoids. YAP suppression markedly impaired organoid formation and growth, suggesting that an appropriate level of YAP activity is required for the maintenance of epithelial architecture. In contrast, YAP overexpression induced nuclear stratification and the formation of multiple lumens, resulting in an intestinalized morphology. This phenotype closely resembled that of patient-derived intestinal-type gastric cancer organoids—characterized by multilayered nuclei and multiple lumens—but was clearly distinct from the compact, poorly lumenized morphology observed in diffuse-type gastric cancer organoids (Figure 3K). Finally, to visualize infection-induced structural disruption at high resolution, 2D organoids cultured under air–liquid interface conditions were exposed to H pylori 60,190 for 48 hours. This resulted in junctional rupture, expansion of intercellular spaces, and colocalization of YAP and CagA, as confirmed by transmission electron microscopy, scanning electron microscopy, and IF analysis (Figure 3L). Collectively, these results demonstrate that H pylori–induced YAP nuclear translocation compromises epithelial tight-junction integrity and promotes intestinal-type metaplastic reprogramming. –Mediated Yes-Associated Protein Activation via Deubiquitination and SUMOylation Drives Metaplastic Reprogramming Given that YAP stability is regulated by post-translational ubiquitin modification, we next examined transcriptional changes in deubiquitination-related genes to identify potential upstream modulators of YAP under H pylori infection. Transcriptome analysis showed changes in the K63-linked deubiquitination gene set and highlighted OTUB2 (Figure 4A). By qRT-PCR, OTUB2 expression was significantly increased in cells infected with H pylori 60,190 (CagA+), compared with both uninfected controls and ΔCagA-infected cells (Figure 4B), and IF demonstrated colocalization of OTUB2 with CagA (Figure 4C).Figure 4OTUB2-mediated YAP stabilization through deubiquitination and SUMOylation. (A) Scatter and heatmap of genes in K63-linked deubiquitination (GO:0070536) from RNA-seq in AGS cells infected with H pylori 60,190; OTUB2 highlighted. (B) qRT-PCR for OTUB2 in AGS cells infected with 60,190 or ΔCagA (GAPDH-normalized; n = 4). (C) Representative IF images for OTUB2 (green) and CagA (red) after 5-hour infection in AGS cells; scale bar, 50 μm. (D) Representative IF images for YAP (green), CDX2 (red), and ZO-1 (red) in AGS cells after recombinant OTUB2 (16 μg/mL) treatment; scale bar, 50 μm. (E) Representative WB for YAP, CDX2, and ZO-1 in AGS cells after rOTUB2 (16 μg/mL) (n = 3). (F) Representative WB for OTUB2 and YAP in AGS cells after siOTUB2 pretreatment followed by H pylori infection (n = 3). (G) Representative IP–WB detecting YAP ubiquitination in AGS cells (n = 3). (H) Representative WB analysis of SUMO2/3 signals in total cell lysates and OTUB2 immunoprecipitates from AGS cells to assess OTUB2 SUMOylation under H pylori infection (n = 3). (I) Representative WB showing the effect of the SUMO-activating enzyme inhibitor TAK-981 on OTUB2 SUMOylation and its association with YAP in AGS cells (n = 3). (J) Dot plot of OTUB2 expression across gastric epithelial cell types and pathologic stages. (K) OTUB2 immunohistochemistry in human CSG, IM, IGC, and DGC samples; scale bar, 100 μm. (L) Schematic of CagA-driven activation of YAP signaling promoting metaplastic changes in the gastric epithelium. Statistics: Data are presented as mean ± SEM. One-way or 2-way ANOVA with Tukey’s post hoc test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. CSG, chronic superficial gastritis; DGC, diffuse-type gastric cancer; IGC, intestinal-type gastric cancer; SEM, standard error of the mean. OTUB2-mediated YAP stabilization through deubiquitination and SUMOylation. (A) Scatter and heatmap of genes in K63-linked deubiquitination (GO:0070536) from RNA-seq in AGS cells infected with H pylori 60,190; OTUB2 highlighted. (B) qRT-PCR for OTUB2 in AGS cells infected with 60,190 or ΔCagA (GAPDH-normalized; n = 4). (C) Representative IF images for OTUB2 (green) and CagA (red) after 5-hour infection in AGS cells; scale bar, 50 μm. (D) Representative IF images for YAP (green), CDX2 (red), and ZO-1 (red) in AGS cells after recombinant OTUB2 (16 μg/mL) treatment; scale bar, 50 μm. (E) Representative WB for YAP, CDX2, and ZO-1 in AGS cells after rOTUB2 (16 μg/mL) (n = 3). (F) Representative WB for OTUB2 and YAP in AGS cells after siOTUB2 pretreatment followed by H pylori infection (n = 3). (G) Representative IP–WB detecting YAP ubiquitination in AGS cells (n = 3). (H) Representative WB analysis of SUMO2/3 signals in total cell lysates and OTUB2 immunoprecipitates from AGS cells to assess OTUB2 SUMOylation under H pylori infection (n = 3). (I) Representative WB showing the effect of the SUMO-activating enzyme inhibitor TAK-981 on OTUB2 SUMOylation and its association with YAP in AGS cells (n = 3). (J) Dot plot of OTUB2 expression across gastric epithelial cell types and pathologic stages. (K) OTUB2 immunohistochemistry in human CSG, IM, IGC, and DGC samples; scale bar, 100 μm. (L) Schematic of CagA-driven activation of YAP signaling promoting metaplastic changes in the gastric epithelium. Statistics: Data are presented as mean ± SEM. One-way or 2-way ANOVA with Tukey’s post hoc test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. CSG, chronic superficial gastritis; DGC, diffuse-type gastric cancer; IGC, intestinal-type gastric cancer; SEM, standard error of the mean. Treatment with recombinant OTUB2 (rOTUB2) alone elevated YAP and CDX2 levels while reducing ZO-1 levels (Figure 4D and E), whereas siOTUB2 pretreatment markedly attenuated the infection-induced increase in YAP levels. These findings suggest that OTUB2 is involved in the process of YAP stabilization induced by H pylori infection (Figure 4F). Immunoprecipitation (IP) of YAP followed by immunoblotting for ubiquitin detected a high-molecular–weight smear of polyubiquitinated YAP that varied across conditions. The smear intensity was lower with H pylori 60,190 infection than with no infection or ΔCagA, indicating that YAP ubiquitination is selectively suppressed under H pylori 60,190 infection (Figure 4G). IP of OTUB2 and blotting with anti-SUMO2/3 identified higher-molecular–weight OTUB2 species consistent with SUMOylated OTUB2, which increased with H pylori 60,190 infection (Figure 4H). Upon