OTUD4 regulates pancreatic cancer progression via Hippo/YAP axis

Abstract

The over-activation of Hippo/YAP axis was often observed in pancreatic adenocarcinoma (PAAD), while the detailed mechanism is not totally understood. Recent studies demonstrated that the ubiquitin modification, which controlled the protein stability of YAP, played important roles in Hippo signaling and PAAD progression. In order to understand the underlying link between YAP protein stability and Hippo activity in PAAD progression, we carried out GSEA bioinformatic analysis coupled with siRNA screening and identified OTUD4 as an important effector for Hippo signaling in PAAD. OTUD4, which was highly expressed in PAAD tissue, correlated with Hippo target gene expression in PAAD tissues. Depletion of OTUD4 significantly reduced the activity of Hippo/YAP axis and hampered PAAD progression. Mechanism studies revealed that OTUD4 could interact with YAP and promote YAP K48-linked poly-ubiquitination and degradation in PAAD. In conclusion, our study identified an interesting regulation mechanism between OTUD4 and Hippo signaling in PAAD, while targeting OTUD4 could be a plausible strategy for PAAD therapy.

Abbreviation: OTUD4, OTU Domain-Containing Protein 4; YAP, Yes-Associated Protein; TEAD, Transcriptional Enhanced Associate Domain transcriptional factor; TCGA, The Cancer Genome Atlas; ATCC, American Type Culture Collection; DMEM, Dulbecco’s Modified Eagle Medium; DUB, Deubiquitinase; GSEA, Gene Set Enrichment Analysis; NES, Normalized Enrichment Score; ECL, Enhanced Chemiluminescence; PVDF, Polyvinylidene Fluoride; PMSF, Phenyl Methane Sulfonyl Fluoride; PFA, Paraformaldehyde; Co-IP, Coimmunoprecipitation; IHC, Immunohistochemistry; CHX, Cycloheximide; IF, Immunofluorescence; GEO, Gene Expression Omnibus

Introduction

Pancreatic cancer is one of the most lethal human malignancies worldwide with the overall 5-year survival <10% [1]. The dys-regulation of Hippo pathway were widely observed in several human malignancies, inducing pancreatic cancer [2]. In physiological conditions, the activity of Hippo pathway was tightly control by a series of phosphorylation kinases, such as MST1/2, LATS1/2 and NF2 [3]. When Hippo pathway was activated, MST1/2 could phosphorylate LATS1/2 [4], which could subsequently induce the phosphorylation of YAP/TAZ [5]. The phosphorylated YAP/TAZ could be retained in the cytosol and subject to protein degradation via several E3 ubiquitin ligases [6,7]. If Hippo pathway was turned off, the un-phosphorylated YAP/TAZ could trans-locate into the nuclear [8,9], interact with several transcriptional factors and induce Hippo target gene expression [10,11]. Interestingly, YAP/TAZ could be over-activated in pancreatic cancer, while the auto-inhibition feedback of these phosphorylation kinases was often still functional [12]. Thus, it might indicate that other forms of protein modification could also regulate YAP function, which subsequently contributed to its over-activation and pancreatic cancer progression.

The biological link between pancreatic cancer progression and Hippo pathway has been established for decades. Several in vivo studies implicated that the activation of YAP was crucial for the pancreatic carcinogenesis [13,14], while YAP depletion could significantly inhibit the oncogenic process of pancreatic cancer in KRAS-driven mice models [15,16]. Besides, YAP activation could enhance cancer cell stemness and DNA repair capability [[17], [18], [19]], which subsequently promote chemotherapy resistance in pancreatic cancer [20]. Moreover, YAP could also induce the expression of PD-L1 in pancreatic cancer, which help to escape surveillance from immune cells [21]. Based on these conclusions, we can expect that target Hippo/YAP axis could be a promising strategy in pancreatic cancer. Several attempts were exerted to targeting the upstream kinases [22,23], while there was still a lack of mature inhibitors for Hippo/YAP targets in cancers.

Recently studies revealed that the post-translational modifications, such as ubiquitination [[24], [25], [26]], played important roles in Hippo signaling activity. The stability of YAP protein and YAP transcriptome are subject to a balance between E3 ubiquitin ligases and deubiquitinases [[27], [28], [29]]. The deubiquitinases, which removes the poly-ubiquitination chains from the substrate, could often stabilize certain targets [30]. Thus, blockage of critical deubiquitinases, which induces the protein degradation of YAP, is a plausible strategy to inhibit Hippo/YAP activity and cancer progression [[31], [32], [33]]. Based on this, we carried out GSEA analysis and siRNA screening of DUBs library, which revealed OTUD4 as a critical modulator for Hippo/YAP axis in pancreatic cancer. OTUD4 interacted with YAP protein and deubiquitinated YAP at several lysine sites, which subsequently stabilized and facilitated YAP function in pancreatic cancer. Our study identified a novel biological link between OTUD4 and Hippo signaling in pancreatic cancer progression, while blockade of OTUD4 could be a plausible strategy for Hippo-driven pancreatic cancer.

