- Email: [email protected]
The Molecular and Functional Characterization of E2F-5 Transcription Factor
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
242, 586–592 (1998)
RC978010
The Molecular and Functional Characterization of E2F-5 Transcription Factor Yashwantrai N. Vaishnav,1 Mahima Y. Vaishnav, and Vinod Pant Virology Group, International Centre for Genetic Engineering and Biotechnology, Aruna Asaf Ali Marg, P. O. Box 10504, New Delhi - 110067, India
Received November 17, 1997
The E2F activity plays a critical role in the control of cell cycle and action of tumor suppressor proteins and is also a target of the transforming proteins of small DNA tumor viruses. We describe here molecular cloning and functional characterization of a fifth member of the E2F family of transcription factors. E2F-5 protein is more homologous to E2F-4 (72% amino acid identity) than to E2F-1, E2F-2, and E2F-3 (35% amino acid identity). Based on structural and functional criteria, the E2F family appears to comprise two distinct sub-families, one composed of E2F-1, E2F-2, and E2F3 and the other composed of E2F-4 and E2F-5, E2F-5 mRNA is expressed in a wide variety of human tissues. The protein is expressed as multiple species ranging in size from 46 to 54 kDa as a result of differential phosphorylation. The expression of a reporter gene containing E2F binding sites in the promoter is transcriptionally activated by E2F-5 in a cooperative manner with the DP-1 protein. The interaction between E2F-5 and DP-1 is demonstrated using a two-hybrid system in mammalian cells. We have also demonstrated the presence of a strong transactivation domain at the carboxy terminus (273-346 amino acid residues) of E2F-5 protein. q 1998 Academic Press
The E2F activity was initially identified as a cellular DNA-binding activity required for the transactivation of viral promoter E2 by adenovirus E1A protein (1). The generic term ‘‘E2F activity’’ is now used to describe a heterodimer containing one subunit derived from the E2F family and another derived from the DP family (reviewed in 2). The E2F/DP heterodimer binds to DNA in a sequence-specific manner with high affinity and results in transcriptional activation of genes (3-6). Several members of the E2F and DP families have been molecularly cloned and characterized (4, 7-18). The con1 Corresponding author. Fax: 91-11-6162316. E-mail: yash@ icgebnd.ernet.in.
0006-291X/98 $25.00
sensus E2F-binding sites are found in the promoters of a variety of genes some of which are specifically expressed in S phase of the cell cycle and are required for DNA synthesis e.g. DNA polymerase a, dihydrofolate reductase and thymidine kinase (19-21). The promoters of certain genes which code for transcriptional regulators of cell growth also harbor E2F sites e.g. N-myc, cmyc, RB, cdc2, E2F-1 and B-myb (22-30). Many of these sites have been shown to be required for transcriptional activation in a cell cycle dependent manner. Overexpression of E2F-1 causes quiescent cells to progress into S phase of the cell cycle (31-33). Several studies suggest that deregulated E2F expression or mutation of E2F so as to release it from negative regulation by pocket proteins is potentially oncogenic (12-13, 34-36). Thus they appear to be critically involved in the control of cell cycle and consequently cell growth. The E2F/DP heterodimer is usually present in a complex with the RB family of tumor suppressor proteins such as pRB (7-8, 10-11), p107 (12-14) or p130 (14-15). The resultant complex is incompetent for transactivation (37-43). The site of interaction with the tumor suppressor proteins is located in the C-terminal end within the transactivation domain of the E2F proteins. The cell cycle dependent phosphorylation-dephosphorylation of tumor suppressor proteins modulate their interaction with E2F (7-9). Transforming proteins of DNA tumor viruses such as adenovirus E1A, SV40/polyoma virus large T antigen and human papilloma virus E7 dissociate E2F from the complex with tumor suppressor proteins (44-45). The ‘‘free’’ E2F activity, thus released, transcriptionally activates various E2F-responsive genes resulting in a loss of cell growth control. This provides a model to explain the transforming activity of these viral proteins. The E2F proteins contain several domains that are evolutionarily conserved among all members of the family (2). The DNA binding domain is composed of basic residues as well as a helix-loop-helix motif (4647). The domain that determines interaction with DP proteins is located downstream of the DNA binding
586
Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 8010
/
6946$$$381
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
domain (DBD) (4-5, 47). The transactivation domain enriched in acidic amino acids is located at the C-terminus (8, 46). Embedded within this domain is a stretch of conserved amino acid residues involved in mediating the interaction of these proteins with tumor suppressor proteins. These features are common to all members of the E2F family of transcription factors. Additionally, various members have certain exclusive features. For example, E2F-1, E2F-2 and E2F-3 have a cyclin binding domain located immediately upstream of the DNA binding domain (48). E2F-4 has a serine rich region that is encoded by varying copies of trinucleotide repeats (12-13). In this communication, we describe the molecular cloning and functional characterization of a fifth member of the E2F family of transcription factors designated as E2F-5. MATERIALS AND METHODS Molecular cloning of E2F-5. The E2F-5 cDNA clone was obtained fortuitously while screening a cDNA expression library in search of a clone for NFRRE , a cellular factor that specifically interacts with the Rev response element (RRE) of the human immunodeficiency virus type 1 (HIV-1) (49). The method used for screening was essentially as described for DNA-binding proteins (50). A lgt11 cDNA expression library prepared from activated HUT78 cells (Clontech) was screened with in vitro transcribed RRE as a probe. The cDNA insert obtained from recombinant phage DNA by digesting with EcoRI was subcloned into pBluescript SK(/) vector (Stratagene). JM109 cells carrying the recombinant plasmid were infected with the helper phage R408 (Clontech) and single-stranded phagemid DNA was prepared from the supernatant. Both strands of the insert were sequenced by the dideoxy chain termination method (51) initially using universal primers (forward and reverse) and subsequently gene specific primers as the sequence information became available. Plasmids. For expression in mammalian cells, E2F-5 open reading frame was subcloned into an eukaryotic expression vector pSGI (52). Expression of the insert from pSGI is controlled by SV40 early promoter/enhancer region. The vector pSG424 (53) expresses the DNA-binding domain (1-147 amino acid residues) of the yeast Gal4 transactivator protein. Gal4-E2F-5 (273-346) and Gal4-E2F-5 (70346) expression constructs were made by cloning E2F-5 cDNA (corresponding to 273-346 or 70-346 amino acid residues respectively) into pSG424 vector at a SmaI site in-frame with Gal4 DBD. E2F4CAT (4), pCMV-DP-1 (4), G5EC (54), Gal4-VP16 (54), have been described. pBluescript SK (/) (Stratagene) and RcCMV (Invitrogen) were obtained commercially. Two hybrid system approach, as adapted for mammalian cells (55), was used for probing protein-protein interactions between DP-1 and E2F-5 or E2F-1. Gal4-DP-1 expression construct was made by cloning an EcoRI DP-1 insert derived from pHB44-DP-124-410 (56) at EcoRI site in pSG424 vector. The vector AASV-VP16 (54) expresses the transactivation domain (413-490 amino acid residues) of the herpesvirus VP16 transactivator protein. VP16-E2F-5 (70-346) chimera was made by cloning an EcoRI E2F-5 fragment derived from Gal4E2F-5 (70-346) at an EcoRI site in AASV-VP16. VP16-E2F-1 chimera was made by cloning an EcoRI E2F-1 insert derived from pGAD10E2F-1 wt (56) in AASV-VP16. Northern blot analysis. The multiple tissue Northern blot (Clontech) containing 2 mg of poly A/ RNA derived from various human tissues was probed with E2F-5 cDNA. The 1.7 kb E2F-5 cDNA insert was 32P-radiolabeled to a specific activity of 21108 cpm/mg by random priming (Boehringer Mannheim) and was used for hybridization.
The blot was subsequently reprobed with a synthetic oligonucleotide (Clontech) corresponding to b-actin gene (5*-GACGACGAGCGCGGCGATATCATCATC-3*). Antibodies. A synthetic peptide (NLPEQHVSER SQALQQTSAT DISSAGSISG D) corresponding to the amino acid residues 271-301 of E2F-5 was coupled to BSA with glutaraldehyde (57) and used for raising polyclonal antibodies in rabbits. The sequence of this peptide was derived from the region of E2F-5 which is highly divergent from E2F-4 as determined by sequence alignment (see Fig. 1). The monoclonal antibody against p53 (DO-1) that recognises the N-terminal epitope (58) was obtained from Santa Cruz Biotechnology. Transfection, radio-immunoprecipitation assay (RIPA), and chloramphenicol acetyltransferase (CAT) assay. COS-1 and U2-OS cells in 60-mm culture dishes were transfected using lipofectin reagent (GIBCO-BRL) according to the manufacturer’s guidelines. The total amount of DNA was kept constant by including pBluescript SK(/). pCMV-b-gal (0.2 mg) was used as an internal control for monitoring the efficiency of transfection. For RIPA, transfected cells were metabolically labeled 48 h post-transfection with 35S-methionine (100-200 mCi/60 mm plate) in methionine-free DMEM for about 4 h and extracts were prepared in RIPA buffer (10 mM Tris-HCl [pH 8.0], 140 mM NaCl, 5 mM iodoacetamide, 0.5% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS], 2 mM phenylmethylsulfonyl fluoride). Immunoprecipitation was carried out as described (59). For CAT assay, the cell extracts were prepared 48 h post-transfection and CAT activity was monitored by standard method (49). Each experiment was repeated at least 3 times and representative data are presented. Nucleotide sequence accession number. The nucleotide sequence of E2F-5 cDNA reported here has been deposited in EMBL databank under accession number Z78409.
