Tim-3 expression in cervical cancer promotes tumor metastasis.

BACKGROUND: T cell immunoglobulin mucin-3 (Tim-3) has been identified as a negative regulator of anti-tumor immunity. Recent studies highlight the important role of Tim-3 in the CD8(+) T cell exhaustion that takes place in both human and animal cancer models. However, the nature of Tim-3 expression in the tumor cell and the mechanism by which it inhibits anti-tumor immunity are unclear. This present study aims to determine Tim-3 is expressed in cervical cancer cells and to evaluate the role of Tim-3 in cervical cancer progression.
METHODOLOGY: A total of 85 cervical tissue specimens including 43 human cervical cancer, 22 cervical intraepithelial neoplasia (CIN) and 20 chronic cervicitis were involved. Tim-3 expression in tumor cells was detected and was found to correlate with clinicopathological parameters. Meanwhile, expression of Tim-3 was assessed by RT-PCR, Western Blot and confocal microscopy in cervical cancer cell lines, HeLa and SiHa. The migration and invasion potential of Hela cells was evaluated after inhibiting Tim-3 expression by ADV-antisense Tim-3.
CONCLUSIONS: We found that Tim-3 was expressed at a higher level in the clinical cervical cancer cells compared to the CIN and chronic cervicitis controls. We supported this finding by confirming the presence of Tim-3 mRNA and protein in the cervical cell lines. Tim-3 expression in tumor cells correlated with clinicopathological parameters. Patients with high expression of Tim-3 had a significant metastatic potential, advanced cancer grades and shorter overall survival than those with lower expression. Multivariate analysis showed that Tim-3 expression was an independent factor for predicting the prognosis of cervical cancer. Significantly, down-regulating the expression of Tim-3 protein inhibited migration and invasion of Hela cells. Our study suggests that the expression of Tim-3 in tumor cells may be an independent prognostic factor for patients with cervical cancer. Moreover, Tim-3 expression may promote metastatic potential in cervical cancers.

Introduction

Current studies continue to demonstrate a strong correlation between Tim-3 expression and tumor-associated immune suppression [1]–[3]. Sakuishi et al [4] found that Tim-3 was expressed on CD8+ tumor-infiltrating lymphocytes (TILs) in mice bearing solid tumors. All Tim-3+ TILs coexpress PD-1, and this kind of TILs represents the predominant fraction of T cells infiltrating tumors. Tim-3+PD-1+ TILs exhibit the most severe exhausted phenotype as defined by failure to proliferate and produce IL-2, TNF and IFN-γ. Combined targeting of the Tim-3 and PD-1 pathways is more effective in restoring anti-tumor immunity than targeting either pathway alone. Zhou et al [5] detected the same phenomenon in mice with disseminated acute myelogenous leukemia. Even in melanoma patients, upregulation of Tim-3 and PD-1 expression is also found to be associated with tumor antigen-specific CD8+ T cell dysfunction [6]. In our previous study, we found that Tim-3 was preferentially expressed in lymphoma-derived endothelial cells and suppressed activation of CD4+ T lymphocytes through the activation of the interleukin-6–STAT3 pathway. Tim-3 also facilitated the establishment of lymphoma immune tolerance [7]. However, whether Tim-3 is also expressed on cancer cells in other nonhematologic cancers remain an open question.

Cervical cancer is a tumor that possesses distinct Tumor-associated antigens (TAAs). Human papillomaviruses (HPVs) have been shown to cause progressive changes in the cervical epithelium, leading to cervical cancer. Greater than 99% of cervical malignancies harbor HPV (mainly HPV-16 and HPV-18). E6 and E7 are the two proteins expressed by the virus that are necessary for the initiation and progression of cancer [8]. Several studies [9], [10] show that there are other potential roles such as Fas-FasL system and loss of HLA class I for immune escape in cervical cancer. Because many pathways are potentially involved in cervical cancer, we wondered if Tim-3, an antigen that has been implicated in the development of various cancers, might also play a role in the development of cervical cancer. In addition, our previous study found that Tim-3 was overexpressed in epithelial cancers. These reasons combined made us interested in investigating the role of Tim-3 in cervical cancer.