Results

OTUD4 regulates the Hippo pathway in PAAD

To identify deubiquitinases (DUBs) that may regulate Hippo pathway, we extracted RNA-seq data from the TCGA database and performed Gene Set Enrichment Analysis (GSEA) of the Hippo signaling pathway. The DUBs were ranked based on their Normalized Enrichment Score (NES), and DUBs with an NES > 1.5 were screened for further analyses. Subsequently, we conducted survival analysis on the initially screened DUBs to identify DUBs associated with poor prognosis. We then transfected a siRNA library targeting these DUBs and examined their effects on TEAD transcriptional activity (Fig. 1A). The results showed that knockdown of OTUD4 significantly suppressed TEAD4 transcriptional activity (Fig. 1B). GSEA revealed that high OTUD4 expression exhibited a strong positive correlation with YAP gene signature (Fig. 1C). Prognostic analysis further indicated that elevated OTUD4 expression was significantly associated with poor prognosis (Fig. 1D). By comparing OTUD4 expression levels between pancreatic cancer tissues and normal tissues, we found that OTUD4 was highly expressed in pancreatic cancer tissues (Fig. 1E). Immunohistochemical (IHC) experiments confirmed that the expression of both OTUD4 and YAP were significantly upregulated in pancreatic cancer patients, and OTUD4 expression was positively correlated with YAP expression (Fig. 1F-1G). Furthermore, we knocked down OTUD4 in pancreatic cancer cell lines and performed RNA sequencing analysis and GSEA analysis. The results demonstrated that OTUD4 knockdown led to a significant downregulation of YAP target genes (Fig. 1H-1I). In summary, these results indicate that OTUD4 plays an important role in regulating Hippo pathway.

Fig. 1.

OTUD4 regulates the Hippo pathway in PAAD.

A. Illustration of target DUB screening. GSEA and prognostic analyses were performed on data obtained from TCGA-PAAD. 10 deubiquitinases with a normalized enrichment score (NES) > 1.5 were selected, and their siRNAs were transfected into the AsPC-1 cells. Relative luciferase activity was measured to evaluate the activation status of the TEAD response element.

B. The relative luciferase activity of target DUBs. Upon knockdown of OTUD4, the activity of TEAD response element was most significantly inhibited.

C. GSEA of TCGA-PAAD data. High OTUD4 expression is positively correlated with the YAP gene signature.

D. Prognostic analysis of TCGA-PAAD data shows the relationship between OTUD4 expression and overall survival.

E. The expression difference of OTUD4 between normal and tumor tissues was analyzed by integrating datasets from both TCGA-PAAD and GTEx. Data were downloaded from UCSC Xena and normalized to ensure comparability.

F-G. Immunohistochemistry (IHC) assay shows the expression of OTUD4 and YAP between normal tissues and pancreatic cancer tissues. OTUD4 and YAP are more highly expressed in pancreatic cancer samples than normal pancreas samples. The scale bars are 400 μm for 10x and 100 μm for 40x. The P value was calculated using chi-square test.

H. GSEA of our RNA-sequencing data (siOTUD4 vs siControl). Gene expression in siOTUD4 group showed a negative correlation with the YAP gene signature.

I. Heatmap results indicated that the expression of genes within the YAP gene set was downregulated in siOTUD4 group.

The experiments were performed in triplicate. All the data are presented as the means ± SDs. Statistical methods: Student’s t-test for B. ***P < 0.001.

OTUD4 is crucial for the progression of PAAD

We next investigated the role of OTUD4 in PAAD progression using the AsPC-1 and SW1990 cell lines. The knockdown efficiency of OTUD4 in these cells was confirmed by RT-qPCR and Western blot assays (Fig. 2A-2C). CCK-8 assay showed that OTUD4 knockdown significantly suppressed cell proliferation (Fig. 2D-2E). Transwell assay indicated that OTUD4 depletion markedly decreased the invasive and migratory abilities of PAAD cells (Fig. 2F-2I). EdU assay revealed a significant reduction in the proportion of proliferating cells upon OTUD4 knockdown (Fig. 2J-2M). Additionally, apoptosis assay showed that OTUD4 knockdown promoted cell apoptosis (Fig. 2N-2Q). To evaluate the in vivo function of OTUD4, we established a subcutaneous xenograft model in nude mice. OTUD4 knockdown resulted in a significant reduction in tumor weight and growth rate (Fig. 2R-2T). IHC analysis of the xenografted tumors showed a decreased percentage of Ki-67-positive cells after OTUD4 knockdown (Fig. 2U-2V). Collectively, these results demonstrate that OTUD4 is essential for pancreatic cancer progression.

Fig. 2.

OTUD4 is crucial for the progression of PAAD.

A. Transfection of siControl or siOTUD4 #1/#2 was performed in PAAD cells. The knockdown efficiency of OTUD4 at the protein level was confirmed by Western blot analysis.

B-C. The relative mRNA expression level of OTUD4 was measured by RT-qPCR following transfection with siControl or siOTUD4 #1/#2.

D-E. Proliferation rate of PAAD cells transfected with siControl or siOTUD4 #1/#2 was measured using the CCK-8 assay.

F-I. The migration and invasion abilities of PAAD cells following transfection with siControl or siOTUD4 #1/#2 was assessed by Transwell assay.

J-M. Proliferation rate of PAAD cells transfected with siControl or siOTUD4 #1/#2 was assessed using the EdU assay. Hoechst 33342 was used for nuclear staining, and the scale bar is 200 μm.

N-Q. Apoptosis of PAAD cells transfected with siControl or siOTUD4 was measured by flow cytometry following staining with FITC and PI.

R-V. SW1990 cells transfected with shControl and shOTUD4 were injected subcutaneously into nude mice. The nude mice were sacrificed after 35 days of injection, and the tumor weight (S) and volume (T) were measured. IHC assays were utilized for evaluating the Ki-67 positive rate (U-V). n = 6. The scale bar is 400 μm for 10X and 100 μm for 40X.

