Synthesis of peptide analogues using the multipin peptide synthesis method

ANALYTICAL

BIOCHEMISTRY

197,168-177

(1991)

Synthesis of Peptide Analogues Peptide Synthesis Method Robert

M. Valerio,

Coselco Mimotopes

Received

March

Mary

Benstead,

Pty Ltd., Cnr Duerdin

Andrew and Martin

Using the Multipin

M. Bray,

Rhonda

A. Campbell,

Street, P.O. Box 40, Clayton,

Victoria

and N. Joe Maeji’ 3168, Australia

4, 1991

using a modular 96-pin format Modification of the multipin peptide synthesis method which allows the simultaneous synthesis of large numbers of different peptide analogues is described. Peptides were assembled on polyethylene pins derivatized with a 4-(&alanyloxymethyl)benzoate (& Ala-HMB) handle. For comparative purposes, peptides were also assembled on the diketopiperazine-forming handle W-(/3-alanyl)lysylprolyloxylactate. In model studies it was demonstrated that &Ala-HMB-linked peptides were cleaved from polyethylene pins with dilute sodium hydroxide or 4% methylaminelwater to yield analogues with &Ala-free acid (&Ala-CO,H) and &Ala-methylamide (&Ala-CONHCH,), respectively. To assess the suitability of this approach for T-cell determinant analysis, analogues of a known T-cell determinant were synthesized with the various C-terminal endings. Peptides were characterized by amino acid analysis and fast atom bombardment-mass spectrometry. HPLC of the crude cleaved peptides indicated that 22 of the 24 peptides were >95% pure. These crude peptide solutions were nontoxic in sensitive cell culture assays without further purification. All three cleavage procedures gave comparable activities in T-cell proliferation assays. These results demonstrate the potential of the multipin peptide synthesis method for the production of large numbers of different peptide analogues. 0 1991

Academic

Press,

Inc.

In recent years, a number of methods (l-4) for the simultaneous synthesis of many peptides have been developed. Some are capable of simultaneous synthesis of as many as 100 peptide resins but all these procedures are limited by the need for individual cleavage and purification which are invariably time consuming. The problem of handling larger numbers of peptides was addressed by the multipin peptide synthesis method of Geysen et al. (5). With this approach, it is possible to assemble and process thousands of peptides

i To whom

correspondence

should

be addressed.

without

specialized

equipment. Importantly, only simple postsynthesis processing steps are required before peptides may be used in solid-phase ELISA2 assays to rapidly screen for peptides of interest. In its original format, the method generated peptides that were covalently bound to the pins, allowing repetitive assays to be performed. More recently, however, the simultaneous preparation of large numbers of cleavable peptides using the Geysen method has been described (6). This application of the multipin method allows the direct cleavage of peptides into neu-

tral pH buffers, which may then be used in biological assays without prior purification. The quality of peptides generated was demonstrated to be comparable to that of conventional solid-phase peptide synthesis procedures (7). Cleavage from pins under very mild conditions was achieved by utilizing a diketopiperazine-forming handle which is stable to strong acids such as trifluoroacetic acid but rapidly cyclizes to a diketopiperazine under neutral or basic conditions with concomitant cleavage. A limitation of this approach is that it yields peptides containing a diketopiperazine moiety at the C-terminus. Although it was demonstrated that these peptides gave excellent results in T-cell proliferation assays (6,15), for some studies, for example, small peptide hormones, free acid or amide carboxy termini are generally necessary for optimum activity. As a result, we have been investigating alternative handles which are cleaved under mild conditions to yield peptides with C-terminal end groups that relate more closely to native structures.

2 Abbreviations used: AC, acetyl; P-Ala, fl-alanine; Boc, tert-butyloxycarbonyl; DCC, dicyclohexylcarbodiimide; DCM, dichloromethane; DIEA, NJ-diisopropylethylamine; DMAP, I-(dimethylamino)pyridine; DMF, dimethylformamide; DMSO, dimethylsulfoxide; DNP, 2,4dinitrophenyl; ELISA, enzyme-linked immunosorbent assay; FAB, fast atom bombardmenti Fmoc, fluorenylmethoxycarbonyl; HOBt, l-hydroxybenzotriazole; HMB, I-(hydroxymethyl)benzoic acid; Lac, L-lactic acid; MeOH, methanol; Pat, phenacyl; PITC, phenyl isothiocyanate; tBu, tertiary butyl; TFA, trifluoroacetic acid.

