RASSF4 Suppresses Gastric Tumor Growth through Activation of Chk2-p53 Signaling Axis

Abstract

Purpose

Ras association domain family 4 (RASSF4) is a putative tumor suppressor that is frequently inactivated in multiple human cancers. However, its candidacy as a suppressor in gastric tumorigenesis remains undefined. To understand the role for RASSF4 in gastric tumorigenesis, we investigated its expression status in cancer cell lines and tissues and regulatory role in tumor growth.

Materials and Methods

RASSF4 expression was analyzed in 13 cancer cell lines and 20 carcinoma tissues using polymerase chain reaction and immunoblot assays. RASSF4 effect on cell proliferation and apoptosis was examined by flow cytometry, colony formation, and [³H]thymidine incorporation assays and its regulation of p53 was determined using cycloheximide chase, promoter reporter, and immunoprecipitation assays. Mouse xenograft assay was performed to verify RASSF4 effect on tumor growth and therapeutic response.

Results

RASSF4 expression is epigenetically inactivated in eight of 13 (61.5%) cancer cell lines and 15 of 20 (75%) primary carcinomas. RASSF4 suppresses cell proliferation by inducing a G2/M cell cycle arrest and enhances apoptotic response to therapeutic drugs. RASSF4 is induced in response to genotoxic agents to facilitate stress-induced apoptosis in a highly p53-dependent fashion. Mechanistically, RASSF4 stabilizes p53 through Chk2 activation and its apoptotic function is profoundly impaired by depletion of either p53 or Chk2. RASSF4 attenuates xenograft tumor growth and enhances tumor response to 5-fluorouracil. Clinically, RASSF4 expression correlates strongly with the overall survival of gastric cancer patients.

Conclusion

RASSF4 suppresses gastric tumor growth through the activation of the Chk2-p53 axis, illuminating the mechanistic consequence of its inactivation in gastric tumorigenesis.

Introduction

Aberrant activation of Ras is implicated in the development and progression of multiple human cancers [1]. Ras transduces growth factor signals to the nucleus via multiple kinases, such as Raf kinases, RalGEFs, and phosphatidylinositol 3’-kinases to control cell response to various extrinsic stimuli [1,2]. Ras also interacts with a diverse array of effector proteins, and most Ras effectors play oncogenic roles in stimulating cell proliferation, survival, and cancer progression [2,3]. However, certain Ras effectors such as the Ras association domain family (RASSF) proteins have growth-inhibitory functions by inducing apoptosis, cell cycle arrest, and senescence [4]. It was thus proposed that inactivation of growth-inhibitory effectors may drive an oncogenic shift of Ras function, which represents a non-mutational activation of Ras signaling in tumors.

The RASSF family comprises 10 members (RASSF1-RASSF10) that share the presence of the RA domain, which potentially associates with the Ras family of GTPase [5]. Several RASSF members serve as tumor suppressors by modulating cell growth signalings through direct or indirect interaction with activated Ras [5-7]. Among these, RASSF1 and RASSF5 (NORE1) are the most thoroughly investigated tumor suppressors that are frequently inactivated by promoter hypermethylation in multiple human cancers [5-8]. A series of investigations uncovers that several RASSF members including RASSF1 and RASSF5 suppress tumor progression through the regulation of a broad spectrum of cell functions such as proliferation, motility, epithelial-mesenchymal transition, and apoptosis [8-10]. Recently, we reported that RASSF1A functions as an antagonist of RhoA’s oncogenic activity while RASSF5 reinforces tumor necrosis factor (TNF)–induced apoptosis through the opposite regulation of BAX and TNF receptor [11,12].

RASSF4 (also known as AD037) is a member of the classical RASSF family of scaffold proteins whose expression is decreased in different tumor types [13-15]. Despite increasing evidence supporting its tumor suppressor role in multiple cancers, the biological characteristics and clinical significance of RASSF4 are poorly understood. Previous studies demonstrated that RASSF4 induces cell cycle arrest and apoptosis by linking Ras to several pro-death pathways or through the regulation of MST1-Hippo or YAP-Bcl-2 signaling [16,17].

Gastric cancer is one of the most commonly diagnosed malignancies with relatively high incidence in certain geographical areas such as Korea, China, and South America, and Chile and the fourth leading cause of cancer-related deaths worldwide, highlighting its significant global health burden [18]. Despite advances in chemotherapy and targeted therapies, the prognosis for advanced gastric cancer remains poor due to drug resistance and limited therapeutic options. Therefore, identifying novel therapeutic targets and understanding the underlying molecular mechanisms are crucial for improving patient outcomes. A number of molecular abnormalities, including genetic and epigenetic alterations of genes such as TP53, CDH1, ARID1A, PIK3CA, FGFR2, and HER2 are involved in the pathogenesis of gastric cancer development [19]. A growing body of evidence points to the key role of RASSF family members in gastric tumorigenesis. Epigenetic inactivation of RASSF1A is commonly observed in primary gastric carcinomas and associated with malignant tumor progression [9]. Altered expression of RASSF2, RASSF6, RASSF8, and RASSF10 is also identified to contribute to gastric tumor progression and poor survival of cancer patients [20-23]. However, the expression status and growth-regulatory function of RASSF4 in tumor cells and its candidacy as a suppressor in gastric tumorigenesis have not been explored yet.

In the present study, we identified that RASSF4 expression is commonly inactivated in gastric cancer cell lines and primary carcinoma tissues and that its alteration contributes to enhanced growth and chemoresistance of tumor cells. Our study also uncovers that RASSF4 activates p53 through checkpoint kinase 2 (Chk2) regulation, illuminating the presence of the RASSF4-Chk2-p53 signaling axis and the mechanistic consequence of its alteration in gastric tumorigenesis.