All the experiments were performed in triplicate. All data are presented as mean ± SDs. Statistical methods: Student’s t-test for B-C, G, I, K, M, O, Q, S and V; two-way ANOVA for D and E. **P < 0.01; ***P < 0.001.

The hydrolytic activity of OTUD4 is essential in regulating YAP function

Since OTUD4 is a deubiquitinating enzyme of the OTU family, we hypothesized it regulates YAP via deubiquitination. Previous studies [34] report that mutation of the 45th cysteine residue to alanine (C45A) significantly impairs the deubiquitinase activity of OTUD4. To further investigate the mechanism by which OTUD4 regulates YAP function, we generated a catalytically impaired mutant, C45A, to examine the specific effect of OTUD4′s deubiquitinase activity on YAP (Fig. 3A). Knockdown of OTUD4 decreased YAP protein levels, as indicated by Western blot (Fig. 3B-3C), and reduced the expression of YAP target genes CTGF and CYR61, as shown by RT-qPCR (Fig. 3D). A dual-luciferase reporter assay also revealed markedly reduced TEAD4 transcriptional activity upon OTUD4 knockdown (Fig. 3E-3F). Overexpression of wild-type OTUD4 increased YAP protein level (Fig. 3G-3H) and the expression of CTGF and CYR61 (Fig. 3I-3J), and promoted TEAD4 transcriptional activity (Fig. 3K-3L). In contrast, the C45A mutant lacked these effects. Together, these results demonstrate that OTUD4 regulates YAP function in a hydrolytic activity-dependent manner.

Fig. 3.

The hydrolytic activity of OTUD4 is essential in regulating YAP function.

A. A schematic diagram depicts the C45A mutation of OTUD4.

B-C. The protein level of YAP after transfecting either siControl or siOTUD4 #1/#2 in AsPC-1 cells (B) and SW1990 cells (C) was measured through western-blot.

D. RT-qPCR assays were used for evaluating the mRNA expression levels of CTGF and CYR61 in PAAD cells after transfection with siControl or siOTUD4.

E-F. Dual-luciferase reporter gene assays were conducted in siControl-transfected or siOTUD4-transfected AsPC-1 (E) and SW1990 cells (F) to evaluate TEAD4 transcriptional activity.

G-H. Immunoblotting assays were performed to observe the protein level of YAP after transfecting Flag or Flag-OTUD4^WT or Flag-OTUD4^C45A in AsPC-1 cells (G) and SW1990 cells (H).

I-J. Quantification of CTGF and CYR61 expression levels by RT-qPCR was performed in PAAD cells transfected with Flag or Flag-OTUD4^WT or Flag-OTUD4^C45A.

K-L. TEAD4 transcriptional activity were measured using a dual-luciferase reporter assay after cells transfection with Flag, Flag-OTUD4^WT, or Flag-OTUD4^C45A plasmids.

All the experiments were performed in triplicate. All the data are presented as the means ± SDs. Statistical methods: Student’s t-test for D, E-F, I and J-K. *P < 0.05; **P < 0.01; ***P < 0.001.

YAP overexpression rescues OTUD4 knockdown-induced suppression of PAAD

To determine whether the tumor-suppressive effects of OTUD4 knockdown are mediated through YAP, we performed several rescue experiments. We knocked down OTUD4 and overexpressed YAP simultaneously. Western blot analysis showed that YAP protein level was decreased following OTUD4 knockdown, and was restored upon YAP overexpression (Fig. 4A). RT-qPCR results indicated that knockdown of OTUD4 downregulated the expression of YAP target genes CTGF and CYR61, and this downregulation was reversed by YAP overexpression (Fig. 4B). Both CCK-8 and EdU assays demonstrated that OTUD4 knockdown inhibited cell proliferation, and this phenotype was rescued by YAP overexpression (Fig. 4C, 4F-4G ). Transwell assays showed that YAP overexpression reversed the impairment in cell invasion and migration caused by OTUD4 knockdown (Fig. 4D-4E). Apoptosis assays indicated that OTUD4 knockdown increased cell apoptosis, which was also reversed by YAP overexpression (Fig. 4H-4I). Additionally, we established a subcutaneous xenograft model in nude mice to validate these findings in vivo. The results showed that YAP overexpression restored the suppressed tumor growth caused by OTUD4 knockdown (Fig. 4J-4L ). Ki-67 staining further confirmed that the anti-proliferative effect of OTUD4 knockdown was reversed by YAP overexpression (Fig. 4M-4N ). Together, these results demonstrate that the tumor-suppressive effects of OTUD4 knockdown are largely mediated through YAP and can be effectively rescued by YAP overexpression.

Fig. 4.

YAP overexpression rescues OTUD4 knockdown-induced suppression of PAAD.

A. SW1990 cells were transfected with siControl, siOTUD4, or siOTUD4+Myc-YAP, respectively. The protein levels of OTUD4 and YAP were measured by immunoblotting.

B. To determine the effect of OTUD4 and YAP, the mRNA levels of CTGF and CYR61 in SW1990 cells were measured using RT-qPCR after the cells were transfected with siRNAs and Myc-YAP plasmids.

C. CCK-8 assay was conducted to evaluate cell proliferation of SW1990 cells after transfection with siControl, siOTUD4+Myc, or siOTUD4+Myc-YAP.