treatment with TAK-981, an inhibitor of the SUMO activating enzyme, these SUMOylated OTUB2 bands markedly decreased, and the coprecipitation of OTUB2 with YAP observed with H pylori 60,190 was reduced (Figure 4I). These findings indicate that OTUB2 undergoes regulated SUMO2/3 conjugation in response to H pylori infection and that this modification promotes the stability of the OTUB2–YAP complex during infection. In single-cell datasets, OTUB2 transcripts were enriched in specific epithelial compartments, with prominent signals in IM-related cells (Figure 4J). In human tissue specimens, OTUB2 expression increased across pathologic stages, underscoring clinical relevance (Figure 4K). Together, these findings indicate that OTUB2 acts as a deubiquitinase hub that channels CagA signaling to YAP stabilization and nuclear translocation (Figure 4L). Comparative Helicobacter Infection in a Chief Cell–Specific YAP Knockout Model SPEM is the metaplasia most consistently observed after Helicobacter infection in mice; thus, we prespecified SPEM as the primary histologic endpoint to test in vivo requirement of YAP signaling. As SPEM arises from the reprogramming of mature chief cells at the base of oxyntic corpus glands, our analyses focused on the corpus. We generated an inducible Mist1-creERT2Cre/+;YAPfl/fl model in adults to selectively remove YAP in chief cells (Figure 5A). Mice were then orally gavaged for 8 weeks with H pylori 60,190 (CagA+), ΔCagA, or H felis (Figure 5B), ensuring equivalent infection durations across groups. Because H pylori shows limited colonization in mice, H felis, which can achieve stable infection and robust pathology, was included as a comparator species to enable more consistent evaluation of strain-dependent tissue responses. In corpus hematoxylin and eosin (H&E) sections, WT mice developed prominent inflammation and glandular distortion after infection with H pylori 60,190 and H felis, whereas these pathologic changes were markedly attenuated in knockout (KO) mice (Figure 5C). Notably, H pylori 60,190 (CagA+) induced more pronounced lesions than its ΔCagA strain, underscoring the contribution of CagA to disease severity. Species-specific in situ hybridization (ISH) probes verified colonization in both genotypes; however, tissue damage and glandular alterations were strongly suppressed in KO animals (Figure 5D). Consistently, H&E-based inflammation scores were elevated in WT mice infected with either H pylori 60,190 or H felis, both showing significant increases relative to phosphate-buffered saline (PBS) or ΔCagA, whereas KO mice remained largely protected (Figure 5E). ISH quantification showed that WT mice infected with H pylori 60,190 exhibited higher gland-base colonization ratios than KO mice and that colonization was significantly greater following H pylori 60,190 infection, compared with the ΔCagA strain (Figure 5F). This analysis provides a framework for subsequent quantification of the SPEM pathology (parietal cell loss, expansion of the neck region, and a shift in the proliferative region) and suggests that strain virulence (CagA+) and interspecies colonization capacity (H felis) may modulate the intensity of tissue responses at the site of chief cell reprogramming.Figure 5Tissue-level changes with chief cell–specific YAP manipulation and comparative Helicobacter infection. (A) Schematic of chief cell–specific conditional deletion using Mist1-creERT2Cre/+;YAPfl/fl, with YAP loss by IHC at the gland base after tamoxifen (inset). (B) Eight-week infection design in WT (Mist1-creERT2Cre/+) and chief cell–specific YAP KO (Mist1-creERT2Cre/+;YAPfl/fl): PBS, H pylori 60,190 (CagA+), ΔCagA, or H felis. (C) Representative H&E of oxyntic corpus at 8 weeks; scale bars, 450 μm and 100 μm. (D) Representative ISH with species-specific probes for H pylori and H felis (red) in the gastric corpus, counterstained with hematoxylin; scale bar, 100 μm. (E) Inflammation scores (0–4; 0 none, 1 minimal, 2 mild, 3 moderate, 4 severe) in the corpus based on H&E (n = 5). (F) ISH-based quantification of foci density (foci/mm2) and gastric gland-base colonization (%) for H pylori and H felis (n = 4). The P value indicates comparison with the control group. Statistics: Data are presented as mean ± SEM. Two-way ANOVA with Tukey’s post hoc test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. SEM, standard error of the mean. Tissue-level changes with chief cell–specific YAP manipulation and comparative Helicobacter infection. (A) Schematic of chief cell–specific conditional deletion using Mist1-creERT2Cre/+;YAPfl/fl, with YAP loss by IHC at the gland base after tamoxifen (inset). (B) Eight-week infection design in WT (Mist1-creERT2Cre/+) and chief cell–specific YAP KO (Mist1-creERT2Cre/+;YAPfl/fl): PBS, H pylori 60,190 (CagA+), ΔCagA, or H felis. (C) Representative H&E of oxyntic corpus at 8 weeks; scale bars, 450 μm and 100 μm. (D) Representative ISH with species-specific probes for H pylori and H felis (red) in the gastric corpus, counterstained with hematoxylin; scale bar, 100 μm. (E) Inflammation scores (0–4; 0 none, 1 minimal, 2 mild, 3 moderate, 4 severe) in the corpus based on H&E (n = 5). (F) ISH-based quantification of foci density (foci/mm2) and gastric gland-base colonization (%) for H pylori and H felis (n = 4). The P value indicates comparison with the control group. Statistics: Data are presented as mean ± SEM. Two-way ANOVA with Tukey’s post hoc test. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. SEM, standard error of the mean. Yes-Associated Protein–Dependent Gastric Lineage Reprogramming Under Pathogenic Helicobacter Infection To further delineate how YAP influences epithelial hierarchy and lineage allocation in the gastric corpus, we analyzed a panel of markers representing proliferative, chief, mucous, and parietal cell compartments. These analyses aimed to determine whether YAP activation under infection shifts the balance from a differentiated to a regenerative or metaplastic state. In uninfected controls, Ki67-positive cells were confined to the neck and isthmus. Following infection, the proliferative region extended down toward the gland base, with significant expansion in H pylori 60,190 (CagA+), but not in KO mice. Conversely, the ΔCagA strain produced no appreciable change, compared with PBS controls (Figure 6A). Markers of mature chief cells (MIST1 and pepsinogen [PGC]) decreased in