JOSD1 stabilizes SULF1 to activate Wnt7B-FZD1 signaling in gastric cancer

Highlights

• •JOSD1 is upregulated in gastric cancer and correlates with advanced stage and poor prognosis.
• •JOSD1 functions as a deubiquitinase that stabilizes SULF1 by removing ubiquitin chains.
• •Stabilized SULF1 directly binds FZD1 to facilitate Wnt7B–FZD1 coupling and activate Wnt/β-catenin signaling.
• •JOSD1-driven oncogenic effects are strictly dependent on SULF1 stabilization.
• •Targeting the JOSD1–SULF1–Wnt7B/FZD1 axis represents a potential therapeutic strategy for gastric cancer.

Abstract

Background

Aberrant activation of Wnt/β-catenin signaling is a major driver of Gastric cancer (GC) progression. However, the upstream mechanisms that sustain receptor–ligand engagement within this pathway remain insufficiently characterized.

Methods

Comprehensive analyses of GC cohorts and tissue microarrays were performed to evaluate Josephin Domain Containing 1 (JOSD1) expression and its clinical significance. The impact of JOSD1 on cell proliferation, migration, invasion, apoptosis, and epithelial mesenchymal transition (EMT) was examined in vitro employing CCK-8, colony formation, Transwell, flow cytometry, Western blotting, and immunofluorescence assays. Subcutaneous xenograft models were used to assess the effects of JOSD1 on tumor growth in vivo. Mechanistic studies, including co-immunoprecipitation, ubiquitination, and rescue experiments, were employed to elucidate the molecular relationship between JOSD1, Heparan sulfate 6-O-endosulfatase 1 (SULF1), and the Wnt7B/FZD1/β-catenin signaling axis.

Results

JOSD1 expression was markedly elevated in GC tissues (log₂ FC > 1, FDR < 0.05) and correlated with advanced stage (P < 0.05) and poor patient prognosis (HR > 1, log-rank P < 0.05). Functionally, JOSD1 promoted GC cell proliferation, invasion, and EMT, while inhibiting apoptosis (P < 0.05). Mechanistically, JOSD1 functioned as a critical deubiquitinase that stabilized SULF1. Stabilized SULF1 directly bound the Wnt co-receptor Frizzled class receptor 1 (FZD1) and facilitated Wnt7B–FZD1 complex formation (P < 0.05), thereby activating canonical Wnt/β-catenin signaling and inducing β-catenin nuclear accumulation (P < 0.05). Ubiquitination and rescue assays confirmed that JOSD1-driven oncogenic effects were strictly dependent on SULF1 stabilization (P < 0.05). In vivo modulation of the JOSD1–SULF1 axis significantly altered tumor growth, apoptotic activity, EMT marker expression, and Wnt pathway activation (P < 0.05).

Conclusion

JOSD1 functions as a critical deubiquitinase that stabilizes SULF1 to activate Wnt/β-catenin signaling, thereby driving GC progression. Targeting the JOSD1-SULF1-Wnt7B/FZD1/β-catenin axis may provide a promising therapeutic strategy for patients with GC.

Graphic abstract

The graphical abstract was prepared using Figdraw (www.figdraw.com) under an official academic license.

Introduction

Gastric cancer (GC) remains a leading cause of cancer-related mortality worldwide, characterized by high recurrence rates and limited treatment options for advanced disease [1]. Despite improvements in surgery, chemotherapy, and targeted therapy, overall survival remains poor, highlighting the urgent need to elucidate molecular mechanisms driving GC progression and to identify novel therapeutic targets [[2], [3], [4]].

Aberrant activation of the Wnt/β-catenin pathway is closely associated with GC initiation and progression ([5,6]), promoting uncontrolled proliferation, apoptosis evasion, and epithelial-mesenchymal transition (EMT) [[7], [8], [9], [10]]. In tumors, genetic and epigenetic alterations frequently lead to the loss or functional inactivation of E-cadherin. As a result, membrane-bound β-catenin is released, accumulates in the nucleus, and drives sustained oncogenic transcriptional programs [11,12]. However, the extracellular processes that sustain Wnt ligand-receptor interactions and maintain chronic pathway activation in GC remain poorly understood.

Heparan sulfate-modifying enzymes, such as the extracellular sulfatase Heparan SULF1, regulate the sulfation status of heparan sulfate proteoglycans and thereby modulate the accessibility of Wnt ligands to their receptors [13]. Accumulating studies have revealed that SULF1 is dysregulated across diverse cancer types, exerting context-dependent effects on tumor behavior. In colorectal carcinomas, SULF1 is upregulated and promotes proliferation, invasion, and metastasis through activation of the FAK/PI3K/AKT/mTOR pathways [14], whereas in GC it is enriched in cancer-associated fibroblasts and contributes to TGF-β driven metastasis and chemoresistance [15,16]. Collectively, these findings highlight the multifaceted, tumor-promoting functions of SULF1. However, its precise role in gastric cancer progression and the mechanisms that sustain its oncogenic activity remain unclear.

