Fig 1.
Pre- and early post-ART viral dynamics.
Average pVL grouped by time to ART initiation (A). Area-under-curve of total pVL was calculated between infection and day 84 post infection (B). CA-DNA (C) and CA-RNA (D) accumulation between infection and day 12 post infection are shown. Groups are color-coded based on days post infection until ART: d3 (blue), d4 (red), d5 (green), d6 (purple), d7 (orange), d9 (black) and d12 (brown).
Fig 2.
Viral decay dynamics affected by timing of ART initiation.
Delay in first phase viral decay following ART initiation (A). The primary phase 1 decay rate using both a maximal decay rate (B) and a 7-day average rate (C) excluding 6 RMs with low pVL where rates could not be estimated. Decay kinetics of CA-DNA (D-F) and CA-RNA (G-I) within PBMC (circle), LNMC (square), DUO (up triangle) and BM (down triangle) from 3-, 70- or 250-days following ART initiation. Unfilled symbols represent samples with no viral signal detected and are plotted at the calculated limit of detection (LOD) for each sample based on specimen input. Groups are color-coded based on days post infection until ART: d3 (blue), d4 (red), d5 (green), d6 (purple), d7 (orange), d9 (black) and d12 (brown).
Fig 3.
SIV CA-DNA, SIV CA-RNA, and IPDA levels prior to ART interruption using CA-DNA and RNA values and IPDA.
Viral RNA and DNA copies/million cells from PBMC, mesenteric LN, and spleen just prior to ART release (A). Unfilled symbols represent samples with no viral signal detected and are plotted at the calculated LOD for each sample based on specimen input. IPDA vs CA-DNA measurements in PBMC (B) and mesenteric LNMCs (C). Groups are color-coded based on days post infection until ART: d3 (blue), d4 (red), d5 (green), d6 (purple), d7 (orange), d9 (black) and d12 (brown).
Fig 4.
Rebound kinetics and time to detection following ART discontinuation.
PVL values starting at treatment interruption through peak rebound (A). Non-rebounding RMs are shown with unfilled symbols and animal sex indicated in legend. Kaplan-Meier plot of time to detectable rebound (50 copies/ml) after treatment interruption (B). Groups are color-coded based on days post infection until ART: d3 (blue), d4 (red), d5 (green), d6 (purple), d7 (orange), d9 (black) and d12 (brown).
Fig 5.
Barcode assessment post rebound.
Rebound kinetics and barcodes for each rebounding RM are presented based on the post-infection day ART was initiated with the rebound pVL of each RM shown on the left and the relative barcode proportion at peak rebound shown on the right (A-F). Representative (d5 ART group) of barcode comparisons across time post rebound and compared to pre-ART population (G). LOD is determined by viral input levels and is shown as elevated x-axis levels. Groups are color-coded based on days post infection until ART: d3 (blue), d4 (red), d5 (green), d6 (purple), d7 (orange), d9 (black) and d12 (brown).
Fig 6.
Reactivation rates (events per day) of individual RMs within each treatment group are shown with RM-specific symbols (A). Reactivation rates from non-rebounding RMs are shown with open symbol set at LOD of 0.01 events/day. The pre-ART per virion reactivation rate is plotted for each individual RM using LOD and open symbols for RMs that did not rebound (B). No significant differences were found between RMs in group 5, 6 and 7 (p>0.4), however they were each significantly different from groups d9 and d12 (p<0.05), and the latter two groups were also significantly different from each other (p = 0.016). The average rate of d5, d6, and d7 was significantly different from the d9 RMs (12-fold, p<0.001) and d12 RMs (74-fold, p<0.001). Groups are color-coded based on days post infection until ART: d3 (blue), d4 (red), d5 (green), d6 (purple), d7 (orange), d9 (black) and d12 (brown).
Fig 7.
Phases of the accumulation of RCVR during primary infection.
The reactivation rate is plotted against pre-ART peak pVL which exhibits a logistic fit (orange curve) (A). The RMs with pre-ART peak VL within the range 4–6 Log10 are in the dynamic range (purple symbols). The rate of change of the best fitting logistic model is shown by green curve. RMs with pVL <4 Log10 (blue symbols) only rarely rebound and fall within the stochastic range, while those with >6 Log10 peak pre-ART pVL fall within the saturation range (pink symbols) where the RCVR accumulation slows. Non-rebounding RMs are shown with unfilled symbols. Comparison of the pre-ART per virion reactivation rate of RMs in the dynamic range (purple) versus the saturation range (pink) show that RMs in the dynamic range have significantly higher pre-ART per virion reactivation rate than those in saturation group (p<0.00001) (B).