|Home | About | Journals | Submit | Contact Us | Français|
We describe here a rapid, high-throughput genotyping procedure that allows the simultaneous detection of 16 high- and low-risk genital human papillomavirus (HPV) types by multiplex PCR in a single reaction tube. Multiplex PCR is based on the amplification of HPV DNA by sets of HPV genotype-specific primers, and the genotypes of HPV are visually identified by the sizes of amplicons after they are separated by capillary electrophoresis. The procedure does not include a hybridization step with HPV-specific probes and is rapid and labor-saving. We detected all 16 HPV genotypes (types 16, 58, 52, 51, 56, 31, 18, 39, 66, 59, 6, 33, 30, 35, 45, and 11) with a high sensitivity and a high degree of reproducibility. By using this newly developed method, we conducted a pilot study to examine the correlation between the prevalence and genotype distributions of HPV and the cytological group classifications for 547 cervical samples. Compared with the group of samples considered normal (14.7%), there was a significant increase in the prevalence of HPV in women with atypical squamous cells of unknown significance (61.3%), low-grade intraepithelial lesions (75.8%), and high-grade intraepithelial lesions (HSILs) (82.2%). The prevalence and distribution of type 58 were correlated with cytological malignancies, with the highest prevalence in women with HSILs. In conclusion, the novel multiplex PCR method described appears to be highly suitable not only for the screening of cervical cancer precursor lesions but also for the characterization of genotype distributions in large-scale epidemiological studies and HPV vaccination trials.
Accumulating evidence indicates that persistent infection with high-risk human papillomaviruses (HPVs) is indeed a major causative factor in the development of cervical intraepithelial neoplasia and invasive cervical carcinoma (42, 10, 8, 5, 40, 27, 11, 30). The HPV family includes over 100 genotypes, 30 to 40 of which are mucosotropic, and at least 15 types of the mucosotropic HPVs have been linked to cervical cancer (5, 8, 10, 42). In addition, some of these types are also related to other cancers of the genital tract (21, 22) and to cancers of other organs (14, 28). Light microscopic examination of a Papanicolaou (Pap)-stained smear is of primary importance for the detection of cervical cancer precursor lesions. It has been demonstrated that concomitant testing for DNA of the high-risk HPV types by the Pap test can clearly identify women at high risk for cervical cancer, particularly if persistent infection by high-risk HPV types is diagnosed (5, 8, 10, 11, 27, 30, 40). Furthermore, HPV genotyping is of critical importance for the investigation of the clinical behavior and the epidemiology of HPV infection, for population studies for HPV vaccination trials, and for monitoring of the efficacy of HPV vaccines. Several genotyping methods have been developed in order to identify high-risk HPV in liquid-based cytology (LBC) samples and tissue samples (1, 12, 34). The molecular techniques that have been applied for HPV DNA detection (20) include direct probe methods with Southern blotting and in situ hybridization, signal amplification methods such as the hybrid capture II assay (29), and target amplification methods by PCR (8, 20). For the genotyping of HPV, the target products amplified by PCR are subjected to sequence analysis (2), restriction fragment length polymorphism analysis (RFLP) analysis (4, 33, 41), and hybridization with type-specific probes (17, 23, 26). Reverse line blot assays have also been developed and validated (15).
In the present study, we have developed a rapid, highly sensitive, multiplex PCR-based HPV genotyping assay that allows 16 genotype-specific primer amplifications in a single step in a single reaction tube. Using this newly developed assay, we investigated the correlation between cytological groups and the prevalence and genotype distributions of HPV in clinical samples.
Cervical samples collected in LBC medium were obtained from Japanese women seeking routine gynecologic care at outpatient clinics. The average age and age distribution of the individuals in each cytological group are shown in Table Table1.1. Cervical cells were collected from the endo- and exocervices with a cytobrush and were stored in either 10 ml of cytoRich (TriPath Imaging, Burlington, NC) or 18 ml of preserveCyt (Cytic Corporation, Boxborough, MA) preservative medium. Pap-stained LBC specimens were screened by cytotechnologists, and suspicious cases were diagnosed by pathologists at the GLab Pathology Center Co., Ltd. (Sapporo, Japan). Smears were classified according to the Bethesda system 2001 (35) into normal cells (normal), atypical squamous cells of unknown significance (ASCUS), and cells with low- and high-grade intraepithelial lesions (LSILs and HSILs, respectively). After preparation of the cytological specimens, the residual cells in the preservative medium were stored at 4°C until DNA extraction. The cytological diagnosis was made without knowledge of the HPV genotyping results. Among 547 clinical samples, 292 samples (53.4%) were considered normal, 62 (11.3%) were considered to have ASCUS, 120 (21.9%) were considered to have LSILs, and 73 (13.4%) were considered to have HSILs.