Materials and Methods

1. Human cancer cell lines and reagents

A total of 36 gastric tissues, including 20 primary carcinomas and 16 normal tissues, were obtained from 20 patients with cancer and 16 patients without cancer by surgical resection at the Kyung Hee University Medical Center (Seoul, Korea). Signed informed consent was obtained from each patient. Tumor specimens composed of at least 70% carcinoma cells and adjacent tissues found not to contain tumor cells were chosen for molecular analysis. Thirteen human gastric cancer cell lines were obtained from Korea Cell Line Bank (Seoul National University, Seoul, Korea) or American Type Culture Collection (Rockville, MD). The cells were maintained in Dulbecco’s modified Eagle medium supplemented with 10% fetal bovine serum (FBS; Gibco) at 37°C in a humidified atmosphere with 5% CO₂. Cycloheximide (CHX), 5-fluorouracil (5-FU), cisplatin, and adriamycin were purchased from Sigma-Aldrich.

2. Transfection of DNA and siRNA

Expression vector encoding RASSF4 was constructed by a polymerase chain reaction (PCR)–based approach using specific primer pairs (sense: 5′-ATGAAGGAAGACTGTCTGCCGAGTTC-3′ and antisense: 5′-CTTGGCCTCCACCAGCTGCTCCAGGCG-3′). The PCR products were cloned into pcDNA3.1 (Invitrogen) or pcDNA3.1-His6-V5 vector (Invitrogen). Expression vectors encoding RASSF1A, p53, and MDM2 were constructed using a PCR-based approach as previously described [11,24]. Transfection was performed using Turbofect (Fermentas), Lipofectamine LTX (Invitrogen), or Electroporation (Neon transfection system, Invitrogen). To generate RASSF4-expressing sublines, AGS cells were transfected with RASSF4 expression vectors and colonies were isolated by hygromycin selection (100-500 μg/mL). RASSF4 knockdown subline was established in MKN1 cells by transfection of shRASSF4 constructs (Genolution Pharmaceuticals Inc.) and Zeocin (Invitrogen) selection. siRNA duplexes against RASSF4, p53, or Chk2 and control siRNA duplex (siControl) which served as negative control were synthesized by Bioneer. Transfection of siRNA was done using electroporation.

3. Semi-quantitative reverse transcription–polymerase chain reaction

Total cellular RNA was extracted from tissues and cell lines by standard method. One microgram of DNase1-treated RNA was converted to cDNA by reverse transcription using random hexamer primers and MoMuLV reverse transcriptase (Life Technologies Inc.). PCR was performed for 36 cycles at 95°C (1 minute), 58°C-62°C (0.5 minutes), and 72°C (1 minute) in 1.5 mM MgCl₂-containing reaction buffer (TAKARA Bio Inc.) with specific primers for RASSF4 (sense: 5′-ACAACTGCTACCATGAGGGCAA-3′ and antisense: 5′-CTCTCAGCCTTGTGCACTGG-3′), TP53 (sense: 5′-ATAGTGTGGTGGTGCCCTATGAGCCG-3′ and antisense: 5′-GTGGGAGGCTGTCAGTGGGAAGAA-3′), CDKN1A (p21^WAF1) (sense: 5′-CTGCGCCAGCTGAGGTGTGAG-3′ and antisense: 5′-GCCGCATGGGTTCTGACGGA-3′), XAF1 (sense: 5′-ATGGAAGGAGACTTCTCGGT-3′ and antisense: 5′-TTGCTGAGCTGCATGTCCAG-3′), BAX (sense: 5′-TGATGGACGGGTCCGGG-3′ and antisense: 5′-TGTCCAGCCCATGATGGTTC-3′), PUMA (sense: 5′-GGGCAGGAAGTAACAATGAGA-3′ and antisense: 5′-CTCCCTGGGGCCACAAATC-3′), PMAIP1 (NOXA) (sense: 5′-GAGGAACAAGTGCAAGTAGCTG-3′ and antisense: 5′-GGAGTCCCCTCATGCAAGTT-3′), and RRM2B (p53R2) (sense: 5′-ATTCTCATCGAGAATGTTCA-3′ and antisense: 5′-CCACAAGTAATCTGTCAGCT-3′). Ten microliters of PCR products were resolved on 2% agarose gels. Quantitation was achieved by densitometric scanning of the ethidium bromide-stained gels. Absolute area integrations of the curves representing each specimen were then compared after adjustment for glyceraldehyde 3-phosphate dehydrogenase (GAPDH) level. Integration and analysis was performed using Molecular Analyst software program (Bio-Rad).

4. 5-Aza-2′-deoxycytidine treatment and bisulfite DNA sequencing

Cells were exposed to 1-10 μM of 5-aza-2′-deoxycytidine (5-Aza-dC; Sigma-Aldrich) for 48 hours and its effect on RASSF4 mRNA expression was determined by reverse transcription–polymerase chain reaction (RT-PCR). For bisulfite DNA sequencing analysis, 2 ng of bisulfite modified DNA was subjected to PCR to amplify the exon 1 region of RASSF4 (+204 to +256 relative to transcription start site) using primers designed to recognize sodium bisulfite-converted. The PCR products were cloned into TA cloning vectors (RBC Bioscience) and five clones of each specimen were sequenced by automated fluorescence based DNA sequencing.