Two recent studies [11], [12] have identified Tim-3 expression on leukemic stem cells (LSC) in patients with acute myeloid leukemia (AML). Tim-3+ AML cells were able to reconstitute AML and anti-human Tim-3 antibody blocked AML engraftment in a xenotransplant model. Although anti-Tim-3 antibodies seem to reduce the metastatic potential of cancer cells, the mechanism of action is not well understood. In this study, we used ADV-antisense Tim-3 to down-regulate Tim-3 in the cervical cancer Hela cell line and then assessed the ability of the cancer cells to migrate and invade.

Our results from this study confirmed the mRNA and protein level expression of Tim-3 in cervical cancer cell lines and revealed for the first time that Tim-3 was preferentially expressed in the clinical primary cervical cancer cells when compared to the CIN and chronic cervicitis. In addition, patients with high expression of Tim-3 had a significantly greater metastatic potential and advanced cancer grades and shorter overall survival than those with lower Tim-3 expression. Therefore, Tim-3 may potentially be an independent prognostic factor for patients with cervical cancer. We also found that ADV-antisense Tim-3 can inhibit migration and invasion of Hela cells. Combined with our previous study, in which we found that Tim-3 activates the IL-6–STAT3 pathway to suppress the activation of CD4+ T lymphocytes, our study opens the possibility that Tim-3 may promote metastasis through the activation of IL-6-STAT3 pathway.

Materials and Methods

Patients

Samples of 43 cervical cancer tissues, 22 CIN tissues and 20 chronic cervicitis tissues were derived from patients that underwent primary surgery for cervical diseases at the Department of Gynecologic Oncology in Tongji Hospital (Huazhong University of Science and Technology, Wuhan, China) from 2004–2006. All of the selected cervical cancer tissues met the following inclusion criteria: no history of any other type of malignant tumor, without neoadjuvant therapy prior to surgery. All patients gave informed written consent for analysis of their tissue for research purposes. This study was approved by the ethics committee of the Tongji Hospital for analysis of human tissues. The patients ranged in age from 27 to 67 years (median age, 39 years). Histological examination of the excised cervical tissues were carried out following hematoxylin & eosin (H & E) staining of paraffin-embedded sections. 31 patients of invasive cervical cancers were classified as grade I, IIa without metastasis, while 12 patients were grade IIb, III and IV with metastasis.

Follow-up

Follow-up data retrieved from the clinical record ranged from 5–60 months post-surgery (median, 45.2 months). Each patient’s overall survival (OS) is calculated as the period from the date of surgery until the date of death.

Immunohistochemical Detection of Tim-3 in Cervical Tissues

Paraffin-embedded tissue sections were dewaxed in xylene and subjected to immunohistochemical analysis as previously described [13]. Anti-Tim-3 goat polyclonal antibody (Santa Cruz Biotechnology, Inc.) and biotinylated secondary antibody were used in the present study. For semiquantitative evaluation, an immunoreactivity-scoring (IRS) system was applied. Intensity of staining was designated as either nonexistent (0), weak (1), moderate (2), or strong (3). The percentage of positive cells was termed as the expression score. The IRS was calculated by multiplying the expression score with the intensity score, and may range from 0–3. The sample with IRS scores of 0–1 points was considered as negative expression of Tim-3, otherwise was designated as positive staining. These data were analyzed along a continuum, and the objective was to use this semiquantitative method to assess differences between various experimental groups. The subcellular localization of the staining (cytoplasmic and/or nuclear) was also observed.

Cell Lines

Cervical cancer cell lines used in the current study were obtained from American Type Culture Collection (ATCC) (HeLa, SiHa). These two cells were cultured in DMEM medium with 10% fetal bovine serum.

Reverse Transcriptase-Polymerase Chain Reaction Detection of Tim-3

Briefly, total RNA was extracted from two cervical cancer cell lines. A 5′ sense primer (5′-CGGAGGTCGGTCAGAATGCCTATC-3′) and a 3′ antisense primer (5′-GGGCTCCTCCA CTTCATATACGTTC-3′) were used to amplify Tim-3 transcripts. The expected product for full-length Tim-3 is 749 bp. A 5′ sense primer (5′-CTCACGAAACTGGAATAAGC-3′) and a 3′ antisense primer (5′-AAGCCACACGTACTAAAGGT-3′) were used to amplify a 180-bp β-actin internal control. Total RNA extracted from PBL was used as positive control. The primers used for RT-PCR detection of Tim-3 were designed to span introns to avoid false positive amplifications resulting from DNA amplifications. Meanwhile we use total RNA product without reverse transcription as templates for PCR as negative control to be sure free of genomic DNA contamination.

