The Evolution of Regulatory Elements in the Emerging Promoter-Variant Strains of HIV-1 Subtype C

Abstract

In a multicentric, observational, investigator-blinded, and longitudinal clinical study of 764 ART-naïve subjects, we identified nine different promoter variant strains of HIV-1 subtype C (HIV-1C) emerging in the Indian population, with some of these variants being reported for the first time. Unlike several previous studies, our work here focuses on the evolving viral regulatory elements, not the coding sequences. The emerging viral strains contain additional copies of the existing transcription factor binding sites (TFBS), including TCF-1α/LEF-1, RBEIII, AP-1, and NF-κB, created by sequence duplication. The additional TFBS are genetically diverse and may blur the distinction between the modulatory region of the promoter and the viral enhancer. In a follow-up analysis, we found trends, but no significant associations between any specific variant promoter and prognostic markers, probably because the emerging viral strains might not have established mono infections yet. Illumina sequencing of four clinical samples containing a coinfection indicated the domination of one strain over the other and establishing a stable ratio with the second strain at the follow-up time points. Since a single promoter regulates viral gene expression and constitutes the master regulatory circuit with Tat, the acquisition of additional and variant copies of the TFBS may significantly impact viral latency and latent reservoir characteristics. Further studies are urgently warranted to understand how the diverse TFBS profiles of the viral promoter may modulate the characteristics of the latent reservoir, especially following the initiation of antiretroviral therapy.

Keywords: HIV-1; evolution; latency; sequence duplication; subtype C.

Figure 1

Profile of HIV-1C promoter variants in the Indian population. (A) The time of sample collection, sample number, and the nature of the promoter configuration are illustrated. The pie charts represent the percentage prevalence of the variant viral strains as color-coded – canonical HHC (beige), FHC (blue), FHHC (green), RR (duplication of RBEIII; red), and duplication of RBEIII like and NF-κB (grey). The data for the periods 2000–2003 and 2010–2011 are adapted from Bachu et al. (2012a) and replotted. The data of the present study are presented in the pie chart 2016–19. (B) The magnitude of TFBS variation in HIV-1C LTR. The upper panel represents the genome organization of HIV-1 followed by the TFBS arrangement in the canonical HIV-1C LTR. Sp1 motifs are depicted as grey circles, RBEIII motifs (R) as red triangles, and TCF-1α/LEF-1 sites (L) as open rectangular boxes. The various types of NF-κB binding sites (H, C, F, and h) are depicted as green square boxes. The various HIV-1C viral strains are classified into three main categories based on the NF-κB and/or RBEIII motif duplication. (I) The 3-κB LTR viral strains. (II) The canonical 4-κB LTR viral strains. (III) The viral strains containing the RBEIII site duplication. The two RBEIII sites are separated by an interceding sequence that constitutes an additional copy of a κB-motif (H), κB-like motif (h), TCF-1α/LEF-1 motif (L), or sequence without a distinct pattern (X). The analysis represents 455 of the 518 LTR sequences, and we could not type 63 other sequences.

Figure 2

The nature of the RBEIII cluster duplication in HIV-1C. (A) A schematic representation of the RBEIII cluster duplication in the RR group of HIV-1C variants. The top panel depicts the arrangement of TFBS in the canonical HHC-LTR. The bottom panel portrays the RBEIII cluster duplication (RR, two RBEIII clusters) in two variant LTRs – the co-duplication of RBEIII and TCF-1α/LEF motifs or RBEIII and NF-κB motifs. Of note, the RBEIII cluster duplication comprises the copying of a sequence that recruits both RBF-2 and AP-1 (c-Jun and ATF) collectively. The original and duplicated sequences are marked with solid and dotted arrows, respectively. The two RBEIII clusters are typically separated from each other by an intervening sequence that constitutes a binding site for NF-κB (a canonical H-κB site or a non-canonical h-κB site), TCF-1α/LEF-1 motif, or a sequence of undefined character. (B) The relative prevalence of the RBEIII and/or NF-κB motif duplication in HIV-1 subtypes. All the available LTR sequences containing the duplication of one or both TFBS were downloaded from the LANL HIV database and categorized under the major HIV-1 subtypes as shown. A single sequence per patient was included in the analysis. The grey and black bars represent the prevalence of duplication of RBEIII and NF-κB motifs, respectively. n: the number of sequences, CRF: Circulating recombinant forms.

Figure 3

Phylogenetic analysis of HIV-1C LTR variant sequences. A total of 455 viral sequences isolated from study participants are included in the analysis. The analysis also includes four HIV-1C reference sequences and three sequences representing each primary genetic subtype of HIV-1, as described in materials and methods. Different LTR variants are represented using different symbols and colors, as depicted. The evolutionary history was inferred by using the maximum likelihood method based on the Tamura-Nei model. The analysis was performed with 1,000 bootstrap values. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. There were a total of 292 positions in the final data set. Evolutionary analyses were performed using MEGA 7.0 software.

Figure 4

Cross-sectional and longitudinal analyses of prognostic markers. (A) Plasma viral load, (B) CD4 cell count, and (C) soluble CD14 levels of the four groups are presented at the baseline (left panels) and follow-up points (right panels). An available-case and complete-case analysis is presented for all the prognostic markers. The sample size used in each evaluation and the corresponding statistics are presented, in the tables. Given the limited sample numbers, the seven RR groups were pooled into a single double-RBEIII arm. Different LTR variant types are represented using different symbols and colors, as depicted. A non-parametric test, that is, the Kruskal-Wallis test, was applied for the statistical analysis of the plasma viral load. One-way ANOVA with Dunnett multiple comparison test was applied to CD4 cell count and sCD14. Two-way ANOVA was used for the comparison of the longitudinal analysis.

Figure 5

The frequencies of single- and double-RBEIII variants in a subset of study participants. (A) Two independent analyses were performed (replicates 1 and 2) using both the whole blood genomic DNA and plasma viral RNA. The samples were collected at six-month intervals, as shown. An asterisk (*) represents the samples collected post-ART. The dark, grey, and hollow bars represent the percentage prevalence of double-RBEIII, single-RBEIII, and minority/un-typable viral strains, respectively. (B) Multiple sequence alignment of single- and double-RBEIII promoter variants in respective subjects, as indicated. TFBS of relevance are marked using open square boxes. The viral variants are aligned with the Indie.C1 reference sequence, which was pulsed in the sequencing sample as an internal control. Dashes and dots represent sequence deletion and sequence homology, respectively. (C) Prognostic markers, plasma viral load (PVL), and CD4 cell count are represented by a filled box with a solid line and an open box with a dotted line, respectively.