HIV-1 pre-integration complexes selectively target decondensed chromatin in the nuclear periphery

Abstract

Integration of the double-stranded DNA copy of the HIV-1 genome into host chromosomal DNA is a requirement for efficient viral replication. Integration preferentially occurs within active transcription units, however chromosomal site specificity does not correlate with any strong primary sequence. To investigate whether the nuclear architecture may affect viral integration we have developed an experimental system where HIV-1 viral particles can be visualized within the nuclear compartment. Fluorescently labeled HIV-1 virions were engineered by fusing integrase, the viral protein that catalyzes the integration reaction, to fluorescent proteins. Viral tests demonstrate that the infectivity of fluorescent virions, including the integration step, is not altered as compared to wild-type virus. 3-D confocal microscopy allowed a detailed analysis of the spatial and temporal distribution of the pre-integration complexes (PICs) within the nucleus at different moments following infection; the fluorescently labeled PICs preferentially distribute in decondensed areas of the chromatin with a striking positioning in the nuclear periphery, while heterochromatin regions are largely disfavored. These observations provide a first indication of how the nuclear architecture may initially orient the selection of retroviral integration sites.

Figure 1. IN-EGFP trans-incorporation through Vpr produces intact and infective viral particles.

(A) IN-EGFP is incorporated into intact HIV-1 virions. Supernatant from un-transfected control cells (lane 1) or cells transfected with pVSV-G and pNL4-3.Luc.R-E- (lane 2), pD64E (lane 3) or pD64E together with pVpr-IN-EGFP (lane 4) were pelleted by ultra-centrifugation through a 20% sucrose cushion and analyzed by western blot using antibodies anti-IN (left panel) or with an anti-HIV-1 human serum (right panel). (B) HIV-IN-EGFP virions are infective. Ultracentrifuged supernatants from cells producing the indicated viruses were normalized for p24 content and used to infect HeLa cells. Seven days post-infection cell lysates were analyzed for luciferase activity. Luciferase activity values were normalized for protein content of lysates and expressed as relative light units (RLU). (C) HIV-IN-EGFP virions integrate in the cellular genome. DNA extracted from HeLa cells infected for 24 hours with a NL4-3.Luc.R-E- (lane 1), D64E (lane 2) or HIV-IN-EGFP (lane 3) viruses pseudotyped with VSV-G were amplified with primers specific for the Alu-LTR (to analyze the fraction of integrated HIV-1 DNA-upper panel) and for the lamin B2 cellular gene (to standardize the total amount of extracted DNA- lower panel).

Figure 2. Visualization of HIV-IN-EGFP virions.

Supernatants from cells transfected with the pD64E, pVpr-IN-EGFP and pVSV-G plasmids were ultra-centrifuged through a 20% sucrose cushion and adhered to glass coverslips. Following rinsing and fixation, immunostaining was performed with specific antibodies against viral proteins p17^MA and p24^CA and then visualized for EGFP fluorescence (green), p17^MA (blue) and p24^CA (red). Left panel is the merged image where white spots result from overlapping green, blue and red signals. Right panels show the individual fluorescent images. Bars, 5 µm.

Figure 3. HIV-IN-EGFP virions can be visualized in infected cells.

(A) Confocal visualization of HeLa cells infected with concentrated supernatants derived from cells transfected with the pD64E, pVpr-IN-EGFP and pVSV-G plasmids. Six hours post-infection cells were immunostained with antibodies against p17^MA and visualized for EGFP fluorescence (green) or p17^MA staining (red). In the merged image, where cell and nuclear shapes are outlined in white, yellow spots indicate full overlapping of green and red signals. Bar, 5 µm. Enlargements of the boxed region show individual color and merged images. Bar, 2 µm. (B) Same as in (A) using antibodies against p24^CA.

Figure 4. HIV-IN-EGFP virions containing neo-synthesized viral c-DNA can be visualized in infected cells.

Confocal visualization of a HeLa cell at 6 hours after infection with HIV-IN-EGFP (green) viral supernatant incubated with deoxynucleotides Alexa-594-dUTP (red). The large panel shows the merged image where yellow indicates overlapping of green and red signals. Cell and nuclear shapes are outlined in white. Bar, 5 µm. Individual color and merged magnification images of the boxed region are shown in the right panels. Bar, 0.5 µm.

Figure 5. Nuclear visualization of the HIV-IN-ECFP virus.

Nucleus of a HeLa cell infected with the HIV-IN-ECFP virus (green) and immunostained with lamin A/C (blue). Acquisition of confocal images was performed as z stack with a z step size of 0.3 µm and was followed by deconvolution based on experimentally determined PSF (see Experimental Procedure). (A) x-y section taken approximately at half high of the analyzed nucleus. (B) Adjacent vertical projections taken every 0.301 µm from upper (top-left panel) to lower side (bottom-right panel) of the nucleus in (A). Bars, 5 µm.

Figure 6. Nuclear HIV-IN-EGFP virions are functionally intact.

(A) Merged image of a confocal section obtained from HeLa cells infected with HIV-IN-EGFP supernatants incubated with deoxynucleotides Alexa-594-dUTP. Yellow indicates full overlapping of EGFP and Alexa-594-dUTP signals. Cell and nuclear shapes are outlined in white. Bars, 5 µm and 1 µm in the whole frame and inset box, respectively. (B) Merged image of a confocal section obtained from HeLa cells infected with HIV-IN-EGFP supernatants (green) and immunostained with antibodies anti-p24^CA (red) and lamin A/C (blue). Yellow indicates full overlapping of EGFP and anti-p24^CA signals. Cell and nuclear shapes are outlined in white. Enlargements of frames 1 and 2 are shown in bottom panels along with the individual color images. Bars, 5 µM and 1 µm in the whole frame and enlarged boxes, respectively. (C) HeLa cells infected with HIV-IN-EGFP virions were fixed and the mean number of intranuclear viral particles per nucleus was quantified at 3 hrs, 6 hrs, 9 hrs, 12 hrs, 16 hrs and 24 hrs. At each time point an average of 30 cells was analyzed and SD represented as error bars.

Figure 7. HIV-IN-ECFP virions preferentially localize in the periphery of the nucleus and outside heterochromatin regions.

(A) Confocal images of HIV-IN-ECFP virions (green) in nuclei of HeLa cells expressing H2B-EYFP (red) and immunostained with antibody against lamin A/C (blue). Bar, 5 µm. (B) H2B-EYFP fluorescence intensity distribution frequency for PIC ROIs (defined by the HIV-IN-ECFP virions, red bars) and for Random ROIs (selected randomly in the same planes, grey bars) (n = 103). Solid lines are obtained by Gaussian fitting. The 3MeK9H3 labeled arrow represents the width at half-height of the H2B-EYFP fluorescence intensity distribution derived by cross analyzing with this heterochromatin marker (see Experimental Procedures). In the inset the distribution cumulative probabilities are plotted for PIC ROIs (red) and Random ROIs (grey) (P<0.001, Kolmogorov–Smirnov test). (C) Distribution of intranuclear PICs distances from the lamin A/C staining.