A Genetic Variant in the Promoter Region of Toll-Like Receptor 9 and Cervical Cancer Susceptibility

Abstract

The Toll-like receptors (TLRs) are important for the innate immune system by recognizing pathogen-associated molecular patterns expressed in infectious agents. E6 and E7 protein from HPV16 suppress the host immune response by regulating the TLR9 transcript. Therefore, we hypothesized that a single nucleotide polymorphism in TLR9 may contribute to cervical cancer. We genotyped TLR9 -1486T/C (rs187084) in a case–control study of 712 cervical cancer cases and 717 cancer-free controls in Chinese women. Logistic regression analyses showed that the rs187084 heterozygote TC was associated with a significantly increased risk of cervical cancer (adjusted OR=1.28, 95% CI=1.01–1.62), compared with the TT genotype. Although the variant homozygote was associated with a nonsignificantly increased cervical cancer risk, the TC/CC genotypes contributed to the risk of cervical cancer in the dominant genetic model (adjusted OR=1.24, 95% CI=1.01–1.53). The findings indicate that TLR9 -1486T/C (rs187084) may contribute to cervical cancer carcinogenesis.

Introduction

C ervical cancer is the second most common cancer among women worldwide, with 83% of the cases occurring in developing countries (Parkin et al., 2005). Epidemiological and laboratory-based studies have identified that human papillomavirus (HPV) infection contributes to cervical cancer (Frazer et al., 1995; Tindle, 2002; zur Hausen, 2002; Chen et al., 2011). HPV can be classified into two groups: low-risk HPVs (e.g., types 6 and 11) that are mainly associated with benign genital warts and high-risk (HR) HPVs (e.g., types 16 and 18) that are the etiological agents of cervical cancer (zur Hausen, 2002). However, only a small percentage of infected persons with high-risk HPV develop low- and/or high-grade cervical intraepithelial neoplasia (CIN), which may regress or progress to invasive carcinoma (Frazer et al., 1995). Evidence suggested that immune regulation may be a key event for cervical cancer development (Tindle, 2002; Nardelli-Haefliger et al., 2005; Yang et al., 2005).

The Toll-like receptors (TLRs), a type of cell-surface signaling molecule, contribute to the host's innate immune system by recognizing pathogen-associated molecular patterns expressed in infectious agents. Once activated, the TLRs can not only induce the inflammatory response, but also affect antigen-specific adaptive immunity (Takeda et al., 2003). Therefore, mammalian TLRs play a key role in host defense to pathogenic infection by regulating and linking the innate and adaptive immune responses (Takeda et al., 2003). Ligand specificity has been elucidated for most members of the TLR family. Among those, TLR-2 and -4, respectively, recognize Gram-positive and Gram-negative bacterial cell wall products, whereas TLR-3, -7, and -8 recognize different forms of microbial-derived nucleic acid. Specially, TLR9 can be activated by bacteria dsDNA-derived CpG motifs or some viral components including those from HPV16 (Beutler, 2004). Studies have reported that E6 and E7 protein from HPV16 suppress the host immune response by regulating the TLR9 transcript (Hasan et al., 2007). Further, deletions in the TLR9 promoter have shown that element(s) involved in the HPV16-mediated inhibition are located within a 2 kb region upstream of the -700 to -1 regulatory region (Hasan et al., 2007). Therefore, variation of the TLR9 promoter may affect the function of HPV16 and subsequently play a role in HPV-related cervical cancer.

