Expression of Stress-Induced Phosphoprotein1 (STIP1) is Associated with Tumor Progression and Poor Prognosis in Epithelial Ovarian Cancer

Abstract

Stress-induced phosphoprotein1 (STIP1) is a candidate biomarker in epithelial ovarian cancer (EOC). In this study, we investigated in detail the expression of STIP1, as well as its functions, in EOC. STIP1 expression was assessed by immunohistochemistry (IHC) and the results were compared with clinicopathologic factors, including survival data. The effects of STIP1 gene silencing via small interfering RNA (siRNA) were examined in EOC cells and a xenograft model. The expression of STIP1 protein in EOC was significantly higher than in the other study groups (P < 0.001), and this increase of expression was significantly associated with tumor stage (P = 0.005), tumor grade (P = 0.029), and lymph node metastasis (P = 0.020). In multivariate analysis, overall survival in EOC was significantly shorter in cases with high STIP1 expression (HR = 2.78 [1.01–7.63], P = 0.047). STIP1 silencing in EOC cells resulted in inhibition of cell proliferation and invasion. In addition, in vivo experiments using STIP1 siRNA clearly showed a strong inhibition of tumor growth and a modulation of expression of prosurvival and apoptotic genes, further suggesting that STIP1 silencing can prevent cell proliferation and invasion. In conclusion, increased STIP1 expression is associated with poor survival outcome in EOC, and STIP1 may represent a useful therapeutic target in EOC patients.

INTRODUCTION

Among gynecologic malignancies, epithelial ovarian cancer (EOC) is the leading cause of death (Kohler et al., 2011). Most EOC cases are diagnosed at an advanced stage of the disease because the premalignant state is poorly understood and an efficient screening strategy is not presently available (Bast et al., 2005). Although improvements in primary treatment, often consisting of cytoreductive surgery and platinum-based chemotherapy, have been made, the majority of these patients continue to experience relapses and eventually die from the disease (Bookman, 2003; Bast et al., 2009). Therefore, there is an urgent need to identify good diagnostic or prognostic markers and new therapeutic strategies to improve the outcome of EOC patients.

Developing blood-based biomarkers for the early detection and prognosis of cancer remains a significant challenge (Kim and Kim, 2011). Potential biomarkers include differentially expressed tumor antigens or corresponding autoantibodies relating to tumor-associated antigens (TAA). As biomarkers, autoantibodies are precise, easily purified from serum, and readily detectable with high-quality secondary reagents (Jaras and Anderson, 2011). While autoantibodies to TAA have been identified in patients prior to their development of cancer, (Trivers et al., 1995; Qiu et al., 2008), few have been evaluated as prognostic biomarkers of cancer survival and/or recurrence. Previously, our group identified autoantibodies against stress-induced phosphoprotein1 (STIP1) as a novel biomarker candidate for ovarian cancer using the two-dimensional differential gel electrophoresis analysis of immunoprecipitated tumor antigens (2D-DITA) method (Kim et al., 2010). Recently, STIP1 was evaluated as a prognostic biomarker of survival in ovarian cancer patients drawn from Taoyuan, Taiwan (Chao et al., 2013). That study sought to validate and explore the role of STIP1 expression by means of immunohistochemistry, comparing borderline ovarian tumor with invasive carcinoma. In addition, they demonstrated that cell proliferation and migration are stimulated by treating ovarian cancer cells with recombinant STIP1. Even though these studies suggested that STIP1 is a promising candidate as a prognostic marker and therapeutic target in ovarian cancer, little is known about the function of STIP1, and its molecular mechanisms are still unclear. In this study, the expression level of STIP1 in clinical specimens at various stages of ovarian cancer was analyzed. We further investigated whether STIP1 knockdown via siRNA influences either cell proliferation or invasion in EOC. Lastly, using a xenograft mouse model, we evaluated whether this decrease in STIP1 expression has effects on tumor growth in vivo.