D-E. To assess migration and invasion, transwell assays were performed using SW1990 cells transfected with the specified siRNA and plasmid.

F-G. The EdU assay was used to further evaluate the proliferative capacity of SW1990 cells following transfection with indicated siRNA and plasmid. The nuclei were stained with Hoechst 33342.The scale bar is 200 μm.

H-I. Flow cytometry was performed to analyze apoptosis in SW1990 cells transfected with specific siRNAs and plasmids.

J-N. SW1990 cells transfected with shControl, shOTUD4, or shOTUD4 + Myc-YAP were subcutaneously injected into nude mice. After 35 days, the mice were euthanized (J), and tumor weight (K) and volume (L) were measured. In addition, IHC staining was performed to assess the expression of Ki-67 (M), and the percentage of Ki-67-positive cells was quantified (N). n = 6.

All the experiments were performed in triplicate. All the data are presented as the means ± SDs. Statistical methods: Student’s t-test for B, E, J, I, K and N. Two-way ANOVA for C and L.**P < 0.01 ***P < 0.001.

OTUD4 interacts with YAP to promote its stability

We further investigated the mechanism by which OTUD4 regulates YAP. Immunofluorescence (IF) assays revealed that YAP is localized to both the cytoplasm and the nucleus, whereas OTUD4 is predominantly cytoplasmic (Fig. 5A). Co-immunoprecipitation (Co-IP) experiments demonstrated an endogenous interaction between OTUD4 and YAP (Fig. 5B). Based on this finding, we constructed truncation mutants of both OTUD4 and YAP to identify the domains responsible for their interaction (Fig. 5D). Subsequently, co-immunoprecipitation assays indicated that the OTU domain of OTUD4 and the WW domain of YAP are essential for their binding (Fig. 5E-5F). This interaction was further validated by molecular docking analysis (Fig. 5C). To examine whether OTUD4 stabilizes YAP through the proteasomal pathway, we knocked down OTUD4 and treated cells with the proteasome inhibitor MG132. The results showed that OTUD4 knockdown reduced YAP expression, and this effect was abolished upon MG132 treatment (Fig. 5G, J). We then performed protein stability assays using cycloheximide (CHX) and found that YAP stability decreased after OTUD4 knockdown (Fig. 5H-5I, 5K-5 L), while overexpression of wild-type OTUD4 enhanced YAP stability. In contrast, overexpression of the catalytically inactive C45A mutant did not significantly alter YAP stability compared to the control (Fig. 5M-5N). These results indicate that OTUD4 interacts with YAP and stabilizes YAP in a proteasome-dependent manner.

Fig. 5.

OTUD4 interacts with YAP to promote its stability.

A. Immunofluorescence (IF) analysis was performed to determine the subcellular localization of OTUD4 and YAP. OTUD4 is shown in green, and YAP is displayed in red. Nuclei were stained with DAPI. The scale bar represents 50 μm.

B. The endogenous interaction between OTUD4 and YAP was investigated by co-IP. Immunoprecipitation was performed using anti-OTUD4 or anti-YAP on lysates of SW1990 cells, and the associated proteins were detected by Western blotting.

C. A schematic diagram of the molecular docking structure, which reveals the interaction between the OTU domain of OTUD4 and the WW domain of YAP, was visualized using PyMOL.

D-F. Schematic representations of the individual domains of OTUD4 and YAP are provided (D). Plasmids encoding full-length OTUD4, YAP, and their respective truncated domains (OTUD4 residues: 1-155, Δ34-155 and 156-1114; YAP residues: 1-171, 1-292, 171-504 and 292-504) were transfected into HEK-293T cells. Co-IP assays show the WW domain of YAP (E) and the OTU domain of OTUD4 (F) are responsible for their interaction.

G, J. MG132 treatment abolished the decrease in YAP expression induced by siOTUD4 in both AsPC-1 (G) and SW1990 (J) cells. After transfection with siControl or siOTUD4, the cells were treated with 10 μM MG132 for 12 hours.

H-I, K-N. The effect of OTUD4 knockdown or overexpression on YAP stability was assessed using cycloheximide (CHX) assays. OTUD4 knockdown reduced YAP stability (H-I, K-L), whereas OTUD4 overexpression enhanced it. In contrast, the C45A mutant of OTUD4 failed to promote YAP stabilization (M-N). A concentration of 100 μM CHX was used in all experiments.

All the experiments were performed in triplicate. All the data are presented as the means ± SDs. Statistical methods: two-way ANOVA for I, L and N. *P < 0.05; ***P < 0.001.

OTUD4 stabilizes YAP via decreasing its K48-linked polyubiquitination

Since OTUD4 is a deubiquitinase of the OTU family, we further explored the underlying mechanism of how OTUD4 stabilizes YAP by examining YAP ubiquitination level in SW1990 cells. The results showed that knockdown of OTUD4 significantly increased both total polyubiquitination and K48-linked polyubiquitination of YAP, while K63-linked polyubiquitination remained largely unaffected (Fig. 6A). To further validate these findings, we employed two dominant-negative ubiquitin mutants, K48R and K63R. Following OTUD4 knockdown, the YAP ubiquitination level remained largely unchanged in the K48R group, whereas it was significantly increased in the K63R group (Fig. 6B). Similarly, overexpression of OTUD4 led to a notable decrease in both total and K48-linked polyubiquitination of YAP, with no obvious effect on K63-linked polyubiquitination (Fig. 6C). These results were further confirmed using K48R and K63R mutant plasmids (Fig. 6D). Additional ubiquitination assays demonstrated that the deubiquitinase activity of OTUD4 is essential for this function, as the catalytically inactive C45A mutant failed to regulate YAP ubiquitination levels (Fig. 6E-6F). Taken together, these findings indicate that OTUD4 stabilizes YAP via decreasing its K48-linked polyubiquitination (Fig. 7).