Human TLR9 is located at 3p21.3, a region frequently deleted in human cancers (Chuang and Ulevitch, 2000). Stimulation of TLR9 can activate human B cells and results in innate immune responses in preclinical tumor models and in patients (Murad et al., 2007; Rozkova et al., 2010 ). Hold et al. first identified a total of 20 TLR9 single nucleotide polymorphisms (SNPs) by direct sequencing in three ethnic samples (European American, African American, and Hispanic American) and a sample of European American asthmatics (Lazarus et al., 2002), but only four were found in the public dbSNP database (A-1923C, rs352144; T-1486C, rs187084; G1174A, rs352139; and G2848A, rs352140). Lately, a new SNP within the TLR9 proximal promoter region, substituting tyrosine to cytosine at nucleotide position −1237 (rs5743836), was reported to be associated with an increased risk to allergic asthma among an American-European cohort (Lazarus et al., 2003). Studies showed that rs5743836 and rs352144 were rare in the Asian population (Hur et al., 2005; Ito et al., 2007; Tao et al., 2007), and TLR9 -1486T/C (rs187084), 1174G/A (rs352139), and 2848G/A (rs352140) were in a high linkage disequilibrium among the Asian population (r ²>0.95) (Mockenhaupt et al., 2006; Ito et al., 2007; Tao et al., 2007). For rs187084 located in the promoter of TLR9, in silico analysis showed that this SNP created a putative Sp1 binding site, which may be functionally relevant (Hamann et al., 2006). In addition, a study found an association between C allele of rs187084 and increased risk of systemic lupus erythematosus (SLE) in the Japanese population, and reporter gene assay showed that this allele downregulated TLR9 expression (Tao et al., 2007). However, no studies have investigated the association between rs187084 and susceptibility to cervical cancer. In the current study, we hypothesized that the potentially functional SNP rs187084 (-1486T/C) may contribute to cervical cancer susceptibility. To test the hypothesis, we performed a case–control study including 712 cervical cancer cases and 717 age frequency-matched cancer-free controls in Chinese women.

Materials and Methods

Participants

The participants of our study were previously described (Chen et al., 2009). Briefly, the 712 patients in whom cervical cancer had been newly diagnosed and histologically confirmed were recruited from the Nantong Tumor Hospital and the First Affiliated Hospital of Nanjing Medical University, Jiangsu, China, between March 2006 and April 2009. The 717 controls were randomly selected from a pool of >20,000 individuals who had participated in a community-based screening program for noninfectious diseases conducted in Wujin county of Changzhou City, Jiangsu Province, China, by adopting a multi-stage and random cluster sampling method. For the 20,000 individuals, we randomly selected 7 towns from 14 towns in the county. Then, from each town, we selected three villages according to economic development status. All the controls were frequency matched to the cases with regard to age (±5 years) and residential areas (urban and rural). All the cases and control subjects were unrelated ethnic Han Chinese women and had no self-reported cancer history. After informed consent was obtained, each subject was interviewed face-to-face by trained interviewers using a structured questionnaire, and a 5-mL venous blood sample was collected. The study was approved by the Institutional Review Board of Nanjing Medical University.

Genotyping

The T-1486C (rs187084) variant was genotyped by using the PCR-restriction fragment length polymorphism assay. The primers were 5′ CATTCATTCAGCCTTCACTC 3′ (forward) and 5′ TTTGCCTGGTCTATGTTTCT 3′ (reverse). After AflII (New England BioLabs, Beverly, MA) digestion, the variant C allele produced one fragment of 268 bp, and the wild-type T allele generated two fragments of 147 and 121 bp. Quality control for genotyping was previously described (Chen et al., 2009), and >10% samples were randomly selected for repetition, thus yielding a 100% concordant. The calling rate of genotyping was 98.6%.

Statistical analysis

Differences in the distributions of demographic characteristics, selected variables, and genotypes of rs187084 between the cases and controls were evaluated using the student's t-test or χ ² test. Logistic regression analyses were used to evaluate the differences in the frequency distributions of genotypes of the polymorphism between the cases and controls by computing the crude and adjusted odds ratios (ORs) and 95% confidence intervals (CIs). The Hardy–Weinberg equilibrium was tested by a goodness-of-fit χ ² test to compare the observed genotype frequencies with the expected ones among the control subjects. We also applied the false positive report probability (FPRP) statistical tool to evaluate noteworthiness of the associations by using the method described by Wacholder et al. (2004). FPRP <0.5 was considered a noteworthy association. All the other statistical analyses were performed with SAS 9.1.3 (SAS Institute, Cary, NC).

Results

Selected characteristics of the 712 cervical cancer cases and the 717 cancer-free controls were described in Table 1. As expected, there was a similar distribution of age in cases and controls (p=0.989). However, the cervical cancer cases had significantly lower age at menarche (p=0.003) and first live birth (p<0.001), more smokers (p=0.002), premenopausal women (p=0.020), parity women (p=0.002), and women with a higher family history of cancers (p<0.001), compared with the control subjects. Among the 712 cases, 636 (89.3%) were squamous cell carcinoma; 50 (7.0%), adenocarcinoma; 7 (1.0%), adenosquamous carcinoma; and 13 (1.8%), others. Only 6 (0.8%) patients had CIN3; 186 (26.1%), stage I carcinoma; 374 (52.5%), stage II; 97 (13.6%), stage III; 8 (1.1%), stage IV; and 41 (5.8%), unknown.