The residual cells in the preservative medium were centrifuged at 13,000 × g for 15 min, and the cell pellets were resuspended in 50 ml of phosphate-buffered saline and centrifuged again at 13,000 × g for 15 min. The washed cell pellets were added to a DNA purification column (Generation capture column kit; Gentra Systems, Inc. Minneapolis, MN). The extraction of total DNA was performed according to the manufacturer's instructions. The concentration and the quality of the extracted DNA were determined by spectrophotometry. Five microliters of DNA solution was used as a template for PCR amplification.
The HPV sequences of different isolates and complete genomes were obtained from GenBank and were used for primer design.
The genome sequences of all HPV genotypes were obtained through the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/) and the Los Alamos National Laboratory HPV Database (http://hpv-web.lanl.gov/stdgen/virus/hpv/) and were aligned by using Genetyx software (version 7.03; Genetyx Corporation, Tokyo, Japan). HPV genotype-specific primers were designed on the basis of the multiple-sequence alignments, as described in the Results. The dissociation temperature and the ability to form duplex structures were determined for each primer sequence by using Primer3 software (http://fokker.wi.mit.edu/primer3/).
Sixteen HPV genotypes (genotypes 6, 11, 16, 18, 30, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 66) were detected by multiplex PCR in a single tube. PCRs were performed with a multiplex PCR kit (Qiagen Inc., Valencia, CA), according to the manufacturer's instructions, with minor modifications. The details are described in the Results.
The data were statistically analyzed by Fisher's exact test with “R” free statistics software (version 2.4.1; http://www.r-project.org/). A P value of <0.05 was considered to indicate statistical significance.
To study the associations among HPV infections, genotype distributions, multiple infections, and cervical cancer precursor lesions, we designed 17 sets of genotype-specific primers that could amplify specific regions of 16 types of HPV DNA. There primers were designed to fulfill several criteria. First, the 3′ regions of the primer sequences should contain as many base pair mismatches as possible among the primer sets to ensure genotype-specific amplification. Second, the melting temperatures of the primers should be higher than 65°C to increase the specificity of PCR amplification. Third, the GC contents of the primers were basically designed to be between 45 and 60%, with a few exceptions. Fourth, the size of each amplification product (amplicon) should differ by more than 20 bp to facilitate identification by capillary electrophoresis. Fifth, the primers should not anneal to human genomic DNA.
Table Table22 shows the primer sets that fulfilled the criteria outlined above and that were used to achieve the type-specific amplification of 16 HPV genotypes. All primer sequences were used to search the entire HPV genomes for genotype-specific sequences, and their target specificities were validated. One additional primer set was designed for HPV type 16 to ensure its identification, since infection with HPV type 16 is known to confer the highest risk for the development of cervical cancer.
