P16ink4a Immunoexpression Profile in HPV-Oral Lesions from HIV-Infected Patients

Rosales-Terrazas Estrella1, Ramírez-Amador Velia1, García-Carrancá Alejandro2,3, Guido-Jiménez Miriam2,3, Irigoyen-Camacho Esther1, Ortíz-Sánchez Elizabeth2 and Anaya-Saavedra Gabriela1* 1Universidad Autónoma Metropolitana-Xochimilco, Mexico 2Unidad de Virus y Cáncer, Instituto Nacional de Cancerología (INCan), Secretaría de Salud, Mexico 3Instituto de Investigaciones Biomédicas, Universidad Nacional Autónoma de México, Ciudad de México

The p16 INK4a gene, located on chromosome 9p21, is a cyclindependent kinase inhibitor that acts as a negative cell cycle regulator [19,20], therefore, is an important biomarker of cellular transformation [21,22]. Besides its role as tumor suppressor gene, p16 INK4a has been associated with HR-HPV infection in cervical cancer [23] and, although controversial results, in a number of oral and oropharyngeal cancer studies [19,24,25]. The HPV-E7 oncogen inactivates pRB, leading to p16 INK4a overexpression via E2F alteration; thus, p16 INK4a overexpression is considered a hallmark of HPV-induced transformation [23,26] and a surrogate marker of HR-HPV oncogenic activity [22,26].
Thus the aim of the present study was to investigate the value of p16 INK4a in HPV-OL considering the finding of HR-HPV in HPV-OLs and the increased risk of oral and oropharyngeal cancer in HIV-patients.

Method
A cross-sectional study was conducted in three HIV/AIDS referral centers in Mexico City (Clínica Especializada Condesa (CEC), Instituto Nacional de Ciencias Médicas y Nutrición Salvador Zubirán and Instituto Nacional de Enfermedades Respiratorias). The study was approved by the review boards of the participating institutions, and the participants read and signed an informed written consent form before their participation in the study.
(CEC) between 2009 and 2011, since these patients were included in a previous report [6]. Demographic data (age, sex, schooling, and occupation), tobacco use, and alcohol consumption were recorded. Clinical and laboratory data including HIV-transmission category, CDC clinical stage [27], lymphocyte CD4 + T-cell count, HIV viral load levels, and type and duration of HAART administration were obtained from the medical records. HAART was defined as the concomitant use of three antiretroviral (ARV) drugs, either a combination of two types of ARVs (nucleoside reverse transcriptase inhibitors [NRTIs], non-nucleoside reverse transcriptase inhibitors [NNRTIs], protease inhibitors [PIs], or a fusion inhibitor [FIs]), or three NRTIs [28]. Patients were considered to be on HAART if they received ARV drugs for more than 30 days.
HPV-OLs were identified according to predetermined clinical criteria [1,29,30] by three Oral Pathology and Medicine specialists with experience in the recognition of these lesions. The clinical characteristics of HPV-OLs were registered; a section of the tissue was obtained using a disposable punch, fixed in a 10% formalin solution, and then the tissue was processed. The histopathological diagnosis was performed by the Oral Pathology and Medicine Department staff of the Universidad Autónoma Metropolitana, based on pre-established criteria [1,29,30].
Although a number of patients showed more than one HPV-OL, the largest, or the one that produced more discomfort was chosen to be removed, and therefore, utilized to the molecular analysis, particularly when secondary lesions were too small or did not cause discomfort, so, surgical treatment was not indicated.
DNA was extracted using the QIAmp DNA FFPE tissue kit (Qiagen, Duesseldorf, Germany) for purification of genomic DNA from formalin-fixed paraffin-embedded tissues, according to the manufacturer's instructions. The total DNA concentration was determined by spectrophotometry and the ratio of 260/280 was measured using a NanoDrop 2000 spectrophotometer (Thermo Scientific/Waltham, MA, USA). To demonstrate the integrity of the purified DNA, a β-globin gene PCR assay (260/280 pb) was performed in each sample, with PCO4/GH20 primers. Genomic HPV-DNA amplification was carried out using MY09/MY11 [31] and GP5+/GP6+ primers [32]. Additionally, in order to identify HPV 16 and 18, E6 specific primers F204/R419 [33] and HZ30/E65 [34] were used.