168 All

Copyright 0 1991 rights of reproduction

0003-2697191 $3.00 by Academic Press, Inc. in any form reserved.

MULTIPIN

PEPTIDE

SYNTHESIS

In this study, we report the use of a hydroxymethylbenzoate (HMB) handle in conjunction with the multipin peptide synthesis method. To demonstrate the viability of this system, we describe the synthesis of known T-cell determinants and assess the effect of the C-terminal groups on peptide activity using T-cell proliferation assays. MATERIALS

AND

METHODS

Fmoc amino acids were purchased from Auspep (Australia), Cambridge Research Biochemicals (UK), and Novabiochem (Switzerland). All solvents used were AR grade unless otherwise stated. DMF (BDH, Australia) was distilled under vacuum from ninhydrin before use. Piperidine (Novabiochem) was distilled from potassium hydroxide before use. 2,4-Dinitrofluorobenzene was purchased from BDH Chemicals (Australia). HMB was obtained from Novabiochem. Phenacyl bromide, DCC, and HOBt were obtained from Fluka (Switzerland). Zinc dust, 40% aqueous methylamine, and DMAP were supplied by the Aldrich Chemical Co. (Milwaukee, WI). NMR spectra. NMR spectra were recorded on a Varian Gemini 200-MHz spectrometer using DMSO-$ as a solvent. ‘H NMR spectra were recorded at 200 MHz and 13CNMR spectra were recorded at 50 MHz. Chemical shifts are given in parts per million relative to tetramethylsilane as internal standard. Mass spectra. FAB mass spectra were recorded on a JEOL-DX303 mass spectrometer fitted with a FAB source. Peptide samples were dispersed in a thioglycerol matrix and bombarded with xenon ions. Spectra were recorded in the positive ion mode. HPLC. HPLC analysis was performed using a Waters Associates liquid chromatography system consisting of two 510 pumps, a WISP 710B autosampler, a Model 440 uv detector (254 nm) with an extended wavelength module (214 nm), and Maxima 820 chromatography workstation. Analytical runs were carried out on a 5-pm Merck Lichrosphere lOORP-18 (250 X 4 mm i.d.) column. The following buffer system was used: A, 0.1% TFA in water; and B, 0.1% TFA in water/acetonitrile (40/60 v/v). Samples were eluted using the following protocol: O-5 min, buffer A, isocratic; 5-20 min, buffer A to buffer B, linear gradient. Amino acid analysis. Amino acid analysis was performed using the precolumn phenylisothiocyanate (PITC) derivatization procedure. Crude peptide solutions were dried and then hydrolyzed in 6 N HCl/O.l% phenol solution (1 ml) under nitrogen at 112°C for 24 h; 0.2-ml aliquots of hydrolysate were dried in a SpeedVac concentrator (Savant Instruments Inc., Farmingdale, NY) under vacuum and redried twice after reconstitution in 40 ~1 water:MeOH:triethylamine (2:2:1). Derivatization was carried out with 20 ~1MeOH:water:triethy1amine:PITC (7:l:l:l) for 20 min at room temperature. The derivatized samples were dried and reconstituted in

