Inhibition of cell proliferation and induction of apoptosis by ... - Nature

2 downloads 0 Views 333KB Size Report
et al., 1995; Morkel et al., 1997; Cartwright et al., 1998;. Gaubatz et al., 1998; ..... kindly provided by Dr John Doorbar (Cambridge, UK) and a. 15-mer random ...
Oncogene (2003) 22, 4943–4952

& 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00 www.nature.com/onc

Inhibition of cell proliferation and induction of apoptosis by novel tetravalent peptides inhibiting DNA binding of E2F Silvia Montigiani1,2, Rolf Mu¨ller*,1 and Roland E Kontermann1,3 1 Institute of Molecular Biology and Tumor Research (IMT), Philipps-University, Emil-Mannkopff-Strasse 2, D-35033 Marburg, Germany

We have isolated several peptides from random peptide phage display libraries that specifically recognize the cell cycle regulatory transcription factor E2F and inhibit DNA binding of E2F/DP heterodimers (E2F-1, E2F-2, E2F-3, E2F-4 or E2F–5, and DP-1). The inhibitory efficiency could be strongly enhanced by generating branched tetravalent molecules. To analyse the biological consequences of peptide-mediated E2F inhibition, we fused two of these branched molecules to a cell-penetrating peptide derived from the HTV-Tat protein. Incubation of human tumor cells with these branched Tat-containing peptides led to an inhibition of cell proliferation and induction of apoptosis. These results provide new insights into the function of E2F and further validate E2F as a potential therapeutic target in proliferative diseases. Oncogene (2003) 22, 4943–4952. doi:10.1038/sj.onc.1206495 Keywords: E2F; phage display; inhibitor of DNA binding; peptidomimetics; tat; cell cycle; apoptosis Introduction The E2F family of transcription factors is a key factor in the regulation of cell proliferation (Zwicker and Mu¨ller, 1996; Muller and Helin, 2000; Nevins, 2001; Phillips and Vousden, 2001; Loughran and La Thangue, 2002). E2F transcription factors are heterodimeric proteins consisting of one member of the E2F family and one member of the DP family. As of yet, six members of the E2F family and three members of the DP family of proteins have been identified (Helin et al., 1992; Ivey-Hoile et al., 1993; Lees et al., 1993; Beijersbergen et al., 1994; Buck et al., 1995; Morkel et al., 1997; Cartwright et al., 1998; Gaubatz et al., 1998; Trimarchi et al., 1998). DNA binding and transcriptional activity require heterodimerization of one E2F family member with one of the DP proteins (Bandara et al., 1993; Helin et al., 1993; Krek et al., 1993). Highly conserved DNA binding and heterodimerization domains have been identified in all E2F and DP members (Sardet et al., 1995).

E2F activity is regulated by members of the retinoblastoma family proteins, which include pRB and the related proteins p107 and p130 (Helin et al., 1992; Kaelin et al., 1992; Sardet et al., 1995; Vairo et al., 1995). The E2F family members interact with specific proteins of the pRB family (pocket proteins) through a small region located near the carboxy terminus of E2F (amino acids (aa) 409–426 in E2F-1). Since the pRB family binding site is embedded within the carboxy-terminal transactivation domain, the binding of a pocket protein to E2F prevents or represses E2F-mediated transactivation. The interaction with pocket proteins results in the repression of E2F-dependent transactivation in early G1 phase of the cell cycle. In contrast, ‘free’ and transcriptionally active E2F prevails as cells proceed from G1 into S phase of the cell cycle (Buchkovich et al., 1989; Chen et al., 1989; DeCaprio et al., 1989; Chellappan et al., 1991; Mittnacht and Weinberg, 1991; Zwicker and Mu¨ller, 1996). The release of E2F/DP from complexes with pRB family members and the consequent accumulation of active E2F is the result of pocket protein phosphorylation by G1 cyclin-dependent kinases. The central role played by E2F and its correlation with cell proliferation and tumorigenesis makes it a promising target to control cellular proliferation. In the present study, we have selected peptides that specifically inhibit DNA binding of E2F from two random peptide phage libraries. We also show that the inhibitory potential of the peptides can be dramatically improved by fusion to a branched scaffold (Tarn and Zavala, 1989) generating tetravalent molecules. For the analysis of the peptides on cell growth and cell viability, these branched molecules were fused to the cell-penetrating region of HIV-Tat (Fawell et al., 1994). We show that incubation of cells with these fusion molecules leads to an inhibition of cell proliferation and the induction of apoptosis, thus providing new insights into the biology of E2F. The peptides described here might also be useful tools to further elucidate the potential of E2F as a therapeutic target. Results

*Correspondence: Dr R Mu¨ller, E-mail: [email protected] 2 Current address: Chiron S.p.A., Via Fiorentina 1, 53100 Siena, Italy 3 Current address: Vectron Therapeutics AG, Rudolf-Breitscheid-Str. 24,35037 Marburg, Germany Received 18 October 2002; revised 13 February 2003; accepted 14 February 2003

Isolation and characterization of peptides binding to E2F-1 A recombinant fragment of E2F-1 (E2F1-N1; aa 64–301) containing the DNA-binding and dimerization

E2F inhibiting peptides S Montigiani et al

4944

domain was used to select peptides from two random peptide phage display libraries composed of 6-mers (Doorbar and Winter, 1994) or 15-mers (Scott and Smith, 1990). After three rounds of selection, three different clones were identified from the 6-mer library and two clones from the 15-mer library (Table 1). The selected phage bound specifically to E2F1-N1, while no binding was observed with various control proteins (Figure 1). Table 1 Selection of peptides directed against E2F1. Name

