Notes
Inhibition of EGFR Activation by Bivalent Ligands Based on a Cyclic Peptide Mimicking the Dimerization Arm Structure of EGFR
2018 Volume 66 Issue 11 Pages 1083-1089
Details
2018 Volume 66 Issue 11 Pages 1083-1089
The epidermal growth factor receptor (EGFR) is a receptor in the ErbB family, and is overexpressed in some cancer cells. Recent research has shown that, since clustering of the EGFR increases the possibility of its dimerization and activation, the dimerization state of the EGFR on the cell surface is important for the recognition of the EGFR. In case a bivalent inhibitor has an optimized linker length, the clusters of the EGFR could be recognized with high affinity and kinase activation, which depends on EGF, could be suppressed. Peptide 1, which is derived from the dimerization arm of the EGFR, has been found previously to inhibit autophosphorylation of the EGFR. In this study, bivalent ligands based on peptide 1 with linkers of poly(L-proline) or poly-[(glycine)4(L-serine)] have been designed and synthesized. Bivalent ligands with polyproline linkers could maintain the distance between the ligand moieties. The inhibitory activity of these bivalent ligands against EGFR autophosphorylation was measured and was found to increase as the linker enlarges up to a 15-mer proline linker. The inhibitory activity of a bivalent ligand 7b is significantly higher compared to the corresponding monomeric peptide 2a. This suggests that bivalent EGFR ligands with optimal and rigid linkers could recognize the clusters of the EGFR with higher affinity and suppress kinase activation involving EGF.
Receptors of the ErbB family are overexpressed in several cancer cells, and the dysregulation of their signaling is implicated in cancerous change.1,2) A number of studies have demonstrated ErbB clustering on cell surfaces,3,4) and environmental scanning electron microscopy (ESEM) of whole cells has detected higher order clusters of the epidermal growth factor receptor (EGFR).5) It has been reported that EGFR clusters might be distributed within outside domains of caveolae.6) The EGFR is known to be enriched in lipid rafts,7,8) whose components include cholesterol, sphingolipids, and gangliosides and which act as platforms for cellular signaling.9,10) It has been suggested that the localization of EGFR clusters in lipid rafts is related to dysregulation of its normal ligand binding and tyrosine kinase activation,8,11–14) and thus EGFR clustering is a significant target for development of cancer therapeutics. Exploitation of bivalent ligands could be a useful strategy with which to enhance binding of drug candidates to therapeutically relevant proteins by taking advantage of a bivalent binding site.15–19) Binding studies have shown improved affinities of bivalent inhibitors toward their targets when compared with those of monovalent inhibitors.20–23) In bivalent ligands derived from the chemokine receptor CXCR4 antagonist T140, the binding affinity for CXCR4 and the anti-chemotactic activity of these bivalent ligands increase as the linker length increases up to approximately 6 nm, which are remarkably higher compared to those of the monomeric T140 derivatives.24,25) As a consequence, bivalent ligands could be potent inhibitors of receptor–ligand interactions, and a strategy which uses bivalent ligands could be effective in drug design. Previously, we designed an inhibitory cyclic decapeptide, CQTPYYMNTC (1), which has a disulfide bridge linking between the two terminal cysteine residues and was derived from a part of the dimerization arm of the EGFR (residues 242–259).26) This peptide (1) may act on the extracellular region of the EGFR and inhibit autophosphorylation of the EGFR by allosteric disruption of receptor dimerization. The cellular uptake and significant suppression of proliferation of a conjugated peptide, which was composed of peptide (1), the pro-apoptotic domain (PAD) peptide and a linker cleavable with a protease, were shown in an EGFR-positive lung carcinoma cell line, A549.27) Peptide (1) is a promising lead compound both as an inhibitor against EGF-dependent signals and as a new intracellular delivery vehicle for therapeutically effective peptides. Since clustering of the EGFR could increase the possibility of its preformed dimerization and its activation,28–31) capture of the preformed dimerization state of the EGFR on the cell surface would be important. In this study, bivalent inhibitors based on the peptide (1) with various linkers have been designed and synthesized in order to explore an optimal linker length with the aim of recognition of preformed dimeric recepters in the EGFR clusters with high affinity and suppression of EGF-dependent kinase activation.