WT mice upon infection with H pylori 60,190, but not with the ΔCagA strain, and these decreases were attenuated in KO mice (Figure 6B and C). The surface mucous marker MUC5AC expanded into deeper glands in WT mice following infection with H pylori 60,190 but remained at baseline with the ΔCagA strain and was suppressed in KO mice (Figure 6D). Consistently, ATP4B+ parietal cells decreased in WT mice, whereas Griffonia simplicifolia II lectin (GSII) lectin binding expanded into the neck/base region, most prominently with H pylori 60,190 infection. Conversely, ΔCagA resembled PBS, and in KO mice, ATP4B was preserved with only limited GSII expansion (Figure 6E). Together, these coordinated changes indicate that YAP activation under pathogenic infection drives a broad remodeling of gastric lineage architecture—marked by loss of mature cell identity, expansion of proliferative and mucous compartments, and emergence of a reparative/metaplastic phenotype. Collectively, these lineage rearrangements were pronounced in WT mice but mitigated in chief cell–specific YAP KO mice despite identical infection stimuli, indicating that epithelial-intrinsic YAP activity amplifies tissue remodeling and that CagA is required for H pylori–driven changes. Notably, infection with H felis, a well-established murine-adapted Helicobacter species that robustly induces corpus pathology, consistently recapitulated these lineage alterations, thereby serving as a positive control confirming the YAP-dependent nature of pathogenic gastric remodeling in vivo.Figure 6Gastric lineage marker rearrangement in the corpus of WT and chief cell–specific YAP KO mice infected with H pylori 60,190 (CagA+), ΔCagA, or H felis. (A–D) Representative DAB IHC staining in the gastric corpus for (A) Ki-67 (proliferation), (B) MIST1 (chief cells), (C) PGC (chief cells), (D) MUC5AC (surface mucous cells); scale bar, 100 μm. (E) IF for ATP4B (parietal cells, red) and GSII lectin (neck mucous cells, green); scale bar, 100 μm. Quantifications are shown below each panel. Statistics: Data are presented as mean ± SEM. Two-way ANOVA with Tukey’s post hoc test for genotype × infection factorial comparisons (n = 3). ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. SEM, standard error of the mean. Gastric lineage marker rearrangement in the corpus of WT and chief cell–specific YAP KO mice infected with H pylori 60,190 (CagA+), ΔCagA, or H felis. (A–D) Representative DAB IHC staining in the gastric corpus for (A) Ki-67 (proliferation), (B) MIST1 (chief cells), (C) PGC (chief cells), (D) MUC5AC (surface mucous cells); scale bar, 100 μm. (E) IF for ATP4B (parietal cells, red) and GSII lectin (neck mucous cells, green); scale bar, 100 μm. Quantifications are shown below each panel. Statistics: Data are presented as mean ± SEM. Two-way ANOVA with Tukey’s post hoc test for genotype × infection factorial comparisons (n = 3). ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. SEM, standard error of the mean. Requirement of Yes-Associated Protein for Chief Cell–Derived Metaplasia (Spasmolytic Polypeptide–Expressing Metaplasia) As SPEM represents a prototypical injury-induced lineage reprogramming that arises from mature chief cells, we tested whether YAP activity within these cells is required for this transition. SPEM serves as a key intermediate state that bridges parietal cell loss and the emergence of intestinal-type lineages, reflecting an early adaptive response that can progress toward preneoplastic remodeling. Therefore, a chief cell–specific YAP KO model provides a direct means to determine whether YAP is indispensable for initiating this regenerative-to-metaplastic switch under chronic Helicobacter infection. We evaluated GSII–gastric intrinsic factor (GIF) double positivity (a hallmark of SPEM), CD44 variant 9 (CD44v9; damage/repair–metaplasia marker), and TFF2 (a trefoil factor expressed by mucous neck and SPEM cells). In WT mice, infection with H pylori 60,190 (CagA+) and with H felis significantly increased GSII+/GIF+ cells as a fraction of total GIF+ chief cells (Figure 7A). Coexpression of YAP and CD44v9 also increased significantly, supporting the activation of chief cell–derived metaplastic change (Figure 7B). Similarly, TFF2 levels significantly increased under the same conditions, reinforcing the histologic consistency of SPEM (Figure 7C). Conversely, infection with the ΔCagA strain produced results comparable to PBS controls, with no significant induction of SPEM markers. In the chief cell–specific YAP knockout, GSII–GIF double positivity, CD44v9, and TFF2 levels remained near baseline, and infection-induced increases were significantly blunted (Figure 6A–C). This attenuation was accompanied by the preservation of glandular morphology and a restricted expansion of the proliferative zone, indicating that YAP loss constrains both structural and transcriptional aspects of metaplastic remodeling. This indicates a genotype-by-infection interaction and demonstrates that YAP is required to initiate chief cell reprogramming. Collectively, these findings reaffirm that the SPEM program triggered by H pylori infection is YAP-dependent.Figure 7Chief cell–specific YAP KO suppresses SPEM in the oxyntic corpus during H pylori and H felis infection. SPEM was assessed based on GSII–GIF double positivity and the expression of established metaplasia-associated markers, including CD44v9 and TFF2. (A) Representative IF for GSII and GIF in WT and chief cell–specific YAP KO mice infected with H pylori 60,190 (CagA+), ΔCagA, or H felis. GSII+/GIF+ double-positive cells per gland were quantified (n = 5). Scale bar, 100 μm. (B) Representative IF for YAP and CD44v9 in the same corpus samples; YAP+ and CD44v9+ cells per sample were quantified (n = 5). Scale bar, 100 μm. (C) Representative TFF2 DAB IHC under the same conditions (n = 3). Scale bar, 100 μm. Statistics: Data are presented as mean ± SEM. Two-way ANOVA with Tukey’s post hoc test for genotype × infection comparisons. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. SEM, standard error of the mean. Chief cell–specific YAP KO suppresses SPEM in the oxyntic corpus during H pylori and H felis infection. SPEM was assessed based on GSII–GIF double positivity and the expression of established metaplasia-associated markers, including CD44v9 and TFF2. (A) Representative IF for GSII and GIF in WT and chief cell–specific YAP KO mice infected with H pylori 60,190 (CagA+), ΔCagA, or H felis. GSII+/GIF+ double-positive cells per gland were quantified (n = 5). Scale bar, 100 μm. (B) Representative IF for YAP and CD44v9 in the same corpus samples; YAP+ and CD44v9+ cells per sample were quantified (n = 5). Scale bar, 100 μm. (C) Representative TFF2 DAB IHC under the same conditions (n = 3). Scale bar, 100 μm. Statistics: Data are presented as mean ± SEM. Two-way ANOVA with Tukey’s post hoc test for genotype × infection comparisons. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. SEM, standard error of the mean. Pharmacologic Modulation of OTU Deubiquitinase, Ubiquitin Aldehyde Binding 2 Reveals the Plasticity of Yes-Associated Protein–Driven Metaplastic Signaling To determine whether YAP signaling, regulated by OTUB2, is pharmacologically tunable, we tested gain- and loss-of-function interventions in vivo. Repeated intraperitoneal dosing of rOTUB2 (2 μg/100 μL in PBS, every other day for 7 doses) enhanced YAP and CD44v9 expression in the oxyntic corpus, accompanied by a shift of ATP4B–GSII markers toward a metaplastic pattern (Figure 8A–C). This treatment also increased the number of GSII+ cells, indicating that enforced OTUB2 activity amplifies YAP-driven reprogramming and proliferation.Figure 8Pharmacologic modulation of OTUB2 regulates YAP-dependent SPEM. (A) rOTUB2 dosing scheme: 7 doses of rOTUB2 (2 μg/100 μL in PBS) were intraperitoneally administered every other day. (B) Representative corpus IF staining in WT mice and in chief cell–specific YAP KO mice following rOTUB2 treatment: CD44v9 (green), ATP4B (red), GSII (green), and YAP (red, DAB); scale bar, 450 μm and 250 μm. (C) Quantification of positive cells for the markers in (B) (n = 5). (D) Inhibitor scheme: LN5P45 (2.5 mg/100 μL, intraperitoneal) pretreatment followed by H pylori infection. (E) Representative corpus IF stains after LN5P45 + H pylori 60,190: CD44v9 (green), ATP4B (red), GSII (green), YAP (red), and OTUB2 (DAB); scale bar, 450 μm and 250 μm. (F) Quantification of positive cells in (E) (n = 5). Statistics: Data are presented as mean ± SEM. Two-way ANOVA with Tukey’s post hoc test for factorial comparisons (eg, treatment × infection); 2-tailed unpaired Student’s t tests for 2-group comparisons. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. SEM, standard error of the mean. Pharmacologic modulation of OTUB2 regulates YAP-dependent SPEM. (A) rOTUB2 dosing scheme: 7 doses of rOTUB2 (2 μg/100 μL in PBS) were intraperitoneally administered every other day. (B) Representative corpus IF staining in WT mice and in chief cell–specific YAP KO mice following rOTUB2 treatment: CD44v9 (green), ATP4B (red), GSII (green), and YAP (red, DAB); scale bar, 450 μm and 250 μm. (C) Quantification of positive cells for the markers in (B) (n = 5). (D) Inhibitor scheme: LN5P45 (2.5 mg/100 μL, intraperitoneal) pretreatment followed by H pylori infection. (E) Representative corpus IF stains after LN5P45 + H pylori 60,190: CD44v9 (green), ATP4B (red), GSII (green), YAP (red), and OTUB2 (DAB); scale bar, 450 μm and 250 μm. (F) Quantification of positive cells in (E) (n = 5). Statistics: Data are presented as mean ± SEM. Two-way ANOVA with Tukey’s post hoc test for factorial comparisons (eg, treatment × infection); 2-tailed unpaired Student’s t tests for 2-group comparisons. ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; ∗∗∗∗P < .0001. SEM, standard error of the mean. In contrast, in chief cell–specific YAP-deficient mice (Mist1-creERT2Cre/+;YAPfl/fl), rOTUB2 administration failed to induce the increase in YAP and CD44v9 expression, parietal cell (ATP4B) loss, or GSII expansion observed in WT mice (Figure 8B), indicating that OTUB2-induced metaplastic changes occur in a YAP-dependent manner in vivo. Conversely, pretreatment with the selective OTUB2 inhibitor LN5P45 (2.5 mg/100 μL intraperitoneally) prior to H pylori 60,190 infection markedly diminished YAP, OTUB2, and CD44v9 signals, while maintaining relatively preserved ATP4B–GSII indices (Figure 8D–F). The inhibitory effect of LN5P45 also limited the expansion of the proliferative zone and preserved glandular polarity, underscoring the functional reversibility of this signaling axis. These reciprocal effects demonstrate bidirectional control of YAP activity and highlight OTUB2 as an upstream modulator of YAP stabilization in vivo. Together with genetic results of the genetic analysis, these data indicate that the pharmacologic perturbation of OTUB2 can dynamically reshape the epithelial response to H pylori infection. Therefore, OTUB2 activation promotes the YAP-dependent epithelial metaplastic program, whereas OTUB2 inhibition or YAP loss suppresses it. Our findings confirm that epithelial-intrinsic and tissue-level reprogramming converge on YAP signaling, with OTUB2 acting as a regulatory amplifier. Discussion In this study, we functionally validated a CagA–OTUB2–YAP signaling axis as the molecular pathway through which H pylori infection drives epithelial junctional disruption and metaplastic reprogramming in the gastric epithelium.12,20 By integrating human epithelial models, single-cell transcriptomic analyses, and in vivo genetic approaches, we demonstrated that CagA-dependent signals promote YAP stabilization through OTUB2 and that epithelial-intrinsic YAP activity plays a functionally important role in Helicobacter-induced metaplastic responses. These findings position YAP as a central regulator of infection-associated epithelial plasticity and provide a mechanistic framework linking bacterial virulence to early gastric disease progression.18,23 Although YAP activation has been repeatedly observed in gastric tissues from patients with chronic H pylori infection, prior studies have largely described this association at a correlative level.19,20 Herein, we showed that infection with CagA-positive H pylori robustly induces YAP stabilization and nuclear accumulation, whereas these effects are markedly attenuated in ΔCagA infection.18 Moreover, the depletion of OTUB2 substantially reduced CagA-induced YAP stabilization, supporting a role for OTUB2 as a functional regulator that transduces CagA signaling to YAP activation. OTUB2 suppresses YAP degradation by removing K48-linked ubiquitin chains.24,25 Our findings extend this model by demonstrating that OTUB2 acts as a functional regulator of YAP stability in the context of CagA-positive