Sequence

Round of selection

Frequency

D6-8 D6-10 D6-11 S15-3 S15-7

FWLRFT WHFIFW WVRWHF IWLSGLSRGVWVSFP GSRILTFRSGSWYAS

3 3 3 3 3

3 2 1 8 2

The selected hexapeptides showed distinct homologies. All three peptides are rich in aromatic amino acids (W, F) and a positively charged amino acid (R) is present in two peptides (D6-10, D6-11). D6-8 and D6-11 share the motif WHF, while the D6-10 and D6-11 peptides contain a motif WXR, where X is, in both peptides, a hydrophobic residue (L in D6-10 and V in D6-11). In both cases, the motif is followed by an aromatic residue (F, W). Some similarities were also observed between the two 15-mer sequences, which are rich in aromatic amino-acid residues (Table 1). Epitope mapping performed by ELISA using different E2F-1 fragments (E2F1-DD dimerization domain, E2F1-DBD DNA-binding domain) showed that the phage selected from the 6-mer library (D6-8, D6-10, D6-11) recognized the E2F-1 DNA-binding domain. The S15-3 and S15-7 peptides did not show a significant binding for any of the E2F-1 DBD or DD fragments (Figure 1), indicating that binding of the peptides to E2F-1 requires the presence of both the DNA binding as well as dimerization domain. Competition ELISA with soluble GST-DP1 at a concentration of 1 mg/ml did not show any inhibition of phage binding to E2F-N1. Furthermore, GST-DP1 was still able to bind to the immobilized E2F1-N1 fragment incubated with synthetic peptides at a concentration of 1 mm, as detected with an anti-GST antibody (data not shown). Therefore, the peptides do not interfere with heterodimerization of E2F-1 and DP-1. Inhibition of DNA binding The inhibitory potential of the peptides on E2F DNA binding was analysed by EMSA. An initial analysis using a B-myb E2F-binding site as a probe (Lam and Watson, 1993) showed the strongest inhibition of DNA binding for peptides D6-11 and S15-7. These peptides were therefore used for further analysis. The IC50 for D6-11 was B100 mm and for S15-7 B50 mm (Figure 2). Since the peptides do not inhibit heterodimerization of

Figure 1 Epitope mapping using different E2F-1 fragments (E2F-1-N1: aa 64–301, E2F-1-DNA-binding domain (DBD): aa 118–191, E2F1 dimerization domain (DD): aa 191–301). All phage selected from the 6-mer library recognized the N1 fragment and the DBD, the S15-3 phage showed a weak binding for E2F-1-DD. S15-7 phage recognized only the N1 fragment. No binding was observed for all the five phages with various control proteins including the DP1GST fusion protein Oncogene

Figure 2 Inhibition of DNA binding of recombinant E2F-1/DP-1 dimers by synthetic peptides D6-11 (670–0.7 mm) or S15-7 (200– 0.7 mm) in a gel retardation assay. IC50 values were B50 mm for S157 and B100 mm for D6-11. Lane 1 DP1-GST; lane 2 E2F-1-N1, lane 3 E2F-1-N1+DP1–GST

E2F inhibiting peptides S Montigiani et al

4945

E2F-1 and DP-1, the inhibition of DNA binding in these experiments is most likely caused by peptide binding to the DNA-binding domain of E2F-1. Mutational analysis of D6-11 and S15-7 To obtain further evidence for the specificity of peptide– E2F interactions and to identify functionally important residues, peptide D6-11 (WVRWHF) was further analysed by alanine-scanning mutagenesis. This showed that the substitution of Trp1 and Trp4 by alanine resulted in a total loss of function. In addition, Phe6 was found to contribute to the inhibition of E2F. No effects were observed upon substituting of the other amino acids (Figure 3a). Deletion analysis of the S15-7 peptide led to the identification of a core region consisting of residues 3–13 (S15-7/7). This region exhibited B50% of the inhibitory activity of S15-7. A shorter peptide (S15-7/2), lacking three amino acids at the C and N terminus was completely inactive (Figure 3b). The S15-7 peptide was further analysed by alanine mutagenesis at various positions (Figure 3c). These experiments showed that the amino acids Phe7 and Trp12 are essential for complete inhibition, and that Arg8 also contributes to the inhibition. No effects were observed upon mutation of Arg3 or Tyr13 to alanine. Thus in both, the hexamer and the pentadecamer, the aromatic amino acids play a crucial role in inhibition of the DNA-binding activity. That both peptides bind to the same region was shown by competition ELISA (Figure 3d). Incubation of phage displaying D6-11 or SI5-7 with excess amounts of peptides D6-11 or SI5-7 resulted in complete inhibition of phage binding (Figure 3d). Inhibition of E2F DNA binding in nuclear extracts We next investigated the effects of peptides D6-11 and SI5-7 on DNA binding of E2F in nuclear extracts prepared from various cell lines. With extracts from F9 embryonal carcinoma cells and HeLa cells, a specific inhibition by S15-7 was observed at an IC50 of B100 mm (data not shown). No inhibition was observed with D611 at a concentration of 500 mm. Thus, inhibition of E2F binding in nuclear extracts requires a higher peptide concentration compared to the inhibition of recombinant E2F complexes. A similar observation has been made by others with a DP-1-derived peptide inhibiting E2F-DP heterodimerization (Bandara et al., 1997). In an attempt to increase the inhibitory activity, we synthesized branched tetravalent molecules consisting of four or eight copies of the same peptide linked by a multiple antigen peptide (MAP) (Tarn and Zavala, 1989). The MAP system consists of a small core matrix of lysine residues with a- and e-amino groups for anchoring multiple copies of the same or different synthetic peptides. The tetravalent peptides showed B100-fold lower IC50 values relative to the monomeric versions when used in EMSAs with recombinant E2F-1/ DP-1, that is, B100mm for D6-11 and B1 mm for MAP4D6-11 (Figure 4a). No further improvement of the

