Genotyping human papillomavirus (HPV) in clinical specimens is important because each HPV type has different oncogenic potential. Amplification of HPV DNA by PCR with the consensus primers that are derived from the consensus sequences of the L1 gene has been used widely for the genotyping. As recent studies have shown that the cervical specimens often contain HPV of multiple types, it is necessary to confirm whether the PCR with the consensus primers amplifies multiple types of HPV DNA without bias. We amplified HPV DNA in the test samples by PCR with three commonly used consensus primer pairs (L1C1/L1C2+C2M, MY09/11, and GP5+/6+), and the resultant amplicons were identified by hybridization with type-specific probes on a nylon membrane. L1C1/L1C2+C2M showed a higher sensitivity than the other primers, as defined by the ability to detect HPV DNA, on test samples containing serially diluted one of HPV16, 18, 51, 52, and 58 plasmids. L1C1/L1C2+C2M failed to amplify HPV16 in the mixed test samples containing HPV16, and either 18 or 51. The three consensus primers frequently caused incorrect genotyping in the selected clinical specimens containing HPV16 and one or two of HPV18, 31, 51, 52, and 58. The data indicate that PCR with consensus primers is not suitable for genotyping HPV in specimens containing multiple HPV types, and suggest that the genotyping data obtained by such a method should be carefully interpreted. (Cancer Sci 2011; 102: 1223–1227)
Human papillomavirus (HPV), composed of an icosahedral capsid and a circular double-stranded DNA genome, is classified into more than 100 genotypes based on the nucleotide sequence homology of the L1 gene encoding the major capsid protein.(1) The HPV types found in lesions of the skin and genital mucosa are grouped as cutaneous and genital HPVs, respectively. Of genital HPVs, 15 types (HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, 68, and 73) that have been found in cervical cancer are called high-risk HPVs(2) and the types, such as HPV6 and HPV11, that have been found in benign genital warts are called low-risk HPVs.(3)
For detection and genotyping of HPV DNA in the clinical specimens, such as cervical swabs and Pap smears, a part of the L1 gene is amplified by PCR then grouped based on the susceptibility to various restriction enzymes, the binding capacity to type-specific probes, or the nucleotide sequences of the amplicons.(4) Several consensus primer pairs have been developed and used as standard primers for PCR-based genotyping of HPV in the clinical specimens. L1C1/L1C2+C2M was developed in 1991,(5) and has been used in more than 10 articles describing HPV prevalence in the Japanese population.(5–16) MY09/11(17) and GP5+/6+(18) were developed in 1989 and 1995, respectively, and have been used in numerous studies worldwide.(19) These primers are derived from the consensus sequences of the L1 gene and the amplicons contain type-specific sequences. Recently new primers, PGMY09/11(20) and modified GP5+/6+ (MGP),(21) which are composed of several type-specific primers, were developed to improve the accuracy of HPV genotyping. The World Health Organization HPV Laboratory Network, which was founded to improve the quality of laboratory services for effective surveillance and monitoring of HPV vaccination impact, recommends PCR with PGMY09/11 followed by reverse blotting hybridization with type-specific probes, as a standard procedure for HPV genotyping.(22)
Recent studies showed that many HPV-positive women are infected with multiple genotypes.(23,24) Therefore, the methods capable of detecting and genotyping HPV DNA of multiple types in a single clinical specimen are necessary to know the precise prevalence of HPV types and the impact of HPV vaccines. As PCR does not always amplify different DNA fragments with equal efficiency, we examined whether PCR with consensus primers can amplify HPV DNA of multiple genotypes in a single sample without bias. We found that PCR with consensus primers sometimes results in mistyping.
Materials and Methods
Plasmids. The pUC plasmid containing the complete genome of HPV16, 18, 31, 51, or 52, or the L1 gene of HPV58 was used. Purified plasmids were quantified with the NanoDrop ND-1000 (Thermo Fisher Scientific, Waltham, MA, USA). The copy numbers of HPV genomes were calculated from the concentration of plasmid expressed as molarities and Avogadro’s number.
Clinical specimens. Cervical exfoliated cells were collected from outpatients who visited the NTT Medical Center Tokyo, with their informed consent. The study design and sample collection were approved by the institutional review board. One case of normal cytology, two cases of cervical intraepithelial neoplasia (CIN) grade 1, two cases of CIN2, two cases of CIN3, and one case of unknown cytology were selected for this study. DNA was purified using the QIAamp DNA blood kit (Qiagen, Hilden, Germany).