H pylori infection. Under CagA-positive infection conditions, reduced YAP ubiquitination coincided with increased OTUB2 SUMOylation, which was sensitive to inhibition by the SUMO E1 inhibitor TAK-981. Although the precise molecular consequences of OTUB2 SUMOylation were not directly unraveled in this study, SUMOylation modulates protein stability, enzymatic activity, and protein–protein interactions, which highlight the plausibility of this modification enhancing OTUB2 function under infection-induced stress.40,41 At the transcriptional level, YAP activation in human gastric epithelial cell lines was associated with a clear epithelial-intrinsic “IM-like” transcriptional shift, characterized by the loss of ZO-1, induction of CDX2, and activation of TEAD-dependent transcription.19,32,42 YAP activity peaked approximately 5 hours after H pylori infection and subsequently declined, while remaining elevated relative to baseline, indicating that YAP is most strongly engaged during the early phase of infection. These findings support a model in which CagA disrupts junctional and inflammatory signaling circuits, mobilizes YAP to the nucleus, and transiently opens an intestinal-type transcriptional program.4,8,9,42 To connect these epithelial-intrinsic transcriptional changes to tissue-level pathology, we focused on the mouse oxyntic corpus and analyzed SPEM as a principal indicator of chief-cell reprogramming.35, 36, 37,43 Infection with CagA-positive H pylori was associated with parietal cell loss, expansion of the neck region, downward displacement of the proliferative zone, and induction of established SPEM markers, including GSII–GIF, CD44v9, and TFF2. In contrast, these metaplastic responses were markedly reduced in chief cell-specific YAP KO mice, demonstrating that epithelial YAP activity is functionally required for SPEM induction.22 Although GKN3 has been proposed as a marker of SPEM, particularly in murine studies, its interpretation in human datasets requires caution, as the human ortholog is annotated as a pseudogene (GKN3P) with uncertain functional conservation.44 In addition, GKN3/GKN3P expression is variably detected in human gastric epithelial populations and is often sparse in single-cell transcriptomic datasets.38 Consistent with this, GKN3/GKN3P was not robustly detected in our dataset. Therefore, we defined epithelial cell identities based on combinatorial marker expression and overall transcriptional context rather than relying on a single marker.38,39 Metaplastic changes were consistently more severe following infection with H pylori 60,190 (CagA+) than with the ΔCagA strain, supporting a role of CagA as a disease-modifying factor that amplifies epithelial reprogramming in the context of H pylori-induced injury.45 In contrast, infection with H felis—a rodent-adapted gastric Helicobacter species that does not encode CagA—markedly induced more pronounced metaplastic changes, demonstrating that lesion severity can be strongly influenced by species-specific pathogenic properties. Accordingly, in this study, H felis was not used to model CagA-dependent mechanisms, but instead served as a rodent-adapted reference (positive control) model that reliably elicits robust inflammatory and metaplastic responses in the murine stomach.34,46,47 In chief cell–specific YAP KO mice, the observation that SPEM was reduced but not completely abolished can be reasonably explained by incomplete tamoxifen-induced recombination and cellular heterogeneity within the gastric epithelium.34,35,48 Although the cellular origin of SPEM remains debatable, multiple lineage-tracing studies have consistently demonstrated that mature chief cells represent a major cellular source of SPEM.23,36,37,43 In this context, our genetic data strongly support a central role for chief cell–intrinsic YAP signaling in Helicobacter-induced metaplastic reprogramming.21,22 These findings are also consistent with the concept of paligenosis, as proposed by the Mills laboratory, wherein differentiated chief cells respond to injury by undergoing dedifferentiation and transcriptional reprogramming en route to a metaplastic state.49, 50, 51 Our data suggest that activation of the CagA–OTUB2–YAP axis represents a molecular mechanism through which microbial infection engages this injury-associated epithelial plasticity program.23,24,49 Beyond its role in cell fate reprogramming, our data further suggest that YAP activation may also amplify inflammatory and tissue remodeling responses. Although inflammation and oxyntic atrophy are well-established upstream drivers of metaplastic reprogramming,23 our data indicate that epithelial-intrinsic YAP activity contributes to sustaining these processes. In vitro, YAP activation was associated with increased expression of inflammatory mediators, including IL-8, and disruption of epithelial barrier integrity, whereas YAP depletion attenuated these responses. These findings support a model in which YAP reinforces epithelial injury signals, thereby promoting sustained inflammatory and metaplastic progression. Accordingly, loss of YAP in chief cells may attenuate epithelial-derived inflammatory signaling, leading to reduced tissue-level inflammation and oxyntic atrophy. From a translational perspective, YAP forms a nexus at which injury responses, regeneration, and metaplasia converge during chronic H pylori infection.18,21,22 By identifying OTUB2 as an upstream regulator of YAP stability, our study highlights potential avenues for therapeutic intervention, including OTUB2 inhibition or YAP pathway blockade.14,26,27 In addition, the coexpression of OTUB2, YAP, and SPEM markers may inform pathology- and transcriptome-based strategies for early risk-stratification.22,38,39 In summary, this study establishes YAP as a central regulator of H pylori-induced epithelial metaplastic reprogramming and delineates a CagA–OTUB2–YAP signaling axis that links epithelial junctional disruption and transcriptional remodeling in cells and organoids to chief cell–derived SPEM in vivo.19,21,22 This pathway represents a common denominator of infection-associated epithelial plasticity and a therapeutically actionable target, providing a molecular foundation for early intervention and prevention strategies in gastric cancer. Materials and Methods Human Gastric Tissues Human gastric tissue specimens, including chronic superficial gastritis, IM, and gastric