RESULTS E2F-5 is highly homologous to E2F-4. While screening a human cDNA expression library in lgt11, we isolated a gene that codes for a fifth member of the E2F family of transcription factors designated as E2F-5. The multiple sequence alignment of E2F proteins by MACAW (60) revealed that E2F-5 is more closely related to E2F-4 (72 % amino acid identity) than to E2F1, E2F-2 or E2F-3 (32 % amino acid identity). The E2F5 cDNA clone is 1698 bp long. It contains a complete ORF of 1038 bp that can potentially code for a protein of 346 amino acids with a predicted molecular mass of Ç38 kDa. The cDNA has 34 bp of 5*-untranslated sequences and 626 bp of 3*-untranslated sequences. The polyA signal (AAUAAA) is missing in E2F-5 cDNA indicating that it does not contain complete 3*-untranslated sequences. The alignment of E2F-4 and E2F-5 amino acid sequences indicated extensive homology between the two proteins (Fig. 1). A large block in E2F5 from amino acid residues 39-268 is highly conserved between E2F-4 and E2F-5. This is followed by a highly divergent sequences. Finally, a short sequence at the extreme C-terminus (amino acid residues 327-345) is well conserved between the two proteins and presumably forms a pocket protein-binding domain. Multiple alignment of various E2F proteins allowed us to predict the position of various functional domains in E2F-5 on the basis of homology with the domains
587
AID
BBRC 8010
/
6946$$$382
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
that is conserved among E2F-1, E2F-2 and E2F-3 proteins, and which has been shown in E2F-1 to bind to cyclin A (48).
FIG. 1. The optimum alignment of E2F-5 and E2F-4 amino acid sequences derived using MACAW. The amino acid residues belonging to homologous blocks are presented in upper case. The identical amino acid residues are boxed.
that are well characterized in other E2F proteins, especially E2F-1. The region between residues 49 to 116 may constitute a DNA binding domain since it is relatively enriched in basic residues and is also homologous to the DNA binding domain as defined for E2F-1 (4, 7, 11). The DNA binding domain is the most highly conserved region between all members of the E2F family. A region from residues 133 to 201 may qualify as the heterodimerization domain based on its sequence homology with the corresponding domain in E2F-1 (4). It includes a leucine heptad repeat between amino acid residues 133-154 which is predicted to form a stable amphipathic a-helix. The leucine residues in this heptad repeat are highly conserved among all five E2F proteins even though amino acid residues present between the conserved leucines are quite divergent suggesting that hydrophobic residues occurring at every 7th position may form an interface required to interact with DP proteins. The region between residues 202 to 318 is quite divergent among members of the E2F family, except for a weak but definite homology in the sequence (214 to 232 residues) earlier designated as ‘marked box’ (11). However, E2F-4 and E2F-5 amino acid sequences exhibit considerable homology not only in the marked box but also adjacent sequences from 202 to 270 residues. The E2F-5 sequence from 271 to 318 residues is completely divergent from E2F-4 as well as other E2F sequences. It includes a serine rich region in E2F-4 that is encoded by varying copies of the trinucleotide repeats (12-13). The C-terminal end from residues 319 to 346 is well conserved between E2F-4 and E2F-5, as well as between all other members of the E2F family. It presumably includes a pocket protein binding region that is involved in interaction with pRB family of tumor suppressor proteins. Both E2F-4 and E2F-5 lack at their N-termini, a short sequence
E2F-5 mRNA is ubiquitously expressed in human tissues. In order to examine the expression of E2F-5 mRNA in various human tissues, a Northern blot containing poly(A)/ RNA derived from various tissues was probed with E2F-5 cDNA. E2F-5 mRNA of Ç2.2 kb was found to be expressed in a wide variety of human tissues (Fig. 2). Since the size of E2F-5 cDNA obtained is Ç1.7 kb, this indicated that the difference could be due to the absence of complete 5*- and 3*- untranslated sequences. Interestingly, skeletal muscles appear to express very low level of the E2F-5 mRNA. The lower signal observed in case of liver, however, was due to a low input of poly(A)/ RNA in that lane as determined by re-probing the same blot for b-actin expression (lower panel, Fig. 2). E2F-5 is differentially phosphorylated. In order to characterize the protein product, the E2F-5 cDNA was subcloned into pSGI eukaryotic expression vector under the control of SV40 early promoter/enhancer. COS1 cells were transfected with pSGI vector or pSGI-E2F5 expression construct, metabolically labeled with 35Smethionine and cell extracts were immunoprecipitated with E2F-5 specific rabbit polyclonal antiserum. As a control, pre-immune rabbit serum was used for precipitation. The E2F-5 specific antibodies immunoprecipitate proteins ranging in size from 46 kDa (the fastest
FIG. 2. The expression of E2F-5 mRNA in various human tissues. Northern blot containing poly(A)/ RNA from various tissues, as indicated, was hybridized to a human E2F-5 cDNA probe. The lower panel shows signals obtained by re-probing the same blot with a b-actin oligonucleotide. Note that the 2.2 kb species of b-actin mRNA is present in all the tissues while a 1.8 kb species of mRNA is present only in cardiac and skeletal muscle tissues.
588
AID
BBRC 8010
/
6946$$$382
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Differential phosphorylation of E2F-5 protein. Metabolically labeled extracts prepared from COS-1 cells transfected with 2.5 mg of either vector alone (pSGI) or E2F-5 expression construct (pSGIE2F-5) were immunoprecipitated with preimmune (PI) serum or polyclonal rabbit immune (I) serum and separated on SDS-7.5 % polyacrylamide gel. In some cases, the immunoprecipitates were treated with l phosphatase in the presence or absence of orthovanadate.
migrating), 49 kDa (intermediate band) to 54 kDa (the slowest migrating) from the cells transfected with pSGI-E2F-5 (Fig. 3). The pre-immune serum did not precipitate these proteins. Post-translation modification could possibly account for such heterogeneity. We examined the phosphorylation status of E2F-5 since it has several potential phosphorylation sites. The immunoprecipitated proteins were treated with l phosphatase (NEB) either in the presence or absence of orthovanadate, a specific inhibitor of phosphatases. As a control, immunoprecipitated proteins were treated with phosphatase buffer alone. The phosphatase treatment abolished 54 kDa and 49 kDa bands and most of the protein could be seen as a single species migrating as a 46 kDa band, thus indicating that the 54 kDa and 49 kDa bands represented differentially phosphorylated species of E2F-5 protein. The presence of orthovanadate during phosphatase treatment effectively prevented this change. The discrepancy between the predicted molecular weight of 38 kDa and the fastest migrating species of 46 kDa could be due to anomalous migration of E2F-5 in SDS-PAGE. The presence of two proline-rich regions (6-38 and 238-262 amino acid residues) in E2F-5 could possibly account for the anomalous migration. E2F-5 activates transcription in a cooperative manner with DP-1. Since all the E2F proteins share a common property of transactivating gene expression in an E2F site-dependent manner, we examined this property of E2F-5. When a reporter CAT construct E2F4CAT was transfected in U2-OS cells along with E2F-5 or DP-1 ex-
pression constructs separately at various concentrations, there was no transactivation (data not shown). However, when E2F4CAT was co-transfected with increasing amount of E2F-5 in presence of constant amount of DP-1, a dose dependent increase in the extent of transactivation could be seen (Fig. 4). Similarly, increasing amount of DP-1 in presence of a constant amount of E2F-5 led to transcriptional activation. However, the saturation was reached earlier as compared to E2F-5 and higher amount of DP-1 led to inhibition. The fact that E2F-5 and DP-1 could transcriptionally activate reporter gene expression only when expressed together indicated a co-operative action. The requirement of co-expression of E2F-5 and DP-1 presumably reflects that heterodimerization between the two proteins is essential for efficient transactivation. The protein-protein interaction between E2F-5 and DP-1 was examined using a mammalian version of the two hybrid system (55). For this purpose, we constructed Gal4-DP-1, VP16-E2F-5 and VP16-E2F-1 chimerae as shown in Fig. 5. None of these chimerae transactivated the reporter construct when transfected alone as expected. However, when Gal4-DP-1 was cotransfected with VP16-E2F-5, there was transactivation indicating the interaction between E2F-5 and DP-1. Similarly, co-transfection with Gal4-DP-1 and VP16-E2F-1 also led to transactivation of the reporter construct and this was used as a positive control. The amount of chimerae was deliberately kept low in this experiment (50 ng of VP16-E2F-5 and 200 ng of VP16E2F-1) in order to obtain CAT activity in linear range for quantitative comparison. The higher CAT activity obtained with VP16-E2F-5 chimera as compared to E2F-1 chimera perhaps reflects inherent differences in the affinity of interaction between DP-1 with E2F-5 or E2F-1 although we have not ruled out the possibility
FIG. 4. E2F-5 is a transcriptional activator and shows cooperativity with DP-1. U2-OS cells were transfected with 1 mg of E2F4CAT reporter construct either alone or with various amounts of E2F-5 or DP-1 expression vectors (pSGI-E2F-5 and pCMV-DP-1, respectively) as shown. Cell extracts were assayed for CAT activity 48 h posttransfection. The total amount of DNA was kept constant by including varying amounts of pBluescript plasmid. pCMV-b-gal (0.2 mg) was used as an internal control for monitoring the efficiency of transfection.
