ABSTRACT
Studies on receptor-ligand interactions are important for the design of agonists or antagonists of natural ligands. We developed a luciferase reporter assay to screen epidermal growth factor receptor (EGFR) binding molecules rapidly for their ability to stimulate or inhibit signal transduction. Human EGF displayed on fd filamentous phage presented an activity similar to soluble EGF when tested for binding to the EGFR, for induction of cell cycle progression or in the luciferase assay. Two libraries of human EGF variants displayed on phage were constructed in which the aspartic acid residue at position 46 or the arginine residue at position 41 were randomised. EGF mutants displayed on phage were screened in parallel for binding to the EGFR using an ELISA assay and for transducing activity using the luciferase assay. Regarding the 46 position, most of the mutants retained the ability to bind the EGFR and their transducing activity corresponded perfectly with their binding. For the more crucial 41 position, only the wild-type EGF was able to bind the EGFR. Our approach allowed a simple determination of crucial positions and paved the way for identification of agonists with altered transduction activity.
Human epidermal growth factor (EGF) is a 53 amino acid peptide which acts as a growth factor on a wide variety of cell types. Binding of the EGF to its membrane receptor results in the activation of intracellular signalling pathways which lead to essential biological functions such as cell growth and differentiation (1 ). EGF binding triggers EGF receptor (EGFR) dimerisation, followed by activation of the receptor's intracellular kinase activity (2 ) and autophosphorylation of the C-terminal ends of the receptor dimer (1 ,3 ). The phosphorylated receptor adopts a conformation which is competent in phosphorylating cellular substrates (4 ). Most of these substrates are kinases involved in different signal transduction pathways (5 ). Following EGF binding, EGF/EGFR complexes are rapidly internalised and enter distinct cellular sorting pathways leading to either degradation or recycling to the cell surface (1 ).
Overexpression or deregulation of the EGFR is thought to play a critical role in several human tumors such as mammary and ovarian carcinoma. The autocrine stimulation of the overexpressed receptor is essential in achieving a transforming effect (6 -8 ). Blocking antibodies, which prevent EGF binding or EGFR autophosphorylation, can inhibit tumor proliferation (9 -11 ). Therefore, the development of novel antagonists has a great clinical potential.
So far, no structure of EGF in complex with its receptor has been determined. However, a model for the receptor binding site was proposed based on the high resolution NMR structure of the free EGF molecule (12 ), sequence comparison and mutational studies (13 -17 ). It was found that several residues were necessary for EGF binding to its receptor: specifically Tyr13, Leu15, His16, Arg41, Gln43 and Leu47.
Phage display technology (18 ) has been extensively used to select specific peptide and protein ligands (19 ). Foreign DNA fragments are inserted into the genome of filamentous phages in fusion with coat protein genes. Through fusion with the minor coat protein g3p, of which three to five copies are displayed on the phage head surface (18 ), the genotype (the specific coding DNA sequence) and phenotype (the foreign protein exposed on the surface) are physically linked. This allows the selection and rescue of rare peptides or proteins with the desired ligand specificity by repeated rounds of affinity purification on immobilized antigens.
In this report, we displayed human EGF on an fd phage surface (phage-EGF), and tested its ability to bind to the EGFR and to induce DNA synthesis, which represents a specific cellular feature of EGFR-induction downstream on the signal transduction pathway. We described an assay for EGFR transduction activity utilising the luciferase gene under control of the c-fos gene enhancer SIE (v-sis inducible element) (20 ,21 ) as a reporter. The test was used to rapidly screen and compare the transduction activity of human EGF mutants displayed on the phage surface.
A synthetic human EGF (22 ) was amplified by PCR with the primers EGFback (5'-GCGGTACCGGCCGACAATTCCGACAGCG-3') and EGFfor (5'-AACTGCAGGCCCCAGAGGCCCCACGCAGTTCCCACC-3') and cloned into the phage fd FUSE5 (23 ) using its EagI/SfiI sites.
