Aberrant GIMAP2 expression affects oral squamous cell carcinoma progression by promoting cell cycle and inhibiting apoptosis.

GTPases of immunity-associated protein 2 (GIMAP2) is a GTPase family member associated with T cell survival. However, its mechanisms of action in oral squamous cell carcinoma (OSCC) remain largely unknown. Therefore, the present study aimed to elucidate the possible role of GIMAP2 in OSCC development by investigating its expression levels and molecular mechanisms in OSCC. Reverse transcription quantitative PCR, immunoblotting and immunohistochemistry indicated that GIMAP2 expression was significantly upregulated (P<0.05) in OSCC-derived cell lines and primary OSCC specimens compared with that in their normal counterparts. GIMAP2-knockdown OSCC cells exhibited decreased cell growth, which was associated with cyclin-dependent kinase (CDK)4, CDK6 and phosphorylated Rb downregulation and p53 and p21 upregulation. In addition to cell cycle arrest, GIMAP2 affected anti-apoptotic functions in GIMAP2-knockdown cells by upregulating Bcl-2 and downregulating Bax and Bak. These findings indicated that GIMAP2 may significantly influence OSCC development and apoptosis inhibition and thus is a potential biomarker of OSCC. Copyright: © Komatsu et al.

Introduction

Oral cancer, the sixth most common cancer, is a severe and growing problem with an estimated incidence of ~275,000 cases annually (1). Oral squamous cell carcinoma (OSCC), the commonest type of oral cancer, can occur via numerous processes during which multiple genetic events alter the normal functioning of oncogenes and tumor suppressor genes. Cancer-related genes display the following six fundamental features: Growth signal self-sufficiency, insensitivity to growth-inhibitory signals, apoptosis evasion, limitless replicative potential, sustained angiogenesis and the ability to invade and metastasize (2). Previous studies have shown that OSCC development is associated with cell proliferation and apoptosis rates (3–5); thus, it is necessary to understand the mechanisms underlying malignant tumors to develop new and effective treatment strategies.

GTPases of immunity-associated proteins (GIMAPs), also known as immunity-associated nucleotide binding proteins or IMAPs, are a family of GTPases found in vertebrates and plants. Humans have seven GIMAPs clustered on chromosome 7, consisting of an amino-terminal guanine-nucleotide binding domain (G-domain) followed by varying C-terminal extensions of 50–100 amino acids long (6,7). GIMAPs have seven isoforms expressed in humans (namely, GIMAP1, GIMAP2, GIMAP4-GIMAP8; GIMAP3 is a pseudogene) (8), regulating T cell survival during their development, selection and homeostasis and they may be linked to the onset of T lymphopenia, leukemia and autoimmunity (9,10). GIMAPs may also be involved in the mitochondrial regulation of lymphocyte apoptosis by interacting with the Bcl-2 family proteins (8). In addition, it has been suggested that GIMAP3 and GIMAP5 are involved in the mitochondria-mediated apoptosis regulatory pathway (8). GIMAP3 and GIMAP5 have similar primary structures and they are both localized in the intracellular membrane fraction, where several Bcl-2 family proteins are located. Furthermore, both GIMAP3 and GIMAP5 co-immunoprecipitate with Bcl-2 and Bcl-xL in T cells (8). Patterson et al (11) indicate that GIMAP5 is associated with lymphocyte survival, autoimmunity and colitis and is essential for the inactivation of glycogen synthase kinase-3β following T cell activation. The preliminary examination of the mRNA expression levels of GIMAP1, GIMAP2, GIMAP4-GIMAP8 in OSCC the present study revealed only GIMAP2 expression to be significantly (P<0.05; Fig. 1A and B) increased compared with that in HNOKs with the other isoforms showing low expression and GIMAP4 and GIMAP7 not being expressed (Fig. S1).

Figure 1.