cancer, were obtained from patients at Severance Hospital, Yonsei University College of Medicine. All patients provided informed consent, and the protocols were approved by the Institutional Review Board (IRB; Nos. 4-2011-0149, 4-2013-0880, and 4-2018-0676). Cell Culture and Helicobacter Infection Human gastric cancer cell lines AGS (CRL-1739, ATCC, RRID:CVCL_0139), AGS-LUC2 (CRL-1739-LUC2, ATCC, RRID:CVCL_A4BR), MKN74 (80,104, KCLB, RRID:CVCL_2791), N87 (CRL-5822, ATCC, CVCL_1603), KATO III (HTB-103, ATCC, RRID:CVCL_0371), the colorectal cancer cell line Caco-2 (HTB-37, ATCC, RRID:CVCL_0025), and HEK293T (CRL-11268, ATCC RRID:CVCL_1926) were cultured in RPMI-1640 (HyClone, SH30027.01) or Dulbecco’s Modified Eagle Medium (HyClone, SH30243.01) supplemented with 10% fetal bovine serum (Thermo Fisher Scientific) and 1 % penicillin–streptomycin (Thermo Fisher Scientific) at 37 °C in a humidified 5% CO2 incubator. H pylori strains—60190 (CagA+, ATCC 49503), ΔCagA mutant (CagA-deletion strain, kindly provided by Dr Masanori Hatakeyama, University of Tokyo), 11,637 (CagA+, ATCC 43504), and 8822 (CagA−, ATCC 51932)—and H felis (ATCC 49179) were cultured on Brucella agar plates supplemented with 10% sheep blood under microaerophilic conditions at 37 °C using the Campy Container system (BBL). For infection experiments, bacterial cells were harvested, suspended in Dulbecco’s PBS (pH 7.4), and adjusted to an optical density of OD600 corresponding to a multiplicity of infection (MOI) of 100. Control cells were treated with PBS under the same conditions. Gastric Organoid Culture Patient-derived gastric organoids were established from gastric tissue biopsies obtained with IRB approval, following previously described protocols.5,29,30 Organoids were embedded in Matrigel (Corning #356231) and cultured in Advanced Dulbecco’s Modified Eagle Medium/F12 (Gibco, Thermo Fisher Scientific) supplemented with penicillin–streptomycin (Gibco #15140-122), HEPES (10 mM; Invitrogen), GlutaMAX (Invitrogen #35050-061), B27 supplement (Invitrogen #17504-044), and N-acetylcysteine (1 mM; Sigma-Aldrich #A7250). Growth factors were added as follows: epidermal growth factor (50 ng/mL; PeproTech #AF-100-15), Noggin (PeproTech #120-10C), R-spondin1 conditioned medium (10%), Wnt-conditioned medium (50%), fibroblast growth factor 10 (200 ng/mL; PeproTech #100-26), gastrin (1 nM; Tocris #3006), and transforming growth factor β inhibitor A83-01 (2 μM; Sigma-Aldrich #SML0788). Optional supplements included nicotinamide (10 mM; Sigma-Aldrich #N0636) and ROCK inhibitor Y-27632 (10 μM; Sigma-Aldrich #Y0503). For 2D culture, organoid-derived single cells were seeded on collagen type I–coated Transwell inserts (15 μg/cm2; Corning #3470) and cultured under air–liquid interface. H pylori infection of 2D organoids was performed on day 7 of culture at an MOI of 100 to 200 using H pylori 60,190 (CagA+) or ΔCagA suspended in PBS. Mouse Models Eight- to 10-week-old C57BL/6J mice, kindly provided by Professor. Ki Taek Nam (Yonsei University), were used. Mist1-creERT2Cre/+ mice were generated using standard embryonic stem cell–targeting techniques, in which the entire coding region of the Mist1 gene was replaced with CreERT2. YAPfl/fl mice were derived from YAP/TAZfl/fl mice originally obtained from The Jackson Laboratory (JAX #030532). To generate a chief cell–specific Yap KO, Mist1-creERT2Cre/+; YAPfl/fl mice were subcutaneously treated with 3 doses of 5 mg tamoxifen. The tamoxifen dosing regimen was selected based on prior studies to achieve efficient recombination in Mist1+ chief cells. Mice were infected with H pylori 60,190 (CagA+ or ΔCagA) or H felis through oral gavage for 8 weeks. All the animal experiments were performed in accordance with institutional guidelines and approved by the Institutional Animal Care and Use Committee (IACUC No. 2022-0314, Yonsei University). Modulation of OTU Deubiquitinase, Ubiquitin Aldehyde Binding 2 Activity In Vivo To evaluate the functional role of OTUB2 in gastric metaplastic responses, rOTUB2 (2 μg/100 μL; Sino Biological, #13177-H07E) or the OTUB2 inhibitor LN5P45 (2.5 mg/100 μL; MedChemExpress, #HY-149482) was intraperitoneally administered to C57BL/6J mice every other day for 7 days. WT mice received rOTUB2 to assess its ability to promote SPEM, whereas H pylori–infected mice were pretreated with LN5P45 to examine the inhibition of OTUB2-mediated metaplastic signaling. Stomach tissues were collected 8 weeks after the final injection. Control groups received equal volumes of PBS. Lentiviral Transduction and Generation of Stable Cell Lines Premade lentiviral short-hairpin RNA (shRNA) constructs and a negative control consisting of a construct identical to the vector system (pLKO.1) were purchased from Addgene (shYAP#1, 42,540; shYAP#2, 42,541; shControl, 8453). Lentiviral particles were produced in vitro using a 3rd Generation Packaging System mix (LV053, Abcam) and lentipectin (G074, Applied Biological Materials). According to the manufacturer’s protocol, cell supernatants were harvested 48 hours after transfection in HEK293T cells and used to infect cells or stored at −80 °C. Puromycin (2 μg/mL) was used 48 hours after lentiviral infection to select AGS and MKN74 cells that stably expressed shRNA constructs. Western blotting (WB) was used to identify and validate the positive clones. Xenograft Mouse Model and Bioluminescence Imaging AGS cells (5 × 106) stably expressing shRNA targeting YAP (shYAP#1 or shYAP#2) or control shRNA (pLKO.1) were subcutaneously injected into the hind limbs of BALB/c nude mice. Mice were maintained in a pathogen-free animal facility at Yonsei University College of Medicine. All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC no. 2023-0047). Tumor growth was monitored every 3 to 5 days using caliper measurements, and tumor volume was calculated using the formula (length × width2)/2. At day 60 after injection, the mice were intraperitoneally injected with D-luciferin (150 mg/kg body weight; dissolved in sterile PBS at 15 mg/mL) for bioluminescence imaging. Luminescent signals were acquired 10 to 15 minutes after injection using an IVIS Spectrum imaging system (PerkinElmer). At the experimental endpoint, tumors were excised for gross imaging. Tumor tissues were either snap-frozen in liquid nitrogen or fixed in 10% formaldehyde for subsequent immunohistochemistry (IHC) and H&E staining. Yes-Associated