Figure 3 Mutational analysis of D6-11 and S15-7. (a) Alaninescanning mutagenesis of D6-11. Single amino acids of D6-11 were substituted by alanine residues and analysed in EMSA at concentration of 400 and 100 mm for inhibition of DNA binding of recombinant E2F-1/DP-1. Results are indicated as (++) complete inhibition of DNA binding, (+) weak inhibition, and () no inhibition. (b) S15-7 deletion analysis. Fragments of S15-7 were analysed in band shift experiments for inhibition of DNA binding. A central core region, consisting of aa 3–13 was identified. (c) Alanine-scanning mutagenesis of S15-7. Several positions of S15-7 were subsituted by alanine residues. The peptides were used at concentrations of 100, 30, and 6 mm (d) Inhibition of phage binding to E2F-1-N1 by synthetic peptides. D6-11 or S15-7 phage was incubated with peptides D6-11 or S15-7 at a concentration of 1 mm and binding of the phage to E2F-1-N1 was detected by ELISA. Both peptides inhibited binding of D6-11 or S15-7 phage

functional affinity was seen when a MAP peptide with eight identical peptides (MAP8) was used (data not shown). In nuclear HeLa cell extracts, half-maximal inhibition was seen at 10–30 mm with MAP4-D6-11 and at 10 mm with MAP4-S15-7. No inhibition was observed with unrelated binding sites (e.g. NF-Y) using MAP4D6-11 or MAP4-S15-7 (data not shown) demonstrating that the peptides specifically inhibit DNA binding of E2F. Oncogene

E2F inhibiting peptides S Montigiani et al

4946

Inhibition of DNA binding of all five E2F family members Using stably transfected NIH3T3 cells overexpressing E2F-1, E2F-2, E2F-3, E2F-4 or E2F-5, we observed inhibition of DNA binding of all the five family members with MAP4-D6-11 and MAP4-S15-7 at a concentration of 50 mm (Figure. 4b). In contrast, the monovalent peptides D6-11 and S15-7 showed only partial or no inhibition at a concentration of 300 mm. No inhibition was observed for the MAP peptides under the same conditions using unrelated binding sites like NF-Y (Figure 4c, d). The inhibition of DNA binding of the five E2F members can be explained by their highly conserved DNA-binding domains (Sardet et al., 1995). This is also reflected by the fact that all E2F members bind to the same E2F-binding sites, although individual preference has been described (Tao et al., 1997). Inhibition ofE2F target gene expression in vivo The peptides were tested in further experiments for their biological effects in vivo. For this purpose, the peptides were conjugated to a cell penetrating HIV-Tat-derived peptide (aa 48–60) (Fawell et al., 1994) to be able to assess their effect on the expression of known E2Fregulated genes. As expected, we observed a clearly decreased level of expression of several known E2Fregulated genes upon treatment with 5 mm of MAP-D611-Tat or MAP4-S15-7-Tat for 5 h, including E2F-1, E2F-2, E2F-3, and DHFR, while other genes (GAPDH, c-fos, actin) were not affected (Figure 5). In line with this observation, we also observed an inhibition of E2Fbinding activity in cells treated with MAP4-D6-11-Tat and MAP4-S15-7-Tat peptides (not shown).

Figure 4 Effect of tetramerization on the inhibitory potency of peptides D6-11 and S15-7. (a) Effect on recombinant E2F. Improved inhibition of E2F DNA binding by a branched tetravalent peptide using a multiple antigen peptide (MAP) system, analysed by EMSA using a B-myb probe. A 100-fold increase in inhibitory efficacy was observed for D6-11 in EMSA with recombinant E2F-1/DP-1. The peptides were used at concentrations between 300 and 3 mm for D6-11 and between 30 mm and 300 nm for MAP4-D6-11.(b) Effect on E2F in nuclear extract from stably transfected NIH3T3 cells overexpressing single E2F members. DNA binding of all the E2F members was inhibited by MAP4-D6-11 and MAP4-S15-7. Peptides D6-11 and S157 were used at 300 mm and the peptides MAP4–D6-11 and MAP4-S15-7 at 50 mm EMSA was performed as in panel a. DNA binding using an NF-Y binding probe, which lacks a E2F-binding sequence, is not affected by MAP4-D6-11 (c), or MAP4–S15-7 (d) (, no peptide; wt, + MAP4 peptides at 50 mm)

Oncogene

Figure 5 RT–PCR analysis of gene expression. HL60 cells were incubated for 5 h without peptide (lane 1) or with 5 mm peptides: MAP4-D6-11-Tat (lane 2), MAP4-D6-11(A1, A4)-Tat, MAP4-S157-Tat (lane 4), or MAP4-S15-7(A7, A12)-Tat (lane 5). A reduction of gene expression is observed for E2F-1, E2F-2, E2F-3, and DHFR in the presence of MAP4-D6-11-Tat or MAP4-S15-7-Tat, while no effects were observed for GAPDH, c-fos or actin

E2F inhibiting peptides S Montigiani et al

4947

We also analysed the expression of E2F-1 and cyclin A protein in both peptide-treated and -untreated cells by immunoblotting (Figure 6a). After 6 h treatment, a slight decrease of the amount of E2F-1 was observed in cells treated with the peptides MAP4-D6-11-Tat or MAP4-S15-7-Tat. After incubation for 24 h, the amount of E2F-1 protein was substantially decreased in cells treated with MAP4-D6-11-Tat and very low in cells treated with MAP4-S15-7-Tat. The amount of cyclin A was already reduced with both peptides after treatment for 6 h. No changes in protein amounts of E2F-1 and cyclin A were observed in experiments with the mutant peptides comparable to experiments without peptide (Figure 6a). In agreement with these observations, we found that the activity of a cyclin A promoter-luciferase promoter construct was inhibited by MAP4-D6-11-Tat and MAP4-S15-7-Tat in transiently transfected HeLa cells by B70%, while no effects were observed with the control peptides MAP4-D6-11(A1, A4)-Tat and MAP4S15-7(A7, A12)-Tat (Figure 6b). In addition, no effects were observed with an SV40 promoter-luciferase construct (Figure 6b).

