Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2013 Jul 16;110(29):12036-41.
doi: 10.1073/pnas.1307157110. Epub 2013 Jul 1.

BET Proteins Promote Efficient Murine Leukemia Virus Integration at Transcription Start Sites

Affiliations
Free PMC article

BET Proteins Promote Efficient Murine Leukemia Virus Integration at Transcription Start Sites

Amit Sharma et al. Proc Natl Acad Sci U S A. .
Free PMC article

Abstract

The selection of chromosomal targets for retroviral integration varies markedly, tracking with the genus of the retrovirus, suggestive of targeting by binding to cellular factors. γ-Retroviral murine leukemia virus (MLV) DNA integration into the host genome is favored at transcription start sites, but the underlying mechanism for this preference is unknown. Here, we have identified bromodomain and extraterminal domain (BET) proteins (Brd2, -3, -4) as cellular-binding partners of MLV integrase. We show that purified recombinant Brd4(1-720) binds with high affinity to MLV integrase and stimulates correct concerted integration in vitro. JQ-1, a small molecule that selectively inhibits interactions of BET proteins with modified histone sites impaired MLV but not HIV-1 integration in infected cells. Comparison of the distribution of BET protein-binding sites analyzed using ChIP-Seq data and MLV-integration sites revealed significant positive correlations. Antagonism of BET proteins, via JQ-1 treatment or RNA interference, reduced MLV-integration frequencies at transcription start sites. These findings elucidate the importance of BET proteins for MLV integration efficiency and targeting and provide a route to developing safer MLV-based vectors for human gene therapy.

