Transactivation Mediated by B-Myb Is Dependent on TAFII250

Transactivation mediated by B-Myb is dependent on TAFII250

Thorsten Bartusel1 and Karl-Heinz Klempnauer*,1

1Institut fu¨r Biochemie, Westfa¨lische-Wilhelms-Universita¨t Mu¨nster, Wilhelm-Klemm-Str. 2, D-48149 Mu¨nster, Germany

B-Myb is a highly conserved member of the Myb family of IID (TFIID), which is composed of the TATA box- transcription factors, which has been implicated in cell binding protein (TBP) and a number of TATA box- cycle regulation. B-Myb is expressed in most proliferating binding protein-associated factors (TAFs).The largest cells and its activity is highly regulated around the G1/S- subunit of the TFIID complex, TAFII250, plays a key phase border of the cell cycle. It is generally assumed that role in this complex by acting as a scaffold for the B-Myb regulates the expression of genes that are crucial assembly of other TAFs and TBP into TFIID and by for cell proliferation; however, the identity of these genes, displaying various enzymatic activities that facilitate the molecular mechanisms by which B-Myb stimulates transcriptional initiation (Albright and Tjian, 2000; their expression and the involvement of other proteins Wassarman and Sauer, 2001).TAF II250 possesses have not been sufficiently clarified. We have employed the acetyltransferase activity, which is located in the central hamster cell line ts13 as a tool to demonstrate a functional domain of the protein and is able to acetylate histones as link between B-Myb and the coactivator TAFII250, a key well as other components of the basal transcriptional component of the transcriptional machinery which itself is machinery (Mizzen et al., 1996; Imhof et al., 1997). In essential for cell proliferation. ts13 cells express a point- addition, TAFII250 contains two independent serine/ mutated version of TAFII250 whose intrinsic histone threonine-specific protein kinase domains located at the acetyl transferase activity is temperature sensitive. N- and C-terminus (Dikstein et al., 1996; O’Brien and Transactivation of Myb-responsive reporter genes by B- Tjian, 1998).Finally, TAF II250 has ubiquitin-activat- Myb is temperature-dependent in ts13 cells but not in ts13 ing/conjugating activity and ubiquitinates histone H1 cells, which have been rescued by transfection with an (Pham and Sauer, 2000). expression vector for wild-type TAFII250. Furthermore, TAFII250 is an essential protein which is required for B-Myb and TAFII250 can be coprecipitated, suggesting progression through the cell cycle.The ts13 hamster cell that both proteins are present in a complex. The formation line harbors a single point mutation in the gene of this complex is dependent on the DNA-binding domain encoding TAFII250, which renders the acetyltransferase of B-Myb and not on its transactivation domain. Taken activity of TAFII250 temperature-sensitive (Hisatake together, these observations provide the first evidence that et al., 1993; Dunphy et al., 2000). When ts13 cells are the coactivator TAFII250 is involved in the activation of grown at the nonpermissive temperature, they arrest in Myb responsive promoters by B-Myb. The finding that B- the G1 phase of the cell cycle and undergo apoptosis.A Myb transactivation is dependent on a key coactivator large number of genes have been identified whose involved in cell cycle control is consistent with and expression is affected at the restrictive temperature strengthens the idea that B-Myb plays a crucial role as (O’Brien and Tjian, 2000), including the genes for a transcription factor in proliferating cells. cyclins A and D1 (Wang and Tjian, 1994; Suzuki- Oncogene (2003) 22, 2932–2941. doi:10.1038/sj.onc.1206494 Yagawa et al., 1997; Wang et al., 1997). A number of studies have shown that, depending on the gene, Keywords: B-Myb; TAFII250; ts13; transactivation TAFII250 acts on its targets by affecting basal promoter activity or by mediating the effects of upstream activators or by doing both.For example, the TAF II250 dependence of cyclin A expression is mediated by the Introduction basal promoter as well as by an upstream sequence element of the cyclin A gene (Wang et al., 1997). These The initiation of transcription by RNA polymerase II is two effects appear to be independent, as upstream a highly regulated process in which the cooperation of binding sites for activators can confer TAF 250 DNA-binding transcription factors with the basal II dependence on promoters that are normally TAF 250 transcriptional machinery and the modulation of II independent (Wang et al., 1997). A number of reports chromatin structure play crucial roles.A central have shown that TAF 250 interacts with and is component of this machinery is the transcription factor II regulated by several activators, such as the HIV Tat (Weissman et al., 1998), adenovirus E1A (Siegert et al., *Correspondence: K-H Klempnauer; E-mail: [email protected] 2000), c-Jun (Lively et al., 2001) and other transcrip- Received 23 December 2002; revised 13 February 2003; accepted 14 tional regulators, including the retinoblastoma (Siegert February 2003 and Robbins, 1999) and MDM2 proteins (Levaillard TAFII250-dependent transactivation T Bartusel and K-H Klempnauer 2933 and Wasylyk, 1997).The identification and analysis of B-Myb is a transcription factor that plays a crucial role genes and of transcriptional activators whose activity in proliferating cells, we were interested in the possibility depends on TAFII250 will undoubtedly provide further that there is a functional link between B-Myb and insight into the role of TAFII250 during the cell cycle. TAFII250.To address this possibility, we investigated The transcription factor B-Myb is a highly conserved whether activation at the Myb responsible reporter gene member of the myb proto-oncogene family, which pGL3-3xATk-Luc by full-length B-Myb or by the consists of A-Myb, B-Myb and c-Myb (Nomura et al., C-terminal deletion mutant B-Myb-D3 is temperature- 1988; Bouwmeester et al., 1992; Foos et al., 1992; Lam dependent in the ts13 hamster cell line, which carries a et al., 1992), B-Myb is expressed in a wide variety of temperature-sensitive allele of the TAFII250 gene dividing cells and a large body of evidence suggests that (Figure 1a).This reporter gene contains three copies of B-Myb plays a crucial role during cell proliferation.B- a Myb-binding site from the chicken mim-1 gene and is myb expression is downregulated in quiescent fibroblasts commonly used to study transactivation mediated by and subsequent re-entry into the cell cycle induces Myb proteins.This experiment demonstrated that expression of B-myb in the late G1/early S phase (Lam transactivation of the reporter gene by B-Myb and et al., 1992; Lam and Watson, 1993). The cell cycle B-Myb-D3 was indeed temperature-sensitive in ts13 regulated B-myb expression is dependent on an E2F- cells.This was particularly evident when the C- binding site within the B-myb promoter and involves terminally truncated form of B-Myb was used.As negative regulation by E2F complexes (Lam and shown before (Ziebold et al., 1997), this form of B-Myb Watson, 1993; Zwicker et al., 1996). During mouse is more active than full-length B-Myb because an embryogenesis, B-myb expression is tightly linked to the inhibitory domain, located at the C-terminus of B- proliferative activity of cells and tissues (Sitzmann et al., Myb, has been deleted in this mutant.The poor activity 1996).B- myb antisense oligonucleotides inhibit cell of the reporter gene at the nonpermissive temperature proliferation (Arsura et al., 1992), while overexpression was not because of insufficient amounts of B-Myb of B-myb enables cells to escape the cell cycle arrest that protein; as demonstrated by the Western blot shown in is induced by overexpression of the tumor suppressor Figure 1c, there was actually an increase in the amount protein p53 and mediated by the Waf1/Cip1 protein of B-Myb and B-Myb-D3 at the nonpermissive tem- (Lin et al., 1994). In addition to the control of B-myb perature, suggesting that the degree of inhibition of mRNA expression, the activity of the B-Myb protein is B-Myb activity at the restrictive temperature is even regulated by post-translational modification.Phosphor- higher than is evident from the data shown in Figure 1a. ylation of B-Myb by cyclin A/Cdk2 in the early S phase We concluded from this experiment that TAFII250 is of the cell cycle relieves repression mediated by an required for efficient transactivation by full-length or inhibitory domain located at the carboxyl-terminus truncated B-Myb.To substantiate this conclusion, and thereby stimulates the transactivation potential of we performed the same experiment using ts13 cells B-Myb (Robinson et al., 1996; Lane et al., 1997; Sala which had been stably transfected with an expression et al., 1997; Ziebold et al., 1997; Ziebold and Klemp- vector for wild-type (wt) TAFII250 (referred to as nauer, 1997; Saville and Watson, 1998; Bartsch et al., ts13HAX250) and which do not show a temperature- 1999).In addition to stimulating B-Myb activity, cyclin sensitive cell-cycle phenotype.As shown in Figure 1a, A/Cdk2-mediated phosphorylation triggers polyubiqiti- B-Myb-mediated transactivation was no longer tem- nation of B-Myb and directs its degradation by the 26S perature-dependent in these cells, confirming that proteasome (Charrasse et al., 2000). Recently, we the decrease in the transactivation potential of showed that cyclin D1, in contrast to cyclin A, inhibits B-Myb and B-Myb-D3 observed in ts13 cells upon the activity of B-Myb in a Cdk-independent manner by temperature shift is indeed because of the lack of physically interacting with B-Myb (Horstmann et al., TAFII250.To corroborate these findings, we also used a 2000a).Finally, the coactivator p300/CBP interacts with reporter gene (pGL3-A(-144)wt-Luc) containing the B-Myb and increases its transcriptional activity (Bessa Myb-responsive promoter of the tom-1 gene (Burk et al., 2001; Johnson et al., 2002). et al., 1997) for a transactivation experiment The long-standing observation that B-Myb, like (Figure 1b).The tom-1 promoter was activated less TAFII250, plays a crucial role in proliferating cells has efficiently by B-Myb-D3 than the tk promoter construct; prompted us to look for a functional link between both however, it is obvious that the activation of the tom-1 proteins.The work presented here directly supports the promoter was also temperature-dependent when ts13 idea of such a link by demonstrating that transactivation cells but not when tsHAX250 cells were used (Figure 1c). mediated by B-Myb is dependent on TAFII250 and that The finding that the activity of the truncated form of both proteins are present in a complex in vivo. B-Myb is impaired at the nonpermissive temperature in ts13 cells is particularly interesting because it suggests that the effect of TAF 250 on B-Myb is unlikely an Results II indirect effect mediated by cyclin A, which is a Transactivation mediated by B-Myb depends on transcriptional target of TAFII250 (Wang and Tjian, functional TAF 250 1994; Wang et al., 1997). The activity of B-Myb-D3is II independent of cyclin A (Ziebold et al., 1997), hence the TAFII250 has been shown to control the expression of decrease in activity of truncated B-Myb at nonpermis- genes that are required during cell proliferation.Since sive temperature cannot be explained by the lack of