In this study, a peptide (2a) with a head to side-chain cyclic structure was synthesized as a precursor unit for bivalent ligands. To prepare easily bivalent ligands having several lengths of linkers, Huisgen cycloaddition is a suitable reaction for conjugation as described later. Thus, a lactam-based cyclic peptide (2a) was synthesized instead of a disulfide-linked peptide (1). Peptide 2a was synthesized by the combination of 9-fluorenylmethyloxycarbonyl (Fmoc)-based solid-phase peptide synthesis (SPPS) and lactam formation between the main chain α-amino group of the N-terminal L-lysine residue and the side-chain γ-carboxy group of the C-terminal L-glutamic acid residue (Chart 1). Each peptide chain was elongated by conventional Fmoc-based SPPS on a 2-chlorotrityl resin, and the protected peptide was ultimately cleaved from the resin. The peptide was cyclized by condensation of the α-amino group of the N-terminal Lys residue with the side-chain carboxy group of the C-terminal Glu residue, and then the protected cyclic peptide was deprotected by treatment with a trifluoroacetic acid (TFA) cocktail. The crude cyclic peptide was purified by preparative reverse-phase (RP)-HPLC, followed by lyophilization to give the desired cyclic peptide (2a).
The peptide chain was elongated by Fmoc-based SPPS using a 2-chlorotrityl chloride resin, and the protected peptide was then cleaved from the resin. The protected peptide was then cyclized by cross-linking the α-amino group of the N-terminal Lys residue to the side-chain carboxyl group of the C-terminal Glu residue. The cyclic peptide formed was deprotected by treatment with a TFA cocktail. Reagents and conditions: (a) 20% piperidine/DMF, 15 min, (b) Fmoc-AA-OH (3.0 eq), DIPCDI (3.0 eq), HOBt·H2O (3.0 eq), DMF, 2 h, (c) TFE/AcOH/DCM (1 : 1 : 3), 3 h, (d) HOAt (3.0 eq), HATU (2.9 eq), DIPEA (6.0 eq), CH2Cl2, 3 h, (e) TFA/EDT/TIS/H2O (10 : 0.75 : 0.25 : 0.1, v/v), 2 h.
The inhibitory effects of peptides 1 and 2a on autophosphorylation of the EGFR were assessed in an intact human epidermoid carcinoma cells (A431) cell line. This cell-based assay was performed as previously reported.26) In the presence of peptide 2a at a concentration of 10 µM, autophosphorylation of the receptor showed an approximately 60% decrease compared with the control, which was treated with EGF alone (Fig. 1). The level of inhibitory activity of peptide 2a was similar to that of peptide 1. These results suggest that peptide 2a can be utilized as a monomeric precursor in place of peptide 1.
(A) Bands of autophosphorylated EGFR were detected using anti-pY1068 EGFR. (B) Inhibitory effects of peptides 1 and 2a (10 µM) on EGFR autophosphorylation in intact A431 cells. EGF-stimulated autophosphorylation in the absence of a test peptide is set at 1.0. The results are shown as the mean±standard deviation for three independent experiments.
Bivalent ligands, in which two peptides of 2a are connected by a poly(L-proline) linker (a Pro linker) or a poly[(glycine)4(L-serine)] linker (a GlySer linker), were designed and synthesized (Fig. 2). Propiolic acid was condensed at the side-chain of the N-terminal L-lysine residue of peptide 2a to obtain an alkyne (3), which was reacted with linkers containing two azide groups, 5a–d and 6a–c, by Huisgen cycloaddition (Chart 2). Pro linkers have been adopted as rigid linkers between two functional units, which produce a predetermined separation of active sites.32,33) Linkers consisting of poly(L-prolines) [Pro numbers: 9–27] are expected to maintain constant distances of 2.5–8.0 nm between the ligands, and poly(L-proline) helices are known to maintain a length of 0.9 nm per turn.34,35) GlySer linkers have been widely used as flexible linkers which require a certain degree of movement.36) These linkers consist of small amino acids which are non-polar (Gly) or polar (Ser). In this experiment, Pro and GlySer linkers with lengths of 2.5-8.0 nm were synthesized. Bivalent ligands 7a–d and 8a–c, which have two types of linkers with various lengths, were synthesized by azide-alkyne Huisgen cycloaddition. Some ligands were expected to be more potent as inhibitors because they can maintain the distance between the two interaction sites of the dimerization arm of the EGFR. In addition, another bivalent ligand (4), a disulfide cross-linked dimer of 2b, was prepared and a monovalent control (9), an analogue of the peptide (7b) with a 15-mer L-proline linker, which was acetylated at the other terminus, was synthesized.
