Inflammation is a crucial driving force in the development of gastric cancers (GCs). Accordingly, persistent activation of STAT3, a transcription factor pivotal in regulating both inflammation and oncogenesis, is often detected in GC, although its mechanism remains elusive. Suppressor of cytokine signaling-1 (SOCS-1) is a negative regulator of proinflammatory cytokine signaling and SOCS-1 gene methylation is frequently detected in various cancers including GC. However, the significance of SOCS-1 methylation in GC cells remains unexplored. Our study is undertaken to evaluate the role of SOCS-1 in GC cell proliferation and its effect on signaling pathways in GC cells. Among five GC cell lines, SOCS-1 gene was methylated in all cell lines and constitutive STAT3 phosphorylation with elevated endogenous IL-6 production was detected in two cell lines (NUGC-3 and AGS). Unexpectedly, anti-IL-6R antibody inhibited neither cell proliferation nor STAT3 phosphorylation in NUGC-3 and AGS. In contrast, enforced SOCS-1 expression by adenoviral vector (AdSOCS-1) markedly suppressed STAT3 phosphorylation and proliferation of NUGC-3 and AGS cells in vitro. Interestingly, the antiproliferative effect of SOCS-1 was attributable not only to the inhibition of STAT3 but also to that of p38 MAPK activity, and chemical inhibitors of JAK/STAT and p38 MAPK signaling effectively suppressed proliferation of these GC cells. Furthermore, treatment with AdSOCS-1 in vivo significantly suppressed GC proliferation in a xenograft model. These results suggest that SOCS-1 gene methylation is a critical step in the development of GC, and enforced expression of SOCS-1 may represent a novel therapeutic approach for the treatment of GC.
Gastric cancer (GC) is the second most common cause of cancer deaths worldwide.1 Recent diagnostic and therapeutic advances have significantly improved prognosis for patients with early GC. However, an effective treatment for patients with advanced GC has not yet been established and prognosis remains poor.2
It has become evident that inflammatory responses can promote cancer development.3 An important factor in gastric carcinogenesis is persistent inflammation in gastric epithelium due to infection of H. pylori.4 Critical contributors involved in both inflammation and tumorigenesis are proinflammatory cytokines and their downstream signaling pathways. Indeed, dysregulated activation of the JAK/STAT signaling pathway, the major downstream pathway of cytokines such as IL-6, has been detected in various cancers including GC.5–10 In particular, constitutive activation of STAT3, an important mediator of both proinflammatory and oncogenic signals of IL-6 family cytokines, is linked to inflammation-associated tumorigenesis6–8, 11 and has been detected in ∼30% of primary GC.12 In addition, activation of other signaling pathways, such as mitogen-activated protein kinase (MAPK) and phosphoinositide 3 kinase (PI3K) pathways, downstream of proinflammatory cytokines are also involved in the progression of GC.13 However, the mechanisms for dysregulated activation of these signaling pathways in cancer cells are largely unknown.
Under homeostatic conditions, cytokine signaling pathways are tightly controlled by negative regulatory mechanisms. The most representative of these mechanisms is the induction of suppressors of cytokine signaling (SOCS) family proteins, which act in a feedback loop to inhibit cytokine responses by terminating the activation of the JAK/STAT and other signaling pathways.14–16 The SOCS family, characterized by a central src homology 2 (SH2) domain and a conserved C-terminus SOCS box, is composed of eight structurally related proteins. Among these, SOCS-1 is known as the most potent negative regulator of proinflammatory cytokine signaling.17 SOCS-1 interacts with phosphorylated tyrosine residues on proteins such as JAK kinases18, 19 to interfere with the activation of STAT proteins or other signaling intermediates. SOCS-1 also recruits the elongin BC-containing E3 ubiquitin-ligase complex via the conserved SOCS box to promote the degradation of target proteins.20 Studies on SOCS-1 deficient mice have indicated that SOCS-1 is essential for the inhibition of excessive immune responses and also are involved in the suppression of tumor development.17, 21 In accordance with this notion, epigenetic silencing of SOCS-1 by methylation of the CpG island is detected in human cancers, such as hepatocellular carcinoma (HCC), multiple myeloma and pancreatic ductal neoplasm22–25 and is implicated in cancer development.