Oncogene TAFII250-dependent transactivation T Bartusel and K-H Klempnauer 2934

Figure 1 B-Myb-mediated transactivation requires TAFII250.( a,b) ts13 cells or ts13HAX250 cells were kept at 331Corat391C and transfected with the Myb-responsive reporter genes pGL3-3xATk-Luc or pGL3-A(-144)wt-Luc, pCMVb and empty expression vectors or expression vectors for full-length (B-Myb) or carboxy-terminally truncated B-Myb (B-Myb-D3), as indicated.Luciferase and b- galactosidase activities were determined 24 h after transfection.Luciferase activities were normalized with respect to the cotransfected b-galactosidase plasmid to correct for different transfection efﬁciencies.Bars show fold activation of the luciferase reporter gene by B- Myb (or B-Myb-D3) relative to an identical transfection containing only empty expression vector and lacking B-Myb.Thin lines show s.d.’s. (c) Western blot analysis of cells transfected with the same amount of B-Myb (lanes 1) or B-Myb-D3 (lanes 2) expression vector as described in (a).Total cell extracts were fractionated by 10% SDS–PAGE and analysed by Western blotting using antibodies against B-Myb.Molecular weight markers are indicated on the left

cyclin A under these conditions.Taken together, these of the B-Myb constructs was able to rescue the mutant results provide the ﬁrst evidence that B-Myb requires cells at nonpermissive temperature (Figure 2), clearly TAFII250 for transactivation of B-Myb-responsive showing that B-Myb does not complement the promoters. TAFII250 deﬁciency in ts13 cells.

