Abstract
Little is known about the potential involvement of the oncoprotein gankyrin in human oral cancer progression. In this study, the levels of gankyrin mRNA and protein expression were assessed in human oral epithelial cell lines, at-risk normal oral tissues, premalignant oral lesions, and primary oral squamous cell carcinomas (OSCCs). Materials and Methods: Biopsies included 6 oral epithelial cell lines, 32 OSCC specimens for qRT-PCR analysis, 27 OSCC specimens and 12 premalignant oral lesions for immunohistochemical analysis. Results: Gankyrin was overexpressed in all tested oral epithelial cell lines and the majority of OSCC specimens (32/32 (100%) and 21/27 (71%) at the mRNA and protein levels, respectively). Moreover, 6/12 of premalignant oral lesions overexpressed gankyrin protein. Conclusion: Gankyrin overexpression is a prevalent event in human oral cancer and occurs during the early stages of oral carcinogenesis, thus being a viable therapeutic or chemopreventive target in oral cancer.
Carcinogenesis is a multistep process through which normal cells are sequentially transformed via the activation of proto-oncogenes and inactivation of tumor suppressor genes into their malignant counterparts. During human oral cancer development, one of the most prominent genetic changes is alteration of the p16CDKN2A tumor suppressor gene. More than 80% of human oral squamous cell carcinomas (OSCCs) (1, 2) and 50% of severe oral dysplasias (3) have been shown to exhibit CDKN2A alterations. Inactivation of the p53 tumor suppressor gene (TP53) is another frequent alteration observed in OSCC development and affects nearly 50% of all cases (4). In contrast to the strikingly prevalent incidence of alterations in these important tumor suppressor genes, knowledge related to the activation of specific proto-oncogenes in human oral cancer remains relatively limited.
Recently, overexpression of a newly defined oncoprotein, gankyrin (PSMD10), has been found to be associated with human cancers through diverse mechanisms (5-9). Gankyrin facilitates phosphorylation and degradation of RB1 protein (pRb) in vitro and in vivo (10, 11), and activates the E2F family of transcription factors. In parallel, gankyrin competes with p16 for binding with cyclin-dependent kinase 4 (CDK4), which precludes the inhibition of p16 in CDK4-mediated phosphorylation of pRb, thus leading to cell cycle progression (8). Furthermore, gankyrin also functions in the degradation of the tumor suppressor protein p53 (12). By binding the ubiquitin ligase, murine double minute 2 (MDM2), gankyrin promotes p53 ubiquitylation and subsequent proteasomal degradation (13). Hence, gankyrin functions as a dual-negative regulator of the two most prominent tumor suppressor pathways: (i) INK4-CDK-pRb, and (ii) ARF-MDM2-p53. Moreover, gankyrin overexpression has been found to be a prevailing event in human hepatocellular carcinomas (HCCs) (6) and esophageal SCCs (ESCCs) (9), and emerging evidence shows that it could be one of the earliest molecular events occurring during epithelial cancer development. In a chemically induced rodent hepatocarcinogenesis model (6, 14), hypermethylation of CDKN2A and mutation of TP53 appear during the later stages (HCC), whereas gankyrin is overexpressed much earlier, shortly after initial carcinogen exposure (liver fibrosis), preceding the loss of pRb (cirrhosis) and hepatocellular adenoma formation. In human HCCs, Tan et al. reported that the frequencies of gankyrin overexpression in Edmondson's grade I/II, III, and IV HCCs were 82%, 63%, and 22%, respectively (15). Consistently, 81% of HCCs at the low TNM stage and 35% of HCCs at the high TNM stage overexpressed gankyrin (16). In addition, a recent study on ESCCs (9) showed that gankyrin overexpression was positively correlated with lower survival rate, extent of the primary tumor, lymph node metastasis, distant lymph node metastasis and stage. These findings further indicate that gankyrin may be important during the development of the malignant potential in ESCC, and may play an important role in its progression (9). Thus, it would appear that gankyrin is one of few oncogenic proteins which negatively modulate both pRb and p53 pathways (11), and its aberrant overexpression could be an early initiating event in carcinogenesis. Since most oral carcinomas have inactivated pRb and p53 pathways either directly or indirectly (1-4), it is of interest to investigate the involvement of gankyrin during oral cancer development.
Here we report the first investigation into the incidence of altered gankyrin expression in oral cancer. Using quantitative real-time RT-PCR (qRT-PCR) and immunohistochemistry, we evaluated the expression of gankyrin in human oral lesions.
