To understand the molecular mechanism of RB1 phosphorylation in basal–parabasal layers of normal cervix and during cervical cancer (CACX) development, we analyzed the alterations (expression/methylation/deletion/mutation) of RB1/phosphorylated RB1 (p-RB1) (ser807/811 and ser567) and two RB1 phosphorylation inhibitors, P16 and RBSP3, in disease-free normal cervical epithelium (n = 9), adjacent normal cervical epithelium of tumors (n = 70), cervical intraepithelial neoplasia (CIN; n = 28), CACX (n = 102) samples and two CACX cell lines. Immunohistochemical analysis revealed high/medium expression of RB1/p-RB1 (ser807/811 and ser567) and low expression of P16 and RBSP3 in proliferating basal–parabasal layers of majority of normal cervical epitheliums, irrespective of HPV16 infection. Interestingly, 35–52% samples showed high/medium expression of P16 in basal–parabasal layers of normal and had significant association with deleterious non-synonimous SNPs of P16. Methylation of P16 and RBSP3 in basal–parabasal layers of normal cervix (32 and 62%, respectively) showed concordance with their respective expressions in basal–parabasal layers. The methylation frequency of P16 and RBSP3 in basal–parabasal layers of normal did not change significantly in CIN and CACX. The deletion frequency of P16 and RB1 increased significantly with CACX progression. While, deletion of RBSP3 was high in CIN and comparable during CACX progression. P16 showed scattered and infrequent mutation in CACX. The alteration of P16 and RBSP3 was synergistic and showed association with overexpression of p-RB1 in tumors and associated with poor prognosis of patients. Thus, our data suggest that overexpression of p-RB1 in basal–parabasal layers of normal cervical epithelium was due to methylation/low functional-linked non-synonimous SNPs of P16 and RBSP3. This pattern was maintained during cervical carcinogenesis by additional deletion/mutation.

Introduction

Worldwide, cervical cancer (CACX) is the third most threatening cancer among women [1]. In India, it stands as the second most frequent female cancer as per incidence rate [2]. Persistent infection by high-risk (hr) human papillomavirus (HPV) 16, 18, 33, etc is the principal cause of CACX along with other etiological factors [3]. The hr-HPV infection in the squamocolumnar junction of normal cervix blocks differentiation and induces transformation of basal stem cells of cervical epithelium to acquire genetic and epigenetic alterations for the development of CACX [4]. Apart from hr-HPV oncoprotein E7-mediated RB1 degradation during CACX development, inactivation of RB1 also occurs through overexpression (60–80%) of phosphorylated RB1 (p-RB1), deletion (32–47%), and promoter methylation (40–61%) in this tumor [58]. Phosphorylation of RB1 occurs initially by cyclinD1–CDK4 complex at ser807/811 residues in early G1 phase, followed by sequential phosphorylation in the late GI-S phase of cell cycle by cyclinE–CDK2 complex at ser567, ser608, etc. [9]. The overexpression of p-RB1 might be due to the inactivation of P16 (inhibitor of CDK4/6) and RBSP3 (dephosphorylate RB1 at ser807/811) by deletion/methylation/mutation as seen in premalignant cervical lesions (cervical intraepithelial neoplasia, CIN) and in subsequent stages of CACX [58,1014]. Apart from the above mechanisms, overexpression of p-RB1 may also occur by functional inactivation of P16, due to non-synonimous SNPs in exon2 as reported in melanoma and other tumors [15,16].

P16 showed variable frequencies of promoter methylation in CIN (23–57%) and CACX (31–93%) followed by 17–22% deletion and 9–15% mutation in CACX [1014]. Similarly, in concordance, several studies revealed low mRNA expression of P16 in CIN (26%) and CACX (75–78%) followed by its reduced protein expression (56–79%) in CACX [8,1719]. However, some reports of overexpression of P16 mRNA (74–81%) and protein (42–94%) in CACX have been seen [2022]. The prevalence of functionally inactivated non-synonimous SNPs of P16 associated with development of CACX is not known. RBSP3 showed frequent deletion in CIN (48%) and CACX (42–45%) followed by ∼26% promoter methylation both in CIN and CACX [5,23]. Highly reduced (64–82%) mRNA expression of RBSP3 was also seen in CACX [5,23]. However, detailed analyses of mutation and protein expression of RBSP3 in CACX have not been done. To the best of our knowledge, the molecular profiles of P16, RBSP3, and RB1 have not been analyzed in same set of cervical lesions to understand the mechanism of p-RB1 overexpression.

On the other hand, like CIN/CACX, overexpression of p-RB1 has also been seen in the proliferating basal–parabasal layer of normal cervical epithelium, which gradually decreased in the differentiated spinous layer [5]. Thus, the overexpression of p-RB1 is needed in both basal–parabasal layers and also during CACX development. However, the mechanism of persistent p-RB1 overexpression in the basal–parabasal layers and CIN/CACX may not be similar. In normal cervical epithelium, low expression of P16 protein and 11% P16 promoter methylation have been reported, while no such data were reported for RBSP3 [11,20].

Thus to understand the molecular mechanism of p-RB1 overexpression in basal–parabasal layers of the normal cervical epithelium and during CACX development, at first, we analyzed protein expression pattern of RB1, p-RB1 (ser807/811 and ser567), P16, RBSP3 by immunohistochemistry (IHC)/immunocytochemistry (ICC) in normal cervical epithelium, cervical lesions at different clinical stages, and CACX cell lines. Then the promoter methylation status of P16 and RBSP3 were analyzed in basal–parabasal and spinous layers of normal cervical epithelium, cervical lesions, and CACX cell lines. Finally, deletion/sequence variation of P16, RBSP3, and RB1 were done in normal cervical epitheliums/cervical lesions, followed by correlation of the alterations of these genes with different clinicopathological parameters.

Our data revealed that overexpression of p-RB1 in the basal–parabasal layers of normal cervical epithelium was due to frequent inactivation of P16 and RBSP3 by their promoter methylation or presence of low functional non-synonimous SNPs of P16. This profile of P16 and RBSP3 was maintained during development of CACX with additional deletion/mutation.

Materials and methods

Collection of clinical specimens

  • Freshly operated/biopsy samples of 28 CIN (11 low-grade CIN I and 17 high-grade CIN II/III) and 102 primary CACX samples (50 stage I/II and 52 stage III/IV tumors) along with 5 ml of corresponding blood and adjacent normal tissues (n = 70) were collected from the hospital section of Chittaranjan National Cancer Institute (CNCI), Kolkata, India (Supplementary Figure S1). In addition, normal cervical tissues (disease free) (n = 9) were collected from patients who underwent hysterectomy due to other gynecological reasons (Supplementary Figure S1). Appropriate approval of the Institutional Ethical Committee and informed consent from the patients were taken. A part of the freshly operated samples were paraffin embedded after formalin fixation for immunohistochemical analyses. Other part of the samples and respective blood were frozen immediately after collection at −80°C until use. The tumors were graded and staged according to International Federation of Gynecology and Obstetrics classification (Supplementary Table S1).

  • The CACX cell lines HeLa and SiHa were purchased from National Centre for Cell Sciences, Pune, India and grown in accordance with the supplier's instructions.

Microdissection and DNA extraction

  • The contaminant normal cells in the cervical lesions were removed by microdissection from cryosections (5 µm) using surgical knives under a dissecting microscope (Leica MZ16, Germany). The representative sections from different regions of the specimens were stained with hematoxylin and eosin (H&E) for diagnosis as well as for marking of the dysplastic epithelium/tumor-rich regions. The samples containing >60–80% dysplastic epithelium/tumor cells were taken for isolation of high-molecular weight DNA by proteinase-K digestion followed by phenol–chloroform extraction according to standard procedure [24,25]. Similarly, DNA from normal cervical tissues (disease free), histopathologically adjacent normal cervical epithelium and blood were also isolated [24,25].