Western Blot Detection of the Expression of Tim-3 Protein

Two kinds of cervical cancer cells were lysed for 30 min at 4°C in a lysis buffer composed of 150 mmol/L NaCl, 50 mmol/L Tris (pH 8.0), 5 mmol/L EDTA, 1%(v/v) NP40, 1 mmol/L phenyl-methylsulfonyl fluoride, 20 µg/ml aprotinin, and 25 µg/ml Leupeptin. Equal amounts of protein extracts (10 µg) were resolved by SDS-PAGE. Following transfer to a nitrocellulose filter, it was blocked for 1 h at room temperature with buffer containing 20 mmol/L Tris-Hcl (pH7.5), 500 mmol/L NaCl, and 5% nonfat milk; incubated with Tim-3 antibody (1∶1000, Santa Cruz Biotechnology, Inc.) for overnight at 4°C; washed; and incubated with a horseradish peroxidase-labeled secondary antibody donkey anti-goat IgG (1∶5000) for 1 h at room temperature. Finally, the blots were developed using an enhanced chemiluminescence detection system (Amersham Life Science). In the preliminary experiment, we found NK-92 cell line predominantly expressed Tim-3 protein, while THP-1 cell line did not express Tim-3 protein. Total protein extracted from NK-92 cell line was used as positive control of Tim-3 protein, while total protein extracted from THP-1 cell line was used as negative control.

Immunofluorescent Detection of HeLa Cell Line

HeLa cells were harvested with a combination of trypsin and ethylenediamine tetraacetic acid (EDTA) and washed in PBS, and grown on coverslips. cells then were fixed in 95% ethnol and blocked with 1% BSA for 1 h, and then incubated with primary antibody at 4°C overnight. The primary antibody used were goat anti-Tim-3 (Santa Cruz Biotechnology, Inc.) at a dilution of 1∶100. The specimens then were washed in PBS for 5 minutes 3 times. Negative control slides were incubated without the primary antibody. Donkey anti-goat IgG conjugated with FITC diluted with 2% BSA/PBS to a dilution of 1∶100 was incubated for 2 h at room temperature, cell nuclear was stained by PI at a dilution of 1∶1000, checked by confocal microscope.

Adenoviral Mutants

The AdEasy system (MP Biomedicals) was used in this study to construct a recombinant replication-deficient adenovirus vector named ADV-antisense Tim-3 which contained a fragment of reverse Tim-3 cDNA (bp 198–1312). Standard protocols were followed as described previously [14]. ADV-GFP containing a GFP gene under the control of a Rous sarcoma virus long terminal repeat promoter in the region of the excised E1 adenoviral genes was used as a control in this study. Hela cells were infected with ADV-antisense Tim-3 or ADV-GFP at a proper MOI determined by Apoptosis assay for 2 h, then cultured for another 24 h and subjected to subsequent experiments.

Apoptosis Assay

Briefly, Hela cells infected with different titers of adenovirus mutants or treated with PBS for 72 h. Then cells were fixed with 70% ethanol for 1 h, washed in PBS and, treated with RNase for 15 minutes at 37°C and then stained with 50 µg/ml propidium iodide. Cells were collected and subjected to FACS analysis of sub-G1 population.

Wound Healing Assay

The cells were grown to confluence in a 6-well culture plate. A linear wound was made by scratching the monolayer with a sterile 10-ul pipette tip. The wounded monolayers were washed 3 times with regular medium and incubated in fresh serum-free medium. Photographs were taken at 0 h, 24 h and 48 h after wounding by phase contrast microscopy.