Materials and Methods
Cell culture and siRNA transfection. Seven oral cell lines were used, including a normal epithelial cell line (TE1177), one premalignant epithelial cell line (SCC-83-01-82), and five malignant cell lines (SCC-83-01-82CA, SCC9, SCC4, CAL27, and UM-SCC-22A) (17-20). SCC9, SCC4, and CAL27 were purchased from the American Type Culture Collection (Manassas, VA, USA); UM-SCC-22A was obtained from the University of Michigan; SCC-83-01-82, SCC-83-01-82CA, and TE1177 were kindly provided by Dr. S.M. D'Ambrosio of The Ohio State University (19, 20). All cells were cultured in Advanced DMEM/F12 medium (Invitrogen, Carlsbad, CA, USA) containing 5% fetal bovine serum (FBS; Invitrogen) in a 90% relative humidity incubator at 37°C supplied with 5% CO2.
Synthetic siRNA targeting human gankyrin (26S proteasome non-ATPase regulatory subunit 10, PSMD10) (sc-72186; Santa Cruz Biotechnology, CA, USA) and control siRNA (sc-37007, Santa Cruz Biotechnology) were transfected into TE1177, SCC-83-01-82, and SCC-83-01-82CA cells using Oligofectamine (Invitrogen) in accordance with the manufacturer's protocol. Briefly, 6×106 cells (in a 10-cm diameter dish, 75% confluency) were transfected with siRNAs at a final concentration of 10 μM to knockdown more than 80% of PSMD10 mRNA expression (data not shown). Upon transfection for 48 hours, the transfected cells were analyzed by flow cytometry to assess potential changes in the cell cycle distribution (propidium iodide staining) and apoptosis (Annexin-V-FLUOS staining) as previously described (21).
Patient sample procurement. All oral cancer and normal tissues were obtained from the Head and Neck Tumor Bank (J.C.L.) or Biorepository and Biospecimen Shared Resource at The Ohio State University Comprehensive Cancer Center following Institutional Review Board approved protocols. Two biospecimen sample sets were used for the retrospective studies: (i) Archival tissue blocks for 27 OSCC cases and 12 preneoplastic oral lesions, (ii) Total cellular RNA for 29 primary OSCCs, 2 recurrent OSCCs, 1 verrucous carcinoma, and 10 histologically normal oral tissues obtained from OSCC patients at the time of surgery (Table I).
Quantitative real-time RT-PCR. Total RNA was isolated from cells and tissues using TRIzol Reagent (Invitrogen) according to the manufacturer's instructions. First-strand cDNA synthesis was performed using PowerScript Reverse Transcriptase (BD Sciences, San Jose, CA, USA) as previously described (22). Real-time RT-PCR reactions were carried out using a SmartCycler system (Cepheid, Sunnyvale, CA, USA), and custom TaqMan assays designed using Primer Express 3.0 (Applied Biosystems, Carlsbad, CA, USA). Each reaction mixture included 200 μM dNTPs, 3 mM MgSO4, 1.25 units of Taq polymerase (Agilent Technologies, Santa Clara, CA, USA), 1× Additive (0.2 mg/ml bovine serum albumin (BSA), 150 mM Trehalose, and 0.2% Tween 20), 0.1 volume of first-strand cDNA, 0.2 μM each primer, and 0.1 μM fluorogenic probe. The samples were amplified at 96°C for 2 minutes, followed by 55 cycles of 30 seconds at 95°C, 30 seconds at 60°C, and 30 seconds at 68°C. Human hypoxanthine phosphoribosyltransferase (HPRT1) was used for gene expression normalization. The primers and fluorogenic probes were as follows: PSMD10, 5’-TCCTCTTCATATTGCGGCTT-3’ (forward), 5’-CTTGAGCACCTTTTCCCAGAA-3’ (reverse), and 5’-TET-TGG CCGGGATGAGATTGTAAAA GCCTAMRA-3’ (probe); CDKN2A, 5’-TGCCCAACGCACCGA-3’ (forward), 5’-GGGCGC TGCCCATCA-3’ (reverse), and 5’-TETTAGTTACGGTCGG AGGCCG ATCCA-TAMRA-3’ (probe); for HPRT1, 5’-CGGCT CCGTTATGGCG-3’ (forward), 5’-GGTCA TAACCTGGTTCAT CATCAC-3’ (reverse), and 5’-FAMCGCAGCCCTGGCGTCGTGA-TAMRA-3’ (probe). Each gene was amplified separately, and all experiments were performed in triplicate.