5. Cell proliferation, colony formation, and apoptosis analysis

Cells were seeded in 60-mm dishes at the density of 0.3×10⁵ cells per dish and were maintained in the presence of 10% FBS. Cell numbers were counted using a hemocytometer for 4 days at 24-hour intervals. For flow cytometry analysis, cells were seeded at the density of 2×10⁵ cells in 60-mm dishes and cultured in medium with 10% FBS for 48 hours. DNA synthesis was measured by determining [³H]thymidine incorporation. Briefly, cells transfected with expression vector or siRNA were pulse-labeled for 4 hours with 1 μCi/m of [³H]thymidine (Amersham) and the radioactivity incorporated into trichloroacetic acid-precipitable materials was counted by a liquid scintillation counter. For colony formation assay, AGS and MKN1 subline cells (1×10³) were seeded in soft agar and maintained in the presence of hygromycin (100 μg/mL) for 7 days. Colonies were fixed with methanol for 15 min and stained with 0.05% crystal violet in 20% ethanol. For flow cytometric analysis of apoptotic sub-G1 fraction, cells were fixed with 70% ethanol and resuspended in 1 mL of phosphate buffered saline containing 50 μg/mL RNase and 50 μg/mL propidium iodide (Sigma-Aldrich). The assay was performed on a FACScan flow cytometer (Becton Dickinson), and the cell cycle profile was analyzed using MultiCycle software (Phoenix Flow Systems).

6. Reporter constructs and luciferase assay

The CDKN1A and BAX promoter region was cloned into the pGL4.14 vector (Promega Corporation). Cells were co-transfected with 200 ng of reporter plasmids and 20 ng of the β-galactosidase expression plasmid. β-galactosidase activity was measured by Mammalian β-Galactosidase Assay Kit (Invitrogen) according to the manufacturer’s protocol and used for normalization. After normalization of each extract of protein content, luciferase activity was measured using Steady-Glo Luciferase Assay System (Promega Corporation) and SpectraMax i3x microplate reader (Molecular Devices).

7. Immunoblot assay

Cells were lysed in buffer containing 20 mM Tris (pH 7.4), 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 50 mM sodium fluoride, 2 mM sodium pyrophosphate, 1 mM sodium orthovanadate, protease inhibitor cocktail, and 1 mM PMSF. The cell lysate was clarified by centrifugation and 20 to 40 μg of total protein was supplemented with Laemmli buffer and loaded on a 10% SDS–polyacrylamide gel for electrophoresis. Immunoblot (IB) analyses were performed using antibodies specific for RASSF4 (Abcam), cyclin D1 (Cell Signaling Technology), cleaved poly(ADP-ribose) polymerase (cl-PARP; Cell Signaling Technology), p53 (Santa Cruz Biotechnology), MDM2 (Santa Cruz Biotechnology), p-ATM (Ser1981, Cell Signaling Technology), p-Chk1 (Ser345, Santa Cruz Biotechnology), p-Chk2 (Thr68, Cell Signaling Technology), GAPDH (Santa Cruz Biotechnology), and β-tubulin (Sigma-Aldrich). Antibody binding was detected by enhanced chemiluminescence (Santa Cruz Biotechnology) using a secondary antibody conjugated to horseradish peroxidase.

8. Immunoprecipitation and ubiquitination assay

Cells were incubated with 5 mM MG132 for 6 hours and lysed in buffer containing complete protease inhibitor (Roche) and ubiquitin aldehyde (Sigma-Aldrich). The lysates were incubated with antibodies specific for p53, MDM2, and RASSF4 overnight at 4°C, and protein complexes were pelleted with protein A-agarose beads (Pierce) and separated by SDS–polyacrylamide gel electrophoresis. Ubiquitinated p53 proteins were precipitated with anti-p53 antibody and analyzed by immunoblotting using anti-His antibody.

9. Mouse tumor xenograft assay

Identical numbers (1×10⁶) of AGS subline cells were injected subcutaneously into 5-week-old immunodeficient female nude mice (nu/nu) (Orient Bio Inc.). Tumor growth was monitored every 6 days after inoculation. Tumor volume (V) was measured using the modified ellipsoidal formula: V=1/2×length×(width)². At day 36, mice were exposed to 5-FU (50 mg/kg) by intratumoral injection and tumor growth was compared regularly. All animal studies were carried out with the approval (KUIACUC 2021-0044) of the Korea University Institutional Animal Care and Use Committee and the Korea Animal Protection Law.

10. Expression and Kaplan-Meier analysis

The data for the Kaplan-Meier curve analysis was obtained from web-based tool OSgc (https://bioinfo.henu.edu.cn/GC/GCList.jsp). The public RNA-seq data from pan-cancer patients (https://tnmplot.com/analysis/) were used to compare RASSF4 mRNA levels in normal (n=294) and cancer (n=375) patients. Log-rank p-value and hazard ratio were calculated with a Mantel-Cox and represented with 95% confidence intervals.

11. Statistical analysis

All experiments, including PCR, colony formation, [³H] thymidine incorporation, reporter luciferase and flow cytometry assays were repeated three times, and the results were presented as mean values±standard deviation (SD). Mean and SD values were calculated using Microsoft Excel software. A student’s t test (GraphPad Prism8 software) was performed to determine the statistical significance. A p-value of less than 0.05 was considered significant.