As EGFR binding moieties, peptides 2a and 2b and a disulfide cross-linked dimer (4) were prepared. Bivalent ligands with EGFR binding moieties on both termini of Pro linkers, (7a–d), and GlySer linkers, (8a–c), were synthesized. As a monomer binding ligand with a linker, an acetylated analogue with a 15-mer L-proline linker (9) was also synthesized.
Reagents and conditions: (a) NHS-propiolic acid (2,5-dioxopyrrolidin-1-yl propiolate) (1.5 eq), dry DMF, 24 h. (b) Peptide 3 (2.0 eq), CuSO4 (1.0 eq), sodium ascorbate (2.0 eq), DMF/H2O (1 : 1), 24 h.
The inhibitory activity of the synthetic ligands against autophosphorylation of the EGFR was examined in intact A431 cells. After the cells were stimulated with EGF (10 nM) in the presence of either peptide 2a, a disulfide bridged dimeric peptide (4), bivalent ligands (7a–d, 8a–c), or a monovalent ligand (9), each whole cell lysate was isolated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and the phosphorylated receptor was then detected by immunoblotting with an anti-phosphotyrosine 1068 EGFR monoclonal antibody (Fig. 3). It was found that bivalent ligands 7a–d, which contain Pro linkers, had higher inhibitory activity against autophosphorylation of the EGFR than peptide 2a. The degree of reduction of the phosphorylated receptor was quantified by densitometry of the protein bands, and the addition of peptide 2a or each of the bivalent ligands 7a–d at 10 µM (monomer unit concentration) was found to suppress the phosphorylation levels to <61, 48, 36, 46 and 42% of the control of EGF alone, respectively (Fig. 3). Among the Pro linker-type ligands, 7b showed the highest inhibitory activity. In contrast, bivalent ligands 8a–c, containing GlySer linkers, suppressed phosphorylation to levels similar to that achieved by peptide 2a. Both 7b and 8b contain 15-mer amino acid linkers but showed different activities. The dimeric peptide (4) was also subjected to the assay, and found to show inhibitory activity similar to that of the monomeric peptide (2a). These results indicate that the inhibitory effects of bivalent ligands of the poly(L-proline) type, which contain two monomeric peptides with a Pro linker, are remarkably higher than that of a monomeric peptide. Although a monovalent ligand (9) contains a poly(L-Pro)15 linker as in 7b, it had inhibitory activity similar to that of peptide 2a (Fig. 4). These results suggest that the Pro linker moiety has no effect on the activity, and that bivalent ligands bearing peptide 2a connected by rigid Pro linkers with certain distances play an important role in the inhibitory activity and the interaction with the EGFR. Since the EGFR is clustered on the cell membrane as preformed dimers,28–31) bivalent ligands with Pro linkers of lengths of 15-mer at least could maintain a distance between the ligand moieties that is optimum for these preformed dimeric receptors.
(A) Bands of autophophorylated EGFR were detected using anti-pY1068 EGFR. (B) Inhibitory effects of bivalent ligands (monomer unit concentration: 10 µM) on EGFR autophosphorylation in intact A431 cells. EGF-stimulated autophosphorylation based on the concentrations of the bands observed in Fig. 3A in the absence of a test peptide is set at 1.0. The results are shown as the mean±standard deviation for at least three independent experiments. * p<0.005 (Student’s t test).
EGF-stimulated autophosphorylation in the absence of a test peptide is defined as 1.0. The results are shown as the mean±standard deviation for at least three independent experiments. * p<0.005 (Student’s t test). A ligand 9 could not be assayed simultaneously with ligands 7a–d or 8a–c. Thus, it was independently assayed with only ligands 2a and 7b.