Transwell Invasion Assay

Cell invasion was assayed using Transwell chambers (Costar, Cambridge, MA, USA) with 8-µm pore polycarbonate filters that were coated with Matrigel™ (BD Biosciences, Franklin Lakes, NJ). Cells infected with adenoviral mutants for 24 h were seeded into the upper chambers in serum-free medium at a density of 2.0×104 per well, and 500 µl of 10% fetal bovine serum-containing medium was placed in the lower chamber as a chemo-attractant. After 48 h at 37°C in 5% CO2, the cells were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet solution. Cells on the upper surface of the filter were removed with cotton buds. Invaded cells on the underside of the filter were photographed and counted by phase contrast microscopy (×200 magnifications). The experiments were performed in triplicate.

Statistical Analysis

The SPSS statistical software program was used to test for correlations between quantitative variables by the establishment of nonparametric linear regression. Data are presented as means ± SD of at least three experiments and were analyzed by one-way analysis of variance followed by the Student-Newman-Keuls test. The Kaplan-Meier method was used to estimate the overall survival rate as a function of time. Survival differences were analyzed using the log-rank test. The Cox proportional hazard model was used in the univariate and multivariate analysis of prognostic factors. All p-values were two-sided, and P<0.05 was considered significant. SPSS software (version 11.5) was used for all statistical procedures.

Results

Tim-3 is Preferentially Expressed in Cervical Cancer Tissue

We analysed sections of tumor tissue from 43 patients that were operated on for cervical cancer, from 22 patients with CIN and from 20 patients with chronic cervicitis. Positive staining for Tim-3 protein was seen in only 15.0% (3 out of 20) of chronic cervicitis, but 50.0% (11 out of 22) of the CIN and 65.1% (28 out of 43) of cervical cancer stained positively for Tim-3 protein (Fig. 1A–D). When the expression of Tim-3 protein was further compared by semiquantitative immunoreactivity H-scoring, cervical cancer and CIN displayed a much higher Tim-3 score than chronic cervicitis tissue. (0.905±0.584, 0.558±0.123 vs 0.102±0.075; P<0.01) (Table 1).

Figure 1

Representative immunohistochemical staining for Tim-3 in cervical tissues.

(A) cervical squamous carcinoma, (B) cervical adenocarcinoma, (C) cervical intraepithelial neoplasia, and (D) chronic cervicitis tissue. Original magnification ×200.

Table 1

The expression of Tim-3 in cervical tissues.

Group	Cases	Tim-3
		positive cases(%)	scorea
Cervicitis	20	3 (15.0)	0.152±0.075b,d
CIN	22	11 (50.0)	0.558±0.123c
Cervical cancer	43	28 (65.1)	0.905±0.584

Expression score (mean ± SE). The expression score represents the expression level of Tim-3 protein in cervical cancer tissue as calculated by the immunoreactivity-scoring system.

Cervical cancer group versus Cervicitis group P = 0.002;

Cervical cancer group versus CIN group P = 0.006;

CIN group versus Cervicitis group P = 0.002.

Tim-3 Expression Correlated with Clinicopathologic Parameters

We correlated the Tim-3 expression data to clinicopathologic characteristics such as age, histological type, clinical grade, histological grade, metastasis and overall survival. High immunoreactivity of Tim-3 was found to be significantly correlated with clinical grade (p = 0.018), histological grade (p = 0.038) and metastasis (p = 0.004). There were no significant correlations with age (p = 0.178) and histological type (p = 0.773) (Table 2).

Table 2

The expression of Tim-3 in cervical cancer correlates with clinical features.

Variable		Cases	Tim-3
			scorea	F	P
Age	<39	21	0.781±0.457	1.882	0.178
	>39	22	1.023±0.673
Clinical stage	I,IIa	31	0.776±0.553	6.066	0.018*
	IIb,III,IV	12	1.238±0.547
Type	Adenocarcinoma	8	0.850±0.657	0.084	0.773
	Squamous cellcarcinoma	35	0.917±0.575
Metastasis	No	31	0.748±0.526	9.59	0.004**
	Yes	12	1.308±0.547
Histology grade	I	13	0.784±0.603	4.589	0.038*
	II,III	30	1.183±0.438

Expression score (mean ± SE). The expression score represents the expression level of Tim-3 protein in cervical cancer tissue as calculated by the immunoreactivity-scoring system. Corresponding F and P values are displayed for each cross tabulation.

P<0.05;

P<0.01.