Evaluation of GIMAP2 expression in OSCC-derived cell lines and primary OSCC specimens. (A) Reverse transcription-quantitative PCR analysis showed that GIMAP2 mRNA expression was significantly upregulated (*P<0.05, Dunnett's test) in the two OSCC cell lines compared with that in HNOKs. (B) Western blotting was conducted three times per cell type using GIMAP2 primer No. 1. The results are expressed as mean ± standard error of the mean of triplicate data. GIMAP2 expression was upregulated (*P<0.05, Dunnett's test) in four OSCC cell lines compared with that in HNOKs. Non-continuous parts of blots probed on the same membrane are indicated using vertical lines. Representative IHC results for GIMAP2 expression in (C) the normal oral tissues (scale bar, 50 µm) and (D) primary OSCC tissues (scale bar, 50 µm). (E) IHC scores showed GIMAP2 expression in primary OSCC (n=100) and normal tissue samples. The GIMAP2 IHC scores of the normal oral tissues and primary OSCC tissues ranged from 21.2 to 134.5 (median, 44.6) and from 30.2 to 148.0 (median, 109.8), respectively. GIMAP2 expression was considerably (*P<0.05, Wilcoxon signed-rank test) higher in OSCC tissues compared in normal oral tissues. (F) The Cancer Genome Atlas data show the GIMAP2 expression status in primary HNSCC; n=499) and normal tissue samples (n=44). GIMAP2 expression was considerably higher in HNSCC tissues than in normal tissues (*P<0.05, Student's t-test). (G) The ROC curve analysis indicated that the cut-off value was 104.1 and the AUC was 0.81. GIMAP2, GTPases of immunity-associated proteins 2; OSCC, oral squamous cell carcinoma; HNOK, human normal oral keratinocyte; IHC, immunohistochemistry; ROC, receiver operating characteristic; AUC, area under the curve; TPF, true-positive fraction; FPF, false-positive fraction.

GIMAP2, expressed in humans with no orthologs in mice and rats, is the second-largest protein in the human GIMAP family containing two C-terminal hydrophobic regions (12). According to BioGPS (https://www.biogps.org), T cells, blood cells, including platelets and the spleen express GIMAP2. To the best of the authors knowledge, there are only two studies on GIMAP2. In one of the studies, Schwefel et al (6) showed that GIMAP2 assembles into a GTP-dependent scaffold and the C-terminal amino acid stretch targets GIMAP2 toward lipid droplets. In the other study, Schwefel et al (7) found that GIMAP2 expression was maintained in all the examined human lymphoma T cell lines, whereas the expression of other GIMAP members was inhibited in these tumor cell lines. This is in line with the observations of the present study and suggests a favorable role of GIMAP2 in cancer cell survival. Certain GIMAPs may be associated with T lymphopenia and leukemia (9–11,13), although the biological functions of most GIMAPs, including GIMAP2, remain to be elucidated. As the primary function of GIMAP2 in the progression of solid cancers, such as OSCC, is currently unclear, remain investigated its expression and molecular mechanisms in OSCC.

Materials and methods

Cell and tissue samples

Human OSCC-derived cell lines, including HSC-2 (RBRC-RCB1945; oral cavity), HSC-3 (JCRB-0623; tongue), HSC-4 (RBRC-RCB1902, tongue), HSC3-M3 (JCRB-1354, tongue), Sa3 (RBRC-RCB0980, gingiva), Ho-1-N-1 (JCRB-0831, buccal mucosa), KOSC2 (JCRB-0126.1, mouth floor), SAS (RBRC-RCB 1974, tongue) and Ho-1-u-1 (RBRC-RCB2102, mouth floor), were purchased from the JCRB Cell Bank and RIKEN BioResource Center. All cancer cells were cultured in low-glucose Dulbecco's modified Eagle's medium (DMEM; Sigma-Aldrich; Merck KGaA) supplemented with 10% heat-inactivated fetal bovine serum (FBS; HyClone; Cytiva) and 50 U/ml of penicillin and streptomycin at 37°C in a humidified 5% carbon dioxide atmosphere. Human normal oral keratinocytes (HNOKs) were obtained from three healthy donors. The donors comprised 2 men and 1 woman. The donors were 27, 28 and 22 years old, respectively, and were recruited between April 2017 and June 2017. HNOKs were cultured in Oral Keratinocyte Medium New Zealand BPE (ScienCell Research Laboratories, Inc.; cat. no. 2611) as described previously (14,15). The ethics committee of the Graduate School of Chiba University approved this study (protocol number 680). All patients provided written informed consent prior to their inclusion in the study.