Results

1. RASSF4 expression is frequently lost or downregulated in gastric cancer

To explore the candidacy of RASSF4 as a suppressor in gastric tumorigenesis, we initially examined its expression status in cancer cell lines. RT-PCR and IB assays showed that RASSF4 mRNA is not expressed or very low in eight of 13 cancer cell lines and its level is well consistent with protein expression level (Fig. 1A). Meanwhile, a DNA-PCR analysis revealed that none of the cell lines, including four non-expressors (SNU484, SNU638, MKN28, and MKN74), have detectable reduction of the gene level, supporting that altered expression of RASSF4 results from transcriptional downregulation rather than gene deletion (Fig. 1A). Next, we examined RASSF4 expression in 20 primary carcinomas, including 15 matched sets from the same patients. While all noncancerous tissues we tested showed easily detectable levels of RASSF4 mRNA, a substantial fraction of cancer tissues displayed markedly reduced expression (Fig. 1B). Moreover, tumor-specific reduction was commonly detected in matched sets (Fig. 1C, S1A Fig.). We arbitrarily classified expression levels less than a half (2.03) of noncancerous means (4.06) as abnormal expression. On this basis, 75% (15 of 20) of primary carcinomas and 61.5% (8 of 13) of cancer cell lines were classified as abnormally low expressors (Fig. 1D). Consistently, analysis of the public RNA-seq data from pancancer patients (https://tnmplot.com/analysis/) revealed a significant reduction of RASSF4 mRNA in gastric tumors compared to normal tissues (Fig. 1E). To address if reduced expression is due to promoter hypermethylation, low expressor cell lines were treated with the demethylating agent 5-Aza-dC for 48 hours. RASSF4 mRNA level was profoundly elevated in all treated cells (Fig. 1F, S1B Fig.). On this basis, we performed a bisulfite DNA sequencing analysis for eight CpG sites within the CpG-enriched exon 1 region (+204 to +256 relative to transcription start site) and identified a tight correlation of methylation contents of the CpG sites with mRNA expression levels (Fig. 1G, S1C Fig.). Analysis of public database (https://depmap.org/portal/gene) also showed that RASSF4 mRNA expression is inversely associated with expression of DNA methyltransferase DNMT3A (S1D Fig.). Together, these indicate that RASSF4 transcription is commonly inactivated by aberrant CpG sites hypermethylation in gastric cancer cells.

Fig. 1.

RASSF4 expression in gastric cancer cell lines and primary tumors. (A) Expression and genomic status of RASSF4 in 13 cancer cell lines. Reverse transcription-polymerase chain reaction and DNA-PCR analyses were performed to determine the transcript and gene levels of RASSF4, respectively. Immunoblot (IB) assay was carried out to detect RASSF4 protein expression. (B) RASSF4 mRNA expression in normal and carcinoma tissues. N, normal tissue; T, tumor tissue. (C) Comparison of RASSF4 mRNA levels in matched cancer and adjacent noncancerous tissue sets obtained from the same cancer patients. (D) Expression levels of RASSF4 mRNA in gastric tissues and cell lines. Bar indicates the mean expression level of each specimen group. Data represent the mean±standard deviation of triplicate assays. ^**p < 0.01 (Student’s t test). (E) Analysis of RNA-Seq data from pan-cancer patients showing a significant reduction of RASSF4 mRNA level in gastric cancer versus normal tissues. (F) Reactivation of RASSF4 mRNA expression in cancer cells exposed to 5-Aza-dC for 48 hours. (G) Methylation status of promoter CpG sites within the exon 1 region and its association with mRNA levels. The transcription start site is indicated by an arrow at +1. Five plasmid clones of PCR products were sequenced for each specimen. Black, gray, and white squares represent complete (4-5 clones), partial (1-3 clones), and no (0 clones) methylation, respectively. mCpG, methylated CpG site.

2. RASSF4 inhibits cell proliferation by inducing G2/M phase cell cycle arrest

To understand the tumor suppression function of RASSF4, we examined its effect on cell proliferation and apoptosis. Cell growth analysis using ATCC (https://www.atcc.org) and Cellosaurus (https://www.cellosaurus.org) database revealed that growth rates (doubling time) of the 13 cancer cell lines we tested are associated with RASSF4 mRNA levels (Fig. 2A, S2 Fig.). Next, we examined RASSF4’s growth-inhibitory effect using AGS (low expressor) and MKN1 (high expressor) cell lines. Cell number counting assay revealed that cell growth is reduced by ectopic overexpression of RASSF4 while it is enhanced by siRNA-mediated depletion of RASSF4 (Fig. 2B). IB assay showed that cyclin D1 level is down- and upregulated by RASSF4 expression and depletion, respectively (Fig. 2C). [³H]thymidine incorporation assay also revealed that DNA replication is decreased by RASSF4 expression and increased by RASSF4 depletion (Fig. 2D). Flow cytometric analysis identified that RASSF4 expression induces an accumulation of the G2/M phase cells whereas its knockdown promotes cell cycle progression (Fig. 2E and F). Finally, clonogenic assay revealed that the colony-forming capability of tumor cells is significantly down- and upregulated by RASSF4 expression and depletion, respectively (Fig. 2G). Together, these indicate that RASSF4 exerts anti-proliferative effect by inducing a G2/M phase cell cycle arrest.

Fig. 2.