Like other cancers, a variety of epigenetic alterations are involved in the development of GC.26–30 Two groups have recently reported that transcriptional inactivation of SOCS-1 gene by hypermethylation is frequently observed in GC cell lines31 and primary GC samples.32, 33 In particular, Oshimo et al. have reported that SOCS-1 gene hypermethylation is not detectable in normal gastric mucosa but is detected in 44% of primary GC tissues and 12% of corresponding non-neoplastic mucosa and is correlated with the progression and lymph node metastasis of GC.33 However, it remains to be clarified whether the inactivation of SOCS-1 gene is truly important for the oncogenesis of GC or which signaling pathways targeted by SOCS-1 are important for GC cell proliferation.
In our study, we demonstrate that SOCS-1 is silenced in GC cell lines and is involved in enhanced STAT3 activation in these cells. We also demonstrate that gene delivery of SOCS-1 in GC cells has a potent antiproliferative effect via the suppression of not only JAK/STAT activation but also inhibition of p38 MAPK signaling pathway. Our results provide new insights into the pathogenesis of GC and may highlight potential molecular targets for therapeutic intervention in patients with GC.
GC: gastric cancer; IL-6R: interleukin-6 receptor; JAK: Janus kinase; p38 MAPK: p38 mitosis-activated protein kinase; SOCS: suppressor of cytokine signaling; STAT3: signal transducer and activator of transcription 3
Material and Methods
Four human GC cell lines NUGC3 (JCRB0822), MKN45 (JCRB0254), NUGC4 (JCRB0834) and MKN7 (JCRB1025) were obtained from the Japanese Collection of Research Bioresources (Osaka, Japan), and AGS was purchased from the American Type Culture Collection (ATCC, Manassas, VA). All cell lines were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) (HyClone Laboratories, Logan, UT) and 1% penicillin–streptomycin (Nacalai Tesque, Kyoto, Japan) at 37°C under a humidified atmosphere of 5% CO2.
Enzyme-linked immunosorbent assay
Cell lines were cultured in six-well plates at a density of 1 × 105 cells per well and incubated in RPMI 1640 containing 1% FCS. The concentrations of IL-6, soluble IL-6 receptor (sIL-6R) in the cell culture supernatant was measured at 48-hr time points using Quantikine enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN) according to the manufacturer's instructions. The ELISA sensitivities for the detection of IL-6 and sIL-6R, as reported by the manufacturer, were 0.7 and 6.5 pg/ml, respectively.
IL-6 and anti-IL-6R antibody treatment of GC cells
After 24 hr of serum starvation, GC cell lines were treated with 20 ng/ml of recombinant IL-6 (PeproTech, Rocky Hill, NJ) and 20 ng/ml of sIL-6R (R&D Systems) and proteins were extracted 15 min after IL-6 stimulation for further analysis. For antibody treatment, 25 and 50 μg/ml of MRA (humanized monoclonal anti-human IL-6R antibody; Chugai Pharmaceutical Co., Tokyo, Japan) was added to cell culture medium with recombinant 20 ng/ml IL-6 and 20 ng/ml sIL-6R. Purified human IgG (Sigma, St. Louis, MO) was used as control. Cells were then harvested for the determination of the phosphorylation status of STAT3.
