Abstract
Aim: The present study aimed to explore the etiological role of human papillomavirus (HPV) in tonsillar squamous cell carcinoma (TSCC). Materials and Methods: HPV status, including viral load and E6 variants, and the expression of P53, p16INK4A, and FANCD2, in tissues of TSCC (n=24) and tonsillitis (n=31) were investigated. Results: The frequency of high-risk HPV (HPV-16) in TSCCs (42%) was higher than that of tonsillitis (16%). HPV-16 genome was partially or fully integrated in all HPV-16-positive TSCCs. However, the viral genome was partially integrated in three out of five HPV-16-positive tonsillitis cases (p=0.037). HPV-16-positive TSCCs showed a higher frequency of p16INK4A expression than HPV-16-negative TSCCs and tonsillitis (p=0.011). Regardless of HPV status, TSCCs had a lower expression of FANCD2 than tonsillitis (p=0.008). Conclusion: The present study supports the etiological role of HPV-16 in the development of TSCC, and p16INK4A overexpression can be applied as a surrogate marker for the detection of high-risk-HPV in TSCC.
The frequency of head and neck squamous cell carcinoma (HNSCC) varies in different countries. It constitutes 3-5% of malignancies in Europe, while it reaches up to 40-50% in South East Asia and India (1). Although the majority of HNSCC cases are considered to be associated with smoking, betel nut chewing, and alcohol drinking, about 15-20% of cases have no history of inveterate tobacco and alcohol use (2). Previous studies have shown an involvement of human papillomavirus (HPV) in the carcinogenesis of a part of HNSCCs, and HPV-16, a high-risk type of HPV, is the most frequent type among all HPV-positive HNSCC cases (3). A recent meta-analysis examined the HPV prevalence among different sub-sites of HNSCC in European populations and reported the highest frequency of 66.4% in tonsillar squamous cell carcinoma (TSCC) among other sub-sites showing a frequency range of 15.3-47.2% (4).
HPV has two major oncoproteins, HPV E6 and E7, which bind to tumor-suppressor proteins, p53 and pRb, respectively, and subsequently lead to degradation of these proteins. Recently, p16INK4a protein expression is known to be up-regulated in cervical carcinomas and other HPV-positive carcinomas (5-7), which is the consequence of E2F release caused by the competing binding of E7 to pRb. On the other hand, there was no difference found in p53 expression between HPV-positive and -negative HNSCCs. Using an animal model, furthermore, Strati and Lambert showed that HPV-16 E7 has a greater potential for carcinogenesis than E6 in HNSCCs (8).
Fanconi anemia (FA) is a heterogeneous and recessive genetic disease characterized by congenital abnormalities, progressive bone marrow failure, and a predisposition to leukemia and solid tumors, and the frequency of HNSCCs in FA patients is significantly higher than that of the general population (9). Kutler et al. (10) reported that over 80% of HNSCCs from FA patients contained a high-risk HPV genome, especially HPV-16. Up to now, 16 cellular genes (fanc genes) have been identified and homozygous mutations in these genes are related to this disease (11). The products of these fanc genes compose FA pathway that is a DNA repair pathway and acts together with homologous recombination repair system to maintain genomic instability (12). FANCD2 is one of the important members of gene products to regulate the G2 checkpoint (13). Recent studies indicated that mono-ubiquitinated FANCD2 plays a critical role to recruit effectors such as the FAN1 nuclease via ubiquitin-binding domains (14-15). Park et al. showed that FA deficiency increased the incidence of HPV-16-E7-driven tumor using fancD2-knockout mice (16). More interestingly, such an event was not observed in HPV-16-E6 transgenic mice (17).
In order to understand the interaction between HPV and tumor suppressor or DNA repair genes in TSCC which shows the highest HPV prevalence among HNSCC sub-sites, we examined the presence, genotype, viral load and physical status of high-risk HPV, and the expression of p16INK4a and FANCD2 proteins in tissue specimens obtained from Japanese patients with TSCCC and tonsillitis.
