The site of HIV-1 integration in the human genome determines basal transcriptional activity and response to Tat transactivation

Abstract

Because of the heterogeneity of chromatin, the site of integration of human immunodeficiency virus (HIV) in the genome could have dramatic effects on its transcriptional activity. We have used an HIV-1-derived retroviral vector, in which the green fluorescent protein is under the control of the HIV promoter, to generate by infection 34 Jurkat clonal cell lines each containing a single integration of the HIV-1 vector. In the absence of Tat, a 75-fold difference in expression level between the highest and lowest expressing clones was observed. Basal promoter activity was low in 80% of the clones and moderate to high in the remaining 20% of clones. We found that differences in expression levels are due to the integration site and are not controlled by DNA methylation or histone acetylation. Tat activated transcription in each clone, and an inverse correlation was observed between basal transcriptional activity and inducibility by Tat. These observations demonstrate that the chromatin environment influences basal HIV gene expression and that the HIV Tat protein activates transcription independently of the chromatin environment.

Figures

Fig. 1. The site of integration of HIV determines the basal rate of viral transcription. (A) Flow cytometry analysis of a Jurkat culture either uninfected or infected with the HIV-derived vector LTR-GFP (m.o.i. = 0.3). The right panel shows the frequency histogram of R2-gated cells after infection. FL2H is the blank channel used to measure autofluorescence. (B) Basal LTR-GFP expression in 34 individual Jurkat clones. The mean GFP level ± SD is shown. The CV is shown in parentheses and indicates the dispersion of results. Clones A–F are representative clones extensively analyzed below. The frequency histogram for three representative clones (A, D and F) is shown inserted. (C) Flow cytometry analysis of LTR-GFP expression in the collection of clones in two different experiments separated in time by 4 weeks. (D) Relative quantification of LTR-GFP transcription by RT real-time PCR correlates with the flow cytometric quantification of the GFP fluorescence. Five representative clones (A, C, D, E and F) growing exponentially were analyzed by flow cytometry (m.f.i.) and in parallel by real-time RT–PCR with primer pairs corresponding to either the LTR sequence or the GFP open reading frame. Results were normalized to the expression of GAPDH used as endogenous control.

Fig. 2. Heterogeneity observed in HIV promoter activity is not secondary to clonal differences. (A) Comparison of expression of a transient LTR-luciferase construct and integrated LTR-GFP in each individual clone. Each clone was tested by flow cytometry to measure GFP expression (upper panel) or transiently transfected by electroporation with an LTR-luciferase construct and an internal transfection control TK-luciferase (Renilla) to correct for transfection efficiency (lower panel). (B) Plotting of luciferase activity (LRU) versus GFP expression (m.f.i.) for each clone demonstrates that there is no correlation between the expression of integrated LTR and transiently transfected LTR. (C) Absence of correlation between LTR-GFP and LTR-YFP integrated in different genomic locations in a single cell. Jurkat cells were infected at an m.o.i. of 0.1 with viral particles containing the retroviral vector LTR-GFP or LTR-YFP, or with both viruses at the same time, and analyzed by flow cytometry 48 h after infection.

Fig. 3. Effect of the HIV promoter integration site on Tat-dependent transcriptional activation. (A) Transcriptional activity of the integrated LTR-GFP construct after transfection of a Tat expression plasmid. Each clone was electroporated with a Tat expression plasmid (pEV280) or the control empty vector pcDNA3.1 (Invitrogen) and a plasmid encoding YFP, pEYFP-C1 (Clontech). LTR-GFP expression was measured in YFP-positive cells (positively transfected) 48 h after transfection. (B) Inverse correlation between HIV basal promoter activity and Tat transactivation. Scatter plot of the Tat induction factor calculated as the ratio of Tat-induced expression versus basal expression against basal LTR-GFP level (m.f.i.). A representative experiment of three is shown.