  • In order to investigate promoter methylation pattern of P16 and RBSP3 in different layers of histopathologically normal cervical epithelium, microdissection of the basal–parabasal and spinous layers were done in paraffin/cryosections of HPV16-negative normal cervical epithelium (disease free; n = 9), HPV16-positive (n = 23) and HPV16-negative (n = 15) histopathologically normal cervical epithelium adjacent to CACX by laser capture microdissection (LCM) microscope (Zeiss Palm/Apotome, Germany). DNA from the microdissected samples were isolated by standard procedure (Supplementary Figure S1). [24,26].

Detection of HPV-16 and HPV-18

HPV infection was detected by polymerase chain reaction (PCR) using primers (MY09 and MY11) from the consensus L1 region followed by typing of HPV16/18 using type-specific primers in the L1-positive samples. The HPV16/18-positive samples and CACX cell lines (HeLa and SiHa) were further confirmed by southern blot hybridization [27].

Expression analysis of RB1, p-RB1, P16, and RBSP3 by IHC/ICC

In order to study the protein expression pattern of P16, RBSP3, RB1, and p-RB1, immunohistochemical analyses were done in HPV16-negative normal cervical epithelium (disease free; n = 9), HPV16-positive (n = 23) and HPV16-negative (n = 47) histopathologically normal cervical epithelium adjacent to tumors, CIN (n = 15), and primary CACX (n = 55) samples according to the standard procedure as described by Ghosh et al. [28]. The tissue sections (paraffin embedded/cryosections) were reacted overnight with primary antibodies P16 (sc-1661), RB1 (sc-7905), p-RB1 ser807/811 (sc-16670-R), p-RB1 ser567 (sc-32824) from Santa Cruz Biotechnology (CA, USA) and RBSP3 (CP-57-09) custom made from M/s IMGENEX Biotech, Bhubaneswar, India, at a dilution 1:80–1:100 at 4°C. HRP-conjugated goat anti-rabbit (sc-2004) for P16, RBSP3, RB1, and p-RB1 ser807/811, and rabbit anti-goat IgG (sc-2768) for p-RB1 ser567 were added in 1:500 dilutions [29]. For permanent staining of the primary tissues, the slides were developed using 3-3′-diaminobenzidine as the chromogen and counterstained with hematoxylin. The staining intensity (1 = weak, 2 = moderate, 3 = strong) and the percentage of positive cells (<1 = 0, 1–20 = 1, 20–50 = 2, 50–80 = 3 and >80 = 4) were detected by two independent observers and by combining the two scores, final evaluation of expression was done (0–2 = low, 3–5 = intermediate, 6–7 = high) [30].

For ICC analysis, cover slip culture of HeLa and SiHa cell lines were reacted with the same dilution of primary antibody of these genes after permeabilization with 0.5% Triton X-100 and blocking with 3–5% BSA. After washing, the coverslips were incubated with FITC-conjugated corresponding secondary antibody goat anti-mouse (sc-2010), goat anti-rabbit (sc-2012) and rabbit anti-goat (sc-2777) at 1:500 dilution and mounted with glycerol after thorough washing. Imaging of the cover slip was performed in florescence microscope (Leica DM4000 B, Germany).

Promoter methylation analysis of P16 and RBSP3

P16 and RBSP3 promoter methylation status were analyzed in basal–parabasal and spinous layers of HPV16-negative normal cervical epithelium (disease free; n = 9), HPV16-positive (n = 23) and HPV16-negative (n = 15) histopathologically normal cervical epithelium adjacent to CACX, CIN (n = 28), primary CACX (n = 102) samples, and two CACX cell lines HeLa and SiHa by PCR-based methylation-sensitive restriction analysis (MSRA; Supplementary Figure S1 and Table S2) [31]. Approximately, 100 ng of DNA samples were individually digested overnight with methylation-sensitive restriction enzymes HpaII (CCGG; Promega, USA) and HhaI (GCGC; Sibenzyme, Russia) separately. The 445 bp fragment of β-3A adaptin gene (K1) and 229 bp fragment of RARβ2 (K2) were used as digestion and integrity controls, respectively [32]. Mock digestion was done with each sample without any restriction enzyme. PCR products were analyzed on 2% agarose gels, visualized under UV illumination, and photographed.

To validate the methylation data of RBSP3 obtained by MSRA, methylation-specific PCR (MSP) in 28 paired cervical lesions and in HeLa and SiHa cell lines were done after bisulfite modification of the DNA by standard procedure [33].

5-aza-2′-deoxycytidine treatment of HeLa and SiHa

HeLa and SiHa cell lines were grown in the absence and presence of different doses (5, 10, and 20 µM) of demethylating agent 5-aza-2′-deoxycytidine (5-aza-dC) for 5 days. RNA was prepared using TRIzol reagent, and real-time quantification of P16 and RBSP3 expression was performed for mRNA expression as described earlier [34,35]. β2-Microglobulin was used as a control for equal loading and RNA integrity. For immunoflorescence analysis after 5-aza-dC treatment, HeLa and SiHa cells were grown over night on cover slip and treated with 20 µm 5-aza-dC for 72 h. Then the cells were fixed with chilled methanol and used for immunoflorescence analysis.

Deletion analysis of P16, RBSP3, and RB1

Deletion analysis of P16, RBSP3, and RB1 were performed using six microsatellite and exonic markers (Ensembl release 83; Genome Database) in CIN (n = 28) and CACX (n = 102) samples by a standard PCR using a [γ-p32] ATP-labeled forward primer as described by Ghosh et al. (Supplementary Figure S1 and Table S2) [36]. PCR products were electrophoresed on 7% denaturing polyacrylamide sequencing gel and autoradiographed [36].

Loss of heterozygosity (LOH) was detected by densitometric scanning (Bio-Rad GS-800), and the scoring of LOH and microsatellite size alteration (MA) was done on autoradiogram as described previously [36]. MA was detected as a shift in the mobility of 1 (MA1) or both (MA2) alleles compared with their normal alleles. MA in 1 allele and loss in other allele was regarded as LOH + MA (LMA).

Sequence variation analysis of P16 and RBSP3

Mutation screening and SNP analysis were done in P16 exon1 and exon2 due to earlier reports of mutational hotspots/inactivating SNPs, by Sanger sequencing using 3130xl-Genetic Analyzer (Applied Biosystems, USA) in disease-free normal samples (n = 9) and paired CACX lesions (n = 70) (Supplementary Figure S1 and Table S2). Mutation screening was also done in RBSP3 exon6 and exon7 due to presence of RB1 phosphatase domain, by single-strand conformation polymorphism (SSCP) analysis in paired CACX lesions (n = 40) as described by Tripathi et al. [10]. Random samples (n = 10) were sequenced for validation.

Statistical analysis

Fisher's exact test was used to determine different clinicopathological association with tumors genetic alterations. All statistical tests were two-sided and considered significant at probability value, P < 0.05. Survival curves were obtained according to Kaplan–Meier method. Overall survival (OS) was measured from the date of surgery to the date of most recent follow-up or death (up to 5 years). Multivariate Cox proportional hazard regression model was used to test the statistical significance of potential prognostic factors. From this model, we estimated the hazard ratio for each potential prognostic factor with a 95% confidence interval. The detailed follow-up records were available for 58 CACX patients (Supplementary Figure S1). All the statistical analyses were performed using statistical programs Epi Info 6.04, SPSS 10.0 (SPSS, Chicago, IL).

Results

Analysis of HPV prevalence in normal cervical epithelium and primary cervical lesions

In normal cervical epithelium (disease free), HPV infection was not detected (Supplementary Table S3). In adjacent normal cervical epithelium (n = 70) of cervical lesions, HPV DNA was detected in 34% (24/70) samples (Supplementary Table S3). Out of the HPV-positive samples, 96% (23/24) samples were HPV16 positive and 4% (1/24) sample was HPVs other than HPV16/18 type (Supplementary Table S3).

