The surveillance of Human papillomavirus (HPV) infection is epidemiologically relevant considering the incorporation of HPV prophylactic vaccines in different countries. In Argentina, where its general prevalence in 2010 was estimated in 20.1%54, vaccination was introduced into the immunization schedule for females aged 11 years in 201137.

Among the over 150 HPV types described, HPV-16 predominates, whereas the prevalence of other HPV types varies according to the geographic region5. It has been reported that about 70% of cervical cancers worldwide are due to HPV-16 and -1820,32, which are the goals of commercial vaccines. However, there are still other prevalent high-risk HPV types, such as HPV-31, -33, -45 and -5853. In some provinces of Argentina (Chaco, Corrientes, Córdoba and Misiones), an important detection of HPV-33 and HPV-58 has been reported2,17,22,23,51.

Although there have been several studies on the genetic diversity and phylogenetics of HPV4,14,19,21,42, most of them focused on the characterization of HPV-16 and HPV-18 variants and their associations with pathogenic potential and immunogenicity18,19.

Within HPV-58, several authors have identified four distinct clusters (A, B, C and D), being lineage A the most prevalent worldwide14,19. Reports about the important detections of this high-risk type in regions of Argentina such as the Northeast22,23,25, which has the highest mortality rates due to cervical cancer1, require a deeper characterization of viruses of HPV-58.

The aim of this work was to study HPV sequences from the Province of Corrientes, Argentina, especially those belonging to HPV-58. We attempted to characterize the viruses from samples taken from different cervical lesions, their nucleotide changes, phylogenetic relationships and the evolutionary dynamics of HPV-58 in the context of the alpha-9 papillomavirus species.

HPV is an important cofactor for cervical neoplasia, and in Argentina it is the most frequent neoplasia (20.9/100,000 women) after mammary cancer (71.2/100,000 women)33. Provinces in the North and Northeastern regions of the country exhibit the highest rates for cervical cancer mortality1, e.g. the Province of Corrientes shows a mortality rate due to cervical cancer of 14/100,000 women, which is three-fold higher than the rate for Buenos Aires, the Capital city of the country, and even exceeds the value of the mortality rate due to mammary cancer estimated by screening studies in Corrientes1,36.

The population under study gathered women with a low socioeconomic status, who are particularly vulnerable to this disease due to reduced access to prevention and control programs.

Despite the studies performed in this region, the number of reported cases is still scarce to evaluate the role of high-risk types2,22–24,47,48. However, it should be noted that we found a high diversity of oncogenic types, with an increased proportion of high-risk and probable high-risk types not included in the current vaccines. This is of concern since there are reports suggesting a limited cross-protection of the vaccines against non vaccine types43,50,53. Therefore, it is necessary to strengthen the monitoring of regional circulation of these types in cases of intraepithelial lesions, contributing to impact assessment of current vaccines against HPV54.

The description of the molecular epidemiology of HPV before the vaccine introduction in 2011 will lead to a better description of the dynamics of the circulation of different genotypes and their variants and also to a better understanding of the relationship between HPV types and cancer in our country, which is useful information for the evaluation of new vaccine formulations. Further studies in post-vaccinal stages will allow to evaluate the dynamics of high-risk HPV types and cross protection among different types10,54. It is worth noting that HPV-58 is included in the nonavalent vaccine V503 formulation, which is still under evaluation16.

The most frequently found type in Corrientes was HPV-16, as also observed all around the world in populations both with normal cytology and cervical abnormalities5,9,34,46. However, the secondary types vary depending on the geographic region under study. In this work, the most important secondary type was HPV-58, as also observed in other studies in the cities of Corrientes and Resistencia22,23. This result differs from the reference meta-analysis conducted by Bruni et al.5 on samples with normal cytology, which showed that HPV-18 was the most important secondary type both globally and in Latin America and the Caribbean. In different regions of the world, and even in other regions of Argentina36,39,45, HPV-58 was found to have low frequency (<2%) when samples with different grade of lesions were considered9,34,46. However, HPV-58 displays a high prevalence in specific regions of the world, especially in Northeast Asia (4–15%), Mexico (24–28%), Costa Rica (12%) and Brazil (12–14%)7,11,28,30,38,52.