MATERIALS AND METHODS

Patients and Tumor Samples

One hundred twenty-six formalin-fixed paraffin-embedded surgical specimens of ovarian cancer, 15 borderline ovarian tumors, 17 benign ovarian tumors, and 9 specimens of normal ovarian epithelial tissues were obtained from the Department of Obstetrics and Gynecology, Gangnam Severance Hospital, Yonsei University College of Medicine, and the Korea Gynecologic Cancer Bank through the Bio and Medical Technology Development Program of the Ministry of Education, Science and Technology, Korea. Tumor staging was performed according to the International Federation of Gynecology and Obstetrics (FIGO) classification. For all study subjects, CA125 levels were measured at primary diagnosis up to 1 week before operation via CA125 II ECLIA (electrochemiluminescence immunoassay) on the Roche/Hitachi Modular Analytics E170 (Roche Diagnostics, Tokyo, Japan). Age, FIGO stage, cell types, and tumor grade were recorded for all EOC patients. Primary treatment of EOC patients consisted of surgical staging procedures for stage I, and maximal cytoreductive surgery for stage II/III/IV, followed by intravenous paclitaxel (175 mg/m²) or docetaxel (75 mg/m²) plus carboplatin (AUC 5–6) combination chemotherapy every 3 weeks for six to nine cycles. Response to therapy was assessed according to Response Evaluation Criteria in Solid Tumors (RECIST; version 1.0), either by computed tomography (CT), positron emission tomography (PET)-CT, or serum CA125 level (Therasse et al., 2000). Overall survival was defined as the time from diagnosis to death; disease-free survival was calculated as the time interval between diagnosis and recurrence. This study was approved by the Institutional Review Boards (IRBs) of Gangnam Severance Hospital.

Cell Lines

SNU840 was purchased from the Korean Cell Line Bank (KCLB, Seoul, Korea), and TOV112D, SKOV3, OVCAR3, and OVCA429 were purchased from the American Type Culture Collection (ATCC, Manassas, VA). YDOV-13 (Brenner tumor), −105 (serous adenocarcinoma), −139 (serous adenocarcinoma), and −151 (mucinous adenocarcinoma) were established and characterized in this laboratory. The general characteristics of these cell lines are described in Supporting Information Table 1. Detailed culture procedures are described elsewhere (Cho et al., 2008, 2009). Briefly, TOV112D, OVCAR3, and OVCA429 were maintained in DMEM/F12 supplemented with 10% fetal bovine serum (FBS) (Gemini Bio-Products, Calabasas, CA) and YDOV-13, −105, −139, and −151 were grown in a mixture of Medium 199 and MCDB 105 (1:1) (Sigma, St Louis, MO) supplemented with 10% FBS (Gemini Bio-Products). SNU840 and SKOV3 were cultured in RPMI-1640 supplemented with 10% FBS (Gemini Bio-Products). All cells were cultured in a humidified incubator at 37°C and 5% CO₂.

SYBR Green Real-Time PCR

Total RNA was extracted from three ovarian cancer cell lines (SKOV3, OVCAR3, and YDOV-13) and xenograft tumors using the RNeasy Mini kit (Qiagen, Valencia, CA). Next, cDNA was generated from 2 μg of total RNA from each sample using the SuperScript™ III first-strand synthesis system (Invitrogen, Carlsbad, CA) following the manufacturer’s suggested protocol. SYBR Green real-time PCR was performed using an ABI 7300 instrument (Applied Biosystems, Forster, CA). Oligonucleotide sequences of the primer sets used were as follows: STIP1 (forward 5′-ATCCTCAGCGTCCTCTTGG-3′, reverse 5′-GGTGGAGGTGTTGCAATCTCTT-3′); ID3 Forward 5′-GAGAGGCACTCAGCTTAGCC-3′, Reverse 5′-TCCTTTTGTCGTTGGAGATGA-3′; Ki-67 (forward 5′-AGAAGACAGTACCGCAGATGA-3′, reverse 5′-CGGCTCACTAATTTAACGCTGG-3′); BCL2 (forward 5′-CGGTGGGGTCATGTGTGTG-3′, reverse 5′-CGGTTCAGGTACTCAGTCATCC-3′); Tp53 (forward 5′-CCGCAGTCAGATCCTAGCG-3′, reverse 5′-AATCATCCATTGCTTGGGAC-3′); and Bax (forward 5′-CTGGACAGTAACATGGAGCTG-3′, reverse 5′-CACTCGGAAAAAGACCTCTCG-3′). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a housekeeping gene, was used as an internal control. Amplification was carried out in 96-well microtiter plates (Applied Bio-systems Incorporated, Foster City, CA). Thermal cycling conditions were as follows: pre-incubation for 2 min at 50°C, then denaturation for 10 min at 95°C followed by 40 cycles of denaturation for 15 sec at 95°C and annealing/extension for 1 min at 60°C. Melting (dissociation) curve analysis was used for all primer sets to exclude nonspecific amplification. Experiments were performed in triplicate. The comparative 2^−ΔΔCt method was used to calculate relative quantification of gene expression as described previously (Livak and Schmittgen, 2001).