In the dimer interface, Asp279 and His280, which are located at the top of a loop under the dimerization arm, play a critical role in the EGFR dimerization. The dimerization arm and the loop containing Asp279 and His280 contribute more than 75% of the EGFR dimerization energy.37) Consequently, while a monomeric peptide alone shows a certain level of inhibitory activity, bivalent ligands with rigid and long Pro linkers have higher inhibitory activity. In the case of a monomeric peptide or bivalent ligands with short linkers, a partial EGFR dimerization might occur. On the other hand, bivalent ligands with longer rigid linkers might not bind to two dimerized receptors but instead recognize two preformed dimeric receptors28–31) with the maintenance of certain distances between two ligand moieties to separate two receptors on the cell membrane (Fig. 5).
Bivalent ligands based on peptide 2a, a derivative of peptide 1, connected by poly(L-proline) or poly[(glycine)4(L-serine)] linkers have been designed, synthesized and tested. Bivalent ligands bearing two cyclic peptides of 2a with a poly(L-proline) linker were shown to have higher inhibitory activity against autophosphorylation of the EGFR than the monomeric peptide (2a). In contrast, bivalent ligands with poly[(glycine)4(L-serine)] linkers have an inhibitory capability similar to that of the monomeric peptide (2a). The inhibitory activity of bivalent ligands with poly(L-proline) linkers against EGFR autophosphorylation increases as the linker length increases up to that of the L-proline 15-mer, but the corresponding monovalent ligand (9) had inhibitory capacity similar to that of the monomeric peptide 2a. Consequently, bivalent ligands might require the linker length of the L-proline 15-mer at least to maintain the distance between the ligand moieties. It has been reported that in the absence of ligands, the EGFR activity depends on the EGFR density on the cell surface.38) Bivalent ligands have the potential to target clustered EGFR on the cell surface. Since the clustered EGFR is composed of preformed dimers on the cell membrane, bivalent ligands with rigid linkers of lengths of a Pro 15-mer could bind to two preformed dimeric receptors while preserving certain distances between two ligand moieties and separate two receptors on the cell membrane. The EGFR also promotes heterodimerization with other members of the ErbB family receptors (ErbB2–4),39) and the heterodimers produced can cause malignant transformation and drug resistance.40) This bivalent ligand approach could be the basis of bispecific ligands which can target heterodimers.
Fmoc amino acid derivatives, Thr(But), Tyr(But), Pro, Met, Cys(Trt), Asn(Trt), and Gln(Trt), Glu-OBut, Lys(Boc), Lys(N2)[ε-diazo-L-lysine], Gly, Ser(But), were purchased from Merck Japan Ltd. (Tokyo, Japan) or KOKUSAN CHEMICAL Co., Ltd. (Tokyo, Japan). A431 were obtained from RIKEN (Tsukuba, Japan) and 10% fetal bovine serum (FBS) from Thermo Fisher Scientific Inc. (Rockford, IL, U.S.A.), respectively. Reagents were purchased from Nacalai Tesque Inc. (Kyoto, Japan) and used without further purification.
Analytical HPLC was performed using a C18 reversed phase column (4.6×250 mm; a COSMOSIL Packed column, Nacalai Tesque, Inc.) on a JASCO PU-2089 plus (JASCO Corporation, Ltd., Tokyo, Japan) with a binary solvent system—a linear gradient of CH3CN in 0.1% aqueous TFA at a flow rate of 1.0 mL/min with detection at 220 nm. Semi-preparative HPLC was carried out using a C18 reversed phase column (20×250 mm; a COSMOSIL Packed column, Nacalai Tesque, Inc.) on a JASCO PU-2086 plus (JASCO Corporation, Ltd.) with a binary solvent system; a linear gradient of CH3CN in 0.1% aqueous TFA at a flow rate of 5 mL/min with detection at 220 nm. The solvents used were HPLC grade. The purified products were lyophilized and identified by electrospray ionization-time-of-flight (ESI-TOF)-MS, which was recorded on a micrOTOF-2focus mass spectrometer (Bruker Daltonics) in positive and negative detection modes.
