The current standard of care for the early detection of cervical cancer based on Pap smear has significant limitations, false negative results being the most important. In view of this, testing for HR-HPVs is increasingly used in the screening or diagnosis of cervical cancer. However, because HR-HPV infections are highly prevalent among sexually active women and most HPV infections, including HPV 16, do not always lead to cervical cancer, this approach also has limitations. Therefore, it is worthwhile to investigate molecular changes in HPVs rather than infections per se as screening or diagnostic biomarkers of CIN or ICC.
CpG methylation within the HPV genome has been reported previously, but documenting its prevalence, biological or clinical significance has begun only recently. The study of Badal et al (8
) indicated that methylation of the long control region (LCR) and the E6 gene decreases with disease progression, but they did not show the functional effect of methylation-mediated inhibition of E6 gene expression. The study of Hublarova et al addressed this limitation by demonstrating that decreased methylation frequency correlates with the progression of CIN and its association with E6 gene expression by quantitative analysis of E6 mRNA levels (13
). In contrast to the two studies cited above, methylation of the E6 LCR has also been shown to increase in cervical cancer. Bhattacharjee and Sengupta (14
) reported that LCR-E6 methylation was higher in normal subjects than in cervical cancer samples, but this difference was not significant. Kalantari et al. (9
) found that while patients with asymptomatic infections generally had lower methylation of the HPV 16 promoter/enhancer region than patients with cervical carcinoma, low- and high-grade CIN lesions showed the lowest methylation frequency of all the specimens tested. In another study, although premalignant cervical cells from clinical samples showed a pattern consistent with slight methylation of the LCR, the predominant methylation in these samples occurred in the HPV 16 late genes and the E5 gene (15
). Studies showing associations between HPV 16 E6 methylation and CIN risk have thus been inconclusive. None of these studies, however, adjusted their results for demographic, lifestyle or risk factors for cervical cancer. Our study which adjusted for these factors demonstrate that a higher degree of methylation in CpG positions of both the promoter and enhancer are independently associated with reduced likelihood of being diagnosed with CIN 2+, but higher methylation status of the promoter and enhancer combined was associated with the most significant reduced risk of CIN 2+. The CpG residue at position 7862 is nearly always unmethylated in CIN and ICC lesions. This CpG dinucleotide is critically positioned between two nucleosomes that make up the enhancer and promoter region of HPV 16 E6. The CpG 7862 residue is part of an E2 binding site that activates HPV 16 replication, and methylation at this site would displace E2 and block replication. Our results showing that CIN 2+ risk is decreased with higher methylation at either E6 enhancer position 7862 or E6 promoter sites, and is significantly decreased with higher methylation at both sites, is consistent with the idea that methylation at one or both sites may be blocking HPV 16 replication and thereby decreasing CIN risk. E2 binding sites span regions 7853–7872 (designated site 3) and 46–65 (designated site 1) (16
). The former region encompasses the 7862 CpG site, while the latter region covers a portion of the five CpG sites in the E6 promoter that we analyzed in the current study. The cellular transcription factor CEF-2 has been reported to bind tightly to E2 site 3, whereas an unrelated transfection factor, CEF-1, binds tightly to E2 site 1 (16
). When binding of CEF-1 to site 1 or CEF-2 to site 3 was blocked by mutating their respective binding sites, there was significantly reduced P97 promoter activity (16
). We hypothesize that increased methylation at both site 1 and site 3 may also block binding of these two transcription factors, and thereby reduce HPV 16 E6 expression and mitigate CIN 2+ risk.
These results suggest that CpG methylation is involved in the biology of HPV-16 as well as in the development of higher grades of CIN. Therefore, evaluation of methylation in these CpG positions of HPV 16 genome is likely to be useful in identifying women with underlying higher grades of cervical lesions that need medical care. Prospective studies are needed to evaluate the usefulness of this biomarker for identifying HPV 16 positive women who may progress toward higher grades of CIN or for identifying recurrence of CIN after HPV 16 positive lesions are treated. The ability to evaluate methylation status in non-invasively obtained exfoliated cervical cells by using a simple to perform and reproducible pyrosequencing technology will enable repeated monitoring of methylation status in clinical settings.
Dietary or life style factors which may alter the methylation status of HPV 16 are virtually unknown. One carbon micronutrients such as folate and vitamin B12 which are involved in the generation of methyl groups for methylation reactions are more likely than other factors to have influence on DNA methylation. Our recent results from the same cohort of women (17
) demonstrated a significant positive association between supraphysiologic concentrations of folate (> 19.8 ng/mL) and sufficient vitamin B12 (≥ 200.6 pg/mL) and global methylation status of host DNA. Because only 13% of women have this combination, the current study did not have adequate power to test the associations between these micronutrients and methylation status of HPV 16. Our results, however, showed that women with a higher overall HEI were significantly more likely to have higher methylation in the HPV 16 enhancer, suggesting that “cancer-protective” micronutrients are likely to have modifying effects on HPV 16 methylation. Future studies with larger sample sizes are needed to evaluate these associations in detail.
Limitations of the study include small sample size and the cross-sectional nature of the study. The sample size was not determined by design, as it was the product of testing for specific HPV types within a group of women who were eligible for inclusion in the parent study (abnormal pap smear or positive for high-risk HPVs). The limited study size is reflected by the relatively wide 95% confidence intervals of the measures of association, which indicate that the results are imprecise even when significance testing allows rejecting the null hypothesis of no association. Confirmation of the findings from this cross-sectional study using prospectively collected samples from the parent study will increase the scientific credibility of the significance of HPV 16 methylation in modifying the development of CIN 2+.
The limitations listed above are offset by considerable strengths. First, this study assessed cervical lesions by employing a rigorous pathology review protocol, used sensitive and reproducible techniques for testing for HPV and for evaluating the degree of methylation in specific regions of the HPV 16 promoter and enhancer and measured several cancer related micronutrients and other risk factors for CIN. To our knowledge, this is the first study to integrate the measurement of HPV 16 DNA methylation with a thorough assessment of potential determinants and confounders. Multiple regression analysis of the data simultaneously controlled life-style factors and micronutrients, thus disentangling the effect of micronutrient levels from the effect of other covariates. Despite the limited study size and the regression adjustment of multiple covariates, the association of HPV 16 DNA methylation with the prevalence of CIN2+ was statistically significant. In conclusion, the present study provides initial evidence that CpG methylation is involved in thebiology of HPV-16 as well as in the development of higher grades of CIN.