Western Blotting

To measure STIP1 expression, cells were lysed in PRO-PRE Protein Extraction Solution (Intron Biotechnology, Seongnam, Korea). Cell lysates (100 μg of total protein) were separated in 10% acrylamide gels by sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes. The membranes were blocked with 5% skim milk in TBST (50 mM Tris, 150 mM NaCl, 0.05% Tween-20, pH 7.5) for 1 h at room temperature, followed by incubation with anti-STIP1 rabbit polyclonal antibody (mouse monoclonal, Cat.# sc-32761, Santa Cruz Biotechnology, Santa Cruz, CA), anti-Smad5 (Cell Signaling, Danver, MA), anti-phosphoSmad1/5(Ser463/465) (Cell Signaling), anti-human β-actin (Cell Signaling), or anti-GAPDH (Santa Cruz Biotechnology) overnight at 4°C. The membranes were then incubated with secondary antibodies conjugated with horseradish peroxidase (GE Healthcare, Munich, Germany) and visualized on AGFA X-ray film (Agfa Health Care, Mortsel, Belgium) using the SuperSignal Chemiluminescence kit (Thermo Scientific, Rockford, IL).

Immunohistochemistry and Scoring

Paraffin sections (5 μm thick) were deparaffinized three times in xylene and subsequently rehydrated in descending gradient alcohol solutions. Antigen retrieval was carried out in a steam pressure cooker with prewarmed pH 6 antigen retrieval buffer (Dako, Carpinteria, CA) at 95°C for 10 min. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 10 min. To minimize nonspecific staining, sections were incubated with Protein Block (Dako) for 10 min. Sections were then incubated with anti-STIP1 antibody (mouse monoclonal, Cat.# sc-32761, Santa Cruz Biotechnology) at a dilution of 1:100 for 1 hr at room temperature followed by antibody detection using the Dako EnVision1 Dual Link System-HRP (Dako). DAB (3,3′-diaminobenzidine; Dako) reagents were applied to detect the signal from the antigen-antibody reaction. Tissue sections were lightly counterstained with hematoxylin, dehydrated in ethanol, and cleared in xylene. Finally, slides were coverslipped and observed under a light microscope (Olympus Corp., Tokyo, Japan).

Staining for STIP1 was scored as positive when tumor or epithelial cells showed cytoplasmic immunoreactivity. The expression level of STIP1 was scored based on (a) intensity [categorized as 0 (absent), 1 (weak), 2 (moderate), or 3 (strong)] and (b) percentage of positively stained epithelial cells [scored as 0 (0–5% positive), 1 (6–25%), 2 (26–50%), 3 (51–75%), or 4 (>75%)]. For the immunostaining score, the intensity and positivity scores were multiplied, resulting in a value between 0 and 12. This overall score for each patient was further simplified by dichotomizing it into being either negative (overall score of ≤3) or positive (score of >3).

Knockdown of STIP1 via RNA Interference

Validated STIP1 siRNAs (HSS116946, 5′-GGGAGCUGAUAGAGCAGCUACGAAA-3′ and HSS174009, 5′-AGGAACCCGAAAGAUGCCAAAUUAU-3′) and negative control siRNA (Stealth RNAi™ siRNA Negative Control Med GC Duplex #2, Cat. # 12935–112) were obtained from Invitrogen. Transfection of each oligonucleotide was performed using Lipofectamine™ RNAi-MAX (Cat.# 03778–015, Invitrogen) according to the manufacturer’s protocols. Cells were treated for 72 h to allow maximum knockdown, after which they were harvested for western blotting or RNA preparation or used for other assays.

Cell Proliferation Assay

To assess the proliferation of each cell line in response to a decrease in STIP1, cells transfected with each siRNA were seeded at a density of 3 × 10³ cells per well in 96-well plates and cultured with 200 μL of complete media. After 72 h, 20 μL of 3-(4,5-dimethylthiazol-2-yl)–2,5-diphenyltetrazolium bromide (MTT, Sigma-Aldrich, St. Luis, MO) solution (5 mg/mL in PBS) was added into each well, and cells were incubated at 37°C, 5% CO₂ for 90 min. After incubation, spent media was removed, 100 μL of DMSO was added to dissolve the formazan crystals, and absorbance was recorded at 562 nm.