At the end of our follow-up period, 15 patients died in Tim-3 positive group while 3 in Tim-3 negative group, the 5-year survival rate was 46.4% vs 80% respectively (P = 0.006) (Fig. 2).

Figure 2

The correlation between Tim-3 expression and survival rate.

A comparison of five-year cumulative-survival curve of cervical cancer patients between Tim-3 positive expression (broken line) and Tim-3 negative expression (thick line) is shown.

Tim-3 is Expressed in Cervical Cancer Cell Lines

RT-PCR and western blot analysis were used to detect Tim-3 mRNA and protein levels in two human cervical cancer cell lines Hela and SiHa. As expected, a 749-bp product was found to be present in these two cell lines, thereby confirming the presence of Tim-3 mRNA expression. β-actin oligonucleotides were used to detected a 180-bp RNA band and were used as a control (Fig. 3A). A 33 kd band was found confirming Tim-3 protein expression (Fig. 3B). Confocal microscopy was used to test subcellular localization of Tim-3. Tim-3 was distributed in the whole cytoplasm (green fluorescence) of the Hela cell line, with no distribution in the nucleus (red) (Fig. 3C).

Figure 3

Expressions of Tim-3 in cervical cancer cell lines.

(A) Tim-3 transcripts were detected in Hela and SiHa cell lines. The templates of different lanes as follows: Lane 1, total RNA of PBL; Lane 2, cDNA of PBL; lane 3, total RNA of SiHa; lane 4, cDNA of SiHa; lane 5,total RNA of Hela; lane 6, cDNA of Hela. (B) Tim-3 protein was determined by Western blotting in Hela and SiHa cell lines. (C) Hela cells stained with immunofluorescent with anti-Tim-3 antibody and observed with a confocal laser scanning microscope. Tim-3 protein (green, arrowheads) is in the cytoplasm of Hela. Cell nuclei (red) were visualized by staining with PI.

Repressing Tim-3 Expression Inhibited Migration and Invasion of Hela Cells

To further understand the correlation of Tim-3 and tumor metastasis, we infected Hela cells with ADV-antisense Tim-3. At a Multiplicity of Infection (MOI) of 1, ADV-antisense Tim-3 infected Hela cells showed significantly decreased Tim-3 expression with a minor cell death response; thus, this concentration was used in the following experiments (Fig. 4A–B). Since cancer cell migration and invasion are directly related to metastasis, a wound healing assay and a cell invasion assay were performed to determine whether repression of Tim-3 expression inhibits Hela cell migration and invasion. Hela cells infected with either ADV-antisense Tim-3 or with ADV-GFP as a control were evaluated for 24 h and 48 h. As shown in Figure 5A–B, at 24 h and 48 h respectively, ADV-antisense Tim-3 infected cells showed 40% and 70% wound closure, while ADV-GFP infected cells showed 60% and 100%, suggesting that inhibition of Tim-3 expression decreased the migration of Hela cells. As shown in Figure 5C–D, the number of Hela cells that passed through the filter in the ADV-antisense Tim-3 group (151±33) was markedly less than that in the ADV-GFP group (545±48), which shows that inhibition of Tim-3 expression suppressed Hela cell invasion in vitro.

Figure 4

Effect of ADV- antisense Tim-3 on Hela cell line.

(A) Western blotting was performed to detect Tim-3 expression in Hela cells infected with ADV-GFP and ADV- antisense Tim-3 or treated with PBS. (B) Typical result of cell apoptosis determined by flow cytometry in Hela cells infected with ADV-antisense Tim-3 and ADV-GFP or treated with PBS. Data are represented as the mean ± SD of triplicates.

Figure 5

Effect of Tim-3 inhibition on Hela cell migration and invasion in vitro.

(A) Cell migration capability was determined with a wound healing assay. Photographs were taken immediately (0 h), at 24 h and 48 h after wounding. (B) Quantification of wound closure. The data present the mean distance of cell migration to the wound area at 24 h and 48 h after wounding in three independent wound sites per group. (C) The ability of the cells to invade Matrigel was analyzed by the transwell invasion assay through a gel matrix. Hela cells were either infected with ADV-GFP or with ADV-antisense Tim-3, After 10 h viable invasive cells were fixed and counted. Values and error bars shown in this graph represent the averages and standard deviations respectively, of three independent experiments. (D) Representative images of the transwell invasion assay.