A12)-Tat containing two alanine substitutions at positions 7 and 12 (Figure 7b). The proliferation of three other human tumor cell lines, that is, HeLa and C33A cervical carcinoma cells and U2-OS osteosarcoma cells, was also efficiently blocked by the MAP4-D6-11-Tat and MAP4-S15-7 peptides (IC50 ¼ 3–10 mm), but not by the mutant variants (no significant inhibition up to 100mm)

Inhibition of cell proliferation and induction of apoptosis Next, we analysed the tetravalent Tat-containing peptides for their effect on the cell cycle. HL60 leukemia cells were treated with varying concentrations of MAP4D6-11 -Tat and a control peptide MAP4-D6-11(A1, A4)-Tat containing two alanine substitutions at positions 1 and 4. Proliferation of this cell line was inhibited by the wild-type peptide, but not by the mutant peptide. Growth curves of HL60 cells treated with peptide MAP4-D6-11-Tat or with the control peptide MAP4D6-11(A1, A4)-Tat at concentrations between 1 and 5 mm are shown in Figure 7a. Inhibitory effects became visible after 5–24 h depending on the concentration of the peptide. In a similar experiment, specific inhibition of cell growth was also observed using MAP4-S15-7-Tat at a concentration between 1 and 5 mm, while no effects were observed with the same concentrations of control peptide MAP4-S15-7(A7, –––––––––––––––––––––––––––––––––––––––––––––––––––––––––––" Figure 6 (a) Inhibition of endogenous cyclin A and E2F-1 expression. Cells treated without peptide (lane 1), with MAP4-D6-11-Tat (lane 2), or MAP4-S15-7-Tat (lane 4) were analysed by immunoblotting for expression levels of cyclin A or E2F-1. A decrease of E2F-1 protein and cyclin A protein was observed after 24 h treatment with the peptides MAP4-D6-11-Tat and MAP4S15-7-Tat, while no effect was observed in cells treated with control peptides (lanes 3 and 5). An antibody directed against b-actin was used as an internal standard. (b) Specific inhibition of transcription from the cyclin A promoter in HeLa cells transiently transfected with a CycA–luciferase construct and treated with 5 mm MAP4-D6-11-Tat. No effect was observed with control peptide MAP4-D6-11(A1, A4)-Tat at the same concentration or with a SV40 promoter construct. (c) Specific inhibition of transcription from the cyclin A promoter in HeLa cells transiently transfected with a CycA–luciferase construct and treated with MAP4S15-7-Tat. No effect was observed with control peptide MAP4-S157(A7, A12)-Tat at the same concentration or with a SV40 promoter construct. Cell extracts were standardized for protein concentration Oncogene

E2F inhibiting peptides S Montigiani et al

4948

Figure 8 Induction of apoptosis by E2F-inhibiting peptides. Characteristic internuclosomal DNA cleavage (DNA laddering) was observed in HL60 cells treated with 5 mm MAP4-D6-11-Tat (a) or MAP4-S15-7-Tat (b). No effects were observed by treating cells with mutant peptides at the same concentration

Figure 7 Inhibition of cell proliferation. Inhibition of proliferation was observed after 5–24 h of treatment with peptides MAP4D6-11-Tat (a) and MAP4S15-7-Tat (b) in a concentrationdependent manner. No effects were observed on treating cells with control peptides MAP4-D6-11(A1, A4)-Tat and MAP4-S15-7(A7, A12)-Tat

In addition to a diminished proliferation, HL60 cells treated with MAP4-D6-11-Tat showed chromatin condensation and nuclear fragmentation as evidenced by internucleosomal DNA cleavage (‘DNA laddering’; Figure 8a,b) and Hoechst 33342/propidium iodide staining (not shown). No effects were observed with the MAP4-D6-11(A1,A4)-Tat control peptide. Analogous findings were made with HL60 cells treated with MAP4-S15-7-Tat and MAP4-S15-7(A7, A12)-Tat peptides (not shown). HL60 cells treated for 20 h with peptides at 5 mm were further analysed by FACS analysis after staining with propidium iodide (Endresen et al., 1995). Cell populations treated with the two peptides MAP4-D6-11-Tat and MAP4-S15-7-Tat contained 35 and 68% of apoptotic cells with a sub-G1 DNA content, respectively (Figure 9), while no increase in the levels of apoptotic cells was seen in the absence of peptides (not shown). Analysis of the cell cycle profiles indicated an S phase arrest of the cells treated with MAP-D6-11-Tat and MAP4-S15-7-Tat and the presence of cells in sub-G1. In contrasts, mutant peptides did not alter the cell cycle profile (Figure 9). Oncogene

Figure 9 Cell cycle profiles of HL60 cells incubated without peptide, with MAP4-D6-11-Tat, MAP4-S15-7-Tat, MAP4-D611(A1,A4)-Tat, or MAP4-S15-7(A7,A12)-Tat. The data show an S phase arrest and an increased population of apoptotic cells in sub-G1 after treatment with the wild-type peptides

Discussion In the present study, we have isolated from two random phage display libraries two different peptides that