Josephin domain–containing protein 1 (JOSD1), a deubiquitinase, has not been well characterized in GC. In this study, we identify deubiquitinase JOSD1 as a critical upstream regulator of SULF1, which it stabilizes to enhance Wnt7B-Frizzled-1 receptor engagement, activate canonical Wnt/β-catenin signaling, and drive gastric cancer cell proliferation, invasion and EMT. Clinical analyses indicated that JOSD1-SULF1-Wnt/β-catenin signaling axis is upregulated in GC and associated with advanced stage and poor prognosis, highlighting a potential therapeutic vulnerability in this signaling axis.

Materials and methods

Reagents and antibodies

RPMI-1640 medium (C11875500BT, Gibco), Lipofectamine™ 3000 transfection reagent (L3000015, Invitrogen), Protein A/G magnetic beads (HY-K0202, MedChemExpress), pCMV-HA-Ub plasmid (18,712, Addgene), Cell Counting Kit-8 (CCK8, G4103–1 M, Servicebio), Annexin V-FITC apoptosis detection kit (AP105,Liankebio), Transwell inserts (3422, Corning), TUNEL assay kit (11,684,795,910, Roche), cycloheximide (CHX) (C7698, Sigma-Aldrich), FastStart Universal SYBR Green Master (4,913,850,001, Roche). Primary antibodies: anti-Flag (F1804, Sigma-Aldrich), anti-β-actin (4970, Cell Signaling Technology), anti-SULF1 (ab252833, Abcam), anti-E-cadherin (3195, Cell Signaling Technology), anti-N-cadherin (13,116, Cell Signaling Technology), anti-Vimentin (5741, Cell Signaling Technology), anti-Snail (3879, Cell Signaling Technology), anti-JOSD1 (HPA036259, Sigma-Aldrich), anti-Frizzled (ab219413, Abcam), anti-Wnt7B (DF7526, Affinity Biosciences), anti-β-catenin (8480, Cell Signaling Technology) and anti-phospho-β-catenin (9561, Cell Signaling Technology).

SULF1, JOSD1, Wnt7B shRNA and overexpression lentiviral vectors were constructed by Shanghai Genechem Co., Ltd..

Bioinformatic analysis

The transcriptomic profiling datasets and corresponding clinical data for GC were retrieved from the GEO database (https://www.ncbi.nlm.nih.gov/geo/). The analysis was performed on an integrated cohort of gastric cancer samples compiled from multiple GEO datasets. Specifically, datasets GSE19826, GSE27342, GSE33335, and GSE56807, which contain both tumor and adjacent tissue samples along with survival information, were accessed and processed using the GEOquery R package. Differential expression analysis between tumor and adjacent normal tissues was performed in R, adopting a |log₂ fold-change| > 1 and a FDR < 0.05 as the significance thresholds. Patients were stratified into high- and low-expression groups based on the median expression level of the gene of interest. Kaplan-Meier survival curves were plotted, and overall survival differences were determined by the log-rank test. Associations between gene expression levels and clinicopathological variables were analyzed using the Chi-square or Fisher’s exact tests. All statistical analyses and visualizations were conducted within the R statistical environment (v4.3.1) [17].

Tissue-Microarray construction and immunohistochemistry

A 1.5-mm tissue microarray was constructed from 94 radical gastrectomy specimens (94 tumour cores and 86 matched adjacent non-tumour cores). Haematoxylin-eosin (H&E) sections were reviewed by two pathologists to delineate tumour versus adjacent mucosa; cores were arrayed in an 18 × 10 grid using an automated tissue arrayer. Sections (4 µm) were mounted on APES-coated slides and stored at 4 °C. Immunostaining for SULF1 was performed within two weeks (citrate-microwave antigen retrieval, HRP-DAB chromogen, haematoxylin counterstain). Positive and negative controls were included in each run. Digital whole-slide images were acquired, and two independent pathologists blinded to clinical data assigned H-scores; cores with >20 % tissue loss were excluded [18].

Cell culture, plasmids and transfection

Human GC HGC-27 and MKN45 cells were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). HGC-27 is a poorly differentiated diffuse-type (Lauren classification) gastric cancer cell line with TP53 mutation, while MKN-45 is a moderately differentiated intestinal-type (Lauren classification) one with wild-type TP53. The cells were cultured in RPMI-1640 supplemented with 10 % foetal bovine serum at 37 °C in 5 % CO₂. Transient transfections of over-expression plasmids or shRNAs were performed with Lipofectamine 2000 in antibiotic-free medium for 6 h, followed by medium replacement and 36 h additional culture.

Western blotting

Total protein was extracted and quantified as previously described. Samples were resolved by SDS-PAGE, transferred to PVDF membranes, probed with primary antibodies and detected with HRP-conjugated secondary antibodies. Protein bands were quantified using ImageJ and normalised to β-actin.

