Assessing Gammapapillomavirus infections of mucosal epithelia with two broad-spectrum PCR protocols

Samples of total DNA (n = 458), extracted from 249 swabs of the anal canal (226 initial and 23 follow-up samples), 94 cervical cancer formalin-fixed paraffin-embedded (FFPE) tissue samples, 21 genital warts (11 swabs and 10 biopsies) from different anatomical sites [foreskin (n = 4), scrotum (n = 2), penile glans (n = 2), perianal area (n = 4) and pubis (n = 9)], and 94 oral swabs collected during routine paternity testing, were obtained from the archival collection of samples of the Instituto de Biología Molecular y Celular de Rosario (Rosario, Argentina) and the Institute of Microbiology and Immunology, Faculty of Medicine, University of Ljubljana (Ljubljana, Slovenia). Before subsequent analyses, all samples were collected, processed, and stored at − 80 °C, as described previously [11, 27, 29‐32]. The adequacy of samples for downstream analyses was determined by PCR amplification of the human beta-globin gene, as previously described [33].

In more detail, and in order to compare the prevalence and persistence of HPV infection in the anal canal of HIV-positive and HIV-negative subjects, samples and data from 226 patients were mostly obtained from our previous studies on Alpha- and Beta-PV infections in the anal canal [11, 27]. Additionally, 23 follow-up samples from eligible individuals that were obtained after the conclusion of two previous studies were included in the present study. Subjects were 18 to 66 years old (median age = 33 years) and 20/226 (9%) represented HIV-1 positive men. As all anal canal swab samples were collected during standard proctologic exams unfortunately information on the presence of specific anal lesions could not be obtained for all patients. For the 133 patients for whom clinical data were available, more than two-thirds (90/133; 67.6%) had a clinically evident HPV-related or HPV-unrelated anal pathology, including anal warts (77/133; 57.9%), hemorrhoids (5/133; 3.8%) and anal fissure (3/133; 2.3%), as reported in our previous study [27], suggesting that anal warts were most probably the predominant HPV-related pathology in our patient population. Additionally, all male participants were MSM and had a history of receptive anal sexual intercourse. Unfortunately, no data on the sexual behavior of women could be obtained. Moreover, a total of 21 subjects (20 males and one female), age range of 22 to 42 years (median age = 30 years), of which two men were HIV-1-positive, were enrolled in the current analysis of HPV persistence in the anal canal, with a follow-up anal swabs collected in period ranging from 2 to 82 months (average = 36 months; median = 43 months). At enrolment, 10/21 (47.6%) of these patients had anal warts, 4/21 (19.0%) had no clinically evident abnormalities, two patients had anal fissure and hemorrhoids, respectively, and no data were available for 5/21 (23.8%) patients (Table S2).

The presence of HPV infection was determined using two well-established HPV generic primer systems, Gamma-PV PCR [21] and CUT PCR [30]. All PCR reactions were performed as described previously [21], using the following reaction controls: a negative control (5 ng of human placental DNA) to check for the reaction’s specificity, a reagent control (H₂O instead of the sample) to check for carry-over contamination, and 100 copies of cloned HPV4 or HPV10 in a background of 5 ng human placental DNA as a positive control per PCR run. All pre- and post-PCR procedures were carried out in separate cabinets and rooms. Amplicons derived from CUT (≈370 bp) and Gamma-PV PCR systems (≈158 bp) were sequenced by Sanger methodology. The obtained nucleotide sequences were compared to available HPV-sequences in the GenBank database (https://​www.​ncbi.​nlm.​nih.​gov/​genbank/​), using the Blast algorithm (https://​blast.​ncbi.​nlm.​nih.​gov/​Blast.​cgi). Following CUT PCR, a novel putative HPV type was determined when the fragment sequence showed less than 90% nucleotide identity to L1 ORFs of any of the previously known HPV types [1]. The same criterion was applied to classify types/putative types derived from the Gamma-PV PCR, which is based on the amplification of partial HPV E1 sequences, due to the similar nucleotide identities obtained in a pairwise comparison analysis between L1 and E1 ORFs, as shown previously [21].

Additionally, in the analysis of HPV persistence in the anal canal, the obtained results were compared to the results of HPV typing using the commercially available Linear Array HPV Genotyping Test (Roche Diagnostics, Mannheim, Germany) and RHA Kit Skin (Beta) HPV assay (RHA; Diassay BV, Rijswijk, The Netherlands), as described previously [11, 27].

Sequences of E1 gene regions of 166 representative HPV types from the Gamma-PV genus, available at http://​www.​nordicehealth.​se/​hpvcenter/​ and Papillomavirus episteme (http://​pave.​niaid.​nih.​gov), were used as a database for the phylogenetic analysis of novel putative HPV types. Multiple sequence and pairwise alignments were constructed using the ClustalW algorithm of the MEGA6 software package [34] at the amino acid (aa) level. The alignment of novel putative HPV E1 sequences was obtained with MAFFT’s “Align fragment sequences to an MSA” tool [35]. The phylogenetic relationships between representative Gamma-PV types and novel putative HPV E1 sequences were inferred by Bayesian analysis using Beast version 1.7.5 [36]. To do so, Markov Chain Monte Carlo (MCMC) simulations were performed during 2 × 10⁷ generations, sampling one state every 1000 generations, with a burnin of 10%. The setting “Create tree log file with branch length in substitutions” was selected to obtain the phylogram log file. The evolutionary substitution model selected for each run was GTR + I + Γ. Statistical convergence of MCMC was assessed visually by the traceplot and by calculating the effective sample size using TRACER v1.4 (available at http://​beast.​bio.​ed.​ac.​uk/​Tracer). The maximum clade credibility tree across all the plausible trees generated by BEAST was then computed with the TreeAnnotator program available in the BEAST package.