mRNA expression analysis

The present study performed reverse transcription quantitative PCR (RT-qPCR) (14–18). Cell were grown to 80% confluence in a 10-cm dish. Total RNA was isolated using TRIzol® Reagent (Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 15596018), according to the manufacturer's instructions. cDNA was generated using ReverTra Ace qPCR RT Master Mix (Toyobo Life Science; cat. no. FSQ-201) according to the manufacturers' instructions. RT-qPCR was performed in a 20-µl reaction volume using FastStart SYBR-Green Master (Roche Diagnostics; cat. no. 4673492001) according to the manufacturer's protocol (17,18). The following primers were used to amplify GIMAP2: GIMAP2 No. 1, forward, 5′-CGATTCAAATGCTTGCTTCC-3′ and reverse, 5′-GGACCAAAATGAACACAGTCAC-3′ (Thermo Fisher Scientific, Inc., Waltham, MA, USA); GIMAP2 No. 2, forward, 5′-TGGAAGGACCACTGTGAAGC-3′ and reverse, 5′-GTCCTGTGAGGTATAGCGGC-3′ (Greiner Bio-One Co Ltd.); and GIMAP2 No. 3 forward, 5′-GGATGCCATGGGACACACAA-3′ and reverse, 5′-TAAAGGCACAGATTCGCCCA-3′ (Greiner Bio-One Co Ltd.).

In addition, the following primers and universal probes were used: GIMAP1, forward, 5′-TCGAGCTCCTCTCTGGTTATG-3′ and reverse, 5′-TGCAGTCTCAGCCTATGCAC-3′ (Thermo Fisher Scientific, Inc.); GIMAP4, forward, 5′-ACACCAGGGGCCAGTTATG-3′ and reverse, 5′-TGCTGTTTCCTGTTGCACTT-3′ (Thermo Fischer Scientific); GIMAP5, forward, 5′-TGGGGGACACACTCCATAAT-3′ and reverse, 5′- GCAGACGCAGTTAAGGAGGA-3′ (Thermo Fisher Scientific, Inc.); GIMAP6, forward, 5′-GATGGAGGAAGAAGAATATGAACAA-3′ and reverse, 5′-TTCTGTTCTTTCTCCCTTAGACCT-3′ (Thermo Fisher Scientific, Inc.); GIMAP7, forward, 5′-CTCTAGAACTTAGGCACGTACAAGAC-3′ and reverse 5′-CTCTAGAACTTAGGCACGTACAAGAC-3′ (Thermo Fisher Scientific, Inc.); GIMAP8, forward, 5′-CAGATATAGTGCCTTCAACTACCG-3′ and reverse, 5′-GGACAATGTTCAGGGTTTCTTT-3′ (Thermo Fisher Scientific, Inc.); and glyceraldehyde 3-phosphate dehydrogenase (GAPDH), forward, 5′-AGCCACATCGCTCAGACA-3′ and reverse, 5′-GCCCAATACGACCAAATCC-3′ (Thermo Fisher Scientific, Inc.). The transcript amounts for the target genes were estimated from the respective standard curves and normalized to the GAPDH.

A LightCycler 480 PCR system (Roche Diagnostics) was used with the following RT-qPCR conditions: initial denaturation at 95°C for 10 min, 45 amplification cycles of denaturation at 95°C for 10 sec, annealing at 60°C for 30 sec and extension at 72°C for 1 sec, followed by a cooling step at 40°C for 30 sec. This experiment was performed in triplicate.

Western blotting

Protein extraction and immunoblotting were conducted as previously described (16,18,19). Cells were washed three times with cold phosphate buffered saline (FUJIFILM Wako Pure Chemical Corporation; cat. no. 045-29795) and gently and briefly centrifuged (11,000 × g; 4°C; 5 min). The cell pellets were then incubated at 4°C for 10 min in a lysis buffer (7 M urea, 2 M thiourea, 4% w/v CHAPS, and 10 mM Tris, pH 7.4). The total protein concentration was measured using a dye-binding method based on the Bradford assay with Bio-Rad Protein Assay Dye Reagent Concentrate (Bio-Rad Laboratories, Inc.; cat. no. 5000006JA). A total of 20 µg of the protein was loaded per lane.