Co-immunoprecipitation (Co-IP)

Cells were seeded in dishes. After 24 h, cells were lysed in RIPA buffer supplemented with protease and phosphatase inhibitors. Lysates were pre-cleared and incubated overnight at 4 °C with 2 µg of the indicated antibody plus 20 µL Protein A/G magnetic beads. Beads were washed four times, and bound proteins were eluted with 1 × SDS loading buffer (100 °C, 10 min) for Western blotting. To assess SULF1 ubiquitylation, MG132 (10 µM) was added for the final 6 h to inhibit proteasomal degradation, and anti-HA immunoblotting was performed on anti-SULF1 or anti-Flag immunoprecipitates.

CCK8 assay

Cells (2 × 10³ per well) were seeded in 96-well plates and assayed at 0, 24, 48, 72 h. Each well received 10 µL CCK8 reagent for 1 h at 37 °C; absorbance was read at 450 nm. Growth curves were plotted as OD _{(450 nm)} versus time.

Flow-Cytometric apoptosis analysis

Cells were harvested, washed twice with ice-cold PBS, and stained with Annexin V-FITC and propidium iodide (PI) according to the manufacturer’s protocol. Samples were analysed on a BD FACSCanto II within 1 h. Early (Annexin V⁺/PI⁻) and late (Annexin V⁺/PI⁺) apoptotic populations were quantified using FlowJo v10.

Transwell migration and invasion assays

Migration: 6 × 10⁴ cells in 200 µL serum-free medium were seeded in the upper chamber; 600 µL medium with 20 % FBS served as chemoattractant.

Invasion: Matrigel (diluted 1:8 in serum-free medium) was polymerised onto the upper surface for 4 h before cell seeding. After 48 h, non-invading cells were removed with cotton swabs. Migratory/invaded cells were fixed, stained and counted in five random fields per insert. Experiments were performed in triplicate [19].

Immunofluorescence

Cells on coverslips were fixed with 4 % paraformaldehyde (20 min), permeabilised with 0.2 % Triton X-100 (5 min), blocked with 5 % BSA (1 h) and incubated overnight at 4 °C with primary antibodies. Alexa-Fluor-conjugated secondary antibodies (1:100) and DAPI were applied sequentially. Images were acquired with a confocal microscope.

Xenograft tumour model

All animal experiments were approved by the Medical Ethics Committee of The First Hospital of Lanzhou University (Approval No: LDYYLL2021–338). Eighteen SPF-grade female Balb/C nude mice (6–8 weeks old, weighing 18–22 g) were purchased from Chengdu Yaokang Biotechnology Co., Ltd. The mice were randomly assigned to the following groups: sh-Ctrl, sh-JOSD1, oe-Ctrl, oe-JOSD1, oe-JOSD1 + sh-SULF1, and sh-JOSD1 + oe-SULF1. Cells from each group were inoculated into the axillary region of the mice at a density of 1 × 10⁷ cells per mouse. After inoculation, tumors derived from sh-Ctrl cells served as the control group. Tumor measurements were initiated when palpable tumors reached approximately 50–100 mm³. The tumor length (a, the longest diameter) and the shortest diameter (b) were measured every 4 days using a caliper, and tumor volume was calculated using the formula V = (a × b²) / 2. Tumor growth was monitored until the volume reached ≥ 2000 mm³, at which point the mice were euthanized. The tumor tissues were then resected and divided into two portions: one was rapidly frozen and stored in liquid nitrogen, and the other was fixed with 4 % paraformaldehyde for immunofluorescence analysis.

TUNEL staining

Paraffin-embedded sections were processed using the TUNEL assay kit according to the manufacturer’s instructions. Apoptotic cells were visualised by fluorescence microscopy.

Statistical analysis

All data are presented as mean ± standard deviation (SD). Statistical analyses were performed using SPSS 25.0 and GraphPad Prism 10. Two-group comparisons were analysed by two-tailed unpaired Student’s t-test or Mann-Whitney U test, depending on normality and homoscedasticity. Multiple groups were compared using one-way ANOVA followed by Tukey’s post-hoc test. A P value < 0.05 was considered statistically significant.

