Loss of p12^CDK2-AP1 Expression in Human Oral Squamous Cell Carcinoma with Disrupted Transforming Growth Factor-β-Smad Signaling Pathway

Abstract

We examined correlations between TGF-β1, TβR-I and TβR-II, p12^CDK2-AP1, p21^WAF1, p27^KIP1, Smad2, and p-Smad2 in 125 cases of human oral squamous cell carcinoma (OSCC) to test the hypothesis that resistance to TGF-β1-induced growth suppression is due to the disruption of its signaling pathway as a consequence of reduced or lost p12^CDK2-AP1. Immunoreactivity for TβR-II decreased in OSCC with increasing disease aggressiveness; however, no differences were observed for TβR-I and TGF-β1. The expression of TβR-II significantly correlated with p12^CDK2-AP1 and p27^KIP1 (P < .001 and P < .01, respectively). Furthermore, there was a significant relationship between TβR-II expression and p-Smad2 (P < .001). The in vivo correlation of the levels of TβR-II, p12^CDK2-AP1, and p27^KIP1 was confirmed in normal and OSCC cell lines. Additionally, in vitro analysis of TGF-β1-treated cells showed that TGF-β1 treatment of normal keratinocytes suppressed cell growth with upregulation of p-Smad2, p12^CDK2-AP1, and p21^WAF1 expression, whereas there was no effect on OSCC cell lines. These results provide evidence of a link between a disrupted TGF-β-Smad signaling pathway and loss of induction of cell cycle-inhibitory proteins, especially p12^CDK2-AP1 in OSCC, which may lead to the resistance of TGF-β1 growth-inhibitory effect on OSCC.

Introduction

Transforming growth factor-β1 (TGF-β1) is a potent growth inhibitor of various cell types, including epithelial, endothelial, and hematopoietic cells [1–5]. TGF-β1 signaling is mediated mainly by two serine threonine kinase receptors, TGF-β receptor I (TβR-I) and TGF-β receptor II (TβR-II), which activate Smad2/3 and Smad4 transcription factors. The phosphorylation and activation of these proteins are followed by the formation of the Smad2/3-Smad4 complex, which translocates to the nucleus regulating transcriptional responses to TGF-β1 [6–8]. Previous studies have demonstrated that loss of function or decreased expression of TβR-II is associated with tumor progression [9,10]. The mechanism of TGF-β1 growth inhibition in epithelial cells is mediated, in part, by induction of p21^WAF1 expression [11]. Recently, TGF-β1 has been found to induce the expression of a novel cyclin-dependent kinase (CDK) 2 inhibitor, p12 CDK2-associating protein 1 (p12^CDK2-AP1), in normal epithelial cells [12]. Mechanistically, p12 plays a role in TGF-β1-mediated growth suppression by modulating CDK2 activities and pRb phosphorylation [12].

Head and neck cancer affects more than half a million patients worldwide yearly; despite improvements in surgery, chemotherapy, and radiation therapy, overall survival has not improved over the past 30 years [13]. Molecular alterations frequently associated with head and neck cancer include cyclin D1 amplification or overexpression, decreased expression of p53 and pRb, and decreased expression of CDK inhibitors p27^KIP1, p21^WAF1, and p16^INK4A [14,15]. More recently, decreased expression of p12^CDK2-AP1 has been associated with the development of oral squamous cell carcinoma (OSCC) [16,17]. p12^CDK2-AP1 is an S-phase-associated growth suppressor that binds with DNA polymerase A/primase and/or CDK2 [18,19]. Studies have shown that p12 is differentially expressed in normal and tumor oral mucosa, with a reduced expression in 14.3% of oral dysplasias and in 64% to 72% of oral cancers, suggesting a potential role for p12^CDK2-AP1 as a tumor suppressor in oral keratinocytes [16,17].