Materials and Methods
Study subjects. The study subjects in this study were 24 and 31 cases with TSCC and tonsillitis, respectively. All patients were treated at the Kagoshima University Hospital, Kagoshima, Japan during the period from 1991 to 2006. Consecutive patients with tonsillitis were recruited from the pathologic archive of Kagoshima University Hospital in 2006, and patients under age 16 were excluded from the present study. Formalin-fixed and paraffin-embedded tissues of TSCC and tonsillitis and the information on smoking and alcohol drinking were retrospectively retrieved. The ethics committee of Kagoshima University Graduate School of Medical and Dental Sciences approved this study.
DNA extraction. DNA samples were prepared as reported in a previous study (6). In brief, 5-μm thick sections of each tissue were deparaffinized by 0.8 ml of lemosol and 0.2 ml of ethanol. The tissues were digested using buffer (50 mmol/L Tris-Cl, pH 8.0, 1 mmol/L EDTA, pH 8.0, 0.5% Tween 20) containing 200 μg of Proteinase K (Invitrogen Corporation, Carlsbad CA, USA) and incubated at 56°C for 24 h. DNA was extracted from samples using phenol-chloroform followed by ethanol precipitation. The presence of amplifiable DNA was confirmed by β-globin gene amplification.
HPV detection and genotyping. The HPV genome was detected with broad-spectrum SPF1/2 HPV primers PCR method as described previously (6). The DNA template was amplified in the following condition: 15 min at 95°C followed by 40 cycles of 1 min at 94°C, 1 min at 45°C, and 1 min at 72°C, and a final extension of 5 min at 72°C. To determine the HPV types we used the INNO-LiPA HPV genotyping v2 kit (Innogenetics NV, Ghent, Belgium), which is based upon the reverse hybridization principle and can detect a total of 25 HPV genotypes (HPV-6, -11, -16, -18, -31 -33, -35, -39, -40, -42, -43, -44, -45, -51, -52, -53, -54, -56, -58, -59, -66, -68, -70, -73 and -74) using specific oligonucleotide probes that are immobilized on a strip. A part of the L1 gene region of the HPV genome is amplified using biotinylated SPF10 primers, denatured, hybridized with the probes on the strip (18).
HPV-16 viral load and physical status. For all HPV-16 DNA positive samples, quantitative real-time PCR analysis was performed using the ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster city, CA, USA) as reported in detail in our previous study (6). PCR analysis was performed on a 25 μl mixture containing 1×TaqMan Master Mix (Applied Biosystems, Foster city, CA, USA), 300 nM of primers as follows: forward 5’-AAC GAA GTA TCC TCT CCT GAA ATTA TTA G-3’ (3361-3389 nt), reverse 5’-CCA AGG CGA CGG CTT TG-3’ (3427-3443 nt) for amplifying a 76 bp ORF of the E2 gene, and forward 5’-GAG AAC TGC AAT GTT TCA GGA CC-3’ (94-116 nt), reverse 5’-TGT ATA AGT TGT TTG CAG CTC TGT GC-3’ (150-169 nt) for the 81 bp ORF of the E6 gene; 100 nM of specific fluorogenic hybridization probes for E2- (FAM-CAC CCC GCC GCG ACC CAT A-TAMRA) (3406-3424 nt) or E6- (FAM-CAG GAG CGA CCC AGA AAG TTA CCA CAG TT-TAMRA) (119-147 nt); and 1-2 μl of DNA template. The PCR conditions were as follows: incubation at 95°C for 10 min, followed by 40 cycles of denaturation at 95°C for 15 sec and annealing-extension at 60°C for 1 min. DNA extracted from SiHa cells was used as a control for E2 (negative) and E6 (positive) amplification. Since SiHa cells contain only the integrated from of viral genome, E2 gene is disrupted. Adjustment for the differences in the amount of input genomic DNA between samples, was done through calculating the number of genome equivalents or cells in each sample by dividing the amount of β-globin DNA present in the sample [obtained by quantitative real-time PCR for human β-globin gene by 2x QuantiTect SYBR Green PCR kit (QIAGEN, Helden, Germany)] by the weight of one genome equivalent (i.e., 6.6 pg/cell) and a factor of two (since there are 2 copies of beta-globin DNA/genome equivalent or cell). The viral load was expressed as the number of HPV-16 copies per cell (6).