Results

SULF1 expression in GC and its impact on patient survival

Integrated analysis across multiple independent GC transcriptomic datasets identified a consensus signature of 29 genes consistently upregulated in gastric cancer. Among these, SULF1, PMEPA1, and LY6E exhibited the most pronounced increases in expression (Fig. 1A-C). Subsequent survival modelling revealed that elevated expression of SULF1 and PMEPA1, but not LY6E, was significantly associated with adverse clinical outcomes. Notably, these associations displayed distinct temporal patterns: patients with high SULF1 expression experienced a marked reduction in 5-year overall survival (OS), whereas PMEPA1 overexpression correlated with a delayed yet significant decline in long-term survival (Fig. 1D, E). In contrast, LY6E expression showed no statistically significant association with patient survival (Fig. 1F). Stage-stratified analyses demonstrated a strong positive correlation between SULF1 transcript abundance and advanced clinical stage. In contrast, LY6E and PMEPA1 expression levels showed no significant stage-dependent variation (Fig. 1G-I). To experimentally validate these observations at the protein level, we performed immunohistochemical analysis on a prospectively collected tissue microarray comprising 94 paired GC and adjacent non-tumour specimens. Quantitative H-score assessment confirmed that SULF1 protein expression was significantly elevated in tumour tissues compared with matched peritumoral mucosa (Fig. 1J-K). Collectively, these multi-level analyses identify SULF1 as a consistently upregulated gene whose expression correlates with disease progression and poor clinical outcome, nominating it as a robust prognostic biomarker in GC.

Fig. 1.

SULF1 expression and prognostic significance in GC. (A-C) Bioinformatics analysis plot; (D-F) Kaplan-Meier survival curves depicting the correlation between gene expression and OS; (G-I) Analysis of gene expression levels stratified by clinical stage; (J) IHC validation of SULF1 protein expression in a clinical tissue microarray; (K) Quantitative H-score analysis in tumor tissues. (n = 3) *p < 0.05, **p < 0.01, ***p < 0.001.

Functional impact of SULF1 on GC cell

To delineate the oncogenic role of SULF1, we first profiled its baseline expression in non-malignant gastric epithelial cells (GES-1) and five GC cell lines (SUN-216, HGC-27, AGS, MKN-45, NCI-N87).Quantitative RT–PCR and immunoblot analyses revealed heterogeneous SULF1 expression across these models, from which two cell lines with intermediate expression levels (MKN-45 and HGC-27) were selected for subsequent gain- and loss-of-function studies (Fig. 2A–C). To modulate SULF1 expression, cells were transduced with lentiviral constructs encoding either three distinct SULF1-targeting shRNAs (shSULF1, shSULF2, shSULF3) or a SULF1 overexpression plasmid (oeSULF1). Efficient knockdown and ectopic expression were confirmed at both the transcript and protein levels by qRT–PCR and Western blotting (Fig. 2D–H). Among the three shRNAs tested, shSULF1–3 achieved the most robust silencing and was therefore used for all subsequent functional assays (hereafter referred to as sh-SULF1). Systematic phenotypic analyses demonstrated that SULF1 depletion significantly suppressed GC cell proliferation, whereas enforced SULF1 expression exerted the opposite effect (Fig. 2I–J). Consistently, wound-healing assays and Transwell invasion assays revealed markedly reduced migratory and invasive capacities following SULF1 knockdown, while SULF1 overexpression substantially enhanced both phenotypes (Fig. 2K–N). These effects were consistently observed in both MKN-45 and HGC-27 cells, underscoring the robustness of the findings. Collectively, these results indicate that SULF1 promotes proliferative and invasive phenotypes in gastric cancer cells, supporting its functional involvement in GC progression.

Fig. 2.

SULF1 drives malignant phenotypes in GC cells. (A-C) Baseline expression of SULF1 in GES-1 and a panel of GC cell lines; (D-F) Western Blot detection of SULF1 protein expression; (G-H) Quantitative RT-PCR analysis of SULF1 mRNA levels; (I, J)CCK-8 assay; (K, L) Wound-healing assay; (M, N) Transwell invasion assay. (n = 3) *p < 0.05, **p < 0.01, ***p < 0.001.

