Isolation and characterization of replication-competent human immunodeficiency virus type 1 from a subset of elite suppressors

Abstract

Elite suppressors (ES) are untreated human immunodeficiency virus type 1 (HIV-1)-infected individuals who control viremia to levels below the limit of detection of current assays. The mechanisms involved in this control have not been fully elucidated. Several studies have demonstrated that some ES are infected with defective viruses, but it remains unclear whether others are infected with replication-competent HIV-1. To answer this question, we used a sensitive coculture assay in an attempt to isolate replication-competent virus from a cohort of 10 ES. We successfully cultured six replication-competent isolates from 4 of the 10 ES. The frequency of latently infected cells in these patients was more than a log lower than that seen in patients on highly active antiretroviral therapy with undetectable viral loads. Full-length sequencing of all six isolates revealed no large deletions in any of the genes. A few mutations and small insertions and deletions were found in some isolates, but phenotypic analysis of the affected genes suggested that their function remained intact. Furthermore, all six isolates replicated as well as standard laboratory strains in vitro. The results suggest that some ES are infected with HIV-1 isolates that are fully replication competent and that long-term immunologic control of replication-competent HIV-1 is possible.

FIG. 1.

Phylogenetic analysis of ES clones. A phylogenetic tree of env sequences from ES1, -4, -8, and -10 (circles) and from U.S. clade B patients with progressive disease (triangles) was constructed by using a maximum-likelihood method. Colors indicate the year of isolation of each virus. Representative clade A and C env's are included for comparison.

FIG. 2.

Nef and LTR sequences of replication-competent viruses from ES. Sequences are aligned with the clade B consensus sequence. Dots indicate identity with the consensus sequence. Dashes indicate gaps. Numbering is by HXB2 coordinates. Polymorphisms detected in at least two of four ES and at least one third of ES clones but less than 10% of clade B sequences from the Los Alamos database are shaded in pink. (A) Nef sequences from ES show no stop codons or major deletions. The virus from ES10 had a two-amino-acid deletion (positions 9 and 10) that is also seen in 3 to 12% of sequences in the database and a seven-amino-acid tandem duplication resulting in an insertion at position 21 where there are insertions in 37% of clade B sequences. ES10 virus also had a single amino acid insertion after position 64. This same glutamic acid insertion is also seen in 17% of clade B sequences. Clones from ES1 and ES10 had a K→Q polymorphism at position 115 that was seen in <10% of clade B sequences in the database. None of these insertions, deletions, or polymorphisms affected Nef function as assessed by CD4 downregulation (see Fig. 2C). (B) Critical regions of the LTR are conserved in viruses from ES. LTR sequences appear normal with the exception that both clones from ES8 had a single nucleotide deletion in one of the two NF-κB binding sites. In addition, they had a C→T mutation in the Tar bulge region which is not seen in clade B viruses but which has been observed in other clades. There was also a single nucleotide insertion in the 5′ untranslated region of the mRNA in these isolates. These polymorphisms did not appear to affect the activation of transcription by NF-κB or Tat (see Fig. 2D).

FIG. 3.

Sequences of HIV-1 clones from ES. Full-length sequences of six HIV-1 clones from ES were aligned with clade B consensus sequences. An alignment of nucleotides sequences is shown for the LTR. For coding regions, alignments of the amino acid sequences are shown. Dots indicate identity to the consensus sequence. Dashes indicate gaps. Numbering is by HXB2 coordinates. Potential N-linked glycosylation sites (N-X-S/T, where “X” is any amino acid except praline) in Env are indicated with blue boxes.

FIG. 3.

Sequences of HIV-1 clones from ES. Full-length sequences of six HIV-1 clones from ES were aligned with clade B consensus sequences. An alignment of nucleotides sequences is shown for the LTR. For coding regions, alignments of the amino acid sequences are shown. Dots indicate identity to the consensus sequence. Dashes indicate gaps. Numbering is by HXB2 coordinates. Potential N-linked glycosylation sites (N-X-S/T, where “X” is any amino acid except praline) in Env are indicated with blue boxes.

FIG. 3.

Sequences of HIV-1 clones from ES. Full-length sequences of six HIV-1 clones from ES were aligned with clade B consensus sequences. An alignment of nucleotides sequences is shown for the LTR. For coding regions, alignments of the amino acid sequences are shown. Dots indicate identity to the consensus sequence. Dashes indicate gaps. Numbering is by HXB2 coordinates. Potential N-linked glycosylation sites (N-X-S/T, where “X” is any amino acid except praline) in Env are indicated with blue boxes.

FIG. 3.

Sequences of HIV-1 clones from ES. Full-length sequences of six HIV-1 clones from ES were aligned with clade B consensus sequences. An alignment of nucleotides sequences is shown for the LTR. For coding regions, alignments of the amino acid sequences are shown. Dots indicate identity to the consensus sequence. Dashes indicate gaps. Numbering is by HXB2 coordinates. Potential N-linked glycosylation sites (N-X-S/T, where “X” is any amino acid except praline) in Env are indicated with blue boxes.

FIG. 3.

Sequences of HIV-1 clones from ES. Full-length sequences of six HIV-1 clones from ES were aligned with clade B consensus sequences. An alignment of nucleotides sequences is shown for the LTR. For coding regions, alignments of the amino acid sequences are shown. Dots indicate identity to the consensus sequence. Dashes indicate gaps. Numbering is by HXB2 coordinates. Potential N-linked glycosylation sites (N-X-S/T, where “X” is any amino acid except praline) in Env are indicated with blue boxes.

FIG. 3.

Sequences of HIV-1 clones from ES. Full-length sequences of six HIV-1 clones from ES were aligned with clade B consensus sequences. An alignment of nucleotides sequences is shown for the LTR. For coding regions, alignments of the amino acid sequences are shown. Dots indicate identity to the consensus sequence. Dashes indicate gaps. Numbering is by HXB2 coordinates. Potential N-linked glycosylation sites (N-X-S/T, where “X” is any amino acid except praline) in Env are indicated with blue boxes.

FIG. 4.

Functional activity of HIV-1 clones isolated from ES. (A) Growth kinetics of six clones from ES compared to the BA-L and IIIB reference strains. (B) Growth kinetics of reference strains and autologous HIV-1 isolates in CD4⁺ T-cell lymphoblasts from ES1, ES4, and ES8.

FIG. 5.

Functional activity of HIV-1 genes with mutations. (A) CD4 downregulation in lymphoblasts infected with either ES10-53 or the IIIB reference strain. Infected cells were stained with fluorescein isothiocyanate-conjugated antibodies to CD4 and then fixed, permeabilized, and stained for intracellular HIV-1 Gag antigen. The numbers indicate the percentage of infected (Gag positive) cells that have either normal or downregulated levels of CD4 on their surface. (B) Functional activity of the LTR from ES8-43 (red) and the reference clone NL4-3 (blue). The fold increase in transcriptional activity in response to TNF-α or Tat stimulation is shown. (C) Functional activity of Vif from clone ES1-16 grown in PBMC. A region of the genome that is subject to a high rate of hypermutation in the absence of Vif was amplified from the DNA of lymphoblasts infected with ES1-16 by using limiting dilution PCR. 30 independent clones were sequenced. Only a single clone (arrow) had evidence of G-to-A hypermutation.