Cell Line and Culture ConditionsHuman epidermoid carcinoma A431 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) containing 4.5 g/L glucose, L-glutamine, and sodium pyruvate supplemented with 10% FBS and 1% (v/v) penicillin–streptomycin mediun (10000 units/mL and 10000 µg/mL, respectively) at 37°C under a humidified atmosphere of 5% CO2.
Peptide SynthesisGeneral Procedure of Peptide SynthesisPeptide synthesis was performed using the Fmoc SPPS method. Fmoc-amino acids were coupled by N,N′-diisopropylcarbodiimide (DIPCDI, 3.0 eq) and N-hydroxybenzotriazole hydrate (HOBt·H2O, 3.0 eq). The Fmoc group was subsequently removed by treatment with 20% (v/v) piperidine/N,N-dimethylformamide (DMF) for 20 min.
Peptide 2a2-Chlorotrityl chloride resin (1.51 mmol/g, 199 mg, 0.30 mmol) was used for the synthesis of peptide 2a. After loading Fmoc-Glu-OBut on the resin followed by preparation of Fmoc-SPPS, the resulting protected peptide was cleaved from the resin with TFE/AcOH/DCM (1 : 1 : 3, v/v, 20 mL) for 3 h. The reaction mixture was filtered, the filtrate was evaporated under vacuum, and the peptide was precipitated as a solid powder. The crude peptide (214 mg, 0.1 mmol, 1.0 eq) was cyclized by treatment with HOAt (40.8 mg, 0.3 mmol, 3.0 eq), HATU (110 mg, 0.29 mmol, 2.9 eq) and N,N-diisopropylethylamine (DIPEA) (105 µL, 0.6 mmol, 6.0 eq) in CH2Cl2 (200 mL) for 3 h. The solvent was evaporated under reduced pressure, and the deprotection was performed by treatment with TFA/EDT/TIS/H2O (10 : 0.75 : 0.25 : 0.1, v/v) for 2 h. After concentration of the reaction mixture under reduced pressure, cold Et2O was added to the resulting residue to produce a precipitate, which was washed with Et2O then dried to give the crude peptide as a white solid. The crude peptide was purified by preparative RP-HPLC to obtain the desired peptide (2a). The purified peptides were lyophilized and identified by ESI-TOF-MS. HPLC room temperature (r.t.) 27.5 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C56H82N13O18S+ [M+H]+: 1256.6. Found: 1256.6 (23% yield).
Peptide 2bPeptide 2b was synthesized and purified similarly to peptide 2a as described above. Its HPLC r.t. was 32.6 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C53H75N12O18S2+ [M+H]+: 1231.5. Found: 1231.6 (36% yield).
Peptide 2a Derivative 3NHS-propiolic acid (2,5-dioxopyrrolidin-1-yl propiolate, 1.4 mg, 8.14 µmo1, 1.0 eq) was added to a solution of peptide 2a (10.2 mg, 8.14 µmol, 1.0 eq) in dry DMF (100 µL), and the mixture was stirred at r.t. After 24 h, the solution was diluted with CH3CN in 0.1% aqueous TFA, and subjected to preparative RP-HPLC to obtain the desired peptide (3). The purified peptide was lyophilized and identified by ESI-TOF-MS. HPLC r.t. 30.9 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C59H82N13O19S+ [M+H]+: 1308.6. Found: 1308.6 (24% yield).
Dimeric Peptide 4The crude peptide 2b (13.7 mg, 11 µmol) in NH4HCO3 buffer (100 mL, pH 7.8, peptide concentration of 0.1 mg/mL) was stirred at r.t. After 24 h, the solution was evaporated and purified by preparative RP-HPLC to obtain the desired peptide 4. The purified peptide was lyophilized and identified by ESI-TOF-MS. HPLC r.t. 41.6 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C106H148N24O36S42+ [M/2+H]2+: 1230.5. Found: 1230.9 (35% yield).