RASSF4 suppression of tumor cell growth. (A) Association of growth rate with RASSF4 mRNA level in 13 gastric cancer cell lines. Doubling time (hr) of the cell lines were obtained from American Type Culture Collection and Cellosaurus database. r, Pearson’s correlation coefficient. (B) Effect of RASSF4 expression on cell growth. AGS and MKN1 cells were transfected with increasing doses of RASSF4 vector and siRASSF4, respectively. Cell numbers were counted at 24-hour intervals for 4 days. Data represent the mean±standard deviation of triplicate assays. (C) Immunoblot showing RASSF4 effect on cyclin D1 expression. (D) [³H]Thymidine uptake analysis of RASSF4 effect on DNA synthesis. Cells transfected with RASSF4 vector or siRASSF4 were pulse-labeled for 4 hours with [³H]thymidine. Data represent the mean±standard deviation of triplicate assays. (E) RASSF4 expression in stable overexpression and knockdown subline cells. (F) Flow cytometric analysis of cell cycle progression of AGS and MKN1 sublines. (G) Effect of RASSF4 expression on colony-forming ability of tumor cells. AGS and MKN1 subline cells were maintained in the presence of hygromycin B for 7 days. Assays were performed in triplicate and the average number of colonies and SD were calculated. ^*p < 0.05, ^**p < 0.01 (Student’s t test).

3. RASSF4 is induced by genotoxic stress to promote apoptosis

Next, we asked if RASSF4 affects tumor cell response to cytotoxic therapeutic drugs. Intriguingly, a strong induction of RASSF4 mRNA expression was observed in AGS cells exposed to 5-FU, cisplatin, adriamycin, and etoposide (Fig. 3A, S3 Fig.). In response to 5-FU and cisplatin, RASSF4 exhibited a dose-associated induction that was accompanied with cl-PARP elevation, suggesting its pro-apoptotic role (Fig. 3A and B). As predicted, siRNA-mediated blockade of RASSF4 induction markedly attenuated apoptosis induction induced by 5-FU and cisplatin while its overexpression further stimulates apoptosis triggered by low dose of drugs (Fig. 3C-F). Likewise, RASSF4 depletion by either transient or stable knockdown profoundly reduced the apoptotic response of MKN1 cells to 5-FU (Fig. 3G and H). These results indicate that RASSF4 expression is induced by cytotoxic chemotherapeutic drugs to promote stress-induced apoptosis.

Fig. 3.

Apoptosis-promoting effect of RASSF4. (A, B) Induction of RASSF4 mRNA and protein in AGS cells exposed to chemotherapeutic drugs for 48 hours. (C, D) Immunoblot and flow cytometry assays showing blockade of 5-fluorouracil (5-FU)–mediated induction of cleaved poly(ADP-ribose) polymerase (cl-PARP) and sub-G1 fraction by RASSF4 depletion. Data represent the mean±standard deviation (SD) of triplicate assays. (E, F) Apoptosis-promoting activity of RASSF4 in AGS cells. RASSF4-transfected cells were exposed to 5-FU or cisplatin as indicated. Apoptosis induction was determined by cl-PARP level and sub-G1 fraction. Data represent the mean±SD of triplicate assays. (G, H) Effect of RASSF4 depletion on 5-FU–induced apoptosis in MKN1 cells. Apoptosis induction was determined by cl-PARP level and sub-G1 fraction after 72-hour exposure to 5-FU (20 μM). Data represent the mean±SD of triplicate assays. ^*p < 0.05, ^**p < 0.01 (Student’s t test).

4. RASSF4 activates p53 signaling to suppress tumor cell growth

Given that RASSF1A regulates cell proliferation and apoptosis by preventing MDM2-mediated p53 degradation, we asked whether RASSF4-induced apoptosis also occurs through p53 activation [25,26]. We initially compared RASSF4’s apoptotic effect between wild-type (wt) p53 cells (AGS and SNU719) and p53-null (SNU620 and KATOIII) cells. Under genotoxic stress conditions, ectopic overexpression of RASSF4 evoked markedly higher apoptosis-stimulating effect in wtp53 cells compared to p53-null cells (Fig. 4A). To elicit whether RASSF4-driven apoptosis is linked to p53 status, we examined its effect in p53-depleted AGS and p53-restored KATOIII cells. RASSF4’s apoptotic activity was substantially attenuated in AGS by p53 depletion while it was strongly elevated in KATOIII by p53 restoration (Fig. 4B). Likewise, RASSF4-induced growth inhibition was attenuated in AGS by p53 depletion (S4A Fig.). Assays using p53^+/+ and p53^–/– sublines of HCT116 colon cancer cells revealed that both anti-proliferative and pro-apoptotic effects of RASSF4 are evoked in a highly p53-dependent manner (S4B and S4C Fig.). Interestingly, both endogenous and transfected p53 levels were profoundly increased by RASSF4, suggesting that RASSF4 may activate p53 (Fig. 4C and D). Quantitative RT-PCR analysis also showed that RASSF4 increases mRNA expression of p53 target genes, including CDKN1A (p21^WAF1), RRM2B (p53R2), BAX, BBC3 (PUMA), PMAIP1 (NOXA), and XAF1, and this activity of RASSF4 is impaired and reinforced by p53 depletion and restoration, respectively (Fig. 4E and F). CDKN1A and BAX are representative transcriptional targets of p53, which play crucial roles in p53-induced cell cycle arrest and apoptosis. We assessed whether RASSF4 affects the promoter activity of CDKN1A and BAX using CDKN1A-Luc and BAX-Luc reporters, which comprise the p53 response element. In p53-null KATOIII cells, RASSF4 expression increased the luciferase activity of both CDKN1A-Luc and BAX-Luc and this effect was drastically elevated in the presence of p53 (Fig. 4G). Together, these results indicate that the pro-apoptotic and anti-proliferative functions of RASSF4 stem, at least in part, from its p53-activating property.

Fig. 4.

