HPV16 is the viral type with the highest association to HSIL and ICC; it was one of the first HPV's to be sequenced [17, 18], and it is well known that infection by certain HPV16 subtypes and variants can lead to faster disease progression in younger women [19, 20].
HPV subtypes have been identified by comparing the sequences from the E6 and L1 genes and of the long control region (LCR). In the case of HPV16 the E6 gene is frequently used because a short and continuous fragment of its sequence contains sufficient information to identify all the subtypes and variants that have been described [10, 21–24]. HPV16 subtypes differ in prevalence, biochemical and biological properties (e.g., replication and expression of AA E6 and E7 oncogenes is more efficient) with uncertain implications in cervical cancer aetiology [12, 25–27].
In this work, the first to approach HPV molecular phylogeny in Central Mexico, we identified HPV16 subtypes and variants by comparing the E6 ORF sequences . Through the use of nested PCR enough E6-2 DNA was obtained to sequence the amplified samples. Amplicon length was heterogeneous because the length of the product ends is variable with the sequencing method used . Sequence analysis confirmed that all samples correspond to HPV16, supporting the specificity of the PCR-RFLP method  we used for genotyping .
Among the HPV16 E6-2 amplicon sequences, 36 corresponded to the E subtype and two to the AA subtype. No African subtypes were identified, as has been the case in previous studies performed in Mexico City [12, 14].
Berumen et al.  found a 1.1% prevalence of the HPV16 AA subtype in controls and 23.2% in ICC cases in Mexico City, whereas del Refugio Gonzalez-Losa et al.  did not find the HPV16 AA subtype. These differences are probably due to the severity of the lesions included since the group of samples analyzed by Berumen et al.  had ICC, whereas in this work LSIL predominated.
We identified four HPV16 variants in San Luis Potosi City: E-P (n = 27, 71.1%), E-T350G (n = 7, 18.4%), E-C188G (n = 2, 5.3%) and AA-a (n = 2, 5.3%), whereas Berumen et al.  observed frequencies of 47% for the E-T350G and 8% for the E-350T variants. The contrast in the variety of subtypes and variants found by us probably derive also from differences in the kind of lesions as observed above; however, it cannot be ruled out that specific population features can be involved, since we have already found divergence among HPV type frequencies in the neighboring Mexican states of San Luis de Potosí and Guanajuato  which also differ from those observed in Mexico City [12, 14].
Besides the known point mutations characteristic of the HPV16 variants identified by us, we detected 24 novel single nucleotide changes, two of which appeared in a considerable proportion of the sequences analyzed. The non synonymous A404T substitution, observed only in one sequence close to the E-P Ref, generates the I101F amino acid change in the E6 protein. The second most frequent novel change, observed in 13 cases (34.2%), consists in a deletion of two contiguous bases (AC) in the nucleotide positions 56 and 57 located in the 5'-untranslated region of the E6 gene.
The synonymous A334G, the most frequent of the novel single nucleotide changes found in 22 E subtype sequences (57.9%) is phylogenetically close to the E-P Ref prototype and appear to identify an HPV16 variant characteristic of the region. The E-P A334G variant appears to be more oncogenic, because the proportion of HSIL and ICC lesions were clearly higher for them than with the E-P Ref variants (Table 4). To verify if E-P A334G sequences indeed correspond to a new variant, the complete viral genome must be cloned and sequenced .