Closed chromatin architecture is induced by an RNA duplex targeting the HIV-1 promoter region

Abstract

In some mammalian systems small interfering RNAs (siRNA) targeting homologous sequences in promoter regions of genes induce transcriptional gene silencing (TGS). We have previously reported the induction of TGS by an siRNA (prom-A siRNA) targeting the tandem NF-kappaB-binding motifs within the human immunodeficiency virus, type 1 (HIV-1), promoter region. Here we report that induction of TGS by prom-A siRNA is accompanied by immediate and sustained local recruitment of Argonaute-1 (Ago1), histone deacetylase-1 (HDAC1), and induction of dimethylation of histone 3 at lysine 9 (H3K9me2), processes known to be associated with transcriptional silencing. Elevated levels of H3K9me2 and HDAC1 spread upstream of the target sequence, and elevated H3K9me2 levels also spread downstream into the coding region. Moreover, this siRNA induces an immediate change in DNA accessibility to restriction enzyme digestion in the region of the transcription initiation site of the HIV-1. This change in accessibility is because of the relocation of a nucleosome known to be associated with this region of the integrated pro-virus. Although there is a theoretical possibility that the observed viral suppression could be mediated by the PTGS mechanism with this siRNA acting at the 3 (R)-long term repeat of the virus, we demonstrate that this siRNA, and three other U3 targeted siRNAs, are inefficient inducers of PTGS. These data strongly suggest that siRNA targeting the promoter region acts predominantly at a site within the 5 (R)-long term repeat of HIV to induce transcriptional silencing and alterations to chromatin structure of the HIV promoter region that extend well beyond the immediate siRNA target site. These induced changes are consistent with those described in latent HIV-1 infection.

FIGURE 1.

HIV replication is inhibited by siRNA targeting the HIV-1 promoter. A, map of the HIV-1 5′-LTR, with the location of sequences targeted by the four siRNAs indicated, prom-A, -B, -C, and -D. The arrow marks the HIV-1 transcription start site. The number indicates the relative location from the transcription start site. B, various levels of inhibition are induced by the four siRNAs targeting the HIV promoter. Reverse transcriptase (RT) levels at day 6 post-transfection of siRNA in supernatant of viral cultures are shown. C, effect of prom-A and -D siRNAs on the time course of HIV-1_NL4-3 production in MAGIC-5 cells. Prom-A siRNA reduces viral production 200-fold by day 11 post-transfection compared with prom-D siRNA or mock transfection. The zero time point, when siRNAs are transfected into the cells, is 2 days after infection of the cells with HIV-1_NL4-3. Reverse transcriptase (RT) levels are shown for cultures transfected with prom-A siRNA (blue circle), prom-D siRNA (red triangle), and mock-transfected cells (black square).

FIGURE 2.

Enrichment of Ago1, H3K9me2, and HDAC1 associated with HIV-LTR in prom-A siRNA-transfected cells. ChIP assays were conducted at days 1 and 11 post-transfection on extracts from formaldehyde-fixed MAGIC-5 cells treated with prom-A, prom-D, or mock transfection. DNA fragments from whole-cell extracts were co-precipitated with antibodies against the following: A, Ago1; B, H3K9me2, and C, HDAC1, and then amplified by PCR using the LTR-forward and LTR-reverse primer pair (the position of this amplicon is between -110 and +80 relative to transcription start site). HIV-1-LTR copy numbers obtained from each immunoprecipitation were normalized against that obtained from whole-cell extracts. Each value shown is the relative enrichment of three separate experiments (mean ± S.E.) normalized to the value obtained from the mock transfection culture.

FIGURE 3.