OF

PEPTIDE

ANALOGUES

169

0.5 M d&odium hydrogen orthophosphate (pH 7.4) containing 5% acetonitrile (v/v). Separation of the phenylthiocarbamyl derivatives was performed using a 5-pm glass-lined octadecylsilane (ClS) column (250 X 4 mm i.d.) (Scientific Glass Engineering, Melbourne, Australia) at 50°C with uv detection at 254 nm. The HPLC system used was identical to that described above. Phenacyl hydroxymethylbenzoate (4) (HMB-OPac). A solution of phenacyl bromide (32.9 g, 165 mmol) in ethyl acetate (100 ml) and triethylamine (23.0 ml, 165 mmol) was added to a suspension of 4-(hydroxymethyl)benzoic acid 3 (25.0 g, 164 mmol) in ethyl acetate (500 ml). After stirring at room temperature for 64 h, the reaction mixture was partitioned between warm water (300 ml) and ethyl acetate (300 ml). The organic phase was sequentially washed with 10% citric acid (2 x 200 ml), 7% sodium bicarbonate (2 X 200 ml), and saturated sodium chloride (2 X 200 ml) and dried over sodium sulfate. The solid obtained upon evaporation was suspended in ether (150 ml) for 30 min and collected by filtration. The solid was washed with ether and dried to yield 4 as a white crystalline solid. Yield: 35.95 g, 81%; mp, 114-115°C. 13C NMR (DMSO-d,): 6 62.6(CH,,HMB), 67.3(CH,,Pac), 127.1(C3,HMB), 128.2128.5(C3,Pac), 129.7(C2,Pac), 130.0(WHMB), (C2,HMB), 134.7(Cl,Pac), 134.8(C4,Pac), 149.6(Cl, HMB), 166.3(C=O,HMB), 194.1(C=O,Pac). Further details regarding derivative 4 will be published elsewhere (10). 4 - (9 - Fluorenylonethoxy carbonyl- /3- alanyloxymethyl) benzoic acid phenacyl ester (Fmoc-@-Ala-HMB-OPac) (5). A mixture of 4 (4.56 g, 16.8 mmol), Fmoc-P-Ala (5.23 g, 16.8 mmol), and DMAP (0.354 g, 2.9 mmol) in DCM (130 ml) was cooled to 4°C in an ice bath with stirring. A solution of DCC (3.46 g, 16.8 mmol) in DCM (20 ml) was added dropwise over 5 min. The mixture was stirred at 4°C for 30 min and at ambient temperature overnight. Precipitated dicyclohexylurea was removed by filtration and the organic layer washed successively with 10% aqueous citric acid (2 X 20 ml), water (20 ml), 7% sodium bicarbonate (2 X 20 ml), and saturated sodium chloride (2 X 20 ml). After drying over sodium sulfate (1 h), the solution was filtered and evaporated to dryness to give a white solid. Yield: 9.38 g, 99%; mp, 133-135°C. ‘H NMR (DMSO-$): 6 2.55(m,2H,CH,), 3.28(m,2H,CH,), 4.224.31(m,3H), 5.2(s,2H,CH,), 5.74(s,2H,CH,), 7.26-8.03(m,l’lH,arom.). 13C NMR (DMSO-$): 6 34.O(CH,), 36.4(CH,), 46.8(CH), 65.1(CH,), 65.6(CH,), 67.4(CH,), 120.8,125.8,127.8,128.3(Fmoc Ct), 128.6(C3 Pat and C3 HMB), 129.4(C4,Pac), 129.7(C2,Pac), 130.3(C2,HMB), 134.7(C4,HMB), 134.8(Cl,Pac), 141.6(Cl,HMB), 142.9, 144.7(Cq,Fmoc), 157.O(C=O,Fmoc), 166.1(C=O,ester), 172.16(C=O,ester), 193.9(C=O,ketone). 4 - (9- Fluorenylmethoxycarbonyl/I - alunyloxymethyl)benzoic acid (Fmoc-P-Ala-HMB-OH) (6). Derivative 5