Fig. 4. Heterogeneity of HIV expression using a Tat-containing HIV vector after integration. Jurkat cells were infected with viral particles containing the retroviral vector LTR-Tat-IRES-GFP at an m.o.i. of 0.1. Six hours later, individual cells were cloned and a collection of 79 clones was obtained after expansion. LTR-GFP expression measured by flow cytometry is shown for all clones.

Fig. 5. Synergistic activation of the integrated HIV promoter in response to Tat, TPA and TSA. (A) Each Jurkat clone was treated with recombinant Tat (25 µg/ml), TPA (10 nM) and TSA (400 nM) for 24 h, and LTR activity was measured by flow cytometry. Fold induction versus control untreated samples are shown immediately above each bar. Synergism factors are shown above each bar (boxed) for the combinations Tat + TPA, Tat + TSA and Tat + TPA + TSA. They are calculated using the formula: induction by A + B/induction by A + induction by B. (B) Scatter plots of induction factors (defined as GFP m.f.i. in response to agents relative to untreated samples) for Tat, TPA or TSA against basal GFP levels in each clone. Scatter plots of the synergism factors (defined above) for the combinations Tat + TPA, Tat + TSA or Tat + TPA + TSA versus GFP m.f.i. for each clone.

Fig. 6. The HIV promoter is stably expressed over time and unresponsive to HDAC inhibitors or DNA methylation inhibitors. (A) Basal activity of the integrated LTR-GFP construct was measured weekly by flow cytometry over a 17-week period in six representative clones (A–F). The basal LTR activity of these individual clones is shown in Figure 1B. (B) The same six clones were treated with TSA (400 nM), 5-azadC (5 µM) or both agents, in the absence or presence of recombinant Tat (12.5 µg/ml). On day 1, 5-azadC was added to an exponentially growing culture; on day 2, Tat, TSA and a second aliquot of 5-azadC were added; on day 3, the experiment was completed and LTR-GFP expression analyzed by flow cytometry.

Fig. 7. The integrated HIV promoter DNA is not methylated in vivo. (A) Schematic representation of the HIV LTR and adjoining sequences in the retroviral vector used to derive our stable Jurkat clones. Restriction sites for the methylation-sensitive endonucleases HpaII, BssHII or EagI and the predicted size of the fragments obtained after a double digestion with NcoI are shown. (B) Genomic DNA purified from six representative clones was assayed for susceptibility to the indicated methylation-sensitive restriction enzyme (HpaII, BssHII or EagI) by Southern blotting. All digestions were further digested with NcoI to define the ends of the hybridization products. Control digestions with NcoI alone or NcoI + MspI (a methylation-insensitive isoschizomer of HpaII) were included.

Fig. 8. Remodeling of nuc-1 correlates with basal and activated promoter activity. (A) Diagram of the LTR promoter in the integrated HIV-derived vector (modified from Verdin et al., 1993). The positions of nucleosomes are indicated with respect to the HIV promoter, including nuc-1, which is remodeled upon HIV transcriptional activation. Positions of restriction sites for AflII and NcoI and the probe used in indirect end labeling are indicated. (B) AflII accessibility of the HIV promoter in five clones under basal conditions and in response to 400 nM TSA (4 h) or Tat. Purified nuclei were digested with AflII, DNA extracted and analyzed by indirect end labeling after NcoI digestion. The larger band in each clone corresponds to the NcoI–NcoI fragment and the shorter band corresponds to the AflII–NcoI fragment. The extent of nuc-1 remodeling was quantified by measuring the ratio of the intensity of the AflII-digested band and the total intensity of the two bands (undigested + digested). (C) The accessibility of nuc-1 to AflII correlates with the transcriptional activity in each clone. (D) Endonuclease accessibility after α-amanitin treatment. The same experimental conditions as in Figure 8B were used, but α-amanitin (10 µg/ml) was added 1 h before TSA addition and 5 h before analyzing nuc-1 accessibility. The response to 10 nM TPA is shown.