1. Introduction
Oncogenic human papillomavirus (HPV) is a major cause of cervical cancer. HPV 16 and HPV 18 are highly prevalent in all regions of the world [1,2], followed by HPV 31, 33. Although rarely found in western countries, HPV 52 and 58 are prevalent in Asian populations, especially in China [3]. Accumulated studies have shown that HPV 52 and 58 are relatively more prevalent among HPV-positive women from Asia than in other places with only 11.5%–28% of prevalence across the full spectrum of cervical neoplasia [3–5]. It has been estimated that worldwide HPV prevalence is 10.4% in women with normal cervical cytology [6], but less than 1% of the infected women develop malignant lesions [7]. It is clear that persistent infection with specific high-risk HPV types is a strong marker for progressive CIN disease [8]. Recent studies have revealed that persistence of high-risk HPV infection might be associated with virus intratype variants. The HPV intratype variants are defined as having nucleotide sequence variations no more than 2% in the coding region and 5% in the noncoding regions of the viral genome with respect to the prototype [9]. Concerning HPV intratype variants, the most extensive studies have been conducted on HPV16 and HPV18 [10,11]. However, few data about HPV 52 intratype variants has been achieved so far.
Amino acid changes may affect the transforming activity of the E6 and E7 oncoproteins; those in the L1 protein may affect the efficacy of viral infection or alter viral antigenicity [12,13]. Mutations in the long control region (LCR) may affect virus replication rates and transcriptional activity of E6 promoters [14,15]. In this study, the genetic variability of LCR, E6, E7, and L1 genomic regions of HPV 52 was analyzed. Studies on HPV 52 variants in strains from cervical disease patients may improve the understanding of the molecular mechanisms underlying disease progression and transformation.
2. Results
2.1. Characteristics of the Study Populations and Distribution of HPV 52
For the HPV types found from the study population, 8.8% (72/815) were solely infected by HPV 52. Samples with multiple HPV positives were excluded to avoid confounding the results. Analyses of the L1, E6, E7 and LCR genes were performed on all the single HPV 52 positive samples where sufficient materials were available. From the 72 samples, 60 sequences of L1 gene, 66 sequences of LCR, E6 and E7 were obtained, respectively. Tables 1 and 2 show the sites of variations and changes in amino acid sequences in L1, LCR, E6 and E7 genes, respectively, compared with the HPV 52 prototype. The histologic diagnoses of the women whose infected strains were sequenced successfully for the L1 gene were as follows: 11 normal cases (18.3%), 30 cervicitis (50.0%), 6 Ascus (10.0%), 3 CIN 1 (5%), 8 CIN 2 to 3 (13.3%) and 2 invasive cervical cancer (ICC) (3.33%) (Table 1); those for the LCR, E6 and E7 genes were as follows: 12 normal cases (18.2%), 34 Cervicitis (51.5%), 7 ASCUS (10.6%), 3 CIN 1 (4.55%), 8 CIN 2 to 3 (12.2%) and 2 ICC (3.03%) (Table 2).
2.2. Analysis of HPV 52 Genetic Variability
Sequence variations observed in HPV 52 clinical strains are summarized in Tables 1 and 2. Compared with the HPV 52 reference sequence (No.NC_001592), two strains were “L1-prototype-like,” and the remaining 58 strains were grouped into three different variants and named according to their frequency of variations as L1-LN-A, B and C (Table 1). The most prevalent HPV 52 variant in women from northeast China belongs to clade L1-LN-A (81.7%, 49/60). The statistical analysis by binary logistic regression showed that the variants of L1-LN-A, L1-LN-B and L1-LN-C were not associated with the severities of the infected women with the odds ratio (OR) of 0.444 (95% CI 0.094–2.093, p = 0.305), 1.741 (95% CI 0.162–18.675, p = 0.647) and 1.278 (95% CI 0.127–12.806, p = 0.835), respectively (Table 1).
Based on their variation rates among the four analyzed genes, the proportion of polymorphic nucleotides was significantly greater in the LCR than in that the other three genes: eight variation sites over 878 nt (0.91%) in HPV 52 LCR variants, compared with two variation sites (0.45%) over 447 nt in E6 variants, two variation sites (0.67%) over 300 nt in E7 variants and thirteen variation sites (0.82%) over 1590 nt in L1 variants. Deletions were found only in LCR. Based on the variations of E6, E7 genes and LCR, the strains were clarified into four groups (Table 2). The statistical analysis by binary logistic regression showed no association between the first two groups and the severities of the infected women with the odds ratio (OR) of 1.626(95% CI 0.380–6.957, p = 0.512) and 0.714 (95% CI 0.166–3.066, p = 0.651), respectively.
2.3. HPV 52 L1 Sequence Variations
Analyses of the complete sequences of L1 gene revealed 13 different nucleotide mutations. Except for C6917A, all the other mutations in HPV 52 L1 gene were not reported previously. Among them, three (23.1%) novel mutations of C5640T, A5641T and G5642A, which led to the Q26L amino acid change, were identified in nine strains (Table 1). The Q26L amino acid mutation was located in the strand β-B1 of L1 protein. All the others were synonymous mutations, including A5571G, T5972C, G6111A, G6218A, T6701G, T6764G, A6794G, C6824T, C6917A and G7802A. Furthermore, two nonsynonymous mutations of A5641T and G5642A were covariations. The localizations of HPV 52 L1 protein mutations are reported in Table 1.