Multiplex PCR was performed with a final volume of 50 μl. Each PCR mixture contained 25 μl of 2× Qiagen multiplex PCR master mixture, 5 μl of template DNA, 5 μl of distilled water, and 5 nM of total primer sets. Amplifications were performed with the following cycling profiling: Qiagen Taq DNA polymerase activation was performed by incubation at 95°C for 15 min, followed by 40 cycles of denaturation for 30 s at 94°C, 1.5 min of annealing at 70°C, and 1 min of elongation at 72°C. The last cycle was followed by a final extension step of 2 min at 72°C. Amplification was performed in a GeneAmp PCR system 9700 apparatus (Applied Biosystems, Foster City, CA). Aminolevulinate delta-synthase 1 (GenBank accession no. NM 000688) was amplified as an internal positive control for the multiplex PCR. As an external positive control for the multiplex PCR, DNA fragments derived from Brevundimonas diminuta cloned to plasmid pCRII-TOPO (Invitrogen Corporation) were added to the PCR mixtures along with external plasmid-specific primers. Two microliters of the amplicons was analyzed by electrophoresis on 1.5% agarose gels and ethidium bromide staining. The amplified DNA bands were excised, extracted, and purified by use of a QIAquick gel extraction kit (Qiagen). The purified DNA was then sequenced with a CEQ DTCS Quick Start Kit on a CEQ8000 (Beckman Coulter, Inc.). The identity between the amplicon obtained and an HPV genotype sequence in the GenBank database was confirmed by BLAST analysis (http://www.ncbi.nlm.nih.gov/BLAST). For the standard rapid screening test for HPV genotypes, the final amplicons were separated and analyzed with a 2100 bioanalyzer and a DNA 500 LabChip kit (Agilent Technologies Inc., Palo Alto, CA). High-resolution and high-sensitivity analyses were performed with the 2100 bioanalyzer. Gel-like images were trimmed and analyzed with Photoshop software (version 6.0; Adobe Systems, San Jose, CA). The HPV genotypes in the samples were identified by the sizes of the amplicons. HPV type 16 produced two separate bands (Fig. (Fig.1).1). HPV type 58 produced a single major band and a nearby minor band. Both bands were found to be specific for HPV type 58 by sequence analysis. All the other HPV types produced a single band, each of a characteristic size.
As shown in Fig. Fig.1,1, the HPV genotypes were visually determined by the detection of two separate bands for each of types 16 and 58 and a single band for all the other types. Multiple infections were identified by the occurrence of multiple bands of the characteristic size. Detection limits were determined by serial dilution of plasmid mixtures, followed by PCR (Fig. (Fig.2).2). For all 16 genotypes, HPV-specific bands were detectable when more than 102 copies of the plasmid mixtures were present.
The prevalence of HPV infection in each cytological group is shown in Table Table2.2. Among the samples considered normal, 43 of 292 samples (14.7%) were HPV positive. The prevalence of HPV infection was significantly increased in women with cytological abnormalities: 61.3% in women with ASCUS, 75.8% in women with LSILs, and 82.2% in women with HSILs. Thus, the prevalence of HPV infection was well correlated with the cytological abnormalities. Compared with the group considered normal (3.4%), multiple infection rates were also significantly increased in women with ASCUS (27.4%), LSILs (20.0%), and HSILs (32.9%). Thus, the prevalence of multiple infections was also positively correlated with cytological abnormalities. Notably, ASCUS showed the second highest frequency. This is probably because samples with koilocytes are usually classified as ASCUS.
The HPV genotype distributions in each cytological group are shown in Table Table3.3. In the group considered normal, HPV type 16 showed the highest prevalence (30.2%), followed by type 52 (20.9%); type 51 (16.3%); types 58 and 18 (14%); types 66 and 59 (7%); type 31 (9.3%); and types 56, 39, 6, 33, and 45 (2.3%). Types 30, 35, and 11 were not detected. In women with ASCUS, HPV type 16 showed the highest prevalence (36.8%), followed by type 58 (28.9%); type 31 (18.4%); types 39 and 66 (15.8%); type 51 (13.2%); types 52 and 56 (10.5%); types 18 and 6 (7.9%); type 59 (5.3%); and types 33, 35, and 45 (2.6%). Note that the prevalence of HPV type 58 in women with ASCUS rose to 28.9%, more than twice as high as that in the normal group (14%). This tendency became more apparent in women with LSILs and HSILs, with the highest prevalences of 27.5 and 38.3%, respectively. In women with LSILs, type 56 was the second frequent genotype detected (24.2%). It was followed in decreasing order of frequency by types 16 and 52 (17.6%), type 51 (11%), type 31 (9.9%), type 18 (8.8%), type 39 (6.6%), types 59 and 66 (5.5%), type 30 (2.2%), and types 6 and 35 (1.1%). Types 33, 45, and 11 were not detected. In women with HSILs, type 58 (38.3%) was followed by type 16 (35.0%); type 52 (23.3%); type 51 (16.7%); type 31 (11.7%); type 18 (8.3%); type 33 (5.0%); type 30 (3.3%); and types 56, 39, 6, and 35 (1.7%). Types 66, 59, 45, and 11 were not detected. It should be noted that the prevalence of type 58, which is comparable to that of type 16, is high in the groups with abnormal cytologies, especially in women with HSILs, which may suggest that type 58 is one of the major high-risk HPV types in Japan. In women with LSILs, type 56 was the second most frequent virus (24.2%). However, since the frequency of type 56 in women with HSILs declined abruptly to 1.7%, type 56 might be a low-risk passenger virus.