A defective TGF-β-Smad signaling pathway, which leads to loss of response to the proliferation-inhibitory effect of TGF-β1, has been found in several human cancers, such as hepatocellular carcinoma, pancreatic cancer, colon carcinoma, and glioblastoma [20–23]. Recently, mutations of Smad2 and Smad4 have been found in human head and neck squamous cell carcinoma cell lines; however, the expression of TGF-β-Smad signaling proteins in OSCC has not been reported. In addition, the biologic effects of TGF-β1 on OSCC remain unknown [24]. In the present study, we tested the hypothesis that resistance to TGF-β1 growth suppression in OSCC is due to the preferential reduction or loss of p12 expression. We investigated the expression of TGF-β1, its receptors, and the intracellular TGF-β-Smad signaling pathway in human OSCC. We also examined whether TGF-β1 is biologically functional in oral cancer.

Materials and Methods

Tissue Samples

Tissue samples were obtained from 125 previously untreated OSCC patients at the Department of Oral and Maxillofacial Surgery, Ehime University School of Medicine (Ehime, Japan) from 1991 to 2003. The median age of the OSCC patients was 68.2 years (SD = 14.1). Twenty-eight patients were in stage I, 41 patients were in stage II, 23 patients were in stage III, and 33 patients were in stage IV, according to tumor-node-metastasis classification [25]. The grade of tumor differentiation was determined according to the criteria proposed by the World Health Organization [26]. Histologic mode of invasion was classified according to the classification of Anneroth et al. [27].

Immunohistochemical Analysis

Tissues were fixed in 10% buffered formalin and embedded in paraffin. Paraffin blocks were cut at 4 µm thickness. Sections were deparaffinized with xylene and rehydrated in graded alcohol. Three percent hydrogen peroxide was then applied to block endogenous peroxidase activity. The sections were incubated overnight at 4°C with anti-human TGF-β1 (diluted 1:100; Santa Cruz Biotechnology, Santa Cruz, CA), TβR-I (diluted 1:100; Santa Cruz Biotechnology), TβR-II (diluted 1:100; Upstate Biotech, Inc., Lake Placid, NY), Smad2 (diluted 1:100; Chemicon, Inc., Temecula, CA), p-Smad2 (diluted 1:100; Santa Cruz Biotechnology), p21^WAF1 (diluted 1:50; DAKO, Carpinteria, CA), p27^KIP1 (diluted 1:50; DAKO), and Ki67 (diluted 1:100; Abcam, Inc., Cambridge, MA). Immunostaining was performed with the Envision system (DAKO, Carpinteria, CA) in accordance with the manufacturer's instructions. Peroxidase activity was visualized by applying diaminobenzidine chromogen containing 0.05% hydrogen peroxidase. The sections were then counterstained with hematoxylin, dehydrated, cleared, and mounted. Negative control staining was carried out by substituting nonimmune goat or mouse serum for primary antibodies. The immunoreactivity of antibodies to TGF-β1 and its receptors was assessed on a visual analogue scale by semiquantifying cytoplasmic staining. Immunoreactivity was scored as (-) absent, (1+) low (< 25% of positive tumor cells), (2+) moderate (26–75% of positive tumor cells), or (3+) diffuse (> 75% of positive tumor cells) [27,28]. As for Smad2, p-Smad2, p21^WAF1, p27^KIP1, p12^CDK2-AP1, and Ki67, nuclear staining of these antibodies was evaluated by estimating the percentage of positive nuclei. OSCCs were assessed based on at least 10 randomly selected fields. All slides were interpreted by two investigators.

Cell Culture

The human immortalized normal oral epithelium cell lines OKF6 tert-1 and OKF6 tert-2 were gifts from Dr. Jim Rhienwald (Brigham and Women's Hospital, Harvard Medical School, Boston, MA). They were maintained in a serum-free keratinocyte growth medium (Cell Applications, Inc., San Diego, CA) containing glutamine, apo-transferrin, bovine pituitary extract, insulin, human epidermal growth factor, and hydrocortisone. The human malignant oral epithelium cell lines SCC4, SCC9, SCC66, UM1, and 1483 were maintained in DMEM/F12 supplemented with 10% fetal bovine serum and 1% PSF. TGF-β1 was obtained from Sigma (St. Louis, MO). Cells were seeded in six-well plates, TGF-β1 was added when cells reached 50% confluence, and the rate of cell duplication was determined by counting cells for three consecutive days after TGF-β1 treatment.