A library of phage-EGF randomly mutated at the aspartate (D) 46 position was prepared by PCR amplification of a synthetic human EGF clone (22 ) using the EGFback primer and the degenerate primer D46EGFfor (5'-AACTGCAGGCCCCAGAGGCCCCACGCAGTTCCCACCATTTCAGMNNTCTGTACTGGCAGCG-3'). M indicates a mixture of C and A; N represents any of the four nucleotides. The PCR fragments were cloned into the phage fd fuse5 vector as above.Phages were produced as described in the literature for fd filamentous phages (24 ).
The same strategy was adopted to construct the arginine (R) 41 randomised library using the degenerate primer R41EGFfor (5'-AACTGCAGGCCCCAGAGGCCCCACGCAGTTCCCACCATTTCAGATCTCTGTACTGGCAMNNTTCACCGATGTAGCC-3').
Two complementary oligonucleotides (5'-GATCCCATTTCCCGTAAATCCATTTCCCGTAAATCCATTTCCCGTAAATC-3') and (5'-AGCTTGATTTACGGGAAATGGATTTACGGGAAATGGATTTACGGGAAAT-3'), containing three copies of the c-fos serum inducible element (SIE) (20 ) were cloned into the BamHI and HindIII restriction sites of the pZlucTK vector (25 ) upstream of the luciferase cDNA yielding the vector pZlucTKSIE.The same cloning strategy was used to construct the control vector pKSSIE using a pBluescript II (KS)+ vector.
Monkey COS-7, human carcinoma A431 cells (expressing ~106 EGFR per cell), NIH 3T3 cells, clone 2.2 cells (NIH 3T3 cells lacking endogenous murine EGFR) and its derivative clone 2.2 wt, expressing the human EGFR (~4 * 105 per cell) (26 ), were grown in DMEM supplemented with 10% fetal calf serum (FCS).
For transient transfection, COS-7 cells were cotransfected by lipofection (LipofectAMINEtm, GIBCO) for 5 h, with 1 [mu]g each of pRK5 EGF-R (27 ) and pZlucTKSIE (or pZlucTK as a negative control). Stably transfected 2.2 SIE fibroblasts were obtained by cotransfection of 2.2 wt cells using 6 [mu]g of the luciferase vector pZlucTKSIE and 0.3 [mu]g of the plasmid pTKhygro conferring hygromycin resistance (28 ). The day after transfection, cells were amplified in 100 mm Petri dishes and 50 [mu]g/ml hygromycin (Boehringer Mannheim) was added. After 12-15 days of selection, clones were amplified and tested for luciferase activity.
Subconfluent NIH 3T3 cells were serum starved for 48 h and then growth stimulated for 18 h in the presence of 100 mM bromodeoxyuridine (BrdU). Cells were stimulated upon addition of either soluble EGF (25 ng/ml = 2.5 * 1012 EGF molecules/ml), phage-EGF (5 * 1012 phage displayed EGF molecules/ml) or control phage displaying Fab fragments (5 * 1012 phage displayed Fab molecules/ml) randomly picked from naïve libraries (24 ). Cells were then fixed with 3% paraformaldehyde for 10 min at room temperature, washed with cold phosphate buffered saline (PBS) and permeabilised with cold acetone for 30 s. After rinsing in cold PBS, cells were incubated 10 min in 4N HCl and washed again with PBS. An anti-BrdU antibody (DAKO M0744) and 10 [mu]M Hoechst dye (Sigma B2883) was added followed by a FITC-conjugated anti-mouse antibody (Sigma F5387). Coverslips were mounted in mowiol and examined with a Zeiss Axiophot microscope using appropriate filters.
4 * 105 cells were fixed on 96-well plates treated with poly-L-lysine (1 mg/ml in PBS) for 15 min at room temperature. After blocking with 2% dry milk in PBS, phage supernatants diluted in 2% dry milk in PBS were incubated for 1 h and rinsed extensively with PBS containing 0.1% Tween 20. An anti-M13 phage antibody conjugated with horse radish peroxidase (HRP) (Pharmacia) was then incubated to detect bound phages. The ELISA was developed with soluble peroxydase substrate (Sigma T3405).