Protein extracts were electrophoresed on 4–12% Bis-Tris gel (Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. NP0336BOX) and transferred to nitrocellulose membranes (Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 77010) and blocked for 1 h at room temperature (25°C) with Blocking One (Nacalai Tesque, Inc.; cat. no. 03953-95). The membranes were then incubated with polyclonal rabbit anti-GIMAP2 antibody (Rabbit polyclonal antibody specific for human GIMAP2; cat. no. HPA013589; 1:100; Atlas Antibodies). In addition, the following antibodies were used: p21 (Mouse monoclonal antibody specific for human p21; cat. no. sc-6246, 1:200; Santa Cruz Biotechnology, Inc., Inc.), cyclin-dependent kinase (CDK)4 (Mouse monoclonal antibody specific for human CDK4; cat. no. sc-23896, 1:500; Santa Cruz Biotechnology, Inc.), CDK6 (Mouse monoclonal antibody specific for human CDK6; cat. no. sc-7961, 1:200; Santa Cruz Biotechnology, Inc.), Cyclin D1 (Mouse monoclonal antibody specific for human Cyclin D1; cat. no. sc-20044, 1:200; Santa Cruz Biotechnology, Inc.), Cyclin E (Mouse monoclonal antibody specific for human Cyclin E; cat. no. sc-377100, 1:100; Santa Cruz Biotechnology, Inc.), CDK2 (Mouse monoclonal antibody specific for human CDK2; cat. no. sc-6248, 1:200; Santa Cruz Biotechnology, Inc.), p53 (Mouse monoclonal antibody specific for human p53; cat. no. sc-393031, 1:100; Santa Cruz Biotechnology, Inc.), Rb (Mouse monoclonal antibody specific for human Rb, NBP2-54476IR, 1:200; Novus Biologicals), phosphorylated (p)-Rb (Ser780; Rabbit polyclonal antibody specific for human p-Rb (Ser780); cat. no. ab47763, 1:500; Abcam), Bcl-2 (Mouse monoclonal antibody specific for human Bcl-2; cat. no. sc-7382, 1:50; Santa Cruz Biotechnology, Inc.), Bak (Mouse monoclonal antibody specific for human Bak; cat. no. sc-517390, 1:200; Santa Cruz Biotechnology, Inc.), Bax (Mouse monoclonal antibody specific for human Bax; cat. no. sc-7480, 1:200; Santa Cruz Biotechnology, Inc.), Bcl-xL (Mouse monoclonal antibody specific for human Bcl-xL; cat. no. sc-8392, 1:200; Santa Cruz Biotechnology, Inc.) and mouse α-tubulin (Mouse monoclonal antibody specific for human α-tubulin; cat. no. sc-5286, 1:1,000; Santa Cruz Biotechnology, Inc.) overnight at 4°C. The membranes were then washed with TBS-T (1% Tween) and incubated with horseradish peroxidase-conjugated anti-rabbit IgG (Promega Corporation; cat. no. W4011) or anti-mouse IgG as a secondary antibody (Promega Corporation; cat. no. W4021), for 1 h at room temperature (25°C). Finally, the membranes were developed using Clarity Western ECL Substrate (Bio-Rad Laboratories, Inc.; cat. no. 170–5061), and immunoblotting was visualized with the ChemiDoc XRS Plus system (Bio-Rad Laboratories, Inc.). The signal intensities were quantitated using the Image Lab system 6.1 Software (Bio-Rad Laboratories, Inc.). Densitometric GIMAP2 protein data were normalized to α-tubulin protein levels.

Immunohistochemistry (IHC) analysis

IHC analysis was performed using 100 tissue samples according to a previously described scoring system (18,20–24). To determine the cut-off value for the GIMAP2 IHC clinical parameter scores, the scores of 100 samples were evaluated by receiver operating characteristic (ROC) curve analysis using a bell curve in Microsoft Excel (Microsoft Corporation) and Excel Statistics (Social Survey Research Information Co., Ltd.). Samples with a score above the cut-off value were defined as GIMAP2-positive. Polyclonal rabbit anti-GIMAP2 antibody (Rabbit polyclonal antibody specific for human GIMAP2; cat. no. HPA013589, 1:50; Atlas Antibodies) was used as the primary antibody and Dako EnVision+ System- HRP Labeled Polymer Anti-Rabbit (Agilent Technologies, Inc.; cat. no. K4003) was used as the secondary antibody.