To determine the HPV-16 physical status, the HPV-16 E2/E6 ratio was calculated. A lack of E2 amplification, in which the E2/E6 ratio is zero, represents HPV DNA integration into the host genome. When the E2/E6 ratio is equal to or higher than unity, the HPV-16 genome is considered in the episomal form (19). The mixed presence of integrated and episomal forms of HPV16 results in an E2/E6 ratio between 0 and 1 (19).
HPV-16 E6 variant analysis. HPV-16 E6 variant was analyzed as reported in the previous study (6). In brief, HPV-16 E6 gene was amplified by two semi-nested PCRs and the final products were directly sequenced by fluorescent dye-labeled dideoxynucleotides and cycle sequencing methods using the Big Dye Terminator Cycle Sequencing Kit (Applied Biosystems, Foster city, CA, USA). The HPV-16 E6 sequences were aligned by CLUSTAL W multiple alignments package, and verified with HPV-16 variants published elsewhere (6).
Immunohistochemistry for p16INK4A, p53, and FANCD2. The immunohistochemistry was conducted by avidin-biotin-peroxidase complex system (Vectastain Elite ABC lit; Vector Laboratories, Burlingame, CA, USA) using the mouse monoclonal antibody against p16INK4A (1:300 dilution, G175-405, BD Pharmingen, San Jose, CA, USA), p53 (1:100 dilution, DO-7, Dako, Kyoto, Japan), and FANCD2 (1:100 dilution, FI-17, Santa Cruz Biotech., CA, USA). Details of the procedure for p16INK4A and p53 were reported in the previous study (20). The same protocol was applied for FANCD2 analysis. The expression of these proteins was classified into the following three groups: <10%, 10-49%, and >50%. The cases with >10% carcinoma cells stained positively were considered as positive.
Sex, age and HPV distribution among the patients with tonsillar carcinoma and tonsilitis.
Statistical analysis. The difference in the proportion and mean values among groups were examined by Fisher's exact test and Mann-Whitney U-test, respectively. All p-values presented were two-sided.
Results
There was a significant male predominance among TSCC patients (88%) in comparison to patients with tonsillitis (Table I). Age distribution was also different between them. Although 71% of TSCC patients were equal to or above 60 years old, most of patients with tonsillitis (97%) were under 60.
Ten out of 24 TSCCs (42%) and seven out of 31 tonsillitis cases (23%) harbored HPV genome. The HPV genotype was HPV-16 in all HPV-positive TSCCs and five out of seven HPV-positive tonsillitis lesions. The remaining two cases in tonsillitis harbored HPV-6. The difference in the high-risk HPV frequency between two groups was marginally significant (p=0.065). Among TSCCs, all female patients (n=3) and 33% of males (7/21) were HPV-16 positive, and there was no statistically significant difference in the age distribution between HPV-16-negative and -positive TSCCs (p=0.427).
P53, P16INK4A, and FANCD2 expressions in tonsillitis and tonsillar carcinoma by HPV status.