Bisulfite modification of genomic DNA was carried out as described previously.22, 31, 33 The bisulfite-treated DNA was amplified with either a methylation-specific or unmethylation-specific primer set using EpiTect methylation-specific PCR (MSP) Kit (Qiagen, Valencia, CA) according to the manufacturer's instructions. The methylation-specific primer sequences in the exon 1 of SOCS-1 CpG island were 5′-TC GTTCGTACGTCGATTATC-3′ (forward) and 5′-AAAAAA ATACCCACGAACTCG-3′ (reverse); sequences of corresponding unmethylation-specific primer sequences were 5′-TATTTTGTTTGTATGTTGATTATTG-3′ (forward) and 5′-AAACTCAACACACAACCACTC-3′ (reverse). The length of PCR products were 132 bp for methylated reaction, and 122 bp for the unmethylated reaction. PCR products were resolved in 2.5% agarose gels, stained with ethidium bromide and visualized under UV illumination. To ensure that the PCR primers are specific for the detection of methylated or unmethylated bisulfite converted DNA, completely methylated or unmethylated bisulfite converted DNAs, and untreated, unmethylated genomic DNA (EpiTect control DNA, Qiagen), were used for control experiments.
Real-time PCR analysis
After 12 hr of serum starvation, GC cell lines (NUGC3, AGS, MKN45, NUGC4 and MKN7) and human PBMC were treated with 10 ng/ml of recombinant human IFN-γ (PeproTech, Rocky Hill, NJ) for 15 min. Total RNA was prepared from cells using an RNeasy Mini Kit (Qiagen) and cDNAs were synthesized from 500 ng of each total RNA preparation using a Quantitect Reverse Transcription Kit (Qiagen), all according to the manufacturers' instructions. The forward and reverse primers were as follows: for human SOCS-1 forward primer, 5′-AGACCCCTTCTCACCTCTTG-3′ and reverse primer, 5′-GCACAGCAGAAAAATAAAGC-3′; for β-actin, 5′-GTGGGGCGCCCCAGGCACCA-3′ and 5′-CTCC TTAATGTCACGCACGATTTC-3′.34 Primers and cDNA were added to SYBR green premix (Invitrogen), which contained all the reagents required for PCR. The PCR conditions of SOCS-1 consisted of 1 cycle at 95°C for 10 min followed by 40–50 cycles of 96°C for 10 sec, 68°C for 15 sec and 72°C for 15 sec; β-actin cycling conditions consisted of 1 cycle at 95°C for 10 min followed by 40–50 cycles of 96°C for 10 sec, 67°C for 30 sec and 72°C for 30 sec. PCR products were measured continuously using the My IQ™ Single-Color Real-Time Detection System (Bio-Rad Laboratories).
Replication-defective recombinant adenoviral vector expressing the mouse SOCS-1 gene was provided by Dr. Hiroyuki Mizuguchi (Osaka University, Osaka, Japan), which was constructed by an improved in vitro ligation method, as described previously.35, 36 An adenoviral vector expressing the LacZ gene was constructed using similar methods. Expression of these genes was regulated by CMV promoter/enhancer and intron A. The viruses were amplified in 293 cells. Viruses were purified by CsCl2 step gradient ultracentrifugation followed by CsCl2 linear gradient ultracentrifugation. The purified viruses were dialyzed against a solution containing 10 mM Tris-HCl (pH 7.5), 1 mM MgCl2 and 10% glycerol and were stored at −80°C. Viral particle and biological titers were determined by a spectrophotometrical method37 and by using QuickTiter (Adenovirus Titer Immunoassay Kit, Cell Biolabs, San Diego, CA), respectively. After 24-hr incubation of GC cells in culture medium containing 10% FCS, adenoviral vectors were infected by distributing suspensions onto cells at a multiplicity of infection (MOI) of 10–160. Adenoviral vectors containing the genes for HA-tagged Y705F dominant-negative STAT3 (AddnSTAT3) were kindly provided by Dr. Akihiko Yoshimura (Keio University, Tokyo, Japan).