The BglII site in HIV-1 promoter is protected after prom-A siRNA transfection as determined by CHART assay. A, design of CHART assay. Schematic figures show the relative positions of nucleosomes within the 5′-LTR of HIV-1. The nucleosomes (Nuc-0, -1, and -2) are precisely positioned in the HIV-1 LTR after integration of viral DNA into host genome (diagram modified from Ref. 41). Nuc-1 is tightly associated with the transcription initiation site, indicated by the arrow, in the silenced template. The position of Nuc-1 is shifted or disrupted following activation of viral transcription. A BglII site is located 20 bases from transcription initiation position close to the putative location of Nuc-1 in latent virus. The NF-Tar PCR primer pair was used to amplify an amplicon spanning the BglII site (indicated by double arrows) from nuclear DNA. DNA accessibility was assessed by the percentage decrease in HIV copy number measured by real time PCR with and without BglII digestion. B, productive infection is associated with an accessible BglII site. The nuclear fraction of MAGIC-5 cells productively infected with HIV-1 was subjected to BglII digestion at 37 °C for periods up to 60 min, and the extent of digestion was assessed by real time PCR. C, EcoRI digestion of the same DNA did not result in any change in real time PCR readout, supporting the specificity of the BglII CHART assay. D, real time PCR of the gag region confirms similar levels of cell-associated virus were present in all samples. In the following experiment, results are reported as the extent of digestion after 60 min of incubation of nuclei with BglII. E, CHART assay was applied to productively infected MAGIC-5 cells at days 1 and 11 after prom-A siRNA, prom-D siRNA, or mock transfection. In productively infected cells, the BglII site is highly susceptible to digestion with close to 100% of LTR copies cut. Within 24 h of transfection of prom-A siRNA, the BglII site is protected, and this protection is maintained for at least 11 days. Transfection with prom-D siRNA is associated with transient partial protection of this site. Protection of the BglII site suggests an alteration in chromatin structure that reduces DNA accessibility because of the re-positioning of Nuc-1 as proposed in A.

FIGURE 4.

Enrichment of H3K9me2 associated with HIV-LTR and HIV-gag in prom-A siRNA-transfected cells indicates extensive regional heterochromatin formation. ChIP assays were conducted as in Fig. 2 at day 11 post-transfection with antibodies against HDAC1 and H3K9me2 for the analysis of the following: A, upstream Nuc-0 region; B, downstream site within the coding region of gag. The Nuc-0 region was amplified using the LTRup-forward and LTRup-reverse primer pair for quantitative analysis by real time PCR (the position of this amplicon is between positions -435 and -299 relative to the transcription start site). The gag region was amplified with the primer pair SK145 and SCCIB (the position of this amplicon is between +1359 and +1513 relative to transcription start site). Values shown are the relative enrichment of three independent experiments (mean ± S.E.) normalized to the value obtained from the mock-transfected culture. C, model of heterochromatin formation induced by siRNA targeting HIV-1 promoter region. siRNA acts as a nucleation center for recruitment of the RNA-induced transcriptional silencing (RITS) complex and closed chromatin formation extends both upstream and downstream to include adjacent promoter and mRNA coding regions. siRNA (purple line) is loaded into Ago1. The RITS machinery, including histone deacetylase (HDAC1) and histone methyltransferase (HMT), induces H3K9me2 (blue flag) and recruits heterochromatin protein 1 (HP1) to the area surrounding the siRNA target site. Complexes including HP1 and HMT spread beyond the initial site of recruitment, creating broad domains of heterochromatin structure as indicated by H3K9me2 status of both the upstream promoter and downstream gag regions.

FIGURE 5.

Short siRNAs targeting the U3 region of the HIV-1 promoter have limited PTGS activity. A, map of the HIV-1 3′-LTR under the control of the immediate early CMV promoter, with the location of sequences targeted by four previously described siRNAs and code-R siRNA. Code-R siRNA has sequence homologous to part of the R region of HIV-1 3′-LTR upstream of the polyadenylation sites. The position of PCR primers used for the detection of mRNA of HIV-LTR is indicated with arrows. B, identification of two clones stably transfected with high expression of HIV-3′-LTR under the immediate early CMV promoter. RT-PCR was used to detect 3′-LTR expression, and two expressing clones are shown. One clone, CMV-3LTR1-4, with the highest level of expression of 3′-LTR messenger RNA, was chosen for further study. C, assessment of the extent of gene silencing by PTGS following transfection of clone CMV-3LTR1-4, with each of four U3 region-targeted siRNAs or an R region-targeted siRNA. LTR mRNA levels were assessed by real time RT-PCR 48 h after transfection. Real time PCR data are shown as a relative reduction in HIV-mRNA levels normalized to the value obtained from mock transfection experiments. Results shown are from three independent experiments (mean ± S.E.).