In primary cervical lesions, HPV DNA was detected in 95.7% (124/130) samples with 89.2% (25/28) in CIN and 97% (99/102) in CACX (Supplementary Table S4). Among the HPV-positive samples in CIN, 88% (22/25) samples were HPV16 positive, no HPV18-positive sample, and 12% (3/25) samples were HPVs other than HPV16/18 type (Supplementary Table S4a). Likewise, among the HPV-positive samples in CACX, 93% (92/99) samples were HPV16 positive, 3% (3/99) samples were HPV18 positive, 1% (1/99) samples had mixed infection of both HPV16/18, and 3% (3/99) samples were HPVs other than HPV16/18 type (Supplementary Table S4b). No significant association was seen between HPV infection, tumor stage, grade, nodal status, parity, and sexual debut (Supplementary Table S1).

Immunohistochemical analysis of RB1, p-RB1 (ser807/811 and ser567), P16, and RBSP3 in normal cervical epithelium and primary cervical lesions

In normal cervical epithelium, expression of RB1/p-RB1 (ser807/811 and ser567) gradually decreased from proliferating basal–parabasal layers to differentiated spinous layer in majority of the samples (Supplementary Figure S2 and Table 1). In disease-free normal (HPV−ve), high/medium nuclear expression of RB1/p-RB1 in basal/parabasal layers was seen in 100% samples followed by 33–56% samples in the spinous layer (Table 1). Similar pattern of RB1/p-RB1 expression was seen in disease-associated adjacent normal epithelium irrespective of HPV16 infection (Table 1). Interestingly, high/medium nuclear expression of p-RB1 (ser807/811) and/or p-RB1 (ser567) in basal–parabasal layers of normal epithelium was seen in 78–100% samples (Table 1). Like high/medium expression pattern of RB1/p-RB1 in basal–parabasal layers of normal epithelium, similar expression pattern was seen in CIN (80–93%) and CACX (74–93%) (Supplementary Figure S2, Table 1 and Supplementary Table S5). Thus, it indicates that majority of the cells in basal–parabasal layers of normal epithelium, CIN, and CACX are in at least G1-S phase of the cell cycle.

Table 1
Comparative expression pattern (%) of RB1, p-RB1 (ser807/811), p-RB1 (ser567), P16 and RBSP3 in different normal cervical epithelium, CIN and CACX.
 Disease-free normal Adjacent normal CIN CACX 
HPV−ve (n = 9) HPV16−ve (n = 47) HPV16+ve (n = 23) (n = 15) (n = 55) 
Basal-parabasal Spinous Basal-parabasal Spinous Basal-parabasal Spinous   
High/medium Low High/medium Low High/medium Low High/medium Low High/medium Low High/medium Low High/medium Low High/medium Low 
RB1 100 55 45 72 28 35 65 74 26 39 61 80 20 74 26 
p-RB1 (ser 807/811) 100 33 67 89 11 40 60 74 26 39 61 93 11 83 17 
p-RB1 (ser 567) 100 56 44 90 10 43 57 100 40 60 93 11 93 
p-RB1 (ser 807/811)/p-RB1 (ser567) 100 89 78 96 13 45 60 100 22 52 65 93 91 18 
P16 44 56 89 11 52 48 91 35 65 83 17 53 47 44 56 
RBSP3 44 56 89 11 29 71 85 15 26 74 78 22 27 73 27 73 
 Disease-free normal Adjacent normal CIN CACX 
HPV−ve (n = 9) HPV16−ve (n = 47) HPV16+ve (n = 23) (n = 15) (n = 55) 
Basal-parabasal Spinous Basal-parabasal Spinous Basal-parabasal Spinous   
High/medium Low High/medium Low High/medium Low High/medium Low High/medium Low High/medium Low High/medium Low High/medium Low 
RB1 100 55 45 72 28 35 65 74 26 39 61 80 20 74 26 
p-RB1 (ser 807/811) 100 33 67 89 11 40 60 74 26 39 61 93 11 83 17 
p-RB1 (ser 567) 100 56 44 90 10 43 57 100 40 60 93 11 93 
p-RB1 (ser 807/811)/p-RB1 (ser567) 100 89 78 96 13 45 60 100 22 52 65 93 91 18 
P16 44 56 89 11 52 48 91 35 65 83 17 53 47 44 56 
RBSP3 44 56 89 11 29 71 85 15 26 74 78 22 27 73 27 73 

On the other hand, in normal cervical epithelium, expression of P16 gradually increased from basal–parabasal layers to spinous layer in majority of the samples (Table 1 and Supplementary Figure S3a). Low nuclear/cytoplasmic expression of P16 in basal–parabasal layers of normal epithelium was seen in 48–65% samples irrespective of HPV16 infection (Table 1 and Supplementary Table S3). Similar expression pattern of P16 was seen in CIN (47%) and CACX (56%) (Supplementary Figure S3b, Table 1 and Supplementary Table S5). Thus, considerable frequency of high/medium expression of P16 was evident in basal–parabasal layers of different types of normal epithelium (35–52%), CIN (53%), and CACX (44%) (Table 1).

Like P16, similar expression pattern of RBSP3 was evident in normal cervical epithelium (Table 1 and Supplementary Figure S3a). Low nuclear/cytoplasm expression of RBSP3 in basal–parabasal layers of normal epithelium was seen in 56–74% samples irrespective of HPV16 infection (Table 1). Similar expression pattern (73%) of RBSP3 was seen in CIN and CACX (Supplementary Figure S3b, Table 1 and Supplementary Table S5).

In basal–parabasal layers of normal epithelium, CIN and CACX low expression of P16 and high/medium expression of p-RB1 (ser807/811) was evident in 33–51% samples (Table 2). However, considerable frequency of high/medium expression of both P16 and p-RB1 (807/811) was also evident in basal–parabasal layers of normal (39%), CIN (60%), and CACX (33%) (Table 2).

Table 2
Relation between expression of P16 and RBSP3 with expression of p-RB1 (ser807/811) in basal-parabasal layers of normal cervical epithelium, CIN and CACX
p-RB1 (ser807/811) p-RB1 (ser807/811) p-RB1 (ser807/811) 
Normal (basal-parabasal) (n = 79) H/M CIN (n = 15) H/M CACX (n = 55) H/M 
P16 H/M 39% 8% P16 H/M 60% P16 H/M 33% 9% 
47% 6% 33% 7% 51% 7% 
 P-value  0.82  P-value  0.83  P-value  0.58 
RBSP3 H/M 25% 4% RBSP3 H/M 27% RBSP3 H/M 23% 4% 
61% 10% 66% 7% 60% 13% 
 P-value  0.83  P-value  0.58  P-value  0.97 
p-RB1 (ser807/811) p-RB1 (ser807/811) p-RB1 (ser807/811) 
Normal (basal-parabasal) (n = 79) H/M CIN (n = 15) H/M CACX (n = 55) H/M 
P16 H/M 39% 8% P16 H/M 60% P16 H/M 33% 9% 
47% 6% 33% 7% 51% 7% 
 P-value  0.82  P-value  0.83  P-value  0.58 
RBSP3 H/M 25% 4% RBSP3 H/M 27% RBSP3 H/M 23% 4% 
61% 10% 66% 7% 60% 13% 
 P-value  0.83  P-value  0.58  P-value  0.97 

On the other hand, low RBSP3 and high/medium expression of p-RB1 (ser807/811) was observed in 60–66% samples (Table 2). While high/medium expression of both RBSP3 and p-RB1 (ser807/811) was evident in 24–27% samples (Table 2).

Promoter methylation pattern of P16 and RBSP3 in normal cervical epithelium and cervical lesions

To understand the mechanism of P16 and RBSP3 expression in basal–parabasal layers of normal cervical epithelium, their promoter methylation pattern was analyzed in normal epithelium (n = 79) at first. Promoter methylation of these genes was evident in 12–14% samples (Figure 1A,B). However, in the microdissected basal–parabasal layers of normal epithelium (n = 47), high promoter methylation of P16 was seen in basal/parabasal layers (32%) of normal epithelium compared with the spinous layer (8%) (Figure 1A). The increased methylation frequency of basal/parabasal layers might be due to the sensitivity of the LCM which enriches sparsely populated basal–parabasal cells for detecting minor frequencies of methylation. Like basal–parabasal layers of normal epithelium, comparable methylation frequency of P16 was seen in CIN (36%), stage I/II tumors (36%), followed by slight increase in stage III/IV tumors (46%) (Figure 1A). The methylation of P16 in basal–parabasal layers of normal epithelium, CIN, and CACX showed concordance with its reduced expression (Supplementary Table S6i). However, no concordance was seen between methylation and expression of P16 in spinous layer of normal epithelium (Supplementary Table S6i).