JOSD1 interacts with SULF1 and modulates gastric cancer cell behaviour

To elucidate the molecular mechanisms underlying SULF1-driven GC progression, we performed co-immunoprecipitation coupled with mass spectrometry (Co-IP–MS) to systematically profile the SULF1 interactome. Protein lysates from HGC-27 cells stably expressing Flag-tagged SULF1 were subjected to immunoprecipitation using an anti-Flag antibody, followed by LC–MS/MS analysis. Among the proteins selectively enriched in the Flag-immunoprecipitated fraction, the deubiquitinase JOSD1 emerged as a prominent candidate interactor (Fig. 3A, Supplementary materials 1). Analysis of the TCGA-STAD cohort revealed that JOSD1 mRNA expression is significantly elevated in gastric tumour tissues compared with matched adjacent non-malignant samples (Fig. 3B). Moreover, correlation analysis demonstrated a significant positive association between JOSD1 and SULF1 expression levels across GC samples (Fig. 3C). To further validate the physical interaction between JOSD1 and SULF1, endogenous co-immunoprecipitation (Co-IP) assays were performed in HGC-27 and MKN-45 cells, which confirmed the interaction between these two proteins (Fig. 3D-E).Given that JOSD1 is a membrane-anchored deubiquitinase implicated in protein stability and signal transduction [20], these observations prompted us to investigate whether JOSD1 functionally contributes to SULF1-associated oncogenic phenotypes.To this end, we established JOSD1 gain- and loss-of-function models in both HGC-27 and MKN-45 cells using lentiviral overexpression (oe-JOSD1) or shRNA-mediated knockdown (sh-JOSD1). Cell viability assays revealed that JOSD1 overexpression significantly enhanced proliferative capacity, whereas JOSD1 depletion markedly suppressed cell growth (Fig. 3F). Consistently, wound-healing and Transwell migration assays demonstrated that oe-JOSD1 substantially increased migratory potential, while sh-JOSD1 impaired cell migration (Fig. 3G–J). In addition, flow cytometric analysis of apoptosis showed that JOSD1 knockdown led to a significant increase in apoptotic cell populations (Fig. 3K–L), indicating a pro-survival role for JOSD1 in GC cells. To further examine whether JOSD1 influences EMT, we assessed the expression of canonical EMT markers by immunoblotting. Compared with control cells, JOSD1 overexpression induced a pronounced EMT-like molecular shift, characterised by downregulation of the epithelial marker E-cadherin and upregulation of mesenchymal markers N-cadherin, Vimentin, and the EMT-associated transcription factor Snail (Fig. 3M–P). Conversely, JOSD1 silencing elicited the opposite expression pattern. Collectively, these data establish that JOSD1 promotes proliferation and migration, inhibits apoptosis, and actively drives EMT in GC cells.

Fig. 3.

JOSD1 interacts with SULF1 and modulates the functions of GC cells. (A) Identification of JOSD1-interacting proteins by Co-IP-MS; (B) Transcrip tomic analysis of TCGA-STAD dataset; (C) Correlation analysis; (D, E) Co-IP of JOSD1 and SULF1 protein; (F) CCK-8 assays; (G, H) Wound-healing migration assays; (I, J) Transwell migration assays; (K, L) Flow cytometric analysis of apoptosis; (M-P) Western blot analysis of EMT markers. (n = 3) *p < 0.05, **p < 0.01, ***p < 0.001.

JOSD1 deubiquitinates and stabilises SULF1 by antagonising proteasomal degradation

To determine whether JOSD1 regulates SULF1 protein stability via its deubiquitinase activity, we performed cellular ubiquitination assays. GC cells were transfected with HA-Ubiquitin and treated with MG132 to block proteasomal degradation. Compared with control cells, JOSD1 overexpression markedly reduced the accumulation of polyubiquitinated SULF1, whereas JOSD1 silencing resulted in a pronounced increase in SULF1 ubiquitination (Fig. 4A). To independently validate these observations, we next assessed SULF1 protein turnover kinetics using cycloheximide (CHX) chase assays at the concentration of 100 μg/mL, with treatment durations set at 0, 2, 4, and 6 h. Immunoblot analyses revealed that JOSD1 overexpression substantially prolonged the half-life of SULF1, with SULF1 protein levels remaining largely preserved even after 6 h of CHX treatment. In contrast, JOSD1 knockdown significantly accelerated SULF1 degradation, resulting in a markedly shortened protein half-life (Fig. 4B–D). Collectively, these data demonstrate that JOSD1 promotes the deubiquitination of SULF1, thereby protecting it from ubiquitin-proteasome system (UPS)-mediated degradation and enhancing its stability in GC cells.

Fig. 4.

JOSD1 deubiquitinates and stabilises SULF1. (A) JOSD1 decreases the poly-ubiquitination level of SULF1; (B-D) Expression of SULF1 protein in GC cells was assessed at indicated time points following treatment with the protein synthesis inhibitor CHX. (n = 3) *p < 0.05, **p < 0.01, ***p < 0.001.

Direct SULF1-FZD1 interaction and its modulation of wnt7b-fzd1 engagement

Given that SULF1 enzymatically remodels heparan sulfate chains and thereby regulates ligand–receptor interactions, particularly within Wnt signalling pathways, we next investigated whether the oncogenic effects of the JOSD1–SULF1 axis are mediated through aberrant activation of Wnt/β-catenin signalling. We first examined whether SULF1 physically associates with the Wnt receptor Frizzled-1 (FZD1). Co-immunoprecipitation assays demonstrated that HA-tagged FZD1 was specifically precipitated by anti-Flag antibody only in cells co-expressing Flag-SULF1 and HA-FZD1, whereas no interaction was detected in cells transfected with either construct alone or in IgG control immunoprecipitates (Fig. 5A), indicating a specific interaction between SULF1 and FZD1. We next assessed whether SULF1 modulates the assembly of the Wnt7B–FZD1 ligand–receptor complex. Co-IP analyses revealed that SULF1 depletion markedly reduced the amount of Wnt7B co-precipitated with FZD1, whereas SULF1 overexpression substantially enhanced Wnt7B–FZD1 association (Fig. 5B). Notably, manipulation of Wnt7B expression levels did not affect the SULF1–FZD1 interaction (Fig. 5C), indicating that SULF1–FZD1 binding occurs independently of ligand availability. Collectively, these results establish SULF1 as a critical facilitator of Wnt7B engagement with FZD1, positioning SULF1 as a structural or biochemical modulator of Wnt receptor complex assembly rather than a canonical ligand component.