589
AID
BBRC 8010
/
6946$$$382
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 5. E2F-5 interacts with DP-1 as examined by a two-hybrid system. The upper panel depicts various chimerae and reporter constructs used in the experiment. U2-OS cells were transfected with 1 mg of G5EC reporter plasmid alone or with chimera constructs expressing Gal4-DP-1 (200 ng), VP16-E2F-5 (50 ng), or VP16-E2F-1 (50 ng) in various combinations as shown.
of differences in the level of expression and/or the stability of chimerae. E2F-5 contains a powerful transactivation domain between amino acid residues 273-346. The presence of a transactivation domain in E2F-5 was examined using the Gal4 fusion approach. The vector pSG424 expresses the N-terminal 147 amino acid residues of the yeast Gal4 protein which constitutes the DNA binding domain (53). Since transactivation domains in other E2F proteins have been mapped to their C-termini (8, 46), we made a construct that expresses the Gal4 DNA-binding domain (DBD) fused in-frame to the C-terminal 74 amino acid residues (273-346) derived from the E2F-5 protein. This chimera strongly activated CAT expression from G5EC, a Gal4-responsive reporter construct (Fig. 6). Thus the C-terminal end of E2F-5 (273-346 residues) contains a powerful transactivation domain. We have also made another chimera that expresses Gal4 DBD fused in-frame to a larger segment of E2F-5 protein from 70-346 amino acid residues. This chimera also transactivated G5EC reporter construct (data not shown). The construct that expresses the Gal4 DBD fused to the transactivation domain of the powerful herpesvirus VP16 transactivator protein acted as a positive control for the experiment. DISCUSSION We report here the isolation and characterization of E2F-5, a fifth member of the E2F family of transcrip-
tion factors. It is more homologous to E2F-4 than to E2F-1, E2F-2 or E2F-3. The multiple sequence alignment of all the E2F members clearly showed that there are two distinct sub-families, one comprised of E2F-1, E2F-2 and E2F-3 and the other comprised of E2F-4 and E2F-5. This evolutionary relatedness is also reflected in the functional similarities between the members of each sub-family. For example, each E2F member specifically interacts with a distinct member of the pRB family of tumor suppressor proteins with regulatory consequences. Specifically, E2F-1, E2F-2 and E2F3 proteins interact with pRB, but not with p107 or p130, whereas E2F-4 protein interacts with p107 and p130, but not with pRB (11-14). While this work was in progress, others have reported the isolation of E2F5 cDNA and have shown the interaction between E2F5 and p130 (14-16). Furthermore, E2F-1, E2F-2 and E2F-3 have been shown to directly bind to cyclin A (48). The cyclin binding domain is missing in E2F-4 and E2F-5 and they presumably do not directly bind to cyclin A. Similarly, E2F-1, E2F-2 and E2F-3 were shown to interact with Sp1 transcription factor through a sequence in the amino terminus, and synergistically activate transcription. In contrast, E2F-4 and E2F-5 do not interact with Sp1 owing to the lack of Sp1 binding sequence (61). When expressed alone E2F-1, E2F-2 or E2F-3 can induce quiescent cells to enter into S phase as well as overcome a G1 arrest mediated by the p16INK4 tumor suppressor protein. On the other hand, simultaneous expression of DP-1 is required to effect the same changes by E2F-4 or E2F-5 (62). Thus structurally as well as functionally E2F-4 and E2F-5 proteins define a distinct sub-family of E2F transcription factors. Although E2F-5 mRNA is expressed ubiquitously in various human tissues, skeletal muscles express very
FIG. 6. E2F-5 contains a potent transactivation domain at its carboxy terminus. U2-OS cells were transfected with 1 mg of G5EC reporter plasmid alone or with 1 mg of constructs expressing Gal4 DBD, Gal4 DBD-E2F-5 (273-346) chimera, or 0.1 mg of Gal4 DBDVP16 chimera.
590
AID
BBRC 8010
/
6946$$$383
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
low level of it. Thus, it appears that there may be some tissue specificity associated with E2F-5 expression. The relative levels of mRNA expression are : kidney ú heart, placenta, lung ú brain, pancreas ú skeletal muscle. Like other members of the E2F family, E2F-5 is also a phosphoprotein and it exists as multiple species due to differential phosphorylation. This is consistent with the presence in E2F-5 sequence of numerous potential sites for phosphorylation by various kinases. Phosphorylation may play an important role in E2F-5 function. We have also shown that E2F-5 is a transcriptional activator in the presence of DP-1. The synergistic effect of DP-1 on transactivation by E2F-5 probably reflects a need by E2F-5 to heterodimerize with a member of the DP family for high affinity DNA recognition, as documented in the case of other E2F proteins (3-4, 6). Consistent with this, we have shown the interaction between E2F-5 and DP-1 using two hybrid system in mammalian cells. We have demonstrated the presence of a powerful transactivation domain in E2F-5 at its C-terminal end (273-346 amino acid residues). Within this domain, a sequence from 273-318 residues is completely divergent from E2F-4 as well as other E2F proteins. The only feature shared by all E2F proteins in this region is the relative enrichment of acidic amino acid residues consistent with its transactivation function. However, the sequence within 319-346 residues of this domain at the extreme C-terminal end is highly conserved in all E2F proteins, and it likely constitutes a pocket protein binding region. The presence of multiple members in the E2F family raises an important question regarding the biological significance of the apparent functional redundancy. Clearly more work is needed in order to address this issue. It will be particularly important to understand the subtle differences, if any, in the cognate DNA recognition sequences, differential regulation of the activities, temporal expression in different phases of cell cycle and tissue specific distribution of the E2F proteins.
2. Slansky, J. E., and Farnham, P. J. (1996) in Current Topics in Microbiology and Immunology: Transcriptional Control of Cell Growth: The E2F Gene Family (Farnham, P. J., Ed., pp. 1–30, Springer Verlag, Berlin/Heidelberg, Germany. 3. Bandara, L. R., Buck, V. M., Zamanian, M., Johnston, L. H., and La Thangue, N. B. (1993) EMBO J. 12, 4317–4324. 4. Helin, K., Wu, C., Fattaey, A. R., Lees, J. A., Dynlacht, B. D., Ngwu, C., and Harlow, E. (1993) Genes Dev. 7, 1850–1861. 5. Krek, W., Livingston, D. M., and Shirodkar, S. (1993) Science 262, 1557–1560. 6. Wu, C. L., Zukerberg, L. R., Ngwu, C., Harlow, E., and Lees, J. A. (1995) Mol. Cell. Biol. 15, 2536–2546. 7. Helin, K., Lees, J. A., Vidal, M., Dyson, N., Harlow, E., and Fattaey, A. (1992) Cell 70, 337–350. 8. Kaelin, W. G. Jr., Krek, W., Sellers, W. R., DeCaprio, J. A., Ajchenbaum, F., Fuchs, C. S., Chittenden, T., Li, Y., Farnham, P. J., Blanar, M. A., Livingston, D. M., and Flemington, E. K. (1992) Cell 70, 351–364. 9. Shan, B., Zhu, X., Chen, P-L., Durfee, T., Yang, Y., Sharp, D., and Lee, W-H. (1992) Mol. Cell. Biol. 12, 5620–5631. 10. Ivey-Hoyle, M., Conroy, R., Huber, H. E., Goodhart, P. J., Oliff, A., and Heimbrook, D. C. (1993) Mol. Cell. Biol. 13, 7802–7812. 11. Lees, J. A., Saito, M., Vidal, M., Valentine, M., Look, T., Harlow, E., Dyson, N., and Helin, K. (1993) Mol. Cell. Biol. 13, 7813– 7825. 12. Ginsberg, D., Vairo, G., Chittenden, T., Xiao, Z-X., Xu, G., Wydner, K. L., DeCaprio, J. A., Lawrence, J. B., and Livingston, D. M. (1994) Genes Dev. 8, 2665–2679. 13. Beijersbergen, R. L., Kerkhoven, R., Zhu, L., Carlee, L., Voorhoeve, P. M., and Bernards, R. (1994) Genes Dev. 8, 2680–2690. 14. Sardet, C., Vidal, M., Cobrinik, D., Geng, Y., Onufryk, C., Chen, A., and Weinberg, R. A. (1995) Proc. Natl. Acad. Sci. USA 92, 2403–2407. 15. Hijmans, E. M., Voorhoeve, P. M., Beijersbergen, R. L., van’t Veer, L. J., and Bernards, R. (1995) Mol. Cell. Biol. 15, 3082– 3089. 16. Buck, V., Allen, K. E., Sorensen, T., Bybee, A., Hijmans, E. M., Voorhoeve, P. M., Bernards, R., and La Thangue, N. B. (1995) Oncogene 11, 31–38. 17. Girling, R., Partridge, J. F., Bandara, L. R., Burden, N., Totty, N. F., Hsuan, J. J., and La Thangue, N. B. (1993) Nature 362, 83–87. 18. Zhang, Y., and Chellappan, S. P. (1995) Oncogene 10, 2085– 2093. 19. Pearson, B. E., Nasheuer, H-P., and Wang, TS-F. (1991) Mol. Cell. Biol. 11, 2081–2095.