Protein–TEA Domain Transcription Factor Luciferase Assay The cells were transduced at an MOI of 5 using a commercial lentiviral construct incorporating a luciferase reporter for TEAD activity (TEAD Luciferase Reporter Lentivirus, 79,833, BPS Bioscience). A commercial kit (Steady-Glo Luciferase Assay System, E2510, Promega) was used to assess luciferase activity, which was normalized to the cell number. Enzyme-Linked Immunosorbent Assay The concentration of IL-8 was measured using direct sandwich ELISA. Each sample was analyzed in quintuplicates. Human CXCL8/IL-8 was quantified using the DuoSET ELISA (DY208-05, R&D Systems) and DuoSET Ancillary Reagent (DY008, R&D Systems) kits. The absorbance was measured at 450 nm using a VersaMax microplate reader. The minimum detection sensitivity of ELISA was 20 pg/mL. Small interfering RNA transfection YAP targeting and control small interfering RNAs (siRNAs) were synthesized by Bioneer (Daejeon, Korea). The sequences used were as follows: Yes-associated protein small interfering RNA #1•Sense: 5′-CAG AAG AUC AAA GCU ACU U-3′•Antisense: 5′-AAG UAG CUU UGA UCU G-3′ •Sense: 5′-CAG AAG AUC AAA GCU ACU U-3′ Sense: 5′-CAG AAG AUC AAA GCU ACU U-3′ •Antisense: 5′-AAG UAG CUU UGA UCU G-3′ Antisense: 5′-AAG UAG CUU UGA UCU G-3′ Yes-associated protein small interfering RNA #2•Sense: 5′-AGA ACC GUU UCC CAG ACU A-3′•Antisense: 5′- UCU GGG AAA CGG UUC U-3 •Sense: 5′-AGA ACC GUU UCC CAG ACU A-3′ Sense: 5′-AGA ACC GUU UCC CAG ACU A-3′ •Antisense: 5′- UCU GGG AAA CGG UUC U-3 Antisense: 5′- UCU GGG AAA CGG UUC U-3 For OTUB2 knockdown, a predesigned siRNA (sc-76016) was purchased from Santa Cruz Biotechnology. Transfection was performed using Lipofectamine RNAiMAX (13,778-150, Invitrogen) according to the manufacturer’s instructions, with 20 nM siRNA applied to AGS cells. Plasmid Transfection Following the manufacturer’s instructions, Lipofectamine 2000 (11,668-019, Invitrogen) was used to transfect AGS cells with each construct. pEGFP-C3-hYAP1 (RRID:Addgene_17843) was purchased from Addgene. Full-length CagA (AF202923) and EPIYA phosphorylation-resistant mutants were provided by Professor M. Hatakeyama of the University of Tokyo. The medium was replaced with fresh medium 6 hours after transfection. Cells were used for all assays 24 hours after transfection. Electroporation The YAP-EGFP vector and YAP siRNA were delivered to normal gastric organoids cultured for more than 7 days through electroporation using NEPA21 (Nepagene), according to the manufacturer’s instructions. Cell Fractionation AGS cells were infected with H pylori 60,190 (CagA-positive) or H pylori ΔCagA (CagA-negative). Following the manufacturer’s instructions, nuclear-cytoplasmic fractionation was performed 5 hours after infection using the NE-PER Nuclear and Cytoplasmic Extraction Reagent Kit (78,833, Thermo Fisher Scientific). Immunoprecipitation Assay AGS cells were transfected with His-tagged ubiquitin (Addgene #99540) for 24 hours prior to infection. Following H pylori infection, YAP or OTUB2 was immunoprecipitated using Protein A/G PLUS agarose beads. WB was performed using anti-Ub and anti-SUMO2/3 antibodies to detect polyubiquitinated YAP and SUMOylated OTUB2, respectively. Western Blotting Cells were collected and resuspended in radioimmunoprecipitation assay buffer containing 25 mmol/L Tris-Cl (pH 7.4), 0.15 mol/L NaCl, 1% NP40, 0.1% sodium dodecyl sulfate (SDS), 1% sodium deoxycholate, and 1 mmol/L EDTA, along with a cocktail of protease and phosphatase inhibitors (P3300; GenDEPOT). The protein concentrations in the lysates were measured, and an SDS sample buffer consisting of 62.5 mM Tris-Cl (pH 6.8), 2% SDS, 10% glycerol, β-mercaptoethanol, and 0.002% bromophenol blue was added. The samples were then heated to 95 °C for 5 minutes. Proteins were separated by SDS–polyacrylamide gel electrophoresis using 7% or 10% gels and subsequently transferred to Immobilon-FL polyvinylidene fluoride membranes (IPFL00010, Merck Millipore). Primary antibodies were incubated overnight at 4 °C with gentle agitation, and the signal was developed using the SuperSignal West Pico PLUS Chemiluminescent Substrate (34,580, Thermo Fisher Scientific) in the dark. Immunostaining and In Situ Hybridization Formalin-fixed paraffin-embedded slides were deparaffinized and rehydrated in a declining graded sequence of ethanol (100%, 90%, and 70%) for immunostaining. A cooker was used to incubate the slides at high pressure after they were submerged in an antigen retrieval solution (S169984-2, Dako). To inhibit endogenous peroxidase activity, the slides were treated with 3% hydrogen peroxide for 30 minutes, after cooling on ice for 1 hour. After 2 PBS washes, the slides were incubated for 1 to 2 hours at room temperature in a humidity chamber with serum-free protein blocks (X0909; Dako). Mouse tissues were treated with mouse-on-mouse (vector) reagent before antibody application. Tissues were incubated with primary antibodies overnight at 4 °C. After 3 PBS washes, the specimens were incubated for 15 minutes at room temperature with either horseradish peroxidase–conjugated anti-mouse secondary antibodies (Dako) or horseradish peroxidase–conjugated anti-rabbit secondary antibodies (Dako). Mayer hematoxylin (Dako) was used for nuclear staining for IHC analysis, whereas diaminobenzidine was used for signal expression. For IF, Texas Red dye (Invitrogen) and fluorescein isothiocyanate (Invitrogen)-, Cy3 (Invitrogen)-, and Cy5 (Invitrogen)-conjugated secondary antibodies were used to detect the primary antibodies derived from various species. 