Nuclear retention of multiply spliced HIV-1 RNA in resting CD4+ T cells

Abstract

HIV-1 latency in resting CD4+ T cells represents a major barrier to virus eradication in patients on highly active antiretroviral therapy (HAART). We describe here a novel post-transcriptional block in HIV-1 gene expression in resting CD4+ T cells from patients on HAART. This block involves the aberrant localization of multiply spliced (MS) HIV-1 RNAs encoding the critical positive regulators Tat and Rev. Although these RNAs had no previously described export defect, we show that they exhibit strict nuclear localization in resting CD4+ T cells from patients on HAART. Overexpression of the transcriptional activator Tat from non-HIV vectors allowed virus production in these cells. Thus, the nuclear retention of MS HIV-1 RNA interrupts a positive feedback loop and contributes to the non-productive nature of infection of resting CD4+ T cells. To define the mechanism of nuclear retention, proteomic analysis was used to identify proteins that bind MS HIV-1 RNA. Polypyrimidine tract binding protein (PTB) was identified as an HIV-1 RNA-binding protein differentially expressed in resting and activated CD4+ T cells. Overexpression of PTB in resting CD4+ T cells from patients on HAART allowed cytoplasmic accumulation of HIV-1 RNAs. PTB overexpression also induced virus production by resting CD4+ T cells. Virus culture experiments showed that overexpression of PTB in resting CD4+ T cells from patients on HAART allowed release of replication-competent virus, while preserving a resting cellular phenotype. Whether through effects on RNA export or another mechanism, the ability of PTB to reverse latency without inducing cellular activation is a result with therapeutic implications.

Figure 1. Both MS and US HIV-1 RNAs Localize to the Nucleus of Resting CD4⁺ T Cells from Patients on HAART

(A) RT-PCR analysis of US and MS HIV-1 RNAs in nuclear (N), cytoplasmic (C), and total (T) RNA fractions isolated from 10⁶ purified resting CD4⁺ T cells from a patient on HAART. Unspliced GAPDH and spliced GAPDH were used as internal controls for fraction purity and RNA integrity. (B) Effect of cellular activation on the localization of HIV-1 RNAs. 10⁶ resting CD4⁺ T cells were stimulated with anti-CD3 and anti-CD28. Nuclear (N), cytoplasmic (C), and total (T) RNA fractions were isolated at given times post-activation and analyzed as above with primers to detect US (shown) and MS (not shown) HIV-1 RNAs. Similar results were obtained with both US and MS primer sets.

Figure 2. PTB Binds Specifically to MS HIV-1 RNA and Is Upregulated upon Cellular Stimulation

(A) PTB binds specifically to MS HIV-1 RNA. Recombinant PTB was pre-incubated with increasing molar equivalents (0X, 1X, 2X, 5X) of non-biotinylated MS HIV-1 RNA (relative to biotinylated MS HIV-1 RNA) or a 5-fold mass excess of non-specific (NS) competitor, total yeast RNA; and then incubated with biotinylated, in vitro-transcribed MS HIV-1 RNA bound to streptavidin-conjugated magnetic particles. PTB protein eluted by RNase digestion was analyzed by Western blot. (B) PTB expression increases in response to T-cell stimulation. Expression of PTB was measured by Western blotting in resting CD4⁺ T cells cultured for 72 h without activating stimuli (−) or with (+) anti-CD3/anti-CD28. Vinculin served as a loading control.

Figure 3. PTB Overexpression Is Sufficient to Reverse HIV-1 Latency in Primary Resting CD4⁺ T Cells

(A) Diagram of expression constructs. The PTB construct contains full-length PTB including a NLS and four RNA recognition motifs (RRM). PTB ser-ala has a single serine to alanine substitution at position 16 in the NES located within the NLS region. PTBT is the 25 kDA isoform which lacks an NLS but contains two intact RRMs. (B) Overexpression of PTB and PTB ser-ala is sufficient to upregulate virus production from resting CD4⁺ T cells. Purified resting CD4⁺ T cells from patients on HAART were transfected with the indicated constructs. Culture supernatants were isolated 72 h post-transfection and analyzed for viral RNA. Dotted line represents the limit of detection of the assay. (C) Other cellular proteins that bind HIV-1 RNA cannot induce virus production from resting CD4⁺ T cells of patients on HAART. Resting CD4⁺ T cells were transfected as described above with empty vector or hnRNP A1 (left panel) and empty vector or La (right panel). HIV-1 RNA levels in the supernatants 72 h after transfection are shown.

