Human immunodeficiency virus type 1 superinfection was not detected following 215 years of injection drug user exposure

Abstract

Evidence for human immunodeficiency virus type 1 (HIV-1) superinfection was sought among 37 HIV-1-positive street-recruited active injection drug users (IDUs) from the San Francisco Bay area. HIV-1 sequences from pairs of samples collected 1 to 12 years apart, spanning a total of 215 years of exposure, were generated at p17 gag, the V3-V5 region of env, and/or the first exon of tat and phylogenetically analyzed. No evidence of HIV-1 superinfection was detected in which a highly divergent HIV-1 variant emerged at a frequency >20% of the serum viral quasispecies. Based on the reported risk behavior of the IDUs and the HIV-1 incidence in uninfected subjects in the same cohort, a total of 3.4 new infections would have been expected if existing infection conferred no protection from superinfection. Adjusted for risk behaviors, the estimated relative risk of superinfection compared with initial infection was therefore 0.0 (95% confidence interval, 0.00, 0.79; P = 0.02), indicating that existing infection conferred a statistically significant level of protection against superinfection with an HIV-1 strain of the same subtype, which was between 21 and 100%.

FIG. 1.

Time span analyzed for HIV-1 superinfection in 37 IDUs.

FIG. 2.

Example of HMA clonotype analysis of p17 sequence variants. PCR products from plasmid subclones were reannealed, and DNA heteroduplexes were resolved through a polyacrylamide gel. (A) Inter-IDU HMA. A clonal PCR product from a subclone from one IDU was reannealed to clonal PCR amplicons from nine other IDUs, showing that different sequence variants produced distinct electrophoretic mobilities. (B) Intra-IDU HMA. A clonal PCR product from one subclone from an IDU was reannealed to PCR products derived from subclones from the other time point. Subclones producing the same mobility heteroduplexes were classified as members of the same clonotype groups (A through E).

FIG. 3.

ML phylogenetic analysis. (A) V3-to-V5 region of env. (B) First exon of tat. (C) p17. The IDU subjects are represented by letters followed by years of sampling and, following the first “x,” the clone number. Bootstrap values for the nodes holding all sequence variants from any one IDU are shown in large font, while other bootstrap values are shown in smaller fonts. For p17 the total number of sequence variants belonging to the same HMA clonotype group is noted following the second “x”; e.g., Y91x3x4 indicates subject Y sampled in 1991, subclone number 3, belonging to an HMA clonotype group consisting of four different subclones.

FIG. 4.

Intra- and interpatient distribution of pairwise genetic distances for the tat (A) and p17 (B) regions. The intrasubject (black lines) and intersubject (white lines) pairwise HIV-1 genetic distances for San Francisco Bay area IDUs whose variants all clustered with bootstrap values >50, as well as the intersubject distances from unrelated sequences from the HIV database (grey lines), are plotted. Arrows, intrasubject tat genetic distances for the two time points sequenced which clustered phylogenetically together with bootstrap values below 50; horizontal lines, means and ranges of the p17 intrapatient genetic distances for variants that clustered phylogenetically together with bootstrap values below 50. The phylogenetically linked subject pair AE and J was omitted from the intersubject tat distance pairs.