PCR-RFLP assay as an option for primary HPV test

Persistent human papillomavirus (HPV) infection is an essential factor of cervical cancer. This study evaluated the analytical performance of restriction fragment length polymorphism polymerase chain reaction (PCR-RFLP) assay compared to PapilloCheck® microarray to identify human papilloma virus (HPV) in cervical cells. Three hundred and twenty-five women were analyzed. One sample was used for conventional cytology and another sample was collected using BD SurePath™ kit for HPV tests. Eighty samples (24.6%) were positive for HPV gene by PCR-Multiplex and were then submitted to PCR-RFLP and PapilloCheck® microarray. There was a genotyping agreement in 71.25% (57/80) on at least one HPV type between PCR-RFLP and PapilloCheck® microarray. In 22 samples (27.5%), the results were discordant and those samples were additionally analyzed by DNA sequencing. HPV 16 was the most prevalent HPV type found in both methods, followed by HPVs 53, 68, 18, 39, and 66 using PCR-RFLP analysis, and HPVs 39, 53, 68, 56, 31, and 66 using PapilloCheck® microarray. In the present study, a perfect agreement using Cohen's kappa (κ) was found in HPV 33 and 58 (κ=1), very good for HPV 51, and good for types 16, 18, 53, 59, 66, 68, 70, and 73. PCR-RFLP analysis identified only 25% (20/80) HPV coinfection, and PapilloCheck® microarray found 62.5% (50/80). Our Cohen's kappa results indicate that our in-house HPV genotyping testing (PCR-RFLP analysis) could be applied as a primary HPV test screening, especially in low income countries. If multiple HPV types are found in this primary test, a more descriptive test, such as PapilloCheck® microarray, could be performed.


Introduction
Cervical cancer was responsible for an estimated 530,000 diagnoses and 266,000 deaths in 2012, the most common type of gynecological tumor worldwide (1). In Brazil, cervical cancer is the third most common tumor in women and the fourth cause of death. In 2013, 5,430 women died from cervical cancer, and in 2016 16,340 new cases are estimated in Brazil (2). Persistent human papillomavirus (HPV) infection is an essential factor of cervical cancer (3). Two hundred different types of HPV have been discovered (4). From those, about 51 types are considered either high-risk (HR) or low-risk (LR) genital HPVs types associated with benign, precancerous or cancer lesions (5). This discovery is changing the target for reducing the high mortality of cervical cancer, such as the introduction of the HPV vaccine and/or application of the HPV test as a primary screening test.
The introduction of HPV vaccine is stimulating developed countries to change their cervical screening program from cytology to HPV testing as a primary screening test (6)(7)(8)(9)(10). However, various strategies have been proposed to achieve better performance in the detection of precancerous lesions and optimize balance between benefits and harms (11)(12)(13). Some countries are presently discussing which test should be used to improve sensitivity and specificity for cervical screening program.

Study population
From November 2011 to March 2013, 990 women were attended for cervical screening in public clinics in São Miguel do Oeste, SC, Brazil. Three hundred and twenty-five of those women accepted to participate in our study. Two samples were collected from each woman. One sample was used for conventional cytology for routine diagnosis. Another sample was collected with BD SurePatht kit for HPV tests. The routine cytological analysis was performed by Papanicolaou staining, analyzed and classified according to the 2001 Bethesda system: NILM (negative for intraepithelial lesion or malignancy), ASC-US (atypical squamous cells of undetermined significance), ASC-H (atypical squamous cellscannot exclude HSIL), LSIL (low-grade squamous intraepithelial lesion), HSIL (high-grade squamous intraepithelial lesion), and ICC (invasive cervical cancer) (15). The absence of biopsies are a limitation of the present study. The liquid based sample was used for an in-house HPV genotyping testing (PCR-RFLP analysis) and PapilloCheck s microarray. For discrepancies of HPV results, DNA sequencing of L1 amplified fragments was performed.

DNA extraction
To select which method should be applied for DNA extraction, we used QiaAmp s DNA mini Kit (Qiagen, Germany) (16), phenol-chloroform extraction, guanidinium thiocyanate extraction (17), and ammonium acetate extraction (18). The best results were obtained by the ammonium acetate method.
DNA was extracted from cervical cytology vials (SurePatht, USA). A 1-mL aliquot was pelleted by centrifugation at 12,000 g for 5 min at 25°C. The preservative fluid was removed and cells were resuspended in 180 mL sterile phosphate buffered saline (PBS). Subsequently, 20 mL of Proteinase K (Qiagen) and 180 mL of AL buffer (Qiagen) were added to these cells, vortexed and heated at 56°C under agitation (1200 rpm) for 1 h and at 90°C for another hour. Two hundred microliters of 2 M ammonium acetate (Sigma, USA) were added to the cell lysate, submitted to an ice bath for 5 min, and centrifuged at 12,000 g for 4 min at 25°C. Then, the supernatant was transferred to another tube and 600 mL of isopropanol (Sigma) was added. The cells lysate was homogenized for 20 inversions and centrifuged at 12,000 g for 4 min at 25°C. The supernatant was discarded; the pellet was washed with 1 mL of 70% ethanol (Sigma) and centrifuged at 12,000 g for 4 min at 25°C. The supernatant was removed and the pellet was maintained at 60°C until complete evaporation of the ethanol. DNA was suspended in 50 mL of elution buffer Tris-EDTA and stored at -20°C until further use. DNA quality control was determined by NanoVue spectrophotometry (GE Healthcare, United Kingdom).