E2F inhibiting peptides S Montigiani et al

4949

specifically interfere with the function of the transcription factor E2F. We demonstrate that these peptides specifically interact with the DNA-binding domain of E2F family members without interfering with DP-1 heterodimerization, and thereby block DNA binding of E2F–DP dimers, both in vitro with recombinant proteins and in vivo. The initially low inhibitory activity of the peptides was increased up to 100-fold by generating small branched molecules carrying four copies of the same peptide linked to a central lysyl core (MAP) (Tam and Zavala, 1989). In other studies, MAP peptides have been demonstrated to be efficient immunogens (Ahlborg, 1995) because of an increase in functional affinity (Tam and Zavala, 1989). Here, we have extended the applications of MAP peptides to improve the inhibitory activity of DNA-binding peptides and confirm that multimeric peptides directed against the same target exhibit a gain in functional affinity. The peptides were conjugated to a HIV-Tat-derived membrane-penetrating peptide known to translocate through the plasma membrane and to localize to the nucleus. A Tat fragment encompassing aa 48–60 and covering the complete basic domain retains full translocation activity and is able to direct nuclear localization (Vives et al., 1997). The observations reported here indicate that the Tat fragment fused to the C terminal of the MAP core is capable of delivering these branched molecules into cells. This is strongly suggested by the ability of the MAP4-D6-11-Tat and MAP4-S15-7-Tat peptides to reduce the expression of E2F-regulated genes, to inhibit cell proliferation and to induce apoptosis. Our work confirms and extents previous observations that inhibition of E2F activity leads to a perturbation of cell cycle progression. Thus, an E2F-1 dominant-negative mutant lacking the DP-binding site inhibited both the transcriptional activity of E2F and cell proliferation (Fan and Bertino, 1997). Likewise, a dominant-negative DP-1 mutant that forms inactive heterodimers with E2F proteins has been reported to induce a cell cycle arrest in G1 (Wu et al., 1996), and RNA aptamers that specifically blocked E2F DNA binding also inhibited cell proliferation after microinjection into primary fibroblasts (Ishizaki et al., 1996). Apart from its essential function in cell cycle progression, a role for E2F in regulating apoptosis is well documented (Krek et al., 1995; Phillips and Vousden, 2001). The physiological relevance of E2F controlled in promoting cell death has been clearly demonstrated by the development of spontaneous tumors in E2F-1 null mice, indicating that E2F-1 can act as a tumor suppressor gene (Field et al., 1996; Yamasaki et al., 1996). This function of E2F-1 is in part because of its ability to activate proapoptotic genes, such as p73 and Apaf-1 (Irwin et al., 2000; Stiewe and Putzer, 2000; Furukawa et al., 2002). The observations made in the present study are in apparent contrast to this proapoptotic function of E2F-1, but is in line with several other reports in the literature.

Thus, a short DP-1-derived peptide linked to penetratin to enable cellular translocation inhibited E2F–DP heterodimerization and caused apoptosis of tumor cell lines (Bandara et al., 1997). More recently, it has been described that E2F activity is essential for the survival of Myc-overexpressing tumor cells regardless of their ARF and p53 status (Santoni-Rugiu et al., 2002). It would thus appear that E2F can exert opposing functions with respect to the regulation of apoptosis, and that the precise outcome of E2F inhibition may depend on the genetic status of the cell and the loss-of-function approach, which may target either specific E2F genes (as in knockout mice) or multiple E2F members (as in the present study). Unexpectedly, the E2F inhibitory MAP4-D6-11-Tat and MAP4-S15-7-Tat peptides induced a cell cycle arrest in S phase. This is in apparent contrast to a previous report by Fabbrizio et al. (1999) where an E2F inhibitory peptide aptamer was shown to induce a G1 arrest. We do not know the precise reason for this discrepancy, but several explanations are possible. First, Fabbrizio et al. (1999) tested their peptide with normal fibroblasts, which may respond differently than the cancer cell used in the present study. Second, the aptamer was not introduced in the cells as a peptide as in our study, but was expressed from a transfected plasmid, which may affect intracellular levels and distribution of the peptide. Third, it is possible that different affinities and intracellular concentrations may produce a different biological outcome, that is, arrest in G1 or apoptosis following entry into S phase. These considerations would be compatible with a model, where a residual low level of E2F, as seen after treatment with our peptide anatagonists (Figures 5 and 6), would still allow for a transition from G1 to S, but would be insufficient for the completion of S phase, the latter being because of the lack of an E2F-mediated antiapoptotic pathway. The nature of such a pathway remains enigmatic at present, but the existence of an E2F-mediated antiapoptotic mechanism is also suggested by other studies (Santoni-Rugiu et al., 2002). It may be possible that the inhibition of cyclin A by our peptide antagonists (Figure 6) plays a role in this context, since cyclin A may be essential for constraining the activity of E2F after S phase entry (Krek et al., 1995). In summary, we have isolated from random phage display libraries two different peptides that inhibit DNA binding of E2F. The inhibitory potential of the peptides could be dramatically increased by the generation of tetravalent molecules. Fusion of these branched molecules to a Tat-derived peptide enabled the delivery of these molecules into mammalian cells and an analysis of the biological consequences, most notably inhibition of cell proliferation and apoptosis. The reagents described in the present study should represent invaluable tools for E2F-related research and might be useful leads for the development of therapeutic peptidomimetics for the treatment of proliferative diseases. Oncogene

E2F inhibiting peptides S Montigiani et al

4950 Materials and methods

drogen proxide as a substrate. DNA of individual positive phage was sequenced using fdSeq1 as a primer.

Materials Plasmids encoding a N-terminal fragment of human E2F-1 (E2F1-N1; aa 64–310), E2F-5 (aa 1–220), and GST-DP-1 (aa 84–249) were a kind gift of Dr Lan Bandara (Prolifix Ltd, UK). A random 6-mer peptide phage display library was kindly provided by Dr John Doorbar (Cambridge, UK) and a 15-mer random peptide library by Dr George P Smith (MI, USA).

