Measuring turnover of SIV DNA in resting CD4+ T cells using pyrosequencing: implications for the timing of HIV eradication therapies

Abstract

Resting CD4+ T cells are a reservoir of latent HIV-1. Understanding the turnover of HIV DNA in these cells has implications for the development of eradication strategies. Most studies of viral latency focus on viral persistence under antiretroviral therapy (ART). We studied the turnover of SIV DNA resting CD4+ T cells during active infection in a cohort of 20 SIV-infected pigtail macaques. We compared SIV sequences at two Mane-A1*084:01-restricted CTL epitopes using serial plasma RNA and resting CD4+ T cell DNA samples by pyrosequencing, and used a mathematical modeling approach to estimate SIV DNA turnover. We found SIV DNA turnover in resting CD4+ T cells was slow in animals with low chronic viral loads, consistent with the long persistence of latency seen under ART. However, in animals with high levels of chronic viral replication, turnover was high. SIV DNA half-life within resting CD4 cells correleated with viral load (p = 0.0052) at the Gag KP9 CTL epitope. At a second CTL epitope in Tat (KVA10) there was a trend towards an association of SIV DNA half-life in resting CD4 cells and viral load (p = 0.0971). Further, we found that the turnover of resting CD4+ T cell SIV DNA was higher for escape during early infection than for escape later in infection (p = 0.0084). Our results suggest viral DNA within resting CD4 T cells is more labile and may be more susceptible to reactivation/eradication treatments when there are higher levels of virus replication and during early/acute infection.

Figure 1. Analysis of escape at the KP9 Gag CTL epitope by pyrosequencing.

A. Estimation of K165R KP9 escape in serial resting CD4+ T cell SIV DNA samples using pyrosequencing compared to KP9-specific qRT-PCR. Six representative macaque examples comparing CTL escape at KP9 from serial resting CD4+ T cell SIV DNA samples after infection with SIV_mac251 determined using the KP9-specific qRT-PCR compared to pyrosequencing. B. KP9 escape in plasma SIV RNA and resting CD4+ T cell SIV DNA for 2 representative macaques using Roche 454 sequencing. Examples of KP9 CTL escape in plasma SIV RNA and resting CD4+ T cell SIV DNA 2 animals using pyrosequencing. The CTL amino acid sequence is shown in the first column, with the % of sequence in the subsequent columns and the time point post SIV challenge at the top of the column. The mutation identified is shown at each time point with the total reads shown in the bottom row. Common variants at each time point are shaded with rarer variants accounting for the remaining sequences.

Figure 2. Estimating the half-life of SIV DNA in resting CD4+ T cells studying KP9 escape using pyrosequencing data.

The proportion of WT virus in plasma (green circles), the fraction of WT virus estimated from area under the curve (AUC) of viral load (blue circles) and the experimentally observed fraction of WT virus SIV DNA in resting CD4+ T cells (red squares) for each animal in the top of each figure. The black line represents the line of best-fit SIV DNA half-life to the observed fraction of WT virus in resting CD4+ T cells for each animal. Animals are arranged in the order of increasing estimated lifespan. Total plasma viral loads (log₁₀ scale, from 10–10⁹) are illustrated in the bottom part of each figure (black triangles). The lack of data at crucial time points made it impossible to estimate the life spans of resting infected cells in 9 out of 20 animals.

Figure 3. Half-life of resting CD4+ T cell SIV DNA decreases with increasing chronic plasma viral load.

The chronic plasma viral load (geometric mean viral load from day 100 post-infection) is significantly negatively correlated with the estimated half-life of SIV DNA for each animal. Half-life estimated using (A) pyrosequencing data (Spearman correlation, r = −0.7817, p = 0.0052) and (B) KP9-specific q-RT-PCR data (Spearman correlation, r = −0.8358, p<0.0001). C. Half-life estimations' (days) using pyrosequencing compared to the KP9-specific qRT-PCR are significantly correlated (Spearman correlation, r = 0.6682, p = 0.0297).

Figure 4. KVA10 escape for plasma SIV RNA and resting CD4+ T cell SIV DNA using pyrosequencing.

Examples of KVA10 CTL escape in plasma SIV RNA and resting CD4+ T cell SIV DNA in 2 animals by pyrosequencing. The CTL amino acid sequence is shown in the first column, with the % of sequence in the subsequent columns and the time point post SIV challenge at the top of the column. The mutation identified is shown at each time point with the total reads shown in the bottom row. Common variants at each time point are shaded and rarer variants account for the remaining sequences.

Figure 5. Estimating the half-life of SIV DNA in resting CD4+ T cells studying KVA10 escape using pyrosequencing data.

A. The proportion of WT virus in plasma (green circles), the fraction of WT virus estimated from area under the curve (AUC) of viral load (blue circles) and the experimentally observed fraction of WT virus SIV DNA in resting CD4+ T cells (red squares) for each animal in the top of each figure. The black line represents the line of best-fit SIV DNA half-life to the observed fraction of WT virus in resting CD4+ T cells for each animal. Animals are arranged in the order of increasing estimated lifespan. Total plasma viral loads (log₁₀ scale, from 10–10⁷) are illustrated in the bottom part of each figure (black triangles). 12 animals with sufficient data on RNA and DNA escape were available to study. B. Correlation between half-life of SIV DNA in resting CD4+ T cells with chronic plasma viral load for KVA10 epitope using pyrosequencing. Spearman correlation, r = −0.4138, p = 0.0971.

Figure 6. Turnover of SIV DNA in resting CD4+ T cells is higher during early infection.

Relationship between timing of escape in plasma SIV RNA (time to reach 50% escape) at the KP9 epitope using the qRT-PCR and turnover of SIV DNA in resting CD4+ T cells. This analysis used data on 18 animals.