The viral load in TSCCs was higher than that of tonsillitis but this difference was not statistically significant. There were six HPV-16-positive cases with low viral load (<1 copy per cell) in TSCCs (n=3) and tonsillitis (n=3). When the HPV-16-positive cases were limited to those with the viral load ≥1 copy per cell, the median of viral load was 206 and 1.76 copies per cell in TSCCs and tonsillitis, respectively (p=0.040). The HPV-16 in TSCCs showed a predominance of E-350G-variant (80%), and two out of three HPV-16 in tonsillitis were the prototype. Regarding the physical status of HPV, all HPV-16 genomes detected in TSCCs were integrated forms, at least partially. On the other hand, none of HPV-16 in tonsillitis lesions was fully integrated, and the difference between two groups was statistically significant (p=0.037).
Table II shows the expression pattern of p53, p16INK4A, and FANCD2 among tonsillitis, and HPV-16-negative and -positive TSCCs. HPV-16-positive TSCC cases tended to show a high proportion of p16INK4A expression with ≥50%, though most tonsillitis and HPV-16-negative TSCCs showed negative expression of p16INK4A (<10%). Furthermore, there was a marginally significant difference in p16INK4A expression between HPV-16-positive and -negative TSCCs (p=0.059 by Fisher's exact test). On the other hand, the proportion FANCD2-positive expression was lower than that of tonsillitis regardless of HPV status. There was no difference in the expression pattern of p53 among the three groups.
To find a lead on the regulation of FANCD2 expression, the associations between FANCD2 expression and other factors were examined in the specimens of tonsillitis. Female and never alcohol drinkers tended to show a high proportion of FANCD2 expression, but these differences were not statistically significant (Table III). The expression of FANCD2 was related to neither age nor smoking status. Regarding the patients with TSCC, the information on smoking and alcohol drinking habits could not be retrieved from 63% (15/24) of TSCC cases. There was no never-drinker among the nine TSCC patients with information on alcohol drinking. The proportion of occasional, habitual, and former alcohol drinkers was 11% (n=1), 56% (n=5), and 33% (n=3), respectively.
Discussion
The present study revealed that TSCC harbors a higher HPV frequency (42%) than tonsillitis (23%), especially high-risk type of HPV; HPV-16 (42% and 16%, respectively). The HPV-16 genome was integrated, at least partially, in all TSCCs, while in none of the tonsillitis specimens did it show fully integration of HPV-16. The HPV-16 detected in TSCCs tended to have a higher viral load and a higher frequency of E-350G variant than tonsillitis. There were significant differences in p16INK4A and FANCD2 expressions among tonsillitis, HPV-16-positive and -negative TSCCs (p=0.011 and p=0.008, respectively, Table III).
Overexpression of p16INK4A has been proposed as a surrogate marker for high-risk-HPV-infection in cervical cancers and other HPV-related malignancies, including TSCCs (5, 7, 21, 22) The results of our study also support this hypothesis. More than 70% of tonsillitis and HPV-16 -negative TSCCs were negative for p16INK4A expression and none of tonsillitis and HPV-16-negative TSCCs showed p16INK4A expression in more than 50% of the lesions. On the other hand, three out of seven HPV-16-positive TSCCs showed p16INK4A overexpression.
Several studies have pointed-out the interaction between high-risk HPV and FA pathway in the carcinogenesis of HNSCCs (10, 16-17, 23-26), and FANCD2 plays a crucial role in this DNA repair system. Spardy et al. (23) suggested that HPV-16 E7 induces chromosomal instability in FA cells, and other studies indicated the increased susceptibility to HPV infection under the condition of the FA-pathway deficiency (16-17, 24-25). In the present study, however, the absence of FANCD2 was related to carcinoma rather than HPV status. Because of the small sample size in this study, further investigations are necessary to reveal the interaction between FANCD2 and high-risk HPV status in carcinogenesis of TSCC.
Alcohol drinking is one of the major risk factors of TSCC. In the present study, alcohol drinkers tended to show lower or absent FANCD2 expression, while never alcohol drinkers showed a high proportion of FANCD2 expression among the patients with tonsillitis (Table III). A similar trend was also observed in the patients with TSCC although the information was limited (data not shown). Absence of FANCD2 expression in the nuclei could be caused by epigenetic changes such as DNA methylation (27), and alcohol consumption is possibly involved in the aberrant methylation (28). Further studies are warranted to elucidate this association.