6-diamino-2-phenylindole (Sigma) was used for nuclear staining. LSM 770 (Zeiss) and EVOS-M5000 (Thermo Fisher Scientific) microscopes were used for confocal and IF imaging, respectively. RNAscope probes specific for H pylori (Probe-B-H. pylori-16S, Cat. No. 542931) and H felis (Probe-B- H. felis-omp2-C1, Cat. No. 1281251-C1), along with the RNAscope 2.5 HD Detection Kit, were purchased from Advanced Cell Diagnostics and used according to the manufacturer’s instructions for ISH. A ready-to-use protease, hydrogen peroxide, and a target retrieval solution were used as pretreatments for the deparaffinized formalin-fixed paraffin-embedded slides. Prewarmed probes were placed on slides and incubated for 2 hours at 40 °C in a HybEZ oven (Advanced Cell Diagnostics). After the signal amplification, rapid red dye development was performed. Counterstaining was performed using Mayer’s hematoxylin. For ISH quantification, gland-base colonization was calculated as the ratio of ISH-positive glands to total glands per field. SPEM was quantified as the percentage of GSII–GIF double-positive cells among GIF-positive chief cells, whereas CD44v9- and TFF2-positive cells were counted per gland. The proliferative zone was evaluated by Ki67 immunostaining based on the distribution of Ki67-positive cells along the gland axis. Scanning Electron Microscopy Two-D organoid samples were fixed in Karnovsky’s fixative (2% glutaraldehyde and 2% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4) for 24 hours. Samples were then washed twice for 30 minutes with 0.1 M phosphate buffer and postfixed with 1% osmium tetroxide (OsO4) for 2 hours. Dehydration was performed using a graded ethanol series (50%–100%). The samples were dried using a LEICA EM CPD300 critical point dryer and coated with platinum using a LEICA EM ACE600 ion-sputter coater. Imaging was conducted using a field-emission scanning electron microscope (MERLIN, ZEISS). Transmission Electron Microscopy Two-D organoids were fixed under the same conditions for 12 hours and postfixed with 1% OsO4 for 2 hours. The samples were dehydrated in a graded ethanol series (50%–100%) and treated with propylene oxide for 10 minutes at each step. The specimens were embedded using a Poly/Bed 812 kit (Polysciences) and polymerized at 65 °C for 12 hours in an electron microscope oven (TD-700, DOSAKA). Semithin sections (∼200 nm) were prepared using a diamond knife and stained with toluidine blue for light microscopy. Areas of interest were then sectioned into ultrathin sections (∼80 nm), mounted on copper grids, and sequentially stained with 3% uranyl acetate for 30 minutes and 3% lead citrate for 7 minutes. Imaging was performed using a JEM-1011 transmission electron microscope (JEOL) equipped with a Megaview III CCD camera (Soft Imaging System) at an accelerating voltage of 80 kV. RNA Sequencing and Transcriptomic Analysis RNA-seq libraries were prepared from AGS cells using the QuantSeq 3′ mRNA-Seq Library Prep Kit (Lexogen) and sequenced on an Illumina platform. Reads were aligned to the human reference genome (GRCh38) using STAR aligner. DEGs were identified using independent t tests and fold-change criteria. GO enrichment was assessed through the GO database (www.geneontology.org), and GSEA (GSEA v4.0.3, Broad Institute) was performed. Data analysis and visualization of DEGs were performed using ExDEGA (Ebiogen). RNA-seq data have been deposited in the Gene Expression Omnibus (accession no. GSE275727). Single-Cell RNA Sequencing Analysis To analyze the expression levels in human gastric cancer, public single-cell RNA sequencing data (GSE150290) were used to reanalyze samples from normal tissues of 5 patients without cancer and paired cancerous and adjacent tissues from 24 patients with gastric cancer. In summary, a gene-count matrix was constructed from the fastq files mapped to the GRCh38 reference genome using Cell Ranger software (v7.0.0) provided by 10× Genomics. Following a previous approach, low-quality cells were filtered based on the following criteria: RNA counts >20,000, RNA feature counts <200 or >2000, and mitochondrial gene and hemoglobin gene percentages >30% and 10%, respectively. The remaining cells were normalized and clustered using the unsupervised clustering method with the FindClusters function (resolution = 0.3) in the Seurat R package (v5.0.0). Clusters were annotated by manually inspecting DEGs compared with previously defined cell type markers. Single-cell analysis consisted of GSE150290 data previously deposited by our research team in the Gene Expression Omnibus database. Single-cell quality control and data preprocessing were performed in the same manner as described in our previous study.39 Quantitative Reverse-Transcription Polymerase Chain Reaction Total RNA was isolated using QIAzol reagent according to the manufacturer’s instructions. Complementary DNA was synthesized from 1 μg of RNA using reverse transcriptase, and RT-PCR was performed using SYBR Green master mix on a real-time PCR system. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as the housekeeping control. OTU deubiquitinase, ubiquitin aldehyde binding 2 Forward: 5′-CCTGCTTTTGACTGGGTTCC-3′ Reverse: 5′-CTCATGGTTCCCCTGACACT-3′ Yes-associated protein Primer set 1 Forward: 5′-AGGGAGGAAGGAAGGAACAA-3′ Reverse: 5′-GAGAAACAGCTCCCAACTGC-3′ Caudal type homeobox 2 Forward: 5′-GACGTGAGCATGTACCCTAGC-3′ Reverse: 5′-GCGTAGCCATTCCAGTCCT-3′ Zonula occludens-1 Forward: 5′-AGAAGATAGCCCTGCAGC-3′ Reverse: 5′-AGTCCATAGGGAGATTCC-3′ Glyceraldehyde-3-phosphate dehydrogenase (housekeeping gene) Forward: 5′-CGACCACTTTGTCAAGCTCA-3′ Reverse: 5′-AGGGGTCTACATGGCAACTG-3′ Antibodies The following primary antibodies were used for IHC, IF, or WB analysis: YAP (Cell Signaling, 4912S, RRID:AB_2218911), TAZ (BD, 560,235, RRID:AB_1645338), YAP/TAZ (Cell Signaling, 8418S, RRID:AB_10950494), LATS2 (Cell Signaling, 5888S, RRID:AB_10835233), phospho-YAP (Ser127) (Cell Signaling, 4911, RRID:AB_2218913), SUMO2/3 (Cell Signaling, 4971, RRID:AB_2198425), CDX2 (Abcam, ab76541, RRID:AB_1523334), TFF2 (Invitrogen, PA5-57781, RRID:AB_2648410), MUC5AC (Abcam, ab77576, RRID:AB_2146994), MUC6 (Abcam, ab212646, RRID:AB_2864391), CD44v9 (CosmoBio, CAC-LKG-M002, RRID:AB_2910608), E-cadherin (Abcam, ab40772, RRID:AB_731493), Mist1 (Santa Cruz, sc-80984, RRID:AB_2065216), PGC (Novus, NBP1-91011, RRID:AB_11023080), ZO-1 (Invitrogen, 339,100, RRID:AB_87181), OTUB2 (ATLAS, HPA002329, RRID:AB_1079537; Invitrogen, PA5-99680, RRID:AB_2818613), CagA (Santa Cruz, sc-28368, RRID:AB_628229), H pylori (Novus, NBP2-29479, RRID:AB_3276838), ubiquitin (Santa Cruz, sc-8017, RRID:AB_2762364), and Lamin B1 (Santa Cruz, sc-374015, RRID:AB_10947408). Statistical Analysis All experiments were performed at least in triplicate. Data are presented as mean ± standard error of the mean. Statistical significance was determined using 2-tailed unpaired Student’s t tests or 1-/2-way analysis of variance (ANOVA) with Tukey’s post hoc test (GraphPad Prism v10). A P value < .05 indicated statistical significance. Statistical significance is denoted by asterisks as follows: ∗P < .05; ∗∗P < .01; ∗∗∗P < .001; and ∗∗∗∗P < .0001. All authors had access to the study data and had reviewed and approved the final manuscript.