Cell Migration and Invasion Assay

Cell migration and invasion were assayed using a CytoSelect™ 24-Well Cell Migration and Invasion Assay kit (Cell Biolabs, Inc., San Diego, CA) according to the manufacturer’s protocol. In brief, SKOV3 cells were treated with either control or STIP1 siRNA for 48 h, and were subsequently suspended in serum free medium (1.0 × 10⁶ cells/mL). Cell suspension (300 μL) was added to the upper chamber. The membrane chambers were then transferred to the lower well of the migration plate, which was filled with 10% fetal bovine serum-containing media. Cells were allowed to migrate for 24 h at 37°C. Migratory cells that had passed through membrane were collected, lysed, and quantified at OD 560 nm using a plate reader.

For invasion assays, cells were treated with either control or STIP1 siRNA for 48 h, and were subsequently suspended in serum free medium (1.0 × 10⁶ cells/mL). Cell suspension (300 μL) was added to the upper chamber, which was coated with a uniform layer of dried basement membrane matrix solution. The membrane chambers were then transferred to the lower well of the invasion plate, which was filled with 10% fetal bovine serum-containing media. Cells were allowed to invade for 24 h at 37°C. Invasive cells that had passed through the basement membrane layer were collected, lysed, and quantified at OD 560 nm using a plate reader.

Ovarian Cancer Xenografts in Nude Mice

Female BALB/c-nu/nu athymic mice (4–5 weeks old) were purchased from Japan SLC, Inc. (Shizuoka Prefecture, Japan) and maintained under specific pathogen-free conditions and cared for in accordance with the American Association for Accreditation of Laboratory Animal Care guidelines, adhering to all national and international standards. Experiments with nu/nu mice were approved and supervised by the Institutional Animal Care and Use Committee of Gangnam Severance Hospital.

For the xenograft tumor growth assay, SKOV3 cells (1 × 10⁷ cells/200 μL in PBS with 50% Matrigel) were injected subcutaneously into the right flanks of the mice. One week after injection, mice were treated with negative control siRNA or STIP1 siRNA (5 mg/kg) every 3 days for 3 weeks using intratumoral injection (6 mice/group). Three weeks after treatment, mice were sacrificed by cervical dislocation and necropsied. Harvested tumors were fixed in formalin and embedded in paraffin. Tumor volumes were calculated using the formula V = (L × W²)/2, where L is length and W is width (Satpathy et al., 2007).

Statistical Analysis

For analysis of the PCR and IHC data in the study samples, the Mann-Whitney or Kruskal-Wallis test was used when homogeneity of variances was not assumed, or one-way analysis of variance (ANOVA) when appropriate. In addition, in order to determine optimal cutoff values of IHC scores, a receiver operating characteristic (ROC) analysis was plotted to maximize the sum of sensitivity and specificity to discriminate cancer from noncancer tissues. Survival curves were calculated using the Kaplan-Meier method, and statistical significance was calculated by the log-rank test. The Cox proportional hazards model was used for multivariate analysis to determine the independent significance of relevant clinical covariates. Generally, P value <0.05 was considered statistically significant. Statistical analyses were performed using SPSS version 18.0 (SPSS Inc., Chicago, IL).

RESULTS

STIP1 Protein Levels are Elevated in Ovarian Cancer Tissue

We found STIP1 mRNA and protein to be over-expressed in ovarian cancer cell lines and tissues in an earlier study (Kim et al., 2010). Chao et al. also demonstrated that high STIP1 immunohistochemical expression is associated with high stage and high grade ovarian cancer (Chao et al., 2013). In the present study, a large cohort of ovarian tumor tissues representing a significantly higher number of samples than was used in our previous study was subjected to western blotting and immunohistochemistry in order to validate the aforementioned results. Patient ages ranged from 14 to 80 years (mean, 50.6 years). The following histologic types were represented: 126 EOC (85 serous, 15 mucinous, 13 endometrioid, 4 clear cell, 2 transitional cell, 1 Brenner tumor, 4 mixed, and 1 undifferentiated carcinomas), 15 borderline ovarian tumors (10 serous and 5 mucinous tumors), and 17 benign tumors (4 serous and 13 mucinous cystadenomas).