The expression levels of PSMD10 and CDKN2A were normalized to that of HPRT1, as well as appropriate calibrators, and the relative gene expression level was determined using a comparative Ct method (23):
ΔΔCt=(CtPSMD10−CtHPRT1)tumor−(CtPSMD10−CtHPRT1)normal, or ΔΔCt=(CtCDKN2A−CtHPRT1)tumor−(CtCDKN2A−CtHPRT1)normal Relative gene expression=2−ΔΔCt. A two-fold increase (≥2) or decrease (≤0.5) was considered significant for mRNA overexpression or down-regulation analysis, respectively (23).
Western blot. Total cellular proteins were extracted from oral normal and OSCC cells using the M-PER Mammalian Protein Extraction Reagent (ThermoScientific/Pierce, Rockford, IL, USA) and Western blotted as previously described (24). The primary and secondary antibodies were rabbit anti-human gankyrin antibody (PW8325; Enzo Life Sciences/BIOMOL, Plymouth Meeting, PA, USA) and peroxidase-conjugated rabbit anti-human IgG (ab6759; Abcam, Cambridge, MA, USA), respectively. Immunoblots were visualized by enhanced chemiluminescence using the SuperSignal West Pico kit (ThermoScientific/Pierce). Protein levels of β-actin were used for normalization.
Gankyrin immunohistochemical (IHC) staining and analysis. IHC detection of gankyrin was conducted on paraffin-embedded sections using standard avidin-biotin peroxidase procedure with the Vectastain Elite ABC kit (Vector Laboratories, Burlingame, CA, USA) and an automated IHC System (BioGenex, San Ramon, CA, USA) as previously described (25). The percentage positive labeling index was calculated for each oral lesion stained utilizing Image-Pro Plus (Media Cybernetics, Bethesda, MD, USA) analysis software. Histologically normal oral epithelial sections were utilized as controls.
Statistical analyses. All data in this study were statistically analyzed using R 2.9.2 software package. A two-sided t-test or χ2 test at α=0.05 is regarded as significant.
Results
PSMD10/Gankyrin overexpression in OSCC cells. We first evaluated the expression of PSMD10/gankyrin in a ‘normal’ oral epithelium cell line (TE1177), a premalignant squamous epithelium cell line (SCC-83-01-82), and five malignant OSCC cell lines (SCC-83-01-82CA, SCC9, SCC4, CAL27, and UM-SCC-22A) (17-20) using qRT-PCR and immunoblotting. As shown in Figure 1A, the premalignant cells and all five OSCC cell lines exhibited significantly higher levels (2-8 fold) of PSMD10 mRNA expression relative to TE1177 expression. Furthermore, results from Western blots (Figure 1B) demonstrated that the levels of gankyrin protein in these neoplastic oral cell lines were higher than that present in TE1177 cells. Therefore, gankyrin is overexpressed at both mRNA and protein levels in premalignant oral epithelial cells and malignant OSCC cell lines.
PSMD10/Gankyrin overexpression in human OSCC specimens. We then determined the expression of PSMD10 mRNA in a retrospective cohort of 29 primary OSCCs, two recurrence OSCCs, and a verrucous carcinoma compared to the average expression in 10 representative histologically normal oral tissue specimens. As shown in Figure 2A, PSMD10 mRNA was up-regulated in all 32 oral malignancies, with transcriptional overexpression ranging from 0.3- to 90-fold. While it remains unknown what biological level of overexpression generates a significant physiological event for PSMD10, using a more rigid criterion for mRNA overexpression (≥2-fold change) (23) demonstrated that 84% of these OSCCs (27/32) overexpressed PSMD10 mRNA. It is worth noting that none of these clinical specimens demonstrated a reduction in PSMD10 expression. Since down-regulation of CDKN2A is one of the most prevalent molecular alterations found in oral cancer (26), we also assessed the expression of CDKN2A mRNA in these clinical samples (Figure 2B). Consistent with our previous studies (25) and the current knowledge base, CDKN2A mRNA was poorly expressed in 65.6% of the specimens (21/32), expressed at normal levels in 25.0% of the samples (8/32), and overexpressed in 9.4% of the specimens (3/32). These findings support existing studies and demonstrate that down-regulation of CDKN2A expression remains one of the major events during oral carcinogenesis (26). However, no significant correlation between PSMD10 overexpression and CDKN2A down-regulation (p=0.35) was demonstrated in these clinical samples, suggesting that these two transcriptional events may be independent biomarkers of oral malignancy in this patient cohort.