B-Myb does not abrogate the TAFII250 deficiency in ts13 Coimmunoprecipitation of TAFII250 and B-Myb cells The experiments illustrated in Figure 1 show that The experiments shown in Figure 1 suggest a functional transactivation by B-Myb depends on TAFII250, but link between B-Myb-dependent transactivation and they do not reveal the nature of this dependence.One TAFII250.To further explore this link, we were obvious possibility is that the activation of a promoter interested to determine whether overexpression of B- by B-Myb requires TAFII250 to be present at that Myb is able to overcome the TAFII250 deficiency in ts13 promoter.In this case, we might expect to find B-Myb cells.To address this issue, we transfected an expression and TAFII250 in a complex in vivo.To investigate vector for B-Myb or B-Myb-D3 into ts13 cells and whether this is the case, we performed coimmunopreci- cultivated them at the nonpermissive temperature.As pitation experiments using extracts from cells expressing controls, we transfected the cells with an expression B-Myb, TAFII250 or both.As illustrated in Figure 3, we vector for wt TAFII250 or with empty expression vector. transfected fibroblasts with different combinations of As expected, wild-type TAFII250 was able to comple- expression vectors for B-Myb and HA-tagged full- ment the ts13 mutant giving rise to colonies of length or partially deleted TAFII250.The cells were transfected cells that grew at 391C.By contrast, none then radiolabeled with [35S]methionine, followed by

Oncogene TAFII250-dependent transactivation T Bartusel and K-H Klempnauer 2935 coprecipitated, whereas a protein that corresponds to the amino-terminus of TAFII250 (DXba) was coprecipitated weakly (Figure 3).Taken together, it appears that two nonoverlapping regions of TAFII250 indepen- dently mediate coprecipitation with B-Myb.One of these regions is located in the central part of TAFII250 and mediates efficient coprecipitation with B-Myb, while the other is located in the amino-terminal domain of the protein and coprecipitates with B-Myb consider- ably weaker. TAFII250 has weak histone acetyltransferase (HAT) activity which has been mapped to the center of the protein.The HAT activity of TAF II250 is temperature sensitive in ts13 cells (Dunphy et al., 2000), suggesting that the acetylation of histones and possibly of other target proteins by TAFII250 plays a role in TAFII250- dependent gene expression (Dunphy et al., 2000). The results presented so far raised the question of whether B- Myb is acetylated by TAFII250.However, we could not detect any acetylation of B-Myb in the presence Figure 2 B-Myb does not rescue ts13 cells at restrictive TAFII250 using antibodies specific for acetylated lysine temperature.ts13 cells were transfected with expression vectors (data not shown).Thus, there is no evidence that B-Myb for B-Myb (a), B-Myb-A3 (b), TAFII250 (c) or empty expression is a substrate for the HAT activity of TAFII250. vector (d).Cells were then cultivated at 39 1C temperature until colonies were visible Coprecipitation of TAFII250 and B-Myb depends on the DNA-binding domain of B-Myb To map the domain of B-Myb, which is involved in immunoprecipitation with antibodies specific for B-Myb complex formation with TAFII250, we expressed full- or the HA-tag.Finally, the immunoprecipitates were length TAFII250 with different deletion mutants of fractionated by SDS–polyacrylamide gel electrophoresis B-Myb, followed by labeling of the cells with (PAGE) and analysed by autoradiography.As shown in [35S]methionine and immunoprecipitation of cell extracts Figure 3, the transfected full-length TAFII250 could with antiserum specific for B-Myb or the HA-tag be efficiently coprecipitated with B-Myb-specific (Figure 4).Panel a shows the total amount of TAF II250 antiserum, suggesting that TAFII250 is present in a detected with anti-HA antiserum.Panel b of Figure 4 complex with B-Myb.Control immunoprecipitations shows the amount of TAFII250, that was coprecipitated of cells transfected with B-Myb or TAFII250 ex- with B-Myb, while panel c shows the B-Myb proteins pression vectors alone showed that TAFII250 could that were precipitated with B-Myb antiserum.This only be precipitated with B-Myb-specific antiserum experiment demonstrates that coprecipitation of B-Myb when B-Myb was present.This confirms that the and TAFII250 depends on the presence of the DNA- observed coprecipitation was specific and was not binding domain of B-Myb.For example, an otherwise caused by crossreactivity of the antisera or insufficient full-length version of B-Myb which carries a deletion of washing of the immunoprecipitates.Surprisingly, the the DNA-binding domain (B-Myb-DDBD) failed to anti-HA antibody did not coprecipitate B-Myb, coprecipitate TAFII250, whereas a truncated form although under the same conditions antibodies against of B-Myb which contains little more than the B-Myb coprecipitated TAFII250.We do not know the DNA-binding domain (B-Myb-DPvu) coprecipitated reason why coprecipitation of both protein works only TAFII250 efficiently (see Figure 4).B-Myb-Ex9A-is a in one direction.Comparison of the relative intensities natural splice variant of B-Myb lacking the so-called of the protein bands corresponding to B-Myb and exon 9A (Kamano et al., 1995). This protein is inactive TAFII250 suggests that B-Myb is in excess over as a transcriptional activator because it lacks a func- TAFII250.It is therefore possible that the amount of tional transactivation domain (Horstmann et al., coprecipitated B-Myb is below the limit of detection. 2000b).B-Myb-Ex9A- was also able to coprecipitate Alternatively, it is also possible that the HA-epitope is TAFII250, demonstrating that the ability to form a not accessible to the antibody in the complex. complex with TAFII250 does not depend on the In the same experiment we also coexpressed B-Myb transactivation domain of B-Myb. with different deletion mutants of TAFII250 to identify To show that the B-Myb DNA-binding domain not the domain of TAFII250, which is responsible for only mediates the formation of a complex between B- coprecipitation with B-Myb.As seen in Figure 3 Myb and TAFII250 but is also crucial for the functional coprecipiation was most effective when the middle cooperation of both proteins, we fused the major region of TAFII250 was present (constructs D400 and transactivation domain of B-Myb (Horstmann et al., DNC).The C-terminal domain ( D1225) was not 2000b) either to the Gal4 DNA-binding domain or to