Via immunoblotting, we found that STIP1 is highly expressed in TOV112D, SNU840, OVCAR3, SKOV3, OVCA429, YDOV-105, YDOV-151, YDOV-13, and YDOV-139 cells, but it was not detected in HOSE cells (Fig. 1A). These observations suggest that the expression of STIP1 protein occurs mostly in fully transformed cells. In subsequent immunostaining analysis, most immunoreactivity was observed in the cytoplasm of tumor cells (Fig. 1B). IHC scoring results are summarized in Table 1. STIP1 protein expression was significantly upregulated in ovarian cancer tissues relative to borderline (P = 0.039), benign (P < 0.001), and normal ovarian tissues (P < 0.001) (Fig. 1B). In addition, STIP1 immunoreactivity positively correlated with clinical features of poor outcome such as advanced tumor stage (P = 0.008) and grade (P = 0.004) (Table 1, Fig. 1C). However, there was no significant difference based on cell type for IHC score (P = 0.584). These data suggest that STIP1 might play a crucial role in tumorigenesis in EOC.

Figure 1.

STIP1 is highly expressed in human ovarian cancer cells and tissue specimens. A. STIP1 protein levels were analyzed by western blot. GAPDH was included as an internal loading control. B. Representative immunohistochemical staining for STIP1 in formalin-fixed paraffin-embedded EOC tissues (400×). C. IHC staining score of STIP1 in EOC samples was significantly higher than that in healthy controls, benign ovarian tumors, or borderline ovarian tumors (P < 0.001, P < 0.001, and P = 0.039 respectively). Differences between tumor stages and grades were statistically significant, with advanced-stage and poor-grade specimens having higher immunoreactivity.

TABLE 1.

Expression of STIP1 in Relation to Clinicopathological Characteristics in IHC Analysis

	No.	%	Mean scores (95% CI)	Range	P
All study subjects	167	100	3.59 (3.10 to 4.09)	0–12
Diagnostic category					<0.001
Normal	9	5.4	0.22 (−0.12 to 0.56)	0–1
Benign	17	10.2	0.82 (0.27 to 1.38)	0–3
Borderline	15	9.0	2.53 (1.01 to 4.06)	0–9
Cancer	126	75.4	4.83 (4.28 to 5.37)	0–12
FIGO stage					0.008
I–II	34	27.0	3.62 (2.77 to 4.47)	0–8
III–IV	92	73.0	5.27 (4.60 to 5.94)	0–12
Cell type					0.584
Serous	84	66.7	4.56 (3.85 to 5.26)	0–12
Mucinous	15	11.9	4.27 (2.39 to 6.14)	0–9
Other	27	21.4	3.78 (2.54 to 5.01)	0–12
Tumor grade					0.004
Borderline	15	13.8	2.53 (1.01 to 4.06)	0–9
Well	11	10.1	4.09 (2.35 to 5.83)	1–8
Moderate	44	40.3	5.18 (4.16 to 6.20)	0–12
Poor	39	35.8	5.87 (4.94 to 6.80)	1–12
CA125					0.814
Negative	17	13.5	4.53 (2.67 to 6.39)	0–12
Positive (>35 U/mL)	109	86.5	4.33 (3.73 to 4.93)	0–12
Chemosensitivity					0.920
Sensitive	72	71.2	4.63 (3.84 to 5.41)	0–12
Resistant	29	28.8	4.55 (3.35 to 5.76)	0–12

STIP1 Overexpression is Associated with Poor Prognosis

We next examined the relationship between STIP1 expression and prognostic outcome in EOC. Only 113 (90.4%) patients with optimally debulked tumors (residual tumor remaining after surgery ≤1 cm) and available outcome data were included in the survival analysis. The follow-up period for patients ranged from 2 to 60 months, with a median of 21.0 months. Fourteen patients (12.4%) died within this period, 34 (30.1%) survived but suffered recurrence or persistent disease, and 65 (57.5%) showed no evidence of disease after treatment. In the recurrent disease group (n = 32), the mean time to recurrence after initial treatment was 20.4 months.

Kaplan-Meier plots demonstrated that patients with advanced disease (FIGO stage III/IV) and STIP1+ status (IHC score of >3) displayed significantly worse overall survival (P = 0.004 and P = 0.019, respectively) (Fig. 2). A Cox multivariate proportional hazards analysis showed that advanced stage (hazard ratio [95% CI] = 2.94 [1.07–8.06], P = 0.036) and STIP1+ status (HR [95% CI] = 3.05 [1.01–9.20], P = 0.047) were independent prognostic factors with respect to overall survival (Table 2).

Figure 2.

Kaplan-Meier plots for patients with EOC stratified according to tumor stage, STIP1 expression, tumor grade, or CA125 expression.

TABLE 2.