For assays in 6-well culture plates, 105 cells in 2 ml medium were serum starved for 24 h and stimulated for 3 h with EGF or phage-EGF. For assays in 96-well culture plates, 4 * 103 cells in 100 [mu]l of medium were starved 12-24 h before stimulation. Cells were rinsed with cold PBS and lysed for 10 min in 250 [mu]l (6-well plate) or 50 [mu]l (96-well plate) 15 mM phosphate buffer, pH 7, containing 2 mM ATP, 1 mM DTT, 8 mM MgCl2, 0.5% Triton X-100. Luciferase activity was assayed using luciferase substrate buffer (0.1 mM EDTA, 20 mM tricine, 1 mM MgCO3, 2.64 mM MgSO4, 33.3 mM DTT, 270 [mu]M acetyl-coenzymeA (Boehringer 101907) and 470 [mu]M d-luciferine (Boehringer 411400).
Competition experiments were performed as described above. Soluble EGF or phage-EGF supernatant diluted in medium culture were mixed with either blocking antibody (IgG 528, Santa Cruz sc120) (31 ) or non-blocking antibody (clone 29.1, Sigma E-2760) before cell stimulation.
Human EGF (22 ) was displayed on the surface of fd phage FUSE5 through fusion to the N-terminus of the minor coat protein g3p. Resulting phage-EGF particles readily infected Escherichia coli indicating that the g3p function was restored. Phage concentration, determined spectrophotometrically or by DNA extraction (29 ), was ~1010 phage particles/ml of culture supernatant, which is typical for fd-tet derivatives. The phage yield obtained by titration is reproducibly underestimated by a factor 100. Assuming five EGF molecules displayed per phage, the number of EGF displayed molecules therefore corresponds to 500 * p.f.u. number, which is ~5 * 1012 molecules/ml of phage preparation.
Phage-EGF and negative control phages displaying no protein or unrelated proteins were tested for binding to the EGFR by ELISA on cells (Fig. 1 A). The assay was performed either on cells overexpressing the EGFR (2.2wt and A431 cells) or on cells not expressing the EGFR (2.2 cells). We ensured that all phages were used at the same concentration by bacteria infection titration and by direct ELISA detecting coated phages with an anti-phage antibody-HRP conjugated. For EGFR overexpressing cell lines, a strong ELISA signal was obtained with the phage-EGF and not with control phage. A low signal was obtained on cells that did not express the EGFR demonstrating the binding specificity to the EGFR.
The biological activity of wild-type phage-EGF opens the possibility of screening activities of EGF mutants as phage displayed molecules, which can easily be purified through PEG precipitation from bacterial supernatants. However, the DNA synthesis assay described above is not suitable for the screening of a large panel of phage-EGF mutants. Therefore, we investigated earlier EGF responses to develop a workable assay for the EGFR transduction activity. To this aim, we devised a novel test, based on the induction of gene expression mediated by the c-fos gene enhancer SIE, which constitutes a marker for transduction through EGFR activation (30 ). The luciferase gene was chosen as a reporter.
We constructed the vector pZlucTKSIE, derived from pZlucTK vector (25 ), in which the luciferase gene reporter is expressed under the control of the TK promoter and three copies of the SIE element. COS-7 cells were transiently cotransfected with pRK5 EGF-R (27 ) and pZlucTKSIE constructs, serum starved for 24 h and stimulated for 5 h with soluble EGF. High levels of luciferase activity were obtained with soluble EGF (Fig. 2 A). No luciferase expression was detected in COS-7 cells cotransfected with pRK5 EGF-R and pZlucTK, which lacks the SIE element, or transfected only with pZlucTKSIE. This indicated that both EGFR expression and the presence of SIE elements are necessary for luciferase expression. The implication of the SIE elements was demonstrated further by competition experiments (Fig. 2 B). COS-7 cells were cotransfected with pRK5 EGF-R, pZlucTKSIE and increasing amounts of pKSSIE, a pBluescript KS(+) vector containing three copies of SIE. Luciferase activity was inhibited by 50% upon addition of a 20-fold molar excess of pKSSIE, and inhibited completely upon addition of a 100-fold molar excess, while no reduction was observed upon the same molar excess of pBluescript KS(+) vector containing no SIE copies.