Four different lineages (A–D) have been identified within HPV-5814,19. All the viruses reported in this work belonged to lineage A, sublineage A2, which is the most prevalent in the Americas14; none of them belonged to lineage D, which was previously reported in some samples from Argentina14.

HPV-58 is considered a high-risk viral type. However, its disease impact shows geographic variation, since cervical cancer and HSIL are more prevalent in Asian women with HPV-58 than in women from other regions46.

The association between pathogenicity and geographic location could be a consequence of specific genetic backgrounds of the host populations but it could also be related to a differential geographic distribution of distinct HPV variants. For other types it has been proposed that different variant lineages might display different pathogenic potentials31,44. In line with this fact, subgroups within lineage A of HPV-58, which display a differential distribution worldwide, could be related to different clinical progression.

Chan et al.14 proposed to further investigate the possible association of oncogenicity with sublineage A1 of HPV-58 in Asia. Indeed, some mutations in the E7 gene showed a trend of association with the severity of the neoplasia12. In our work, we found no association between high pathogenicity and HPV-58 infections, neither with the mutations reported as involved in this phenotype; however, it should be noted that only sublineage A2 was detected and that the samples belonged to women under 35 years old, and therefore not included in the age range (35–64 years old) strongly associated with progression to cervical cancer32.

The analysis of genome sequences showed that E7 mutations (the only ones rendering amino acid changes) were more frequent than E6 mutations. Neither mutations previously associated with the severity of neoplasia12,15,27, nor cervical cancer associated to HPV-58 were found. However, other mutations not reported previously (such as C767A [T65K] and G859A [A96T]) were found among the samples and deserve further attention given that they localize in a highly reactive epitope among patients with cervical cancer13. Additional studies in the E6 protein assessing the presence of mutations in the protein binding motifs PDZ might help to study their impact on the low oncogenicity of HPV-58 in this population49.

The evolutionary dynamics of the alpha-9 species group and particularly, of type HPV-58 was analyzed. The alpha-9 species separated into two highly supported groups, one formed by types 16, 31, 35 and, the other, by types 33, 52, 58, 67.

Different substitutions rates – needed for phylodynamic analysis calibration – have been proposed for papillomaviruses, such as 1.0×10−7s/s/y (originally proposed for the L1 gene of mucosal HPV)29, 2.4×10−8 and 1.4×10−8s/s/y (from the Feline papillomavirus (FPV) for E6 and E7 genes, respectively)41. The use of these rates resulted in different diversification times of the viral types that belong to the alpha-9 species group, including HPV-58. For instance, at type level, the tMRCA of HPV-16, -33, -52 and -58 showed ancestral times within the last million years (∼0.6 to ∼0.8My) using the substitution rates proposed by Rector et al. or about 0.10–0.15My using those proposed by Halpern. Particularly, lineage A within HPV-58 would have diversified ∼65,800 or ∼300,000 years ago (depending on the calibration used), and the common ancestor of sublineages A1–A3 could have been dated ∼40,000 or ∼200,000 years ago, suggesting that their dispersion – even using the fastest substitution rate – accompanied or preceded human migration across the globe6.

This hypothesis could explain the worldwide distribution of sublineage A2, whereas other factors might be involved in the emergence and dispersion of the others groups, as previously proposed18. Is it worth noting that none of the substitutions rates used to calibrate are proposed for the E6 or E7 genes in HPV, thus these estimations should be taken with caution.

Summarizing, this study describes the molecular epidemiology of HPV in a region of Argentina with a high rate for cervical cancer mortality. The secondary type was HPV-58 (after HPV-16), sublineage A2. Although we found no association between this type and high pathogenicity, there are still some concerns about the extent of cross protection of the vaccines against the types not included in their design. In addition, we described genomic variants of HPV-58 and characterized the ancient origin of this type and its diversification, in the context of the alpha-9 papillomavirus species phylodynamics.

The authors declare that they have no conflicts of interest.