Protein Preparation and Western Blot Analysis

Subconfluent cells were harvested by trypsinization and lysed with lysis buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 2 mM EDTA, 2 mM EGTA, 25 mM NaF, 25 mM β-glycerophosphate, 0.1 mM sodium orthovanadate, 0.1 mM PMSF, 5 µg/ml leupeptin, 0.2% vol/vol Triton X-100, and 0.5% vol/vol Nonidet P-40). Lysates were incubated for 30 minutes at 4°C and were then centrifuged for 15 minutes at 4°C. Protein concentrations of lysate were determined using the Bio-Rad DC protein assay system (Bio-Rad, Hercules, CA), with bovine serum albumin as standard. Equal amounts of proteins were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (8–15%) and transferred to PVDF membranes. Membranes were incubated overnight with antibodies specific for p12 (pAb86 rabbit polyclonal antibody; Strategic Biosciences, Newark, DE), p27, p21, TβR-I, TβR-II, Smad2, p-Smad2/3 (Santa Cruz Biotechnology), and β-actin (Sigma), washed, and incubated with secondary antibody against mouse or rabbit IgG conjugated with horseradish peroxidase (Amersham Biosciences, Piscataway, NJ) for 60 minutes. After washing, visualization was performed by ECL (Amersham Biosciences) according to the manufacturer's recommendations. Band densities were measured with UN-SCAN-IT software (Silk Scientific, Orem, UT).

Statistical Analysis

The correlation between clinicopathological parameters (i.e., surgical staging) and immunohistochemical analysis results was assessed, as was the correlation between TGF-β1, its receptors, and cell cycle-related proteins, and that between these proteins and the Ki67 labeling index. Comparative data were analyzed using Mann-Whitney U test. For all statistical analyses, P < .05 was considered statistically significant.

Results

Abrogated Expression of TβR-II, p-Smad2, and p12^CDK2-AP1 in Human OSCC In Vivo

TGF-β1 signaling is mediated mainly by two serine threonine kinase receptors, TβR-I and TβR-II. Loss of function or decreased expression of TβR-II has been reported in several tumors; however, the level of TβR-I, TβR-II, and TGF-β1, and their association with OSCC carcinogenesis and tumor progression have not been fully described. In this study, we first examined the expression levels of TGF-β1, TβR-I, and TβR-II in 125 cases of human OSCC by immunohistochemistry. Table 1 shows the characteristics of tissue samples used for the immunohistochemical study. Oral squamous epithelium obtained in surgical procedures not associated with squamous neoplasia has consistently revealed moderate to intense homogenous cytoplasmic expression of TGF-β1 and TGF-β receptors I and II. Homogeneous, moderate, intense cytoplasmic expression of TGF-β1 and TβR-I was also shown in OSCC tissues; however, there was significant weak expression or no expression of TβR-II proteins shown in OSCCs (Figure 1). According to the progression of clinical stage, the expression of TβR-II decreased. The intensity of TGF-β1, TβR-I, and TβR-II expression in different clinical tumor stages is summarized in Table 2. There was no correlation between the expression of TGF-β1, TβR-I, or TβR-II and clinicopathological parameters such as differentiation and mode of invasion (data not shown). These results suggest that, in OSCC, there is a significant progressive loss of expression of TβR-II, which associates with highly aggressive lesions. There is no significant loss of expression of TβR-I in OSCC; TβR-I expression is not directly related to the degree of tumor differentiation or to local aggressiveness.

Table 1.

Characteristics of Tissue Samples Used for Immunohistochemical Analysis.

Age (years)	23–94
Gender
Male	70
Female	55
T classification
T1	29
T2	64
T3	9
T4	23
N classification
N0	83
N1	27
N2	15

Figure 1.

Decreased expression of TβR-II in human OSCC tissues. The levels of TGF-β1 and its associated receptors (TβR-I and TβR-II) in human OSCC tissues were compared to those in normal tissues, as described in the Materials and Methods section. The levels of TGF-β1 and TβR-I did not show any significant alterations in OSCC, but there was significant reduction of TβR-II in OSCCs compared to that in the normal epithelium.

Table 2.

Immunohistochemical Analysis of TGF-β1 and Its Receptors in OSCC Tissues.