We further examined the presence of gankyrin protein using IHC in a distinct retrospective OSCC patient cohort (n=27) for which only archival paraffin tissue blocks were available (24). Figure 3 shows representative IHC results for gankyrin protein expression in normal oral and OSCC tissues. Mild staining is apparent in the normal tissue (Figure 3A and 3B), which is consistent with one of the basal activity roles of gankyrin as a transient subunit of the proteasome complex (11). In contrast, the majority of OSCCs (21/27, 71%) exhibited clear overexpression of gankyrin protein (Figure 3C-H). The difference in incidence of overexpression of PSMD10 mRNA and gankyrin protein (84% vs. 71%) is not statistically significant (p=0.75), providing support for a possible simple mechanism linking transcription and translation. These results demonstrate that overexpression of gankyrin protein occurs frequently during oral carcinogenesis and may be the result of deregulation at the transcriptional level (i.e. overexpression of PSMD10 mRNA) resulting in elevated gankyrin protein levels.
Gankyrin overexpression in preneoplastic oral lesions. Since gankyrin overexpression has been reported to occur during the earliest stages of hepatocellular carcinogenesis, prior to prominent cancer-promoting events such as pRb degradation and hypermethylation of CDKN2A and TP53 (6, 14) we postulated that activation of proto-oncogenic gankyrin could be an early event in the development of oral cancer. To address the temporal alteration of gankyrin expression during oral malignant neogenesis, we examined the presence of gankyrin protein in oral tissues at different stages of cancer development using IHC. Figure 4 shows gankyrin protein expression in representative normal oral tissue (Figure 4A and 4B), mild/low-grade dysplasia (Figure 4C), severe/high-grade dysplasia (Figure 4D and 4E), and proliferative verrucous leukoplakia (PVL) (Figure 4F). Mild staining was apparent in the normal tissue, which is consistent with the constitutive function of gankyrin as a component of the 26S proteasome complex. In contrast, gankyrin protein was intensely expressed in the premalignant mild/low-grade dysplasia (Figure 4C), severe/high-grade dysplasia (Figure 4D and 4E), and PVL lesion (Figure 4F). In this premalignant oral tissue group we demonstrate gankyrin protein overexpression with respect to normal oral epithelium in 6/12 of the cases. These results provide support for a role of gankyrin protein overexpression early during oral carcinogenesis, and establish a foundation for examining gankyrin as one of the earliest events in oral epithelial cell deregulation.
Gankyrin promotes the proliferation of premalignant and malignant oral cancer cells in vitro. To evaluate the effect of gankyrin expression in oral cancer cells, we transfected three oral cell lines, TE1177 (normal), SCC-83-01-82 (premalignant), SCC-83-01-82CA (malignant) with siRNA targeting human gankyrin (sc-72186; Santa Cruz Biotechnology) and analyzed changes in cell cycle distribution using flow cytometry. The results are listed in Table II. As mentioned earlier (Figure 1), gankyrin was overexpressed at both mRNA and protein levels in SCC-83-01-82 and SCC-83-01-82CA cells (in comparison with TE1177). In the absence of any treatment, about 72.0% of TE1177 cells were in the G0/G1 phase, compared to 51.2% of SCC-83-01-82 and 49.8% of SCC-83-01-82CA cells. While gankyrin-targeting siRNA treatment did not bring about any significant changes in the cell cycle distribution of TE1177 cells, siRNA-mediated down-regulation of gankyrin expression in premalignant SCC-83-01-82 cells led to a significant increase of cells in the G0/G1 phase (51.2% vs. 58.7%, p<0.05). Such increase in the G0/G1 population was even greater in siRNA-treated malignant SCC-83-01-82CA cells (49.8% vs. 62.6%, p<0.01). In contrast, control siRNA did not bring about significant changes in the cell cycle distribution of any of these oral cell lines. Additionally, no significant changes in apoptosis were observed in any of these cell lines upon transfection of gankyrin-targeting siRNA (data not shown). Hence, down-regulation of gankyrin expression in premaligant and malignant oral cancer cells primarily prohibited cell proliferation. This finding is consistent with previous studies showing that gankyrin enhances the growth and proliferation, not apoptosis, of pancreatic and lung cancer cells (21, 27). Taken together, these results suggest that aberrant gankyrin expression contributes to human cancer progression through promoting cell proliferation.
Discussion
The development of oral cancer proceeds through well-defined sequential stages from normal epithelium, through atypic and hypeplastic phenotypes, mild, moderate, and severe dysplasias, and culminates in OSCC with possible metastatic disease (3, 25). While there are established genetic landmarks during oral cancer progression, the earliest initiating events remain less well characterized and continue to be of particular interest in cancer prognostic, diagnostic, and prevention strategies. Our current studies provide significant foundational evidence for an early role of oncogenic gankyrin in oral cancer development, and provide support for examining the potential of gankyrin as an early biomarker of oral cancer progression.