ACKNOWLEDGMENTS We thank Ed Harlow for providing E2F4CAT and pCMV-DP-1; Joseph Nevins for pHB44-DP-124-410 and pGAD10-E2F-1 wt; G. Nallur and H. Vasavada for G5EC, pSG424, AASV-VP16, and Gal4VP16; V. Kumar for pSGI vector; and N. Jayasuryan for help in raising antibodies. We also thank S. Jameel, S. Sopori, and S. Mukherjee for critical reading of the manuscript. Technical assistance by R. Verma and S. Sehrawat and cell culture maintenance by R. Kumar are gratefully acknowledged. This work was supported initially by Universitywide AIDS Research Program (UARP) Grant R92-SD-137 and a scholar award from the American Foundation for AIDS Research (AmFAR) to Y.N.V. and subsequently by internal funds from the ICGEB.
20. Blake, M. C., and Azizkhan, J. C. (1989) Mol. Cell. Biol. 9, 4994– 5002. 21. Kim, Y. K., and Lee, A. S. (1991) Mol. Cell. Biol. 10, 2296–2302. 22. Takehana, K., Nakada, S., Hara, E., Taya, Y., Sekiya, S., and Oda, K. (1991) Gene 103, 219–225. 23. Hiebert, S. W., Lipp, M., and Nevins, J. R. (1989) Proc. Natl. Acad. Sci. USA 86, 3594–3598. 24. Park, K., Choe, J., Osifchin, N. E., Templeton, D. J., Robbins, P. D., and Kim, S-J. (1994) J. Biol. Chem. 269, 6083–6088. 25. Shan, B., Chang, C. Y., Jones, D., and Lee, W. H. (1994) Mol. Cell. Biol. 14, 299–309. 26. Dalton, S. (1992) EMBO J. 11, 1797–1804. 27. Furukawa, Y., Terui, Y., Sakoe, K., Ohta, M., and Saito, M. (1994) J. Biol. Chem. 269, 26249–26258.
REFERENCES 1. Kovesdi, I., Reichel, R., and Nevins, J. R. (1986) Cell 45, 219– 228.
28. Johnson, D. G., Ohtani, K., and Nevins, J. R. (1994) Genes Dev. 8, 1514–1525.
591
AID
BBRC 8010
/
6946$$$383
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
29. Neuman, E., Flemington, E. K., Sellers, W. R., and Kaelin, W. G., Jr. (1994) Mol. Cell. Biol. 14, 6607–6615. 30. Lam, EW-F., and Watson, R. J. (1993) EMBO J. 12, 2705–2713.
46.
31. Johnson, D. G., Schwarz, J. K., Cress, W. D., and Nevins, J. R. (1993) Nature 365, 349–352.
47.
32. Qin, X-Q., Livingston, D. M., Kaelin, W. G., Jr., and Adams, P. (1994) Proc. Natl. Acad. Sci. USA 91, 10918–10922.
48.
33. Shan, B., and Lee, W-H. (1994) Mol. Cell. Biol. 14, 8166–8173.
49.
34. Singh, P., Wong, S., and Hong, W. (1994) EMBO J. 13, 3329– 3338.
50.
35. Johnson, D. G., Cress, W., Jakoi, L., and Nevins, J. R. (1994) Proc. Natl. Acad. Sci. USA 91, 12823–12827.
51.
36. Xu, G., Livingston, D. M., and Krek, W. (1995) Proc. Natl. Acad. Sci. USA 92, 1357–1361. 37. Bremner, R., Cohen, B. L., Sopta, M., Hamel, P. A., Ingles, C. J., Gallie, B. L., and Phillips, R. A. (1995) Mol. Cell. Biol. 15, 3256– 3265. 38. Flemington, E. K., Speck, S. H., and Kaelin, W. G., Jr. (1993) Proc. Natl. Acad. Sci. USA 90, 6914–6918.
52. 53. 54. 55.
39. Smith, E. J., and Nevins, J. R. (1995) Mol. Cell. Biol. 15, 338– 344. 40. Vairo, G., Livingston, D. M., and Ginsberg, D. (1995) Genes Dev. 9, 869–881. 41. Zamanian, M., and La Thangue, N. B. (1993) Mol. Cell. Biol. 4, 389–396. 42. Zhu, L., van den Heuvel, S., Helin, K., Fattaey, A., Ewen, M., Livingston, D. M., Dyson, N., and Harlow, E. (1993) Genes Dev. 7, 1111–1125. 43. Helin, K., Harlow, E., and Fattaey, A. (1993) Mol. Cell. Biol. 13, 6501–6508. 44. Bagchi, S., Raychaudhuri, P., and Nevins, J. R. (1990) Cell 62, 659–669.
56. 57. 58. 59.
60. 61. 62.
45. Chellappan, S., Kraus, V. B., Kroger, B., Munger, K., Howley,
P. M., Phelps, W. C., and Nevins, J. R. (1992) Proc. Natl. Acad. Sci. USA 89, 4549–4553. Cress, W. D., Johnson, D. G., and Nevins, J. R. (1993) Mol. Cell. Biol. 13, 6314–6325. Jordan, K. L., Haas, A. R., Logan, T. J., and Hall, D. J. (1994) Oncogene 9, 1177–1185. Krek, W., Ewen, M. E., Shirodkar, S., Arany, Z., Kaelin, W. G., Jr., and Livingston, D. M. (1994) Cell 78, 161–172. Vaishnav, Y. N., Vaishnav, M., and Wong-Staal, F. (1991) New Biologist 3, 142–150. Vinson, C. R., LaMarco, K. L., Johnson, P. F., Landschulz, W. H., and McKnight, S. L. (1988) Genes Dev. 2, 801–806. Ausbel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K. (1989) Current Protocols in Molecular Biology, Wiley, New York. Jameel, S., Zafrullah, M., Ozdener, M. H., and Panda, S. K. (1996) J. Virol. 70, 207–216. Sadowski, I., and Ptashne, M. (1989) Nucleic Acids Res. 17, 7539. Sadowski, I., Ma, J., Triezenberg, S., and Ptashne, M. (1988) Nature 335, 563–564. Vasavada, H. A., Ganguly, S., Germino, F. J., Wang, Z. X., and Weissman, S. M. (1991) Proc. Natl. Acad. Sci. USA 88, 10686– 10690. Cress, W. D., and Nevins, J. R. (1994) J. Virol. 68, 4212–4219. Reichlin, M. (1980) Methods Enzymol. 70, 159–165. Vojtesk, B., Bartek, J., Midgley, C. A., and Lane, D. P. (1992) J. Immunol. Methods 151, 237–244. Harlow, E., and Lane, D. (1988) Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Schuler, G. D., Altschul, S. F., and Lipman, D. J. (1991) Proteins Structure Function Genetics 9, 180–190. Karlseder, J., Rothender, H., and Wintersberger, E. (1996) Mol. Cell. Biol. 16, 1659–1667. Lukas, J., Petersen, B. O., Holm, K., Bartek, J., and Helin, K. (1996) Mol. Cell. Biol. 16, 1047–1057.