Peptide synthesis

Olignucleotides

ELISA

For the cloning of E2F-1-DD (aa 191–301): E2F-1-DD back: 50 -TAT ACC ATG GGT GTG AAA TCC CCG GGG GAG AAG TTC A-30 and E2F-1-DD for: 50 -CTC GAG TGC GGA CGC CAG CCA CTG GAT GTG GTT CTT-30 . For the cloning of E2F-1–DBD (aa 118–191): E2F-1-DBD back: 50 -TAT ACC ATG GGC AGC CAC ACC ACA GTG GGC GTC GGC-30 and E2F-1-DBD for: 50 -CTC GAG TGC GGC CGC CTC CTC AGG GCA CAG GAA AAC-30 . For fdSeq1 see Griffiths et al. (1994).

The inhibition of binding of the peptides displayed on phage by the free peptide was analysed by phage-ELISA as described above, adding free peptides at 1 mm to the phage supernatants. The competition phage ELISA was performed incubating phage supernatant and DP1-GST at 1 mg/ml in PBS, 2% skimmed milk powder for 2 h at RT. Phage binding was detected using a horseradish peroxidase-conjugated anti-M13 antibody as described above. The competition peptide ELISA was performed by preincubating coated E2F-1 with the synthetic peptides at 1 mm in PBS, 2% skimmed milk powder for 30 min.After addition of DP1-GST at 1 mg/ml, the plate was incubated for two additional hours at RT. DP-1 binding was detected using an anti-GST-mouse antibody (Santa Cruz Biotechnology) and a horseradish peroxidase-conjugated antimouse antibody (Sigma).

Protein expression and purification Fragments encoding the dimerization domain of E2F-1 (E2F1–DD; aa 191–301) or the DNA-binding domain of E2F-1 (E2F-1-DBD; aa 118–191) were amplified by PCR using the above-indicated primers, digested with NcoI and NotI and cloned into bacterial expression vector pET28b (Novagen). Recombinant proteins carrying hexahistidyl tags were expressed in BL21 DE3 cells (Novagen). After induction of expression with 1 mm IPTG, cells were grown at 201C for additional 3 h or overnight and proteins were purified by immobilized metal affinity chromatography (IMAC). Cells were centrifuged at 4000 r.p.m. for 20 min and the pellet was resuspended in IMAC loading buffer (50 mm phosphate buffer pH 7.5, 500 mm NaCl, 20 mm imidazole). Resuspended pellet was sonicated three times for 1 min on ice and centrifuged at 12 000 r.p.m. for 20 min.Supernatant was loaded onto a NiNTA column (Quiagen) equilibrated with loading buffer. The column was washed with IMAC washing buffer (50 mm phosphate buffer pH 7.5, 500 mm NaCl, 35 mm imidazole). Proteins were eluted with IMAC elution buffers (50 mm phosphate buffer pH 7.5, 500 mm NaCl) containing various amount of imidazole (0.1–1 m imidazole). DPI-GST was expressed and purified as described (Smith and Johnson, 1988).

Peptides were synthesized on a solid phase using a multiple peptide synthesizer (Applied Biosystems, Peptide Synthesizer 432 A, Synergy) employing Fmoc/t-butyl protecting groups (Fields and Noble, 1990). MAP peptides and MAP–peptide– Tat fusion molecules were synthesized by Dr G Bloomberg (University of Bristol, UK).

DNA-binding assay Purified E2F1-N1 (64–301) and DP1-GST (84–249) were incubated at 10 mg/ml in a buffer containing 10 mm HEPES, 100 mm KCl, 1 mm EDTA, 1 mm DTT, 33 mg/ml BSA, and 0.1 mg. of sonicated salmon sperm DNA for 10 min on ice (Liu et al., 1997). Nuclear extracts (3 mg) were incubated in 20 ml buffer containing 50 mm Tris pH 7.9, 0.2 mm EDTA, 1 mm DTT, 15% glycerol, 1 mg of sonicated salmon sperm DNA on ice, then for 10 min at 301C and after that cooled on ice. 32Plabelled probe was added, the reaction mixtures were incubated for another 10 min at 301C, and then cooled on ice. Probes were labelled by filling in overhanging ends of 4–7 bases. Samples were run on 4% nondenaturing polyacrylamide gel in 0.5% TBE at 41C. Gels were exposed to X-ray films. DNA probes: B-myb: 50 -CGA CGC GCT TGG CGG GAG ATA GAA AAG TGC-30 ; NF-Y: 50 -GAT TTT TCT GAT TGG TTA AAA GT-30 ; Sp1: 50 - ATG GGG CGG AGA-30 ; ATF: 50 - CGC CTT GAA TGA CGT CAA GGC CGC GA-30 .

Phage selection A random 6-mer peptide phage library (Doorbar and Winter, 1994) and a random 15-mer peptide phage library (Scott and Smith, 1990) were used for selections. Phage were selected by panning on immunotubes (Maxisorb, Nunc) coated with recombinant E2F-1 at 10 mg/ml. Selections were performed essentially as described (Griffiths et al., 1994). Positive clones were identified by screening phage produced from single colonies for specific binding to E2F-1-N1 in ELISA (phage ELISA). An ELISA plate was coated overnight at 41C with E2F-1 at 10 mg/ml in 50 mm carbonate buffer pH 9.6. After blocking the plate with PBS, 2% skimmed milk powder for 2 h at RT, 50 ml of phage supernatant was added to 50 ml of PBS, 4% skimmed milk powder and incubated for 2 h at RT. The plate was then washed three times using PBS–0.1% Tween-20 and three times with PBS. Phage binding was detected using horseradish peroxidase-conjugated anti-M13 antibody (Pharmacia) with 30 ,30 ,50 ,50 -tetramethylbenzidine (Pierce) and hyOncogene