Samples	OSCC [n (%)]
	I (n = 28)	II (n = 41)	III (n = 23)	IV (n = 33)
TGF-β1
0 to 1+	2 (7.1)	9 (21.9)	3 (13.0)	7 (21.2)
2+ to 3+	26 (92.9)	32 (78.1)	20 (87.0)	26 (78.8)
TβR-I
0 to 1+	4 (14.3)	10 (24.4)	7 (30.4)	11 (33.3)
2+ to 3+	24 (85.7)	31 (75.6)	16 (69.6)	22 (66.7)
TβR-II
0 to 1+	13 (46.6)	28 (68.3)	18 (78.3)	24 (72.7)
2+ to 3+	15 (53.6)	13 (31.7)	5 (21.7)	9 (27.3)

Smad2 and Smad4 transcription factors are mediators of TGF-β1 signaling [6–8]. Phosphorylation activation of Smad2 is a key step in TGF-β1 signaling for the induction expression of downstream molecules, such as cell cycle inhibitors p21 and p27 [11,29,30]. In our study, Smad2 and p-Smad2 were shown in the nuclei of the normal epithelium (Figure 2). Although Smad2 expression was similarly shown in the nuclei of OSCC, p-Smad2 expression was significantly decreased compared to that in the normal epithelium (Figure 2). Statistical analysis showed that there was a significant correlation between the expressions of TβR-II and p-Smad2 in OSCC (P < .001), whereas there was no correlation between the expressions of TGF-β1, TβR-I, and p-Smad2 in OSCC (Table 3).

Figure 2.

Significant reduction of p-Smad2 in human OSCC tissues. The levels of Smad2 and p-Smad2, known regulators in TGF-β1-mediated growth suppression, were analyzed in human OSCC tissues. Immunohistochemical analysis showed nuclear staining of Smad2 and p-Smad2. There was significant reduction of p-Smad2 in OSCC, but there was no significant alteration in Smad2.

Table 3.

Statistical Analysis of the Correlation of TGF-β1 and Its Receptors with Smad2 and p-Smad2 in OSCC.

Samples	Smad2	P	p-Smad2	P
TGF-βI
0 to 1+ (n = 21)	31.7 ± 3.6	NS	16.6 ± 1.7	NS
2+ to 3+ (n = 104)	33.9 ±		1.6 16.8 ± 2.6
TβR-I
0 to 1+ (n = 32)	30.5 ± 2.9	NS	15.3 ± 2.1	NS
2+ to 3+ (n = 93)	34.6 ± 1.7		17.1 ± 1.2
TβR-II
0 to 1+ (n = 83)	22.3 ± 1.8	NS	12.3 ± 1.1	< .001
2+ to 3+ (n = 42)	36.1 ± 2.5		25.2 ± 1.6

NS, not significant.

Cell cycle-regulatory proteins are downstream molecules of TGF-β-Smad signaling. The growth-inhibitory effect of TGF-β1 is mediated by inducing the expression of important cell cycle-inhibitory proteins such as p21^WAF1, p12^CDK2-AP1, and p27^KIP1 [11,12,29]. Our immunohistochemistry study showed that there was loss of expression of p12^CDK2-AP1 and p21^WAF1 in OSCC (Table 4). The mean values of the labeling indices for p12^CDK2-AP1, p21^WAF1, p27^KIP1, Ki67, Smad2, and p-Smad2 in OSCC are summarized in Table 4. A decrease in TβR-II immunoreactivity had a statistically significant correlation with decreased p12^CDK2-AP1 and p27^KIP1 expression (P < .001 and P < .01, respectively) (Table 5). No correlation between other TGF-β receptors and p21^WAF1 was found. In addition, no correlation between TGF-β1, TβR-I, and cell cycle-related proteins was observed.

Table 4.

Levels of Known Cell Cycle Regulators and Signaling Molecules in the TGF-β1-Mediated Pathway in OSCC.