The association between FANCD2 expression by age, sex, smoking, and alcohol drinking in patients with tonsilitis
The HPV prevalence in TSCCs is higher than those of other oral SCCs (4, 21). Some studies suggested that the HPV-related carcinogenesis initially arise in tonsillar crypts (29, 30), which may be the most likely reason why the HPV prevalence in TSCCs is higher than those in other HNSCCs. On the contrary, the HPV-16 frequency is low in tonsillitis and normal tonsillar mucosa although it was relatively high in our study (16%). A review of the studies published until the end of 2002 revealed that the detection rate of HPV genome was 8.5% in 200 specimens from normal mucosa or benign lesions of the tonsil (31). A recent Japanese study also reported 2% (one out of 47) of HPV detection in tonsillitis (21). More recently, Palmer et al. examined the HPV genome in non-malignant tonsil tissue using a large sample size (n=3,377), and found no HPV-positive case among them (32). These observations indicate that certain specific condition(s) in tonsillar mucosa may be required to allow the persistent HPV infection.
In the present study, the viral load in TSCCs ranged from 0.12 to 767 copies per cell, which almost overlapped with those reported in previous studies (21, 33). The viral load at one copy per cell might be the reasonable threshold to distinguish the viral transcription activity (7). When the HPV-16-positive cases were limited to those with viral load ≥1 copy per cell, HPV-16 frequency was 29% and 6% in TSCCs and tonsillitis, respectively, (p=0.032), and the significant associations with physical status (p=0.028) and p16INK4A expression (p=0.004) still remained. Furthermore, HPV-16 E-350G-variant showed a higher viral load (median: 533 copies per cell) than those with the prototype (median: 3.7 copies per cell), and this difference was statistically significant (p=0.025), which was consistent with the results in the previous study (6).
The HPV-16 E-350G variant has a polymorphism at residue 83, leucine to valine (L83V), and this variant, which differs geographically, may affect the pathological risk (34, 35). In cervical cancer, the E-350G variant was reported to increase the persistent HPV infection and risk of cancer progression (34, 36). As observed in this study, this E-350G variant was the most frequent variant as in other studies of HNSCC and TSCC (37, 38).
HPV integration in specific region(s) of host genome has been considered as a key event of carcinogenesis of the cervix (39). The HPV-16 genome was integrated, at least partially, into the host genome in all HPV-16-positive TSCCs but not in tonsillitis. Deng et al. (21) also reported a predominance of mixed or integrated forms in TSCC (75%). These observations suggest that HPV integration is the critical step in the development of HPV-related TSCCs.
As compared to previous studies, HPV frequency in our study was relatively low, 42%. A recent meta-analysis in European patients reported 66.4% HPV frequency in TSCCs (4), and a Japanese study also found a similar HPV prevalence, 69.6% in TSCCs (21). Although the exact reasons of the low HPV frequency in this study are unclear, it could be partly explained by the sensitivity of HPV-detection method and the quality of tissue specimens since the Japanese study examined HPV genome in frozen tissue specimens.
In conclusion, the present study supported the etiological role of HPV-16 in the development of TSCC. Therefore, cervical carcinoma, p16INK4A overexpression can be applied as a surrogate marker for the infection of high-risk-HPV in TSCC. On the other hand, the interaction between FANCD2 and high-risk HPV status should be further clarified.
Conclusion
The present study supports the etiological role of HPV-16 in the development of TSCC, and that p16INK4A overexpression can be applied as a surrogate marker for the detection of high-risk HPV in TSCC.
- Received August 13, 2015.
- Revision received September 13, 2015.
- Accepted September 14, 2015.
- Copyright© 2015 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.