Univariate and Multivariate Analyses of the Associations Between Prognostic Variables and Overall Survival in 113 Cases of Epithelial Ovarian Cancer

	Univariate analysis		Multivariate analysis
	Hazard ratio [95% CI]	P	Hazard ratio [95% CI]	P
FIGO stage (≥111)	3.19 [1.17–8.73]	0.023	2.94 [1.07–8.06]	0.036
Cell type (serous)	NS		NA
Tumor grade (poor)	NS		NA
CA125+ (>35 U/mL)	NS		NA
STIP1+ (>3)	3.39 [1.13–10.14]	0.029	3.05 [1.01–9.20]	0.047
STIP1+/CA125+	3.47 [1.26–9.53]	0.015	2.88 [1.02–8.06]	0.044

CI, confidence interval; FIGO, International Federation of Gynecology and Obstetrics; NS, not significant; NA, not applicable.

STIP1 Knockdown in Ovarian Cancer Cells Inhibits Cell Proliferation and Invasion

To investigate STIP1’s role in the malignant transformation of EOC, siRNA targeting STIP1 or negative control siRNA was transfected into SKOV3, OVCAR3, and YDOV-13 cells, the results of which, after subsequent immunoblotting assays, revealed higher endogenous STIP1 expression in these cells than in HOSE cells (Fig. 1A). The two most effective knockdown constructs (siSTIP1#1 and siSTIP1#2) were selected to investigate phenotypic changes due to STIP1 downregulation. STIP1 expression levels in transfected cells were detected by performing real-time PCR (Fig. 3A) and western blotting (Fig. 3B). Next, we analyzed the effect of STIP1 knockdown on cell proliferation. The MTT assay revealed that cell growth rates in STIP1-transfected SKOV3 (siSTIP1#1) and YDOV-13 (siSTIP1#1) cells were significantly lower than in control groups (siControl) (Fig. 4A). These overall results suggest that STIP1 plays a vital role in the proliferation of EOC cells.

Figure 3.

STIP1 knockdown by siRNA in EOC cells. Whole cell lysates and total RNA were collected from SKOV3, OVCAR3, or YDOV-13 cells following STIP1 knockdown by siRNA. Expression of STIP1 mRNA was measured by real-time PCR (A) and protein level was analyzed by immunoblotting (B). β-actin mRNA levels were used to normalize STIP1 mRNA expression. An asterisk (*) indicates a P value <0.05.

Figure 4.

Effects of STIP1 on cell proliferation, migration, and invasiveness of EOC cells. A. Cell proliferation of SKOV3, OVCAR3, and YDOV-13 was determined by MTT assay 72 h after siRNA transfection. Error bars represent the SD of triplicate experiments. Cell migration (B) and invasion (C) analyses of SKOV3 cells were performed. Migrating or invasive cells were quantified using a plate reader at OD 560 nm. D. STIP1 knockdown (siSTIP1#1 and siSTIP1#2) decreased mRNA level of ID3 in OVCAR3 cells. E. OVCAR3 cells were assayed for phospho-Smad1/Smad5. STIP1 knockdown (siSTIP1#1 and siSTIP1#2) reduced phosphorylation of Smad1/Smad5. The total amount of Smad5 and actin was used as the protein loading control. An asterisk (*) indicates a P value <0.05.

Because STIP1 displayed its most prominent proliferative effect in SKOV3 cells, further in vitro studies were performed using SKOV3 cells (Fig. 4A). To investigate STIP1’s effects on the migration capability of EOC cells, a cell migration assay was conducted. The results of the assay showed that STIP1 knockdown could inhibit cell motility in SKOV3 cells relative to the control cells, although that effect was not found to be statistically significant (Fig. 4B). To evaluate whether STIP1 could also enhance the invasiveness of EOC cells, a cell invasion assay was performed separately. In contrast to the migration assay, STIP1 knockdown in SKOV3 cells (siSTIP1#1, #2) led to significantly decreased invasion (Fig. 4C). Overall, the above results suggest that while STIP1 has minimal effect on the migration of EOC cells, it has a net positive effect on cell invasion via other factors.

STIP1 Knockdown Decreases Expression of ID3 and Phosphorylation of Smad1/Smad5

Because STIP1 was shown to be involved in the activation of the ID3 gene (Tsai et al., 2012), we conducted real-time PCR in whole cell lysates that were taken from OVCAR3 cells. STIP1 knockdown (siSTIP1#1, #2) led to significantly decreased ID3 mRNA level compared with the control cells (Fig. 4D). In addition, to verify a correlation between the STIP1 and Smad pathway, we conducted Western blotting. STIP1 knockdown in OVCAR3 cells (siSTIP1#1, #2) reduced phosphorylation of Smad1/Smad5 (Fig. 4E). These data suggest that STIP1 potentially modulates the activity of Smad pathway.

STIP1 Knockdown in SKOV3 Cells Inhibits Tumor Growth in Xenografts in Nude Mice

To explore whether STIP1 can affect tumor growth in vivo, we inoculated SKOV3 cells with or without STIP1 knockdown as xenografts into nude mice. The tumor volume of dominant masses was measured weekly. Starting at 1 week after administration of siSTIP1, the mean tumor volumes at days 7, 14, and 21 (37.9 ± 6.6 mm³) in mice receiving siSTIP1#1 were significantly smaller (57%) than those in the negative control group (66.5 ± 6.4 mm³) (Fig. 5A). To measure the expression of Ki-67, BCL2, Tp53, and Bax, we collected all of the xenograft tumors and examined the mRNA levels of the aforementioned genes. Real-time PCR showed a significant decrease in Ki-67 expression in siSTIP1 transfected cells (siSTIP1#1) relative to the control group on the third day. In contrast, Tp53 and Bax expression significantly increased in the siRNA-transfected group (siSTIP1#1) (Fig. 5B). These findings further support our initial claim that STIP1 plays a functional role in the malignant transformation of EOC.

Figure 5.

STIP1 knockdown inhibits in vivo xenograft tumor growth. For the xenograft tumor growth assay, treatments were started 1 week after tumor cell injection, and siControl or siSTIP1 was injected every 3 days for 3 weeks at a dose of 5 mg/kg body weight. A. STIP1 knockdown reduces the growth rate of xenografted SKOV3 cells in BALB/c-nu mice. Mean tumor volume ± SD for each group was calculated at each week. B. Ki-67, BCL2, Tp53, and Bax mRNA levels were analyzed by qRT-PCR and normalized to β-actin mRNA levels. An asterisk (*) indicates a P value <0.05.

DISCUSSION

Human STIP1 is often referred to as heat shock protein (HSP) organizing protein (Hop) and mediates the association of the molecular chaperones HSP70 and HSP90 (Carrigan et al., 2006). The main role of STIP1 is to link HSP70 and HSP90, and its primary action is as an adapter that leads HSP90 to HSP70 client proteins in the cytoplasm. Furthermore, recent evidence suggests that STIP1 can control the chaperone activities of these HSPs (Odunuga et al., 2004). Although the role of STIP1 in tumor pathogenesis is not fully elucidated, STIP1 has been reported to be associated with various cancers, especially melanoma (Carta et al., 2005), hepatocellular carcinoma (Sun et al., 2007), and pancreatic cancer (Walsh et al., 2009) and involved in many crucial process in carcinogenesis such as growth signaling, stabilization of mutant proteins, angiogenesis, and metastasis (Averna et al., 2008).

We recently reported STIP1 overexpression in EOC and that STIP1 might serve as a putative diagnostic and prognostic marker in EOC. To further validate our previous results and to investigate whether or not overexpression of STIP1 is involved in the progression of EOC, we first examined STIP1 protein expression by western blotting and IHC. Our western blotting results indicated that STIP1 was highly expressed in all EOC cells, irrespective of histological subtype, but not in HOSE cells. In IHC analysis, STIP1 immunoreactivity was negative or low in all normal ovarian epithelial tissues and benign tumors. In contrast, STIP1 protein expression was significantly upregulated in the majority of EOC specimens (Table 1). These results were similar to our previous observations of EOC, in which STIP1 expression was significantly elevated in 55 ovarian cancer tissues, and confirmed that STIP1 protein expression is elevated in EOC patients. In agreement with our study, Wang et al. also reported higher expression of STIP1 protein in both tissues and sera of ovarian cancer patients (Wang et al., 2010). This group also reported quite recently that their IHC analysis of 330 ovarian tumor samples indicated higher STIP1 expression in ovarian cancer tissue than in borderline ovarian tumors (Chao et al., 2013). Our results correlate with this finding although some important differences in these two studies need to be noted. While the previous investigation by Chao et al. was limited to tumor tissues only, in this study, we investigated STIP1 expression across various stages from normal to benign tissues and from borderline tumors to EOC. Thus, we were able to more clearly show a progressive increase in STIP1 expression along these transitions, and from this slight change in the experimental design we were able to underscore the important role that STIP1 plays in the development and pathogenesis of EOC.

