Background — HIV-1 nucleotide substitution rates are central for understanding the evolution of HIV-1. Their accurate estimation is critical for analysis of viral dynamics, identification of divergence time of HIV variants, inference of HIV transmission clusters, and modeling of viral evolution.

Methods — Intra-patient nucleotide substitution rates in HIV-1C gag and env gp120 V1C5 were analyzed in a longitudinal cohort of 32 individuals infected with a single viral variant. Viral quasispecies were derived by single genome amplification/sequencing from serially sampled blood specimens collected at median (IQR) of 5 (4–6) times per subject from enrollment (during Fiebig stages II to V) over a median (IQR) of 417 (351–471) days post-seroconversion (p/s). HIV-1C evolutionary rates were estimated by BEAST v.1.6.1 using a relaxed lognormal molecular clock model. The effect of antiretroviral therapy (ART) on substitution rates in gag and env was assessed in a subset of six individuals who started ARV therapy during the follow-up period.

Results — During primary HIV-1C infection, the intra-patient substitution rates were estimated at a median (IQR) of 5.22E-03 (3.28E-03–7.55E-03) substitutions per site per year of infection within gag, and 1.58E-02 (9.99E-03–2.04E-02) substitutions per site per year within env gp120 V1C5. The substitution rates in env gp120 V1C5 were higher than in gag (p<0.001, Wilcoxon signed rank test). The median (IQR) relative rates of evolution at codon positions 1, 2, and 3 were 0.73 (0.48–0.84), 0.67 (0.52–0.86), and 1.54 (1.21–1.71) in gag, and 1.01 (0.86–1.15), 1.05 (0.99– 1.21), and 0.86 (0.67–0.94) in env gp120 V1C5, respectively. A first to the third position codon rate ratio > 1.0 within env was found in 25 (78.1%) cases, but only in 4 (12.5%) cases in gag, while a second to the third position codon rate ratio > 1.0 in env was observed in 26 (81.3%) cases, but in gag only in 2 (6.3%) cases (p<0.001 for both comparisons, Fisher’s exact test). No ART effect on substitution rates in gag and env was found, at least within the first 3–4 months after ART initiation. Individuals with early viral set point ≥ 4.0 log10 copies/ml had higher substitution rates in env gp120 V1C5 (median (IQR) 1.88E-02 (1.54E-02–2.46E-02) vs. 1.04E (7.24E-03–1.55E-02) substitutions per site per year; p=0.017, Mann-Whitney sum rank test), while individuals with early viral set point ≥ 3.0 log10 copies/ml had higher substitution rates in gag (median (IQR) 5.66E-03 (3.45E-03–7.94E-03) vs. 1.78E-03 (4.57E-04–5.15E-03); p=0.028; Mann-Whitney sum rank test).

Conclusions —The results suggest that in primary HIV-1C infection, (1) intra-host evolutionary rates in env gp120 V1C5 are about 3-fold higher than in gag; (2) selection pressure in env is more frequent than in gag; (3) initiation of ART does not change substitution rates in HIV-1C env or gag, at least within the first 3–4 months after starting ART; and (4) intra-host evolutionary rates in gag and env gp120 V1C5 are higher in individuals with elevated levels of early viral set point.