170

VALERIO

(9.02 g, 16 mmol) was suspended in ethyl acetate (150 ml) and heated to 70°C with stirring. Acetic acid (150 ml) was added followed by activated zinc powder (5 g) and water (20 ml). The mixture was heated at 70°C for 18 h, cooled, filtered to remove zinc, and evaporated to dryness under reduced pressure. The residue was dissolved in ethyl acetate (120 ml), washed with saturated sodium chloride (2 X 20 ml), and dried over sodium sulfate (1 h). The solution was filtered and evaporated to dryness, and diethyl ether (50 ml) was added. The mixture was stored at -15°C for 3 h after which the solid was filtered and air-dried. Yield: 6.93 g, 97%; mp: l55157°C. ‘H NMR (DMSO-$): 6 2.53(m,2H), 3.26(m,2H), 4.25(m,3H), 5.16(s,2H), 7.27-7.95(m,12H,arom.). 13C NMR (DMSO-$): 6 34.O(CH,), 36.4(CH,), 468(CH), 65.1 ( CH, ) ,65.6 ( CH, ) ,120.8,125.8,127.8 ( Ct,Fmoc ), 128.3(Ct,Fmoc and C3,HMB), 130.2(C2,HMB), 131.1(C4,HMB), 141.6(Cl,HMB), 141.9,144.8(Cq,Fmoc). L-Lactic Acid Phenacyl Ester (Lac-OPac) (7). Triethylamine (35 ml, 251 mmol) and phenacyl bromide (50.0 g, 251 mmol) were added to a stirring solution of L-lactic acid (27.0 g of an 85% solution in water, 255 mmol) in ethyl acetate (500 ml). After 40 h at room temperature, the reaction mixture was extracted with hot water (500 ml). The organic phase was sequentially washed with 10% citric acid (100 ml), 7% sodium bicarbonate (100 ml), and saturated sodium chloride (100 ml) and dried over sodium sulfate (1 h). Evaporation of the solution afforded a gum which was dissolved in ether (100 ml). On addition of petroleum ether (100 ml) and stirring at room temperature for 4 h, a white solid formed and was filtered and air-dried. Yield: 42.17 g, 81%; mp:182183°C. ‘H NMR (DMSO-$): 6 1.35(d,3H,Lac CH,,J= 6 Hz), 4.29(quint,lH,Lac CH,J = 6 Hz), 5.52(Br s,3H,Pac CH,, and Lac OH), 7.55(t,2H,Pac H3,J = 7 Hz), 7.69(t,lH,Pac H4,J = 7 Hz), 7.96(d,2H,Pac H2,J = 7 Hz). 13C NMR (DMSO-d&6 20.4(C3,Lac), 66.O(C2, Lac), 66.5(CH,,Pac), 128.5(C3,Pac), 129.6(C2,Pac), 134.7(C4,Pac), 134.8(Cl,Pac), 175.3(C=O,Lac), 193.9(C=O,Lac). (S) - 2 - (9 - Fluorenylmethoxycarbonyl -L -prolyloxy)propanoic acid phenacyl ester (Fmoc-Pro-Lac-OPac) (8). DCC (2.06 g, 10 mmol) was added to a stirred solution of Fmoc-Pro-OH (3.37 g, 10 mmol), 7 (2.08 g, 10 mmol), and DMAP (0.24 g, 2.0 mmol) in DCM (50 ml) maintained at 4°C in an ice bath. The mixture was stirred at room temperature for 22 h and filtered, and the filtrate was evaporated to dryness under reduced vacuum. The resulting gum was dissolved in ethyl acetate and the solution washed sequentially with 10% citric acid (20 ml), 4% sodium hydroxide (2 X 20 ml), and saturated sodium chloride (20 ml) and dried over sodium sulfate (1 h). Evaporation yielded the product as a yellow gum. Yield: 4.36 g, 83%. 13C NMR (DMSO-$): 6 16.7(C3, Lac), 22.6,23.6(C4,Pro), 28.9,30.O(C3,Pro), 46.2(C5,

E’I’

AL.