Figure 4. PTB Expression Specifically Upregulates HIV-1 Gene Expression without Inducing Cellular Activation

(A) Expression of activation markers on PTB-transfected primary resting CD4⁺ T cells. PTB-transfected cells were analyzed 24, 48 (shown), and 72 h post-transfection for surface expression of CD25 and HLA-DR. Cells were stained with phycoerythrin-conjugated isotype control antibody, anti-CD25, or anti-HLA-DR. Cell size is indicated by the forward scatter (FSC) values on the Y-axis. (B) PTB-transfected cells do not transition to the G_1b stage of the cell cycle. Mock-transfected and PTB-transfected cells and PHA-stimulated cells were stained with PY and 7AAD to measure RNA and DNA levels, respectively. Dye binding was assessed by flow cytometry.

Figure 5. Induction of Virus Production by PTB or Tat

High-level expression of PTB or Tat in resting CD4⁺ T cells from patients on HAART induces virus production that is comparable to that seen following PHA stimulation. Resting CD4⁺ T cells purified from patients on HAART were transfected as described above with expression vectors for PTB or Tat, or as a positive control, were stimulated with PHA. For all transfections, supernatant HIV-1 levels were measured at 72 h post-stimulation. For PHA-stimulated cells, supernatant levels of HIV-1 RNA were measured every day for a 6-d period. Levels shown reflect peak virus release over the 6-d period.

Figure 6. PTB Localization in Transfected Primary Cells

Resting CD4⁺ T cells were mock-transfected or were transfected with the expression construct indicated on far left and stained as indicated on the top. Cells were stained with DAPI (blue) and with antibodies to actin (green) and PTB (red). Immunofluorescence of transfected resting CD4⁺ T cells demonstrated strong nuclear localization of both PTB and PTB ser-ala (speckled pink/blue pattern in nucleus). Transfected cells are indicated with arrows.

Figure 7. Overexpression of PTB Does Not Increase in Total Intracellular Concentrations of HIV-1 RNA

(A) Steady-state levels of HIV-1 RNA are not increased by overexpression of PTB. Resting CD4⁺ T cells (4 × 10⁶) were transfected as described above. 1 × 10⁶ cells were isolated at various times post-transfection. Total cellular RNA was isolated, and quantitative real-time RT-PCR was performed using primers that detect total HIV-1 transcripts. (B) PTB has no major effect on HIV-1 RNA stability. Resting CD4⁺ T cells (4 × 10⁶) were transfected as described above. At 24 h post-transfection, α-amanitin was added to cells to block de novo RNA synthesis. Total RNA was isolated at given times and real-time RT-PCR was performed to detect total HIV-1 RNAs.

Figure 8. PTB Expression Alters the Subcellular Distribution of HIV-1 RNAs in Primary Resting CD4⁺ T Cells from patients on HAART

Nuclear (N), cytoplasmic (C), and total (T) RNA fractions were isolated from mock-transfected (left) or PTB-transfected (right) resting CD4⁺ T cells 24 and 48 h post-transfection. RNA fractions were analyzed with primers to detect US HIV-1 RNA.

Figure 9. PTB Expression Is Sufficient to Reverse HIV-1 Latency and to Induce the Production of Replication-Competent Virus

Resting CD4⁺ T cells from patients on HAART were transfected with empty vector, or a PTB expression vector, or were stimulated with PHA. Where indicated, cells were then co-cultured with uninfected, activated CD4⁺ T cells. Supernatant p24 levels were measured daily over a 2-wk period. Peak p24 levels are shown for each condition. Results from two representative patients are shown.