Oncogene TAFII250-dependent transactivation T Bartusel and K-H Klempnauer 2936

Figure 3 Coimmunoprecipitation of B-Myb and TAFII250.Full-length HA-tagged TAF II250 and deletion mutants of TAFII250 are shown schematically at the top.Functional domains of TAF II250 are highlighted by black and hatched boxes.QT6 cells were transfected with expression vectors for B-Myb and wild-type or mutant TAFII250, as indicated at the bottom.At 24 h after transfection, the cells were radiolabeled with [35S]methionine and cell extracts were immunoprecipitated using antibodies speciﬁc for B- Myb (left panel) or the HA-tag (right panel), followed by SDS–PAGE and autoradiography.The bands corresponding to B-Myb and TAFII250 are marked.TAF II250-DNC is expressed only in small amounts and is marked by an arrow in both the panels.Additional bands that are visible in all lanes of the gels are background bands caused by nonspeciﬁc precipitation.Molecular weight markers are indicated on the left

the B-Myb DNA-binding domain.These proteins assessed in ts13 cells at permissive and nonpermissive (designated as Gal4-D2 and B-Myb-D2) were then used temperatures.As expected, the activity of the construct to transactivate reporter genes consisting of the Herpes containing the B-Myb DNA-binding domain was simplex thymidine kinase promoter fused to Gal4- or signiﬁcantly impaired at nonpermissive temperature. Myb-binding sites (Figure 5).Transactivation was By contrast, the activity of the Gal4/B-Myb fusion

Oncogene TAFII250-dependent transactivation T Bartusel and K-H Klempnauer 2937

Figure 5 TAFII250 dependence of B-Myb transactivation is mediated by the DNA-binding domain of B-Myb.ts13 cells were transfected with the luciferase reporter genes illustrated schematically at the top, pCMV3 and expression vectors for fusion proteins consisting of the major transactivation domain of B-Myb (referred to as D2) fused to the B-Myb DNA-binding domain (B-Myb-D2) or the Gal4 DNA-binding domain (Gal4-D2).Cells were kept at 33 or 391C, as indicated.Luciferase and b-galactosidase activities were determined 24 h after transfection.Luciferase activities were normalized with respect to the cotransfected b-galactosidase plasmid to correct for different transfection efﬁciencies.Bars depict the activation of the luciferase reporter gene by B-Myb-D2 or Gal4-D2 relative to a identical transfections containing only empty expression vector and lacking B-Myb or Gal4 fusion proteins.The activation observed at permissive temperature (331C) was designated as 100%. Thin lines show standard s.d.’s

construct was virtually identical at both temperatures. This result clearly shows that the DNA-binding domain of B-Myb is required for cooperation with TAF 250. Figure 4 Coimmunoprecipitation of TAFII250 with deletion II mutants of B-Myb.Top: Schematic illustration of full-length and To substantiate our results and provide additional partially deleted B-Myb proteins.The DNA-binding domain and evidence that B-Myb and TAFII250 are present in a the exon 9A region of B-Myb are highlighted.Bottom panels: QT6 complex also in untransfected cells, we performed cells were transfected with expression vectors for HA-tagged full- coimmunoprecipitation experiments with Hela cells, length TAFII250 and full-length or partially deleted B-Myb, as indicated below the lanes.At 24 h after transfection, the cells were which express both B-Myb and TAFII250 endogenously. radiolabeled with [35S]methionine and cell extracts were immuno- Extracts from these cells were first immunoprecipitated precipitated using antibodies specific for the HA-tag or for B-Myb, with B-Myb-specific antibodies or with an anti-His-tag followed by SDS-PAGE and autoradiography.( a) Immunopreci- antibody as an irrelevant control.The immunoprecipi- pitation with anti-HA antibodies.Only the part of the protein gel tates were then fractionated by SDS–PAGE and containing TAFn250 is shown.( b,c) Immunoprecipitation with anti-B-Myb antibodies.For convenience (because of the large size analysed by Western blotting using TAFII250-specific differences between the TAFII250 and the different B-Myb antibodies.Figure 6 shows that a small fraction of proteins), the parts of the gel containing TAF 250 (b) and differ- II the endogeous TAFII250 was indeed coprecipitated ent B-Myb proteins (c) are shown in separate panels.The with the B-Myb antiserum but not with the control bands corresponding to TAFII250 and the different B-Myb proteins are marked.Molecular weight markers are indicated on antiserum.Thus, this experiment shows that B-Myb and the right a small fraction of the TAFII250 are present in a