Pro), 46.6(Ct,Fmoc), 46.7(C5,Pro), 58.0,58.6(C2,Pro), 66.7,67.O(CH,,Fmoc), 67.3(CH,,Pac), 68.7(C2,Lac), 120.8(Ct,Fmoc), 125.7,125.8(Ct,Fmoc), 127.8,127.9(Ct,Fmoc), 128.4(C3,Pac), 128.5(C2,Pac), 129.5,129.7(Ct,Fmoc), 134.4(Cl,Pac), 134.9(C4,Pac), 141.6(Cq, Fmoc), 144.7,144.8(Cq,Fmoc), 154.5,154.9(C=O,Fmoc), 170.8,170.9(C=O,Lac), 172.5,172.6(C=O,Pro), 193.4(C=O,Pac). (S) - 2 - (9 - Fluorenylmethoxycarbonyl -L -prolyloxy)propanoic acid dicyclohexylamine salt (Fmoc-Pro-Lac-OH . DCHA) (9). Phenacyl ester 8 (4.36 g, 8.3 mmol) was dissolved in ethyl acetate (30 ml) and acetic acid (100 ml) and water was (30 ml) added. Activated zinc dust (6 g) was added and the suspension was stirred for 16 h at room temperature and then filtered to remove zinc. The gum obtained upon evaporation of the filtrate was partitioned between ethyl acetate (150 ml) and water (200 ml). The organic phase was washed with saturated sodium chloride (25 ml), 10% citric acid (25 ml), and saturated sodium chloride (25 ml) and dried over sodium sulfate (1 h). The pale yellow oil obtained upon evaporation was dissolved in ether (50 ml). A solution of dicyclohexylamine (2.0 g, 11 mmol) in petroleum ether 4060 (50 ml) was slowly added and a white precipitate formed. On standing, a white crystalline solid formed and was collected by filtration and washed with ether (3 X 50 ml) and air-dried. Yield: 4.46 g, 95%; mp: 152153°C. 13C NMR (DMSO-$): 6 17.8(C3,Lac), 22.8,23.7(C4,Pro), 24.4(C3,DCHA), 24.6(C4,DCHA), 25.3(C2, DCHA), 29.8(C3,Pro), 46.4(C5,Pro), 46.5(Ct,Fmoc), 46.9(C5,Pro), 52.1(Cl,DCHA), 58.5,59.O(C2,Pro), 66.0, 67.O(CH2,Fmoc), 71.1,71.8(C2,Lac), 120.5(Ct,Fmoc), 125.4,125.6(Ct,Fmoc), 127.5,127.6(Ct,Fmoc), 128.1(Ct, Fmoc), 141.1(Cq,Fmoc), 144.4(Cq,Fmoc), 154.3(C=O, Fmoc), 172.0,172.3(C=O,Lac), 173.2,173.3(C=O,Pro). Preparation of pins. Peptides were assembled on polyethylene pins that had been radiation grafted with acrylic acid and functionalized with a P-Ala-hexamethylenediamine spacer. Detailed procedures for the preparation of pins have been described elsewhere (5,6). Kits containing functionalized pins for research purposes only are available from Coselco Mimotopes Pty. Ltd. or from Cambridge Research Biochemicals Limited (Cheshire, UK). Preparation of HMB functionalized pins. Pins functionalized with Fmoc - fi - Ala - hexamethylenediamine were deprotected with 20% piperidine/DMF (30 min) and washed successively with DMF (2 min) and MeOH (3 X 2 min) and allowed to air-dry (30 min) in an acidfree fumehood. A 60 mM solution of Fmoc-@-Ala-HMB-OH 6 containing HOBt (2 eq) and DCC (1 eq) was reacted with the pins at 25°C for 18 h and washed with DMF (2 min) and MeOH (10 min), air-dried (30 min), and stored at 4°C until required. Preparation of pins functionalized with the dihetopiperazine-forming handle 2. Pins functionalized with