Table 1.

Demographic and Selected Variables in Cervical Cancer Cases and Controls

Variables	Cases (n=712) n (%)	Controls (n=717) n (%)	p-value
Age, year (mean±SD)	54.61±12.77	54.62±11.76	0.989
Age at menarche, year (mean±SD)	15.99±2.02	16.30±1.91	0.003
Age at menopause, year (mean±SD)^a	48.84±3.80	49.19±4.00	0.192
Age at first live birth, year (mean±SD)^b	22.40±2.95	24.13±3.19	<0.001
Smoking status			0.002
Smoker	46 (6.5)	21 (2.9)
Nonsmoker	666 (93.5)	696 (97.1)
Menopausal status			0.020
Premenopausal	309 (43.4)	268 (37.4)
Postmenopausal	403 (56.6)	449 (62.6)
Parity			0.002
0–1	270 (37.9)	330 (46.0)
≥2	442 (62.1)	387 (54.0)
Family history of cancer			<0.001
No	504 (70.8)	587 (81.9)
Yes	208 (29.2)	130 (18.1)
Histological types
CIN3	6 (0.8)
Squamous cell carcinoma	636 (89.3)
Adenocarcinomas	50 (7.0)
Adenosquamous carcinoma	7 (1.0)
Others	13 (1.8)
Stage
CIN3	6 (0.8)
I	186 (26.1)
II	374 (52.5)
III	97 (13.6)
IV	8 (1.1)
Unknown	41 (5.8)

Information was available in postmenopausal women (399 cases and 436 controls).

Information was available in 705 cases and 684 controls with parity.

CIN, cervical intraepithelial neoplasia.

The genotype distributions of rs187084 in the cases and controls were described in Table 2. The observed genotype frequencies for the SNP in the controls were in consistence with the Hardy–Weinberg equilibrium (p=0.220). The logistic regression analyses showed that the rs187084 heterozygote TC was associated with a significantly increased cervical cancer risk (adjusted OR=1.28, 95% CI=1.01–1.62), compared with the TT genotype. Although the variant homozygote was associated with a nonsignificantly increased cervical cancer risk, the TC/CC genotypes contributed to the risk of cervical cancer in the dominant genetic model (adjusted OR=1.24, 95% CI=1.01–1.53).

Table 2.

TLR-9 -1486T>C Polymorphism and Cervical Cancer Risk

TLR9 -1486T>C	Cases n (%)	Controlsn (%)	OR (95% CI)	OR (95% CI) ^a	p^a
	694	715
TT	246 (35.45)	289 (40.42)	1.00	1.00
TC	346 (49.86)	319 (44.62)	1.27 (1.01, 1.60)	1.28 (1.01, 1.62)	0.037
CC	102 (14.70)	107 (14.97)	1.12 (0.81, 1.54)	1.15 (0.83, 1.60)	0.488
TC/CC	448 (64.55)	426 (59.58)	1.24 (1.00, 1.53)	1.24 (1.01, 1.53)	0.049

Adjusted by age at menarche, smoking status, menopausal status, family history of cancer, and parity.

OR, odds ratio; CI, confidence interval.

Further stratified analyses showed that the risk effect of rs187084 on cervical cancer was slightly more evident among subjects with an earlier age at menarche (adjusted OR=1.54, 95% CI=1.15–2.05), a positive family history of cancers (adjusted OR=1.65, 95% CI=1.03–2.63), and patients with an early cancer stage (adjusted OR=1.44, 95% CI=1.00–2.06). However, the heterogeneity test was only significant for the stratum of age at menarche (p=0.03) (Table 3). Thus, we conducted the interaction analysis and found a significant interaction between the genotypes of rs187084 and age at menarche, and rs187084 TC/CC carriers who had an earlier age at menarche were associated with a 1.54-fold increased risk for cervical cancer compared with TT carriers who had a later age at menarche (Table 4).

Table 3.