592
AID
BBRC 8010
/
6946$$$383
01-07-98 09:34:11
bbrcg
AP: BBRC
242, 586–592 (1998)
RC978010
The Molecular and Functional Characterization of E2F-5 Transcription Factor Yashwantrai N. Vaishnav,1 Mahima Y. Vaishnav, and Vinod Pant Virology Group, International Centre for Genetic Engineering and Biotechnology, Aruna Asaf Ali Marg, P. O. Box 10504, New Delhi - 110067, India
Received November 17, 1997
The E2F activity plays a critical role in the control of cell cycle and action of tumor suppressor proteins and is also a target of the transforming proteins of small DNA tumor viruses. We describe here molecular cloning and functional characterization of a fifth member of the E2F family of transcription factors. E2F-5 protein is more homologous to E2F-4 (72% amino acid identity) than to E2F-1, E2F-2, and E2F-3 (35% amino acid identity). Based on structural and functional criteria, the E2F family appears to comprise two distinct sub-families, one composed of E2F-1, E2F-2, and E2F3 and the other composed of E2F-4 and E2F-5, E2F-5 mRNA is expressed in a wide variety of human tissues. The protein is expressed as multiple species ranging in size from 46 to 54 kDa as a result of differential phosphorylation. The expression of a reporter gene containing E2F binding sites in the promoter is transcriptionally activated by E2F-5 in a cooperative manner with the DP-1 protein. The interaction between E2F-5 and DP-1 is demonstrated using a two-hybrid system in mammalian cells. We have also demonstrated the presence of a strong transactivation domain at the carboxy terminus (273-346 amino acid residues) of E2F-5 protein. q 1998 Academic Press
The E2F activity was initially identified as a cellular DNA-binding activity required for the transactivation of viral promoter E2 by adenovirus E1A protein (1). The generic term ‘‘E2F activity’’ is now used to describe a heterodimer containing one subunit derived from the E2F family and another derived from the DP family (reviewed in 2). The E2F/DP heterodimer binds to DNA in a sequence-specific manner with high affinity and results in transcriptional activation of genes (3-6). Several members of the E2F and DP families have been molecularly cloned and characterized (4, 7-18). The con1 Corresponding author. Fax: 91-11-6162316. E-mail: yash@ icgebnd.ernet.in.
0006-291X/98 $25.00
sensus E2F-binding sites are found in the promoters of a variety of genes some of which are specifically expressed in S phase of the cell cycle and are required for DNA synthesis e.g. DNA polymerase a, dihydrofolate reductase and thymidine kinase (19-21). The promoters of certain genes which code for transcriptional regulators of cell growth also harbor E2F sites e.g. N-myc, cmyc, RB, cdc2, E2F-1 and B-myb (22-30). Many of these sites have been shown to be required for transcriptional activation in a cell cycle dependent manner. Overexpression of E2F-1 causes quiescent cells to progress into S phase of the cell cycle (31-33). Several studies suggest that deregulated E2F expression or mutation of E2F so as to release it from negative regulation by pocket proteins is potentially oncogenic (12-13, 34-36). Thus they appear to be critically involved in the control of cell cycle and consequently cell growth. The E2F/DP heterodimer is usually present in a complex with the RB family of tumor suppressor proteins such as pRB (7-8, 10-11), p107 (12-14) or p130 (14-15). The resultant complex is incompetent for transactivation (37-43). The site of interaction with the tumor suppressor proteins is located in the C-terminal end within the transactivation domain of the E2F proteins. The cell cycle dependent phosphorylation-dephosphorylation of tumor suppressor proteins modulate their interaction with E2F (7-9). Transforming proteins of DNA tumor viruses such as adenovirus E1A, SV40/polyoma virus large T antigen and human papilloma virus E7 dissociate E2F from the complex with tumor suppressor proteins (44-45). The ‘‘free’’ E2F activity, thus released, transcriptionally activates various E2F-responsive genes resulting in a loss of cell growth control. This provides a model to explain the transforming activity of these viral proteins. The E2F proteins contain several domains that are evolutionarily conserved among all members of the family (2). The DNA binding domain is composed of basic residues as well as a helix-loop-helix motif (4647). The domain that determines interaction with DP proteins is located downstream of the DNA binding
586
Copyright q 1998 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 8010
/
6946$$$381
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
domain (DBD) (4-5, 47). The transactivation domain enriched in acidic amino acids is located at the C-terminus (8, 46). Embedded within this domain is a stretch of conserved amino acid residues involved in mediating the interaction of these proteins with tumor suppressor proteins. These features are common to all members of the E2F family of transcription factors. Additionally, various members have certain exclusive features. For example, E2F-1, E2F-2 and E2F-3 have a cyclin binding domain located immediately upstream of the DNA binding domain (48). E2F-4 has a serine rich region that is encoded by varying copies of trinucleotide repeats (12-13). In this communication, we describe the molecular cloning and functional characterization of a fifth member of the E2F family of transcription factors designated as E2F-5. MATERIALS AND METHODS Molecular cloning of E2F-5. The E2F-5 cDNA clone was obtained fortuitously while screening a cDNA expression library in search of a clone for NFRRE , a cellular factor that specifically interacts with the Rev response element (RRE) of the human immunodeficiency virus type 1 (HIV-1) (49). The method used for screening was essentially as described for DNA-binding proteins (50). A lgt11 cDNA expression library prepared from activated HUT78 cells (Clontech) was screened with in vitro transcribed RRE as a probe. The cDNA insert obtained from recombinant phage DNA by digesting with EcoRI was subcloned into pBluescript SK(/) vector (Stratagene). JM109 cells carrying the recombinant plasmid were infected with the helper phage R408 (Clontech) and single-stranded phagemid DNA was prepared from the supernatant. Both strands of the insert were sequenced by the dideoxy chain termination method (51) initially using universal primers (forward and reverse) and subsequently gene specific primers as the sequence information became available. Plasmids. For expression in mammalian cells, E2F-5 open reading frame was subcloned into an eukaryotic expression vector pSGI (52). Expression of the insert from pSGI is controlled by SV40 early promoter/enhancer region. The vector pSG424 (53) expresses the DNA-binding domain (1-147 amino acid residues) of the yeast Gal4 transactivator protein. Gal4-E2F-5 (273-346) and Gal4-E2F-5 (70346) expression constructs were made by cloning E2F-5 cDNA (corresponding to 273-346 or 70-346 amino acid residues respectively) into pSG424 vector at a SmaI site in-frame with Gal4 DBD. E2F4CAT (4), pCMV-DP-1 (4), G5EC (54), Gal4-VP16 (54), have been described. pBluescript SK (/) (Stratagene) and RcCMV (Invitrogen) were obtained commercially. Two hybrid system approach, as adapted for mammalian cells (55), was used for probing protein-protein interactions between DP-1 and E2F-5 or E2F-1. Gal4-DP-1 expression construct was made by cloning an EcoRI DP-1 insert derived from pHB44-DP-124-410 (56) at EcoRI site in pSG424 vector. The vector AASV-VP16 (54) expresses the transactivation domain (413-490 amino acid residues) of the herpesvirus VP16 transactivator protein. VP16-E2F-5 (70-346) chimera was made by cloning an EcoRI E2F-5 fragment derived from Gal4E2F-5 (70-346) at an EcoRI site in AASV-VP16. VP16-E2F-1 chimera was made by cloning an EcoRI E2F-1 insert derived from pGAD10E2F-1 wt (56) in AASV-VP16. Northern blot analysis. The multiple tissue Northern blot (Clontech) containing 2 mg of poly A/ RNA derived from various human tissues was probed with E2F-5 cDNA. The 1.7 kb E2F-5 cDNA insert was 32P-radiolabeled to a specific activity of 21108 cpm/mg by random priming (Boehringer Mannheim) and was used for hybridization.
The blot was subsequently reprobed with a synthetic oligonucleotide (Clontech) corresponding to b-actin gene (5*-GACGACGAGCGCGGCGATATCATCATC-3*). Antibodies. A synthetic peptide (NLPEQHVSER SQALQQTSAT DISSAGSISG D) corresponding to the amino acid residues 271-301 of E2F-5 was coupled to BSA with glutaraldehyde (57) and used for raising polyclonal antibodies in rabbits. The sequence of this peptide was derived from the region of E2F-5 which is highly divergent from E2F-4 as determined by sequence alignment (see Fig. 1). The monoclonal antibody against p53 (DO-1) that recognises the N-terminal epitope (58) was obtained from Santa Cruz Biotechnology. Transfection, radio-immunoprecipitation assay (RIPA), and chloramphenicol acetyltransferase (CAT) assay. COS-1 and U2-OS cells in 60-mm culture dishes were transfected using lipofectin reagent (GIBCO-BRL) according to the manufacturer’s guidelines. The total amount of DNA was kept constant by including pBluescript SK(/). pCMV-b-gal (0.2 mg) was used as an internal control for monitoring the efficiency of transfection. For RIPA, transfected cells were metabolically labeled 48 h post-transfection with 35S-methionine (100-200 mCi/60 mm plate) in methionine-free DMEM for about 4 h and extracts were prepared in RIPA buffer (10 mM Tris-HCl [pH 8.0], 140 mM NaCl, 5 mM iodoacetamide, 0.5% Triton X-100, 1% sodium deoxycholate, 0.1% sodium dodecyl sulfate [SDS], 2 mM phenylmethylsulfonyl fluoride). Immunoprecipitation was carried out as described (59). For CAT assay, the cell extracts were prepared 48 h post-transfection and CAT activity was monitored by standard method (49). Each experiment was repeated at least 3 times and representative data are presented. Nucleotide sequence accession number. The nucleotide sequence of E2F-5 cDNA reported here has been deposited in EMBL databank under accession number Z78409.