Promoter methylation analysis of P16 and RBSP3.

Figure 1.
Promoter methylation analysis of P16 and RBSP3.

(A) (i) Schematic representation of the promoter region of P16 illustrates distribution of HpaII (CCGG, marked by arrow) and HhaI (GCGC, marked by star) restriction sites. The positions of primers designed for MSRA are shown by black arrowheads. (ii) Representative tumor samples (#6583) showing methylated status by MSRA. Corresponding normal cervical tissue was unmethylated. (iii) Methylation frequencies observed in different normal cervical epitheliums, CIN and CACX stage I/II and III/IV samples. (B) (i) Schematic representation of the promoter region of RBSP3. (ii) Representative tumor sample (#6261) showing methylated status by MSRA. Corresponding normal cervical tissue was unmethylated. (iii) Methylation frequencies observed in different normal cervical epitheliums, CIN and CACX stage I/II and III/IV samples. K1 and K2: controls for DNA digestion and integrity, respectively.

Figure 1.
Promoter methylation analysis of P16 and RBSP3.

(A) (i) Schematic representation of the promoter region of P16 illustrates distribution of HpaII (CCGG, marked by arrow) and HhaI (GCGC, marked by star) restriction sites. The positions of primers designed for MSRA are shown by black arrowheads. (ii) Representative tumor samples (#6583) showing methylated status by MSRA. Corresponding normal cervical tissue was unmethylated. (iii) Methylation frequencies observed in different normal cervical epitheliums, CIN and CACX stage I/II and III/IV samples. (B) (i) Schematic representation of the promoter region of RBSP3. (ii) Representative tumor sample (#6261) showing methylated status by MSRA. Corresponding normal cervical tissue was unmethylated. (iii) Methylation frequencies observed in different normal cervical epitheliums, CIN and CACX stage I/II and III/IV samples. K1 and K2: controls for DNA digestion and integrity, respectively.

Like P16, RBSP3 showed high methylation in the basal–parabasal layers (62%) compared with spinous layer (25%) (Figure 1B). Similar to basal–parabasal layers of normal epithelium, comparable methylation pattern of RBSP3 was seen in CIN (50%), stage I/II tumors (42%), and in stage III/IV tumors (51%) (Figure 1B). Concordance existed between two methylation analysis techniques, MSRA and MSP (Supplementary Table S7). The methylation of RBSP3 in basal–parabasal layers of normal epithelium, CIN, and CACX showed concordance with its reduced expression (Supplementary Table S6ii). However, no concordance was seen between methylation and expression of RBSP3 in spinous layer of normal epithelium (Supplementary Table S6ii). P16 was found to be unmethylated both in HeLa and SiHa cell lines. While RBSP3 was methylated only in SiHa cell line.

Treatment with 5-aza-dC increases expression of P16 and RBSP3 in HeLa and SiHa cells

To confirm inactivation of P16 and RBSP3 by promoter hypermethylation, expression of the genes was analyzed by quantitative RT-PCR after treatment of HeLa and SiHa cells with different doses (5, 10, 20 µM) of 5-aza-dC. Dose-dependent reactivation of P16 and RBSP3 expression was observed in the cell lines with respect to untreated controls. At the high 20 µM 5-aza-dC, gradual up-regulation in the expression of P16 and RBSP3 were seen (Figure 2A). Thus, our data validate promoter methylation as one of the inactivating mechanisms of P16 and RBSP3 in CACX development.

Demethylation experiments and ICC in HeLa and SiHa cell lines.

Figure 2.
Demethylation experiments and ICC in HeLa and SiHa cell lines.

(A) The mRNA expression of (i) P16 and (ii) RBSP3 were analyzed in HeLa and SiHa cell lines after treatment with different concentrations (5 µm, 10 µm, 20 µm) of 5-aza-dC by QRT-PCR. Fold change of mRNA expression was compared with the mRNA of untreated cell lines. P16 and RBSP3 mRNA expression were up-regulated in a dose-dependent manner. ICC analysis of P16, RBSP3, p-RB1 (ser807/811 and ser567), and RB1 in (B) HeLa and (C) SiHa: P16 and RBSP3 showed up-regulation of cytoplasmic/nuclear expression after 5-aza-dC treatment (ii) compared with untreated cells (i). On the contrary, p-RB1 (ser807/811 and ser567) showed reduced expression in treated cells (ii) compared with untreated cells (i). Magnification: ×40, Scale bars: 50 µm.

Figure 2.
Demethylation experiments and ICC in HeLa and SiHa cell lines.

(A) The mRNA expression of (i) P16 and (ii) RBSP3 were analyzed in HeLa and SiHa cell lines after treatment with different concentrations (5 µm, 10 µm, 20 µm) of 5-aza-dC by QRT-PCR. Fold change of mRNA expression was compared with the mRNA of untreated cell lines. P16 and RBSP3 mRNA expression were up-regulated in a dose-dependent manner. ICC analysis of P16, RBSP3, p-RB1 (ser807/811 and ser567), and RB1 in (B) HeLa and (C) SiHa: P16 and RBSP3 showed up-regulation of cytoplasmic/nuclear expression after 5-aza-dC treatment (ii) compared with untreated cells (i). On the contrary, p-RB1 (ser807/811 and ser567) showed reduced expression in treated cells (ii) compared with untreated cells (i). Magnification: ×40, Scale bars: 50 µm.

In ICC analysis, increased nuclear/cytoplasmic expression of P16 and RBSP3 proteins was seen in HeLa and SiHa cells after treatment with 20 mm 5-aza-dC for 72 h, whereas expression of p-RB1 (ser807/811 and ser567) was considerably reduced (Figure 2B,C).

Deletion analysis of P16, RBSP3, and RB1

In microsatellite-based deletion analysis of P16, RBSP3, and RB1, frequent deletion and infrequent MA (0–2%) was seen (Figure 3). Deletion frequency of P16 was seen in 20% CIN samples followed by significant increase from stage I/II tumors (28%) to stage III/IV tumors (43%; P = 0.04) (Figure 3A and Supplementary Table S4). The deletion of P16 in CIN and CACX showed concordance with its reduced expression (Supplementary Table S8i). The frequency of biallelic microsatellite alterations (LMA + MAII) of P16 was seen in only 0% CIN and 5% CACX samples (Supplementary Table S4). On the other hand, both deletion and methylation of P16 was seen in 11% CIN and 19% CACX samples (Supplementary Table S9i).

Representative autoradiograph showing deletion and MA of P16, RBSP3, and RB1 in different cervical lesions.

Figure 3.
Representative autoradiograph showing deletion and MA of P16, RBSP3, and RB1 in different cervical lesions.

(A) (i) P16. LOH, MA and LMA in different samples (#4241, #6949, #1419). (ii) Deletion frequency of P16 observed in CIN and CACX stage I/II and III/IV samples. (B) (i) RBSP3. LOH, MA, HD and HED in different samples (#6559, #3068, #4440, #4477) (ii) Deletion frequency and (iii) Overall alteration frequency of P16 and RBSP3 observed in CIN and CACX stage I/II and III/IV samples. (C) (i) RB1 LOH in #6906 (ii) Deletion frequency of RB1 observed in CIN and CACX stage I/II and III/IV samples. LOH, loss of heterozygosity; MA-I, microsatellite size alteration of 1 allele; MA-II, microsatellite size alteration of both alleles; LMA, LOH + LMAI; HD, homozygous deletion; HED, hemizygous deletion; T: DNA of the dysplastic/tumor cells after microdissection; N: DNA of the corresponding normal tissue.

Figure 3.
Representative autoradiograph showing deletion and MA of P16, RBSP3, and RB1 in different cervical lesions.

(A) (i) P16. LOH, MA and LMA in different samples (#4241, #6949, #1419). (ii) Deletion frequency of P16 observed in CIN and CACX stage I/II and III/IV samples. (B) (i) RBSP3. LOH, MA, HD and HED in different samples (#6559, #3068, #4440, #4477) (ii) Deletion frequency and (iii) Overall alteration frequency of P16 and RBSP3 observed in CIN and CACX stage I/II and III/IV samples. (C) (i) RB1 LOH in #6906 (ii) Deletion frequency of RB1 observed in CIN and CACX stage I/II and III/IV samples. LOH, loss of heterozygosity; MA-I, microsatellite size alteration of 1 allele; MA-II, microsatellite size alteration of both alleles; LMA, LOH + LMAI; HD, homozygous deletion; HED, hemizygous deletion; T: DNA of the dysplastic/tumor cells after microdissection; N: DNA of the corresponding normal tissue.

The deletion of RBSP3 was high in CIN (46%) and comparable in stage I/II tumors (48%), followed by slight increase in stage III/IV tumors (54%) (Figure 3B). The deletion of RBSP3 in CIN and CACX showed concordance with its reduced expression (Supplementary Table S8ii). The frequency of biallelic microsatellite alterations of RBSP3 was seen in only 0% CIN and 2% CACX samples (Supplementary Table S4). On the other hand, both deletion and methylation of RBSP3 was seen in 21% CIN and 33% CACX samples (Supplementary Table S9ii).

The deletion of RB1 was low in CIN (5%), followed by significant increase in stage I/II tumors (31%) (P = 0.002) and comparable in stage III/IV tumors (33%) (Figure 3C). Both deletion and reduced expression of RB1 was seen in 7% CIN and 13% CACX samples (Supplementary Table S8iii).

Sequence variation analysis of P16 and RBSP3

Mutation analysis of P16 and RBSP3

In mutation analysis of P16 exon1 and exon2, both transition (C → T) and transversion (G → T, G → C, T → A, C → G, C → A) have been seen, resulting in missense and silent mutation in 4% (3/70) and 7% (5/70) samples, respectively (Figure 4A,B). The mutation of both exon1 and exon2 was seen in only one sample (#5886; Figure 4B). These mutations were distributed between codon 35–49 and codon 57–135 in exon1 and exon2, respectively (Figure 3B). The overall mutation frequency for P16 was seen in 10% (7/70) samples (Figure 4B and Supplementary Table S5). The mutated samples showed methylation and/or deletion in majority of the cases (6/7) with low (4/7) or medium (3/7) protein expression (Figure 4B).

Mutation analysis of P16.
Figure 4.
Mutation analysis of P16.

(A) Representative chromatograms showing mutations in P16 exon1 and exon2; (B) compilation of P16 mutations exon1 and exon2, their coding effect in CACX with respect to their expression, methylation, and deletion. M+/-, methylation positive/negative; D+/-, deletion positive/negative

Figure 4.
Mutation analysis of P16.

(A) Representative chromatograms showing mutations in P16 exon1 and exon2; (B) compilation of P16 mutations exon1 and exon2, their coding effect in CACX with respect to their expression, methylation, and deletion. M+/-, methylation positive/negative; D+/-, deletion positive/negative

In mutation analysis of RBSP3 exon6 and exon7, no band shifts have been observed in SSCP analysis of the samples (0/40). In addition, random sequencing of those samples (0/10) revealed no base changes. Therefore, mutation in RBSP3 may be a rare phenomenon in CACX.

Association of non-synonimous SNPs of P16 with its protein expression in normal cervical epithelium and cervical lesions

To understand the mechanism of high/medium expression (35–52%) of P16 in basal–parabasal layers of the normal cervical epitheliums and corresponding cervical lesions, we have genotyped the non-synonimous SNPs of P16 exon1 and exon2 in the disease-free normal epithelium (n = 9), in the adjacent normal samples (n = 70) and corresponding cervical lesions and compared them with their respective protein expression in basal–parabasal layers of the normal epitheliums and corresponding cervical lesions.

Samples with high/medium P16 protein expression in basal–parabasal layers of normal cervical epithelium and corresponding cervical lesions showed significant association with minor alleles of 55 non-synonimous SNPs of P16 exon1 and exon2 (Ensembl release 83; Genome Database, Figure 5A and Supplementary Table S10). Among these SNPs, 31 were found to be deleterious/possibly damaging in terms of protein function due to change in amino acid as per SIFT and POLYPHEN score, and haplotype analysis revealed linkage between them (Ensembl Release 83, Figure 5B and Supplementary Table S10) [37]. Interestingly, among these deleterious non-synonimous SNPs, rs11552823 (P81L), rs104894094 (G101R), rs121913386 (P114L), and rs104894098 (V126D) of P16 exon2 (reportedly associated with melanoma) [15] showed high prevalence in the normal samples and cervical lesions (72–75%) and had concordance with high/medium P16 expression in basal–parabasal layers of normal cervical epithelium and cervical lesions (Figure 5C). The P16 protein was stable due to the above amino acid changes [15]. The remaining 27 deleterious SNPs also showed similar pattern, and change in the amino acids leads to increased stability of P16 protein as per SVM analysis (data not shown) [38]. Thus, it indicates that functionally deleterious/damaging effect of these linked non-synonimous SNPs might increase the half-life of P16 protein.

SNP analysis of P16.
Figure 5.
SNP analysis of P16.

(A) Non-synonimous SNPs of P16 exon1 and exon2 and their relation with high/medium expression of P16 in basal–parabasal layers of normal cervical epithelium and cervical lesions. The shaded SNPs are deleterious/possibly damaging as per SIFT/POLYPHEN score. (B) linkage disequilibrium (LD) plot of 31 non-synonimous deleterious SNPs of P16 exon1 and exon2. The numbers indicate each deleterious SNPs and score in the boxes indicate LD score. (C) Relationship between four deleterious SNPs of P16 (exon2) with protein expression in basal/parabasal layers of normal cervical epithelium and in cervical lesions.

Figure 5.
SNP analysis of P16.

(A) Non-synonimous SNPs of P16 exon1 and exon2 and their relation with high/medium expression of P16 in basal–parabasal layers of normal cervical epithelium and cervical lesions. The shaded SNPs are deleterious/possibly damaging as per SIFT/POLYPHEN score. (B) linkage disequilibrium (LD) plot of 31 non-synonimous deleterious SNPs of P16 exon1 and exon2. The numbers indicate each deleterious SNPs and score in the boxes indicate LD score. (C) Relationship between four deleterious SNPs of P16 (exon2) with protein expression in basal/parabasal layers of normal cervical epithelium and in cervical lesions.

Clinicopathological association with overall alterations of P16 and RBSP3

P16 showed 32% methylation in basal–parabasal layers of normal epithelium followed by gradual increase in its alterations (methylation/deletion/mutation) with progression from CIN (43%) to stage I/II tumors (48%) and stage III/IV tumors (64%) (Figures 1Aiii and 3Biii). The alterations of P16 were concordant with its reduced expression in CIN and CACX (Supplementary Table S11i). While alterations of P16 showed concordance with high/medium expression of p-RB1 (ser807/811) and p-RB1 (ser567) in CIN and CACX samples (Supplementary Table S12i,ii).

RBSP3 showed 62% methylation in basal–parabasal layers of normal epithelium followed by comparable alterations (methylation/deletion/mutation) with progression from CIN (67%) to stage I/II tumors (58%) and stage III/IV tumors (69%) (Figures 1Biii and 3Biii). The alteration of RBSP3 was concordant with its reduced expression in CIN and CACX (Supplementary Table S11ii). While alterations of RBSP3 showed concordance with high/medium expression of p-RB1 (ser807/811) in CIN and CACX samples, but no so such association was evident with expression of p-RB1 (ser567) (Supplementary Table S12iii,iv).

Significant association was seen with alterations of P16 and RBSP3 in CACX only, suggesting synergistic effect of the alterations of these genes in the development of invasive lesions of CACX (Supplementary Table S13).