Oncogene TAFII250-dependent transactivation T Bartusel and K-H Klempnauer 2938 resulting in a protein whose HAT activity is temperature-sensitive.By contrast, ts13 cells which have been rescued by overexpression of wt TAFII250 no longer show temperature-dependent activity of B-Myb.These results provide strong evidence that B-Myb-mediated transactivation is dependent on the HAT activity of TAFII250.Secondly, we have shown that B-Myb and TAFII250 can be coprecipitated from cells transfected with expression vectors for both proteins as well as from cells expressing these proteins endogenously.The amount of endogenous TAFII250 which could be coprecipitated with B-Myb was very low, which is not Figure 6 Coimmunoprecipitation of B-Myb and TAFII250 from Hela cell extracts.Extract prepared from approximately 2 Â 108 surprising, given the fact that TAFII250 is involved in Hela cells was immunoprecipitated using an irrelevant control the transcription of numerous genes (O’Brien and Tjian, antibody (rabbi anti His-tag; lanes 1 and 2) or rabbit antiserum 2000).TAF II250 therefore probably is involved in against B-Myb (lane 3 and 4).Lane 5 shows total cell extract from complex formation with other transcription factors Hela cells loaded directly onto the gel withour prior immunoprecipitation.Proteins were then analysed by SDS–PAGE and besides B-Myb.It is also possible that coprecipitation Western blotting using antibodies against TAFII250.The band of B-Myb and TAFII250 is less efficient in the corresponding to TAFII250 is labeled by an arrow.The position of endogenous situation than in the transfection experi- a molecular weight marker is indicated ment, because most of the endogenous TAFII250 will be embedded in large complexes with other proteins.By contrast, because of overexpression there is likely a large complex also in untransfected cells expressing both pool of free TAFII250 in the transfected cells.The proteins endogenously. presence of other proteins in the B-Myb/TAFII250 complex in the endogenous situation might hinder the binding of the antibodies used for coprecipitation.We Discussion do not know whether B-Myb and TAFII250 are direct interaction partners; although our in vivo studies provide B-Myb is a highly conserved vertebrate member of the evidence for interaction, binding could be mediated by Myb transcription factor family, is expressed in virtually another proteins. In vitro protein-binding experiments all proliferating cell types.B- myb mRNA expression using bacterially expressed B-Myb and in vitro trans- peaks around the G1-/S-phase transition of the cell lated TAFII250 failed to show specific binding of B-Myb cycle.In addition, the activity of the B-Myb protein is to TAFII250 (data not shown), suggesting that the regulated by cyclin A/Cdk2-dependent phosphorylation interaction might indeed be indirect, as in the case of as well as by direct interaction with cyclin D1 (Lam the transcription factor Sp1, whose interaction with et al., 1992; Robinson et al., 1996; Ziebold et al., 1997; TAFII250 is mediated by TAFII110, another component Saville and Watson, 1998; Horstmann et al., 2000a). of TFIID (Gill et al., 1994). However, the lack of Since B-Myb expression and function are strictly binding of B-Myb and TAFII250 in vitro might also be correlated with cell proliferation and highly regulated because of the limitations inherent in this type of assay. during the cell cycle, it is generally assumed that B-Myb For example, it is possible that the in vitro expressed plays an important role as a transcriptional regulator in proteins do not fold properly and therefore are not able proliferating cells.B-Myb has been shown to interact to bind to each other.Irrespectively of whether the with p300/CBP and poly(ADP-ribose) polymerase interaction of B-Myb and TAFII250 is direct or indirect, (PARP), both of which stimulate the transactivation our results provide the first evidence for a functional potential of B-Myb, as well as with the nuclear link between these two proteins.Since TAF II250 is corepressors N-CoR and SMRT, which downmodulate essential for cell proliferation, these results are consis- B-Myb activity (Cervellera and Sala, 2000; Bessa et al., tent with and strengthen the idea that B-Myb plays a 2001; Johnson et al., 2002; Li and McDonnell, 2002). key role as a transcription factor in proliferating cells. The physiological relevance of the interaction of B-Myb Furthermore, our observations suggest that at least with these proteins is not yet fully understood.It is also some of the genes whose expression is affected in ts13 not known whether these are the only coactivators or cells will be targets for B-Myb.The cyclin D1 gene is a corepressors recruited by B-Myb. candidate for this group of genes as it is downregulated Here, we have employed the hamster cell line ts13 as a in ts13 cells at non-permissive temperature (Suzuki- tool to demonstrate a functional link between B-Myb Yagawa et al., 1997; Wang et al., 1997) and has been and the coactivator TAFII250, a key component of the postulated to be activated by B-Myb (Sala and transcriptional machinery, which is itself essential for Calabretta, 1992). cell proliferation.The evidence for this link is based on We have shown that it is the DNA-binding domain of two major observations.Firstly, we have shown that B-Myb that is required for the interaction with transactivation mediated by B-Myb is temperature- TAFII250.The amino-terminal part of B-Myb which sensitive in the ts13 hamster cell line.ts13 cells carry a contains the DNA-binding domain is not only necessary point mutation in the endogenous TAFII250 gene but appears also to be sufficient for the formation of a