RESULTS E2F-5 is highly homologous to E2F-4. While screening a human cDNA expression library in lgt11, we isolated a gene that codes for a fifth member of the E2F family of transcription factors designated as E2F-5. The multiple sequence alignment of E2F proteins by MACAW (60) revealed that E2F-5 is more closely related to E2F-4 (72 % amino acid identity) than to E2F1, E2F-2 or E2F-3 (32 % amino acid identity). The E2F5 cDNA clone is 1698 bp long. It contains a complete ORF of 1038 bp that can potentially code for a protein of 346 amino acids with a predicted molecular mass of Ç38 kDa. The cDNA has 34 bp of 5*-untranslated sequences and 626 bp of 3*-untranslated sequences. The polyA signal (AAUAAA) is missing in E2F-5 cDNA indicating that it does not contain complete 3*-untranslated sequences. The alignment of E2F-4 and E2F-5 amino acid sequences indicated extensive homology between the two proteins (Fig. 1). A large block in E2F5 from amino acid residues 39-268 is highly conserved between E2F-4 and E2F-5. This is followed by a highly divergent sequences. Finally, a short sequence at the extreme C-terminus (amino acid residues 327-345) is well conserved between the two proteins and presumably forms a pocket protein-binding domain. Multiple alignment of various E2F proteins allowed us to predict the position of various functional domains in E2F-5 on the basis of homology with the domains
587
AID
BBRC 8010
/
6946$$$382
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
that is conserved among E2F-1, E2F-2 and E2F-3 proteins, and which has been shown in E2F-1 to bind to cyclin A (48).
FIG. 1. The optimum alignment of E2F-5 and E2F-4 amino acid sequences derived using MACAW. The amino acid residues belonging to homologous blocks are presented in upper case. The identical amino acid residues are boxed.
that are well characterized in other E2F proteins, especially E2F-1. The region between residues 49 to 116 may constitute a DNA binding domain since it is relatively enriched in basic residues and is also homologous to the DNA binding domain as defined for E2F-1 (4, 7, 11). The DNA binding domain is the most highly conserved region between all members of the E2F family. A region from residues 133 to 201 may qualify as the heterodimerization domain based on its sequence homology with the corresponding domain in E2F-1 (4). It includes a leucine heptad repeat between amino acid residues 133-154 which is predicted to form a stable amphipathic a-helix. The leucine residues in this heptad repeat are highly conserved among all five E2F proteins even though amino acid residues present between the conserved leucines are quite divergent suggesting that hydrophobic residues occurring at every 7th position may form an interface required to interact with DP proteins. The region between residues 202 to 318 is quite divergent among members of the E2F family, except for a weak but definite homology in the sequence (214 to 232 residues) earlier designated as ‘marked box’ (11). However, E2F-4 and E2F-5 amino acid sequences exhibit considerable homology not only in the marked box but also adjacent sequences from 202 to 270 residues. The E2F-5 sequence from 271 to 318 residues is completely divergent from E2F-4 as well as other E2F sequences. It includes a serine rich region in E2F-4 that is encoded by varying copies of the trinucleotide repeats (12-13). The C-terminal end from residues 319 to 346 is well conserved between E2F-4 and E2F-5, as well as between all other members of the E2F family. It presumably includes a pocket protein binding region that is involved in interaction with pRB family of tumor suppressor proteins. Both E2F-4 and E2F-5 lack at their N-termini, a short sequence
E2F-5 mRNA is ubiquitously expressed in human tissues. In order to examine the expression of E2F-5 mRNA in various human tissues, a Northern blot containing poly(A)/ RNA derived from various tissues was probed with E2F-5 cDNA. E2F-5 mRNA of Ç2.2 kb was found to be expressed in a wide variety of human tissues (Fig. 2). Since the size of E2F-5 cDNA obtained is Ç1.7 kb, this indicated that the difference could be due to the absence of complete 5*- and 3*- untranslated sequences. Interestingly, skeletal muscles appear to express very low level of the E2F-5 mRNA. The lower signal observed in case of liver, however, was due to a low input of poly(A)/ RNA in that lane as determined by re-probing the same blot for b-actin expression (lower panel, Fig. 2). E2F-5 is differentially phosphorylated. In order to characterize the protein product, the E2F-5 cDNA was subcloned into pSGI eukaryotic expression vector under the control of SV40 early promoter/enhancer. COS1 cells were transfected with pSGI vector or pSGI-E2F5 expression construct, metabolically labeled with 35Smethionine and cell extracts were immunoprecipitated with E2F-5 specific rabbit polyclonal antiserum. As a control, pre-immune rabbit serum was used for precipitation. The E2F-5 specific antibodies immunoprecipitate proteins ranging in size from 46 kDa (the fastest
FIG. 2. The expression of E2F-5 mRNA in various human tissues. Northern blot containing poly(A)/ RNA from various tissues, as indicated, was hybridized to a human E2F-5 cDNA probe. The lower panel shows signals obtained by re-probing the same blot with a b-actin oligonucleotide. Note that the 2.2 kb species of b-actin mRNA is present in all the tissues while a 1.8 kb species of mRNA is present only in cardiac and skeletal muscle tissues.
588
AID
BBRC 8010
/
6946$$$382
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Differential phosphorylation of E2F-5 protein. Metabolically labeled extracts prepared from COS-1 cells transfected with 2.5 mg of either vector alone (pSGI) or E2F-5 expression construct (pSGIE2F-5) were immunoprecipitated with preimmune (PI) serum or polyclonal rabbit immune (I) serum and separated on SDS-7.5 % polyacrylamide gel. In some cases, the immunoprecipitates were treated with l phosphatase in the presence or absence of orthovanadate.
migrating), 49 kDa (intermediate band) to 54 kDa (the slowest migrating) from the cells transfected with pSGI-E2F-5 (Fig. 3). The pre-immune serum did not precipitate these proteins. Post-translation modification could possibly account for such heterogeneity. We examined the phosphorylation status of E2F-5 since it has several potential phosphorylation sites. The immunoprecipitated proteins were treated with l phosphatase (NEB) either in the presence or absence of orthovanadate, a specific inhibitor of phosphatases. As a control, immunoprecipitated proteins were treated with phosphatase buffer alone. The phosphatase treatment abolished 54 kDa and 49 kDa bands and most of the protein could be seen as a single species migrating as a 46 kDa band, thus indicating that the 54 kDa and 49 kDa bands represented differentially phosphorylated species of E2F-5 protein. The presence of orthovanadate during phosphatase treatment effectively prevented this change. The discrepancy between the predicted molecular weight of 38 kDa and the fastest migrating species of 46 kDa could be due to anomalous migration of E2F-5 in SDS-PAGE. The presence of two proline-rich regions (6-38 and 238-262 amino acid residues) in E2F-5 could possibly account for the anomalous migration. E2F-5 activates transcription in a cooperative manner with DP-1. Since all the E2F proteins share a common property of transactivating gene expression in an E2F site-dependent manner, we examined this property of E2F-5. When a reporter CAT construct E2F4CAT was transfected in U2-OS cells along with E2F-5 or DP-1 ex-
pression constructs separately at various concentrations, there was no transactivation (data not shown). However, when E2F4CAT was co-transfected with increasing amount of E2F-5 in presence of constant amount of DP-1, a dose dependent increase in the extent of transactivation could be seen (Fig. 4). Similarly, increasing amount of DP-1 in presence of a constant amount of E2F-5 led to transcriptional activation. However, the saturation was reached earlier as compared to E2F-5 and higher amount of DP-1 led to inhibition. The fact that E2F-5 and DP-1 could transcriptionally activate reporter gene expression only when expressed together indicated a co-operative action. The requirement of co-expression of E2F-5 and DP-1 presumably reflects that heterodimerization between the two proteins is essential for efficient transactivation. The protein-protein interaction between E2F-5 and DP-1 was examined using a mammalian version of the two hybrid system (55). For this purpose, we constructed Gal4-DP-1, VP16-E2F-5 and VP16-E2F-1 chimerae as shown in Fig. 5. None of these chimerae transactivated the reporter construct when transfected alone as expected. However, when Gal4-DP-1 was cotransfected with VP16-E2F-5, there was transactivation indicating the interaction between E2F-5 and DP-1. Similarly, co-transfection with Gal4-DP-1 and VP16-E2F-1 also led to transactivation of the reporter construct and this was used as a positive control. The amount of chimerae was deliberately kept low in this experiment (50 ng of VP16-E2F-5 and 200 ng of VP16E2F-1) in order to obtain CAT activity in linear range for quantitative comparison. The higher CAT activity obtained with VP16-E2F-5 chimera as compared to E2F-1 chimera perhaps reflects inherent differences in the affinity of interaction between DP-1 with E2F-5 or E2F-1 although we have not ruled out the possibility
FIG. 4. E2F-5 is a transcriptional activator and shows cooperativity with DP-1. U2-OS cells were transfected with 1 mg of E2F4CAT reporter construct either alone or with various amounts of E2F-5 or DP-1 expression vectors (pSGI-E2F-5 and pCMV-DP-1, respectively) as shown. Cell extracts were assayed for CAT activity 48 h posttransfection. The total amount of DNA was kept constant by including varying amounts of pBluescript plasmid. pCMV-b-gal (0.2 mg) was used as an internal control for monitoring the efficiency of transfection.