Cells and tumor tissues from xenograft model were lysed in RIPA buffer (10 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 0.1% sodium deoxycholate, 0.1% SDS, 1× phosphatase inhibitor cocktail (Nacalai Tesque) and 1× protease inhibitor cocktail (Nacalai Tesque) followed by centrifugation (13,200 rpm, 4°C, 15 min), after which the supernatants were stored at −80°C until use. Protein concentrations were determined with the DC Protein Assay kit (Bio-Rad Laboratories), using BSA as the concentration standard. Extracted proteins were used for SDS-PAGE or immunoprecipitation assay. For immunoprecipitation assay, extracted proteins were incubated with primary antibody coupled Protein G-Sepharose (BioVision, Mountain View, CA) for several hours at 4°C with rotating. The samples were washed several times with RIPA buffer, and proteins were extracted using SDS sample buffer (0.125 M Tris-HCl, pH 6.8, 10% 2-mercaptoethanol, 4% SDS, 10% glycerol and 0.004% bromophenol blue). Proteins were resolved using 5–20% gradient SDS-PAGE gels (Wako Pure Chemical Industries, Osaka, Japan) and subsequently transferred to PVDF membranes (Millipore, Bedford, MA). The membranes were blocked with 1% BSA in PBS containing 0.1% Tween 20 (PBST) and incubated with the respective antibodies against different targets. The following antibodies were used: anti-SOCS-1, 1:500 (IBL, Fujioka, Japan), anti-HA (Sigma), anti-phospho-STAT3, 1:1,000 (Cell Signaling Technology, Danvers, MA); anti-STAT3, 1:1,000; anti-GAPDH, 1:2,000 (all from Santa Cruz Biotechnology, Santa Cruz, CA).
Next, the membranes were incubated with horseradish peroxidase-conjugated sheep anti-mouse IgG or horseradish peroxidase-conjugated donkey anti-rabbit IgG (GE Healthcare, Little Chalfont, Buckinghamshire, UK). Finally, the signals were visualized by means of an enhanced chemiluminescence (ECL) reaction system (Perkin-Elmer Life Sciences, Boston, MA).
Cell proliferation assay
NUGC3 and AGS cells were plated in 96-well plates at a density of 5 × 102 cells per well and incubated in RPMI 1640 supplemented with 10% FCS for 24 hr. Then, cells were treated with 2.5 μM (NUGC3) and 5.0 μM (AGS) of JAK inhibitor I (Jak inhibitor; Calbiochem, San Diego, CA), with 10 μM (NUGC3) and 20 μM (AGS) of SB203580 (p38 MAP kinase inhibitor; Calbiochem), MRA (anti-human IL-6R antibody; Chugai Pharmaceutical Co.) and dimethyl sulfoxide (DMSO) or human IgG (Sigma) alone, followed by incubation at 37°C in RPMI 1640 supplemented with 5% FBS. Cell proliferation was evaluated with the WST-8 [2-(2-methoxy-4-nitrophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium, monosodium salt] assay (Cell Counting Kit-SF; Nacalai Tesque, Kyoto, Japan) at indicated period after treatment. The absorption of WST-8 was measured at a wavelength of 450 nm with a reference wavelength of 630 nm using a microplate reader Model 680 (Bio-Rad Laboratories, Hercules, CA). Growth rate was expressed as the percentage of absorbance reading for treated cells vs. control cells. Each value is the average ± standard deviation (SD) of triplicate wells.
Measurement of p38 MAP kinase and ERK activation
Kinase assays of p38MAPK were performed using commercial kits (Cell Signaling Technology). Cells were seeded 24 hr before the assay in a 100 × 100 mm2 polystyrene nonpyrogenic dish at a density of 5 × 105 cells per dish. Cells were harvested at 48 hr after infection of adenoviral vectors and lysed with buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1 mM Na3VO4, 1 mg/ml leupeptin and 1 mM PMSF). Five hundred microliters of cell lysates containing 200 μg total protein were incubated with 20 μL of immobilized phospho-p38MAPK monoclonal antibody for p38MAPK assay. The mixtures were then incubated at 4°C overnight and centrifuged to obtain a cell pellet. The pellet was then washed two times with wash buffer and further two times with a kinase buffer (5 mM Tris, pH 7.5, 10 mM MgCl2, 2 mM DTT and 0.1 mM Na3VO4). Kinase reactions were carried out in kinase buffer supplemented with 100 mM ATP and 2 μg activating transcription factor-2 (ATF-2) fusion protein as a substrate for p38MAPK assay. Reactions were performed at 30°C for 30 min and terminated by adding 5× SDS-PAGE sample buffer. Total ATF2 and phosphorylated substrates were analyzed by Western-blotting analysis using phospho-ATF-2 (Cell signaling Technology) antibody for kinase assay of p38MAPK.