We then assayed the ability of phage-EGF to activate luciferase expression (Fig. 2 C). COS-7 cells cotransfected with pRK5 EGF-R and pZlucTKSIE constructs were serum starved for 24 h and stimulated for 5 h with soluble EGF (25 ng/ml), phage EGF (1010 p.f.u./ml) or control phage (1010 p.f.u./ml). Similar levels of luciferase activity were observed with soluble EGF or phage EGF, whereas stimulation with the control phage did not produce any luciferase expression.
The results obtained with transient transfectants show that luciferase expression depends on overepression of EGFR and on the presence of the SIE element. The luciferase is therefore a suitable reporter gene for the monitoring of EGFR activation when put under the control of the TK promoter and the SIE enhancer cassette.
The efficiency of the 96-well plate luciferase assay was then tested in the screening of a library of EGF mutants displayed on phage. Residue Asp46 was randomised to study the binding and transducing activity of the mutated EGF molecules. This residue was chosen because of its conservation throughout the EGF-like family of growth factors. Residue 46 links the second [beta]-sheet of hEGF (residues 37-45) to the C-terminal helix (residues 47-51) (12 ). We constructed a D46 random library containing 104 clones, a number ensuring the presence of all possible mutants (33 ). Sequence analysis of 72 randomly picked clones revealed the presence of 18 different residues at this position (only glutamate and lysine were not found), indicating a good library complexity. Each mutant (titration and direct ELISA on coated phages were used to estimate the relative concentration of each phage), was tested in duplicate, either for binding to the EGFR by ELISA or for transduction activity using the luciferase assay (Fig. 5 ).
We showed that human EGF displayed on phage is able to transduce EGF growth signalling at least up to the DNA replication step and, therefore, that phage display can be used to study the biological activity of EGF mutants. Furthermore, a luciferase reporter assay was developed and proven to be a suitable tool for the analysis of EGF signalling transduction. High signals were obtained with phage-EGF while no detectable signal was observed using irrelevant phages. Using previously characterised antibodies directed against the EGFR, the ability of the luciferase assay to detect EGF antagonists has also been demonstrated. Analysis of a panel of EGF variants mutated at position 46 showed a good correlation between signal transduction activity as detected by the luciferase assay and cell binding as detected by phage ELISA. The randomisation of the 41 crucial position severely affected the binding as only wild-type phages were able to bind to the EGFR.
Besides the applicability of the luciferase reporter assay to screen for EGF agonists and antagonists, preliminary experiments indicate that the assay could also be used to study signal transduction by human TGF[alpha] (another EGFR ligand), by murine interferon [alpha]11 and by other ligand/receptor complexes, whose signalling pathways include activation through the SIE elements.
We thank Greg Winter, Lutz Riechmann and Andrew Griffiths for fruitful discussions on phage display, Katherine Ellis for the construction of the 2.2 SIE cell line, and Cecile Gauthier-Rouviere and Catherine Salvat for helpful advice in cell biology. The 2.2 wt cells are a kind gift from A. Chantry. The pRK5 EGF-R vector is a kind gift of Ronald Herbst. This work was supported by the Centre National de la Recherche Scientifique, the Ministère de l'Enseignement Supérieur et de la Recherche, the Ministère des Affaires Etrangères (Programme d'Actions Intégrées Franco- britannique), the Association pour la Recherche sur le Cancer and the Société Française d'Immunologie. M.W. was the recipient of an EMBO short term fellowship.
*To whom correspondence should be addressed. Tel: +33 4 67 61 36 39; Fax: +33 4 67 04 02 31; Email: weill@igm.cnrs_mop.fr
REFERENCES
Return