HIV co receptor tropism was assessed from the filtered
HIV-1 co-receptor AICAR manufacturer was assessed from the filtered V3 MPS data using the g2p  tool and classified as X4 when there were more than 2% of the sequences with g2p false positivity rate (FPR) cutoffs of 3.75%, 2.5% and 1%. More detailed analysis was performed at FPR set at 2.5% to increase the X4 detection sensitivity without affecting the specificity. The values of the FPR used here are based on several studies that indicate the capacity of g2p algorithms to provide reliable discrimination between R5 and X4 sequences when FPR is set at lower values . In this study, a minority variant was defined as a variation detected at ≥2% and <20% of the virus MPS reads, corresponding to those mutations that cannot be established using the conventional sequencing technology. The 2% cutoff was established because it was found to be optimal to predict the clinical response .
Results Phylogenetic analysis of the NFLGs and larger partial fragment that cover the env gp 120 region were performed in HIV-1 infected donors with LS infection (n=193) and an unknown clinical stage (n=64) (submitted manuscript). After removal of scaffolding reads not covering the complete V3 region from partial fragments and exclusion of MPS data with low coverage reads (<500x) and poor quality reads, the number of samples was dropped to 84 samples and these were considered for analysis. The coverage after mapping of the sample to its corresponding consensus sequence varied among the patients showing an overall median sequence depth of 1030 (range: 509-6883). Of the 84 investigated subjects, 69 (82.1%) belonged to individuals carrying HIV-1 subtype B, 11 (13.4%) subclade F1, and 4 (4.8%) subtype C. The g2p algorithm1% predicted the occurrence of X4 strains in >94% of generated MPS reads of three (3.6%) participants (coverage depth range: 636–2300) (Table 1). No minority variants (X4 viruses at a frequency below 20%) were observed under this algorithm. At a g2p algorithm 2.5% and 3.75%, the CXCR4-using viruses were predicted in 13 (15.5%) and 20 (23.8%) subjects (coverage depth range: 564–2937), respectively. At the setting of g2p algorithm 2.5%, three participants (3.7%) had detectable X4 viruses in >99% of their MPS reads (coverage depth range: 636–2300). Furthermore, seven (8.3%) subjects were found to harbor minority X4 viruses at a frequency rate below 20% within their viral population (coverage depth range: 564–1868). Besides the seven participants with X4 minority variants, our analysis revealed other three other participants namely: 10BR_PE044, 10BR_PE106, and 10BR_PE024 to harbor CXCR4-using viruses at a frequency of 35.5%, 57.9%, and 65.1%, respectively. Taken together, these results indicate a high rate (11.9%) of R5×4 mixed infection in the studied population. Of note, all the 13 MPS data found to contain CXCR4-using viruses were characterized by phylogenetic analysis of the V3 region to belong to subtype B viruses. Analysis of the MPS data from the 63 subjects with LS infection revealed that 9 (14.3%) participants had CXCR4-using proviruses (coverage depth range: 564–2937). Among the 63 subjects, the minority variants and R5×4 viruses were detected in 4 (6.3%) and 6 (9.5%) subjects, respectively.
Discussion Recently we reported our initial findings of the prevalence of coreceptor tropism of the archived strains at the time of primary infection using a total of 45 MPS data from HIV-1 recently infected Brazilian first-time blood donors . In the present study, we expanded our previous work by determining the prevalence of CXCR4-using viruses in MPS data from a total of 84 Brazilian first-time blood donors consisting of 63 subjects with LS infection and 21 with unknown stage information. Some studies have reported the presence of X4 strains in recent HIV-1 seroconverter Spanish, Brazilian and French subjects [16,22] and in drug-naive chronically HIV-infected individuals , in immunosuppressed patients with a shorter history of viremia suppression , in patients failing antiretroviral therapy , and in patients with detectable HIV-1 subtype B RNA receiving highly active antiretroviral therapy . However, to our knowledge, only one relevant study to date has explored the coreceptor use in MPS data but from recently infected and therapy naive first-time Brazilian blood donors .