Gene expression data for patients with head and neck squamous cell carcinoma (HNSCC) was downloaded from The Cancer Genome Atlas (TCGA) project webpage (http://cancergenome.nih.gov). In total, 499 patients with complete data were selected (i.e., each had a dataset of microRNA expression and publicized clinical information).

Transfection

In a 6-well tissue culture plate, HSC-2 and HSC-3 cells were cultured to a 50–70% confluency in antibiotic-free DMEM supplemented with FBS. Stable knockdown transfectants were established by transfecting the cell lines (HSC-2 and HSC-3) with GIMAP2-targeting short hairpin (sh)RNA [shGIMAP2; GIMAP2 shRNA Plasmid(h): cat. no. sc-89424-SH; Santa Cruz Biotechnology, Inc.] and control shRNA (shMock) (Control shRNA Plasmid-A: cat. no. sc-108060; Santa Cruz Biotechnology, Inc.) using Lipofectamine® 3000 (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's instructions. The concentration of the shRNA plasmid was 0.1 µg/µl. The transfection was carried out at 37°C for 48 h. Two to three weeks after transfection, viable colonies were transferred to new dishes. shGIMAP2- and shMock-transfected cells were used for further experiments.

The vector GIMAP2 Human Untagged Clone (OriGene Technologies, Inc.; cat. no. SC101332) was transiently transfected into stable transfectants to confirm the effects of GIMAP2 knockdown. The circular untagged cloning vector PCMV6-XL4 (OriGene Technologies, Inc.; cat. no. PCMV6XL4) was used as the negative control. Stable transfectants were isolated using low-glucose DMEM (Sigma-Aldrich; Merck KGaA) supplemented with 10% heat-inactivated FBS, 50 U/ml penicillin and streptomycin and 1 µg/ml puromycin (Santa Cruz Biotechnology, Inc.) at 37°C in a humidified 5% carbon dioxide atmosphere.

Proliferation assay and cell cycle analysis

Proliferation assays were performed as previously described (16,18,19,25). Cell cycle was analyzed by flow cytometry with a BD Accuri C6 Flow Cytometer (Becton-Dickinson and Company) and FlowJo 10.5.3 software (FlowJo LLC) as previously described (21,26–28).

Caspase 3/7 activity assay

GIMAP2-knockdown cells (HSC-2 shGIMAP2 and HSC-3 shGIMAP2) and shMock cells (HSC-2 shMock and HSC-3 shMock) (2×103 cells/well) were seeded in white-walled 96-well plates and cultured for 2, 4, or 6 days. The activity levels of caspases-3/7 were analyzed using the Caspase-Glo 3/7 assay system, according to the manufacturer's instructions (Promega Corporation; cat. no. G8091). Briefly, the plates were equilibrated to room temperature (25°C). Caspase-Glo 3/7 reagent (100 µl) was added into each well. Following incubation at room temperature for 30 min, the luminescence signal was detected with Synergy HTX (BioTek Instruments, Inc.).

Statistical analysis

Wilcoxon signed-rank test (P<0.05) was performed to identify significant associations and ROC curve analysis was used to define a cut-off value to confirm whether samples were GIMAP2-positive or -negative for the classified clinical parameters (Fig. 1E). Wilcoxon signed-rank test was used to compare the median values of paired samples. Furthermore, areas under the curve were determined to confirm the usefulness of this method (Fig. 1G). Dunnett's test was used for the analysis of data shown in Figs. 1A and B, 2A and B, 3A, 4, S1, S2, S3 and S5. Dunnett's post hoc test was performed after the one-way ANOVA. Two-tailed Student's t-test was used for the analysis of data shown in Figs. 1F, 3B and C, 5 and S2.