Fig. 5.

SULF1 interacts directly with FZD1 and modulates Wnt7B-FZD1 engagement. (A) Co-IP assay detecting the interaction between Flag-tagged SULF1 and HA-tagged FZD1 in transfected cells using an anti-Flag antibody; (B) Co-IP analysis of the Wnt7B-FZD1 interaction in cells with sh-SULF1 or oe-SULF1, quantitative analysis of the binding intensity is presented on the right; (C) Co-IP analysis of the SULF1-FZD1 interaction in cells with sh-Wnt7B or oe-Wnt7B, quantitative analysis of the binding intensity is presented on the right. (n = 3) *p < 0.05, **p < 0.01, ***p < 0.001.

Functional dependence of gastric cancer cells on the JOSD1–SULF1 axis

To functionally validate the hierarchical relationship between JOSD1 and SULF1 in gastric cancer, we performed systematic epistasis and rescue experiments in MKN-45 and HGC-27 cells. Experimental cohorts included blank control, negative control (sh-Ctrl/oe-Ctrl), single-gene JOSD1 perturbation (sh-JOSD1/oe-JOSD1), and dual-gene rescue groups (oe-JD1+sh-SF1 and sh-JD1+oe-SF1). Functional assays demonstrated that JOSD1 depletion markedly suppressed cell proliferation and migration while inducing apoptosis (Fig. 6A–G). Conversely, JOSD1 overexpression robustly enhanced proliferative and migratory capacities, with no significant impact on basal apoptotic rates. Importantly, these JOSD1-dependent phenotypes were critically contingent on SULF1 expression. Enforced SULF1 expression (oe-SULF1) effectively rescued the proliferative and migratory defects and reversed apoptosis induction caused by JOSD1 knockdown. In contrast, SULF1 silencing abrogated the oncogenic phenotypes elicited by JOSD1 overexpression and restored apoptotic sensitivity (Fig. 6A–G).

Fig. 6.

JOSD1-SULF1 Axis drives malignant phenotypes in GC cells. (A, B) wound-healing assay; (C, D) Transwell invasion assay; (E, F) Apoptosis analysis; (G)CCK-8 assay. (n = 3) *p < 0.05, **p < 0.01, ***p < 0.001.

Activation of Wnt7b/β-catenin signalling by the JOSD1–SULF1 axis drives EMT in gastric cancer cells

We next investigated whether activation of the JOSD1–SULF1 axis regulates EMT and downstream canonical Wnt/β-catenin signalling. Immunoblot analyses revealed that JOSD1 depletion induced an epithelial phenotype, characterised by upregulation of E-cadherin and concomitant downregulation of mesenchymal markers N-cadherin and Vimentin. In contrast, JOSD1 overexpression promoted a mesenchymal shift, inducing the opposite expression pattern. Notably, these EMT-associated changes were fully reversible upon modulation of SULF1 levels, indicating that JOSD1 regulates EMT in a SULF1-dependent manner (Fig. 7A–F). Mechanistically, we found that the JOSD1–SULF1 axis activates canonical Wnt/β-catenin signalling. Specifically, JOSD1 knockdown increased β-catenin phosphorylation, reduced total β-catenin abundance, and impaired its nuclear accumulation, consistent with attenuated Wnt signalling. Conversely, JOSD1 overexpression decreased phosphorylated β-catenin levels, increased total β-catenin, and promoted its nuclear translocation. Importantly, all of these effects were abrogated upon SULF1 silencing and restored by SULF1 overexpression, demonstrating that SULF1 is required for JOSD1-mediated activation of β-catenin signalling (Fig. 7G–L). Consistent with these observations, co-immunoprecipitation assays revealed that JOSD1 enhances the interaction between Wnt7B and the Frizzled-1 (FZD1) receptor in a SULF1-dependent manner (Fig. 7M), providing a mechanistic link between JOSD1-mediated SULF1 stabilisation and potentiation of upstream Wnt ligand–receptor coupling. Collectively, these data demonstrate that the JOSD1–SULF1 axis promotes EMT by facilitating Wnt7B–FZD1 engagement and activating canonical β-catenin signalling, thereby driving aggressive phenotypes in gastric cancer cells.

Fig. 7.

JOSD1 drives GC EMT via SULF1-dependent activation of WNT7B/β-catenin signaling. (A-F) Western blot analysis of EMT markers; (G-L) Analysis of Wnt/β-catenin signaling components and β-catenin localization. (M) Co-IP assay detecting the binding of Wnt7B to Frizzled. Lysates were immunoprecipitated with an anti-Wnt7B antibody, with normal IgG serving as a control. (n = 3) *p < 0.05, **p < 0.01, ***p < 0.001.