Stratified Analyses of TLR-9 -1486T>C (TC/CC vs. TT) Polymorphism and Cervical Cancer

Variables	Cases n	Controls n	OR (95% CI) ^a	p^b
Age				0.81
<55	136/235	159/224	1.31 (0.97, 1.77)
≥55	110/213	130/202	1.24 (0.89, 1.73)
Family history of cancer				0.19
Yes	69/136	58/71	1.65 (1.03, 2.63)
No	177/312	231/355	1.15 (0.89, 1.48)
Menopausal status				0.76
Premenopausal	116/187	112/155	1.21 (0.86, 1.71)
Postmenopausal	130/261	177/271	1.30 (0.97, 1.75)
Parity				0.94
0–1	93/172	131/198	1.24 (0.88, 1.74)
≥2	153/276	158/228	1.22 (0.91, 1.65)
Age at menarche				0.03
≤16	146/298	166/229	1.54 (1.15, 2.05)
>16	100/150	123/197	0.92 (0.65, 1.31)
Age at first live birth				0.61
<25	157/311	106/177	1.09 (0.79, 1.50)
≥25	86/133	172/227	1.23 (0.88, 1.74)
Stage
I	63/117	289/426	1.44 (1.00, 2.06)	0.66
II	136/229	289/426	1.17 (0.89, 1.52)
III/IV	36/67	289/426	1.24 (0.79, 1.94)
Histology
SC	219/402	289/426	1.26 (1.00, 1.58)	0.90
Other	20/34	289/426	1.21 (0.68, 2.15)

Adjusted by age at menarche, smoking status, menopausal status, family history of cancer, and parity.

p for heterogeneity test.

Table 4.

Interaction Between TLR-9 -1486T>C Polymorphism and Age at Menarche on Cervical Cancer

TLR9 -1486T>C	Age	Case n (%)	Control n (%)	OR (95% CI) ^a
TT	>16	100 (14.41)	123 (17.20)	1.00
TC/CC	>16	150 (21.61)	197 (27.55)	0.93 (0.66, 1.32)
TT	0–16	146 (21.04)	166 (23.22)	1.01 (0.70, 1.44)
TC/CC	0–16	298 (42.94)	229 (32.03)	1.54 (1.11, 2.13)
		p for interaction		0.03

Adjusted by smoking status, menopausal status, family history of cancer, and parity.

Discussion

To the best of our knowledge, this is the first study to provide evidence that the potentially functional SNP TLR9 T-1486C plays a role in cervical carcinogenesis and interacts with a known cervical cancer risk factor at an earlier age at menarche.

TLRs induce an inflammatory response by recognizing pathogens and play an important role in both innate and adaptive immune responses. TLR9 recognizes CpG motifs related to viral dsDNA (such as that from HPV and herpes simplex virus [HSV]), thereby resulting in the secretion of IFN-α and IL12 and the enhancement of the Th1 immune process, which are important for HPV clearance (Robinson and O'Garra, 2002; Latz et al., 2004). Further, it has been reported that HPV16 E6 and E7 oncoproteins interfere with the activation of the innate immune response via inhibiting TLR9 promoter activity and down-regulating TLR9 mRNA (Hasan et al., 2007), which might be an important mechanism for HPV escape from host immune surveillance. Similar findings were reported in HPV16-positive cancer-derived cell lines and primary cervical cancers, thus suggesting that variants of TLR9 promoter activity affect the function of HPV and development of cervical cancer (Hasan et al., 2007).

TLR9 rs187084 (T-1486C) is a potentially functional variant located in the promoter region, which is close to the region that interacts with HPV16 E6 and E7 oncoproteins (Hamann et al., 2006). Therefore, this SNP may regulate not only TLR9 basal transcript level (Ng et al., 2005; Tao et al., 2007), but also transcription during HR HPV infection. Several studies have investigated the effect of rs187084 (T-1486C) on human diseases, such as SLE (Hur et al., 2005; Ng et al., 2005; Tao et al., 2007), atherosclerosis (Hamann et al., 2006), severe malaria (Campino et al., 2009), Graves' ophthalmopathy (Liao et al., 2010), and rheumatoid arthritis (Jaen et al., 2009), but most of them showed no significant associations. In the current study, we found that rs187084 (T-1486C) affected susceptibility to cervical cancer in Chinese women, which needs to be validated by other larger studies. Further, we also observed a significant interaction between rs187084 (T-1486C) and earlier menarche age on the risk of cervical cancer, which may imply a role of earlier exposure of estrogen.