In every PCR reaction, water and DNA from the HeLa cell line were used as a negative and positive control, respectively. PCR was performed in a programmable thermal cycler (Mastercycler gradient; Eppendorf, Westbury, NY). All the reagents used were prepared and stored in a PCR-amplified products free area. In order to visualize the PCR products after amplification, a 5 µl aliquot of each sample was analyzed by electrophoresis in a 1.5% agarose gel with ethidium bromide staining (10 mg/ml) (Sigma Chemical Co., St. Louis, MO, USA) and was then visualized by UV light.
Positive PCR products were purified using the DNA Clean and Concentrator TM -5 purification protocol (Zymo Research Corp/USA), and subsequently sent to an external service (Macrogen Inc., Seoul, Korea) for automated DNA sequencing by using one of the PCR primers as a sequencing primer (GP5+). The obtained sequences were matched and compared with the GenBank database sequences (National Center for Biotechnology Information, Bethesda, MD, USA) using the BLAST program (http://blast.ncbi.nlm.nih.gov/), blinded to the histopathological results.
Additionally, for the immunohistochemical detection of p16 INK4a , three-micrometer paraffin-embedded HPV-OL sections were placed on Kling-on HIER supercharged slides (Biocare Medical) and immunostained with the anti-p16 mousse monoclonal antibody (SAB3300036 Sigma, monoclonal anti-p16 antibody, Sigma-Aldrich Co. LLC., MO, USA) at a 1:250 dilution. The staining was performed on the Ventana BenchMark GX (Roche Diagnostics International Ltd., Rotkreuz, Switzerland) automated staining system, according to manufacturer's instructions. The slides were counterstained with hematoxylin for 4 minutes, and post-counterstained with Bluing Reagent, a buffered lithium carbonate solution (Ventana Medical Systems, Inc) for other 4 minutes. A cervical cancer sample and a p16 INK4a negative oral cancer sample were used as positive and negative controls, respectively. Additionally, in order to have comparison parameters, four normal oral mucosa samples from equivalent locations (labial, lingual, buccal, and palatal mucosa) obtained from two oral mucoceles and two fibrous hyperplasias samples were also processed.
All immunohistochemical slides were examined for positive staining by light microscopy by three blinded independent observers. Each slide was placed on the optical photomicroscope (Olympus BX43; Olympus, San Diego, CA, USA), and four digital pictures were taken with 10x objective on each sample; two marginal and two central epithelium fields were selected and stored as jpeg files. To calculate the percentages of positive cells, the numbers for both negative and positive cells were counted in each image using a 6x6 grid. Afterwards, the percentage of positive cells was calculated as follows: positive cell nuclei⁄total cell nuclei × 100. Positivity of p16 INK4a was considered when >75% of cells showed a moderate/strong nuclear (with or without cytoplasmic) stain [35].
Descriptive data were summarized through medians and interquartile intervals. In order to determine the association between categorical variables and the dependent variables, the X 2 and Fisher's exact (when necessary) tests were used; for dimensional variables the U-Mann-Whitney was applied. Risk factors were assessed using logistic regression models, and odds ratios and 95% confidence intervals for the corresponding categories were constructed. A 2-sided p value ≤0.05 was considered significant. The statistical analysis was done in the SPSS (v.20) program.  (Table 1).