Fig. 6.

OTUD4 stabilizes YAP via decreasing its K48-linked polyubiquitination.

A-B. Knockdown of OTUD4 specifically enhanced K48-linked polyubiquitination of YAP without altering K63-linked polyubiquitination. To assess ubiquitination patterns, SW1990 cells were transfected with plasmids encoding HA-tagged wild-type ubiquitin or K48 or K63 ubiquitin (A) or K48R mutant ubiquitin or K63R mutant ubiquitin (B). Following transfection, cells were treated with 20 μM MG132 for 8 hours before harvesting. YAP was immunoprecipitated using an anti-YAP antibody, and its ubiquitination levels were detected by immunoblotting assay with an anti-HA antibody.

C-D. Overexpression of OTUD4 reduced K48-linked polyubiquitination of YAP without affecting K63-linked polyubiquitination. SW1990 cells were co-transfected with plasmids encoding Flag-OTUD4^WT and Myc-YAP, along with HA-tagged wild-type ubiquitin or K48 or K63 ubiquitin (C) or K48R mutant ubiquitin or K63R mutant ubiquitin (D). After transfection, the cells were treated with 20 μM MG132 for 8 hours, harvested, and subjected to immunoprecipitation using an anti-YAP antibody. Ubiquitination levels of YAP were then examined by Western blot with an anti-HA antibody.

E-F. Overexpression of OTUD4 decreased K48-linked polyubiquitination of YAP without altering K63-linked polyubiquitination; however, overexpression of the C45A mutant of OTUD4 did not affect either K48-linked or K63-linked polyubiquitination. SW1990 cells were co-transfected with plasmids encoding either Flag-OTUD4^WT or Flag-OTUD4^C45A and Myc-YAP, together with HA-tagged ubiquitin or K48 ubiquitin or K63 ubiquitin (E) or K48R mutant ubiquitin or K63R mutant ubiquitin (F). After transfection, cells were treated with 20 μM MG132 for 8 hours, harvested, and subjected to immunoprecipitation using an anti-YAP antibody. YAP ubiquitination levels were then analyzed by Western blot with an anti-HA antibody.

Fig. 7.

OTUD4 can stabilize YAP and promote the progression of PAAD.

OTUD4 interacts with YAP and stabilizes YAP by reducing its K48-linked polyubiquitination. The stabilized YAP forms a complex with TEAD to activate the transcription of downstream oncogenic target genes such as CTGF and CYR61, thereby promoting the progression of PAAD.

Discussion

The ubiquitin process is a critical step for protein sorting and degradation. while the dys-function of UPS (ubiquitin-proteasome system) is widely observed in human malignancies [[35], [36], [37]]. For example, the poly-ubiquitination of certain tumor suppressors or de-ubiquitination of certain oncogenic proteins could drive the cancer progression in special context. One important example is MDM2, which is an important E3 ubiquitin ligase for P53 degradation, while MDM2 over-expression or gene amplification in human cancer significantly reduces P53 protein level, which fails to induce cell cycle arrest or apoptosis [[38], [39], [40], [41]]. On the other hand, de-ubiquitinases could also play important role in cancer progression via stabilizing critical oncogenes [[42], [43], [44], [45], [46]]. In our study, we identified an interesting deubiquitinase OTUD4 for YAP stabilizing and pancreatic cancer progression. In clinical samples, OTUD4 were elevated in pancreatic cancer and correlated with poor survival. All these data could implicate that the abnormality of UPS could be an important hallmark in pancreatic cancer progression.

OTUD4 (OTU domain-containing protein 4) belongs to the super-family of OTU containing domain, which was firstly discovery from ovary cancer cells in 2003 [47]. Further studies revealed that OTUD4 was a functional deubiquitinase, which exerted its enzyme activity via its OTU domain [48]. Previous studies revealed that OTUD4 participated in several physiological process. For example, OTUD4 could stabilize STING protein and positively regulates anti-virus response in immune system [49,50]. Besides, OTUD4 could also play an important role in neurological diseases, while genetic mutation of OTUD4 could cause progressively worsening neurodegenerative diseases, including ataxia, dementia, and hypogonadotropism [51]. In human cancer, OTUD4 was reported to modulate several important oncogenes. For example, OTUD4 was found elevated in colorectal cancer, in which OTUD4 could stabilize GPX4 and promote cancer progression [43]. While in breast cancer, OTUD4 could accumulate CD73 and inhibit immune response and facilitate breast cancer progression [52]. However, little is known about OTUD4 function in pancreatic cancer. In our study, we found that OTUD4 is necessary for pancreatic cancer progression. The molecular mechanism study showed that OTUD4 could stabilizes YAP protein and enhance the activation of Hippo/YAP axis. Our study added novel knowledge for OTUD4 in regulation cancer progression.