Figure 2.

shRNA-mediated GIMAP2 knockdown in OSCC cells (HSC-2 and HSC-3-derived transfectants). (A) GIMAP2 mRNA expression was significantly lower (*P<0.05, Dunnett's test) in shGIMAP2 cells compared with shMock cells. (B) Immunoblotting analysis showed that the GIMAP2 level was lower (*P<0.05, Dunnett's test) in shGIMAP2 cells compared with in shMock cells. Western blotting was conducted three times per cell type. sh, short hairpin; GIMAP2, GTPases of immunity-associated proteins 2; OSCC, oral squamous cell carcinoma.

Figure 3.

Cell proliferation assay and flow cytometric analysis. (A and B) shGIMAP2 and shMock cells were counted on seven consecutive days. The results are expressed as mean ± standard error of the mean of triplicate experiments. Cell growth was significantly inhibited (*P<0.05, Dunnett's test) in shGIMAP2 cells after 144 h of culture in HSC-2 cells and after 168 h of culture in HSC-3 cells. Cell growth was significantly (†P<0.05, Student's t-test) higher in GIMAP2 transiently transfected cells (HSC-2 shGIMAP2-1 rescue and HSC-3 shGIMAP2-1 rescue) than in shGIMAP2 control cells (HSC-2 shGIMAP2-1 rescue control and HSC-3 shGIMAP2-1 rescue control) after 24 h of culture in HSC-2 cells and after 48 h of culture in HSC-3 cells. (C) Flow cytometric analysis showed that the percentage of shGIMAP2 cells in the G1 phase was higher compared with shMock cells. sh, short hairpin; GIMAP2, GTPases of immunity-associated proteins 2.

Figure 4.

(A, B) Western blotting of proteins associated with cell cycle and apoptosis regulation. Western blotting analyses showed that the p53 and p21 levels were upregulated and CDK4, CDK6 and p-Rb (S780) levels were downregulated in shGIMAP2 cells compared with those in shMock cells (*P<0.05, Dunnett's test). Western blotting analyses showed that the Bak and Bax levels were upregulated and the Bcl-2 level was downregulated in shGIMAP2 cells compared with those in shMock cells (*P<0.05, Dunnett's test). Western blotting was conducted three times per cell type. (C) Western blot analysis revealed the p-Rb protein levels and then normalized to total Rb (*P<0.05, Dunnett's test). CDK, cyclin-dependent kinase; sh, short hairpin; GIMAP2, GTPases of immunity-associated proteins 2; p-, phosphorylated.

Figure 5.

Caspase 3/7 activity. Caspase 3/7 activity in shGIMPA2 HSC-2 and HSC-3 at 2, 4 and 6 days after cell seeding. Data are reported as mean ± SD (n>3) relative to the control (*P<0.05, Student's t-test). GIMAP2, GTPases of immunity-associated proteins 2; sh, short hairpin.

Results

GIMAP2 upregulation in OSCC-derived cells

To evaluate GIMAP2 expression, RT-qPCR and western blotting analyses of nine OSCC-derived cell lines and HNOKs was performed. The expression of GIMAP2 mRNA was significantly upregulated (P<0.05) in two OSCC cell lines compared with that in HNOKs (Fig. 1A and S2). The GIMAP2 level was significantly upregulated (P<0.05, Dunnett's test) in the four OSCC cell lines compared with that in HNOKs (Fig. 1B). Protein level prediction based on the mRNA level is inaccurate because the mRNA and protein levels do not strictly correlate. As one of the reasons is presumably a mutation in the primer-binding site, the mRNA level was verified in Ho-1-u-1 cells by PCR using two additional primer sets targeting different coding regions. As shown in Fig. S2A, GIMAP2 was not expressed in Ho-1-u-1 cells, resulting in a significant difference compared with that in HNOKs.