Samples	Normal	OSCCs
	(n = 28)
		I (n = 28)	II (n = 41)	III (n = 23)	IV (n = 33)
Ki67	6.3 ± 1.9	26.6 ± 1.8	28.9 ± 1.5	34.2 ± 1.9	36.0 ± 1.7
p12CDK2-AP1	31.3 ± 1.9	17.5 ± 1.8	13.7 ± 1.5	13.6 ± 1.9	14.3 ± 1.6
p21WAF1	25.8 ± 2.3	8.0 ± 1.3	8.9 ± 1.0	7.0 ± 1.4	6.2 ± 1.2
p27KIP1	30.1 ± 3.2	24.3 ± 2.3	25.6 ± 1.9	20.0 ± 2.6	23.1 ± 2.1
Smad2	41.2 ± 3.3	40.0 ± 3.1	32.1 ± 2.6	30.3 ± 3.4	32.0 ± 2.8
p-Smad2	32.8 ± 4.1	24.5 ± 2.1	17.0 ± 1.7	13.5 ± 2.3	12.0 ± 1.9

Table 5.

Statistical Analysis of the Correlation of TGF-β1 and Its Receptors with Cell Cycle Inhibitors in OSCC.

Samples	p12CDK-AP1	P	p21CIP/WAF1	P	p27KIP1	P
TGF-β1
0 to 1+ (n = 21)	13.9 ± 2.0	NS	5.7 ± 1.4	NS	23.1 ± 1.2	NS
2+ to 3+ (n = 104)	14.7 ± 0.9		7.8 ± 0.6		26.5 ± 2.7
TβR-I
0 to 1+ (n = 32)	13.4 ± 1.7	NS	8.0 ± 0.7	NS	23.6 ± 1.3	NS
2+ to 3+ (n = 93)	15.0 ± 1.0		5.5 ± 1.1		23.8 ± 2.2
TβR-II
0 to 1+ (n = 83)	10.7 ± 0.8	< .001	8.7 ± 1.0	NS	20.2 ± 1.9	< .01
2+ to 3+ (n = 42)	22.2 ± 1.2		6.7 ± 0.7		25.3 ± 1.3

NS, not significant.

Our immunohistochemistry results showing the loss of expression of TβR-II, p-Smad2, p21^WAF1, and p12^CDK2-AP1 in OSCC provide evidence of aberrant TGF-β-Smad pathway in this human cancer and also indicate the important role of p21^WAF1 and p12^CDK2-AP1, potentially inhibiting tumor formation or progression.

Decreased Expression of TβR-II, p12, Smad2, and p-Smad2/3 in Human OSCC Lines In Vitro

The role of the TGF-β-Smad signaling pathway in the carcinogenesis of head and neck cancer has not been fully evaluated. To examine this signaling pathway in human OSCC, we first examined the basal expression levels of TβR-I, TβR-II, Smad-2, and p-Smad2/3 in human OSCC cell lines by immunoblotting. Compared with normal keratinocytes, an increasing level of TβR-I in OSCC was detected (Figure 3A); however, the expression of TβR-II was decreased in OSCC (Figure 3A). We also detected decreasing levels of Smad2 and p-Smad2/3 inOSCC(Figure 3B). p27^KIP1, p21^WAF1, and p12^CDK2-AP1 are negative cell cycle regulators and growth inhibitors, and p21^WAF1 and p12^CDK2-AP1 are also implicated to be induced transcriptionally by TGF-β1 in epithelial cells [11,12]. To study the correlation between TGF-β-Smad signaling and these growth regulators, we also examined the basal expression levels of p27^KIP1, p21^WAF1, and p12^CDK2-AP1. Compared with normal oral keratinocytes, p12^CDK2-AP1 expression decreased in most OSCC cell lines (Figure 3A); however, decreased expression of p27^KIP1 and p21^WAF1 was not consistently found in OSCC cell lines. These results indicate that a defective TGF-β-Smad signaling pathway exists in human OSCC and provides a possible link between loss of p12^CDK2-AP1 expression and defective TGF-β-Smad signaling in OSCC.

Figure 3.