2.4. HPV 52 E6 and E7 Sequence Variations
The complete E6 and E7 open reading frames were analyzed in strains from 66 patients. Compared to the reference sequence, all the obtained sequences harbored the nucleotide variation of G350A and 65 had the nonsynonymous mutation of A379G (K93R) in the E6 gene (Table 2). The A379G (K93R) mutation was located in the strand H1 and the third predicted zinc finger of E6 protein. Analysis of the complete E7 open reading frame showed two synonymous substitutions of C751T and A801G.
2.5. HPV 52 LCR Sequence Variations
Compared to the reference sequence, LCR sequences had eight nucleotide mutations (Table 2). Among them, the CTT 7681–7683 deletion was a novel mutation found in all of the studied strains. Besides the CTT 7681–7683 deletion, another three mutations of G7622A, T7624G/C and G7861A were present in all obtained sequences, too. The other frequent polymorphic sites included A7657C, T7659C, G7712C and A7865G, that were detected in more than 64 strains. One covariation pattern and a statistically significant association were found among A7657C, T7659C, G7712C and A7865G mutations. (p = 0.01, phy = 1 for the associations).
All of the eight point mutations identified in LCR encompassed, and thus potentially affected, the proposed binding sites for transcription factors. In particular, the nucleotide substitutions of G7622A and T7624G/C were located in the TATA binding site, as well as C/EBP and SRY binding sites. The CTT7681–7683 deletion was located in the NF-E2 and AP-1 binding sites; the nucleotide substitution of G7861A was located in the Oct-1 binding site. The variations of A7657C, T7659C, G7712C and A7865G, which were predicted at the binding sites for AP-1, HSF, MAT alp, Skn-1 and HFS, introduced additional putative binding sites for cellular proteins, such as Cap, Cdxa, Skn-1, and Oct-1 or HSF. Putative binding sites for cellular proteins SRY, and others, were also affected by less frequently encountered variations (data not shown). However, the binding sites for the HPV E2 protein were conserved in all the strains.
3. Discussion
Globally, the most prevalent HPV type is HPV-16, detected in approximately 40% of high-grade cervical lesions [17], and followed by HPV 18. The two viruses are responsible for about 70.1% of cervical cancers in the world. Therefore, research efforts have focused mainly on these two viral types and a prophylactic vaccine is available currently for prevention of HPV 16 and 18 infections [18,19].
However, the prevalence of other high-risk HPV types varies among different countries and even throughout regions of the same country [20]. In recent years, it has been reported that HPV 52 ranks fourth in cervical cancer cases within some Asian counties [17], and the fourth most prevalent types in patients with high-grade cervical lesions in Northern China [5].
Researches on genetic variability of HPV variants may increase the understanding of the molecular mechanisms underlying disease progression and transformation. Indeed, a number of studies suggest that variants of the same HPV type are biologically distinct and may confer differential pathogenic risks [21]. In this study, the genetic variability of LCR, L1, E6 and E7 genes of HPV 52 was analyzed. Compared to the published data [22–26], several novel mutations were discovered either in the LCR and the coding regions of HPV type 52.
In a recent study, an increase prevalence of CIN3 was reported to associate with HPV 52 lineage C [26]. Results of another study showed that all cases of CIN3 or worse were associated with a combined group of lineages A and B and C, but not with the group of lineage D, resulting in an odds ratio (OR) estimate of infinity [27]. Furthermore, a study found that the nonprototypic LCR variant was associated with the persistence of HPV 52 infection compared to the prototype [28]. In our study populations, HPV 52 showed four main variants inferred on L1 sequences. The most prevalent HPV 52 variant was L1-LN-A. However, no association between the variants and the grade of cervical lesion of infected patients was found. Whether or not the different findings are the result of the variants found in different geographical areas still needs to be studied.
Except for C6917A, all the other mutations in the HPV 52 L1 gene were not described elsewhere [25]. The novel mutations of C5640T, A5641T and G5642A, which lead to amino acid change of Q26L, were detected in nine strains from patients with HSIL. The Q26L mutation was located in the strand β-B1 of the L1 protein, which may suggest that the variants were established to escape neutralization. This mutation may influence the folding of the L1 protein, with possible consequences on the immunogenicity of neutralizing epitopes, thus favoring persistence of the infection by viral evasion from neutralizing antibody responses [12,29].
HPV E6 and E7 genes are believed to be the main oncoproteins. In this study, the E6 and E7 gene sequences were relatively conserved, for only four mutations were identified in most of the studied HPV 52 strains, and no significant trends accordant with severity of cervical neoplasia were observed. Among the four mutations, only A379G (K93R) was nonsynonymous. The mutation G350A detected in the E6 gene has been reported as G350T in the study of Ding et al.[30]. In accordance with Xin’s report [22], the unique A379G (K93R) variation was located in the strand H1 and the third predicted zinc finger of E6 protein and occurred in almost all of the HPV 52 positive samples.