Significantly poor OS was evident for CACX patients with deletion/methylation/alterations of P16 and/or RBSP3 (Figure 6A), indicating their prognostic importance. Multivariate Cox regression model showed relative risks of several prognostic factors (clinical stage, lymph node involvement, parity, age at sexual debut, and HPV infection) on OS of the patients (Figure 6B).

Clinicopathological association with overall alterations of P16 and RBSP3.

Figure 6.
Clinicopathological association with overall alterations of P16 and RBSP3.

(A) Kaplan–Meier 5-year survival probability curves with cumulative survival of CACX patients by (i–iii) P16 methylation, deletion, and overall alteration, (iv–vi) RBSP3 methylation, deletion, and overall alteration, and (vii) P16 and/or RBSP3 overall alteration. Survival time was defined as the time from the date of surgery to the date of last follow-up, known recurrence or death (up to 5 years). The smooth line represents survival probability without molecular alterations and the dotted line represents the same probability with molecular alterations. n, total number of CACX samples. (B) Multivariate analysis of OS of cervical cancer patients with different clinicopathological parameters and P16 and RBSP3 alterations using Cox proportional hazard model.

Figure 6.
Clinicopathological association with overall alterations of P16 and RBSP3.

(A) Kaplan–Meier 5-year survival probability curves with cumulative survival of CACX patients by (i–iii) P16 methylation, deletion, and overall alteration, (iv–vi) RBSP3 methylation, deletion, and overall alteration, and (vii) P16 and/or RBSP3 overall alteration. Survival time was defined as the time from the date of surgery to the date of last follow-up, known recurrence or death (up to 5 years). The smooth line represents survival probability without molecular alterations and the dotted line represents the same probability with molecular alterations. n, total number of CACX samples. (B) Multivariate analysis of OS of cervical cancer patients with different clinicopathological parameters and P16 and RBSP3 alterations using Cox proportional hazard model.

Discussion

The aim of the present study was to understand the molecular mechanisms behind p-RB1 (ser807/811 and ser567) overexpression in basal–parabasal layers of normal cervical epithelium and how the pattern is maintained during CACX progression.

IHC revealed high/medium expression of RB1 (72–100%) and p-RB1 (ser807/811 and ser567; 74–100%) irrespective of HPV16 infection in proliferating basal–parabasal layers of normal cervical epithelium, while frequent low expression of P16 (48–65%) and RBSP3 (56–74%) was seen in the samples (Table 1). However, low RB1/p-RB1 expression was seen in the spinous layer of normal epitheliums along with high/medium expression of P16 and RBSP3 in majority of the samples (Table 1). This suggests an inverse correlation of the p-RB1 (ser807/811 and ser567) expressions in normal epithelium with P16 and RBSP3 expressions. Similar expression pattern of RBSP3 and RB1/p-RB1 was reported earlier in head and neck tissue and normal cervix [5,29]. The reduced expression of P16 and RBSP3 in the basal–parabasal layers of normal cervical epithelium was due to high promoter methylation (32% and 62%, respectively) (Figure 1a,b). However, high/medium expression of P16 and RBSP3 in the spinous layer might be due to low frequency of methylation as seen in our study (Table 1). To the best of our knowledge, reports related to methylation of P16 and RBSP3 in different layers of normal cervical epithelium are unavailable, although low frequency of P16 methylation was reported earlier in normal cervix [11] and also in our study. On the other hand, high/medium expression of P16 in basal–parabasal layers was seen in some (35–52%) normal cervical epithelium and primary cervical lesions (53–44%) (Table 1). Interestingly, these samples showed significant association with minor alleles of 31 deleterious non-synonimous SNPs of P16 exon1 and exon2. Earlier, it was reported that four deleterious SNPs (rs11552823, rs104894094, rs121913386, and rs104894098) of P16 exon2 could lead to intact but functionally impaired P16 protein in melanoma [15]. But, for the first time, we reported high prevalence of these four SNPs along with the other 27 deleterious SNPs in normal cervix and primary cervical lesions. Interestingly, high linkage existed between them, indicating additive effect of these SNPs might lead to reduced function of overexpressed P16 protein in basal/parabasal layers and primary cervical lesions (Figure 5B). Functional inactivation occurs due to protein misfolding by keeping secondary structure unaltered. Also inactivating SNPs induces the formation of large aggregates of P16 protein, reducing the effective concentration of P16 protein monomers and makes CDK-binding sites less available [15]. As per our knowledge, this important mechanism of P16 inactivation other than its genetic/epigenetic alterations leading to overexpression of p-RB1 (ser807/811 and ser567) has never been investigated earlier. In addition, high/medium expression of RBSP3 was seen in basal/parabasal layers of some normal cervical epithelium (27–44%) (Table 1). This might be due to the prevalence of inactivating splice variant, shorter isoform of RBSP3 [5,29]. The inactivating isoform of RBSP3 was prevalent in head and neck squamous cell carcinoma (HNSCC) and CACX, though its presence was seen in some respective normal epithelium [5,29]. Thus, it can be suggested that promoter methylation/deleterious SNPs/inactivating splice variants of P16 and RBSP3 in the basal–parabasal epitheliums of normal cervix might lead to the overexpression of p-RB1 (ser807/811 and ser567).

In CIN and CACX, high/medium expression of RB1 (74–80%) and p-RB1 (ser807/811 and ser567) (83–93%) was seen, while low expression of P16 (47–56%) and RBSP3 (73%) was evident in those samples (Table 1). The expression patterns of the above proteins were comparable to their respective expression pattern in basal–parabasal layers of normal epithelium (Table 1). The reduced expression of P16 and RBSP3 in CIN and CACX samples was due to frequent promoter methylation (36–46% and 51–62%, respectively; Figure 1 and Supplementary Table S6). Promoter methylation of P16 and RBSP3 in cervical lesions has been reported in earlier studies [5,10]. Methylation of RB1 was not done due to its low frequency reported earlier [6]. Thus, high expression of p-RB1 (ser807/811 and ser567) and low expression/high methylation of P16 and RBSP3 might be a molecular signature of normal basal–parabasal cervical epithelium which was maintained during cervical carcinogenesis. Increase in P16 and RBSP3 mRNA expression after 5-aza-dC treatment confirms the promoter methylation as one of the causes of reduced expression of these genes during CACX progression (Figure 2A). Similarly in ICC analysis, increase in P16 and RBSP3 expression and decrease in p-RB1 (ser807/811 and ser567) expression were markedly visualized after 5-aza-dC treatment (Figure 2b,c). This clearly suggests that up-regulation of P16 and RBSP3 could down-regulate the phosphorylation of RB1. Deletion of P16 and RB1 increased significantly in the invasive stages of CACX, while deletion of RBSP3 was high in the CIN samples. P16 showed similar frequency of deletion in CIN, but comparatively low frequency in CACX in earlier studies [10]. While, similar frequency of RB1 and RBSP3 deletion in CACX was reported in our earlier studies [5]. But in the present study, we did deletion analyses for P16 and RBSP3 in same set of samples. Like promoter methylation, deletions of P16 and RBSP3 showed association with their reduced expression (Supplementary Table S8). As a result, there was overexpression of inactive p-RB1 (ser807/811 and ser567) during the development of CACX. Deletion of both P16 and RBSP3 showed association with their respective promoter methylation indicating inactivation of their both alleles (Supplementary Table S9). In addition, the hemizygous deletion of RB1 seen during development of CACX might provide an additional growth advantage of the tumor (Supplementary Table S4).

Mutation of P16 in cervical lesions was infrequent and comparable with earlier reports (Figure 4B) [10]. The mutation of P16 scattered in both exon1 and exon2 suggests truncation of ankyrin repeats and cyclin-dependent kinase inhibitory domain of P16. These mutations were earlier reported in cervical cancer, HNSCC, melanoma, etc. (Ensembl release 83; Genome Database, cosmic) [10,34]. Majority of the mutated samples showed biallelic alterations having methylation and/or deletion (Figure 4B). Most of the mutated samples showed low expression of P16 (Figure 4B). On the other hand, samples with medium expression showed presence of minor alleles of deleterious SNPs earlier reported in melanoma (Figure 4B).