GIMAP2 expression in primary OSCCs

Representative IHC data for GIMAP2 immunoreactivity in normal oral tissues and OSCC samples (magnification, ×400) are shown in Fig. 1C and D; strong cytoplasmic staining for GIMAP2 was detected in OSCC samples, whereas the normal oral tissues showed negative immunoreactivity. The IHC scores of tissue specimens from 100 patients with OSCC were used to investigate the clinical correlations between GIMAP2 expression and pathological characteristics. GIMAP2 expression was significantly higher in OSCC tissues compared with normal oral tissues (P<0.05; Fig. 1E). The GIMAP2 IHC scores of the adjacent normal tissues ranged from 21.2 to 134.5 (median, 44.6); whereas those of the OSCC tissues ranged from 30.2 to 148.0 (median, 109.8). Gene expression data analysis of patients from TCGA revealed that GIMAP2 expression was significantly higher in HNSCC tissues compared with normal oral tissues (P<0.05; Fig. 1F). To determine the cut-off value for the GIMAP2 IHC scores, a ROC curve analysis was performed, which yielded an area under the curve of 0.81 and a cut-off value of 104.1 (Fig. 1G). The clinical classifications of GIMAP2-positive OSCC were significantly associated (P<0.05) with T-primary tumors and the OSCC stage (Table I).

Table I.

Correlation between GIMAP2 expression and the clinical classification of oral squamous cell carcinoma.

	Results of immunostaining No. of patients (%)
Variable	Total	GIMAP2 negative	GIMAP2 positive	P-value
Age at surgery (years)
>70	41	14 (34)	27 (66)	0.580[a]
60-70	35	16 (46)	19 (54)
<60	24	10 (42)	14 (58)
Sex
Male	57	21 (37)	36 (63)	0.458[a]
Female	43	19 (44)	24 (56)
T-primary tumor
T1+T2	51	32 (63)	19 (37)	0.0004[a,b]
T3+T4	49	8 (16)	41 (84)
N-regional lymph node
Negative	63	24 (38)	39 (62)	0.612[a]
Positive	37	16 (43)	21 (57)
Stage
I	15	10 (67)	5 (33)	0.004[a,b]
II	20	12 (60)	8 (40)
III	18	7 (39)	11 (61)
IV	47	11 (23)	36 (77)
Histopathologic type
Good	68	28 (41)	40 (59)	0.811[a]
Moderate	28	11 (39)	17 (61)
Poor	4	1 (25)	3 (75)
Vascular Invasion
Negative	62	24 (39)	38 (61)	0.737[a]
Positive	38	16 (42)	22 (58)

χ2 test;

P<0.05. GIMAP2, GTPase of immunity-associated protein 2.

Establishment of GIMAP2-knockdown cells

As GIMAP2 was significantly upregulated in OSCC-derived cells (Figs. 1A and B and S2), its expression in GIMAP2-knockdown cells (HSC-2 shGIMAP2 and HSC-3 shGIMAP2) was investigated. GIMAP2 mRNA and protein expressions were significantly lower in shGIMAP2 cells compared with shMock cells (P<0.05; Figs. 2A and B and S3). As other GIMAP isoforms were not highly expressed in the OSCC cell lines examined, the GIMAP2 shRNA used might not have affected other GIMAP isoforms, as it is specific for GIMAP2 (P<0.05; Fig. S1).

Growth of GIMAP2-knockdown cells

To investigate the effect of GIMAP2 knockdown on cell growth, a cell proliferation assay was performed and it was found that cell growth was significantly lower in shGIMAP2 cells (HSC-2 shGIMAP2 and HSC-3 shGIMAP2) compared with shMock cells (P<0.05; Fig. 3A). For validation, the vector GIMAP2 Human Untagged Clone was transiently transfected into stable transfectants to rescue GIMAP2 expression. The present study confirmed that the expression of GIMAP2 protein increased in GIMAP2-transfected shMock cells (HSC-2 shMock GIMAP2 overexpression and HSC-3 shMock GIMAP2 overexpression), but not in the control shMock cells (HSC-2 shMock control and HSC-3 shMock control; Fig. S4). In addition, GIMAP2 was expressed in GIMAP2-transfected cells (HSC-2 shGIMAP2-1 rescue and HSC-3 shGIMAP2-1 rescue), but not in shGIMAP2 cells (HSC-2 shGIMAP2-1 rescue control and HSC-3 shGIMAP2-1 rescue control; Fig. S4). In addition, cell growth was significantly (P<0.05) higher in GIMAP2-transfected shGIMAP2 cells (HSC-2 shGIMAP2-1 rescue and HSC-3 shGIMAP2-1 rescue) than in shGIMAP2 cells (HSC-2 shGIMAP2-1 rescue control and HSC-3 shGIMAP2-1 rescue control; Fig. 3B).

