SIRT1 Regulates HIV Transcription via Tat Deacetylation - 《科学公立图书馆生物学》干细胞之家



免疫细胞治疗专区	欢迎关注干细胞微信公众号

12 3 4 5 6 7 8 9 10 ... 23 下一页

返回列表

查看: 345816\|回复: 229	go SIRT1 Regulates HIV Transcription via Tat Deacetylation [复制链接]

春天的风筝

新手上路

Rank: 1

积分: 0
威望: 0
包包: 0

楼主

发表于 2009-4-23 08:44 |只看该作者 |倒序浏览 |打印

1 Gladstone Institute of Virology and Immunology, University of California, San Francisco, California, United States of America,2 Applied Tumorvirology, Deutsches Krebsforschungszentrum, Heidelberg, Germany,3 Institute of Biochemistry, Humboldt University, Berlin, Germany,4 Department of Pathology, University of Pittsburgh, Pittsburgh, Pennsylvania, United States of America,5 Ottawa Regional Cancer Centre, Ottawa, Canada,6 Department of Pharmaceutical Sciences, Albert-Ludwigs-University, Freiburg, Germany
! U; u" s: @; Q  G! `1 g4 S
The human immunodeficiency virus (HIV) Tat protein is acetylated by the transcriptional coactivator p300, a necessary step in Tat-mediated transactivation. We report here that Tat is deacetylated by human sirtuin 1 (SIRT1), a nicotinamide adenine dinucleotide-dependent class III protein deacetylase in vitro and in vivo. Tat and SIRT1 coimmunoprecipitate and synergistically activate the HIV promoter. Conversely, knockdown of SIRT1 via small interfering RNAs or treatment with a novel small molecule inhibitor of the SIRT1 deacetylase activity inhibit Tat-mediated transactivation of the HIV long terminal repeat. Tat transactivation is defective in SIRT1-null mouse embryonic fibroblasts and can be rescued by expression of SIRT1. These results support a model in which cycles of Tat acetylation and deacetylation regulate HIV transcription. SIRT1 recycles Tat to its unacetylated form and acts as a transcriptional coactivator during Tat transactivation.
/ ^3 t) c+ c) ^9 ~) O: @" {/ G  F# r1 E% ^
Introduction/ ^8 x9 c/ r. T8 `* Q3 M, _

The Tat protein of human immunodeficiency virus 1 (HIV-1) is essential for the transcriptional activation of the integrated HIV-1 provirus. Without Tat, HIV transcriptional elongation is inefficient and results in abortive transcripts that cannot support viral replication [1,2]. Tat is produced early after infection from rare full-length genomic transcripts generated despite the elongation defect. These transcripts lead to the synthesis of a few Tat molecules sufficient to stimulate HIV transcription elongation, leading to the production of additional Tat transcripts and protein.+ G& s( _6 s& v1 u7 {0 [5 b+ ~9 x! t

Tat activates HIV transcription through the trans-acting responsive element (TAR), an RNA stem-loop structure that forms at the 5∩ end of all viral transcripts [3,4]. The TAR stem contains a three-nucleotide bulge structure recognized by the arginine-rich motif (ARM) in Tat (amino acids 49–57). In vivo, Tat binding to TAR requires cyclinT1, a cofactor that interacts cooperatively with both the N-terminal transactivation region of Tat (amino acid 1–48) and loop sequences at the top of the TAR stem-loop structure [5]. CyclinT1, a component of pTEFb (the positive transcription elongation factor b), recruits the cyclin-dependent kinase 9 (CDK9) to the HIV promoter. CDK9 hyperphosphorylates the C-terminal domain of RNA polymerase II, which potently enhances the processivity of the RNA polymerase II complex [6].

We and others have shown that Tat is acetylated at lysine 50 by the transcriptional coactivators p300 and human GCN5 (general control of amino acid synthesis 5) [7,8,9,10]. Tat acetylation is important for Tat activity and defines a critical cyclinT1-independent step in Tat transactivation [11]. Tat acetylated at lysine 50 cannot form a ternary complex with cyclinT1 and TAR RNA. It dissociates from TAR and binds instead to the p300/CREB-binding protein-associated factor (PCAF) via its bromodomain [12,13].

Our current working model is that Tat acetylation disrupts the Tat/TAR/cyclinT1 complex and leads to the transfer of Tat and PCAF to the elongating polymerase. According to this model, both forms of Tat, unacetylated and acetylated, play distinct roles in the HIV promoter transcriptional cycle and lead to the sequential recruitment of the cofactors cyclinT1 and PCAF. Because of the limiting amounts of Tat protein in the early stages of HIV infection and the critical role of unacetylated Tat for pTEFb recruitment to TAR, the question arises whether Tat acetylation can be reverted via a cellular Tat deacetylase.! _' s. C5 O, T* f" N) j/ \
7 n6 B: o; _' }$ T
There are three distinct classes of human histone deacetylases (HDACs) based on their homology with yeast transcriptional repressors. Class I and II HDACs are homologous to the yeast proteins Rpd3p (reduced potassium dependency 3) and Hda1p (histone deacetylase A1), respectively [14,15]. The deacetylase activity of class I and II HDACs is efficiently inhibited by trichostatin A (TSA) and other related hydroxamate-based inhibitors.
7 k; Y2 U: I' b* [
Class III HDACs, also named sirtuins (SIRTs), are homologous to the yeast transcriptional repressor silent information regulator 2p (Sir2p) [16]. Sir2p is a TSA-insensitive histone deacetylase that requires nicotinamide adenine dinucleotide (NAD ) as a cofactor [17,18,19]. Seven homologs of Sir2p have been identified in the human genome. Called SIRT1–7, they all contain a highly conserved catalytic domain [20]. Despite their enzymatic activity on histone substrates in vitro, recent experimental evidence suggests that SIRT proteins predominantly target nonhistone proteins for deacetylation, in both the nucleus and the cytoplasm. The nuclear SIRT1 protein deacetylates p53 [21,22,23], TAFI68 (Tata box-binding protein-associated factor I of 68 kDa) [24], PCAF and myoblast determination protein (MyoD) [25], p300 [26] and Forkhead transcription factors [26,27], the p65 subunit of nuclear factor kappa B (NF-百B) [28], and the Ku70 telomeric protein (also known as the thyroid autoantigen of 70 kDa or Ku antigen) [29]. The cytoplasmic SIRT2 protein is found associated with the microtubule network and deacetylates lysine 40 of 汐-tubulin [30]. SIRT3 is a mitochondrial matrix protein whose target has not been identified [31,32].
* j. f0 r7 m& `. ~
Here, we identify the class III HDAC SIRT1 as a specific Tat deacetylase and demonstrate that SIRT1 is a novel cofactor necessary for efficient Tat-mediated transactivation of the HIV promoter.
% p& G' L* R9 X' F
Results
+ C; x1 W3 R! i) {5 r
To test the ability of SIRT1–7 to deacetylate Tat in vitro, we transfected HEK 293 cells with expression vectors for human SIRT1–7 and immunoprecipitated the FLAG-tagged proteins (Figure 1A). The immunoprecipitated material was incubated with a full-length synthetic Tat protein carrying an acetylated lysine at position 50 (AcTat). The extent of Tat deacetylation was determined by Western blot (WB) with antibodies specific for the acetylated ARM in Tat [11]. Incubation of AcTat with immunoprecipitated SIRT1, SIRT2, and SIRT3 resulted in deacetylation of Tat lysine 50 (Figure 1B). These enzymes also deacetylate histones as determined in a standard histone deacetylase assay (Figure 1B). All reactions contained equal amounts of AcTat as determined by immunoblotting with streptavidin-horseradish peroxidase conjugate (SA-HRP), which recognized the biotin label attached to the N terminus of AcTat (SA-HRP in Figure 1B). SIRT enzymes in the reactions were visualized by immunoblotting with FLAG antibodies (FLAG in Figure 1B).
# {8 t" j  a) k0 Z! a% R
(A) Scheme of Tat deacetylation assay with immunoprecipitated SIRT1–7 proteins. Expression vectors for FLAG-tagged SIRT proteins were transfected into HEK 293 cells, immunoprecipitated, and incubated with synthetic Tat (72 amino acids) carrying an N-terminal biotin label and an acetyl group at position 50 (AcTat) in the presence of NAD . Immunoprecipitated material was also analyzed in a radioactive (3H) histone deacetylase assay using an H3 peptide as a substrate.
8 H* a6 n0 l1 Z2 S
(B) WB analysis of deacetylation reactions with antibodies specific for acetylated lysine 50 in Tat (汐-AcTat), with SA-HRP, or with 汐-FLAG antibodies.9 `& Q. G: a2 [9 J; A

(C) WB of Tat deacetylation by immunoprecipitated SIRT1 in the presence or absence of NAD , TSA, or nicotinamide (Nic)./ ]8 q# O4 c& P0 M, p% n
9 \- j5 V& i: `* g) P6 e
SIRT2 and SIRT3 proteins are localized primarily in the cytoplasm and the mitochondria [30,31], and SIRT1 resides in the cell nucleus [23,33]. Since Tat is a predominantly nuclear protein, we focused our efforts on SIRT1. The SIRT1-mediated deacetylation of Tat was dependent on NAD  and completely inhibited by nicotinamide, an inhibitor for class III HDACs [34,35]. TSA, a specific inhibitor of class I and II HDACs, had no effect (Figure 1C). These results demonstrate that the Tat deacetylase activity within immunoprecipitated SIRT1 material can be solely attributed to SIRT1 and not to a contaminating class I or II HDAC.% k% H* |) r  U9 k

To test whether Tat and SIRT1 interact, Tat/FLAG and SIRT1/influenza hemagglutinin (HA) were overexpressed in HEK 293 cells, and cellular lysates subjected to coimmunoprecipitation assays. Tat was detected with an 汐-FLAG antiserum in material immunoprecipitated with SIRT1 by the 汐-HA antibody in cells transfected with SIRT1- and Tat expression vectors, but no signal was obtained when SIRT1 or Tat alone was expressed (IP: 汐-HA in Figure 2A). Conversely, SIRT1 also specifically coimmunoprecipitated with Tat/FLAG (IP: 汐-FLAG in Figure 2A). The same was observed when Tat/T7 was coexpressed with SIRT1/FLAG and was immunoprecipitated with 汐-T7 antibodies (Figure 2B). No coimmunoprecipitation of Tat was observed with SIRT2 and SIRT6 (Figure 2B), two SIRT proteins that can also localize to the cell nucleus (BN and EV, personal communication), or any other SIRT protein (unpublished data). Furthermore, Tat coimmunoprecipitated with endogenous SIRT1 in Tat-expressing, but not in vector-transfected, HEK 293 cells (Figure 2C). No SIRT1- or Tat-specific signals were obtained after immunoprecipitations (IPs) in the absence of 汐-SIRT1 antibodies, excluding nonspecific binding of Tat to the Sepharose beads to which the antibodies were bound.& ^! b( [+ [/ _, r' J
% w% ^1 T# v+ Q% x1 Z
(A) Immunoprecipitation (IP) and WB of FLAG-tagged Tat (Tat-FLAG) and HA-tagged SIRT1 (SIRT1-HA) after transfection of corresponding expression vectors ( ) or empty vector controls () into HEK 293 cells.9 r; h0 k" h, }% y5 X* l6 b
4 r. n+ i( E; H( R: ^$ ^/ b
(B) The same experiments as in (A) performed with T7-tagged Tat and FLAG-tagged SIRT1, SIRT2, and SIRT6.
! ^$ D) A7 x2 z; p1 q
(C) Coimmunoprecipitation of FLAG-tagged Tat with endogenous SIRT1 in HEK 293 cells transfected with the Tat expression vector or the empty vector control. IPs were performed with or without rabbit 汐-SIRT1 antibodies.
. M5 M1 |4 `# W, S! E" }" H" Y' @
(D) WB of recombinant SIRT1 protein after pulldown with synthetic biotinylated Tat or AcTat. Tat proteins were detected with antibodies specific for acetylated lysine 50 in the Tat ARM (汐-AcTat) or SA-HRP.
6 m0 V: j! `8 L/ e
(E) Immunoprecipitation/WB of FLAG-tagged Tat or TatK50R and HA-tagged SIRT1. WT, wild type.

To test whether Tat and SIRT1 interact directly, increasing amounts of biotinylated synthetic Tat (72 amino acids) were incubated with recombinant full-length SIRT1. After pulldown with streptavidin-conjugated agarose, SIRT1 coimmunoprecipitated with Tat in a dose-dependent manner (Figure 2D). Recombinant SIRT1 bound equally well to acetylated and unacetylated synthetic Tat, indicating that the interaction occurred independently of the acetylation state of Tat (Figure 2D). WB with AcTat antibodies showed that AcTat remained acetylated during incubation with the SIRT1 enzyme (Figure 2D). Re-blotting with SA-HRP detected both Tat proteins in equivalent amounts in the binding reactions (Figure 2D). We also tested the ability of a Tat mutant protein (termed TatK50R to indicate mutation of lysine to arginine at position 50 of the Tat protein) to interact with SIRT1. This mutation preserves the basic charge at position 50, but cannot be acetylated. After transfection into HEK 293 cells, TatK50R accumulated to lower concentrations than wild-type Tat, but was bound to SIRT1 efficiently in coimmunoprecipitation assays (Figure 2E). These results collectively indicate that Tat binds SIRT1 directly and independently of lysine 50.

(A) Cotransfection of SIRT1 or the catalytically inactive SIRT1 mutant SIRT1H363Y with the HIV LTR luciferase construct and increasing amounts of a Tat expression vector (RSV-Tat: 0, 2, 20, and 200 ng), an HIV LTR luciferase construct containing mutated binding sites for the transcription factor NF-百B and RSV-Tat (20 ng), or with an RSV-luciferase construct (200 ng) in HeLa cells. The average of three experiments is shown (± standard error of the mean [SEM]).
' l: x( d) ~  W. \0 a
(B) WB analysis of HeLa cells 72 h after transfection of siRNAs directed against SIRT1 or GL3 control siRNAs.

(C) Cotransfection of the HIV LTR luciferase construct with increasing amounts of CMV-Tat or CMV-TatK50R (0, 50, 100, 200, 400, and 800 ng) 48 h after transfection of double-stranded siRNAs directed against SIRT1 or GL3 control siRNAs in HeLa cells. Luciferase activity was measured 24 h after plasmid transfection and 72 h after siRNA transfection. Note that all luciferase reporter vectors used in this study are based on the pGL2 luciferase vector, which is not affected by GL3-specific siRNAs [36]. The average of three experiments is shown (± SEM).

(D) The transcriptional activity of increasing amounts of the CMV-luciferase reporter (0, 50, 100, 200, 400, and 800 ng) was similar in SIRT1 knockdown or GL3-treated control cells. The average of two experiments performed in duplicate is shown (± SEM)." I1 E* S% G" e3 f0 G
; C0 _4 d7 K! [) d, d- }, i! E
(E) WB of endogenous SIRT1 or actin 72 h after transfection of siRNA directed against SIRT1 or mutated SIRT1 siRNA.7 d9 f* X- L& E2 J) l( e

(F) Cotransfection of the HIV LTR luciferase with increasing amounts of CMV-Tat (0, 2, 20, and 200 ng) in HeLa cells pretransfected with wild-type or mutant SIRT1 siRNA oligonucleotides as described in (C). WT, wild-type.# A; t4 V1 V# h! E8 ^

The effects of SIRT1 on Tat function were assessed after transfection into HeLa cells. SIRT1 modestly, but reproducibly, enhanced Tat-mediated transactivation of an HIV promoter luciferase construct (HIV LTR in Figure 3A). In contrast, expression of a catalytically inactive SIRT1 protein (termed SIRT1H363Y to indicate mutation of histidine to tyrosine at position 363 of the SIRT1 protein) suppressed Tat transactivation in a dominant-negative manner, indicating that the catalytic activity of SIRT1 is necessary for Tat transactivation. Similar results were obtained when an HIV promoter reporter construct containing mutant binding sites for the transcription factor NF-百B was used (HIV LTR 忖NF-百B in Figure 3A). This result indicates that the superinduction of Tat activity by wild-type SIRT1 and the suppression of Tat activity by catalytically inactive SIRT1 were dependent on the interaction between SIRT1 and Tat rather than on the interaction between SIRT1 and NF-百B/p65 [28]. Importantly, SIRT1 (both wild-type and SIRT1H363Y mutant) had no effect on the transcriptional activity of the Rous sarcoma virus (RSV) LTR, a promoter used to drive Tat expression in these cotransfection experiments.
2 t6 @3 j# y$ G' h8 J% e6 H
(A) Nuclear microinjection of HIV LTR luciferase, RSV-Tat, and a human cyclinT1-expressing construct into MEFs derived from SIRT /  or SIRT/ mice. In all experiments, a fixed amount of DNA was injected by adding the empty vector control. Cells were coinjected with CMV-GFP, and the luciferase activity per GFP-positive cell was calculated. An average of two injections is shown.

(B) The HIV LTR luciferase construct together with RSV-Tat and the cyclinT1-expressing construct were coinjected into SIRT/ MEFs in the presence of increasing amounts of a plasmid expressing human SIRT1. The average of three experiments is shown (± SEM).* C# g# I- `2 l: L5 O
9 O) M4 `! Q1 t0 u& }7 D) }; o
(C) Coinjection of the human SIRT1 expression vector together with the 5xUAS luciferase construct containing five Gal4 binding sites upstream of the thymidine kinase promoter and a Gal4-VP16 expression plasmid into SIRT1/ MEFs. The average of three experiments is shown (± SEM).3 O; `- @: S0 J* B

The effect of SIRT1 on Tat transactivation was further examined using small interfering RNA (siRNA)-mediated knockdown of SIRT1. HeLa cells were transfected with double-stranded RNA oligonucleotides directed against SIRT1 or against firefly luciferase expressed from the pGL3 vector as a control. All luciferase reporter constructs described in this study are based on the pGL2 vector, which is not affected by GL3 siRNAs (siRNAs directed against firefly luciferase expressed from the pGL3 vector) [36]. Levels of endogenous SIRT1 were markedly reduced at 72 h after transfection of siRNAs specific for SIRT1 (Figure 3B). At that time, a significant decrease in Tat transactivation was noted in cells that had received the SIRT1 siRNA, but not the GL3 siRNA (Figure 3C). The SIRT1 siRNA slightly enhanced the basal HIV promoter activity without Tat, and had no effect on the transcriptional activity of TatK50R, the Tat mutant that cannot be acetylated (Figure 3C). Loss of SIRT1 had no effect on the transcriptional activity of the immediate early promoter of the cytomegalovirus (CMV) used to drive Tat expression in these experiments (Figure 3D). In addition, Tat levels in HeLa cells transfected with SIRT1 siRNAs were comparable to Tat levels detected in cells transfected with GL3 siRNAs as determined by WB (unpublished data). To confirm the specificity of the SIRT1 siRNA, mutant double-stranded siRNA oligonucleotides were generated which contained a two-nucleotide mismatch between the target mRNA for SIRT1 and the antisense strand of the siRNA. Transfection of mutant SIRT1 siRNA did not affect expression of endogenous SIRT1 protein in HeLa cells, indicating that the mutation abrogated SIRT1 mRNA cleavage (Figure 3E). SIRT1 siRNA, but not mutant siRNA, suppressed Tat transactivation of the HIV LTR luciferase construct, confirming that the observed suppression was dependent on the loss of SIRT1 protein (Figure 3F).% G( P5 ^: n1 C( x
+ p# L0 U/ f. \5 Z+ [1 ^' n  n
(A) AcTat functions through TAR and cyclinT1 binding. Nuclear microinjection of increasing amounts of synthetic Tat or AcTat together with wild-type (wt TAR), TAR 忖bulge, or TAR 忖loop mutant HIV LTR luciferase constructs into HeLa cells. Cells were coinjected with CMV-GFP, and luciferase activity was calculated per GFP-positive cell. An average of three experiments is shown (± SEM).0 W3 r4 ?2 I% k9 z7 T% l9 J
" Z$ ?: _  q) G& {+ _
(B) AcTat transactivation requires CDK9. HeLa cells microinjected with Tat or AcTat (each 30 ng/μl) and the HIV LTR luciferase reporter were treated with increasing amounts of DRB, a known CDK9 inhibitor, for 4 h.9 C) d/ w; D" M) k& c+ s+ f1 |
4 H1 H3 I1 `- V7 F' Z6 x. ?8 W- w
(C) AcTat transcriptional activity is inhibited by nicotinamide, but not TSA. HeLa cells injected with HIV LTR luciferase and increasing amounts of AcTat were treated with TSA (400 nM) or nicotinamide (5 mM) for 4 h. The average of two experiments is shown.
# a$ A$ |! {$ l7 i( p+ f
(D) SIRT1 is necessary for AcTat, but not Tat function. HeLa cells were transfected with siRNAs specific for SIRT1 or GL3 control siRNAs 48 h before microinjection of HIV LTR luciferase and Tat or AcTat (each 30 ng/μl). The average of three experiments is shown (± SEM).
& S# ]# R' B; W: S
Since SIRT1 only modestly enhanced Tat transactivation in HeLa cells, which already express significant levels of SIRT1, we examined the effect of SIRT1 on Tat transactivation in a SIRT1-negative background. We obtained mouse embryonic fibroblasts (MEFs) derived from SIRT1 knockout mice [37]. The HIV LTR luciferase reporter and the Tat expression vector were introduced into these cells by nuclear microinjections because of their low responsiveness to various transfection protocols. Because murine cyclinT1 does not support Tat transactivation [38,39], an expression vector for human cyclinT1 was included in the microinjections. A 120-fold increase in HIV promoter luciferase activity was detected in the presence of Tat and human cyclinT1 in SIRT1 /  MEFs (Figure 4A). In contrast, Tat-mediated transactivation of the HIV LTR was reduced in SIRT1/ MEFs (Figure 4A). Ectopic expression of increasing amounts of human SIRT1 resulted in a dose-dependent increase of Tat transactivation in SIRT1/ MEFs (Figure 4B). In contrast, transactivation of the 5xUAS promoter by Gal4-VP16 was reduced in response to SIRT1 (Gal4-VP16 is a fusion between the binding domain of the DNA-binding transcription factor required for the activation of the GAL genes in response to galactose in Saccharomyces cerevisiae [termed Gal4], and the activator domain of the herpes simplex virus transactivator protein [designated VP16]) (Figure 4C). These results collectively demonstrate that SIRT1 represents a positive factor for Tat function.0 Z& K' c$ a; |" z; p, u0 z. F) M

This model was further tested in nuclear microinjection experiments using synthetic full-length Tat and AcTat. Microinjection of increasing amounts of either Tat or AcTat proteins into HeLa cells caused a marked transactivation of the HIV LTR luciferase reporter in a dose-dependent manner (Wt TAR in Figure 5A). AcTat transactivated the HIV promoter approximately 1.5–3-fold better than Tat. Transactivation by Tat and AcTat was dependent on the bulge and loop regions of TAR, indicating that transactivation by both proteins required the formation of an intact Tat/TAR/cyclinT1 complex [4,5,40] (TAR 忖Bulge and TAR 忖Loop in Figure 5A). In agreement with this conclusion, transactivation by both Tat proteins was inhibited in a dose-dependent manner by 5,6-dichlorobenzimidazole riboside (DRB), a CDK9 inhibitor known to block Tat function (Figure 5B) [6].7 }" h( ]* k/ F0 C1 g9 R

(A) In vitro histone deacetylation assays including recombinant SIRT1, radioactively labeled histone H3 peptide, and indicated concentrations of splitomicin or HR73. The average (± SEM) of two experiments performed in duplicate is shown for each point.
/ t5 Y- H2 Y8 d) N2 W' I
(B) Chemical structures of splitomicin and its derivative HR73.
0 v  _5 x# T' o: }" n
(C) Inhibition of Tat transactivation by HR73. RSV-Tat (0, 20, and 200 ng) and HIV LTR luciferase (200 ng) or RSV-luciferase (200 ng) vectors were transfected into HeLa cells. Transfected cells were treated with indicated concentrations of HR73 or DMSO for 8 h.
$ o; @$ c; L, l/ V2 [( N
(D) Inhibition of HIV gene expression by HR73. GFP expression in Jurkat T cells infected with HIVNL4–3 containing the GFP open reading frame in place of the viral nef gene or with an HIV-based lentiviral vector expressing GFP from the heterologous EF-1汐 promoter. Treatment with HR73 (1 μM in DMSO) or DMSO was performed for 36 h after overnight infection. The average (± SEM) of four experiments is shown.7 q  l# `# F3 O/ Z4 M- A  W: r

According to our working model, AcTat represents a second step in the transactivation cycle [11]. Since AcTat cannot form the trimolecular complex with cyclinT1 and TAR RNA in vitro, we hypothesized that AcTat becomes partially deacetylated by the Tat deacetylase upon microinjection. This would allow the initiation of the transactivation process by unacetylated Tat binding to TAR with cyclinT1 and CDK9. To further test this hypothesis, we treated cells with deacetylase inhibitors after microinjection of AcTat and the HIV promoter construct. Treatment with TSA, an inhibitor of class I and II HDACs, enhanced the transcriptional activity of AcTat as well as the basal HIV promoter activity (TSA in Figure 5C). In contrast, nicotinamide, an inhibitor of class III deacetylases, blocked transactivation of the HIV promoter by AcTat while stimulating basal HIV promoter activity (Nicotinamide in Figure 5C). Similarly, knockdown of SIRT1 using siRNA inhibited transcriptional activity of AcTat, while slightly enhancing Tat-mediated or basal transcriptional activity of the HIV promoter (Figure 5D). These results support the model that the transcriptional activity of AcTat depends on deacetylation by SIRT1 in cells.* r% f/ W  o) e& R& C$ B0 B$ p

The identification of SIRT1 as an enzyme that catalyzes an important step in HIV transcription suggests that it could be targeted therapeutically. Splitomicin was identified as a small molecule inhibitor of the S. cerevisiae Sir2p protein [41]. While splitomicin did not inhibit human SIRT1, we identified a splitomicin derivative, called HR73, which is structurally related to a previously described inhibitor of Hst1, a homolog of Sir2p in yeast [42]. HR73 effectively inhibited the histone deacetylase activity of SIRT1 in vitro with an IC50 (concentration causing 50% inhibition) of less than 5 μM (Figure 6A and 6B). Treatment of HeLa cells with HR73 suppressed Tat-dependent HIV transcription in a dose-dependent manner (3-fold at approximately 1 μM) after transfection of the Tat vector and the HIV LTR luciferase construct (HIV LTR in Figure 6C). In separate experiments, HR73 induced hyper(Sara Pagans, Angelika Ped)

收藏0 分享0 顶0 踩0