Multiple NF-κB sites in HIV-1 subtype C long terminal repeat confer superior magnitude of transcription and thereby the enhanced viral predominance

Abstract

We demonstrate that at least three different promoter variant strains of HIV-1 subtype C have been gradually expanding and replacing the standard subtype C viruses in India, and possibly in South Africa and other global regions, over the past decade. The new viral strains contain an additional NF-κB, NF-κB-like, or RBEIII site in the viral promoter. Although the acquisition of an additional RBEIII site is a property shared by all the HIV-1 subtypes, acquiring an additional NF-κB site remains an exclusive property of subtype C. The acquired κB site is genetically distinct, binds the p50-p65 heterodimer, and strengthens the viral promoter at the levels of transcription initiation and elongation. The 4-κB viruses dominate the 3-κB "isogenic" viral strains in pairwise competition assays in T-cell lines, primary cells, and the ecotropic human immunodeficiency virus mouse model. The dominance of the 4-κB viral strains is also evident in the natural context when the subjects are coinfected with κB-variant viral strains. The mean plasma viral loads, but not CD4 counts, are significantly different in 4-κB infection suggesting that these newly emerging strains are probably more infectious. It is possible that higher plasma viral loads underlie selective transmission of the 4-κB viral strains. Several publications previously reported duplication or deletion of diverse transcription factor-binding sites in the viral promoter. Unlike previous reports, our study provides experimental evidence that the new viral strains gained a potential selective advantage as a consequence of the acquired transcription factor-binding sites and importantly that these strains have been expanding at the population level.

FIGURE 1.

Transcription factor-binding site polymorphism in the viral promoter of HIV-1 subtype C. A, progressive increase in the prevalence of the variant viral strains across India during the periods 2000–2003, 2005, and 2010–2011. The time of sample collection, geographic location, sample number, and the nature of the insertion have been illustrated. All the clinical samples below the dashed horizontal line were collected during 2010–2011 from four different clinics in India. The 2000–2003 (southern India) and 2005 (Jawaharlal Nehru Centre for Advanced Scientific Research) clinical cohorts have been reported previously. The nature of the sequence insertion in the viral promoter of each viral isolate was determined by sequencing the PCR product directly or after cloning into a plasmid vector. The pie charts represent the percentage prevalence of the variant viral strains in a color-coded fashion, wild type 3-κB (cream), 4-κB (red), κB-like (pink), RBE-III (green), and κB-like plus RBE-III (yellow). B, molecular nature of the sequence insertion in a South African clinical cohort as reported in Rousseau et al. (34).

FIGURE 2.

Molecular nature of the NF-κB site acquisition and distribution among HIV-1 genetic subtypes. A, multiple sequence alignment of the viral enhancer and upstream sequences of the representative 4-κB viral isolates. The sequences presented here are representative of 159 sequences determined through this work and 54 sequences downloaded from the databases. The κB motifs have been highlighted in red shading, and the RBE-III site in green shading The original (blue arrow) and the duplicated (red arrow) 21-residue sequences are shown. The C and T variations of the respective κB sites are highlighted in black shading and yellow letters. The cohort identity, sample collection time, geographic location, and the total number of sequences available are shown. The asterisk represents the viral sequences derived through this work. B, global distribution of the MFNLP insertions. A large number of the LTR sequences belonging to the various genetic subtypes of HIV-1 and containing any one of the three (NF-κB, NF-κB like, or RBEIII) sequence insertions were downloaded from the extant databases. The color code of the sequence insertions is consistent with that of the Fig. 1. The subtype identity, total number of sequences used in the present analysis, and the number of viral isolates containing additional NF-κB or RBE-III sites are shown. The percent prevalence of the variant viral isolates under each genetic subtype is illustrated.

FIGURE 3.

F-κB site is an authentic NF-κB motif. A, H- and F-κB probes bind cellular factors in the electrophoretic mobility shift assay. Radiolabeled double-stranded probes containing the H- or F-κB sites were incubated with Jurkat nuclear extracts prepared from cells after 60 min of TNFα activation. Cold competition was performed by preincubating the nuclear extracts with 25 m excess of cold oligonucleotides consisting of H, F, IκB-α, or a mutant (Mut) κB site. A probe for the Oct-1 cellular factor was used as a loading control. Free probe (FP) and nonspecific (NS) and specific complexes are indicated. B, H- and F-κB probes bind the p50-p65 heterodimer in the supershift assay. Nuclear extracts prepared from control or TNFα-treated Jurkat cells were preincubated with affinity-purified rabbit antibodies specific to the Rel family members as indicated at the top of the lanes. EMSA was performed as described in A. C, binding affinity determination of the H- and F-κB probes for the recombinant p50 protein in isothermal calorimetry. Double-stranded H-, F- or a mutant κB site containing oligonucleotides were used in the assay. The raw ITC traces from a representative titration are presented with respect to the base line, and the heat change versus the molar ratio of the titrated products is plotted. D, chromatin immunoprecipitation analysis. A schematic representation of the VSV-G pseudotyped viruses expressing a dual-expression cassette of EGFP and Tat is illustrated in the upper panel. Jurkat cells were infected at an m.o.i. of 1 and 72 h later stimulated with TNF-α for 60 min (left panel) or left without activation (middle panel). Immunoprecipitation of the complexes was performed using 5 μg of antibodies as indicated. A reference IκB-α cellular promoter containing three NF-κB sites was used as a positive control for the ChIP assay. Enrichment of the NF-κB subunits and transcription-competent RNA polymerase II on the viral (HHHH- or FFFF-LTR) or the cellular promoter was evaluated using PCR. One-tenth of the input chromatin was uncross-linked and used as an input control. IgG, isotype-matched control antibody. Enrichment of the Ser-2 phosphorylated form of the RNA polymerase II subunit was evaluated by targeting the Tat region located ∼3,000 bp downstream of the TSS of the viral promoters (right panel). The GAPDH cellular promoter was used as a positive control.