Polyproline Linker Derivatives 5a–eA Representative Compound (5a)5a was elongated on a Novasyn TGR resin (0.25 mmol/g, 120 mg, 0.03 mmol) by Fmoc-based SPPS. The coupling of Fmoc-amino acids was performed as described above. After the coupling of C-terminal Fmoc-Lys(N2)-OH, successive Fmoc-Pro-OH, and an N-terminal Fmoc-Lys(N2)-OH, the N-terminal amino group was acetylated by treatment with Ac2O/pyridine/DMF (1 : 1 : 3) for 1 h. For cleavage from the resin and deprotection, the peptide resin was treated with TFA/TIS/H2O (10 : 0.25 : 0.1) for 2 h, followed by filtration. After concentration of the filtrate under reduced pressure, the residue was washed with cold Et2O and dried under reduced pressure. Purification by RP-HPLC gave 5a in 21% yield. HPLC r.t. 36.3 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C59H89N18O12S+ [M+H]+: 1241.7. Found: 1241.6. In the same way, 5b was prepared in 19% yield. HPLC r.t. 37.5 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C89H131N24O18+ [M+H]+: 1824.0. Found: 1824.8; 5c was prepared similarly in 25% yield. HPLC r.t. 39.2 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C119H174N30O242+ [M/2+H]2+: 1203.7. Found: 1203.9; 5d was prepared in the same way in 28% yield. HPLC r.t. 40.4 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C149H216N36O302+ [M/2+H]2+: 1494.9. Found: 1494.9; and in the synthesis of 5e, after the coupling of C-terminal Fmoc-Lys(N2)-OH and and successive Fmoc-Pro-OH, the N-terminal amino group was acetylated by treatment with Ac2O/pyridine/DMF (1 : 1 : 3) for 1 h. The cleavage from the resin and the deprotection were performed similarly, and 5e was prepared in 30% yield. HPLC r.t. 29.6 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C83H121N20O17+ [M+H]+: 1669.6. Found: 1669.5.
GlySer Linker Derivatives 6a–cA Representative Compound 6a6a was elongated on a Novasyn TGR resin (0.25 mmol/g, 120 mg, 0.03 mmol) by Fmoc-based SPPS. The coupling of Fmoc-amino acids was performed as described above. After the coupling of C-terminal Fmoc-Lys(N2)-OH, successive Fmoc-Ser(But)-OH/Fmoc-Gly-OH, and N-terminal Fmoc-Lys(N2)-OH, the N-terminal amino group was acetylated by treatment with Ac2O/pyridine/DMF (1 : 1 : 3) for 1 h. For the cleavage from the resin and the deprotection, the peptide resin was treated with TFA/TIS/H2O (10: 0.25 : 0.1, v/v) for 2 h, followed by filtration. After concentration of the filtrate under reduced pressure, the residue was washed with cold Et2O and dried under reduced pressure. Purification by RP-HPLC gave 6a in 35% yield. HPLC r.t. 25.6 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C25H43N14O9+ [M+H]+: 683.3. Found: 683.7. In the same way, 6b was prepared in 32% yield. HPLC r.t. 19.6 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C47H77N24O21+ [M+H]+: 1313.6. Found: 1313.7; and 6c was prepared in 28% yield. HPLC r.t. 17.7 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C69H111N34O33+ [M+H]+: 1943.8. Found: 1943.9.
Bivalent Ligands 7a–dA Representative Compound 7aCopper(II) sulfate (0.35 µmol, 2.0 eq) in H2O (5 µL) and sodium ascorbate (0.70 µmol, 4.0 eq) in H2O (5 µL) were added to a solution of 3 (0.46 mg, 0.35 µmol, 2.0 eq) and 5a (0.22 mg, 0.18 µmol, 1.0 eq) in DMF (10 µL) and the mixture was stirred at room temperature. After 24 h, the solution was diluted with CH3CN in 0.1% aqueous TFA, and purified by semi-preparative RP-HPLC to obtain the desired peptide. Purification gave 7a in 10% yield. HPLC r.t. 27.9 min [C18 reversed phase column, 1 mL/min, CH3CN (15–45%)/60 min], ESI-TOF-MS; Calcd for C177H253N44O50S23+ [M/3+H]3+: 1286.3. Found: 1286.2. 7b was prepared in 39% yield. HPLC r.t. 33.4 min [C18 reversed phase column, 1 mL/min, CH3CN (15–45%)/60 min], ESI-TOF-MS; Calcd for C207H295N50O56S23+ [M/3+H]3+: 1480.4. Found: 1480.4; 7c was prepared in 45% yield. HPLC r.t. 34.3 min [C18 reversed phase column, 1 mL/min, CH3CN (15–45%)/60 min], ESI-TOF-MS; Calcd for C237H337N56O62S23+ [M/3+H]3+: 1674.5. Found: 1674.6; and 7d was prepared in 66% yield. HPLC r.t. 35.0 min [C18 reversed phase column, 1 mL/min, CH3CN (15–45%)/60 min], ESI-TOF-MS; Calcd for C267H379N62O68S23+ [M/3+H]3+: 1868.6. Found: 1868.6.