The genetic frequency of the variant allele varied among populations. For example, in the case–control study by Campino et al. (2009), the C allele frequency of TLR9 rs187084 in the African population was about 0.23–0.27, which was significantly different from those in European populations (C allele: 0.44) (Hamann et al., 2006) and our Chinese population (C allele: 0.37).

Some limitations in our study need to be addressed. First, inherent selection bias cannot be completely excluded, because our study is a hospital-based case–control study and the study subjects, particularly cancer cases, may not be representative for the target population. However, potential confounding might be minimized by matching the controls to the cases with regard to age and residential areas and by adjusting for some potential confounders in final data analyses. Second, only one SNP was genotyped in our study, and we cannot comprehensively investigate the association between SNPs of TLR9 and cervical cancer susceptibility. Finally, the sample number of subjects in our study is moderate, and the statistical power of the study is limited. Based on the sample size of this study (712 cases and 717 controls), we had 48.8% power to detect an OR as 1.24 for carriers of TC/CC genotypes in the dominant model (TC/CC vs. TT). We then applied FPRP to evaluate the noteworthiness of the association found in our study. Since the cumulative evidence supported the hypothesis that rs187084 (located in the promoter of TLR9 and created a putative Sp1 binding site) may be functionally relevant, a previous probability was set as 0.1. The results showed that the association for rs187084 variant genotypes (TC/CC vs. TT) and risk of cervical cancer was a noteworthy finding (<0.5) with the FPRP value being 0.295, assuming that the OR was 1.5. Therefore, other larger, preferably population-based studies are needed to validate our findings.

Footnotes

Acknowledgments

This work was supported in part by the Program for Changjiang Scholars and Innovative Research Team in University (IRT0631) and a Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions.

Disclosure Statement

No competing financial interests exist.

References

Beutler

2004. Inferences, questions and possibilities in Toll-like receptor signalling. Nature, 430:257–263.

Campino

, Forton

, Auburn

, Fry

, Diakite

, Richardson

et al. 2009. TLR9 polymorphisms in African populations: no association with severe malaria, but evidence of cis-variants acting on gene expression. Malar J, 8:44.

Chen

H.C.

, Schiffman

, Lin

C.Y.

, Pan

M.H.

, You

S.L.

, Chuang

L.C.

et al. 2011. Persistence of type-specific human papillomavirus infection and increased long-term risk of cervical cancer. J Natl Cancer Inst, 103:1387–1396.

Chen

, Han

, Wang

, Zhou

, Zhang

, Dong

et al. 2009. Interactions of IL-12A and IL-12B polymorphisms on the risk of cervical cancer in Chinese women. Clin Cancer Res, 15:400–405.

Chuang

T.H.

, Ulevitch

R.J.

2000. Cloning and characterization of a sub-family of human toll-like receptors: hTLR7, hTLR8 and hTLR9. Eur Cytokine Netw, 11:372–378.

Frazer

I.H.

, Leippe

D.M.

, Dunn

L.A.

, Liem

, Tindle

R.W.

, Fernando

G.J.

et al. 1995. Immunological responses in human papillomavirus 16 E6/E7-transgenic mice to E7 protein correlate with the presence of skin disease. Cancer Res, 55:2635–2639.

Hamann

, Glaeser

, Hamprecht

, Gross

, Gomma

, Schumann

R.R.

2006. Toll-like receptor (TLR)-9 promotor polymorphisms and atherosclerosis. Clin Chim Acta, 364:303–307.

Hasan

U.A.

, Bates

, Takeshita

, Biliato

, Accardi

, Bouvard

et al. 2007. TLR9 expression and function is abolished by the cervical cancer-associated human papillomavirus type 16. J Immunol, 178:3186–3197.

Hur

J.W.

, Shin

H.D.

, Park

B.L.

, Kim

L.H.

, Kim

S.Y.

, Bae

S.C.

2005. Association study of Toll-like receptor 9 gene polymorphism in Korean patients with systemic lupus erythematosus. Tissue Antigens, 65:266–270.

10.

Ito

, Ota

, Katsuyama

, Inoko

, Ohno

, Mizuki

2007. Lack of association of Toll-like receptor 9 gene polymorphism with Behcet's disease in Japanese patients. Tissue Antigens, 70:423–426.

11.