Basal expression level of TGF-β-Smad signaling proteins in OSCC. (A) Western blot analysis of TβR-I, TβR-II, p12^CDK2-AP1, and p27^KIP1 in the normal oral keratinocytes OKF6 tert1 and OKF6 tert2, as well as in OSCC. Expression of TβR-I increased in OSCC compared with that in normal keratinocytes. Expression of TβR-II and p12^CDK2-AP1 decreased in OSCC. Multiple bands of TβR-II represent different isoforms. (B) Western blot analysis of Smad2 and p-Smad2/3 in OKF6 tert1 and OSCC shows decreased Smad2 and p-Smad2/3 expression in OSCC. Forty micrograms of total proteins was used for Western blot analysis. The level of actin is shown as loading control. Protein levels were quantitated by scanning and analysis with UN-SCAN-IT software. Values indicate total pixel values of a particular protein/actin. The same findings were made from three independent experiments.

Loss of the Growth-Inhibitory Effect of TGF-β1 in Human OSCC

TGF-β1 is a potent growth inhibitor of epithelia. To test the response of human OSCC with a disrupted TGF-β-Smad pathway to TGF-β1, we compared the growth pattern of OSCC with that of normal oral keratinocytes after TGF-β1 treatment. Growth suppression was observed in normal human oral keratinocyte after 24 hours of TGF-β1 exposure (Figure 4A); however, five malignant SCC cell lines keep growing even after 72 hours of TGF-β1 treatment (Figure 4, B–F). Among the five tumor cell lines, only UM1 showed partial growth inhibition by TGF-β1 (Figure 4F). The result confirmed that human OSCC has significant resistance to TGF-β1 growth-inhibitory effect.

Figure 4.

Loss of the growth-inhibitory effect of TGF-β1 on OSCC. Cells were seeded in six-well plates, 1 ng/ml TGF-β1 was added when cells reached 50% confluence, and the rate of cell duplication was determined by counting cells for three consecutive days after TGF-β1 treatment. Values represent cell number, and data are expressed as the mean and SD of triplicate experiments. (A) Complete growth suppression by 24 hours of TGF-β1 exposure in normal keratinocyte OKF6 tert1. (B–E) Loss of growth suppression even by 72 hours of TGF-β1 exposure in OSCC. (F) OSCC UM1 exhibited partial growth inhibition after 48 hours of TGF-β1 exposure.

The key step of TGF-β-Smad signaling is mediated by the phosphorylation activation of p-Smad2/3 and the induction expression of downstream cell cycle inhibitors, such as p21^WAF1 and p12^CDK2-AP1, in epithelial cells. To understand TGF-β1 resistance in OSCC, we further examined the induction levels of p-Smad2/3, p21^WAF1, and p12^CDK2-AP1, as well as the components of the TGF-β-Smad pathway in OSCC compared with normal oral keratinocytes. Increased levels of p-Smad2/3, as well as p21^WAF1 and p12^CDK2-AP1, were found in normal keratinocytes after 24 hours of TGF-β1 exposure; however, among the five OSCC lines, p21^WAF1 could only be induced in UM1, and p12^CDK2-AP1 induction was found only in SCC66 (Figure 5, A–D). Besides, a complete failure of p-Smad2/3 induction was found in all OSCCs examined (Figure 5, A and B). TβR-I, Smad2, and p27^KIP1 levels have no consistent changes among the fiveOSCCcell lines. These results strongly suggest that the failure of induction of p12^CDK2-AP1 or p21^WAF1 expression is an important event in the TGF-β1 resistance observed in human OSCC.

Figure 5.

Induction level of TGF-β-Smad signaling proteins by TGF-β1 in normal oral keratinocytes and OSCCs. Cells exposed to 1 ng/ml TGF-β1 for 24 hours were used for Western blot analysis. Forty micrograms of total proteins was loaded. The level of actin is shown as loading control. Protein levels were quantitated by scanning and analysis with UN-SCAN-IT software. Values indicate the total pixel values of a particular protein/actin. (A) Western blot analysis of TβR-I, Smad2, p-Smad2/3, p12^CDK2-AP1, p21^WAF1, and p27^KIP1 in normal oral keratinocyte OKF6 tert1 and OSCC with or without TGF-β1 exposure. (B) Increased p-Smad2/3 expression in normal oral keratinocyte OKF6 tert1 under TGF-β1, but complete loss of p-Smad2/3 induction was found in all OSCCs. (C) Induction of p12^CDK2-AP1 in OKF6 tert1 under TGF-β1; however, p12^CDK2-AP1 was induced only in SCC66. (D) Induction of p21^WAF1 in OKF6 tert1 under TGF-β1; however, in OSCC, this TGF-β1 effect was found only in UM1. The results were confirmed in three independent experiments.