Polymerase chain reaction. Table 1 shows the sequences of primers used in this study: three consensus primer pairs, L1C1/L1C2+C2M,(5) MY09/11,(17) and GP5+/6+;(18) and two mixtures of type-specific primers, PGMY09/11(20) and modified GP5+/6+ (MGP).(21) Polymerase chain reaction amplification was done in a 50 μL reaction mixture containing 1× PCR buffer II (Life Technologies, Carlsbad, CA, USA), 1.25 units AmpliTaq Gold DNA polymerase (Life Technologies), and 50 ng cellular DNA extracted from human HaCaT cells. The 5′-end of either the forward or reverse primer was biotinylated. The concentrations of MgCl2, dNTPs, and primers, and the reaction temperature were adjusted to those used in the original article describing the method.(5,17,18,20,21)Figure 1 shows the location of primers on the HPV16L1 ORF and the size of amplicons.
Table 1. Nucleotide sequences of primers used in this study
MGP, modified GP5+/6+.
Reverse blotting hybridization. Reverse blotting hybridization was done as described previously.(22) Briefly, 15 μL denatured amplicons, of which the 5′-ends were labeled with biotin, were allowed to hybridize with the type-specific probes immobilized on a nylon membrane using the Miniblotter MN45 (Immunetics, Cambridge, MA, USA). The hybridized amplicon was detected using streptavidin–HRP (GE Healthcare Bio-Sciences, Piscataway, NJ, USA) and ECL detection reagents (GE Healthcare Bio-Sciences). The chemiluminescence was detected with the Light-Capture AE-6972 (ATTO, Tokyo, Japan). The intensities of dots were quantified by ImageJ software (National Institutes of Health, Bethesda, MD, USA). The specific density was calculated by the subtraction of the background from the integrated density. Samples showing the specific densities of more than 1000 intensity units were defined as positives.
The nucleotide sequences of type-specific probes for MY09/11, GP5+/6+, PGMY09/11, and MGP primers were described previously.(25,26) Type-specific probes for L1C1/L1C2+C2M primers were newly designed in this study. The nucleotide sequences were as follows: HPV16, 5′-GTTATTGTTAGGTTTTTTAA; HPV18, 5′-CCACCACCTGCAGGAACCCT; HPV31, 5′-AGGATTGTCAGATTTAGGTA; HPV51, 5′-TAGCAGCACGCGTTGAGGTT; HPV52, 5′-ACCATTACCACTACTGGTGT; and HPV58, 5′-TATTGTTATTGGGACTTTTG.
Real-time PCR. Copy numbers of HPV DNA in a clinical sample were determined by real-time PCR using type-specific primers and SYBR-green dye. A reaction mixture (20 μL) containing 2 μL sample, 10 μL Thunderbird SYBR qPCR Mix (Toyobo, Osaka, Japan), 0.4 μL ROX reference dye, and 0.9 μM each primer was subjected to PCR with the Applied Biosystems 7900HT (Life Technologies). The reaction was done in triplicate. The copy number was calculated with the standard curve obtained by using serially diluted HPV plasmids. The nucleotide sequences of type-specific primers are as follows: HPV16 forward, 5′-CAGAACCATATGGCGACAGC and reverse, 5′-GTACATTTTCACCAACAGCA; HPV18 forward, 5′-GATTATTTACAAATGTCTGCA and reverse, 5′-GCACAGTGTCACCCATAGTA; HPV31 forward, 5′-GATTATCTTAAAATGGTTGCT and reverse, 5′-GGACCGATTCACCAACCGTG; HPV51 forward, 5′-AGCTATGGATTTCGCTGCCC and reverse, 5′-AGCAAAGATTTGCTCCCTGC; HPV52 forward, 5′-GATTATTTGCAAATGGCTAGC and reverse, 5′-GGCACAGGGTCACCTAAGGTA; HPV58 forward, 5′-AGTGAACCTTATGGGGATAG and reverse, 5′-AAAGGTCATCCGGGACAGCC. Polymerase chain reaction with these primer sets amplified target HPV DNA without non-specific reaction. For example, PCR with the primers for HPV16 and a test sample containing HPV16 produced a single DNA fragment that formed a single peak in the dissociation curve.