589
AID
BBRC 8010
/
6946$$$382
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 5. E2F-5 interacts with DP-1 as examined by a two-hybrid system. The upper panel depicts various chimerae and reporter constructs used in the experiment. U2-OS cells were transfected with 1 mg of G5EC reporter plasmid alone or with chimera constructs expressing Gal4-DP-1 (200 ng), VP16-E2F-5 (50 ng), or VP16-E2F-1 (50 ng) in various combinations as shown.
of differences in the level of expression and/or the stability of chimerae. E2F-5 contains a powerful transactivation domain between amino acid residues 273-346. The presence of a transactivation domain in E2F-5 was examined using the Gal4 fusion approach. The vector pSG424 expresses the N-terminal 147 amino acid residues of the yeast Gal4 protein which constitutes the DNA binding domain (53). Since transactivation domains in other E2F proteins have been mapped to their C-termini (8, 46), we made a construct that expresses the Gal4 DNA-binding domain (DBD) fused in-frame to the C-terminal 74 amino acid residues (273-346) derived from the E2F-5 protein. This chimera strongly activated CAT expression from G5EC, a Gal4-responsive reporter construct (Fig. 6). Thus the C-terminal end of E2F-5 (273-346 residues) contains a powerful transactivation domain. We have also made another chimera that expresses Gal4 DBD fused in-frame to a larger segment of E2F-5 protein from 70-346 amino acid residues. This chimera also transactivated G5EC reporter construct (data not shown). The construct that expresses the Gal4 DBD fused to the transactivation domain of the powerful herpesvirus VP16 transactivator protein acted as a positive control for the experiment. DISCUSSION We report here the isolation and characterization of E2F-5, a fifth member of the E2F family of transcrip-
tion factors. It is more homologous to E2F-4 than to E2F-1, E2F-2 or E2F-3. The multiple sequence alignment of all the E2F members clearly showed that there are two distinct sub-families, one comprised of E2F-1, E2F-2 and E2F-3 and the other comprised of E2F-4 and E2F-5. This evolutionary relatedness is also reflected in the functional similarities between the members of each sub-family. For example, each E2F member specifically interacts with a distinct member of the pRB family of tumor suppressor proteins with regulatory consequences. Specifically, E2F-1, E2F-2 and E2F3 proteins interact with pRB, but not with p107 or p130, whereas E2F-4 protein interacts with p107 and p130, but not with pRB (11-14). While this work was in progress, others have reported the isolation of E2F5 cDNA and have shown the interaction between E2F5 and p130 (14-16). Furthermore, E2F-1, E2F-2 and E2F-3 have been shown to directly bind to cyclin A (48). The cyclin binding domain is missing in E2F-4 and E2F-5 and they presumably do not directly bind to cyclin A. Similarly, E2F-1, E2F-2 and E2F-3 were shown to interact with Sp1 transcription factor through a sequence in the amino terminus, and synergistically activate transcription. In contrast, E2F-4 and E2F-5 do not interact with Sp1 owing to the lack of Sp1 binding sequence (61). When expressed alone E2F-1, E2F-2 or E2F-3 can induce quiescent cells to enter into S phase as well as overcome a G1 arrest mediated by the p16INK4 tumor suppressor protein. On the other hand, simultaneous expression of DP-1 is required to effect the same changes by E2F-4 or E2F-5 (62). Thus structurally as well as functionally E2F-4 and E2F-5 proteins define a distinct sub-family of E2F transcription factors. Although E2F-5 mRNA is expressed ubiquitously in various human tissues, skeletal muscles express very
FIG. 6. E2F-5 contains a potent transactivation domain at its carboxy terminus. U2-OS cells were transfected with 1 mg of G5EC reporter plasmid alone or with 1 mg of constructs expressing Gal4 DBD, Gal4 DBD-E2F-5 (273-346) chimera, or 0.1 mg of Gal4 DBDVP16 chimera.
590
AID
BBRC 8010
/
6946$$$383
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
low level of it. Thus, it appears that there may be some tissue specificity associated with E2F-5 expression. The relative levels of mRNA expression are : kidney ú heart, placenta, lung ú brain, pancreas ú skeletal muscle. Like other members of the E2F family, E2F-5 is also a phosphoprotein and it exists as multiple species due to differential phosphorylation. This is consistent with the presence in E2F-5 sequence of numerous potential sites for phosphorylation by various kinases. Phosphorylation may play an important role in E2F-5 function. We have also shown that E2F-5 is a transcriptional activator in the presence of DP-1. The synergistic effect of DP-1 on transactivation by E2F-5 probably reflects a need by E2F-5 to heterodimerize with a member of the DP family for high affinity DNA recognition, as documented in the case of other E2F proteins (3-4, 6). Consistent with this, we have shown the interaction between E2F-5 and DP-1 using two hybrid system in mammalian cells. We have demonstrated the presence of a powerful transactivation domain in E2F-5 at its C-terminal end (273-346 amino acid residues). Within this domain, a sequence from 273-318 residues is completely divergent from E2F-4 as well as other E2F proteins. The only feature shared by all E2F proteins in this region is the relative enrichment of acidic amino acid residues consistent with its transactivation function. However, the sequence within 319-346 residues of this domain at the extreme C-terminal end is highly conserved in all E2F proteins, and it likely constitutes a pocket protein binding region. The presence of multiple members in the E2F family raises an important question regarding the biological significance of the apparent functional redundancy. Clearly more work is needed in order to address this issue. It will be particularly important to understand the subtle differences, if any, in the cognate DNA recognition sequences, differential regulation of the activities, temporal expression in different phases of cell cycle and tissue specific distribution of the E2F proteins.
2. Slansky, J. E., and Farnham, P. J. (1996) in Current Topics in Microbiology and Immunology: Transcriptional Control of Cell Growth: The E2F Gene Family (Farnham, P. J., Ed., pp. 1–30, Springer Verlag, Berlin/Heidelberg, Germany. 3. Bandara, L. R., Buck, V. M., Zamanian, M., Johnston, L. H., and La Thangue, N. B. (1993) EMBO J. 12, 4317–4324. 4. Helin, K., Wu, C., Fattaey, A. R., Lees, J. A., Dynlacht, B. D., Ngwu, C., and Harlow, E. (1993) Genes Dev. 7, 1850–1861. 5. Krek, W., Livingston, D. M., and Shirodkar, S. (1993) Science 262, 1557–1560. 6. Wu, C. L., Zukerberg, L. R., Ngwu, C., Harlow, E., and Lees, J. A. (1995) Mol. Cell. Biol. 15, 2536–2546. 7. Helin, K., Lees, J. A., Vidal, M., Dyson, N., Harlow, E., and Fattaey, A. (1992) Cell 70, 337–350. 8. Kaelin, W. G. Jr., Krek, W., Sellers, W. R., DeCaprio, J. A., Ajchenbaum, F., Fuchs, C. S., Chittenden, T., Li, Y., Farnham, P. J., Blanar, M. A., Livingston, D. M., and Flemington, E. K. (1992) Cell 70, 351–364. 9. Shan, B., Zhu, X., Chen, P-L., Durfee, T., Yang, Y., Sharp, D., and Lee, W-H. (1992) Mol. Cell. Biol. 12, 5620–5631. 10. Ivey-Hoyle, M., Conroy, R., Huber, H. E., Goodhart, P. J., Oliff, A., and Heimbrook, D. C. (1993) Mol. Cell. Biol. 13, 7802–7812. 11. Lees, J. A., Saito, M., Vidal, M., Valentine, M., Look, T., Harlow, E., Dyson, N., and Helin, K. (1993) Mol. Cell. Biol. 13, 7813– 7825. 12. Ginsberg, D., Vairo, G., Chittenden, T., Xiao, Z-X., Xu, G., Wydner, K. L., DeCaprio, J. A., Lawrence, J. B., and Livingston, D. M. (1994) Genes Dev. 8, 2665–2679. 13. Beijersbergen, R. L., Kerkhoven, R., Zhu, L., Carlee, L., Voorhoeve, P. M., and Bernards, R. (1994) Genes Dev. 8, 2680–2690. 14. Sardet, C., Vidal, M., Cobrinik, D., Geng, Y., Onufryk, C., Chen, A., and Weinberg, R. A. (1995) Proc. Natl. Acad. Sci. USA 92, 2403–2407. 15. Hijmans, E. M., Voorhoeve, P. M., Beijersbergen, R. L., van’t Veer, L. J., and Bernards, R. (1995) Mol. Cell. Biol. 15, 3082– 3089. 16. Buck, V., Allen, K. E., Sorensen, T., Bybee, A., Hijmans, E. M., Voorhoeve, P. M., Bernards, R., and La Thangue, N. B. (1995) Oncogene 11, 31–38. 17. Girling, R., Partridge, J. F., Bandara, L. R., Burden, N., Totty, N. F., Hsuan, J. J., and La Thangue, N. B. (1993) Nature 362, 83–87. 18. Zhang, Y., and Chellappan, S. P. (1995) Oncogene 10, 2085– 2093. 19. Pearson, B. E., Nasheuer, H-P., and Wang, TS-F. (1991) Mol. Cell. Biol. 11, 2081–2095.