Preparation of nuclear extracts Nuclear extracts were prepared from HeLa cells as described (Dignam et al, 1983), except that the dialysis step at the end of the procedure was omitted. Nuclear extracts were prepared from NIH3T3 cells stably overexpressing individual E2F family members. Harvested cells were resuspended in 20 mm HEPES pH 7.8, 450 mm NaCl, 25% glycerol, 0.2 mm EDTA, 0.5 mm PMSF, 0.5 mg/ml leupeptin, 0.5 mg/ml aprotinin. The cells were frozen and thawed three times, centrifuged again and the supernatant was collected. F9 cell nuclear extract was a kind gift of Dr Lan Bandara (Prolifix Ltd, UK). Immunoblot analysis Cell extracts were prepared as described above and analysed by SDS–PAGE and immunoblot analysis as described in Delia et al. (1997). The reagent used were monoclonal

E2F inhibiting peptides S Montigiani et al

4951 antibodies recognizing E2F-1 (C20; Santa Cruz), cyclin A (H432; Santa Cruz), or actin (C4; Boehringer, Mannheim, Germany). Transient transfections HeLa cells were transfected with 2 mg of plasmid CycA-pGL3 containing the luciferase gene under the control of the human cyclin A promoter or with 2 mg of plasmid SV40 pGL3 containing the luceriferase gene under the control of the SV40 promoter (Zwicker et al, 1995). Transfections were performed using Superfectt (Qiagen) according to the manufacturer’s protocol. After 24 h, cells were incubated with peptides at a concentration of 5 mm for 7 h at 371C. Cells were then analysed for luciferase activity as described (Zwicker et al, 1995). Cell extracts were standardized for protein concentration to eliminate any differences in luciferease activity related to cell growth (consequently SV40 pGL3 gives similar luciferase activities in all samples tested).

Agarose gel electrophoresis of genomic DMA Treated cells were collected, washed with PBS, resuspended in lysis buffer (50 mm Tris-HCl pH 7.9,100 mm EDTA, pH 8, 100 mm NaCl, 1% SDS) and 500 mg/ml proteinase K and incubated at 551C for 18 h. Cells were then incubated with 300 mg/ml RNase A for 1 h at 371C, genomic DNA was purified and examined by 1% agarose gel electrophoresis (Herrmann et al., 1994). FACS analysis Cells were treated without peptide or with peptides at a concentration of 5 mm for 20 h. Cells were then fixed with 1% paraformaldehyde for 30 min at 41C. Cells were transferred into a reaction tube, washed with PBS and resuspended in 500 ml PBS containing 50 mg/ml propidium iodide and 100 mg/ ml RNase, and finally incubated for 30 min at 371C. FACS analysis was performed using a FACStar Plus (Becton Dickinson). RT–PCR

Measurement of cell proliferation Cells were maintained at 371C and 5% CO2 in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal calf serum (FCS), except for HL60 cells, which were cultured in RPMI 1640 with 10% FCS. For peptide treatment, about 2  106 cells were plated in a 3 cm dish and cultured overnight. Tat-conjugated peptides were added to the medium at a concentration of 5 mm. Cells were collected by mild centrifugation, washed with PBS and stained with Hoechst 33342 and PI (10 mm) to evalutate the nuclear morphology. Cell proliferation assays were performed as described above treating the cells with peptide MAP4-D6-11-Tat, MAP4-D6-11 (A1 ,A4)-Tat, MAP4-S15-7-Tat, and MAP4-S15-7(A7,A12)Tat at concentrations 1–5 mm. Cell numbers were counted after various times.

Cells were incubated for 5 h without peptide or with 5 mm of wild-type or mutant peptides. Treated cells were collected and RNA was prepared using a RNAeasy kit (Qiagen). cDNA was synthesized from 4 ml of RNA with MM LV reverse transcriptase (Gibco) and oligo(dT) primer. PCR was performed with 1 ml of cDNA and primers specific for E2F-1, E2F-2, E2F3, DHFR, GAPDH, c-fos, or actin.

Acknowledgements We thank Marylou Zuzarte for FACS analyses and Dr M Krause (IMT) for oligonucleotide and peptide synthesis. This work was supported by a grant from the Deutsche Forschungsgemeinschaft to REK (SFB 397) and the Dr Mildred Scheel Stiftung fu¨r Krebsforschung to RM.

References Ahlborg N. (1995). J. Immunol. Methods, 179, 269–275. Bandara LR, Buck VM, Zamanian M, Johnston LH and La Thangue NB. (1993). EMBO J., 12, 4317–4324. Bandara LR, Girling R and La Thangue NB. (1997). Nat. Biotechnol., 15, 896–901. Beijersbergen RL, Kerkhoven RM, Zhu L, Carlee L, Voorhoeve PM and Bernards R. (1994). Genes Dev., 8424, 2680–2690. Buchkovich K, Duffy LA and Harlow E. (1989). Cell, 58, 1097–1105. Buck V, Allen KE, Sorensen T, Bybee A, Hijmans EM, Voorhoeve PM, Bernards R and La TNB. (1995). Oncogene, 11, 31–38. Cartwright P, Muller H, Wagener C, Holm K and Helin K. (1998). Oncogene, 17, 611–623. Chellappan SP, Hiebert S, Mudryj M, Horowitz JM and Nevins JR. (1991). Cell, 65, 1053–1061. Chen PL, Scully P, Shew JY, Wang JYJ and Lee WH. (1989). Cell, 58, 1193–1198. Delia D, Goi K, Mizutani S, Yamada T, Aiello A, Fontanella E, Lamorte G, Iwata S, Ishioka C, Krajewski S, Reed JC and Pierotti MA. (1997). Oncogene, 14, 2137–2147. DeCaprio JA, Ludlow JW, Lynch D, Furukawa Y, Griffin J, Piwnica-Worms, H, Huang CM and Livingston DM. (1989). Cell, 58, 1085–1095.