Amplification of HPV DNA in a test sample containing a single HPV genotype. Figure 2 shows the results of reverse blotting hybridization of the amplicons obtained by PCR from test samples (50 μL) containing 6, 60, 600, or 6000 copies of the plasmid having genomic DNA of HPV16, 18, 51, 52, and 58, all prevalent types among Japanese women, and 50 ng sheared human DNA. Three consensus primer pairs (L1C1/L1C2+C2M, MY09/11, and GP5+/6+) and two sets of mixed multiple primers (PGMY09/11 and MGP) were used for PCR. The biotinylated amplicons were allowed to hybridize with the type-specific probes immobilized on a nylon membrane and the biotin on the membrane was detected by streptavidin labeled with peroxidase.
Polymerase chain reaction with L1C1/L1C2+C2M produced detectable amplicons of HPV16, 18, 51, 52, and 58 from the samples containing 60, 6, 6, 60, and 60 copies of HPV DNA, respectively. Polymerase chain reaction with MY09/11 produced detectable amplicons of HPV16, 18, and 58 from the samples containing 600 copies of HPV DNA but did not produce detectable amplicons from the samples containing 6000 copies of HPV51 and 52. Polymerase chain reaction with GP5+/6+ produced detectable amplicons of HPV16, 18, 52, and 58 from the samples containing 60, 60, 6000, and 6000 copies, respectively, but failed to produce detectable amplicons from the sample containing 6000 copies of HPV51. Thus, PCR with L1C1/L1C2+C2M amplified HPV in the test samples more efficiently than the other PCR with the consensus primers.
Polymerase chain reaction with PGMY09/11 produced detectable amplicons of HPV16, 18, 51, 52, and 58 from the samples containing 6, 6, 60, 60, and 60 copies of HPV DNA, respectively. The PCR with MGP produced detectable amplicons of HPV16, 18, 51, 52, and 58 from samples containing 6, 60, 60, 6, and 600 copies of HPV DNA, respectively.
Amplification of HPV DNA in a mixed test sample containing HPV16, and either HPV18, 51, 52, or 58. Figure 3 shows the results of reverse blotting hybridization of the amplicons obtained by PCR from test samples containing 6000 copies of HPV16 and 6, 60, 600, and 6000 copies of either HPV18, 51, 52, or 58 and 50 ng sheared human DNA.
Polymerase chain reaction with L1C1/L1C2+C2M failed to amplify HPV16 DNA from the samples containing 6000 copies of HPV18 or 51 and the level of HPV16 amplicons was greatly reduced in the presence of 600 copies of HPV18 or 51. However, the amplification of HPV18 and 51 DNA was not influenced by the presence of 6000 copies of HPV16 DNA (Figs 2,3). The PCR with L1C1/L1C2+C2M amplified HPV16 DNA in the presence of HPV52 or 58 DNA. The data clearly indicate that amplification of HPV16 DNA by PCR with L1C1/L1C2+C2M was significantly interfered with by the presence of HPV18 or 51 DNA.
Polymerase chain reaction with MY09/11, GP5+/6+, PGMY09/11, or MGP amplified HPV16 DNA in the presence of HPV18, 51, 52, or 58 DNA but the level of HPV16 amplicons was reduced by co-existence of 6000 copies of HPV18 DNA. The PCR with GP5+/6+, PGMY09/11, or MGP produced reduced levels of HPV18, 52, or 58 amplicons in the co-existence of 6000 copies of HPV16 DNA (Figs 2,3).
Amplification of HPV DNA in clinical specimens containing two or three HPV genotypes. Table 2 shows the detection and genotyping of HPVs in clinical specimens using PCR with consensus primers. Eight clinical samples in which two or three HPV types had been detected by PCR with PGMY09/11 or MGP were selected, and the copy numbers of the detected HPV DNA in the samples were measured by real-time PCR using type-specific primers. Then the HPV in the samples was examined by PCR with L1C1/L1C2+C2M, MY09/11, or GP5+/6+. The HPV types detected with consensus primers are listed in decreasing order of amplicon levels.