A p53-dependency of RASSF4-induced apoptosis. (A) Comparison of RASSF4’s apoptosis-promoting activity between wtp53 and p53-null or mtp53 cell lines. Cells transfected with RASSF4 were exposed to 5-fluorouracil (5-FU) for 48 hours and apoptosis was determined by flow cytometric measurement of sub-G1 fraction. Data represent the mean±standard deviation (SD) of triplicate assays. (B) p53 effect on RASSF4 stimulation of apoptosis. p53 was depleted and restored in AGS and KATOIII cells, respectively and the cells were exposed to 5-FU (5 μM, 48 hours) for apoptosis induction. Data represent the mean±SD of triplicate assays. (C, D) Immunoblot assay showing the p53-dependency of RASSF4-mediated apoptosis. cl-PARP, cleaved poly(ADP-ribose) polymerase. (E, F) Reverse transcription–polymerase chain reaction analysis showing RASSF4 enhancement of 5-FU–induced p53 target expression. (G) RASSF4 activation of the CDKN1A-Luc and BAX-Luc reporters containing the p53 response element. KATOIII cells were co-transfected with 1 μg of p53 and RASSF4 as indicated. The cells were transfected with the reporters and exposed to 5-FU (5 μM, 24 hours). Data represent the mean±SD of triplicate assays. ^*p < 0.05, ^**p < 0.01 (Student’s t test).

5. RASSF4 stabilizes p53 by preventing MDM2-mediated ubiquitination

To understand the molecular mechanism underlying RASSF4 activation of p53, we tested if RASSF4 affects the protein stability of p53. IB and RT-PCR assays of wtp53 cells (AGS and SNU719) revealed that RASSF4 increases p53 protein but not mRNA in a dose-associated manner (Fig. 5A). As predicted, a higher level of p53 protein was detected in AGS-RASSF4 compared to AGS-pcDNA and MKN1-shControl compared to MKN1-shRASSF4 (Fig. 5B). A CHX chase experiment revealed that p53 degradation is inhibited by RASSF4, indicating that RASSF4 stabilizes p53 by preventing its degradation (Fig. 5C and D). Moreover, RASSF4 depletion-induced p53 reduction was impeded by MG132 (proteasome inhibitor) but not affected by leupeptin (lysosome inhibitor), supporting that RASSF4 inhibits the ubiquitin-mediated proteasomal degradation of p53 (Fig. 5E). Given that RASSF1A activates p53 by destabilizing MDM2, a key ubiquitin E3 ligase against p53, we asked whether RASSF4 also has an activity to downregulate MDM2 [25,26]. However, RASSF4 showed no detectable effect on MDM2 protein level (Fig. 5F). However, immunoprecipitation assay revealed that MDM2-mediated p53 ubiquitination is impeded by RASSF4 (Fig. 5G). To elucidate the mechanistic basis for the RASSF4 protection of p53 from MDM2-mediated ubiquitination, we tested if RASSF4 binds to either p53 or MDM2 to interfere with p53-MDM2 interaction but failed to detect its interaction with p53 and MDM2 (S5A and S5B Fig.). Nevertheless, MDM2 binding to p53 was shown to be attenuated by RASSF4 expression (Fig. 5H). Moreover, in response to 5-FU exposure, MDM2 interaction with p53 was more profoundly reduced in AGS-RASSF4 compared to AGS-pcDNA (Fig. 5I). Together, these indicate that unlikely RASSF1A, RASSF4 does not interact with and destabilize MDM2, suggesting that RASSF4 may regulate upstream signaling molecule(s) involved in p53 activation.

Fig. 5.

RASSF4 stabilization of p53. (A) RASSF4 upregulation of p53 protein level. AGS and SNU719 cells were transfected with increasing doses of RASSF4 and its effect on p53 expression was determined by reverse transcription–polymerase chain reaction (RT-PCR) and immunoblot assays. (B) Comparison of p53 protein levels in stable RASSF4 expression and knockdown subline cells. (C, D) A cycloheximide (CHX) chase experiment showing the p53-stabilizing effect of RASSF4. AGS-pcDNA and AGS-RASSF4 subline cells were exposed to CHX (40 μM) for indicated times. (E) Impairment of RASSF4 depletion–driven p53 reduction by MG132. AGS-RASSF4 cells were transfected with siRASSF4 and exposed to MG132 (10 μM) or leupeptin (10 μM) for 6 hours. (F) No effect of RASSF4 on MDM2 protein expression. RASSF1A was transfected for comparison. (G) Immunoprecipitation (IP) assay showing RASSF4 inhibition of MDM2-mediated p53 ubiquitination. Cells were transfected with Flag-MDM2 and/or RASSF4-V5 as indicated. The transfected cells were incubated with MG132 for 6 hours before harvest. IB, immunoblot; Ub, ubiquitin; WCL, whole cell lysate. (H) IP assay showing RASSF4 attenuation of p53-MDM2 interaction. (I) Comparison of p53-MDM2 interaction in AGS-pcDNA and AGS-RASSF4 cells following 5-fluorouracil (5-FU) treatment.

6. RASSF4 stimulates Chk2-mediated p53 activation

p53 is stabilized by phosphorylation by multiple kinases, including ATM, ATR, and Chk1/2 [27]. We thus assessed if RASSF4 affects p53 stabilization by these kinases. In both AGS and SNU719 cells, RASSF4 transfection led to a dose-associated increase in Chk2 phosphorylation while it did not influence phosphorylation of ATM and Chk1 (Fig. 6A and B). Consistently, a higher level of Chk2 phosphorylation was detected in AGS-RASSF4 versus AGS-pcDNA subline and in MKN1-shControl versus MKN1-shRASSF4 subline (Fig. 6C). Furthermore, RASSF4 induction of p53 and RASSF4 inhibition of cell growth were significantly impaired by Chk2 depletion, indicating the presence of the RASSF4-Chk2-p53 signaling axis (Fig. 6D and E). Likewise, Chk2 phosphorylation induced by 5-FU was attenuated by RASSF4 depletion while RASSF4 stimulation of 5-FU–induced apoptosis was abolished by Chk2 depletion (Fig. 6F-H). These indicate that RASSF4 induces Chk2 phosphorylation to activate p53.