Another of our findings was that STIP1 overexpression in EOC predicts poor overall survival. According to univariate analysis, STIP1+ status and advanced tumor stage are associated with increased risk of death from EOC. Moreover, STIP1 and advanced tumor stage are independent prognostic factors for overall survival in a multivariate analysis of a Cox hazard model. We also observed the same trend but it was not statistically significant when we analyzed only patients with serous ovarian cancer (data not shown). Similarly, it has been shown that STIP1 may be related to prognosis in ovarian cancer; Chao et al. demonstrated that high STIP1 immunohistochemical expression is associated with reduced overall survival of ovarian cancer patients (Chao et al., 2013). These findings indicate that STIP1 is potentially associated with increased risk of death from ovarian cancer. Therefore, examination of STIP1 expression by IHC can be used to stratify patients into subgroups, some of which would benefit from more aggressive adjuvant therapies.

In order to examine the role of STIP1 in tumor cell growth and metastasis, STIP1 protein was downregulated in SKOV3, OVCAR3, and YDOV-13 cells (cells with high expressions of endogenous STIP1) by siRNA transfection. In an MTT assay, we found that siRNA-mediated STIP1 knockdown led to suppression of cell proliferation in SKOV3 and YDOV-13 cells in vitro. In addition to decreased cell growth after siSTIP1 transfection, our results indicated that STIP1 has notable effects on the migration and invasiveness of EOC cells. In a cell migration assay, STIP1 knockdown may have inhibited cell motility in SKOV3 cells, although the difference was not significant. On the other hand, STIP1 knockdown significantly inhibited the invasiveness of SKOV3 cells. To date, the molecular mechanisms by which STIP1 regulates cell growth and/or invasion is unknown. However, a recent study by Tsai et al. indicates that STIP1 stimulates proliferation of ovarian cancer cells by activating the Smad-ID3 signaling pathway (Tsai et al., 2012). To determine the underlying mechanism by which STIP1 confers a growth and invasion advantage to EOC cells, we next investigated whether the Smad pathway was compromised upon STIP1 knockdown. Interestingly, we found that the expression of ID3 that promotes cell proliferation is decreased after STIP1 knockdown in OVCAR3 cells. Furthermore, we found that STIP1 knockdown inhibited phosphorylation of Smad1/Smad5 in OVCAR3 cells. Taken together, our data suggest that STIP1 mediates the malignant phenotype of EOC cells through the activation of the Smad signaling pathway.

A few groups have recently reported a possible role of STIP1 in cancer cells. Walsh et al. observed that STIP1 knockdown reduced cell invasion as matrix metalloproteinase-2 (MMP-2) expression decreased in the pancreatic cancer cell lines, Panc1 and BxPc-3 (Walsh et al., 2011). Chao et al. reported the enhancement of cell proliferation and migration in the ovarian cancer cell lines, BG1 and MDAH2774, by addition of human recombinant STIP1 (Chao et al., 2013). Consequently, it is reasonable to suggest that STIP1 is among the critical molecules related to EOC progress. All of the previous studies were performed in in vitro systems; however, we established xenograft mouse models using SKOV3 cells with or without STIP1 knockdown to verify STIP1’s function in an in vivo system. The exact mechanisms of STIP1 in cell migration and invasion are still unknown, but our results, along with the previously mentioned studies, clearly demonstrate a critical role of STIP1 in cancer cell proliferation and invasion which can further contribute to the growth and progression of EOC.

To confirm the effects of STIP1 that were found in in vitro functional studies, we injected nude mice subcutaneously with cells with or without STIP1 knockdown. Suppression of STIP1 by siRNA in SKOV3 cells (siSTIP1#1) resulted in a significant inhibition of tumor volume increase. The possible mechanism of STIP1 in tumor growth was suggested by assessment of critical molecules of cell proliferation and survival. Expression of wild type Tp53, a cell cycle regulator, was restored by STIP1 knockdown in siSTIP1#1 tumors, and expression of Ki-67, a proliferative index of cancer cells, decreased in these tumors. On the other hand, the expression levels of BCL2 family members were also modulated by STIP1 knockdown in xenografted tumors. Although the expression of the prosurvival protein, BCL2, was not modulated in this study, the expression of the Bax protein, which can disrupt BCL2 function through interaction between the two proteins, was markedly increased in siSTIP1#1 tumors. These results strongly suggest that STIP1 enhances tumor growth by manipulating the cancer cell cycle, proliferation, and apoptotic signaling. However, future study of the molecular mechanisms by which STIP1 promotes tumor growth in EOC will be needed to develop appropriate anticancer therapies.