Polymerase chain reaction (PCR)
Multiplex PCR with PGMY09/11 and PCO4/GH20 primers (19) was performed on a final reaction volume of 25 mL. The reaction was carried out with 20 mM Tris-HCl, pH 8.4, 50 mM KCl, 5.7% glycerol, 0.04 mM of each PGMY09/11 primer (Life Tecnologiest, USA), 0.2 mM of each PCO4/GH20 primer (Life Tecnologiest), 2.5 mM of MgCl 2 , 200 mM of dNTP (Life Tecnologiest), 2U Taq Platinum DNA Polymerase (Life Tecnologiest, Brazil), and 5.0 mL of DNA. The target DNA was amplified by PCR (Mastercycle Personal s Eppendorf) and reaction was carried out with a denaturation step at 95°C for 10 min, 40 cycles of 1 min at 95°C, 1 min at 55°C, 1 min at 72°C, and final extension at 72°C for 10 min. Plasmids containing HPV-33 L1 gene were used as positive control and DNase-and RNase-free water was used as negative control in all amplifications.

PapilloCheck s microarray
All positive HPV PCR-Multiplex samples were submitted to PapilloCheck s microarray (Greiner Bio-One, Germany) 5mL of DNA eluate was used in the Papillo-Check s microarray for each reaction. Specimens containing the target DNA are hybridized to specific oligonucleotide probes immobilized on a DNA chip and detected by the binding of a Cy5-dUTP labeled oligonucleotide probe to the tag sequence. The DNA chip was scanned by the CheckScanner apparatus at 532 and 635 nm wave lengths. This test detects HPV genotypes 6,11,16,18,31,33,35,39,40,42,43,44,45,51,52,53.56, 58, 59, 66, 68, 70, 73, and 82. In addition, human ADAT1 gene (adenosine deaminase, tRNA specific 1) was used as an internal control to assess the quality of the DNA

DNA sequencing
DNA sequencing of PGMY09/11 PCR fragments was performed for samples with discordant results between the two genotyping methods. The amplicons were purified with the PureLink s PCR Purification kit (Life Tecnologiest, Germany) or with QIAquick Gel Extraction kit (Qiagen), according to the manufacturers' instructions. Automated DNA sequencing was performed in an ABI 3730 Genetic Analyzer sequencer (Applied Biosystems, USA). The accuracy of the DNA sequencing was evaluated through the CAP3 program based on Phred quality score (21).
Sequences were aligned and compared to those available in the GenBank database using the software Chromas Lite 2.1 (Technelysium, Australia). HPV type was identified based on 490% sequence homology over 449-458 nucleotides.

Statistical analysis
To determine the correlation between PCR-RFPL and PapilloCheck s microarray, Kappa test was performed and the reference values adopted were determined as proposed by Altman (22).

Ethical approval
This cross-sectional descriptive study was approved by Research Ethical Committee of the Universidade Federal de Santa Catarina (process No. 2155), and participants provided written informed consent to the study protocol.
All 325 samples were positive for b-globin gene (control) and 80 samples (24.6%) were positive for HPV gene by PCR-Multiplex. Those 80 samples were then submitted to PCR-RFLP and PapilloCheck s microarray. Of those, 22 (35.0%) were additionally analyzed by DNA sequencing.
Kappa test was applied to analyze the agreement level between PCR-RFLP and PapilloCheck s microarray. The genotyping agreement between the methods was considered    (Table 3).
In this study, HPV 53 was found in 12.5% by PCR-RFLP analysis, while in 22.5% of samples by Papillo-Check s microarray. Using PapilloCheck s microarray, HPV 39 was revealed in 23.8%, but only in 5% of samples by PCR-RFLP analysis. Other types of HPVs, such as 68 and 18, did not show relevant differences in their prevalence. Martins  Some studies have reported that HPV 56 was the second most frequent type (27,33,34) while others revealed low frequency, which corroborate to the current study (35,36). This discrepancy could be attributed to a higher sensitivity of PapilloCheck s microarray compared to PCR-RFLP analysis or due to its reduced specificity (37).
DNA sequencing was applied for discordant results from PCR-RFLP and PapilloCheck s microarray. In one case, HPV 68 was found by PCR-RFLP and DNA sequencing, while HPV 39 was found only by Papillo-Check s microarray. Those HPV types belong to same family, alphapapillomavirus, genus a7 in the phylogenetic tree (23,38), which could justify the different results. To confirm the DNA sequencing, cloning was performed using 450 bp fragment amplified by primers and subsequently sequenced using PGMY09/11, which revealed HPV types 61 (5 cases), HPV 62 (4 cases), and HPV 32 (1 case) in the clones. In the present study, a perfect agreement was found for HPV 33 and 58 (k=1), very good for HPV 51, and good for types 16,18,53,59,66,68,70, and 73. PCR-RFLP analysis identified only 25% (20/80) HPV coinfection, and PapilloCheck s microarray found 62.5% (50/80). Results indicated that our in-house HPV genotyping testing (PCR-RFLP analysis) could be applied as a primary HPV test screening, especially in low income countries. Probably, the small number of patients limited our results. However, if multiple HPV types are found in this primary test, a more descriptive test, such as PapilloCheck s microarray, could be performed. Currently, there is no gold standard for HPV typing (39,40) and the method should be chosen for clinical purpose based on its advantages and disadvantages.