The overall alterations of P16 increased gradually from basal–parabasal layers to CIN and CACX (Figures 1Aiiiand 3Biii). While alterations of RBSP3 was comparable throughout CACX progression (Figures 1Biii and 3Biii). Likewise, overall alterations of P16 and RBSP3 showed association with their respective reduced expressions (Supplementary Table S12), leading to overexpression of p-RB1 (ser807/811 and ser567) during disease progression. Interestingly, synergistic alterations of P16 and RBSP3 in CACX (Supplementary Table S13) suggest that down-regulation of P16 and/or RBSP3 is necessary for the development of CACX. Significant association of P16 and/or RBSP3 alterations with poor patient survival indicates that alteration of either of the genes may be used as a prognostic marker in CACX (Figure 6).

Thus, our data suggest that overexpression of p-RB1 (ser807/811 and ser567) in basal–parabasal layers of normal cervical epithelium was due to inactivation of P16 and RBSP3 by methylation/low functional non-synonimous SNPs. This inactivation was maintained during CACX progression by additional deletion/mutation. However, more studies are needed to put insights on RB1 phosphorylation pathway in cervical carcinogenesis.

Abbreviations

CACX, cervical carcinoma; CDK2, cyclin dependent kinase 2; CDK4, cyclin dependent kinase 4; CIN, cervical intraepithelial neoplasia; HNSCC, head and neck squamous cell carcinoma; hr, high risk; HPV, human papillomavirus; ICC, immunocytochemistry; IHC, immunohistochemistry; LCM, laser capture microdissection; LOH, loss of heterozygosity; MA, microsatellite size alteration; MSP, methylation-specific; MSRA, methylation-sensitive restriction analysis; OS, overall survival; PCR, polymerase chain reaction; POLYPHEN, polymorphism phenotyping; p-RB1, phosphorylated RB1; SNP, single nucleotide polymorphism; SSCP, single-stranded confirmation polymorphism.

Author Contribution

C.C. performed all the experiments and wrote the paper. A.R.C. and S.S. helped in DNA sequencing and quantitative real-time analysis. A.R. did the histopathological examination of the tissue sections and checked the stages and grades of tumor. R.K.M. and P.B. provided all the clinical samples. S.R. helped in experimental design and C.K.P. designed the whole study and corrected the manuscript.

Funding

This work was supported by CSIR (Council of Scientific and Industrial Research, Government of India)-JRF/NET grant [File No.09/030(0059)/2010-EMR-I] to Mr C.C., grant [Sr. No. 2121130723] from UGC (University Grants Commission, Government of India) to Mr S.S., grant [SR/SO/HS-116/2007] from DST (Department of Science and Technology, Government of India) to Dr C.K.P. and Dr S.R. and grant [No. 60(0111)/14/EMR-II of dt.03/11/2014] from CSIR (Council of Scientific and Industrial Research, Government of India) to Dr C.K.P.

Acknowledgments

The authors thank the Director of Chittaranjan National Cancer Institute, Kolkata, India. We are grateful to Professor (Dr) H. Zur Hausen and Professor (Dr Mrs) E.M. de Villiers for their generous gift of HPV-16/18 plasmids.

Competing Interests

The Authors declare that there are no competing interests associated with the manuscript.