Oncogene TAFII250-dependent transactivation T Bartusel and K-H Klempnauer 2939 complex with TAFII250 in vivo, as evidenced by the was also found that the upstream element, which efficient coprecipitation of TAFII250 with a truncated B- contains ATF binding sites, could confer TAFII250- Myb protein (B-Myb-DPvu).The B-Myb DNA-binding dependence on a heterologous, TAFII250-independent domain has previously been shown to bind another promoter.Thus, the results presented here provide protein, PARP, via protein–protein interactions.Inter- another example in which upstream binding sites (Myb estingly, it has been shown that PARP stimulates B- sites in this case) in the presence of a cognate Myb-dependent transactivation of a Myb-responsive transcription factor (B-Myb) confer TAFII250 depen- reporter (Cervellera and Sala, 2000), suggesting that dence on a basal promoter.Further analysis of the PARP acts as a coactivator when bound to B-Myb. molecular details of the interplay of B-Myb and Thus, the DNA-binding domain of B-Myb is not only TAFII250 will probably shed light on the mechanisms required for protein–DNA interaction but also stimu- by which B-Myb regulates the expression of cellular lates transcription via specific protein–protein interac- genes and the role of the HAT activity of TAFII250. tions involving at leat two coactivators.That the Myb DNA-binding domain is involved in protein–protein interactions has also been shown in case of v-Myb and c- Materials and methods Myb (Mink et al., 1996; Tahirov et al, 2002), whose interaction with members of the C/EBP transcription Cell culture factor family is mediated by their DNA-binding The ts13 cell line and its derivative ts13HAX250 have been domain.Previous work has shown that B-Myb directly described (Talavera and Basilico, 1977; Wang et al., 1997). The interacts with the coactivator p300/CBP via the central cells were grown at 33.51C (permissive temperature) or at part of B-Myb, which contains the transactivation 39.51C (nonpermissive temperature) in DMEM supplemented domain (Horstmann et al., 2000b). B-Myb, therefore, with 10% fetal calf serum.The quail fibroblast cell line QT6 appears to communicate with the transcriptional ma- (Moscovici et al., 1977) was grown in Iscove’s modified chinery by at least two routes involving different DMEM supplemented with 8% fetal calf serum and 2% coactivators and two distinct domains of the protein chicken serum.Trichostatin A (TSA) was added to some (the DNA-binding and the transactivation domain).A cultures at a concentration of 100 ng/ml. complex strategy for gene activation like this might increase the regulatory flexibility of the system and Eukcaryotic expression vectors provide a means for increasing the specificity of B-Myb Expression vectors for full-length mouse B-Myb (pMuB- for particular target genes. MEx9A+) and for deletion mutants B-Myb D3 (pMuBMEx9- The observation that B-Myb-dependent transactiva- a+D3) and B-Myb DFsp (pMuBMDFsp) have been described tion is decreased at the restrictive temperature in ts13 before (Kamano and Klempnauer, 1997).An expression vector cells suggests that the HAT domain of TAFII250 and for the Gal4/B-Myb fusion protein Gal4/B-Myb-D2 was presumably its enzymatic activity is required for B-Myb described elsewhere (Horstmann et al., 2000b). pB-Myb-D2 activity.At present, it is not clear what the exact role of encodes a fusion protein of the B-Myb DNA-binding and the the HAT activity is, that is, which protein(s) must be major transactivation domain of B-Myb and was constructed as follows: B-Myb sequences encoding the DNA-binding acetylated by TAFII250 as a prerequisite for B-Myb- domain (amino acids 1–190 of mouse B-Myb) were PCR- dependent stimulation of the tk promoter.The HAT amplified using the primers MuBM143fd GCGGATC- activity of TAFII250 apparently is not required for the CATGTCTCGGCGGACGCGC and MuBM712bd GCGAA basal promoter function because the tk promoter by TTCTGGGAAACCTCCCGTGTCGAC.The sequences en- itself is not temperature-dependent in ts13 cells. coding domain D2 of B-Myb were generated similarly using Temperature-dependent activity is only seen in the the primers MuBM859fd GCATGAATTCCGAGGAGGA- presence of Myb-binding sites and B-Myb protein. GATTGCCATAG and MuBM1271bd GCATTCTAGAC- TAAAGGCGGTACTCAGTCACA.The DNA fragments Therefore, the HAT activity of TAFII250 is likely required for the communication of B-Myb with the were digested with HindIII and EcoRI or EcoRI and XbaI, transcriptional machinery that assembles at the basal tk respectively, and cloned between the HindIII and XbaI sites of pCDNA3 (Invitrogen).The resulting plasmid encodes a promoter.Since B-Myb itself apparently is not acety- protein consisting of amino acids 1–190 and 240–376 of mouse lated by TAF 250 at detectable levels, it might be II B-Myb.The TAF II250 expression vector pCS2+ speculated that TAFII250 acetylates (an) additional, as HA-WTTAF250 (encoding HA-tagged full-length human yet unknown protein(s) involved in B-Myb-dependent TAFII250) was obtained from EH Wang.Expression vectors transactivation.However, at present we cannot dismiss for HA-tagged deletion mutants of TAFII250 were generated the alternative possibility that B-Myb just interacts with as follows.First, a CMV-based expression vector for HA- the HAT domain of TAFII250 and that the interaction tagged full-length TAFII250 was constructed by cloning a itself is temperature-sensitive because of conformational double-stranded oligonucleotide encoding the HA-tag and the changes. complete coding region of TAFII250 (obtained as a NdeI/SpeI The scenario described here for B-Myb is somewhat fragment from pCS2+HA-WTTAF250) between the HindIII and XbaI sites of pCDNA3 (Invitrogen).The HA-tag reminiscent of the role of TAFII250 at the cyclin A oligonucleotide was obtained by annealing the single strands promoter.In that case, it was shown that one function AGCTTGGATCCACCATGGGCTACCCCTACGACGTGC of the HAT activity of TAFII250 was to mediate the CCCGACTACGCCATCGAAGGCCGCCA and TATGGC stimulatory effect of an upstream element on the activity GGCCTTCGATGGCGTAGTCGGGCACGTCGTAGGGG of the basal cyclin A promoter (Wang et al., 1997). It TAGCCCATGGTGGATCCA.The resulting plasmid, pHAh-