Keywords: retroviral gene therapy; virus–host interactions.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.
Fig. 1.
BET proteins specifically interact with MLV integrase. (A) GST pull-down of NIH 3T3 cell lysate with either GST-MLV IN or GST–HIV-1 IN and immunoblotting with Brd2, -3, and -4 antibodies. (B) Coomassie-stained SDS/PAGE gel of GST pull-down products from A showing that similar levels of GST-MLV IN or GST–HIV-1 IN bound to glutathione Sepharose beads. (C) GST pull-down of HEK293T cell lysate expressing FLAG-Brd3(1-726) with GST-MLV IN, GST-MLV IN CTD, GST-MLV IN NTD, and GST-MLV IN NTD-CCD and immunoblotting with FLAG antibody. (DF) Affinity pull-down with GST-MLV IN and HEK293T cell lysate expressing FLAG-Brd3(1-726) (D), GFP or GFP-Brd3(420-726) (E), and FLAG-Brd3(1-419) (F), respectively. Immunoblotting with FLAG or GFP antibody. “Input lysate (10%)” indicates 10% of indicated whole-cell lysate used for pull-down. “Beads + lysate” indicates control pull-down without GST-tagged protein.
Fig. 2.
Fig. 2.
BET proteins directly bind and stimulate MLV integrase activity in vitro. (A) GST pull-down of purified recombinant His-Brd4(1-720) (5 μM) with either GST-MLV IN or GST–HIV-1 IN. GST-MLV IN and GST–HIV-1 IN inputs indicate control pull-down without His-Brd4(1-720). Also shown is the 10% of His-Brd4(1-720) as input. (B) GST pull-down of purified recombinant His-LEDGF/p75 (5 μM) with either GST-MLV IN or GST–HIV-1 IN. GST-MLV IN and GST–HIV-1 IN inputs indicate control pull-down without His-LEDGF/p75. Also shown is the 10% of His-LEDGF/p75 as input. (C) Affinity pull-down of increasing concentrations (lanes 2–8) of purified recombinant His-Brd4(1-720) with GST-MLV IN. Control pull-down is shown with 125 nM His-Brd4(1-720) without GST-MLV IN (lane 1) to rule out nonspecific binding to glutathione beads. (D) Graphical representation of immunoblot shown in C to determine the apparent Kd of binding of ∼33 nM. The intensities of GST-MLV IN-bound His-Brd4(1-720) bands were quantified using ImageJ software, and data were fit to the Hill equation. (E) Concerted integration of 5′-Cyanine 5 (Cy5)–labeled viral donor DNA (1 μM) into target DNA (pBR322; 300 ng) by purified recombinant His-MLV IN (0.3 μM). Purified recombinant His-Brd4(1-720) was added to the reactions at the indicated concentrations. The image represents Cy5 signal as detected by Typhoon 9410 Imager. Indicated are the “one-end” and the biologically relevant concerted “two-end” integration products. (F and G) Effects of Brd4(1-720) and LEDGF/p75 on in vitro strand transfer (ST) activity of MLV IN (F) and HIV-1 IN (G). The ST products were detected by measuring the HTRF signal. Recorded signals were normalized using 100% ST activity for integrase alone. Bars represent means ± SD [n = 3; *P < 0.05; **P < 0.01; ***P < 0.001 by Student t test for IN alone vs. indicated MLV IN+Brd4(1-720) or HIV-1 IN+LEDGF/p75 samples, respectively].
Fig. 3.
Fig. 3.
Inhibition of BET proteins reduces MLV integration. (A) Dose-dependent effect of JQ-1 on HIV-1 or MLV expression. HEK293T cells infected with luciferase reporter HIV-1 (HIV-1–Luc) or transduced with MLV-Luc vector in the absence or presence of indicated concentrations of JQ-1 inhibitor. Luciferase assay was performed at 48 h postinfection or posttransduction. The luciferase signal obtained at 0 nM JQ-1 (DMSO alone) was set to 100% (values represent mean ± SD; n = 3). (BE) qPCR analysis of JQ-1–treated (indicated as “+”; 1,000 nM) or nontreated (indicated as “–”; DMSO) HEK293T cells infected with vesicular stomatitis virus-glycoprotein G (VSV-G) pseudotyped MLV. Bar graphs indicate the amount of PCR products relative to nontreated sample at 24 h postinfection for MSSS (B), PSE (C), and 2-LTR circle (D) products. (E) Bar graph indicates the integrated provirus relative to nontreated sample at 10 d postinfection. (FH) qPCR analysis of JQ-1 treated (indicated as “+”; 1,000 nM) or nontreated (indicated as “–”; DMSO) HEK293T cells infected with VSV-G pseudotyped HIV-1–Luc. Bar graphs indicate the amount of PCR products relative to nontreated sample at 24 h postinfection for late reverse-transcription (Late RT) (F) and 2-LTR circle (G) products. (H) Bar graph indicates the integrated provirus relative to nontreated sample at 10 d postinfection. All bars represent means ± SD (n = 3; *P < 0.05; **P < 0.01; ***P < 0.001 by Student t test). (I and J) HEK293T cells transduced with MLV-LTR GFP vector in the absence or presence of indicated concentrations of JQ-1 inhibitor. (I) Dose-dependent effect of JQ-1 on expression of MLV-LTR GFP vector. The percentages of GFP-positive cells were quantified by FACS analysis at 48 h posttransduction. The number of GFP-positive cells for 0 nM JQ-1 (DMSO alone) was set to 100%. (J) Dose-dependent effect of JQ-1 on integration of MLV-LTR GFP vector. Genomic DNA was harvested 15 d posttransduction, and the integrated copies of MLV-LTR GFP vector were measured. Bar graph indicates the integrated vector relative to 0 nM JQ-1 (DMSO alone) sample. All bars represent means ± SD (n = 3; ***P < 0.0001 measured by one-way ANOVA; multiple comparisons of the JQ-1 treatment to the DMSO control used Dunnett simultaneous test).
Fig. 4.
Fig. 4.
Antagonism of BET proteins reduces MLV-integration frequencies at the transcription start sites. (A and B) Analysis of integration frequencies of ASLV, HIV-1, and MLV relative to BET proteins (Brd2, -3, and -4) or HP1α/β chromatin sites in HEK293T cells. The chromatin sites and promoters bound by BET proteins or HP1α/β were quantified using ChIP-Seq data (21). (A) Heatmap depicting association of integration sites with BET proteins or HP1α/β chromatin sites. (B) Heatmap depicting association of integration sites with promoters bound by BET proteins or HP1α/β. The frequency of integration sites relative to the matched random controls was quantified using the receiver operating characteristic area method (44). The color key depicts enrichment or depletion of chromatin sites or promoters bound by indicated protein near integration sites. All comparisons of MLV to HIV or ASLV achieved P < 0.001 (Wald statistic). (CF) Percentage of MLV-integration sites found within each interval surrounding RefSeq transcription start sites (TSSs) in HEK293T cells. The integration sites near TSSs were compiled onto a single start site, and the frequencies were mapped. The x axis depicts the distance (in kb) relative to the TSSs (set at 0). The y axis depicts the percentage of integration sites in the indicated window. For comparison, integration sites of HIV-1 (9) and MLV (45, 46) in HEK293T cells are shown. (C) Dose-dependent effect of JQ-1 on MLV-integration frequencies at the TSSs. Percentage of MLV-integration sites within each interval surrounding TSSs in HEK293T cells treated with indicated concentrations of JQ-1 inhibitor or DMSO (indicated as “0 nM JQ-1”). (D) Percentage of MLV integration within 2-kb distance from TSSs. All samples achieved statistical significant (***P < 0.001; Fisher’s exact test) compared with 0 nM JQ-1 treatment. (E) Effect of concurrent down-regulation of BET proteins on MLV-integration frequencies at the TSSs. Percentage of MLV-integration sites within each interval surrounding TSSs in HEK293T cells transfected with scrambled siRNA (indicated as “Sci”) or a pool of Brd2, -3, and -4 siRNAs [indicated as “Brd(2+3+4)i”]. (F) Percentage of MLV integration within 2-kb distance from TSSs. Brd(2+3+4)i achieved statistical significant (**P = 0.009; Fisher’s exact test) compared with Sci.

Similar articles

See all similar articles

Cited by 86 articles

See all "Cited by" articles

Publication types

MeSH terms

LinkOut - more resources

Feedback