Comparison of five different polymerase chain reaction methods for detection of human papillomavirus in cervical cell specimens

Abstract

The polymerase chain reaction (PCR) methods enable the detection of large number of human papillomavirus (HPV) genotypes that infect the anogenital tract. In this study, two groups of cervical scrapes with abnormal cytomorphology were analysed. The first group was tested with three sets of consensus primers located within the L1 region of HPV genome, MY09/MY11 (i.e. MY), L1C1/L1C2-1/L1C2-2 (i.e. LC) and pI-1/pI-2 (i.e. pI) primer sets, while the second group of samples, which were all negative with the MY primers, was tested further with the LC primers, as well as with the GP5/GP6 (i.e. GP) primers. The GP primers were used in the nested PCR following amplification with the MY primers (i.e. MY/GP nested PCR). Samples from both groups were also tested with type-specific primers for HPV types 6/11, 16, 18, 31 and 33. In the first study group (N=164) there were 76.2% positive results obtained with at least one set of consensus primers. There were 62.2, 39, 62.2 and 59.1% positive results obtained with the MY, the pI, the LC and the HPV type-specific primer sets, respectively. The best results were obtained when both the MY and the LC primer sets were used, because in combination they detected 75% positive samples compared to 62.2% when used alone. There were 2.4% samples negative with all consensus primers, but positive with one of the HPV type-specific primers, which increased the overall positivity rate to 78.6%. In the second study group (N=250) there were 8.4, 38.8 and 4% samples positive with the LC primers, the nested MY/GP and the HPV type-specific primer sets, respectively. Thus, the use of the MY/GP nested PCR increased significantly the positivity rate of HPV DNA detection and should be used for samples with a low copy number of HPV DNA. In conclusion, the following diagnostic protocol would be appropriate for detection of cancer-related HPVs: preselection of samples with the MY and the LC primers, additional amplification of the MY- and the LC-negative samples with the MY/GP nested PCR and HPV typing of consensus PCR-positive samples with the HPV type-specific primers.

Introduction

Human papillomaviruses (HPVs) play a central role in cervical carcinogenesis. The presence of >70 HPV genotypes, of which >30 infect the anogenital tract, has complicated the design of polymerase chain reaction (PCR) methods for HPV DNA detection. Consensus or universal PCR methods enable the detection of large number of HPV types in one reaction.

The genome of HPVs consists of seven early (E1–E7) and two late (L1 and L2) open reading frames (ORFs). Beside many variable regions, most HPVs share high homology in certain regions of their genome. Such regions are used for construction of consensus or universal primers, which are able to amplify many different genotypes of HPV. There are several consensus primers complementary to L1 (Manos et al., 1989; Yoshikawa et al., 1990; Snijders et al. 1990) and E6/E7 regions of HPV genome (Resnick et al. 1990). During the integration into the host DNA, parts of the L1 region could be deleted (Schwarz et al. 1985; Choo et al. 1988) and consensus primers, which are not located in the L1 region, are thus preferred. But, in comparison to E6 primers, L1-PCR has a larger spectrum of HPV detection and it is commonly used for HPV DNA detection.

The MY09 and MY11 (in the text referred to as MY) primers (Manos et al. 1989) are located within the L1 region of HPV genome, they are degenerate primers, i.e. they consist of 24 pairs of primers and are able to amplify >25 genital HPVs.

Novelli et al. (1992) constructed pI-1 and pI-2 (in the text referred to as pI) primers similar to the MY primers, in which inosine was incorporated in the degenerated positions. Inosine is a neutral base, which is found commonly in tRNA and forms stable base pairs with the four conventional bases with similar pairing strength.

The GP5 and GP6 (in the text referred to as GP) primers (Snijders et al. 1990) are complementary to the part of the L1 region located inside the sequence recognized by the MY primers. Therefore, they can be used either as single primers or in the nested PCR after the amplification with the MY primers (Evander et al. 1992).

Yoshikawa et al. (1990) constructed a set of consensus primers which are also complementary to the L1 region of HPV genome, but located upstream of the MY and GP sequences. By using the L1C1 and L1C2-1 primers, at least nine genital HPVs can be efficiently amplified (HPV 6, 11, 16, 18, 31, 33, 42, 52 and 58) and two (HPV 34 and 36) amplified less efficiently. The third primer L1C2-2 that recognizes HPV 58 was introduced, and L1C1, L1C2-1 and L1C2-2 (in the text referred to as LC) primers were used in a ratio 2:1:1, in order to increase the sensitivity of HPV 58 DNA amplification, without decreasing the sensitivity of other HPV types detection.

In this study, the results obtained with three sets of consensus primers (MY, pI and LC) and the type-specific primers for HPV 6/11, 16, 18, 31 and 33 were compared. In order to increase the sensitivity of HPV detection, the GP primers were introduced in the nested PCR after amplification with the MY primers. The results of the nested PCR were compared to those obtained with the LC and the HPV type-specific primers.