Oncogene TAFII250-dependent transactivation T Bartusel and K-H Klempnauer 2940 TAF250, was used to construct expression vectors encoding 24 h, the cells were radiolabeled for 6 h with 100 mCi of various deletions of TAFII250.pHAhTAF250- DBgl, encoding [35S]methionine ml (specific activity, >800 Ci/mmol), washed a HA-tagged protein which consists of amino acids 1–1419 of with PBS, and lysed in ice-cold ELB lysis buffer (50 mm Tris- TAFII250, was obtained by cloning a 4.3 kb BamHI/BgIII HCl, PH 7.5; 120 mm NaCl; 1 mm EDTA; 6 mm EGTA; 20 mm fragment from pHAhTAF250 into the BamHI site of NaF; 15 mm sodium pyrophoshate; 1 mm PMSF; 0.1 % (v/v) pCDNA3.pHAhTAF250- DXba (HA-tagged amino acids 1– NP-40).All subsequent steps were carried out on ice.After 401 of TAFII250) was generated by the HA-tag oligonucleo- shearing the DNA with a syringe, insoluble material was tide and a NdeI/XbaI fragment from pCS2+HA-WTTAF250 removed by centrifugation for 30 min at 41C and 14 000 r.p.m. (encoding the first 401 amino acids of TAFII250) between the The supernatant was divided into two aliquots, and antibodies HindIII and XbaI sites of pCDNA3.To generate N-terminal against mouse B-Myb (N19, St Cruz) or the HA-tag (HA.11, deletion mutants, the full-length TAFII250 coding region BabCO) were added.After incubation overnight at 4 1C, (including the HA-tag) was subcloned as a HindIII/SpeI protein A sepharose was added, incubated for 1 h and washed fragment between the HindIII and XbaI sites of pbluescript. three times with lysis buffer and once with PBS.Finally, the The resulting plasmid (pblueHAhTAF250) was then digested immunoprecipitated proteins were analysed by SDS–PAGE with NdeI and XbaI, followed by fill-in of the restriction sites and autoradiography. and religation to generate pblueHAhTAF250-D400 (lacking To detect complexes of endogenous B-Myb and TAFII250, the first 400 amino acids of TAFII250) or with NdeI and Hela cells were lysed in ELB lysis buffer.Aliquots of the cell EcoRI, followed by fill-in of the restriction sites and religation extract were incubated at 41C with antibodies against the N- to obtain pblueHAhTAF250-D1225 (lacking the first 1225 terminus of B-Myb (N19; St Cruz) or against the His-tag (H- amino acids of TAFII250).The insert fragments of these 15; St Cruz).Immunoprecipitates were collected by incubation plasmids were then subcloned as BamHI/NotI fragments with protein A sepharose, followed by three washes with ELB between the BamHI and NotI sites of pCDNA3 to obtain buffer.The immunoprecipitates were then analysed by SDS– pHAhTAF250-D400 and pHAhTAF250-D1225.In addition, a PAGE and Western blotting using antibodies against B-Myb BamHI/BglII fragment of pblueHAhTAF250-D400 was cloned (N19; St Cruz) or TAFII250 (6B3; St Cruz). into the BamHI site of pCDNA3 to obtain pHAhTAF250- DNC.These plasmids encode deleted forms of TAFII250 Western blotting containing amino acids 401–1893 (pblueHAhTAF250-D400), amino acids 1226–1893 (pHAhTAF250-D1225) and amino Immunostaining of mouse B-Myb proteins transferred to acids 401–1420 (pHAhTAF250-DNC).All deletion mutants of nitrocellulose was performed as described (Kamano et al., TAFII250 carry an N-terminal HA-tag. 1995).B-Myb proteins were detected with a rabbit antiserum raised against the amino-terminus of B-Myb (N19, St Cruz). Reporter genes, transfections, luciferase and b-galactosidase TAFII250 was detected with the HA-specific monoclonal assays antibody HA.11 (BabCO) or with the TAFII250-specific antibody 6B3 (St Cruz).Acetylated B-Myb was detected with The Myb-inducible reporter plasmids pGL3-3xATk-Luc and antibodies against acetyl-lysine (Ac-K-103, Cell Signaling pG5-tk-Luc have been described (Horstmann et al., 2000b; Technology). Yang et al., 2001). pCMVb was obtained from Clontech.Cells were transfected as described (Burk et al., 1993) and harvested 24 h after transfection.Preparation of cell extracts, luciferase and b-galactosidase assays were performed as described (Burk Acknowledgements et al., 1993). We thank S Horstmann for valuable discussions, EH Wang for providing cell lines and plasmids and for critical comments on this manuscript and A Brehmer-Fastnacht for excellent In vivo coimmunoprecipitation technical assistance.This study was supported by a grant To detect complexes of B-Myb and TAFII250 in vivo, QT6 cells from the DFG (KL 461/9-2) and by the Fonds der chemischen were transfected with the appropriate expression vectors.After Industrie.

References

Albright SR and Tjian R.(2000). Gene, 242, 1–13. Charrasse S, Carena I, Brondani V, Klempnauer KH and Arsura M, Introna M, Passerini F, Mantovani A and Golay J. Ferrari S.(2000). Oncogene, 19, 2986–2995. (1992). Blood, 79, 2708–2716. Dikstein R, Ruppert S and abd Tjian R.(1996). Cell, 84, 781– Bartsch O, Horstmann S, Toprak K, Klempnauer K-H and 790. Ferrari S.(1999). Eur. J. Biochem., 260, 384–391. Dunphy EL, Johnson T, Auerbach SS and Wang EH.(2000). Bessa M, Saville MK and Watson RJ.(2001). Oncogene, 20, Mol. Cell. Biol., 20, 1134–1139. 3376–3386. Eckner R, Ewen ME, Newsome D, Gerdes M, DeCaprio JA, Bouwmeester T, Guehmann S, el-Baradi T, Kalkbrenner F, Lawrence JB and Livingston DM.(1994). Genes Dev., 8, van Wijk I, Moelling K and Pieler T.(1992). Mech. Dev., 37, 869–884. 57–68. Foos G, Grimm S and Klempnauer KH.(1992). EMBO J., 11, Burk O, Mink S, Ringwald M and Klempnauer KH.(1993). 4619–4629. EMBO J., 12, 2027–2038. Gill G, Pascal E, Tseng H and Tjian R.(1994). Proc. Natl. Burk O, Worpenberg S, Haenig B and Klempnauer K-H. Acad. Sci. USA, 91, 192–196. (1997). EMBO J., 16, 1371–1380. Hisatake K, Hasegawa S, Takada R, Nakatani Y, Horikoshi Cervellera MN and Sala A.(2000).Poly(ADP-ribose) poly- M and Roeder RG.(1993). Nature, 362, 179–181. merase is a B-Myb coactivator. J. Biol. Chem., 275, 10692– Horstmann S, Ferrari S and Klempnauer KH.(2000a). 10696. Oncogene, 19, 5428–5434.