Multiple infections, that is, infections with more than two types of HPV in a single sample, were rather common; and as many as five types of HPV were simultaneously detected in a single sample (Fig. (Fig.3).3). Among a total of 232 HPV-positive samples, 75 were shown to be multiple infections (Table (Table4).4). The remaining 157 samples were infected by a single type of HPV (67.4%). Among the 75 samples with multiple infections, 54 were infected by two types of HPV (72.0%), followed by 14 samples with three types (18.7%), 5 samples with four types (6.7%), and 2 samples with five types (2.7%).
We next examined whether certain combinations of HPV genotypes could influence the risk of development of cervical lesions. The occurrences of multiple infections involving more than two of the four high-risk HPV genotypes, types 16, 51, 52, and 58 (Table (Table4),4), were 40.0% in the normal group, 47.0% in women with ASCUS, 37.5% in women with LSILs, and 58.3% in women with HSILs. Although the prevalence rate was the highest in women with HSILs, there were no significant correlations between the frequency of multiple infections and the progression of cytological abnormalities. Further follow-up studies with a larger number of clinical samples will be needed to clarify the relationship between genotype combinations and the development of cervical lesions.
The detection of cervical cancer precursor lesions by the use of Pap-stained cytological smears, followed by establishment of a pathological diagnosis with biopsy specimens, is of primary importance for the identification of women at risk for cervical cancer. Since persistent infection (3, 7, 18, 19, 27) and chromosomal integration (24) with high-risk HPV types have been demonstrated to be major causative factors in the development of cervical cancer, determination of the HPV types present in a cervical smear is important in evaluating a woman's risk of developing cervical cancer. Moreover, HPV genotyping is of critical importance for characterizing the population in HPV vaccination trials and for monitoring the efficacy of HPV vaccines. There are essentially three types of HPV DNA detection techniques: direct nucleic acid probe methods, hybridization signal amplification methods, and target amplification methods (8, 20). Direct nucleic acid probe methods use Southern blotting and in situ hybridization; they are time-consuming and require relatively large amounts of highly purified DNA (20). Among the hybridization signal amplification methods, the hybrid capture II assay (Digene Corporation, Silver Spring, MD) is widely used; it is an antibody capture/solution hybridization/signal amplification assay that uses chemiluminescence to qualitatively detect the presence of high- and low-risk HPV types (20). The hybrid capture II test provides an excellent tool for the triage of patients with minor cytological abnormalities on Pap tests but cannot identify specific HPV genotypes. PCR-based target amplification methods are the most sensitive and flexible among all DNA analysis techniques. The amplification of target DNA is achieved with either general HPV-specific primers (13, 16, 25, 31, 33, 41) or genotype-specific primers (38, 39). In the former method, the detection of individual genotypes can be accomplished by several methods, including RFLP analysis, cycle sequencing and assignment of genotypes by sequence comparison (8, 20), and reverse hybridization assays. RFLP analysis and direct sequencing may sometimes lead to incorrect conclusions, especially when infections with multiple types are encountered. Since a single band separated by electrophoresis may contain multiple PCR products of the same size or of similar sizes, the results of direct sequencing of the PCR products are difficult to interpret. Subcloning is then needed before cycle sequencing. On the other hand, reverse hybridization assays can detect and classify infections with single and/or multiple HPV genotypes with a high sensitivity (15). However, the reverse hybridization method is rather laborious. Although type-specific PCR primer sets allow the identification of individual genotypes of HPV, the need to perform multiple and parallel amplifications for each sample imposes severe constraints on throughputs. Recently, two independent multiplex PCR methods have been reported and allow the detection and genotyping of mucosotropic HPV, including infections with multiple genotypes (6, 36). These simple PCR-based methods are capable of detecting multiple HPV genotypes in a single reaction tube on the basis of the product size. However, similar to the type-specific PCR, both methods need nested PCR and require the performance of multiple and parallel amplifications for each sample.