Targeting the JOSD1–SULF1 axis governs gastric cancer growth and metastatic potential in vivo

To confirm the oncogenic function of the JOSD1-SULF1 axis in vivo, we established orthotopic xenograft models. Growth-curve analysis revealed that JOSD1 expression levels robustly determined tumor growth dynamics. Compared with control tumours, JOSD1 depletion markedly suppressed tumour growth and reduced final tumour burden, whereas JOSD1 overexpression significantly accelerated tumour expansion (Fig. 8A–B). Critically, the tumorigenic activity of JOSD1 was strictly dependent on SULF1 status: silencing SULF1 in the oe-JOSD1 context abolished tumor promotion, while ectopic SULF1 expression rescued the growth inhibition induced by sh-JOSD1 (Fig. 8A, B). Consistent with these growth phenotypes, TUNEL staining revealed that JOSD1 promotes tumor growth by suppressing apoptosis in a SULF1-dependent manner (Fig. 8I, J). Western blot analysis of xenograft tumor tissues confirmed that JOSD1 and SULF1 protein levels were modulated consistent with the experimental design (Fig. 8C-E). JOSD1 regulates EMT programmes in vivo through SULF1. sh-JOSD1 tumors exhibited elevated E-cadherin and reduced N-cadherin and Vimentin levels, whereas oe-JOSD1 tumors displayed the opposite phenotype. Notably, these EMT-associated alterations were fully reversed upon modulation of SULF1 expression (Fig. 8K–L). In parallel, immunoblot analysis revealed that oe-JOSD1 markedly enhanced the nuclear accumulation of β-catenin, whereas sh-JOSD1 led to its cytoplasmic retention, confirming JOSD1-mediated activation of canonical Wnt signaling in a SULF1-dependent manner (Fig. 8F-H). Finally, dual immunofluorescence staining of tumor sections demonstrated that JOSD1 enhances the spatial colocalisation of Wnt7B and the Frizzled-1 (FZD1) receptor in vivo via SULF1, an effect that was lost upon SULF1 depletion (Fig. 8M–N), providing in vivo confirmation of enhanced Wnt ligand–receptor coupling. Collectively, these in vivo findings establish that the JOSD1–SULF1–Wnt7B/β-catenin signalling axis is a critical driver of EMT, tumour growth, and malignant progression in GC.

Fig. 8.

Orthotopic xenografts validate the JOSD1-SULF1-Wnt7B/β-catenin axis. (A) Representative photographs of tumors from the indicated experimental groups at the endpoint of the study; (B) Tumor growth curves of orthotopic xenografts over time. (C, D) TUNEL assay for detecting apoptotic cells in tumor sections; (E, F) Western blot analysis of JOSD1 and SULF1 protein expression in excised tumor tissues; (G, H) Immunofluorescence analysis of EMT markers in tumor sections; (I, J) Analysis of β-catenin localization; (K, L) Dual immunofluorescence staining showing the co-localization of Wnt7B and Frizzled-1 in tumor tissues. (n = 3) *p < 0.05, **p < 0.01, ***p < 0.001.

Discussion

GC remains a major global health challenge, characterized by high metastasis rates and poor responses to frontline chemotherapy, resulting in 5-year survival rates below 30 %. Understanding the molecular mechanisms driving GC progression and identifying novel therapeutic targets are therefore critical [[21], [22], [23]]. In this study, we systematically delineated the hierarchical interactions within the JOSD1-SULF1-FZD1-Wnt7B/β-catenin signaling axis in GC. We clarify how this axis coordinates SULF1 stability, modulates FZD1 receptor function, and activates a specific Wnt branch, thereby offering a novel molecular explanation for GC progression and metastasis [11,20].

We confirmed that SULF1 is markedly upregulated in GC tissues and positively correlates with adverse prognosis and advanced tumor stage. While an earlier report by Hur et al. suggested a tumor-suppressive role for SULF1 in GC via inhibition of interfering growth factor signaling [24], our comprehensive data derived from large-scale TCGA cohorts, independent clinical samples, and in vivo xenograft models firmly establish an oncogenic function for SULF1 in our experimental context. This context dependent duality of SULF1 resembles that of other signaling modulators like TGF-β, which can switch from tumor suppressive to oncogenic depending on the cellular context and Wnt ligand availability [25]. Consistent with our findings, recent studies in hepatocellular carcinoma and esophageal cancer have also identified SULF1 as a driver of tumorigenesis and drug resistance, supporting the notion that SULF1 acts as an oncoprotein in advanced malignancies [26,27]. This finding elevates SULF1 from a classical extracellular-matrix remodeling enzyme to a direct regulator of the Wnt receptor complex, providing a new perspective on its pro-tumour activities [28].