Oncogene TAFII250-dependent transactivation T Bartusel and K-H Klempnauer 2941 Horstmann S, Ferrari S and Klempnauer KH.(2000b). Robinson C, Light Y, Groves R, Mann D, Marais R and Oncogene, 19, 298–306. Watson R.(1996). Oncogene, 12, 1855–1864. Imhof A, Yang XJ, Ogryzko VV, Nakatani Y, Wolffe AP and Sala A, Calabretta B (1992). Proc. Natl. Acad. Sci. USA, 89, Ge H.(1997). Curr. Biol., 7, 689–692. 10415–10419. Johnson LR, Johnson TK, Desler M, Luster TA, Nowling T, Sala A, Kundu M, Casella I, Engelhard A, Calabretta B, Lewis RE, Rizzino A.(2002). J. Biol. Chem., 277, Grasso L, Paggi MG, Giordano A, Watson RJ, Khalili K 4088–4097. and Peschle C.(1997). Proc. Natl. Acad. Sci. USA, 94, 532– Kamano H, Burk B, Noben-Trauth K and Klempnauer KH. 536. (1995). Oncogene, 11, 2575–2582. Saville MK and Watson RJ.(1998). Oncogene, 17, 2679–2689. Kamano H and Klempnauer KH.(1997). Oncogene, 14, 1223– Siegert JL and Robbins PD.(1999). Mol. Cell. Biol., 19, 846– 1229. 854. Lam EW, Robinson C and Watson RJ.(1992). Oncogene, 7, Siegert JL, Rushton JJ, Seller WR, Kaelin WG and Robbins P. 1885–1890. (2000). Oncogene, 19, 5703–5711. Lam EW and Watson RJ.(1993). EMBO J., 12, 2705–2713. Sitzmann J, Noben-Trauth K, Kamano H and Klempnauer Lane S, Farlie P and Watson R.(1997). Oncogene, 14, 2445– KH.(1996). Oncogene, 12, 1889–1894. 2453. Suzuki-Yagawa Y, Guermah M and Roeder RG.(1997). Mol. Levaillard T and Wasylyk B.(1997). J. Biol. Chem., 272, Cell. Biol., 17, 3284–3294. 30651–30661. Tahirov TH, Sato K, Ichikawa-Iwata E, Sasaki M, Inoue- Li X and McDonnell DP.(2002) Mol. Cell. Biol., 22, 3663–3673. Bungo T, Shiina M, Kimura, K, Takata S, Fujikawa A, Lin D, Fiscella M, O’Connor PM, Jackman J, Chen M, Luo Morii H, Kumasaka T, Yamamoto M, Ishii S and Ogata K. LL, Sala A, Travali S, Appella E and Mercer WE.(1994). (2002). Cell, 108, 57–70. Proc. Natl. Acad. Sci. USA, 91, 10079–10083. Talavera A and Basilico C.(1977). J. Cell. Physiol., 92, Lively TN, Ferguson HA, Galasinski SK, Seto AG and 425–436. Goodrich JA.(2001). J. Biol. Chem., 276, 25582–25588. Wang EH and Tjian R.(1994). Science, 263, 811–814. Mink S, Kerber U and Klempnauer K-H.(1996). Mol. Cell. Wang EH, Zou S and Tjian R.(1997). Genes Dev., 11, Biol. 16, 1316–1325. 2658–2669. Mizzen CA, Yang X-J, Kokubo T, Brownell JF, Bannister AJ, Wassarman DA and Sauer F.(2001). J. Cell Sci., 114, Owen-Hughes T, Workman J, Wang L, Berger SL, 2895–2902. Kouzarides T, Nakatani Y and Allis CD.(1996). Cell, 87, Weissman JD, Brown JA, Howcroft TK, Hwang J, Chawla A, 1261–1270. Roche PA, Schiltz L, Nakatani Y and Singer DS.(1998). Moscovici C, Moscovici MG, Jiminez H, Lai MMC, Hayman Proc. Natl. Acad. Sci. USA, 95, 11601–11606. MJ and Vogt PK.(1977). Cell, 11, 95–103. Yang SH, Vickers E, Brehm A, Kouzarides T and Sharrocks Nomura N, Takahashi M, Matsui M, Ishii S, Date T, AD.(2001). Mol. Cell. Biol., 21, 2802–2814. Sasamoto S and Ishizaki R.(1988). Nucleic Acids Res., 16, Ziebold U, Bartsch O, Marais R, Ferrari S and Klempnauer 11075–11089. KH.(1997). Curr. Biol., 7, 253–260. O’Brien T and Tjian R.(1998). Mol. Cell., 1, 905–911. Ziebold U and Klempnauer KH.(1997). Oncogene, 15, O’Brien T and Tjian R.(2000). Proc. Natl. Acad. Sci. USA., 1011–1019. 97, 2456–2461. Zwicker J, Liu N, Engeland K, Lucibello FC and Muller R. Pham A-D and Sauer F.(2000). Science, 289, 2357–2360. (1996). Science, 271, 1595–1597.

Oncogene