Mouse xenograft model
All animal experiments were conducted according to the institutional ethical guidelines for animal experimentation of the National Institute of Biomedical Innovation (Osaka, Japan). Male ICR nu/nu mice, 4–5 weeks of age, were obtained from Charles River Japan (Yokohama, Japan). For subcutaneous xenograft experiments, we injected 2 × 106 cells in a total volume of 100 μL of 1/1 (v/v) PBS/Matrigel (Becton Dickinson, Bedford, MA) in the flank of ICR nu/nu mice. After 1 week when the tumor sizes reached to ∼100 mm3, 2 × 108 plaque-forming units (pfu)/50 μL of AdSOCS1 or AdLacZ was injected intratumorally twice per week. Tumor volumes were determined weekly by measuring in two dimensions, length (L) and width (W), and calculating volume as (W2 × L)/2.
Statistical analyses were performed using the StatView 5.0 software package (Abacus Concepts, Berkeley, CA). One-way ANOVA followed by a Scheffe's test or Mann–Whitney U tests were used to evaluate the significance of differences. In all analyses, p < 0.05 was considered statistically significant.
Methylation status of SOCS-1 CpG islands in GC cell lines
Since it was reported that SOCS-1 gene methylation is frequently observed in primary GC,33 we initially screened five GC cell lines for methylation of the SOCS-1 gene. By MSP analysis using primers selected from the CpG islands inside exon 1 of the SOCS-1 gene, we detected SOCS-1 methylation in all five (NUGC3, AGS, MKN45, NUGC4 and MKN7) GC cell lines (Fig. 1a). By real-time PCR analysis, we also found that these five GC cell lines, but not human PBMC, failed to upregulate SOCS-1 expression in response to IFN-γ, indicating that transcription of the SOCS-1 gene is inhibited by gene methylation (Fig. 1b).
Constitutive activation of STAT3 in GC cell lines
We then evaluated the activation status of signal transducer and activator of transcription 3 (STAT3), since STAT3 acts as an important transcriptional mediator of proinflammatory cytokine signaling pathways and contributes to oncogenesis by both preventing apoptosis and enhancing cell proliferation.11 STAT3 was constitutively phosphorylated in three of five GC cell lines, NUGC3, AGS and MKN45 cells (Fig. 1c), raising the possibility that epigenetic silencing of SOCS genes may lead to the aberrant activation of the JAK/STAT pathway in these GC cells. In other cell lines, STAT3 phosphorylation was induced strongly in NUGC4 cells after the stimulation with IL-6, but it was not induced in MKN7 cells (Fig. 1c).
No inhibitory effect of anti-IL-6R antibody on GC cell growth
Previous studies have revealed that IL-6 levels are elevated in cancer tissues38 and in sera39, 40 of GC patients. We next quantitated levels of IL-6 and soluble IL-6 receptor (sIL-6R) in 48-hr culture supernatants of GC cell lines by sandwich ELISA. As shown in Figure 2a, elevated IL-6 levels were observed in NUGC3 cells (1348.6 ± 91.2 pg/ml) and AGS cells (166.3 ± 68.1 pg/ml), while sIL-6R secretion was comparable among all the cell lines tested (Fig. 2b). These results suggest that spontaneous production of IL-6 may be crucial for aberrant STAT3 phosphorylation in NUGC3 and AGS cells.