Discussion

TGF-β receptors mediate the biologic effects of TGF-β1. The direct involvement of TβR-I and TβR-II in TGF-β signal transduction suggests that changes in the expression of these two receptors may lead to loss of TGF-β sensitivity and may ultimately negate the autocrine growth-regulatory mechanism imposed by TGF-β [31]. TGF-β is a potent regulator of proliferation, growth, and differentiation in normal squamous epithelium. TGF-β exerts its antiproliferative effect through TβR-II. Loss of function or decreased expression of TβR-II has been reported in several tumors, and it has been described as having an association with carcinogenesis and tumor progression [9,10,32,33]. Recently, Grady et al. [34] reported that mutations in TβR-II coincide with the transformation of benign adenomas to malignant carcinomas in colon cancers with microsatellite instability. These observations have led to the hypothesis that escape from TGF-β-mediated negative growth control is an important aspect of malignant phenotype in many epithelial neoplasms.

In the present study, we have analyzed the expression of TGF-β1 and the direct involvement of TβR-I and TβR-II in OSCC. Our results indicate that expression of TβR-II inversely correlates with clinical stages and suggest that aberrant TβR-II expression is a contributing factor to OSCC development. To investigate whether anomalies in TβR-II expression are specific for this tumor, we have also analyzed the expression of TβR-I, another component of the TGF-β signal transduction pathway. The normal squamous epithelium had moderate to high levels of both receptors. Furthermore, TβR-I was consistently expressed in all tumors analyzed and in all areas within the tumors. In contrast to TβR-I, the distribution of TβR-II expression was markedly different. In the majority of OSCCs, particularly in infiltrating regions, TβR-II expression was either markedly reduced or completely absent. The mechanisms of resistance to the antiproliferative effects of TGF-β are unknown. However, a decrease in the number of TβR-II-binding sites has been associated with loss of responsiveness to the TGF-β1 cell line in the present study. Similar findings were observed in a number of studies [35–39]. This decrease in TβR-II expression correlates with tumorigenicity in many cell lines. Moreover, an abnormal TGF-β signal transduction pathway could contribute to progressive loss of differentiation in carcinomas.

Previous studies revealed cell type-dependent variations in the effect of TGF-β1 on cell cycle regulators. p27^KIP1 and p21^WAF1, both CDK inhibitors, have also been reported to have an essential role in cell cycle inhibition induced by TGF-β1 [29,40–42]. Recently, we have suggested that p12^CDK2-AP1 may play a role in TGF-β1-mediated growth suppression [12]. This study revealed that TβR-II, p12^CDK2-AP1, p21^WAF1, p27^KIP1, and p-Smad2 were concordantly decreased, supporting the view that TβR-II and cell cycle control proteins may play an important role in the process of cell progression in OSCC. In addition, this result may provide a link between the effects of growth factor and cell cycle control in this tumor. Smad proteins play a critical role in transmitting TGF-β superfamily signals from the cell surface to the nucleus [43]. By forming a heteromeric complex composed of Smad2/3/4, Smad proteins transduce TGF-β or activin signals [30]. In the current study, immunohistochemistry and Western blot analysis showed the correlation between the abrogated expression of TβR-II, p-Smad2, and TβR-II, and cell cycle control proteins such as p27^KIP1, p21^WAF1, and p12^CDK2-AP1. Our data provide evidence of a disrupted TGF-β-Smad signaling pathway; consequently, abrogated expression of cell cycle inhibitors such as p12^CDK2-AP1 may lead to loss of response to TGF-β1 growth-suppressive effects on OSCC.

Abbreviations

TGF-β

transforming growth factor-β

TβR-I

TGF-β receptor I

TβR-II

TGF-β receptor II

p12^CDK2-AP1

p12 CDK2-associating protein 1

OSCC

oral squamous cell carcinoma