References

References
1
Arbyn
,
M.
,
Castellsague
,
X.
,
de Sanjose
,
S.
,
Bruni
,
L.
,
Saraiya
,
M.
,
Bray
,
F.
et al. 
(
2011
)
Worldwide burden of cervical cancer in 2008
.
Ann. Oncol.
22
,
2675
2686
doi:
2
WHO/ICO Information Centre on Human Papilloma Virus (HPV) and Related Diseases report in India. Summary Report.
2014
3
Snijders
,
P.J.F.
,
Steenbergen
,
R.D.M.
,
Heideman
,
D.A.M.
and
Meijer
,
C.J.L.M.
(
2006
)
HPV-mediated cervical carcinogenesis: concepts and clinical implications
.
J. Pathol.
208
,
152
164
doi:
4
López
,
J.
,
Ruíz
,
G.
,
Organista-Nava
,
J.
,
Gariglio
,
P.
and
García-Carrancá
,
A.
(
2012
)
Human papillomavirus infections and cancer stem cells of tumors from the uterine cervix
.
Open Virol. J.
6
,
232
240
doi:
5
Mitra
,
S.
,
Mazumder Indra
,
D.
,
Bhattacharya
,
N.
,
Singh
,
R.K.
,
Basu
,
P.S.
,
Mondal
,
R.K.
et al. 
(
2010
)
RBSP3 is frequently altered in premalignant cervical lesions: clinical and prognostic significance
.
Genes Chromosomes Cancer
49
,
155
170
doi:
6
Bahnassy
,
A.A.
,
Zekri
,
A.R.N.
,
Saleh
,
M.
,
Lotayef
,
M.
,
Moneir
,
M.
and
Shawki
,
O.
(
2007
)
The possible role of cell cycle regulators in multistep process of HPV-associated cervical carcinoma
.
BMC Clin. Pathol.
7
,
4
doi:
7
Takaaki
,
S.
,
Tetsunarl
,
O.
,
Kenji
,
K.
,
Toshio
,
F.
and
Takashi
,
N.
(
1998
)
Immunohistochemical overexpression of pl6 protein associated with intact retinoblastoma protein expression in cervical cancer and cervical intraepithelial neoplasia
.
Pathol. Int.
48
,
580
585
doi:
8
Jin
,
S.L.
,
Zhang
,
Z.S.
and
Yang
,
Y.H.
(
2000
)
[The expression of p16 and Rb protein in the tissue of cervical squamous carcinoma]
.
Hunan Yi Ke Da Xue Xue Bao
25
,
393
396
PMID:
[PubMed]
9
Buchkovich
,
K.
,
Duffy
,
L.A.
and
Harlow
,
E.
(
1989
)
The retinoblastoma protein is phosphorylated during specific phases of the cell cycle
.
Cell
58
,
1097
1105
doi:
10
Tripathi
,
A.
,
Banerjee
,
S.
,
Roy
,
A.
,
Roychowdhury
,
S.
and
Panda
,
C.K.
(
2003
)
Alterations of the P16 gene in uterine cervical carcinoma from Indian patients
.
Int. J. Gynecol. Cancer
13
,
472
479
doi:
11
Carestiato
,
F.N.
,
Afonso
,
L.A.
,
Moyses
,
N.
,
Almeida Filho
,
G.L.
,
Velarde
,
L.G.
and
Cavalcanti
,
S.M.
(
2013
)
An upward trend in DNA p16ink4a methylation pattern and high risk HPV infection according to the severity of the cervical lesion
.
Rev. Inst. Med. Trop. Sao Paulo
55
,
329
334
doi:
12
Jha
,
A.K.
,
Nikbakht
,
M.
,
Jain
,
V.
,
Capalash
,
N.
and
Kaur
,
J.
(
2012
)
p16INK4a and p15INK4b gene promoter methylation in cervical cancer patients
.
Oncol. Lett.
3
,
1331
1335
doi:
13
Wong
,
Y.F.
,
Chung
,
T.K.H.
,
Cheung
,
T.H.
,
Nobori
,
T.
,
Yu
,
A.L.
,
Yu
,
J.
et al. 
(
1999
)
Methylation of p16INK4A in primary gynecologic malignancy
.
Cancer Lett.
136
,
231
235
doi:
14
Park
,
J.S.
,
Dong
,
S.M.
,
Kim
,
H.S.
,
Lee
,
J.Y.
,
Um
,
S.J.
,
Park
,
I.S.
et al. 
(
1999
)
Detection of p16 gene alteration in cervical cancer using tissue microdissection and LOH study
.
Cancer Lett.
136
,
101
108
doi:
15
Zhang
,
B.
and
Peng
,
Z.-y.
(
1996
)
Defective folding of mutant p16(INK4) proteins encoded by tumor-derived alleles
.
J. Biol. Chem.
271
,
28734
28737
doi:
16
Koh
,
J.
,
Enders
,
G.H.
,
Dynlacht
,
B.D.
and
Harlow
,
E.
(
1995
)
Tumour-derived p16 alleles encoding proteins defective in cell-cycle inhibition
.
Nature
375
,
506
510
doi:
17
Nakashima
,
R.
,
Fujita
,
M.
,
Enomoto
,
T.
,
Haba
,
T.
,
Yoshino
,
K.
,
Wada
,
H.
et al. 
(
1999
)
Alteration of p16 and p15 genes in human uterine tumours
.
Br. J. Cancer
80
,
458
467
doi:
18
Yuan
,
J.R.
,
Li
,
Y.
,
Hu
,
S.-J.
and
Li
,
P.
(
2005
)
[Methylation and aberrant expression of the p16 gene in cervical carcinoma]
.
Yi Chuan
27
,
39
43
PMID:
[PubMed]
19
Ruan
,
Y.H.
,
Wei
,
W.L.
,
Zhang
,
H.X.
,
Liang
,
Z.N.
,
Liu
,
B.Y.
and
Chen
,
Y.
(
2003
)
[Comparison and analysis of expression of c-myc and p16 in cervical carcinoma]
.
Ai Zheng
22
,
602
606
PMID:
[PubMed]
20
Lin
,
J.
,
Albers
,
A.E.
,
Qin
,
J.
and
Kaufmann
,
A.M.
(
2014
)
Prognostic significance of overexpressed p16INK4a in patients with cervical cancer: a meta-analysis
.
PLoS ONE
9
,
e106384
doi:
21
Kanao
,
H.
,
Enomoto
,
T.
,
Ueda
,
Y.
,
Fujita
,
M.
,
Nakashima
,
R.
,
Ueno
,
Y.
et al. 
(
2004
)
Correlation between p14ARF/p16INK4A expression and HPV infection in uterine cervical cancer
.
Cancer Lett.
213
,
31
37
doi:
22
Lesnikova
,
I.
,
Lidang
,
M.
,
Hamilton-Dutoit
,
S.
and
Koch
,
J.
(
2009
)
p16 as a diagnostic marker of cervical neoplasia: a tissue microarray study of 796 archival specimens
.
Diagn. Pathol.
4
,
22
doi:
23
Anedchenko
,
E.A.
,
Kiseleva
,
N.P.
,
Dmitriev
,
A.A.
,
Kiselev
,
F.L.
,
Zabarovskii
,
E.R.
and
Senchenko
,
V.N.
(
2007
)
[Tumor suppressor gene RBSP3 in cervical carcinoma: copy number and transcriptional level]
.
Mol. Biol. (Mosk.)
41
,
86
95
PMID:
[PubMed]
24
Dasgupta
,
S.
,
Mukherjee
,
N.
,
Roy
,
S.
,
Roy
,
A.
,
Sengupta
,
A.
,
Roychowdhury
,
S.
et al. 
(
2002
)
Mapping of the candidate tumor suppressor genes’ loci on human chromosome 3 in head and neck squamous cell carcinoma of an Indian patient population
.
Oral Oncol.
38
,
6
15
doi:
25
Sambrook
,
J.
,
Fritsch
,
E.F.
and
Maniatis
,
T.
(
1989
)
Molecular Cloning: A Laboratory Manual
,
2nd edn
,
Cold Spring Harbor Laboratory
,
Cold Spring Harbor, NY
26
Chakraborty
,
C.
,
Dutta
,
S.
,
Mukherjee
,
N.
,
Samadder
,
S.
,
Roychowdhury
,
A.
,
Roy
,
A.
et al.  (2015
)
Inactivation of PTCH1 is associated with the development of cervical carcinoma: clinical and prognostic implication
.
Tumor. Biol.
36
,
1143
1154
doi:
27
Singh
,
R.K.
,
Maulik
,
S.
,
Mitra
,
S.
,
Mondal
,
R.K.
,
Basu
,
P.S.
,
Roychowdhury
,
S.
et al. 
(
2006
)
Human papillomavirus prevalence in postradiotherapy uterine cervical carcinoma patients: correlation with recurrence of the disease
.
Int. J. Gynecol. Cancer
16
,
1048
1054
doi:
28
Ghosh
,
S.
,
Ghosh
,
A.
,
Maiti
,
G.P.
,
Alam
,
N.
,
Roy
,
A.
,
Roy
,
B.
et al. 
(
2008
)
Alterations of 3p21.31 tumor suppressor genes in head and neck squamous cell carcinoma: correlation with progression and prognosis
.
Int. J. Cancer
123
,
2594
2604
doi:
29
Maiti
,
G.P.
,
Ghosh
,
A.
,
Chatterjee
,
R.
,
Roy
,
A.
,
Sharp
,
T.V.
,
Roychoudhury
,
S.
et al.  (2012
)
Reduced expression of Limd1 in ulcerative oral epithelium associated with tobacco and areca nut
.
Asian Pac. J. Cancer Prev.
13
,
4341
4346
doi:
30
Perrone
,
F.
,
Suardi
,
S.
,
Pastore
,
E.
,
Casieri
,
P.
,
Orsenigo
,
M.
,
Caramuta
,
S.
et al. 
(
2006
)
Molecular and cytogenetic subgroups of oropharyngeal squamous cell carcinoma
.
Clin. Cancer Res.
12
,
6643
6651
doi:
31
Loginov
,
V.I.
,
Maliukova
,
A.V.
,
Seregin
,
I.A.
,
Khodyrev
,
D.S.
,
Kazubskaia
,
T.P.
,
Ermilova
,
V.D.
et al. 
(
2004
)
[Methylation of the promoter region of the RASSF1A gene, a candidate tumor suppressor, in primary epithelial tumors]
.
Mol. Biol. (Mosk.)
38
,
654
667
PMID:
[PubMed]
32
Ivanova
,
T.
,
Petrenko
,
A.
,
Gritsko
,
T.
,
Vinokourova
,
S.
,
Eshilev
,
E.
,
Kobzeva
,
V.
et al. 
(
2002
)
Methylation and silencing of the retinoic acid receptor-β2 gene in cervical cancer
.
BMC Cancer
2
,
4
doi:
33
Herman
,
J.G.
,
Graff
,
J.R.
,
Myohanen
,
S.
,
Nelkin
,
B.D.
and
Baylin
,
S.B.
(
1996
)
Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands
.
Proc. Natl Acad. Sci. USA
93
,
9821
9826
doi:
34
Ghosh
,
A.
,
Ghosh
,
S.
,
Maiti
,
G.P.
,
Sabbir
,
M.G.
,
Alam
,
N.
,
Sikdar
,
N.
et al. 
(
2009
)
SH3GL2 and CDKN2A/2B loci are independently altered in early dysplastic lesions of head and neck: correlation with HPV infection and tobacco habit
.
J. Pathol.
217
,
408
419
doi:
35
Livak
,
K.J.
and
Schmittgen
,
T.D.
(
2001
)
Analysis of relative gene expression data using real-time quantitative PCR and the 2−ΔΔCT method
.
Methods
25
,
402
408
doi:
36
Ghosh
,
A.
,
Ghosh
,
S.
,
Maiti
,
G.P.
,
Sabbir
,
M.G.
,
Zabarovsky
,
E.R.
,
Roy
,
A.
et al.  (2010
)
Frequent alterations of the candidate genes hMLH1, ITGA9 and RBSP3 in early dysplastic lesions of head and neck: clinical and prognostic significance
.
Cancer Sci.
101
,
1511
1520
doi:
37
Barrett
,
J.C.
,
Fry
,
B.
,
Maller
,
J.
and
Daly
,
M.J.
(
2005
)
Haploview: analysis and visualization of LD and haplotype maps
.
Bioinformatics
21
,
263
265
doi:
38
Cheng
,
J.
,
Randall
,
A.
and
Baldi
,
P.
(
2006
)
Prediction of protein stability changes for single-site mutations using support vector machines
.
Proteins
62
,
1125
1132
doi:

Supplementary data