ACKNOWLEDGMENTS We thank Ed Harlow for providing E2F4CAT and pCMV-DP-1; Joseph Nevins for pHB44-DP-124-410 and pGAD10-E2F-1 wt; G. Nallur and H. Vasavada for G5EC, pSG424, AASV-VP16, and Gal4VP16; V. Kumar for pSGI vector; and N. Jayasuryan for help in raising antibodies. We also thank S. Jameel, S. Sopori, and S. Mukherjee for critical reading of the manuscript. Technical assistance by R. Verma and S. Sehrawat and cell culture maintenance by R. Kumar are gratefully acknowledged. This work was supported initially by Universitywide AIDS Research Program (UARP) Grant R92-SD-137 and a scholar award from the American Foundation for AIDS Research (AmFAR) to Y.N.V. and subsequently by internal funds from the ICGEB.
20. Blake, M. C., and Azizkhan, J. C. (1989) Mol. Cell. Biol. 9, 4994– 5002. 21. Kim, Y. K., and Lee, A. S. (1991) Mol. Cell. Biol. 10, 2296–2302. 22. Takehana, K., Nakada, S., Hara, E., Taya, Y., Sekiya, S., and Oda, K. (1991) Gene 103, 219–225. 23. Hiebert, S. W., Lipp, M., and Nevins, J. R. (1989) Proc. Natl. Acad. Sci. USA 86, 3594–3598. 24. Park, K., Choe, J., Osifchin, N. E., Templeton, D. J., Robbins, P. D., and Kim, S-J. (1994) J. Biol. Chem. 269, 6083–6088. 25. Shan, B., Chang, C. Y., Jones, D., and Lee, W. H. (1994) Mol. Cell. Biol. 14, 299–309. 26. Dalton, S. (1992) EMBO J. 11, 1797–1804. 27. Furukawa, Y., Terui, Y., Sakoe, K., Ohta, M., and Saito, M. (1994) J. Biol. Chem. 269, 26249–26258.
REFERENCES 1. Kovesdi, I., Reichel, R., and Nevins, J. R. (1986) Cell 45, 219– 228.
28. Johnson, D. G., Ohtani, K., and Nevins, J. R. (1994) Genes Dev. 8, 1514–1525.
591
AID
BBRC 8010
/
6946$$$383
01-07-98 09:34:11
bbrcg
AP: BBRC
Vol. 242, No. 3, 1998
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
29. Neuman, E., Flemington, E. K., Sellers, W. R., and Kaelin, W. G., Jr. (1994) Mol. Cell. Biol. 14, 6607–6615. 30. Lam, EW-F., and Watson, R. J. (1993) EMBO J. 12, 2705–2713.
46.
31. Johnson, D. G., Schwarz, J. K., Cress, W. D., and Nevins, J. R. (1993) Nature 365, 349–352.
47.
32. Qin, X-Q., Livingston, D. M., Kaelin, W. G., Jr., and Adams, P. (1994) Proc. Natl. Acad. Sci. USA 91, 10918–10922.
48.
33. Shan, B., and Lee, W-H. (1994) Mol. Cell. Biol. 14, 8166–8173.
49.
34. Singh, P., Wong, S., and Hong, W. (1994) EMBO J. 13, 3329– 3338.
50.
35. Johnson, D. G., Cress, W., Jakoi, L., and Nevins, J. R. (1994) Proc. Natl. Acad. Sci. USA 91, 12823–12827.
51.
36. Xu, G., Livingston, D. M., and Krek, W. (1995) Proc. Natl. Acad. Sci. USA 92, 1357–1361. 37. Bremner, R., Cohen, B. L., Sopta, M., Hamel, P. A., Ingles, C. J., Gallie, B. L., and Phillips, R. A. (1995) Mol. Cell. Biol. 15, 3256– 3265. 38. Flemington, E. K., Speck, S. H., and Kaelin, W. G., Jr. (1993) Proc. Natl. Acad. Sci. USA 90, 6914–6918.
52. 53. 54. 55.
39. Smith, E. J., and Nevins, J. R. (1995) Mol. Cell. Biol. 15, 338– 344. 40. Vairo, G., Livingston, D. M., and Ginsberg, D. (1995) Genes Dev. 9, 869–881. 41. Zamanian, M., and La Thangue, N. B. (1993) Mol. Cell. Biol. 4, 389–396. 42. Zhu, L., van den Heuvel, S., Helin, K., Fattaey, A., Ewen, M., Livingston, D. M., Dyson, N., and Harlow, E. (1993) Genes Dev. 7, 1111–1125. 43. Helin, K., Harlow, E., and Fattaey, A. (1993) Mol. Cell. Biol. 13, 6501–6508. 44. Bagchi, S., Raychaudhuri, P., and Nevins, J. R. (1990) Cell 62, 659–669.
56. 57. 58. 59.
60. 61. 62.
45. Chellappan, S., Kraus, V. B., Kroger, B., Munger, K., Howley,
P. M., Phelps, W. C., and Nevins, J. R. (1992) Proc. Natl. Acad. Sci. USA 89, 4549–4553. Cress, W. D., Johnson, D. G., and Nevins, J. R. (1993) Mol. Cell. Biol. 13, 6314–6325. Jordan, K. L., Haas, A. R., Logan, T. J., and Hall, D. J. (1994) Oncogene 9, 1177–1185. Krek, W., Ewen, M. E., Shirodkar, S., Arany, Z., Kaelin, W. G., Jr., and Livingston, D. M. (1994) Cell 78, 161–172. Vaishnav, Y. N., Vaishnav, M., and Wong-Staal, F. (1991) New Biologist 3, 142–150. Vinson, C. R., LaMarco, K. L., Johnson, P. F., Landschulz, W. H., and McKnight, S. L. (1988) Genes Dev. 2, 801–806. Ausbel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A., and Struhl, K. (1989) Current Protocols in Molecular Biology, Wiley, New York. Jameel, S., Zafrullah, M., Ozdener, M. H., and Panda, S. K. (1996) J. Virol. 70, 207–216. Sadowski, I., and Ptashne, M. (1989) Nucleic Acids Res. 17, 7539. Sadowski, I., Ma, J., Triezenberg, S., and Ptashne, M. (1988) Nature 335, 563–564. Vasavada, H. A., Ganguly, S., Germino, F. J., Wang, Z. X., and Weissman, S. M. (1991) Proc. Natl. Acad. Sci. USA 88, 10686– 10690. Cress, W. D., and Nevins, J. R. (1994) J. Virol. 68, 4212–4219. Reichlin, M. (1980) Methods Enzymol. 70, 159–165. Vojtesk, B., Bartek, J., Midgley, C. A., and Lane, D. P. (1992) J. Immunol. Methods 151, 237–244. Harlow, E., and Lane, D. (1988) Antibodies: A Laboratory Manual, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Schuler, G. D., Altschul, S. F., and Lipman, D. J. (1991) Proteins Structure Function Genetics 9, 180–190. Karlseder, J., Rothender, H., and Wintersberger, E. (1996) Mol. Cell. Biol. 16, 1659–1667. Lukas, J., Petersen, B. O., Holm, K., Bartek, J., and Helin, K. (1996) Mol. Cell. Biol. 16, 1047–1057.
592
AID
BBRC 8010
/
6946$$$383
01-07-98 09:34:11
bbrcg
AP: BBRC