The GenBank/EMBL/DDBJ accession numbers for the novel HPV putative types reported in this paper are: EP26 (MK510728), EP27 (MK510729), EP28 (MK510730), EP29 (MK510731).

A total of 71 different HPV types/putative types (27 Alpha-, 7 Beta- and 37 Gamma-PV), clustering to 30 PV species (9 Alpha-, 3 Beta- and 18 Gamma-PV), were identified in HPV-positive mucosal samples using both broad-spectrum PCR protocols (Table 1, Table S1). It should be noted that only five HPV types (HPV38, HPV133, HPV161, HPV135 and HPV180) were consistently simultaneously detected in the same samples by both primer systems. In addition, four novel putative HPV types (EP26, EP27, EP28, EP29) were identified in this set of samples, all of them being found in anal canal swabs with the Gamma-PV PCR primer system (Table 1, Table S1). Phylogenetic relationships between these novel putative HPV types and sequences of E1 gene regions of 166 representative HPV types from the Gamma-PV genus are shown in Fig. 1. Putative HPV type EP26 clusters within the species Gamma-20 and shares a 90% E1 ORF nucleotide identity with HPV163. EP27 belongs to the species Gamma-9 and shows the highest nucleotide identity with HPV216 (88%). On the other hand, EP28 exhibits an 87% nucleotide identity with HPV-mSE379, clustering within the Gamma-8 species, while EP29 belongs to the Gamma-10 species, sharing a 90% nucleotide identity with HPV180.

Overall, HPV DNA was present in 55% (253/458) of all tested mucosal samples and HPV prevalence was determined at 58% (253/435) (Table 1). Specifically, HPV prevalence in anal canal swabs, cervical cancer biopsies, genital warts and oral swabs was estimated at 85% (192/226), 47% (44/94), 62% (13/21) and 4% (4/94), respectively. Additionally, 25% (116/458) of included samples contained multiple HPV infections, most frequently in swabs of the anal canal (105/226; 48%) (Table 1).

As shown in Fig. 2 and Table S1, in swabs of the anal canal members of the Alpha-PV genus were the most frequently detected (113/192; 59%), followed by Gamma- (71/192; 37%) and Beta-PV genera (8/192; 4%). Alpha-PV genus types were also the most frequently detected in cervical cancer biopsies (43/44; 98%) and genital wart samples (13/14; 93%). While HPV types clustering to the species Alpha-10 were the most frequent in anal canal swabs (50/192; 26%) and genital wart samples (11/14; 79%), HPVs grouped within the species Alpha-9 were most frequently identified in cervical cancer biopsies (36/44; 82%). The prevailing HPV type detected in both anal canal swabs (37/192; 19%) and genital wart samples (7/14; 50%) was HPV6 (Alpha-10) and, as expected, HPV16 (Alpha-9) was the most prevalent in cervical cancer biopsies (32/44; 73%). Nevertheless that a wide variety of Gamma-PVs, clustering to different Gamma-PV species, were detected in the anal canal (Fig. 2, Table S1), those grouped within the species Gamma-6 were the most frequently identified (11/192; 6%). It should be noted that unclear electropherograms from which the prevailing HPV type/s and species could not be determined (Gamma-X) were obtained from 18 Gamma-PV-positive anal canal swab samples (Fig. 2, Table S1). The frequency of HPV infection in oral swab samples was relatively low and the majority of positive samples contained Alpha-PVs [HPV10 (Alpha-2) and HPV18 (Alpha-7)] or Beta-PVs [HPV38 (Beta-2)] (Fig. 2, Table S1).

The overall prevalence of HPV infection in the anal canal of the 226 patients included in the study was estimated at 79% (Table 2). Particularly, Alpha-, Beta- and Gamma-PVs were present in 59, 3 and 35% of the initial samples obtained from all study participants at the first visit, respectively. In addition, multiple infections were detected in 43% of included samples.

In comparison to women, the prevalence of overall HPV, Gamma-PV and multiple HPV infections was significantly higher in anal canal swab samples of men (82% vs. 69%, p = 0.034; 38% vs. 22%, p = 0.027 and 49% vs. 26%, p = 0.003, respectively) (Table 2). Multiple HPV infections were slightly more common among 18-30 and 31-40-year-old patients (45 and 49%, respectively), in comparison to individuals older than 40 years (27%) (p = 0.05). A significantly higher prevalence of Gamma-PVs and multiple HPV infections was observed in HIV-1-positive vs. HIV-1-negative individuals (65% vs. 32%, p = 0.003 and 65% vs. 41%, p = 0.04, respectively). All differences between gender, age and HIV-1 status observed during univariate analyses were additionally confirmed using multivariate analyses (Table 3). Thus, men had a higher risk for overall HPV infections, Gamma-PV infections and multiple HPV infections than women, and HIV-1-positive subjects had a higher risk for infections with Gamma-PV and for multiple HPV infections than HIV-1-negative individuals.

As shown in Table S2, out of 21 patients with follow-up anal swabs, persistent infection with the same HPV type was detected only in a single patient (HPV58, patient No.18; Table S2). Additionally, two individuals were HPV-negative throughout the follow-up period (patient No. 8 and No. 21; Table S2). A total of 52% (11/21) individuals were infected with different HPV types/putative types in the initial and follow-up anal samples, and 29% (6/21) had HPV-positive initial samples but tested negative in the follow-up samples. Only one patient tested HPV-positive in follow-up sample, while being previously HPV-negative (patient No. 2; Table S2).