Aberrant gankyrin expression at the transcriptional and translational levels represents a prevalent molecular event during oral carcinogenesis. Similar to the observations of gankyrin overexpression in human and rat HCC and human ESCCs, we observed significant overexpression of PSMD10 mRNA in 6/6 of OSCC cell lines and 27/32 (84%) of malignant oral lesions. The striking incidence of PSMD10 mRNA overexpression is recapitulated by our studies on gankyrin protein expression in OSCC cell lines and tumor tissues. Mimicking their PSMD10 overexpression, all six OSCC cell lines overexpressed gankyrin protein. Similarly, 71% of OSCC tissues had intense staining of gankyrin protein compared to patient-matched normal oral tissues (25). Furthermore, activation of gankyrin (overexpression) occurs during the early stages of oral cancer development. In the established rat model of chemically induced HCC, distinct molecular events occur sequentially during the multi-step process of hepatocellular carcinogenesis. These initiating and promoting actions include gankyrin overexpression, pRb degradation, and hypermethylation of CDKN2A and TP53 (6, 14). In our current studies, gankyrin also maintains a prominent early role in the multistep process of oral carcinogenesis, with 50% of oral preneoplastic lesions demonstrating a higher level of gankyrin protein expression compared to their normal tissue counterparts.
While alterations in CDKN2A mRNA and p16 protein represent early biomarkers of oral cancer development and are components of a common cell cycle regulatory pathway, our current studies do not demonstrate a direct association between gankyrin overexpression and the functional status of the CDKN2A gene in this OSCC patient cohort. While both up-regulation of PSMD10/gankyrin and down-regulation of CDKN2A occur in these retrospective cohorts, it is possible that these events represent independent biomarkers of oral carcinogenic progression. The genetic status of CDKN2A varies among the gankyrin overexpressing-OSCC cell lines. UM-SCC-22A cells display a homozygous CDKN2A deletion, while SCC9 cells have a point mutation in the CDKN2A promoter, and CAL27 cells lack functional p16 due to unknown mechanisms (20). Conversely, SCC4, SCC83, and SCC83CA cells have an intact CDKN2A gene (17-20). Since both overexpression of gankyrin and loss of p16 expression contribute to control of cell cycle progression through altering the CDK4-mediated phosphorylation of pRb protein (8), it is of interest whether it is necessary to have an additional mechanism of pRb inactivation (gankyrin up-regulation) in a pathway that already exhibits a prevalent inactivating event (loss of functional p16). Support for the value of gankyrin overexpression as an important biomarker during oral carcinogenesis comes from the following observations: (i) Gankyrin overexpression occurs during the earliest stages of oral cancer progression, possibly in advance of the known biomarker p16 deregulation; (ii) Previous studies have shown that approximately 40% of OSCCs contain intact CDKN2A or maintain functional p16 despite the presence of missense mutations (26). In these cases, functional p16 would remain subject to gankyrin-mediated cell cycle interference, and initiated premalignant cells could continue through the cell cycle (8); (iii) Gankyrin functions through distinct mechanisms and regulatory pathways, including the cell cycle via CDK4 interference, MDM2-mediated protein degradation of pRb and p53 via the proteasome complex, and p53-mediated apoptotic suppression (5-9). Early alterations in PSMD10/gankyrin can generate diverse multifactorial consequences independently capable of modulating cell growth and carcinogenic progression.
In conclusion, our results demonstrate for the first time that gankyrin is frequently overexpressed in premalignant and malignant oral tissues, further suggesting that activation of gankyrin is an early event during oral carcinogenesis. The sensitivity and specificity of gankyrin as a putative oral cancer biomarker will need to be evaluated in larger populations and prospective studies. The specific environmental or genotoxic exposures responsible for activation of gankyrin in oral cancer are unknown; however, it is known that tobacco smoke is a significant risk factor for both OSCC and HCC. Given the striking incidence of gankyrin overexpression in HCCs and the common risk factor of smoking, it is possible that tobacco carcinogens could activate the proto-oncogenic PSMD10 gene in oral epithelial cells in the field of cancerization, and provide the metabolic advantage necessary to initiate and promote deregulated cell growth. These findings provide novel insights into early molecular markers of oral cancer development and emphasize the need for additional studies on the potential of gankyrin as a biomarker, therapeutic, or chemopreventive target.
Acknowledgements
This work was supported by research grants from NIH: R01 CA69472 (J.L.), R01 DE011943 (C.M.W.), R21 DE016361 (C.M.W.)
Footnotes
-
↵* Both Authors contributed to this work equally.
-
This article is freely accessible online.
- Received March 12, 2011.
- Revision received June 9, 2011.
- Accepted June 10, 2011.
- Copyright© 2011 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved