- Email: [email protected]
Extensive interactions between HIV TAT and TAFII250
Biochimica et Biophysica Acta 1546 (2001) 156^163 www.bba-direct.com
Extensive interactions between HIV TAT and TAFII 250 Jocelyn D. Weissman *, Jae Ryoung Hwang, Dinah S. Singer Experimental Immunology Branch, NCI, NIH, 9000 Rockville Pike, Bethesda, MD 20892, USA Received 27 October 2000; received in revised form 28 December 2000; accepted 28 December 2000
Abstract The HIV transactivator, Tat, has been shown to be capable of potent repression of transcription initiation. Repression is mediated by the C-terminal segment of Tat, which binds the TFIID component, TAFII 250, although the site(s) of interaction were not defined previously. We now report that the interaction between Tat and TAFII 250 is extensive and involves multiple contacts between the Tat protein and TAFII 250. The C-terminal domain of Tat, which is necessary for repression of transcription initiation, binds to a segment of TAFII 250 that encompasses its acetyl transferase (AT) domain (885^1034 amino acids (aa)). Surprisingly, the N-terminal segment of Tat, which contains its activation domains, also binds to TAFII 250 and interacts with two discontinuous segments of TAFII 250 located between 885 and 984 aa and 1120 and 1279 aa. Binding of Tat to the 885^984 aa segment of TAFII 250 requires the cysteine-rich domain of Tat, but not the acidic or glutamine-rich domains. Binding by the N-terminal domain of Tat to the 1120^1279 aa TAFII 250 segment does not involve the acidic, cysteine- or glutamine-rich domains. Repression of transcription initiation by Tat requires functional TAFII 250. We now demonstrate that transcription of the HIV LTR does not depend on TAFII 250 which may account for its resistance to Tat mediated repression. ß 2001 Elsevier Science B.V. All rights reserved. Keywords: TAFII250; HIV Tat
1. Introduction Transcription of eukaryotic genes requires the ordered interaction of a series of general transcription factors (GTF) that result in the recruitment of the RNA polymerase II (RNAP) complex, transcription initiation and subsequent elongation [1,2]. The process is initiated by the binding of TFIID to the promoter, which in turn triggers the formation of a preinitiation complex consisting of TFIIB, TFIIE, TFIIF and TFIIH (reviewed in [3]). Interactions between the RNA polymerase complex and the GTFs
* Corresponding author. Fax: +1-301-480-8499; E-mail: [email protected]
result in phosphorylation of the CTD of the polymerase and transcription initiation and elongation (reviewed in [4,5]). Phosphorylation is e¡ected by a kinase activity in TFIIH [6,7]; elongation is regulated by TFIIF [1,8,9]. Initiation depends on TFIID, which contains the TATA-binding protein (TBP) in association with the TBP-associated factors (TAFs) [10]. TAFII 250 is the largest component of TFIID, and is known to contain acetyl transferase activity (AT) [11]. The acetyl transferase activity of TAFII 250 is presumed to be important for initiation, although neither its substrate nor its mechanism of action is known. The HIV gene product, Tat, functions as a potent activator of the viral LTR increasing promoter activity by up to 100-fold [12^17]. Activation, which is
0167-4838 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 0 1 ) 0 0 1 3 5 - 2
BBAPRO 36373 7-3-01
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
mediated by the N-terminal 67 amino acids (aa) of the 101 amino acid full-length Tat, depends on the interaction between Tat, the P-TEFb complex component cyclin T1, and the HIV TAR RNA element [18]. This interaction stimulates the CTD kinase [19,20], resulting in increased phosphorylation of the RNAP CTD and enhanced elongation. In addition, Tat binds to and inhibits a CTD phosphatase, reducing dephosphorylation and thereby also enhancing rates of elongation [21,22]. Thus, the major e¡ect of Tat on HIV LTR promoter activity is to enhance transcription elongation (reviewed in [23]). Analysis of the Tat protein has demonstrated that two N-terminal domains of the Tat protein are necessary for this activation: an acidic domain located between aa 2 and 6 and a cysteine-rich domain spanning aa 20^48. Additionally, a glutamine-rich domain, aa 57^67, contributes to Tat's activation. Cterminal sequences, extending beyond amino acid 72 that are encoded by the second exon of the Tat gene, are not necessary for activation of the LTR by Tat. Although Tat stimulates transcription elongation of the HIV LTR in the presence of the HIV TAR element, it can repress transcription initiation of other promoters, in the absence of TAR. We have shown previously that Tat represses the activity of the promoters of major histocompatibility complex (MHC) class I, L2 -microglobulin and certain viral genes [24^26]. This repression occurs at the level of transcription initiation, is independent of TAR and is mediated through the interaction of Tat with TAFII 250, which inhibits the intrinsic AT activity of TAFII 250 [27]. Both repression of transcription and of TAFII 250 AT activity are mediated by the C-terminal polypeptide (67^101 aa) of Tat, which is not required for activation of the HIV LTR [24,25,28^30]. Thus, the activities of activation and repression are separable. Previous studies have demonstrated that Tat interacts with TAFII 250 in vivo both in HeLa cells and yeast, and in vitro in GST pull-down assays [27]. However, ¢ne mapping of the interaction sites of the two molecules was not performed. The present studies were undertaken to further characterize the interaction of Tat with TAFII 250. We report that at least three distinct regions of the Tat protein mediate binding to TAFII 250. Two of these reside in the Nterminal region of Tat, which is not involved in re-
157
pression, and anchor the binding of Tat to the TAFII 250. Surprisingly, binding of the N-terminal domains of Tat to TAFII 250 occurs at two discontinuous sites, one of which spans the binding site for Rap74. The third Tat region is in the C-terminal peptide which binds to the AT domain of TAFII 250, consistent with its inhibition of AT activity and repression of transcription initiation. Interestingly, transcription of the HIV LTR does not depend on TAFII 250, consistent with Tat's failure to repress the HIV LTR. 2. Materials and methods 2.1. Cell culture and transfections The tsBN462 cells containing a point mutation in the TAFII 250 gene, derived from BHK cells as described [32,33], were obtained from T. Sekiguschi (Salk Institute). The cells were maintained at 32³C, 7.5% CO2 in Dulbecco's modi¢ed Eagle's medium (Bio£uids) with 10% fetal calf serum. Transfections were essentially as previously described. Brie£y, cells were transfected with 5 Wg of pBennCAT(HIVLTR) or -68CAT (class I promoter with 68 bp of upstream £anking sequence) [34] using CaPO4 at 32³C. RSVluciferase was used as an internal transfection e¤ciency control. Following transfection, cells were incubated at 32³C for 24 h, at which time they were refed with fresh medium and either left at 32³C or shifted to 39.5³C. After an additional 16 h, cells were harvested and assayed for CAT activity as previously described [27]. Activity was corrected to protein concentration. 2.2. DNA constructs The TAFII 250 clone was isolated from a yeast two-hybrid screen using Tat as bait, as previously described [27]. TAFII 250 subfragments were constructed via PCR ampli¢cation using the TAFII 250 clone as template as follows. The 5P primer for each subfragment (see Fig. 2) contains the consensus Kozak sequence (ACC GCC ATG GGA CAC) followed by speci¢c sequence #848: GAA GGC CTC TGT GCC, #885: TGC TGT GCT TAT TAT AGC ATG ATA GC, #984: GTG AAA AAG ACA GTC ACA GGA AC, #1009: AAA TTT GGT GTG CCT
BBAPRO 36373 7-3-01
158
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
GAG GAG GAG, #1120: GAG CGG GAG GAG CAG GAA CGG AAG C. The 3P primers were as follows: #984: ACC GCC TTA CAC TGG CTG AGG CTC GGG ATC, #1009: TTA CTA TTT ACG AAG AAG CTG CTT GGC, #1034: ACC GCC TTA TGT TGA CAT TGT ACG TAC CAC, #1120: TTA CTA CTC CCG AGA CAG CTG AGA ACT TG, #1279: TCA TTA GTG CCC GAT GGC ACC ACA T. PCR fragments were directly cloned into pCR3.1 TA cloning vector (Invitrogen). Orientation and sequence of subfragment clones were con¢rmed by sequencing reactions (Taqfs Perkin-Elmer) on ABI Model 373, (Perkin-Elmer Bioscience). The GST-Tat subfragments were previously described [31]. 2.3. Production of GST-TAT fusion proteins GST-Tat subfragment constructs were used to transform BL21(DE3)pLysS cells (Promega). Protein production and puri¢cation on GST-Sepharose 4B beads (Amersham-Pharmacia) were carried out by standard methods as described [31].
cells (Invitrogen) by one cycle of freeze-thaw in bu¡er B (20 mM Tris-Cl, pH 8.0/5 mM MgCl2 /10% glycerol/0.1% NP-40) supplemented with 420 mM KCl, 1 mM PMSF, 20 Wg/ml aprotinin, 5 Wg/ml leupeptin, and 10 Wg/ml pepstatin. 5 Wg of the GST-Tat subfragments were incubated with 100 Wl of the HAhTAFII 250 containing High5 extract, with 30 Wl GST-Sepharose beads (50% slurry) at 4³C for 60 min. The beads were washed, samples were resolved on reducing SDS-PAGE gels and transferred to nitrocellulose membranes. After blocking with 5% milk in TBST (10 mM Tris-Cl, pH 8.0/150 mM NaCl/ 0.05% Tween 20), the blot was incubated at room temperature with 100 Wg of monoclonal anti-HA antibody HA.11 (Covance) in 25 ml of 1% milk/ TBST for 2 h, washed in TBST twice, and incubated at room temperature for 30 min with goat antimouse horseradish peroxidase conjugated anti-mouse IgG (Santa Cruz Biotechnology) at 1/2500 dilution in 1% milk/TBST. The ¢lter was washed three times with TBST, once with TBS and was developed with Super Signal West Pico Chemiluminescent Substrate
2.4. GST pull-downs The various TAFII 250 subfragments were translated in vitro (2 Wg/100 Wl reaction), in the TnT Coupled Reticulocyte Lyste System (Promega) from the T7 promoter with [35 S]methionine (ICN). GSTSepharose 4B beads (Amersham-Pharmacia) were prewashed in 15 ml cold BB (20 mM HEPES, pH 7.9/100 mM KCl/12.5 mM MgCl2 /0.1 mM DTT/ 0.2% Non-Idet P-40/17% glycerol) with 0.5 mg/ml BSA, spun at 2000 rpm, and resuspended in 1 ml BB without BSA at 50% v/v. For the immunoprecipitations, 5 Wg of the GST fusion protein was combined with 24 Wl reaction mix of 35 S-TAFII 250 fragments and 30 Wl of the prewashed GST-Sepharose beads (50% slurry); the ¢nal volume was adjusted to 200 Wl. The reaction was incubated for 2 h at 4³C. The beads were washed twice in Wash Bu¡er (50 mM Tris-Cl pH 7.9/150 mM NaCl/0.2% NP-40), and samples were resolved on reducing SDS-PAGE gels and quanti¢ed by phosphorimaging (AmershamPharmacia). HA-tagged full-length human TAFII 250 was prepared from recombinant baculovirus-infected High5
Fig. 1. Tat1-67 binds to hTAFII 250. GST-Tat101 and Tat subfragments GST-Tat1-67, GST-Tat67-101 and control GSTSnap23 were tested for their ability to interact with hTAFII 250. The relative strength of the interactions is shown normalized to GST-Tat67-101. The location of the TAFII 250 AT domain and Rap74-binding region is schematized at the top of the ¢gure.
BBAPRO 36373 7-3-01
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
159
Fig. 2. Two segments of Tat bind to two discontinuous segments of TAFII 250. The TAFII 250 fragment between amino acid residues 848 and 1279 contains both the AT domain and Rap74-binding region of TAFII 250. Fragments within this region were generated, in vitro translated in the presence of [35 S]methionine and tested for their binding to either GST alone, GST-Tat101, GST-Tat1-67 or GST-Tat67-101 by pull-down assays. Bound material was analyzed by polyacrylamide gel electrophoresis. Binding was visualized, as shown on the bottom left and quantitated by phosphorimaging, as summarized on the upper right. The data shown are representative ; binding of each fragment was tested at least three times. +, signi¢cant binding above the background GST alone control; 3, no binding above background.
(Pierce). Quanti¢cation of bands was determined by densitometry (Amersham-Pharmacia). 3. Results and discussion 3.1. Tat binds to two distinct domains within the TAFII 250 molecule A yeast two-hybrid screen recently identi¢ed the TFIID component, TAFII 250, as a Tat-interacting protein [27]. Preliminary studies mapped Tat binding to a 433 amino acid fragment of TAFII 250 that extends from residues 848 to 1279 (Fig. 1, top). The present studies were undertaken to ¢ne map the regions of interaction in both Tat and TAFII 250. The Tat bait used in the original yeast two-hybrid screen was a C-terminal polypeptide extending from amino acid 67 to 101 (Tat67-101), which is responsible for Tat101-mediated repression of transcription
of a class I promoter [35]. The interaction between Tat67-101 and TAFII 250, although weak, was con¢rmed in subsequent in vitro GST pull-down assays where a GST-Tat67-101 fusion protein retained the TAFII 250 fragment [27]. However, full-length Tat101 fused to GST (GST-Tat101) bound TAFII 250 consistently and signi¢cantly better than the GST-Tat67101 (Fig. 1). This ¢nding raised the possibility that the N-terminal peptide 1-67 of Tat either stabilizes the binding of the C-terminal 67^101 segment to TAFII 250 or binds to it independently or both. To explore these possibilities, the ability of the GSTTat1-67 fragment to bind TAFII 250 was examined (Fig. 1). Surprisingly, whereas the original C-terminal Tat67-101 bait used to identify TAFII 250 in vivo binds TAFII 250 weakly in vitro on a GST pull-down, the N-terminal Tat1-67, which was not part of the original bait, binds TAFII 250 strongly in vitro, and as e¡ectively as intact Tat101. Thus, the Tat1-67 peptide is able to bind TAFII 250 independent of
BBAPRO 36373 7-3-01
160
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
Fig. 3. Tat1-67 binds to both the TAFII 250 AT domain and the Rap74 region, although the binding requirements di¡er. The binding of in vitro translated, [35 S]methionine labeled TAFII 250 fragments 848^1034 (lower left) or 1120^1279 (lower right) were tested for their ability to bind various deletion constructs of GST-Tat, as schematized in the upper part of the ¢gure.
the binding of the Tat67-101 peptide. Indeed, the Nterminal domain binding appears to contribute most of the binding observed with the intact Tat101. The possibility that the 1-67 peptide stabilizes the binding of the 67^101 segment has not been excluded. Indeed, this is likely, but remains to be demonstrated. The TAFII 250 fragment isolated in the yeast twohybrid screen spanned the segment 848^1279 amino acids. The region includes both the acetyl transferase domain and a region that binds to the Rap74 component of the general transcription factor, TFIIF, which plays pivotal roles in both transcription elongation and initiation [1,8,9,36]. Since the above studies demonstrated that the N- and C-terminal segments of Tat bind to TAFII 250 independently, we considered the possibility that they might be interacting with di¡erent regions of the TAFII 250 fragment. Therefore, it became of interest to determine where the Tat segments bind within the TAFII 250 fragment. To this end, a series of overlapping TAFII 250 fragments, spanning the length of the
TAFII 250 fragment, were generated, as schematized in Fig. 2. Each of the TAFII 250 subfragments was assayed for its ability to be bound by GST fusions of Tat101, Tat1-67, or Tat67-101. The GST fusion protein containing the full-length Tat101 binds to TAFII 250 segments derived from both the AT domain (Fig. 2, fragments A, AP, B, D, E and F) and the Rap74-binding region (Fig. 2, fragment R). Surprisingly, Tat101 did not interact with fragments I and J, which are located in the middle of the extended TAFII 250 fragment (Fig. 2). Taken together, these observations demonstrate that Tat101 binds to two discontinuous segments within the TAFII 250 fragment: 885^984 aa in the AT domain and 1120^1279 aa in the Rap74-binding region. Since both the Tat1-67 and Tat67-101 fragments bind to TAFII 250, we next mapped each of their binding sites on TAFII 250. The GST fusion protein containing the Tat67-101 fragment binds TAFII 250 fragments spanning the region 885^1034 aa, containing the AT enzymatic activity of the molecule (Fig. 2,
BBAPRO 36373 7-3-01
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
fragments A, AP and D). Although binding is weak, it is reproducible and above the background binding of GST alone. This mapping is consistent with the isolation of TAFII 250 using the Tat67-101 fragment as bait in the original yeast two-hybrid screen [27]. In addition, we have demonstrated previously that Tat67-101 inhibits the AT activity of TAFII 250 [27], further supporting the conclusion that the Cterminal domain of Tat binds to the AT domain of TAFII 250. In contrast to the Tat67-101 fragment, the Tat1-67 fragment binds to two discontinuous segments of TAFII 250 (Fig. 2, fragments A, AP B, D, E, F and R). Thus, Tat1-67 interacts with fragments spanning the region 885^984 aa, and 1120^1279 aa. It does not bind to the TAFII 250 segment between amino acids 984 and 1120, which contains much of the AT domain (see fragments E and F, Fig. 2). In conclusion, we have identi¢ed three interaction sites of Tat101 on TAFII 250 (see ahead to Fig. 4). 3.2. The cysteine-rich domain of Tat is necessary for binding to TAFII 250 Since Tat1-67 has two distinct binding sites within the TAFII 250 fragment, we next determined which of Tat's structural domains were necessary for this interaction. The N-terminal Tat protein, extending from amino acids 1 to 72, transactivates the HIV LTR [37]. It contains three domains involved in transactivation: an acidic domain (aa 2^6), a cysteine-rich domain (aa 18^36) and a glutamine-rich domain (aa 56^67) (Fig. 3). The acidic and cysteine domains are essential for transactivation, while the glutamine-rich domain augments transactivation but is not absolutely required [37]. To further characterize the binding of Tat to TAFII 250, deletions of each of these domains were tested for their ability to bind to TAFII 250 (Fig. 3). As demonstrated above, Tat101 and Tat1-67 bound to both TAFII 250 fragments 848^1034 aa and 1120^1279 aa (Fig. 3, lanes 2 and 5). Deletion of the glutamine-rich domain (Fig. 3, lanes 3) did not a¡ect the binding of Tat to either TAFII 250 fragment. Thus, the glutamine domain does not appear to be necessary for either interaction. Binding to the AT domain of TAFII 250 (848^ 1034 aa) is abrogated by deletion of the cysteine-rich domain (Fig. 3, lane 7) but not the acidic domain
161
Fig. 4. Representation of the interaction of HIV Tat with TAFII 250. HIV Tat (represented by grey ribbon) interacts with TAFII 250 (represented by solid black line) at three distinct sites: Tat aa 18^36 (cysteine region) interacts but does not inhibit within TAFII 250 AT region (885^984), Tat aa 48^56 (basic region) interacts within TAFII 250 Rap74-binding region (1120^1279), and Tat aa 67^101 interacts and inhibits TAFII 250 AT domain (885^1034 aa).
(Fig. 3, lane 6). In contrast, Tat interaction with the TAFII 250 Rap74-binding domain (1120^1279 aa) does not depend on the acidic, cysteine-rich or glutamine domains (Fig. 3, lanes 3, 6, 7, 8). Taken together, these data demonstrate distinct requirements for the interaction of Tat with TAFII 250. Binding to the AT domain of TAFII 250 is dependent on the cysteine-rich domain of Tat. The interaction of Tat1-67 with the Rap74-binding region maps to the Tat segment aa 36^56, which spans the basic domain (48^56 aa) of Tat. 3.3. HIV LTR does not require TAFII 250 The ability of Tat to bind to TAFII 250 and inhibit its AT activity is consistent with Tat's marked repression of transcription initiation of some cellular promoters, such as the MHC class I promoter [27]. However, it is not consistent with the well characterized ability of Tat to activate transcription of the HIV LTR. One possible explanation for this disparity might be that the HIV LTR does not depend on TAFII 250, whereas the class I promoter does. To examine this possibility, the ability of each of these two promoters to function in the absence of a functional TAFII 250 was determined. The hamster cell line, tsBN462, contains a temperature sensitive mutation in the CCG1 gene that encodes TAFII 250 [32]. At the permissive temperature of 32³C, TAFII 250 is functional and the cells grow normally. However, at
BBAPRO 36373 7-3-01
162
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
Table 1 HIV LTR, unlike the class I promoter, does not depend on TAFII 250 Promoter HIV LTR Class I
Promoter activitya 32³C
39³C
31.0 þ 1.3 71.3 þ 1.5
63.3 þ 5.4 19.4 þ 3.5
Rb 2.0 0.27
TsBN462 cells, containing a point mutation in the TAFII 250 gene, were transfected with either the HIVLTR (pBennCAT) or a class I promoter construct (-68CAT). Twenty-four hours after transfection, cells were either shifted to the restrictive temperature (39³C) or left at the permissive temperature (32³C) for an additional 24 h. CAT activity was determined, as described in Section 2. Results are presented as CAT activity at the permissive and restrictive temperatures and as the ratio of CAT activity at the restrictive temperature, relative to the permissive temperature. a CAT activity (% acetylation of chloramphenicol) generated by the promoter at the indicated temperature. b R = relative promoter activity calculated as corrected promoter activity at 39³C/corrected promoter activity at 32³C.
the restrictive temperature of 39.5³C, TAFII 250 is not functional and the transcription of a large number of genes ^ particularly those involved in cell cycle regulation ^ ceases and the cells arrest at G1 [32]. Therefore, the activities of the HIV LTR and the class I promoter were compared at the permissive and restrictive temperature. As shown in Table 1, the MHC class I promoter is active in these cells at the permissive temperature, but signi¢cantly repressed at the restrictive temperature, consistent with its dependence on TAFII 250. The HIV LTR, in the absence of Tat, displayed basal activity in the tsBN462 at the permissive temperature. However, in marked contrast to the MHC class I promoter, the HIV LTR not only retained activity at the restrictive temperature of 39³C, but was signi¢cantly more active than at the permissive 32³C. Tat further enhances HIV LTR activity in these cells at both temperatures (data not shown). These data demonstrate that the HIV LTR does not depend on a functional TAFII 250, and provide a possible explanation for the failure of Tat to repress viral promoter activity. 3.4. Summary of Tat binding to the TAFII 250 fragment The above studies demonstrate that the interaction
of Tat with TAFII 250 is complex and involves three distinct binding sites. A pictorial representation of our ¢ndings is shown in Fig. 4. (It is important to note that we have no evidence that these interactions occur simultaneously.) The C-terminal segment of Tat (Tat67-101) binds to the AT domain of TAFII 250 between aa 885 and 1034, resulting in inhibition of the enzymatic activity which is correlated with Tat's repression of TAFII 250-dependent promoters. The N-terminal domain of Tat (Tat1-67) binds to two discontinuous sites on the TAFII 250 fragment: 885^984 aa within the AT domain and 1120^1279 aa within the Rap74-binding region. The cysteine-rich region of Tat is required for binding to the AT domain, whereas the basic domain appears to be necessary for binding to the Rap74-binding region. Although Tat binds to the Rap74-binding region, we did not observe any competition for binding to the TAFII 250 fragment between Tat101 and Rap74 (data not shown) nor did Rap74 phosphorylate Tat101 in the presence of the TAFII 250 fragment (data not shown). These data would suggest that Tat binding to the Rap74-binding domain is to a distinct site. In summary, the present study demonstrates that the interaction between Tat and TAFII 250 is complex, involving at least three distinct sites of interaction. The binding of the C-terminal segment of Tat, 67^101 aa, to the AT domain of TAFII 250 results in inhibition of AT activity and consequent repression of transcription initiation of TAFII 250-dependent promoters [27]. The N-terminal segment of Tat, 1^ 67 aa, binds to two discontinuous sites on TAFII 250 ^ the AT domain and the RAP74-binding region ^ which may stabilize the interaction and e¡ect greater repression. The failure of Tat to repress the HIV LTR promoter may be explained by the observation that transcription initiation of the HIV LTR does not require TAFII 250. Acknowledgements The authors gratefully acknowledge Kevin Howcroft, John Brady and Fatah Kashanchi for helpful discussions, and Klaus Strebel for critical review of the manuscript.
BBAPRO 36373 7-3-01
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
References [1] R.S. Chambers, B.Q. Wang, Z.F. Burton, M.E. Dahmus, J. Biol. Chem. 270 (1995) 14962^14969. [2] R.D. Kornberg, Trends Cell Biol. 9 (1999) M46^M49. [3] S.R. Albright, R. Tjian, Gene 242 (2000) 1^13. [4] O. Bensaude, F. Bonnet, C. Casse, M.F. Dubois, V.T. Nguyen, B. Palancade, Biochem. Cell Biol. 77 (1999) 249^255. [5] D. Reines, R.C. Conaway, J.W. Conaway, Curr. Opin. Cell Biol. 11 (1999) 342^346. [6] H. Lu, L. Zawel, L. Fisher, J.M. Egly, D. Reinberg, Nature 358 (1992) 620^621. [7] R. Skiekhattar, F. Mermelstein, R.P. Fisher, R. Drapkin, B. Dynlacht, H.C. Wessling, D.O. Morgan, D. Reinberg, Nature 374 (1995) 283^287. [8] C.-h. Chang, C.F. Kostrub, Z.F. Burton, J. Biol. Chem. 268 (1993) 20482^20489. [9] T. O'Brien, R. Tjian, Mol. Cell 1 (1998) 905^911. [10] G. Orphanides, T. Lagrange, D. Reinberg, Genes Dev. 10 (1996) 2657^2683. [11] C. Mizzen, X. Yang, T. Kokubo, J. Brownell, A. Bannister, T. Owen-Hughes, J. Workman, L. Wang, S. Berger, T. Kouzarides, Y. Nakatani, C.D. Allis, Cell 87 (1996) 1261^1270. [12] B. Cullen, FASEB J. 5 (1991) 2361^2368. [13] W.A. Haseltine, FASEB J. 5 (1991) 2349^2360. [14] K. Sastry, H. Raghava, R. Pandita, K. Tatpal, B. Aggarwal, J. Biol. Chem. 265 (1990) 20091^20093. [15] G. Scala, M. Ruocco, C. Ambrosino, M. Mallardo, V. Giordano, F. Baldasarre, E. Dragonetti, I. Quinto, S. Venuta, J. Exp. Med. 179 (1994) 961^971. [16] W.P. Tansey, S. Ruppert, R. Tjian, W. Herr, Genes Dev. 8 (1994) 2756^2769. [17] M.O. Westendorp, V.A. Shatrov, K. Schulze-Ostho¡, F. Rainer, M. Kraft, M. Los, P.H. Krammer, W. Droge, V. Lehmann, EMBO J. 14 (1995) 546^554. [18] M. Garber, T. Mayall, E. Suess, J. Meisenhelder, N. Thompson, K. Jones, Mol. Cell. Biol. 20 (2000) 6958^6969. [19] C.A. Parada, R.G. Roeder, Nature 384 (1996) 375^378.
163
[20] L. Garcia-Martinez, G. Mavankal, J. Neveu, W. Lane, I. Ivanov, R. Gaynor, EMBO J. 16 (1997) 2836^2850. [21] R.S. Chambers, M.E. Dahmus, J. Biol. Chem. 269 (1994) 26243^26248. [22] N. Marshall, G.K. Dahmus, M.E. Dahmus, J. Biol. Chem. 273 (1998) 31726^31730. [23] R. Taube, K. Fujinaga, J. Wimmer, M. Barboric, B.M. Peterlin, Virology 264 (1999) 245^253. [24] T.K. Howcroft, K. Strebel, M. Martin, D.S. Singer, Science 260 (1993) 91^93. [25] T.K. Howcroft, L. Palmer, J. Brown, B. Rellahan, F. Kashanchi, J. Brady, D.S. Singer, Immunity 3 (1995) 127^138. [26] I. Carroll, J. Wang, T.K. Howcroft, D.S. Singer, Mol. Immun. 35 (1998) 1171^1178. [27] J.D. Weissman, J. Brown, T.K. Howcroft, J. Hwang, A. Chawla, P. Roche, L. Schiltz, Y. Nakatani, D.S. Singer, Proc. Natl. Acad. Sci. USA 95 (1998) 11601^11606. [28] E. Pocsik, M. Higuchi, B. Aggarwal, Lymph. Cyto. Res. 11 (1992) 317^325. [29] S.F. Purvis, D. Georges, T. Williams, M. Lederman, Cell. Immunol. 144 (1992) 32^42. [30] S. Flores, J. Marecki, K. Harper, S. Bose, S. Nelson, J. McCord, Proc. Natl. Acad. Sci. USA 90 (1993) 7632^7636. [31] J. Brown, T.K. Howcroft, D.S. Singer, J. AIDS 17 (1998) 9^ 16. [32] T. Hayashida, T. Sekiguchi, E. Noguchi, H. Sunamoto, T. Ohba, T. Nishimoto, Gene 141 (1994) 267^270. [33] T. Nakashima, T. Sekiguchi, H. Sunamoto, K. Yura, S. Tomoda, M. Go, J. Kere, D. Schlessinger, T. Nishimoto, Gene 141 (1994) 193^200. [34] J. Leonard, C. Parrot, A. Buckler-White, W. Turner, E. Ross, M. Martin, A. Rabson, J. Virol. 63 (1989) 4919^4924. [35] T.K. Howcroft, J. Richardson, D.S. Singer, EMBO J. 12 (1993) 3163^3169. [36] J. Archambault, G. Pan, G.K. Dahmus, M. Cartier, N.F. Marshall, S. Zhang, M.E. Dahmus, J. Greenblatt, J. Biol. Chem. 273 (1998) 27593^27601. [37] B.R. Cullen, Cell 46 (1986) 973^982.
BBAPRO 36373 7-3-01
Extensive interactions between HIV TAT and TAFII 250 Jocelyn D. Weissman *, Jae Ryoung Hwang, Dinah S. Singer Experimental Immunology Branch, NCI, NIH, 9000 Rockville Pike, Bethesda, MD 20892, USA Received 27 October 2000; received in revised form 28 December 2000; accepted 28 December 2000
Abstract The HIV transactivator, Tat, has been shown to be capable of potent repression of transcription initiation. Repression is mediated by the C-terminal segment of Tat, which binds the TFIID component, TAFII 250, although the site(s) of interaction were not defined previously. We now report that the interaction between Tat and TAFII 250 is extensive and involves multiple contacts between the Tat protein and TAFII 250. The C-terminal domain of Tat, which is necessary for repression of transcription initiation, binds to a segment of TAFII 250 that encompasses its acetyl transferase (AT) domain (885^1034 amino acids (aa)). Surprisingly, the N-terminal segment of Tat, which contains its activation domains, also binds to TAFII 250 and interacts with two discontinuous segments of TAFII 250 located between 885 and 984 aa and 1120 and 1279 aa. Binding of Tat to the 885^984 aa segment of TAFII 250 requires the cysteine-rich domain of Tat, but not the acidic or glutamine-rich domains. Binding by the N-terminal domain of Tat to the 1120^1279 aa TAFII 250 segment does not involve the acidic, cysteine- or glutamine-rich domains. Repression of transcription initiation by Tat requires functional TAFII 250. We now demonstrate that transcription of the HIV LTR does not depend on TAFII 250 which may account for its resistance to Tat mediated repression. ß 2001 Elsevier Science B.V. All rights reserved. Keywords: TAFII250; HIV Tat
1. Introduction Transcription of eukaryotic genes requires the ordered interaction of a series of general transcription factors (GTF) that result in the recruitment of the RNA polymerase II (RNAP) complex, transcription initiation and subsequent elongation [1,2]. The process is initiated by the binding of TFIID to the promoter, which in turn triggers the formation of a preinitiation complex consisting of TFIIB, TFIIE, TFIIF and TFIIH (reviewed in [3]). Interactions between the RNA polymerase complex and the GTFs
* Corresponding author. Fax: +1-301-480-8499; E-mail: [email protected]
result in phosphorylation of the CTD of the polymerase and transcription initiation and elongation (reviewed in [4,5]). Phosphorylation is e¡ected by a kinase activity in TFIIH [6,7]; elongation is regulated by TFIIF [1,8,9]. Initiation depends on TFIID, which contains the TATA-binding protein (TBP) in association with the TBP-associated factors (TAFs) [10]. TAFII 250 is the largest component of TFIID, and is known to contain acetyl transferase activity (AT) [11]. The acetyl transferase activity of TAFII 250 is presumed to be important for initiation, although neither its substrate nor its mechanism of action is known. The HIV gene product, Tat, functions as a potent activator of the viral LTR increasing promoter activity by up to 100-fold [12^17]. Activation, which is
0167-4838 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 0 1 ) 0 0 1 3 5 - 2
BBAPRO 36373 7-3-01
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
mediated by the N-terminal 67 amino acids (aa) of the 101 amino acid full-length Tat, depends on the interaction between Tat, the P-TEFb complex component cyclin T1, and the HIV TAR RNA element [18]. This interaction stimulates the CTD kinase [19,20], resulting in increased phosphorylation of the RNAP CTD and enhanced elongation. In addition, Tat binds to and inhibits a CTD phosphatase, reducing dephosphorylation and thereby also enhancing rates of elongation [21,22]. Thus, the major e¡ect of Tat on HIV LTR promoter activity is to enhance transcription elongation (reviewed in [23]). Analysis of the Tat protein has demonstrated that two N-terminal domains of the Tat protein are necessary for this activation: an acidic domain located between aa 2 and 6 and a cysteine-rich domain spanning aa 20^48. Additionally, a glutamine-rich domain, aa 57^67, contributes to Tat's activation. Cterminal sequences, extending beyond amino acid 72 that are encoded by the second exon of the Tat gene, are not necessary for activation of the LTR by Tat. Although Tat stimulates transcription elongation of the HIV LTR in the presence of the HIV TAR element, it can repress transcription initiation of other promoters, in the absence of TAR. We have shown previously that Tat represses the activity of the promoters of major histocompatibility complex (MHC) class I, L2 -microglobulin and certain viral genes [24^26]. This repression occurs at the level of transcription initiation, is independent of TAR and is mediated through the interaction of Tat with TAFII 250, which inhibits the intrinsic AT activity of TAFII 250 [27]. Both repression of transcription and of TAFII 250 AT activity are mediated by the C-terminal polypeptide (67^101 aa) of Tat, which is not required for activation of the HIV LTR [24,25,28^30]. Thus, the activities of activation and repression are separable. Previous studies have demonstrated that Tat interacts with TAFII 250 in vivo both in HeLa cells and yeast, and in vitro in GST pull-down assays [27]. However, ¢ne mapping of the interaction sites of the two molecules was not performed. The present studies were undertaken to further characterize the interaction of Tat with TAFII 250. We report that at least three distinct regions of the Tat protein mediate binding to TAFII 250. Two of these reside in the Nterminal region of Tat, which is not involved in re-
157
pression, and anchor the binding of Tat to the TAFII 250. Surprisingly, binding of the N-terminal domains of Tat to TAFII 250 occurs at two discontinuous sites, one of which spans the binding site for Rap74. The third Tat region is in the C-terminal peptide which binds to the AT domain of TAFII 250, consistent with its inhibition of AT activity and repression of transcription initiation. Interestingly, transcription of the HIV LTR does not depend on TAFII 250, consistent with Tat's failure to repress the HIV LTR. 2. Materials and methods 2.1. Cell culture and transfections The tsBN462 cells containing a point mutation in the TAFII 250 gene, derived from BHK cells as described [32,33], were obtained from T. Sekiguschi (Salk Institute). The cells were maintained at 32³C, 7.5% CO2 in Dulbecco's modi¢ed Eagle's medium (Bio£uids) with 10% fetal calf serum. Transfections were essentially as previously described. Brie£y, cells were transfected with 5 Wg of pBennCAT(HIVLTR) or -68CAT (class I promoter with 68 bp of upstream £anking sequence) [34] using CaPO4 at 32³C. RSVluciferase was used as an internal transfection e¤ciency control. Following transfection, cells were incubated at 32³C for 24 h, at which time they were refed with fresh medium and either left at 32³C or shifted to 39.5³C. After an additional 16 h, cells were harvested and assayed for CAT activity as previously described [27]. Activity was corrected to protein concentration. 2.2. DNA constructs The TAFII 250 clone was isolated from a yeast two-hybrid screen using Tat as bait, as previously described [27]. TAFII 250 subfragments were constructed via PCR ampli¢cation using the TAFII 250 clone as template as follows. The 5P primer for each subfragment (see Fig. 2) contains the consensus Kozak sequence (ACC GCC ATG GGA CAC) followed by speci¢c sequence #848: GAA GGC CTC TGT GCC, #885: TGC TGT GCT TAT TAT AGC ATG ATA GC, #984: GTG AAA AAG ACA GTC ACA GGA AC, #1009: AAA TTT GGT GTG CCT
BBAPRO 36373 7-3-01
158
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
GAG GAG GAG, #1120: GAG CGG GAG GAG CAG GAA CGG AAG C. The 3P primers were as follows: #984: ACC GCC TTA CAC TGG CTG AGG CTC GGG ATC, #1009: TTA CTA TTT ACG AAG AAG CTG CTT GGC, #1034: ACC GCC TTA TGT TGA CAT TGT ACG TAC CAC, #1120: TTA CTA CTC CCG AGA CAG CTG AGA ACT TG, #1279: TCA TTA GTG CCC GAT GGC ACC ACA T. PCR fragments were directly cloned into pCR3.1 TA cloning vector (Invitrogen). Orientation and sequence of subfragment clones were con¢rmed by sequencing reactions (Taqfs Perkin-Elmer) on ABI Model 373, (Perkin-Elmer Bioscience). The GST-Tat subfragments were previously described [31]. 2.3. Production of GST-TAT fusion proteins GST-Tat subfragment constructs were used to transform BL21(DE3)pLysS cells (Promega). Protein production and puri¢cation on GST-Sepharose 4B beads (Amersham-Pharmacia) were carried out by standard methods as described [31].
cells (Invitrogen) by one cycle of freeze-thaw in bu¡er B (20 mM Tris-Cl, pH 8.0/5 mM MgCl2 /10% glycerol/0.1% NP-40) supplemented with 420 mM KCl, 1 mM PMSF, 20 Wg/ml aprotinin, 5 Wg/ml leupeptin, and 10 Wg/ml pepstatin. 5 Wg of the GST-Tat subfragments were incubated with 100 Wl of the HAhTAFII 250 containing High5 extract, with 30 Wl GST-Sepharose beads (50% slurry) at 4³C for 60 min. The beads were washed, samples were resolved on reducing SDS-PAGE gels and transferred to nitrocellulose membranes. After blocking with 5% milk in TBST (10 mM Tris-Cl, pH 8.0/150 mM NaCl/ 0.05% Tween 20), the blot was incubated at room temperature with 100 Wg of monoclonal anti-HA antibody HA.11 (Covance) in 25 ml of 1% milk/ TBST for 2 h, washed in TBST twice, and incubated at room temperature for 30 min with goat antimouse horseradish peroxidase conjugated anti-mouse IgG (Santa Cruz Biotechnology) at 1/2500 dilution in 1% milk/TBST. The ¢lter was washed three times with TBST, once with TBS and was developed with Super Signal West Pico Chemiluminescent Substrate
2.4. GST pull-downs The various TAFII 250 subfragments were translated in vitro (2 Wg/100 Wl reaction), in the TnT Coupled Reticulocyte Lyste System (Promega) from the T7 promoter with [35 S]methionine (ICN). GSTSepharose 4B beads (Amersham-Pharmacia) were prewashed in 15 ml cold BB (20 mM HEPES, pH 7.9/100 mM KCl/12.5 mM MgCl2 /0.1 mM DTT/ 0.2% Non-Idet P-40/17% glycerol) with 0.5 mg/ml BSA, spun at 2000 rpm, and resuspended in 1 ml BB without BSA at 50% v/v. For the immunoprecipitations, 5 Wg of the GST fusion protein was combined with 24 Wl reaction mix of 35 S-TAFII 250 fragments and 30 Wl of the prewashed GST-Sepharose beads (50% slurry); the ¢nal volume was adjusted to 200 Wl. The reaction was incubated for 2 h at 4³C. The beads were washed twice in Wash Bu¡er (50 mM Tris-Cl pH 7.9/150 mM NaCl/0.2% NP-40), and samples were resolved on reducing SDS-PAGE gels and quanti¢ed by phosphorimaging (AmershamPharmacia). HA-tagged full-length human TAFII 250 was prepared from recombinant baculovirus-infected High5
Fig. 1. Tat1-67 binds to hTAFII 250. GST-Tat101 and Tat subfragments GST-Tat1-67, GST-Tat67-101 and control GSTSnap23 were tested for their ability to interact with hTAFII 250. The relative strength of the interactions is shown normalized to GST-Tat67-101. The location of the TAFII 250 AT domain and Rap74-binding region is schematized at the top of the ¢gure.
BBAPRO 36373 7-3-01
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
159
Fig. 2. Two segments of Tat bind to two discontinuous segments of TAFII 250. The TAFII 250 fragment between amino acid residues 848 and 1279 contains both the AT domain and Rap74-binding region of TAFII 250. Fragments within this region were generated, in vitro translated in the presence of [35 S]methionine and tested for their binding to either GST alone, GST-Tat101, GST-Tat1-67 or GST-Tat67-101 by pull-down assays. Bound material was analyzed by polyacrylamide gel electrophoresis. Binding was visualized, as shown on the bottom left and quantitated by phosphorimaging, as summarized on the upper right. The data shown are representative ; binding of each fragment was tested at least three times. +, signi¢cant binding above the background GST alone control; 3, no binding above background.
(Pierce). Quanti¢cation of bands was determined by densitometry (Amersham-Pharmacia). 3. Results and discussion 3.1. Tat binds to two distinct domains within the TAFII 250 molecule A yeast two-hybrid screen recently identi¢ed the TFIID component, TAFII 250, as a Tat-interacting protein [27]. Preliminary studies mapped Tat binding to a 433 amino acid fragment of TAFII 250 that extends from residues 848 to 1279 (Fig. 1, top). The present studies were undertaken to ¢ne map the regions of interaction in both Tat and TAFII 250. The Tat bait used in the original yeast two-hybrid screen was a C-terminal polypeptide extending from amino acid 67 to 101 (Tat67-101), which is responsible for Tat101-mediated repression of transcription
of a class I promoter [35]. The interaction between Tat67-101 and TAFII 250, although weak, was con¢rmed in subsequent in vitro GST pull-down assays where a GST-Tat67-101 fusion protein retained the TAFII 250 fragment [27]. However, full-length Tat101 fused to GST (GST-Tat101) bound TAFII 250 consistently and signi¢cantly better than the GST-Tat67101 (Fig. 1). This ¢nding raised the possibility that the N-terminal peptide 1-67 of Tat either stabilizes the binding of the C-terminal 67^101 segment to TAFII 250 or binds to it independently or both. To explore these possibilities, the ability of the GSTTat1-67 fragment to bind TAFII 250 was examined (Fig. 1). Surprisingly, whereas the original C-terminal Tat67-101 bait used to identify TAFII 250 in vivo binds TAFII 250 weakly in vitro on a GST pull-down, the N-terminal Tat1-67, which was not part of the original bait, binds TAFII 250 strongly in vitro, and as e¡ectively as intact Tat101. Thus, the Tat1-67 peptide is able to bind TAFII 250 independent of
BBAPRO 36373 7-3-01
160
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
Fig. 3. Tat1-67 binds to both the TAFII 250 AT domain and the Rap74 region, although the binding requirements di¡er. The binding of in vitro translated, [35 S]methionine labeled TAFII 250 fragments 848^1034 (lower left) or 1120^1279 (lower right) were tested for their ability to bind various deletion constructs of GST-Tat, as schematized in the upper part of the ¢gure.
the binding of the Tat67-101 peptide. Indeed, the Nterminal domain binding appears to contribute most of the binding observed with the intact Tat101. The possibility that the 1-67 peptide stabilizes the binding of the 67^101 segment has not been excluded. Indeed, this is likely, but remains to be demonstrated. The TAFII 250 fragment isolated in the yeast twohybrid screen spanned the segment 848^1279 amino acids. The region includes both the acetyl transferase domain and a region that binds to the Rap74 component of the general transcription factor, TFIIF, which plays pivotal roles in both transcription elongation and initiation [1,8,9,36]. Since the above studies demonstrated that the N- and C-terminal segments of Tat bind to TAFII 250 independently, we considered the possibility that they might be interacting with di¡erent regions of the TAFII 250 fragment. Therefore, it became of interest to determine where the Tat segments bind within the TAFII 250 fragment. To this end, a series of overlapping TAFII 250 fragments, spanning the length of the
TAFII 250 fragment, were generated, as schematized in Fig. 2. Each of the TAFII 250 subfragments was assayed for its ability to be bound by GST fusions of Tat101, Tat1-67, or Tat67-101. The GST fusion protein containing the full-length Tat101 binds to TAFII 250 segments derived from both the AT domain (Fig. 2, fragments A, AP, B, D, E and F) and the Rap74-binding region (Fig. 2, fragment R). Surprisingly, Tat101 did not interact with fragments I and J, which are located in the middle of the extended TAFII 250 fragment (Fig. 2). Taken together, these observations demonstrate that Tat101 binds to two discontinuous segments within the TAFII 250 fragment: 885^984 aa in the AT domain and 1120^1279 aa in the Rap74-binding region. Since both the Tat1-67 and Tat67-101 fragments bind to TAFII 250, we next mapped each of their binding sites on TAFII 250. The GST fusion protein containing the Tat67-101 fragment binds TAFII 250 fragments spanning the region 885^1034 aa, containing the AT enzymatic activity of the molecule (Fig. 2,
BBAPRO 36373 7-3-01
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
fragments A, AP and D). Although binding is weak, it is reproducible and above the background binding of GST alone. This mapping is consistent with the isolation of TAFII 250 using the Tat67-101 fragment as bait in the original yeast two-hybrid screen [27]. In addition, we have demonstrated previously that Tat67-101 inhibits the AT activity of TAFII 250 [27], further supporting the conclusion that the Cterminal domain of Tat binds to the AT domain of TAFII 250. In contrast to the Tat67-101 fragment, the Tat1-67 fragment binds to two discontinuous segments of TAFII 250 (Fig. 2, fragments A, AP B, D, E, F and R). Thus, Tat1-67 interacts with fragments spanning the region 885^984 aa, and 1120^1279 aa. It does not bind to the TAFII 250 segment between amino acids 984 and 1120, which contains much of the AT domain (see fragments E and F, Fig. 2). In conclusion, we have identi¢ed three interaction sites of Tat101 on TAFII 250 (see ahead to Fig. 4). 3.2. The cysteine-rich domain of Tat is necessary for binding to TAFII 250 Since Tat1-67 has two distinct binding sites within the TAFII 250 fragment, we next determined which of Tat's structural domains were necessary for this interaction. The N-terminal Tat protein, extending from amino acids 1 to 72, transactivates the HIV LTR [37]. It contains three domains involved in transactivation: an acidic domain (aa 2^6), a cysteine-rich domain (aa 18^36) and a glutamine-rich domain (aa 56^67) (Fig. 3). The acidic and cysteine domains are essential for transactivation, while the glutamine-rich domain augments transactivation but is not absolutely required [37]. To further characterize the binding of Tat to TAFII 250, deletions of each of these domains were tested for their ability to bind to TAFII 250 (Fig. 3). As demonstrated above, Tat101 and Tat1-67 bound to both TAFII 250 fragments 848^1034 aa and 1120^1279 aa (Fig. 3, lanes 2 and 5). Deletion of the glutamine-rich domain (Fig. 3, lanes 3) did not a¡ect the binding of Tat to either TAFII 250 fragment. Thus, the glutamine domain does not appear to be necessary for either interaction. Binding to the AT domain of TAFII 250 (848^ 1034 aa) is abrogated by deletion of the cysteine-rich domain (Fig. 3, lane 7) but not the acidic domain
161
Fig. 4. Representation of the interaction of HIV Tat with TAFII 250. HIV Tat (represented by grey ribbon) interacts with TAFII 250 (represented by solid black line) at three distinct sites: Tat aa 18^36 (cysteine region) interacts but does not inhibit within TAFII 250 AT region (885^984), Tat aa 48^56 (basic region) interacts within TAFII 250 Rap74-binding region (1120^1279), and Tat aa 67^101 interacts and inhibits TAFII 250 AT domain (885^1034 aa).
(Fig. 3, lane 6). In contrast, Tat interaction with the TAFII 250 Rap74-binding domain (1120^1279 aa) does not depend on the acidic, cysteine-rich or glutamine domains (Fig. 3, lanes 3, 6, 7, 8). Taken together, these data demonstrate distinct requirements for the interaction of Tat with TAFII 250. Binding to the AT domain of TAFII 250 is dependent on the cysteine-rich domain of Tat. The interaction of Tat1-67 with the Rap74-binding region maps to the Tat segment aa 36^56, which spans the basic domain (48^56 aa) of Tat. 3.3. HIV LTR does not require TAFII 250 The ability of Tat to bind to TAFII 250 and inhibit its AT activity is consistent with Tat's marked repression of transcription initiation of some cellular promoters, such as the MHC class I promoter [27]. However, it is not consistent with the well characterized ability of Tat to activate transcription of the HIV LTR. One possible explanation for this disparity might be that the HIV LTR does not depend on TAFII 250, whereas the class I promoter does. To examine this possibility, the ability of each of these two promoters to function in the absence of a functional TAFII 250 was determined. The hamster cell line, tsBN462, contains a temperature sensitive mutation in the CCG1 gene that encodes TAFII 250 [32]. At the permissive temperature of 32³C, TAFII 250 is functional and the cells grow normally. However, at
BBAPRO 36373 7-3-01
162
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
Table 1 HIV LTR, unlike the class I promoter, does not depend on TAFII 250 Promoter HIV LTR Class I
Promoter activitya 32³C
39³C
31.0 þ 1.3 71.3 þ 1.5
63.3 þ 5.4 19.4 þ 3.5
Rb 2.0 0.27
TsBN462 cells, containing a point mutation in the TAFII 250 gene, were transfected with either the HIVLTR (pBennCAT) or a class I promoter construct (-68CAT). Twenty-four hours after transfection, cells were either shifted to the restrictive temperature (39³C) or left at the permissive temperature (32³C) for an additional 24 h. CAT activity was determined, as described in Section 2. Results are presented as CAT activity at the permissive and restrictive temperatures and as the ratio of CAT activity at the restrictive temperature, relative to the permissive temperature. a CAT activity (% acetylation of chloramphenicol) generated by the promoter at the indicated temperature. b R = relative promoter activity calculated as corrected promoter activity at 39³C/corrected promoter activity at 32³C.
the restrictive temperature of 39.5³C, TAFII 250 is not functional and the transcription of a large number of genes ^ particularly those involved in cell cycle regulation ^ ceases and the cells arrest at G1 [32]. Therefore, the activities of the HIV LTR and the class I promoter were compared at the permissive and restrictive temperature. As shown in Table 1, the MHC class I promoter is active in these cells at the permissive temperature, but signi¢cantly repressed at the restrictive temperature, consistent with its dependence on TAFII 250. The HIV LTR, in the absence of Tat, displayed basal activity in the tsBN462 at the permissive temperature. However, in marked contrast to the MHC class I promoter, the HIV LTR not only retained activity at the restrictive temperature of 39³C, but was signi¢cantly more active than at the permissive 32³C. Tat further enhances HIV LTR activity in these cells at both temperatures (data not shown). These data demonstrate that the HIV LTR does not depend on a functional TAFII 250, and provide a possible explanation for the failure of Tat to repress viral promoter activity. 3.4. Summary of Tat binding to the TAFII 250 fragment The above studies demonstrate that the interaction
of Tat with TAFII 250 is complex and involves three distinct binding sites. A pictorial representation of our ¢ndings is shown in Fig. 4. (It is important to note that we have no evidence that these interactions occur simultaneously.) The C-terminal segment of Tat (Tat67-101) binds to the AT domain of TAFII 250 between aa 885 and 1034, resulting in inhibition of the enzymatic activity which is correlated with Tat's repression of TAFII 250-dependent promoters. The N-terminal domain of Tat (Tat1-67) binds to two discontinuous sites on the TAFII 250 fragment: 885^984 aa within the AT domain and 1120^1279 aa within the Rap74-binding region. The cysteine-rich region of Tat is required for binding to the AT domain, whereas the basic domain appears to be necessary for binding to the Rap74-binding region. Although Tat binds to the Rap74-binding region, we did not observe any competition for binding to the TAFII 250 fragment between Tat101 and Rap74 (data not shown) nor did Rap74 phosphorylate Tat101 in the presence of the TAFII 250 fragment (data not shown). These data would suggest that Tat binding to the Rap74-binding domain is to a distinct site. In summary, the present study demonstrates that the interaction between Tat and TAFII 250 is complex, involving at least three distinct sites of interaction. The binding of the C-terminal segment of Tat, 67^101 aa, to the AT domain of TAFII 250 results in inhibition of AT activity and consequent repression of transcription initiation of TAFII 250-dependent promoters [27]. The N-terminal segment of Tat, 1^ 67 aa, binds to two discontinuous sites on TAFII 250 ^ the AT domain and the RAP74-binding region ^ which may stabilize the interaction and e¡ect greater repression. The failure of Tat to repress the HIV LTR promoter may be explained by the observation that transcription initiation of the HIV LTR does not require TAFII 250. Acknowledgements The authors gratefully acknowledge Kevin Howcroft, John Brady and Fatah Kashanchi for helpful discussions, and Klaus Strebel for critical review of the manuscript.
BBAPRO 36373 7-3-01
J.D. Weissman et al. / Biochimica et Biophysica Acta 1546 (2001) 156^163
References [1] R.S. Chambers, B.Q. Wang, Z.F. Burton, M.E. Dahmus, J. Biol. Chem. 270 (1995) 14962^14969. [2] R.D. Kornberg, Trends Cell Biol. 9 (1999) M46^M49. [3] S.R. Albright, R. Tjian, Gene 242 (2000) 1^13. [4] O. Bensaude, F. Bonnet, C. Casse, M.F. Dubois, V.T. Nguyen, B. Palancade, Biochem. Cell Biol. 77 (1999) 249^255. [5] D. Reines, R.C. Conaway, J.W. Conaway, Curr. Opin. Cell Biol. 11 (1999) 342^346. [6] H. Lu, L. Zawel, L. Fisher, J.M. Egly, D. Reinberg, Nature 358 (1992) 620^621. [7] R. Skiekhattar, F. Mermelstein, R.P. Fisher, R. Drapkin, B. Dynlacht, H.C. Wessling, D.O. Morgan, D. Reinberg, Nature 374 (1995) 283^287. [8] C.-h. Chang, C.F. Kostrub, Z.F. Burton, J. Biol. Chem. 268 (1993) 20482^20489. [9] T. O'Brien, R. Tjian, Mol. Cell 1 (1998) 905^911. [10] G. Orphanides, T. Lagrange, D. Reinberg, Genes Dev. 10 (1996) 2657^2683. [11] C. Mizzen, X. Yang, T. Kokubo, J. Brownell, A. Bannister, T. Owen-Hughes, J. Workman, L. Wang, S. Berger, T. Kouzarides, Y. Nakatani, C.D. Allis, Cell 87 (1996) 1261^1270. [12] B. Cullen, FASEB J. 5 (1991) 2361^2368. [13] W.A. Haseltine, FASEB J. 5 (1991) 2349^2360. [14] K. Sastry, H. Raghava, R. Pandita, K. Tatpal, B. Aggarwal, J. Biol. Chem. 265 (1990) 20091^20093. [15] G. Scala, M. Ruocco, C. Ambrosino, M. Mallardo, V. Giordano, F. Baldasarre, E. Dragonetti, I. Quinto, S. Venuta, J. Exp. Med. 179 (1994) 961^971. [16] W.P. Tansey, S. Ruppert, R. Tjian, W. Herr, Genes Dev. 8 (1994) 2756^2769. [17] M.O. Westendorp, V.A. Shatrov, K. Schulze-Ostho¡, F. Rainer, M. Kraft, M. Los, P.H. Krammer, W. Droge, V. Lehmann, EMBO J. 14 (1995) 546^554. [18] M. Garber, T. Mayall, E. Suess, J. Meisenhelder, N. Thompson, K. Jones, Mol. Cell. Biol. 20 (2000) 6958^6969. [19] C.A. Parada, R.G. Roeder, Nature 384 (1996) 375^378.
163
[20] L. Garcia-Martinez, G. Mavankal, J. Neveu, W. Lane, I. Ivanov, R. Gaynor, EMBO J. 16 (1997) 2836^2850. [21] R.S. Chambers, M.E. Dahmus, J. Biol. Chem. 269 (1994) 26243^26248. [22] N. Marshall, G.K. Dahmus, M.E. Dahmus, J. Biol. Chem. 273 (1998) 31726^31730. [23] R. Taube, K. Fujinaga, J. Wimmer, M. Barboric, B.M. Peterlin, Virology 264 (1999) 245^253. [24] T.K. Howcroft, K. Strebel, M. Martin, D.S. Singer, Science 260 (1993) 91^93. [25] T.K. Howcroft, L. Palmer, J. Brown, B. Rellahan, F. Kashanchi, J. Brady, D.S. Singer, Immunity 3 (1995) 127^138. [26] I. Carroll, J. Wang, T.K. Howcroft, D.S. Singer, Mol. Immun. 35 (1998) 1171^1178. [27] J.D. Weissman, J. Brown, T.K. Howcroft, J. Hwang, A. Chawla, P. Roche, L. Schiltz, Y. Nakatani, D.S. Singer, Proc. Natl. Acad. Sci. USA 95 (1998) 11601^11606. [28] E. Pocsik, M. Higuchi, B. Aggarwal, Lymph. Cyto. Res. 11 (1992) 317^325. [29] S.F. Purvis, D. Georges, T. Williams, M. Lederman, Cell. Immunol. 144 (1992) 32^42. [30] S. Flores, J. Marecki, K. Harper, S. Bose, S. Nelson, J. McCord, Proc. Natl. Acad. Sci. USA 90 (1993) 7632^7636. [31] J. Brown, T.K. Howcroft, D.S. Singer, J. AIDS 17 (1998) 9^ 16. [32] T. Hayashida, T. Sekiguchi, E. Noguchi, H. Sunamoto, T. Ohba, T. Nishimoto, Gene 141 (1994) 267^270. [33] T. Nakashima, T. Sekiguchi, H. Sunamoto, K. Yura, S. Tomoda, M. Go, J. Kere, D. Schlessinger, T. Nishimoto, Gene 141 (1994) 193^200. [34] J. Leonard, C. Parrot, A. Buckler-White, W. Turner, E. Ross, M. Martin, A. Rabson, J. Virol. 63 (1989) 4919^4924. [35] T.K. Howcroft, J. Richardson, D.S. Singer, EMBO J. 12 (1993) 3163^3169. [36] J. Archambault, G. Pan, G.K. Dahmus, M. Cartier, N.F. Marshall, S. Zhang, M.E. Dahmus, J. Greenblatt, J. Biol. Chem. 273 (1998) 27593^27601. [37] B.R. Cullen, Cell 46 (1986) 973^982.
BBAPRO 36373 7-3-01