Since OTUD4 could be a promising target for pancreatic cancer, the specific inhibitor for OTUD4 is still unavailable. There are still several challenges exist for the drug development and clinical application. Firstly, since several DUB family members share similar active pocket structures, it is technically challenge to develop specific inhibitors for OTUD4. On the other hand, OTUD4 could been a tumor suppressor on certain context. For example, OTUD4 could stabilize P53 in some cancers [53], which rise the concerns if OTUD4 is a valuable cancer drug target. However, P53 is often mutated in pancreatic cancer [54,55], while this could warrant its effectiveness and get rid of the worry of inhibiting P53. Besides, in our data, we showed that OTUD4 depletion could dramatically inhibit pancreatic cancer progression, which proved that the drug development of OTUD4 is still a valuable work. In the future, the application of High throughput chemical screening or PROTAC (Proteolysis-Targeting Chimeras) technology could be helpful for discovery effective OTUD4 inhibitors for cancer therapy.

Methods and materials

Reagents

In this essay, cycloheximide (CHX, HY-12320) and MG132 (HY-13259) were both purchased from MedChemExpress (MCE).

Cell culture

The AsPC-1, SW1990, HEK293T cell lines were all originated from American Type Culture Collection (ATCC). RPMI-1640 (R8758, Sigma-Aldrich) was used for the culture of AsPC-1 cell line, DMEM (D6459, Sigma-Aldrich) was used for SW1990 and HEK293T cell lines. 10% Fetal Bovine Serum (FBS, A5670701, Gibco) and Penicillin-Streptomycin Solution (C100C5, NCM) was added to the medium. Cells were cultured in a 37°C incubator supplied with 5%CO₂.

To establish stable cell lines, HEK293T cells were employed to produce shOTUD4 lentivirus. The target plasmid, envelope plasmid psPAX2 and pMD2G were mixed and transfected into HEK293T cells. The cells were then incubated for 48 h to produce the desired lentivirus. Subsequently, SW1990 cells were transfected with these lentiviruses. After 48 h of the transfection, puromycin (2 μg/mL) was added to screen stable transfected SW1990 cell lines.

Target DUBs screening

TCGA database was employed for the preliminary screening of the target deubiquitinases (DUBs). Specifically, patients from TCGA-PAAD cohort were divided into two groups based on the expression level of each DUB. Gene Set Enrichment Analysis (GSEA) was then performed for each DUB using GSEA software (version 4.3.3). The Normalized Enrichment Score (NES) of ‘CORDENONSI_YAP_CONSERVED_SIGNATURE’ gene set was calculated, and DUBs with an NES > 1.5 and negative correlation between their expression and prognosis were selected for further siRNA library screening. Luciferase plasmid and siRNAs were both transfected into cells for 48 h, then the luciferase assay was performed and the relative luciferase activity of candidate DUBs was ranked to identify target DUB.

Mouse xenograft tumor model

The BALB /c-nude mice were purchased from SPF Biotechnology (Beijing, China). Six-week-old nude mice were randomly allocated into each treatment group (n = 6 per group). The mice were injected subcutaneously with 3 × 10^6 SW1990 cells re-suspended in 100ul sterile PBS. The tumor-implanted mice were monitored continuously for six weeks, and the tumor volume was recorded weekly. Tumor volume was calculated with the following formula: Volume=length × width²/2.

Plasmids and siRNA

The Flag, Flag-OTUD4^WT, Flag-OTUD4^C45A and OTUD4 truncations were generated through PCR and inserted into pcDNA 3.1 vector. The HA-Ub, HA-K48-Ub, HA-K48R-Ub, HA-K63-Ub, HA-K63R-Ub, Myc-YAP and its mutant plasmids were stored in our laboratory and have been described in our previous works [56]. For plasmid transfection, Lipofectamine 3000 (7E732H3, Vazyme) was added, Moreover, T3000 enhancer reagent (7E732H3, Vazyme) was added as well. For siRNA transfection, RNAiMAX reagent (13778150, invitrogen) was used. The following sequences of the OTUD4 siRNAs were listed: siControl: UUC UCC GAA CGU GUC ACG U siOTUD4#1: GGG UAG GAC AAG UGG AAA U siOTUD4#2: CUG UAU GAG AAG GUA UUU A

Luciferase reporter assay

A luciferase reporter plasmid containing TEAD response element was used for this assay. The cells transfected with siOTUD4 or siControl, together with the reporter plasmid and Renilla were incubated for 24 hours and then collected and subjected to a luciferase assay via the Dual-Luciferase® Reporter Assay System (E1090, Promega).

RNA extraction and real-time PCR

RNA extraction was performed using an RNA isolator (R401, Vazyme). The cDNA was then subsequently reverse transcribed from the isolated RNA. After that, primers and SYBR GREEN (Q711-02, Vazyme) were mixed with cDNA and subjected to real-time qPCR. The Quantstudio^TM Design & Analysis Software (Applied Systems, thermo fisher scientific) was used for experiment designing and data analyzing. To normalize the results, 36B4 was employed as an internal reference. The primers used were as follows: 36B4: Forward: GAT GGC AAA TGT AGA CCC TAC C; Rerverse: AAG GCC CGG AGT ATC ACG A; CTGF: Forward: CAG CAT GGA CGT TCG CTG; Reverse: AAC CAC GGT TTG GCC TTG G; CYR61: Forward: CTC GCC TTA GTC GTC ACC C; Reverse: CGC AGT TGC ATT CCA G; OTUD4: Forward: TTC TGA TGG ATT ACA GAG GGC; Reverse: ACG CAT GTT GTC TTA CTC CTG A.