Jaen

, Petit-Teixeira

, Kirsten

, Ahnert

, Semerano

, Pierlot

et al. 2009. No evidence of major effects in several Toll-like receptor gene polymorphisms in rheumatoid arthritis. Arthritis Res Ther, 11:R5.

12.

Latz

, Schoenemeyer

, Visintin

, Fitzgerald

K.A.

, Monks

B.G.

, Knetter

C.F.

et al. 2004. TLR9 signals after translocating from the ER to CpG DNA in the lysosome. Nat Immunol, 5:190–198.

13.

Lazarus

, Klimecki

W.T.

, Raby

B.A.

, Vercelli

, Palmer

L.J.

, Kwiatkowski

D.J.

et al. 2003. Single-nucleotide polymorphisms in the Toll-like receptor 9 gene (TLR9): frequencies, pairwise linkage disequilibrium, and haplotypes in three U.S. ethnic groups and exploratory case-control disease association studies. Genomics, 81:85–91.

14.

Lazarus

, Vercelli

, Palmer

L.J.

, Klimecki

W.J.

, Silverman

E.K.

, Richter

et al. 2002. Single nucleotide polymorphisms in innate immunity genes: abundant variation and potential role in complex human disease. Immunol Rev, 190:9–25.

15.

Liao

, Chen

, Lin

, Liu

, Chen

, Tsai

et al. 2010. Toll-like receptor gene polymorphisms are associated with susceptibility to Graves' ophthalmopathy in Taiwan males. BMC Med Genet, 11:154.

16.

Mockenhaupt

F.P.

, Cramer

J.P.

, Hamann

, Stegemann

M.S.

, Eckert

, Oh

N.R.

et al. 2006. Toll-like receptor (TLR) polymorphisms in African children: Common TLR-4 variants predispose to severe malaria. Proc Natl Acad Sci USA, 103:177–182.

17.

Murad

Y.M.

, Clay

T.M.

, Lyerly

H.K.

, Morse

M.A.

2007. CPG-7909 (PF-3512676, ProMune): toll-like receptor-9 agonist in cancer therapy. Expert Opin Biol Ther, 7:1257–1266.

18.

Nardelli-Haefliger

, Lurati

, Wirthner

, Spertini

, Schiller

J.T.

, Lowy

D.R.

et al. 2005. Immune responses induced by lower airway mucosal immunisation with a human papillomavirus type 16 virus-like particle vaccine. Vaccine, 23:3634–3641.

19.

M.W.

, Lau

C.S.

, Chan

T.M.

, Wong

W.H.

, Lau

Y.L.

2005. Polymorphisms of the toll-like receptor 9 (TLR9) gene with systemic lupus erythematosus in Chinese. Rheumatology (Oxford), 44:1456–1457.

20.

Parkin

D.M.

, Bray

, Ferlay

, Pisani

2005. Global cancer statistics, 2002. CA Cancer J Clin, 55:74–108.

21.

Robinson

D.S.

, O'Garra

2002. Further checkpoints in Th1 development. Immunity, 16:755–758.

22.

Rozkova

, Novotna

, Pytlik

, Hochova

, Kozak

, Bartunkova

et al. 2010. Toll-like receptors on B-CLL cells: expression and functional consequences of their stimulation. Int J Cancer, 126:1132–1143.

23.

Takeda

, Kaisho

, Akira

2003. Toll-like receptors. Annu Rev Immunol, 21:335–376.

24.

Tao

, Fujii

, Tsukumo

, Maekawa

, Kishihara

, Kimoto

et al. 2007. Genetic variations of Toll-like receptor 9 predispose to systemic lupus erythematosus in Japanese population. Ann Rheum Dis, 66:905–909.

25.

Tindle

R.W.

2002. Immune evasion in human papillomavirus-associated cervical cancer. Nat Rev Cancer, 2:59–65.

26.

Wacholder

, Chanock

, Garcia-Closas

, El Ghormli

, Rothman

2004. Assessing the probability that a positive report is false: an approach for molecular epidemiology studies. J Natl Cancer Inst, 96:434–442.

27.

Yang

, Wheeler

C.M.

, Chen

, Uematsu

, Takeda

, Akira

et al. 2005. Papillomavirus capsid mutation to escape dendritic cell-dependent innate immunity in cervical cancer. J Virol, 79:6741–6750.

28.

zur Hausen

2002. Papillomaviruses and cancer: from basic studies to clinical application. Nat Rev Cancer, 2:342–350.