Dignam JD, Lebovitz RM and Roeder RG. (1983). Nucl. Acids Res., 11, 1475–1489. Doorbar J and Winter G. (1994). J. Mol. Biol., 244, 361–369. Endresen PC, Prytz PS and Aarbakke J. (1995). Cytometry, 20, 162–171. Fabbrizio E, Le Cam L, Polanowska J, Kaczorek M, Lamb N, Brent R and Sardet C. (1999). Oncogene, 18, 4357–4363. Fan J and Bertino JR. (1997). Oncogene, 14, 1191–1200. Fawell S, Seery J, Daikh Y, Moore C, Chen LL, Pepinsky B and Barsoum J. (1994). Proc. Natl. Acad. Sci. USA, 91, 664–668. Field SJ, Tsai FY, Kuo F, Zubiaga AM, Kaelin Jr WG, Livingston DM, Orkin SH and Greenberg ME. (1996). Cell, 85, 549–561. Fields GB and Noble RL. (1990). Int. J. Pept. Protein Res., 35, 161–214. Furukawa Y, Nishimura N, Satoh M Endo H, Iwase S, Yamada H, Matsuda M, Kano Y and Nakamura M. (2002). J. Biol. Chem., 30, 30. Gaubatz S, Wood JG and Livingston DM. (1998). Proc. Natl. Acad. Sci. USA, 95, 9190–9195. Griffiths AD, Williams SC, Hartley O, Tomlinson IM, Waterhouse P, Crosby WL, Kontermann RE, Jones PT, Low NM, Allison TJ. (1994). EMBO J., 13, 3245–3260. Oncogene

E2F inhibiting peptides S Montigiani et al

4952 Helin K, Lees JA, Vidal M, Dyson N, Harlow E and Fattaey A. (1992). Cell, 70, 337–350. Helin K, Wu CL, Fattaey AR, Lees JA, Dynlacht BD, Ngwu C and Harlow E. (1993). Genes Dev., 7, 1850–1861. Herrmann M, Lorenz HM, Voll R, Grunke M, Woith W and Kalden JR. (1994). Nucleic Acids Res., 22, 5506–5507. Irwin M, Marin MC, Phillips AC, Seelan RS, Smith DI, Liu W, Flores ER, Tsai KY, Jacks T, Vousden KH and Kaelin Jr WG (2000). Nature, 407, 645–648. Ishizaki J, Nevins JR and Sullenger BA. (1996). Nat. Med., 2, 1386–1389. Ivey-Hoile M, Conroy R, Huber HE, Goodhart PJ, Oliff A and Heimbrook D.C. (1993). Mol. Cell. Biol., 13, 7802–7812. Kaelin WGJ, Krek W, Sellers WR, DeCaprio JA, Ajchenbaum F, Fuchs CS, Chittenden T, Li Y, Farnham PJ, Blanar MA, Livingston DM and Flemington EK. (1992). Cell, 70, 351– 364. Krek W, Livingston DM and Shirodkar S (1993). Science, 262, 1557–1560. Krek W, Xu G and Livingston DM. (1995). Cell, 83, 1149– 1158. Lam EW and Watson RJ. (1993). EMBO J., 12, 2705–2713. Lees JA, Saito M, Vidal M, Valentine M, Look T, Harlow E, Dyson N and Helin K. (1993). Mol. Cell. Biol., 13, 7813– 7825. Liu N, Lucibello FC, Ko¨rner K, Wolfraim LA, Zwicker J and Mu¨ller R. (1997). Nucl. Acids Res., 25, 4915–4920. Loughran O and La Thangue NB. (2002). Curr. Biol., 12, R377. Mittnacht S and Weinberg RA. (1991). Cell, 65, 381–393. Morkel M, Wenkel J, Bannister AJ, Kouzarides T and Hagemeier C. (1997). Nature, 390, 567–568.

Oncogene

Muller H and Helin K. (2000). Biochim. Biophys. Acta, 1470, M1–M12. Nevins JR. (2001). Hum. Mol. Genet., 10, 699–703. Phillips AC and Vousden KH. (2001). Apoptosis, 6, 173–182. Santoni-Rugiu E, Duro D, Farkas T, Mathiasen IS, Jaattela M, Bartek J and Lukas J. (2002). Oncogene, 21, 6498–6509. Sardet C, Vidal M, Cobrinik D, Geng Y, Onufryk C, Chen A and Weinberg RA. (1995). Proc. Natl. Acad. Sci. USA, 92, 2403–2407. Scott JK and Smith GP. (1990). Science, 249, 386–390. Smith DB and Johnson KS. (1988). Gene, 67, 31–40. Stiewe T and Putzer BM. (2000). Nat. Genet., 26, 464–469. Tam JP and Zavala F. (1989). J. Immunol. Methods, 124, 53– 61. Tao Y, Kassatly RF, Cress WD and Horowitz JM. (1997). Mol. Cell. Biol., 17, 6994–7007. Trimarchi JM, Fairchild B, Verona R, Moberg K, Andon N and Lees JA. (1998). Proc. Natl. Acad. Sci. USA, 95, 2850– 2855. Vairo G, Livingston DM and Ginsberg D. (1995). Genes Dev., 9, 869–881. Vives E, Brodin P and Lebleu B. (1997). J. Biol. Chem., 272, 16010–16017. Wu CL, Classon M, Dyson N and Harlow E. (1996). Mol. Cell. Biol., 16, 3698–3706. Yamasaki L, Jacks T, Branson R, Goillot E, Harlow E and Dyson NJ. (1996). Cell, 85, 537–548. Zwicker J, Lucibello FC, Wolfraim LA, Gross C, Truss M, Engeland K and Mu¨ller R. (1995). EMBO J., 14, 4514–4522. Zwicker J and Mu¨ller R. (1996). Progress in Cell Cycle Research, Vol. 1. Meijer L, Guidet S, Tung HYL (eds). Plenum Press: New York, pp. 91–99.