Cell cycle analysis of GIMAP2-knockdown cells

Cell cycle analysis showed that the percentage of shGIMAP2 cells in the G1 phase was significantly higher than that of shMock cells (P<0.05; Fig. 3C). In addition, G1 phase-related protein expression in shGIMAP2 cells revealed CDK4, CDK6 and phosphorylated (p-)Rb (S780) to be downregulated, whereas p53 and p21 were upregulated (P<0.05; Figs. 4 and S5), indicating that shGIMAP2 suppressed proliferation by arresting the cell cycle in the G1 phase.

Apoptosis-related protein expression

The GIMAP family, including GIMAP2, is hypothesized to be associated with apoptosis (12). The evaluation of apoptosis-related protein expression in shGIMAP2 cells showed that Bcl-2 was significantly downregulated, whereas Bak and Bax were upregulated (P<0.05; Figs. 4 and S5), suggesting that GIMAP2 may be associated with apoptosis inhibition.

To determine whether the expression of GIMAP2 was associated with the inhibition of apoptosis, the caspase-3/7 activity in shGIMAP2 cells was evaluated. Fig. 5 shows that shGIMAP2 cells exhibited significant activation of caspase-3/7 compared with that in shMock cells on days 4 and 6.

Discussion

To the best of the authors' knowledge, the present study is the first to demonstrate that GIMAP2 is upregulated in HNSCC and is positively associated with TNM classification. The knockdown of GIMAP2 revealed that it droves cell proliferation by arresting the cell cycle in the G1/S phase and that it may inhibit apoptosis via Bcl-2 upregulation and Bak and Bax downregulation, suggesting that GIMAP2 serves an important role in TNM classification in human OSCC. The cell cycle analysis showed that GIMAP2 knockdown induced G1/S phase arrest in OSCC cells and decreased CDK4/CDK6 and p-Rb activities. Recently, the changes in Rb phosphorylation via CDK4/CDK6 signaling have been reported to control tumor cell proliferation by dysregulating cell cycle progression in several types of tumors, including lung, prostate, head and neck cancers (29–33). Accordingly, the present study confirmed that the expression of GIMAP2 activated the cyclin D1/CDK4/CDK6 complex, which phosphorylates Rb and results in OSCC proliferation. In addition, increased p53 and p21 expression was observed in GIMAP2-knockdown cells; a study reports that abnormal p53 and p21 signaling correlates with tumor cell proliferation via the Rb/E2F pathway (34), in which E2F activation suppresses the function of p53 and p21 and induces excessive OSCC proliferation. Thus, the data of the present study suggested that GIMAP2 may be associated with cell cycle progression via the Rb/E2F pathway. In addition to cell cycle arrest, p53 regulates apoptosis via the release of cytochrome c and the transcriptional regulation of pro-apoptotic genes (35). Thus, GIMAP2-knockdown OSCC cells exhibited decreased cell growth, which was associated with CDK4, CDK6 and p-Rb downregulation and p53 and p21 upregulation. The present study found that not only the pro-apoptotic proteins Bak and Bax, but also caspase-3/7 was upregulated in GIMAP2-knockdown cells, whereas the anti-apoptotic protein Bcl-2 was downregulated, consistent with previous findings for other GIMAP family proteins (35). The Bcl-2 family members are associated with cancer development (36), whereas GIMAP2 may be associated with apoptotic signals (37) and inhibit Bcl-2 family-mediated apoptotic signals to support cancer cell survival and development.

Therefore, GIMAP2 may regulate OSCC tumor growth partly via the inhibition of apoptosis; however, further studies are warranted to elucidate the mechanism underlying GIMAP2 anti-apoptotic effects. GIMAP2 expression in OSCC cells was associated with high CDK4, CDK6 and p-Rb expression and low p53 and p21 expression, demonstrating an essential role of GIMAP2 in growth regulation. In addition, GIMAP2 expression was positively associated with Bcl2 and inversely associated with Bak and Bax expression, which may be indicative of a secondary function of GIMAP2 in controlling apoptosis. Thus, GIMAP2 expression could be a biomarker of OSCC progression and apoptosis inhibition.