We then assessed the impact of IL-6 blockade on STAT3 phosphorylation and proliferation of NUGC3 and AGS cells, which have demonstrated constitutive STAT3 activation concomitant with high level of IL-6 production (Figs. 1c and 2a). Western-blot analysis confirmed that humanized anti-IL-6R monoclonal antibody (MRA), which inhibits IL-6 function by competing for the membrane bound and the soluble forms of the human IL-6 receptor,41 effectively inhibited IL6-induced phosphorylation of STAT3 in NUGC4 cells. However, the same treatment could not reduce STAT3 phosphorylation in NUGC3 and AGS cells (Fig. 2b). In addition, cell proliferation assay using WST-8 showed that treatment with MRA did not suppress cell proliferation of IL6-producing GC cell lines (Fig. 2c). These in vitro results suggest that cell proliferation and constitutive STAT3 phosphorylation in NUGC3 and AGS cells occur independently of autocrine IL-6 production, and that signaling pathways other than IL-6 signaling may be involved in cell proliferation of these cells.
Antiproliferative effect of SOCS-1 gene delivery in GC cells
We next investigated the role of SOCS-1 in regulation of intracellular signaling cascades and GC cell proliferation. For this purpose, we used replication-defective recombinant adenoviral vectors carrying SOCS-1 in cell proliferation assays. Given the established role of STAT3 in tumor development, we also assessed the effect of the inhibition of STAT3-depended pathways, using an adenoviral vector expressing dominant negative STAT3 (dnSTAT3). Immunoblotting analysis showed that adenovirus-mediated gene delivery could induce dose dependent expression of SOCS-1 and dnSTAT3 in both NUGC3 and AGS cells (Fig. 3a). As shown in Figure 3b, WST-8 assay revealed that adenovirus-mediated SOCS-1 gene delivery markedly decreased cell proliferation of IL-6 producing NUGC3 and AGS cells. In addition, the magnitude of antiproliferative effects of AdSOCS-1 on GC cells was higher than that of AddnSTAT3. These results suggest that both SOCS-1 and dnSTAT3 can suppress proliferation of NUGC3 and AGS cells, but SOCS-1 is likely to exert additional effects on these GC cells.
Effect of SOCS-1 gene delivery on JAK/STAT3, MAPK and PI3K pathways
We next determined the activation status of signaling molecules in GC cells infected with AdLacZ, AdSOCS-1 and AddnSTAT3. As shown in Figure 4a, immunoblotting analysis showed that phosphorylation levels of STAT3 were effectively decreased in NUGC3 and AGS cells treated with either AdSOCS-1 or AddnSTAT3. Since AdSOCS-1 inhibited GC cell proliferation more effectively than AddnSTAT3, we next investigated if AdSOCS-1 also downregulates STAT3-independent signaling pathway(s) in these cells. Our screening analyses indicated that AKT, SAPK/JNK and p44/p42 MAPK pathways were not affected by AdSOCS-1 infection, although these molecules were constitutively phosphorylated in NUGC3 and AGS cell lines (data not shown). We therefore examined another MAPK pathway, the p38MAPK pathway, by evaluating kinase assays with ATF-2 as substrate. Interestingly, the activation of ATF2 in NUGC3 and AGS cells was reduced by AdSOCS-1 infection but not by AddnSTAT3 infection (Fig. 4b). Our results indicate that forced expression of SOCS-1 suppresses both JAK/STAT3 and p38 MAPK pathway in NUGC3 and AGS cells.