FIGURE 4.

F-κB site confers quantitative gain-of-function on the C-LTR. A, schematic representation of the dual-expression vectors. The isogenic LTRs originated from a representative subtype C LTR BL42 (FHHC). Reference LTRs from subtype B (NL4.3) and subtype C (Indie-C1) were also included for comparison. B, induced reporter gene expression from the viral promoters in the absence of Tat. Jurkat cells were transfected with one of the reporter vectors illustrated in A above and 12 h later were subjected to diverse activation conditions as indicated, and the luciferase secretion at 24 h was evaluated from the medium. Each assay was performed in triplicate wells, and the data are presented as mean relative light units ± S.D. The data are from one of the three representative experiments. C, induced reporter gene expression from the viral promoters in the presence of Tat. Jurkat cells were co-transfected with a plasmid pool containing a Tat expression vector and one of the reporter vectors. D, induced reporter gene expression from the viral promoters under synergistic activation conditions. Jurkat cells were treated as in B above and subjected to combinations of two or three different activation conditions as indicated.

FIGURE 5.

4-κB viruses out-compete the 3-κB strains in T-cells A, schematic representation of the paired isogenic viral constructs used in the competition assays. The 22-bp sequence consisting of the F-κB site was engineered into the 3′-LTR of the Indie-C1 molecular clone (HHC) to generate the 4-κB LTR (FHHC). Replication profiles of the viral strains in vitro using CEM-CCR5 T-cells (B) or the CD8-depleted, mitogen-activated PBMC (C) from a representative subject. The cells were infected with 500 infectious units of FHHC or HHC viruses, and the secretion of p24 into the medium was monitored for several weeks as indicated. The data are presented as the mean of triplicate wells ± 1 S.D. and are representative of three independent experiments. Pairwise competition between the HHC and FHHC isogenic viruses in CEM-CCR5 cells (D) or activated PBMC (E). The paired viruses were competed against each other at different ratios of m.o.i. as schematically depicted in supplemental Fig. 4. Genomic DNA was extracted on day 10 following the viral infection and subjected to the HTA analysis. The heteroduplex band intensities are plotted as relative values compared with the monoinfections (see formula supplemental Fig. 4D).

FIGURE 6.

4-κB viruses out-compete the 3-κB strains in a small animal model and in the natural infection. A, schematic representation of the paired isogenic viral constructs used in the competition assays. The 22-bp sequence consisting of the F-κB site was engineered into the 3′-LTR of the EcoHIV molecular clone (HHC) to generate the 4-κB LTR (FHHC). B, pairwise competition between the HHC and FHHC isogenic viruses in the EcoHIV-mouse model. Mice, four animals per group, were infected intraperitoneally with a total of 5 μg of p24 equivalent of the viruses using the strategy schematically depicted in supplemental Fig. 4A. Two weeks later, genomic DNA was extracted from splenocytes and subjected to the HTA analysis. The data are representative of three independent experiments. C, relative prevalence of the 3- and 4-κB viral variants in the genomic DNA of the clinical samples. Genomic DNA was extracted from PBMC of six different subjects all containing a mixed infection with 3- and 4-κB viral strains. The relative band intensities of the HHC and FHHC viruses were determined in HTA. The mean value of band intensities of the 4-κB strains is statistically significant, p < 0.001. D, relative prevalence of the 3- and 4-κB viruses in the plasma viral RNA of three different subjects. Plasma viral RNA was reverse-transcribed, and the proportion of the 3- and 4-κB viruses was determined as above. Longitudinal analysis of the plasma samples separated by a year from a single patient is shown (right panel).

FIGURE 7.

Comparison of the post-entry events between the 3- and 4-κB isogenic viruses. A, schematic representation of the early events of HIV replication in CEM-CCR5 T-cells following infection with the HHC or FHHC isogenic viruses. Viruses were generated as in Fig. 5A. The filled boxes represent regions targeted for amplification. B, quantification of the late reverse transcription products. Cellular DNA was extracted from infected CEM-CCR5 cells 12 h following viral infection, and a 142-bp sequence within the U5-ψ region was amplified. The data are presented as the mean of three independent reactions ± S.D. C, quantification of the 2-LTR circles. Cellular DNA was extracted from infected CEM-CCR5 cells 24 h following viral infection, and a 223-bp sequence across the U5-U3 junction was amplified. D, quantification of the integrated proviruses using a nested Alu-LTR PCR. Cellular DNA was extracted from infected CEM-CCR5 cells 48 h following viral infection, and a 200-bp sequence within the R-U5 region was amplified. ns, not significant. E, quantification of the proximal and distal viral transcripts. Forty eight hours following infection of CEM-CCR5 cells with the HHC or FHHC isogenic viruses, total RNA was extracted; first strand cDNA was synthesized, and the quantitative real time PCR was performed as described. Proximal transcripts were detected with primers amplifying an 89-bp fragment in the TAR region, and the distal transcripts with primers to amplify a 226-bp fragment in Tat.

FIGURE 8.

Comparative analysis of the correlates of disease progression between the 3- and 4-κB viral infections. Subjects categorized by the presence of 3- or 4-κB viral infections have been drawn from all the four Indian clinical cohorts. Each point in the plot represents an individual HIV-1 seropositive subject. A, plasma viral load; B, CD4 cell count of the subjects are presented in each group. The number of subjects, means, and medians of the data and standard deviation values under each group are shown. The horizontal lines represent group median values. C, schematic model depicting a novel strategy the 4-κB viral strains use to enhance their replication competence by enhancing the infectivity.