Western blotting

The reagents for western blotting were listed as follows: RIPA buffer (SW104-01, SEVEN), phenyl methane sulfonyl fluoride (PMSF, SW106-01, SEVEN), phosphatase inhibitors, enhanced chemiluminescence (ECL) luminescence reagent (MA0186, Meilunbio), and polyvinylidene fluoride (PVDF) membranes (E801-01, Vazyme). The following antibodies were used: anti-OTUD4 (HPA036623, Sigma-Aldrich), anti-Flag (AE005, AbClonal), anti-Myc (AE070, AbClonal), anti-YAP (ET1068-30, HUABIO) and anti-HA (90513, Biolegend). The secondary antibodies used were HRP-labeled goat anti-rabbit IgG (H + L) (A0208, Beyotime) and HRP-labeled goat anti mouse IgG (H + L) (A0216, Beyotime). The protein samples were extracted using a mixture of RIPA buffer, PMSF, and phosphatase inhibitors at a ratio of 100:1:1, mixed with loading buffer and denatured by heating at 99°C for 10 minutes. Then the samples were subjected to SDS-PAGE electrophoresis and transferred onto PVDF membranes. Subsequently, the membranes were blocked with 5% skim milk and then incubated with primary antibody and secondary antibody. The results were analyzed via enhanced chemiluminescence (ECL) system (Bio-Rad ChemiDoc).

CCK8 assay

Before the CCK-8 assay, the AsPC-1 cells and SW1990 cells were transfected with siOTUD4 or siControl and were incubated for 48 h. Then the cells were seeded into 96-well plates with 2,000 cells per well. After cell adhesion,10ul CCK-8 reagent (KA303002, Targetmoi) was added into indicated wells for 2 hours at 37°C, 5% CO₂. The absorbance of 450 nm were used for evaluating the proliferation rate among the negative control groups and OTUD4 knockdown groups. The absorbance values of 450 nm were measured and calculated at 0 h, 24 h, 48 h, 72 h.

Transwell assay

The transfected AsPC-1 and SW1990 cells were digested and re-suspended in 1 ml serum-free medium, then the cells were counted and diluted to 2 × 10^5 /ml and seeded into transwell chambers (32123009, Cornig) with 100 μl cell suspension. 500 μl of completed DMEM or RPMI-1640 were added to the lower chamber. In invasion assays, 100μl matrix gel was added into each chamber before the cells were seeded. The cells were incubated for 12-16 h, fixed with 4% paraformaldehyde (PFA) for 15 minutes and stained with crystal violet for 15 minutes as well. The photo was captured by EVOS M7000 (Invitrogen, thermo fisher scientific), Image J was applied for counting crystal violet stained cells.

Apoptosis analysis

After the transfection of AsPC-1 and SW1990 cell lines, the cells were collected and treated with an apoptosis detection reagent (556547, BD sciences), each sample was treated by a mixture consisting of 100μl 1 × BD buffer, 5μl FITC-Annexin V, 5μl PI, light-shielded. After 15 minutes, the samples were added with 400μl 1 × BD buffer per sample and translocated into flow tubes in the dark and analyzed immediately with flow cytometry (BD Bioseiences). The results were analyzed via FlowJo ver10.0.7.

EdU assay

After the cells were transfected with indicated siRNA or plasmid for 48 h, the cells were digested and seeded into 96-cell plates with 1 × 10⁴ cells per well. Cell-Light^TM Apollo567 in vitro kit (C10310-1, ribobio) was applied for EdU assay. The experiment was performed according to the instructions provided with the EdU assay kit. Hoechst 33342 was applied for nuclei staining. The results were captured by EVOS FL and analyzed by ImageJ.

Immunofluorescence assay

The AsPC-1 and SW1990 cells were fixed with 4% PFA for 15 minutes, then the cells were permeabilized with 0.1% Triton X-100 for 15 minutes. After washed by PBS, the cells were blocked with 5% BSA (ST025, Beyotime) for 1 hour. Several antibodies were utilized for immunofluorescence, these antibodies were: anti-OTUD4 (HPA036623, Sigma-Aldrich), anti-YAP (sc-101199, Santa-Cruz). The cells were incubated with primary antibodies at 4°C overnight and secondary antibodies for 1 hours. At last, the cells were re-stained with DAPI for 10 minutes in order to stain the nuclei. A confocal laser scanner was made use of capturing the images. Additionally, the secondary antibodies used in this experiment were: Goat anti-Mouse IgG (H + L) Highly Cross-Absorbed Secondary Antibody, Alexa Flour^TM594 (A-11005, invitrogen); Goat anti-Rabbit IgG (H + L) Highly Cross-Absorbed Secondary Antibody, Alexa Flour^TM488 (A-11008, invitrogen).

Coimmunoprecipitation assay

Cells were lysed by IP lysis buffer (P0013, Beyotime) and centrifuged to get the supernatant liquid. Then the supernatant was divided to IP, IgG and Input groups, followed by incubating with antibodies or IgG (Mouse: A-7028, Beyotime Rabbit: A-7016, Beyotime) and protein A/G (P2108, Beyotime) at 4°C overnight. On the next day, the protein A/G beads were centrifuged and washed by lysis buffer, then, 2 × SDS-PAGE loading buffer was added to resuspend the beads, then the mixture was heated at 99°C for 10 minutes to denature the protein. After heating, western blotting was performed to evaluate the result.