Inhibition of JAK kinase- and p38 MAP kinase-induced suppression of cell proliferation in GC cells
To confirm whether the activities of JAK/STAT and p38 MAP kinase signaling pathways regulate proliferation of these GC cancer cells, cells were treated with the JAK inhibitor (JAK Inhibitor I), p38 MAP kinase inhibitor (SB203580) or both. Cell proliferation assays revealed that JAK inhibitor I markedly suppressed proliferation in NUGC3 cells (Fig. 5a), and moderately suppressed proliferation in AGS cells (Fig. 5b). SB203580 suppressed cell proliferation of both NUGC3 and AGS cells more effectively than JAK inhibitor I (Figs. 5a and 5b). In both NUGC3 and AGS cells, combined treatment with JAK inhibitor I and SB203580 further suppressed the proliferation of these GC cell lines (Figs. 5a and 5b). Thus, our results suggest that both JAK/STAT3 and p38 MAPK signaling pathways play crucial roles in the proliferation of NUGC3 and AGS cells.
SOCS-1 exhibits antitumor activity in a GC xenograft model
We also evaluated the therapeutic effect of AdSOCS-1 injection on the growth of GC cells in vivo. For this purpose, we established a xenograft model of ICR nu/nu mice in which NUGC3 cells were subcutaneously implanted. Injection of AdSOCS-1 vector (1 × 108 pfu/50 μL) intratumorally twice per week suppressed tumor growth compared to tumor volumes in control AdLacZ-injected animals (Figs. 6a and 6b). AdSOCS-1 in vivo could modulate intracellular signaling in GC cells as in vitro, since Western-blot analysis showed that phosphorylation levels of STAT3 were decreased in the NUGC3 tissues from AdSOCS-1 injected animals (Fig. 6c).
Proinflammatory cytokines induced by H. pylori infection are critical for both chronic inflammation in gastric mucosa and the initiation and progression of GC. It is thus reasonable to speculate that the changes in sensitivity of gastric epithelial cells to proinflammatory cytokines greatly contribute to the development of GC. In our study, we provide evidence that silencing of SOCS-1, a negative regulator of cytokine signaling, is profoundly involved in the development of GC.
We identified SOCS-1 gene methylation in GC cell lines (Fig. 1a). This result is consistent with the previous findings that hypermethylation of the SOCS-1 gene is detected in GC cell lines and primary GC tissues.31, 33 To et al. have previously shown that demethylation treatment can increase SOCS-1 expression and reduce STAT3 activation in GC cell lines.32 In our study, we demonstrated that forced SOCS-1 expression can inhibit STAT3 activation in GC cells and suppress their proliferation. Our results together with To et al. suggest that SOCS-1 functions as a tumor suppressor in GC cells, and its gene silencing should promote GC development and progression.
In our study, we showed that high levels of IL-6 are spontaneously produced in NUGC3 and AGS cell lines. Previous studies have shown that IL-6 facilitates GC cell invasion in vitro42 and serum IL-6 levels correlate well with disease progression and recurrence in GC.39, 40, 43 Moreover, the methylation of SOCS-1 gene is associated with lymph node metastasis and advanced tumor stage in GC.33 Thus, NUGC3 and AGS cell lines share two features (IL-6 production and SOCS-1 methylation) of GC with poor prognosis. While the frequency of GC that possesses these two features is currently unknown, future studies on both IL-6-producing capacity and SOCS-1 gene methylation status in GC may help to predict the prognosis of GC.
We showed here that STAT3 is persistently activated in several GC cell lines. Similar activation of STAT3 has been reported in other cancer cell lines in which SOCS-1 gene expression is downregulated by DNA methylation.22, 24, 25 One possible mechanism for STAT3 activation in these cells is constitutive production of cytokines such as IL-6. However, in IL-6-producing NUGC3 and AGS cells, anti-IL-6R antibody failed to reduce STAT3 phosphorylation (Fig. 2), suggesting that IL-6 does not act as an autocrine growth factor in these GC cells. Interestingly, conditioned culture media of NUGC3 cells could induce STAT3 phosphorylation in anti-IL-6R antibody-treated human umbilical vein endothelial cells (Supporting Information Fig. 1), suggesting the presence of unknown soluble factor(s) contributing to the STAT3 phosphorylation in NUGC3 cells. However, IL-11, another member of IL-6 family possibly involved in the development of GC,44 was not elevated in conditioned media of NUGC3 cells (data not shown). These results, nevertheless, suggest that the blockade of IL-6 may have limited therapeutic effects on the growth signals in established GC cells.