Discussion

In our previous work, we found that Tim-3 was preferentially expressed in lymphoma-derived endothelial cells (ECs), and that the level of Tim-3 in B cell lymphoma endothelium was closely correlated to both dissemination and poor prognosis. Tim-3+ ECs modulated T cell response to lymphoma surrogate antigens by suppressing activation of CD4+ T lymphocytes through the activation of the interleukin-6–STAT3 pathway, inhibiting Th1 polarization, providing protective immunity, and facilitating the establishment of lymphoma immune tolerance. Although studies suggest that Tim-3 is involved in the immune regulation of tumors, its direct expression in the tumor cell and its function in tumor metastasis were still unknown.

In the present study, we found that Tim-3 was preferentially expressed in cervical cancer tissues, and its expression was significantly correlated with advanced cancer grades (p = 0.018), histological grades (p = 0.038), metastasis (p = 0.004) and shorter survival (p = 0.006). The presence of Tim-3 mRNA and protein in two cervical cancer cell lines (Hela and Siha) in addition to the localization of Tim-3 to the cytoplasm of the Hela cell confirmed that Tim-3 was indeed expressed in the cervical cancer cell. To clarify the link between Tim-3 and tumor metastasis, we used ADV-antisense Tim-3 to do a wound healing assay and transwell invasion assay. We found that ADV-antisense Tim-3 infected cells showed 40% and 70% wound closure at 24 h and 48 h respectively, compared to the 60% and 100% seen in ADV-GFP infected cells. In addition, markedly fewer Hela cells passed through the filter in the ADV-antisense Tim-3 group (151±33) than that in the ADV-GFP group (545±48). These results strongly suggest that down-regulating the expression of Tim-3 decreases the migration and invasion of Hela cells significantly.

In line with our study, Wiener et al [15] also found that Tim-3 was expressed not only in mast cells around melanomas, but also in tumor cells in tissue sections and human melanoma cell lines WM35 and HT168-M1. Meanwhile, Kikushige and Jan [11], [12] identified Tim-3 expression on leukemia stem cells (LSC) in patients with acute myeloid leukemia. To determine whether Tim-3 is universally expressed in tumor cells, additional studies on other cancer cells are required.

Despite the progress made in understanding the involvement of Tim-3 in tumor immunity, the link between Tim-3 expression and tumor cell itself has not yet been defined. One of our most striking findings is that when we used ADV-antisense Tim-3 to down-regulate the expression of Tim-3 in HeLa cells, both the migration and invasion of HeLa cells were inhibited significantly. Indeed, high score expression of Tim-3 was significantly correlated with metastasis (p = 0.004) in cervical cancer patients. The next step would be to determine how Tim-3 expression is correlated with tumor metastasis and which pathways are involved? In our previous work, we have verified that Tim-3 can activate the IL-6-STAT3 pathway. Tim-3 expression increased the EC-derived production of IL-6 by almost 10-fold, and the addition of IL-6 significantly increased the level of phosphorylated STAT3 (p-STAT3). According to published studies [16]–[19], the IL-6-STAT3 pathway plays an important role in tumor metastasis. STAT3 can promote premetastatic niche formation and metastatic cells can escape the pro-apoptotic effects of TNF-α through increased autocrine IL-6-STAT3 signalling. In addition, inhibition of p-STAT3 enhances IFN-α efficacy against metastatic melanoma in a murine model. Based on these previous findings and our results from this study, we hypothesize that Tim-3 might facilitate tumor metastasis through the IL-6-STAT3 pathway. Because distant metastasis is a major factor in the survival of individuals with cervical cancer, Tim-3 may be a critical prognostic marker for cervical cancer.

Taken together, our study suggests that Tim-3 not only negatively regulates anti-tumor immunity, but also influences cancer development directly via its expression in cancer cells. For the first time, we have associated the expression of Tim-3 in tumor cells with worse clinical pathological parameters in cervical cancer. In addition, we found that the inhibition of Tim-3 protein expression can prevent tumor metastasis. Thus, it is rational for future experiments to explore Tim-3 as a target for anti-cancer therapy, tumor immunotherapy, and in the control of metastatic diseases.