Ubiquitination assay

SW1990 cells were transfected with indicated plasmids and OTUD4 siRNA. After transfection, SW1990 cells were treated with 20 μM MG132 for 8 hours. The cells were obtained after 48 hours of transfection and pre-cleared with protein A/G for 2 hours. Subsequently, the pre-cleared lysates were mixed with anti-Myc or anti-YAP antibodies and incubated overnight at 4°C. On the next day, the beads were washed, and resuspended with 2 × loading buffer, followed by heating at 99°C for 10 minutes. The ubiquitination level of YAP was detected via immunoblotting.

Molecular docking analysis

The structures of OTUD4 and YAP were downloaded from uniprot (https://www.uniprot.org/uniprotkb), then the structures were optimized via Discovery Studio 2019 software. The results of optimized structures were visualized through PyMOL.

Immunohistochemistry (IHC)

The tissues were fixed with 4% PFA and embedded in paraffin wax. Then tissues were sliced into pieces using S700 paraffin microtome (RWD, USA). The sections were treated with xylene for 20 minutes to hyalinize the tissue, next 100%, 95%, 75% ethanol were utilized for isolate the xylene and treated histologic sections were stained with an immunohistochemical kit (PV-9000, ZSBG-BIO). The primary antibodies were incubated at 4°C overnight, and then the secondary antibody was incubated at room temperature for 20 minutes. Diaminobenzidine Horseradish Peroxidase Color Development Kit was utilized for ensuring the antigen-antibody complexes visible. After the coloration, the sections were re-stained with hematoxylin staining solution (BA4025, BASO). Finally, the results were captured using EVOS M7000. Anti-Ki-67 (ZM-0166, ZSBG-BIO) was employed for ki-67 staining. The Ki-67 positive cells were counted using ImageJ.

Cycloheximide (CHX) assay

The AsPC-1 cells and SW1990 cells were transfected with siControl, siOTUD4, while SW1990 cells were additionally transfected with Flag, Flag-OTUD4^WT and Flag-OTUD4^C45A plasmids for 24 hours. Then the transfected cells were treated with 100 μM CHX (HY-12320, MCE) and the samples were harvested at 0 h, 4 h, 8 h, 12 h. Subsequently, immunoblotting was employed to evaluate the protein levels of OTUD4 and YAP. The relative grey value of YAP was measured using ImageJ.

Clinical Samples

Paraffin-embedded tissue blocks were obtained from the Pathology Department of Shengjing Hospital. A total of 26 normal pancreatic tissues and 33 pancreatic cancer tissues were collected to analyze the expression levels of OTUD4 and YAP in cancer tissues and normal tissues. During the experiment, the personal information of the patients was well preserved, and no extra troubles were introduced to the patients.

Analysis of publicly available clinical data

Gene expression data from GTEx database and TCGA-PAAD cohort were obtained from UCSC Xena (xena.ucsc.edu). The expression level of OTUD4 was calculated and visualized using GraphPad Prism 9.5. The survival analysis was performed via R studio (version.4.2.1). The R package utilized for analysis were: survival (version.3.3.1), surviminer (version.0.4.9), ggplot2 (version.3.4.4).

RNA-sequencing and data analysis

AsPC-1 cells of three biological replicates were transfected with siControl or siOTUD4. After 48 h the RNA was isolated. The samples were transferred to BGI (Shenzhen, China) for RNA sequencing analysis. Differentially expressed genes were calculated using DESeq2, and genes with log₂FC >1 and P < 0.05 were considered statistically significant. Heatmap was plotted using the online platform SRplot.

Data statistics

SPSS 26.0 was used for data analysis. All the data are presented as the means ± SD from a minimum of triplicated assays, with at least six replicates used for animal analysis. Student’s t-test was utilized to observe the differences between two groups, while two-way ANOVA was used for analyzing the changing indexes over time. P < 0.05 was defined as statistically significant.

Ethics approval and consent to participate

All the experiments in this article were conducted in strict compliance with relevant laws and institutional guidelines and were approved by the Ethics Committee of Shengjing Hospital, China Medical University. The ethnical code number is 2025PS1271K for the use of clinical samples and 2025PS711K for animal experiments.

Data availability

The RNA-seq data can be found in the GEO database (GSE308205).

Consent for publication

All authors consent to publication.

Competing interest

The authors declare that they have no conflict of interest.

Funding

The project was supported by the Liaoning Provincial Key Research and Development Project (2023JH2/101700184, Xiaodong Tan) and Liaoning Provincial Key Research and Development Project (2023JH2/20200003, Jian Zhu).

CRediT authorship contribution statement

Shuaike Qu: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Software, Validation, Visualization, Writing – original draft. Zhihao Liu: Conceptualization, Data curation, Formal analysis, Resources, Validation, Writing – original draft, Writing – review & editing. Beibei Wang: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Resources, Validation, Visualization. Xin Li: Conceptualization, Data curation, Methodology, Resources, Validation, Writing – review & editing. Zhongbo Li: Investigation, Methodology, Resources, Software, Writing – review & editing. Yingwen Gai: Investigation. Yanan Sun: Methodology, Resources, Software. Qi Zhang: Investigation. Yaguang Sun: Investigation. Wei Pan: Investigation. Jian Zhu: Conceptualization, Data curation, Funding acquisition, Resources, Supervision, Writing – original draft, Writing – review & editing. Xiaodong Tan: Conceptualization, Data curation, Funding acquisition, Project administration, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no conflict of interest.