Fig. 6.

RASSF4 activation of the Chk2-p53 axis. (A, B) RASSF4 induction of Chk2 phosphorylation. Cells were transfected with RASSF4 (1 μg) and its effect on ATM, Chk1, and Chk2 phosphorylation was determined by immunoblot assay. (C) Comparison of Chk2 phosphorylation level in RASSF4 expression and depletion subline cells. (D, E) Attenuation of RASSF4 induction of p53 and inhibition of cell growth by Chk2 depletion. AGS cells were transfected with siChk2 and RASSF4 as indicated. Cell numbers were counted at 72 hours after transfection. Data represent the mean±standard deviation (SD) of triplicate assays. (F) Blockade of 5-fluorouracil (5-FU)–induced Chk2 phosphorylation by RASSF4 depletion. cl-PARP, cleaved poly(ADP-ribose) polymerase. (G, H) Attenuation of RASSF4 stimulation of 5-FU–induced apoptosis by Chk2 depletion. AGS cells transfected with siChk2 and/or RASSF4 were exposed to 5-FU (5 μM, 48 hours). Data represent the mean±SD of triplicate assays. ^**p < 0.01 (Student’s t test).

7. RASSF4 suppresses gastric tumor growth

To elucidate a role for RASSF4 in tumor growth and response to chemotherapeutic drug, mouse tumor xenograft assays were performed using AGS-pcDNA and AGS-RAS-SF4 sublines. Both subline tumors were exposed to 5-FU at day 36 after inoculation and tumor growth was compared. As predicted, a higher growth rate was observed in pcDNA versus RASSF4 subline tumors (Fig. 7A and B). Compared with AGS-pcDNA tumors, AGS-RASSF4 tumors displayed markedly higher regression rate following 5-FU injection (Fig. 7C and D). IB assay of tumor tissues detected higher levels of p53, Chk2 phosphorylation, and cl-PARP in RASSF4 versus pcDNA subline tumors, supporting the implication of the RASSF4-Chk2-p53 signaling axis in chemotherapy-induced tumor regression (Fig. 7C). Finally, a targeted prognostic analysis of gastric cancer patients using the TCGA database (https://bioinfo.henu.edu.cn/GC/GCList.jsp) revealed that RASSF4 expression is strongly associated with overall survival of cancer patients, further supporting its clinical significance in gastric cancer (Fig. 7D). Collectively, our data show that RASSF4 suppresses gastric tumorigenesis through activation of the Chk2-p53 axis (Fig. 7E). This study also uncovers that RASSF4-Chk2-p53 signaling represents a novel mechanism that dictates stress response, suggesting the implication of its alteration in the development of chemoresistant tumors.

Fig. 7.

RASSF4 enhancement of tumor response to therapeutic drug. (A, B) Mouse tumor xenograft assay showing RASSF4 effect on tumor response to 5-fluorouracil (5-FU). Tumors were derived from pcDNA and RASSF4 sublines of AGS cells and exposed to saline or 5-FU (50 mg/kg) at day 36 by intratumoral injection. Representative photographs of xenograft tumors at day 54 were shown. Data represent the mean± standard deviation (n=6 per group). ^*p < 0.05, ^**p < 0.01 (Student’s t test). (C) Comparison of Chk2 phosphorylation, p53, and cleaved poly(ADP-ribose) polymerase (cl-PARP) levels in AGS-pcDNA and AGS-RASSF4 tumors. (D) The Cancer Genome Atlas database analysis showing a significant association of RASSF4 expression with the overall survival of gastric cancer patients. RASSF4 high (n=711) and RASSF4 low (n=270) patients were analyzed. (E) Schematic representation of the signaling pathway underlying RASSF4 activation of the Chk2-p53 axis and its tumor-suppressive function. CI, confidence interval; HR, hazard ratio.

Discussion

Accumulating evidence indicate that epigenetic alteration of several RASSF family genes contributes to the development and malignant progression of a variety of human cancers [4-8]. Our previous study demonstrated that aberrant promoter hypermethylation of RASSF1A is associated with the malignant progression of gastric cancers [9]. Epigenetic inactivation of RASSF2, RASSF6, RASSF8, and RASSF10 has also been shown to associate with gastric tumor progression and poor survival of cancer patients [20-23]. However, their roles as tumor suppressors in the pathogenesis of gastric cancer remain largely undefined.

RASSF4 is a putative tumor suppressor gene located on chromosome 10q11.21 and its expression is frequently inactivated in multiple human malignancies [13-17]. Although altered expression of RASSF4 mRNA was observed in some gastric cardia adenocarcinoma, its candidacy as a tumor suppressor in gastric tumorigenesis has not been explored yet [21]. In the present study, we identified that RASSF4 mRNA expression is abnormally downregulated in eight of 13 (61.5%) cancer cell lines and 15 of 20 (75%) of primary carcinomas, indicating that RASSF4 inactivation is a common event in gastric tumorigenesis. Moreover, bisulfite DNA sequencing analysis showed that the methylation status of eight CpG sites within the exon 1 region is tightly associated with mRNA expression levels. Therefore, our study verifies that RASSF4 undergoes epigenetic transcriptional silencing in a considerable proportion of gastric carcinomas by aberrant CpG sites hypermethylation.