Analysis of HPV 52 LCR revealed one covariation pattern. Notably, the novel CTT7681–7683 deletion is firstly reported being detected in all the studied strains in Northern China. The biological meaning and geographical areas of such finding remains to be established. In addition, some of the mutations identified in LCR encompassed the proposed binding sites for YY1 and SRY transcription factors. It is suspected that these mutations may affect the transcription of the downstream E6 gene, though it remains to be determined by in vitro studies. Among HPV 52 isolates, the LCR T7624G/C substitution was detected in all of the samples tested. Interestingly, this mutation was suggested to be associated with the persistence of HPV 52 infection by Aho et al. [28].
The genetic variations of HPV 52 were first documented in this report in Northeast China. Additionally, the frequency distributions of HPV 52 variants in Northeast China were different from those reported in European and American populations.
4. Materials and Method
4.1. Study Populations
Cervical swabs of the studied populations were collected from patients who referred to the Department of Obstetrics and Gynecology, Shengjing Hospital of China Medical University for routine gynecological detections. The median age of the studied populations was 38 years (range, 20–66 years) at the time, the cervical scrapings were obtained. After giving informed consents, a total of 815 patients without a history of hysterectomy received an examination with the papanicolaou (Pap) smear, which collected cervical cells using ViraPap kits (Digene Diagnostic, Silver Spring, MD, USA) at study entry. Pap smears were graded according to the Bethesda system, which categorizes cytology grades into normal, atypical squamous cells of undetermined significance (ASCUS), low grade squamous intraepithelial lesion (LSIL), high grade cervical squamous intraepithelial lesion (HSIL) and cancer. Patients with LSIL or worse lesions were referred for a biopsy and treatment. Histological data were reviewed by expert pathologists to confirm the disease outcomes. Subjects gave a signed informed consent. The study protocol was approved by institutional ethical and research review boards of the participating institutions in the northeast of China.
4.2. HPV DNA Detection and Typing
Viral types were detected using the HPV GenoArray test kit (Hybribio Ltd, Hong Kong) according to the manufacturer’s instructions as described in the previous study [5]. In this method, flow-through hybridization was used based on the principle of Reverse Dot Blot Assay which was described in previous studies [16]. The method could classify 21 HPV subtypes including 12 established high-risk (HR) types (HPV-16, −18, −31, −33, −35, −39, −45, −51, −52, −56, −58, −59), three probable HR types (HPV-53, −66, −68) and six established low-risk (LR) types (HPV-6, −11, −42, −43, −44, and CP8304).
4.3. PCR Amplification
Samples positive with HPV 52 were further amplified with primer sets specifically designed for the regions containing the partial LCR fragment and complete E6, E7 and L1 ORF, respectively. PCR reactions were done in a 50 μL reaction volume containing 1× PCR buffer, 200 μM of each dNTP, 2 mM MgCl2, 20 pmol of each primer and 1 unit of Taq DNA polymerase (Takara, Japan). PCR amplicons were separated on 2% agarose gels and visualized by ethidium bromide staining under UV transillumination. In the case of no observed band on the gel, 2 μL PCR products obtained with outer primer pairs were used as a template for amplification with inner primer pairs. The primer sets and PCR amplification profiles are shown in the supplementary Table S1. A reaction mixture without template DNA was included as a negative control in every set of PCR.
4.4. HPV DNA Sequence Analysis
PCR amplicons were purified and applied to enzymatic extension reactions for DNA sequencing using the ABI PRISM Big-Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Bedford, MA, USA). Both strands of the recovered DNA were sequenced with the same forward and reverse primers as those used for PCR amplification. The sequencing reactions were run on an ABI 3730 XL DNA Analyzer (Applied Biosystems). The obtained HPV 52 sequences were aligned with the reference sequence (Genbank Accession No.NC_001592) on NCBI [31]. The multiple alignments were further refined by manual intervention.
4.5. Analysis of Transcription Factors Binding Sites in <italic>LCR</italic> Region
Potential binding sites for cellular and viral transcriptional factors within the HPV 52 LCR regions were searched by the TFSEARCH software, including sites for AP-1, E2, GRE, NF-1, Oct-1, TATA, YY1, C/EBP, Sp1, SRY, AML-1a and c-Myc/c-Max. Cut-off values and coincidence levels between consensus binding sites and the LCR sequence of HPV 52 type were adjusted in order to minimize both the number of negative and positive faults [32].
4.6. Statistical Analysis
Statistical analysis was performed using the SPSS software (version 17.0) [33]. The magnitude of the associations between HPV variants and HSILs of patients was assessed by binary logistic regression with odds ratios (ORs) and respective 95% confidence intervals (CIs). For examining distributions of HPV 52 variations with respect to disease severity, spearman correlation was employed. Two-sided, p < 0.05 was considered to be statistically significant.