In the present study, we have developed a highly sensitive, low-cost screening assay that allows the rapid and specific detection of 16 high-risk and low-risk HPV genotypes in conjunction with LBC. Since HPV genotype 16 is known to have many variant forms, two primer sets were designed for two different regions of type 16 in order to avoid false-negative results. The method developed here is simple, rapid, and reliable: it does not require probe hybridization, and the whole procedure can be completed in approximately 6 h (1 h for DNA extraction, 3 h for PCR, 1 h for electrophoresis, and 1 h for final genotyping). Moreover, this screening system enables one to conduct both Pap-stained cytology and HPV genotyping simultaneously with a single LBC sample (34). The identification of high-risk HPV genotypes in abnormal smears is important for the identification of patients who are at increased risk for cervical cancer. Thus, our system may also contribute to the reduction of unnecessary referrals to colposcopy clinics and help to reduce health care costs.
Nationwide surveys for the prevalence and genotype distributions of HPV have not yet been fully conducted in Japan. Miura et al. reviewed and summarized 14 Japanese studies that reported on the prevalence and genotype distributions of HPV in clinical samples with LSILs, HSILs, cervical intraepithelial neoplasia, or invasive cervical cancer (32). The prevalence of HPV was reported to be 10.2% in the normal group, 79.4% in women with LSILs, 89.0% women with in HSILs, and 87.4% in women with invasive cervical cancer. In that meta-analysis, the HPV genotype distributions in the normal, LSIL, and HSIL groups were also summarized. The prevalence of HPV type 16 was 7.2% in the normal group, 13.8% in the LSIL group, and 34.3% in the HSIL group. The prevalence of type 52 was 12.1% in the normal group, 11.1% in the LSIL group, and 15.0% in the HSIL group. The prevalence of type 58 was 3.4% in the normal group, 5.8% in the LSIL group, and 6.7% in the HSIL group. The prevalence of type 18 was 4.0% in the normal group, 4.0% in the LSIL group, and 4.7% in the HSIL group. The prevalence of type 56 was 4.8% in the normal group, 9.0% in the LSIL group, and 2.8% in the HSIL group. In our study, the overall prevalence of HPV in the normal group was 14%, which is slightly higher than the data (10.2%) compiled by Miura et al. (32). Except for this difference, the prevalence of HPV and the genotype distributions in our study are generally in agreement with the data of Miura et al. (32), in which types 16, 52, and 58 have been identified as the major HPV types involved with HSILs. Like Miura et al. (32), we also found that the prevalence of type 56 is high in women with LSILs and declines in women with HSILs. In our study, type 58 was the most frequently found HPV type in women with LSILs and HSILs. This may be correlated to the findings that the prevalence of type 58 in women with LSILs is second to the prevalence of type 16 and is apparently higher than the prevalence of other genotypes except type 16 in Asia compared with the prevalence of other genotypes in North America, Europe, South and Central America, and Africa (9).
Recently, Trottier et al. reported in a 4-year follow-up study of 2,462 Brazilian women that infections with multiple types appear to act synergistically in cervical carcinogenesis and the risk of progressing to HSIL increases markedly as the number of infecting HPV types increases (37). A synergistic incremental risk was clearly shown for infections with multiple types, including type 58 as well as type 16 with other types of HPV (37). Our study showed a positive correlation between infections with multiple HPV types and cytological malignancy. However, we could not draw any definitive conclusions concerning the synergistic roles of particular genotype combinations in the development of cervical lesions. Further follow-up studies with a larger number of clinical samples are required to address this issue.
Lastly, since vaccine trials for HPV types 16 and 18 are now in progress in Japan, nationwide surveys for the prevalence and genotype distributions of HPV are urgently required. The high-throughput HPV genotyping method described here may contribute greatly to such nationwide surveys and may be useful for the population studies required for HPV vaccination trials and for the monitoring of the efficacy of HPV vaccines.
Published ahead of print on 30 January 2008.