In contrast to IL-6 blockade, adenovirus-mediated forced expression of SOCS-1 or dnSTAT3 successfully suppressed cell proliferation of NUGC3 and AGS cells (Fig. 3). The suppressive effect of STAT3 inhibition on GC cell proliferation has been reported previously.12 Importantly, however, the magnitude of antiproliferative effects of AdSOCS-1 was superior to that of AddnSTAT3, suggesting that AdSOCS-1 has additional mechanism(s) of action distinct from that of AddnSTAT3. Indeed, proliferation of MKN7 cells, in which constitutively phosphorylated STAT3 was not detected (Fig. 1c), was successively inhibited by AdSOCS-1, but not by AddnSTAT3 (Supporting Information Fig. 2). Furthermore, proliferation of MKN45 cells, which exhibited constitutive STAT3 phosphorylation (Fig. 1c), was also inhibited by AdSOCS-1 but, surprisingly, not by AddnSTAT3 (Supporting Information Fig. 2). These results indicate that SOCS-1 inhibits not only JAK/STAT3-dependent pathway but also STAT3-independent growth signal pathways. In accordance with this, we found that p38MAPK pathway in GC cells is downregulated by SOCS-1. Moreover, inhibition of these two pathways using JAK inhibitor I and SB203580 had significant antiproliferative effects on GC proliferation. Thus, our results suggest that the potent antiproliferative effect of SOCS-1 is associated with combined inhibition of JAK/STAT3 and p38 MAPK pathways in GC cells.
MAPK family including p38MAPK is activated by a variety of environmental stresses and inflammatory cytokines, and traditionally is thought to play a role in differentiation, growth arrest, inflammation, immune activation and apoptosis. In vivo, p38MAPK may function as a tumor suppressor, since mice lacking p38α are sensitized to lung and liver tumors.45 However, for the development of H. pylori-associated GC, p38 MAPK/ATF-2-mediated COX-2 was reported to be necessary.46 Further studies are needed to elucidate the exact role of p38 MAPK signaling pathway for the growth of GC cells.
In our study, the mechanism by which SOCS-1 suppress the activation of p38 MAPK was not determined. One possible target of SOCS proteins is apoptosis signal-regulating kinase 1 (ASK1), an upstream activator of both the p38 MAPK and JUN N-terminal kinase (JNK) cascades. Supporting this possibility, it has been reported that SOCS-1 regulates the activation of stress-activated MAPKs by binding to ASK1.47 Moreover, a recent study has shown that ASK1 contributes to tumor growth in GC.48 SOCS-1 may also target other proteins in GC cells, since SOCS-1 can interact with a variety of tyrosine-phosphorylated proteins via their SH2 domain and promote the degradation of target proteins via the SOCS box.21 Recently, we showed that adenoviral gene delivery of SOCS-3, one of the genes of the SOCS family, could inhibit growth of malignant pleural mesothelioma cells in vitro and in vivo via the modulation of multiple pathways involving JAK/STAT3, ERK, FAK and p53.41 Although further studies are required, SOCS-1 may also act as an inhibitor of a wide variety of growth signals and may be applicable to the treatment of various types of GC.
In conclusion, we identified SOCS-1 gene methylation in GC cell lines and highlighted a potent antiproliferative effect of SOCS-1 on GC cells both in vitro and in vivo, via the inhibition of JAK/STAT3 and p38 MAPK activation. Epigenetic silencing of SOCS-1 may thus represent a critical step in the development of human GC and forced expression of SOCS-1 may represent a novel therapeutic approach for various types of GC through the inhibition of JAK/STAT3 and p38 MAPK signaling pathways.
We thank Y. Ito, N. Kawakami and Y. Kanazawa for secretarial assistance, and M. Urase for technical assistance.