The deubiquitinase JOSD1 is also highly expressed in GC [29]. JOSD1 regulates SULF1 at the post-transcriptional level by inhibiting its polyubiquitination and proteasomal degradation, which accounts for the sustained high expression of SULF1 in the GC microenvironment [20,30]. This post-transcriptional mechanism explains the persistent high levels of SULF1 observed in the gastric-cancer microenvironment. The positive-feedback stabilisation-accumulation loop between JOSD1 and SULF1 constitutes the core engine driving sustained downstream signalling. Downstream, we demonstrate that the axis converges on the canonical Wnt/β-catenin pathway [31]. Accumulated SULF1 enhances Wnt7B-FZD1 engagement, effectively promoting me-a key step endowing GC with high metastatic potential. We thereby link JOSD1 and SULF1 to the Wnt7B/FZD1 specific signalling module, offering a refined regulatory map of aberrant Wnt signalling in GC [32,33]

From a translational perspective, our findings provide actionable insights for the development of targeted therapies for GC. Currently, targeting the Wnt pathway is challenging due to the toxicity associated with global Wnt blockade [34]. In contrast, small-molecule inhibitors designed to specifically disrupt the JOSD1-SULF1 interaction or the SULF1-FZD1 binding could offer a more selective and less toxic alternative to global Wnt pathway blockad [35]. While our study delineates this axis in the models employed, it is important to acknowledge that its activity may vary across the molecularly heterogeneous landscape of gastric cancer. Future studies should prioritize evaluating the relevance of this signaling axis across different molecular subtypes of GC, expanding the number of experimental animals, and employing clinically relevant patient-derived xenograft (PDX) models to validate its role in resistance to chemotherapy or immunotherapy. These efforts are critical for translating our mechanistic findings into clinically applicable strategies, thereby improving the prognosis of patients with GC.

These efforts are critical for translating our mechanistic findings into clinically applicable strategies and thereby improving the prognosis of gastric cancer patients.

Conclusions

We identify a critical JOSD1–SULF1–FZD1–Wnt7B signaling axis that drives GC progression. JOSD1 directly binds and deubiquitinates SULF1, preventing its proteasomal degradation and stabilizing its protein levels. Stabilized SULF1 translocates to the plasma membrane, where it interacts with FZD1 and enhances Wnt7B binding, thereby activating canonical Wnt/β-catenin signaling. This cascade promotes EMT, proliferation, migration, and invasion of GC cells while inhibiting apoptosis. In summary, our study not only clarifies the previously unrecognized oncogenic functions and precise molecular mechanisms of JOSD1 and SULF1 in GC but also establishes JOSD1 and SULF1 as novel prognostic biomarkers and potential therapeutic targets for this disease. Targeting this signaling axis may serve as promising potential targets for gastric cancer diagnosis and therapy.

Ethical statements

This study strictly adhered to internationally accepted standards for animal experimentation and implemented the 3Rs principles. All experiments involving live animals were reported following internationally recognized guidelines (ARRIVE 2.0) to promote ethical research practices.

Human Subjects Statement: All clinical tissue samples used in this study were obtained with patient informed consent, and the research protocol was reviewed and approved by the Ethics Committee of Shanghai Outdo Biotech Co., Ltd. (Approval No: YB M-05–02). All procedures involving human samples were conducted in strict accordance with the ethical principles of the Declaration of Helsinki.

Animal Experiment Statement:The animal experiments described in this study were approved by the Animal Ethics Committee of The First Hospital of Lanzhou University (Approval No: LDYYLL2021–338). All procedures were performed in compliance with institutional guidelines for the care and use of laboratory animals.

Funding

This work was supported by National Natural Science Foundation of Gansu province (No. 26JRRA335), Key Research and Development Program of Gansu Province (No. 26YFFA004), and Lanzhou University Student Innovation and Entrepreneurship Action Plan (No. 20260060199, No. 20260060068).

Availability of data and material

All data generated or analyzed during this study are included in this work and are accessible upon request from the authors.

CRediT authorship contribution statement

Lixin Liu: Writing – review & editing, Writing – original draft, Visualization, Resources, Formal analysis, Data curation, Conceptualization. Zhijian Ma: Writing – review & editing, Visualization, Software, Investigation, Data curation, Conceptualization. Yaqing Zhang: Writing – review & editing, Software, Resources, Methodology, Formal analysis, Conceptualization. Zhenzhen Ye: Writing – review & editing, Visualization, Software, Investigation, Formal analysis, Conceptualization. Hongbin Li: Writing – review & editing, Visualization, Resources, Methodology, Formal analysis, Conceptualization. Wengui Shi: Writing – review & editing, Supervision, Methodology, Formal analysis, Data curation, Conceptualization. Zuoyi Jiao: Writing – review & editing, Validation, Supervision, Project administration, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.