RASSF4 shares approximately 60% identity with RASSF2 and 25% identity with RASSF1 and interacts with activated Ras through its effector domain [13]. Despite its emerging role as a growth inhibitor, the biological characteristics of RASSF4 are poorly defined. Studies showed that RASSF4 inhibits malignant behavior through YAP/Bcl-2 signaling in colorectal cancer cells and induces G2-phase cell cycle arrest and apoptosis by linking Ras to several pro-death pathways in multiple myeloma cells [15,16]. Recently, RASSF4 was identified to interact with MST1 to inhibit YAP nuclear translocation through the Hippo pathway [17]. In this study, we observed that RASSF4 suppresses cell proliferation by inducing a G2/M phase cell cycle arrest, which is accompanied with reduction of cyclin D1 and DNA synthesis. Interestingly, RASSF4 mRNA expression is induced in response to cytotoxic chemotherapeutic drugs, such as 5-FU, cisplatin, adriamycin, and etoposide, and blockade of its induction attenuates drug-induced apoptosis. Given that RASSF family members, including RASSF1, RASSF3, and RASSF5, are upregulated by p53, it is plausible that RASSF4 is also activated by p53. However, our observation of its induction in mtp53 cells as well as wtp53 cells suggests that multiple transcription factors including p53 might be involved in the transcriptional activation of RASSF4 under stressful conditions. Although additional studies are required to define whether RASSF4 regulation by DNA-damaging agents is governed by methylation, our finding suggests that epigenetic inactivation of RASSF4 might be implicated in the development of tumor resistance to therapeutic drugs. In this context, RASSF4 methylation could potentially serve as a predictive biomarker for chemotherapy response in gastric cancer patients.

It is well documented that under physiological conditions, p53 is destabilized by the ubiquitin E3 ligase MDM2 while in response to various stresses, p53 is phosphorylated by multiple kinases, such as ATM, ATR, and Chk1/2, to protect from MDM2-mediated ubiquitination [27,28]. RASSF1A stabilizes p53 by promoting the self-ubiquitination of MDM2 through the interaction with DAXX and activates the p53-p21^WAF1 pathway by promoting the proteasomal degradation of MDM2 [25,26]. In this study, we found that RASSF4 regulates cell proliferation and apoptosis in a highly p53-dependent manner. Moreover, RASSF4 induction causes p53 stabilization, thereby upregulating expression of p53 target genes, including CDKN1A, BAX, BBC3, and PMAIP1. However, unlikely RASSF1A, RASSF4 does not interact with and destabilize MDM2 but attenuates its binding to p53, raising the possibility that RASSF4 activates upstream kinases of p53. Indeed, we found that RASSF4 increases Chk2 phosphorylation and stabilizes p53 in a Chk2-dependent fashion, identifying RASSF4 as a novel activator of the Chk2-p53 axis and elucidating the mechanism underlying RASSF4-mediated tumor suppression. Chk2 is phosphorylated mainly by ATM kinase in response to DNA double-strand breaks, and RASSF1A is well characterized to augment ATM, thereby reinforcing the DNA damage response [27,29]. However, we failed to detect ATM elevation or phosphorylation by RASSF4, suggesting an ATM-independent link between RASSF4 and Chk2. Considering that RASSF family proteins do not possess kinase activity and that RASSF4 is predicted to act as a scaffolding protein, it is hypothesized that RASSF4 recruits signaling factors capable of activating Chk2 [15]. In this context, it is noteworthy that RASSF4 activates the JNK/c-Jun pathway via the interaction with the Hippo kinase MST1 and thereby increases H2B and H2AX phosphorylation, resulting in DNA fragmentation and apoptosis [15]. Given that MST1/2 are also known to cooperate with the Chk1 and Chk2 pathways to regulate the DNA damage response, it is likely that RASSF4 may activate Chk2 through the Hippo and JNK pathways [30].

Genomic integrity is critical for preventing tumor development. Among a variety of intricate mechanisms for maintaining genome stability, the Chk2-p53 axis plays a pivotal role in detecting DNA double-strand breaks and facilitating their repair [31]. In many cancer cells, the DNA damage response is often impaired due to the mutational alteration or dysregulation of the Chk2-p53 axis, contributing to unchecked cell proliferation and tumor progression [27,32]. Therefore, the molecular understanding of the functionality of this pathway could provide valuable insights into the mechanisms underlying tumor progression and highlight potential therapeutic targets that could improve the effectiveness of cancer treatments. Cisplatin and 5-FU are well-established chemotherapeutic agents for gastric cancer treatment while etoposide and adriamycin hold potential for synergy with novel therapeutic strategies. In this context, our study suggests that RASSF4-mediated activation of the Chk2-p53 axis through RASSF4 demethylation may open new therapeutic avenues for gastric cancer.

In conclusion, we demonstrate that RASSF4 functions as a tumor suppressor in gastric tumorigenesis through activation of the Chk2-p53 axis, establishing that RASSF4 represents one critical activator of p53 signaling. Although the physiologic role of the RASSF4-Chk2-p53 signaling pathway is only beginning to be determined, the evidence we obtained here illuminates the mechanistic consequence of its alteration in gastric tumorigenesis.

Electronic Supplementary Material

Supplementary materials are available at Cancer Research and Treatment website (https://www.e-crt.org).