Table 2. Genotyping of clinical samples containing multiple types of human papillomavirus (HPV)
Sample no. (cytology)
Genotypes: copy numbers†
†Genotypes detected using PGMY09/11 and modified GP5+/6+ primers. Copy numbers were determined by real-time PCR and expressed as copy numbers subjected to PCR for genotyping. ‡The HPV types detected with consensus primers are listed in decreasing order of amplicon levels.
In no. 165 containing HPV16 (16 000 copies), 18 (5200), and 31 (3100), HPV18 was detected by PCR with L1C1/L1C2+C2M but neither HPV16, which was three times more abundant than HPV18, nor 31 were detected. All three HPVs were detected by PCR with MY09/11. HPV16 and 18 were detected by PCR with GP5+/6+ but HPV31, the least component, was not detected by PCR with GP5+/6+.
In no. 258 containing HPV52 (1700), 16 (1300), and 58 (180), all three HPVs were detected by PCR with L1C1/L1C2+C2M. HPV16 and 58 were detected by PCR with MY09/11 but HPV52, the most abundant type, was not detected. Only HPV16 was detected by PCR with GP5+/6+.
In no. 352 containing HPV16 (200) and 52 (25), neither HPVs were detected by PCR with L1C1/L1C2+C2M. Only HPV16 was detected by PCR with MY09/11 and PCR with GP5+/6+.
In no. 402 containing HPV52 (7300), 51 (120), and 16 (110), HPV16 was not detected by PCR with L1C1/L1C2+C2M. HPV16 and 52, but not HPV51, were detected by PCR with MY09/11. Only HPV16 was detected by PCR with GP5+/6+.
In no. 933 containing HPV16 (5600) and 52 (3300), and in no. 996 containing HPV16 (47 000) and 52 (25 000), HPV52 was not detected by PCR with MY09/11 and GP5+/6+.
In no. 1061 containing HPV18 (4300) and 16 (290), HPV16 was not detected by PCR with L1C1/L2C2+C2M.
In no. 1245 containing HPV16 (78 000) and 58 (78 000), HPV58 was not detected by PCR with GP5+/6+. Thus, amplification of HPV DNA in the clinical specimens containing multiple HPV genotypes by PCR with consensus primers was biased and sometimes caused misjudgment in typing.
We evaluated HPV consensus primers for PCR amplification of HPV DNA in specimens containing multiple HPV types and concluded that PCR with consensus primers is not suitable for simultaneous amplification of multiple types of HPV DNA. The low sensitivity of HPV51 and 52 detection by PCR with MY09/11 and GP5+/6+ is consistent with previous reports.(20,21,27) Polymerase chain reaction with L1C1/L1C2+C2M, which amplified HPV DNA tested more efficiently than PCR with MY09/11 or GP5+/6+ (Fig. 2), failed to amplify HPV16 DNA when HPV18 or 51 DNA coexisted in the samples (Fig. 3, Table 2).
Polymerase chain reaction with L1C1/L1C2+C2M amplifies HPV18 and 51 DNA very efficiently, as shown in Figure 2. The resultant abundant HPV18 and 51 amplicons might inhibit reactions amplifying the other HPV types in the samples. Similarly, it is reported that interference in PCR was observed even between two closely related plant virus isolates that have identical binding sites for consensus primers, although the mechanism of the interference is not fully explained.(28)
As PCR with L1C1/L1C2+C2M has been used widely,(5–16) the biased amplification may have caused mistyping in the previous numerous studies of genotyping of clinical specimens. Previous studies using L1C1/L1C2+C2M showed that the rate of multiple infections among HPV-positive Japanese women was 10–20%.(15,16) In other countries, studies using PGMY09/11 showed that 30–45% of HPV-positive women were infected with multiple genotypes.(20,29–31) It is possible that reevaluation of HPV prevalence in Japanese women by using mixtures of type-specific primers, such as PGMY09/11 and MGP, increases the rate of multiple infections. The data of this study suggest that the genotyping data obtained by PCR with consensus primers should be carefully interpreted.
We acknowledge the World Health Organization’s HPV Laboratory Network for technical transfer of reverse blotting hybridization into our laboratory. This work was supported by a grant-in-aid from the Ministry of Health, Labour and Welfare for the Third-Term Comprehensive 10-year Strategy for Cancer Control.
The authors have no conflict of interest to declare.