Twenty
As it is shown in Table 2, 18 (62.1%) of the 29 patients showed more than one HPV-OL, being MEH the most frequent type of lesion. Patients with multiple HPV-OL had a longer median time of HAART use (44 months, Q 1 -Q 3 : 8-87) than those with solitary lesions (7 months, Q 1 -Q 3 : 2-16 months) (p=0.015), and were in more advanced stages of the disease (100%) than those with one HPV-OL (61%) (p=0.026).

P16 INK4a immunoexpression
All the normal oral mucosa samples were negative for p16 INK4a according to the criteria described in the material and methods section: the samples showed a weak staining in the basal cell layer, which gradually decreased (until disappearance) towards to superficial layers. In contrast, at it is explained in Table 4, all HPV-OL showed a moderate/ strong nuclear p16 INK4a immunostaining (median of 97.9%). It is remarkable that despite the total and intense nuclear expression, none of the HPV-OL samples exhibited cytoplasmic p16 INK4a immunostaining ( Figure 2).
No correlation was observed between the p16 INK4a immunostaining and the specific type of lesion, the HPV type (low or high risk), or the clinical characteristics of the patients.

Discussion
In the present study, though low risk-HPV was identified in the majority of HPV-OLs from HIV/AIDS patients, HR-HPV sequences were found in 10.3% of cases, always in combination with LR-HPV. Additionally, a moderate/strong p16 INK4a nuclear immunostaining was found, in contrast with the weak or absent expression observed in normal oral mucosa. The comparable p16 INK4a staining in both low and high risk-HPV types, as well as the absence of cytoplasmic expression, suggest that the identification of HR-HPV in these lesions was not biologically significant.
While in HIV-patients there are few detailed reports on the epidemiology and behaviour of HPV-OLs [36], the prevalence found in the present study (3.4%) correspond to previous reported frequencies (3-6.9%) [6,37]. Despite the relatively low frequency of HPV-OL in HIV-patients, the lesions could exhibit a widespread distribution and high rates of recurrence, that usually implies complications in its management and treatment [38].
Most published studies have used the generic term "oral warts" to describe HPV-OLs [7][8][9]12,37,[39][40][41]; only few studies [6,[42][43][44][45], as this one, indicate the specific type of HPV-OL, so, the possibility of making comparisons is limited. However, in the present study, like in a previous one [6], the most common lesion was MEH. In Mexico [46][47][48], like in other Latino American countries such as Colombia [49] and Peru [50], MEH may be present in specific ethnic and racial groups, particularly in childhood [51], which may explain the high frequency found in our population.
The labial mucosa was the most frequently affected site (55.2%), similar to previous reports [6,52,53]. According with Mravak-Stipetic et al. [54], the labial mucosa is one of the most frequently exposed locations to microtrauma, which is a requirement for HPV transmission in different oral sites.  MEH=multifocal epithelial hyperplasia; SCP=squamous cell papilloma; VV=verruca vulgaris; AC=accuminated condyloma. a The sample had also b The sample had also HPV-16; c The sample had also HPV-32; d The sample had also HPV-42. Patients with more than one HPV-OL had been on HAART for longer periods of time than those with single lesions (44 vs. 7 months, p=0.015). It has been mention that some of the factors that could influence the augmented frequency of HPV-OL in HIV patients in the post-HAART era could be the increased survival of patients and the altered immune status for long periods of time, rather than the prolonged administration of antiretrovirals [6,55].
The majority of the samples (82.7%) had LR-HPV, higher than the 65.4% [6] and 47.1% [12] found in two previous studies that included 55 and 34 HPV-OL affected HIV-patients, respectively. The two cited studies, as well as the present one, used similar same sample collection techniques, the L1 consensus primers, and comparable HPV detection assays (PCR), so, the different frequencies could be explained by the inherent bias associated with small sample sizes, rather than with methodological differences.
In immune competent individuals HPV-13 and 32 have been identified in 75-100% [1] of MEH, HPV-13 being the commonest strain reported in most studies [1,48], including the Mexican population [47]. In contrast in our results, similar to other reports [56,57], HPV-32 was identified in most cases. The possible association between HPV-32 oral infection in HIV-infected patients merits future studies.
In this respect, several studies [58][59][60][61] have found HPV-infection in the normal oral mucosa of HIV/AIDS patients, in frequencies that varies from 20 [61] to 80% [58]. Moreover, current studies have reported the frequency of HR-HPV, specifically the type 16, in values that fluctuate from 0.6% [62] to 2.8% [63]. Interestingly, most of infections cleared within one year [62]; thus, probably the 10.3% of the HPV-OLs containing HR-HPV in our study represents a transient HPV infection.
According with the established criteria detailed in the material and methods section [35], that considered positivity of p16 when >75% of cells showed a moderate/strong nuclear (with or without cytoplasmic) stain, we assessed the normal oral mucosa samples in the present study as p16INK4a negative, similar to previous reports [64][65][66][67] in immunocompetent individuals. In contrast, and in agreement with recent evidence, the identification of >75% of nuclear and cytoplasmic p16 INK4a staining represents a good cutoff for HPV presence [35]. In consequence, whilst more than 95% of the samples in our results showed p16 INK4a nuclear positivity, the immunoexpression profile did not fulfill the criteria to consider it as a marker of HR-HPV infection, given that none of the samples exhibited cytoplasmic staining.
Additionally, p16 INK4a immunostaining in both LR and HR-HPV samples was comparable in the present study, which could be explained by the highly proliferative activity of the lesions. Physiological stress, oncogene-driven senescence and replicative senescence due to DNA damage or oxidative stress have been proposed to explain p16 INK4a overexpression in normal tissues [24]. Although there is less data about the molecular interactions made by E6 and E7 proteins of the LR-HPV, it is known that E7 protein from LR-HPV, as in HR-HPV, can also inactivate the RB protein and consequently, resulting in over expression of p16 INK4a [68]. According with a recent review [68], there is sufficient evidence explaining that despite considerable differences in the outcomes of low and high risk mucosotrophic HPVs, both strains infect and replicate in the same general manner, and presumably encounter the same cellular environments.
In conclusion, the nuclear immunoexpression of p16 INK4a in HPV-OLs did not accomplish the criteria for HPV oncoprotein driven overexpression, which is critical in the context of the increased cancer incidence reported in HIV/AIDS patients [15][16][17][18]. In HPV-OLs from HIV-infected patients, the comparable p16 INK4a immunoexpression, independently of the specific HPV-type, as well as the absence of cytoplasmic staining, may suggest a lack of HR-HPV activity.
Our results emphasize that HPV-OL in patients with HIV infection are benign, and could bring information that clinicians can employ to comfort patients that develop multiple HPV-OL in the setting of HIV infection and HAART. Longitudinal studies based on viral transcriptional activity are warranted.