Bivalent Ligands 8a–cA Representative Compound 8aCopper(II) sulfate (0.35 µmol, 2.0 eq) in H2O (5 µL) and sodium ascorbate (0.70 µmol, 4.0 eq) in H2O (5 µL) were added to a solution of 3 (0.46 mg, 0.35 µmol, 2.0 eq) and 6a (0.12 mg, 0.18 µmol, 2.0 eq) in DMF (10 µL) and the mixture was stirred at room temperature. After 24 h, the solution was diluted with CH3CN in 0.1% aqueous TFA, and purified by semi-preparative RP-HPLC to obtain the desired peptide. Purification gave 8a in 19% yield. HPLC r.t. 24.2 min [C18 reversed phase column, 1 mL/min, CH3CN (10–40%)/60 min], ESI-TOF-MS; Calcd for C143H206N40O47S22+ [M/2+H]2+: 1649.7. Found: 1649.1. 8b was prepared in 34% yield. HPLC r.t. 22.0 min [C18 reversed phase column, 1 mL/min, CH3CN (15–45%)/60 min], ESI-TOF-MS; Calcd for C165H241N50O59S33+ [M/3+H]3+: 1310.2. Found: 1310.3; 8c was prepared in 17% yield. HPLC r.t. 20.6 min [C18 reversed phase column, 1 mL/min, CH3CN (15–45%)/60 min], ESI-TOF-MS; Calcd for C187H275N60O71S23+ [M/3+H]3+: 1520.3. Found: 1520.3.
Monovalent Ligand 9Copper(II) sulfate (0.37 µmol, 1.0 eq) in H2O (4 µL) and sodium ascorbate (0.74 µmol, 2.0 eq) in H2O (4 µL) were added to a solution of 3 (0.48 mg, 0.37 µmol, 1.0 eq) and 5e (0.62 mg, 0.37 µmol, 1.0 eq) in DMF (8 µL) and the mixture was stirred at room temperature. After 24 h, the solution was diluted with CH3CN in 0.1% aqueous TFA, and purified by semi-preparative RP-HPLC to obtain the desired peptide (9). Purification gave 9 in 33% yield. HPLC r.t. 19.2 min [C18 reversed phase column, 1 mL/min, CH3CN (15–45%)/60 min], ESI-TOF-MS; Calcd for C142H203N33O36S2+ [M/2+H]2+: 1489.3. Found: 1489.3.
Autophosphorylation AssayTreatment of A431 cells with a test peptide (10 µM) followed by treatment with EGF (20 nM), cell-lysis, separation by SDS-PAGE and blotting onto a polyvinylidene difluoride membrane, incubation with an anti-EGFR [pY1068] ABfinity™ recombinant rabbit monoclonal antibody followed by treatment with an anti-rabbit immunoglobulin G (IgG) antibody, and final visualization and quantification were conducted as described in our earlier paper.26) The results are shown as the mean±standard deviation for at least three independent experiments.
This work was supported by JSPS KAKENHI Grant Numbers JP15H04652 to H.T., and JP17J08315 to K.T., the Platform for Drug Discovery, Informatics, and Structural Life Science of Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan, and the Cooperative Research Project of Research Center for Biomedical Engineering.
The authors declare no conflict of interest.
The online version of this article contains supplementary materials. HPLC charts of peptides 2a–9.