MULTIPIN

PEPTIDE

SYNTHESIS

Fmoc-/3-Ala-hexamethylenediamine were deprotected as described above. A 60 mM solution of 9 (this was liberated to the free acid by treatment with aqueous citric acid and extraction into ethyl acetate prior to use) containing HOBt (2 eq) and DCC (1 eq) was reacted with the pins at 25°C for 18 h and washed as described above. Identical conditions were then used to sequentially couple Boc-Lys(Fmoc)-OH and Fmoc-P-Ala-OH to give Boc-Lys(Fmoc-P-Ala)-Pro-Lac functionalized pins. Cleavage test. To determine the efficiency of cleavage of the handle, Fmoc-P-Ala-HMB functionalized pins were Fmoc deprotected as described and reacted with 2,4dinitrofluorobenzene (30 mM in DMF with 1 eq DIEA, 4 h, 25°C). The pins were washed with DMF (2 min) and MeOH (10 min), air-dried (30 min), and stored at 4°C. The DNP functionalized pins were treated with aqueous methylamine (1, 4, 10, 20, and 40%) and aqueous sodium hydroxide (0.1, 0.25, 1.0 M) in 96-well microtiter plates using 0.15 ml of solution per well. The effectiveness of cleavage was assessed by measuring the increase in absorbance at 405 nm due to the DNP chromophore by using a Whittaker M.A. Bioproducts MA310 EIA reader. Quantitation of amounts cleaved per pin (in nmol) was determined from a standard curve based on various concentrations of DNP-P-Ala (10) in the range lo-70 nmol per microtiter plate well (150 ~11 well). The values were as follows: 10 nmol (A,,, 0.281, 20(0.55), 30(0.84), 40(1.11), 50(1.38), 60(1.6), 70(1.85). Peptide synthesis. Peptides were synthesized using N*-Fmoc protected amino acids. Amino acids requiring side-chain protection were as follows: Lys(Boc), Ser(tBu), Thr(tBu), Tyr(tBu). Pins functionalized with the appropriate handle (either 1 or 2) were treated with 20% piperidine/DMF (30 min) to remove the Fmoc group. Deprotected pins were washed with DMF (2 min) and MeOH (3 X 2 min) and air-dried in an acid-free fume hood (30 min). Pins were soaked in DMF (5 min) immediately prior to coupling. Amino acid couplings were performed in polyethylene 96-well plates using 0.15 ml of a 60 mM solution of the Fmoc amino acid containing DCC (1 eq) and HOBt (2 eq) for each pin. Couplings were allowed to proceed for 18 h at 25°C after which the pins were washed with DMF (2 min) and MeOH (10 min) and air-dried (30 min). The deprotection, coupling, and washing cycles were repeated until the desired sequences were assembled. Where necessary, as a final step, the N-termini of peptides were acetylated using acetic anhydride/triethylamine/DMF (5/l/50) for 90 min at 25°C. The four sequences YSYFPSV (lo), IYSYFPSV (ll), YSYFPSVI (12), and TKIYSYFP (13) were assembled with variations in the end groups at the C-terminus (P-Ala-free acid, methylamide, or cyclo(Lys-Pro)) and the N-terminus (free amino or acetylated), giving a total of 24 different peptide analogues. Deprotection and cleavage. Pins were soaked in a

OF

PEPTIDE

ANALOGUES

171

plastic bath containing a mixture of TFAlphenollethanedithiol (95/2.5/2.5, v/w/v) (75 ml for 96 pins) for 4 h at room temperature to remove side-chain protecting groups. The pins were air-dried (10 min), vacuum desiccated (1 h), sonicated in 0.1% HCl in water/MeOH (l/l) (15 min), and soaked in 0.1 M phosphate-citrate buffer (pH 3) for 3 h. These steps totally remove noncovalently bound nonpeptide material. Peptides attached to pins with the diketopiperazine-forming handle were cleaved by soaking in 0.1 M sodium phosphate buffer pH 7 (0.15 ml/pin) at 25°C overnight to yield peptides with the diketopiperazine group at the C-terminus. Peptides linked to pins with the fl-Ala-HMB handle were cleaved in two ways: FREE ACID ENDING. Deprotected pins were soaked in 0.1 M NaOH (0.15 ml/pin) in a microtiter plate at room temperature for 3 h. Solutions were neutralized by addition of 0.6 M sodium dihydrogen orthophosphate (0.03 ml) to each well and were suitable for direct use in biological assays. METHYLAMIDE ENDING. Deprotected pins were soaked in 4% methylamine/water (0.15 ml/pin) at room temperature for 3 h (fumehood). Peptide solutions (in microtiter plates) were transferred to a desiccator containing phosphorus pentoxide and placed under efficient vacuum (~15 Torr) until dry (usually 5-6 h). The peptides were reconstituted in the buffer of choice for assay or analytical purposes. Conventional solid-phase peptide synthesis. Control peptide H-IYSYFPSVI-NH, was synthesized on a Milligen 9050 synthesizer (Milligen/Biosearch, Burlington, MA). A standard l-h coupling cycle was employed. Amino acids were coupled to Pepsyn KB resin as the Fmoc protected-0-pentafluorophenyl ester derivatives in the presence of HOBt. Side-chain protecting groups were removed by treatment of the resin with TFA/water (95/5) for 90 min at room temperature. The peptide was cleaved from the resin (saturated ammonia/MeOH, room temperature overnight) and purified by reversephase chromatography. Amino acid analysis: serine, 2.18(2); proline, 1.13(l); tyrosine, 2.06(2); valine, 0.89(l); isoleucine, 1.69(2); phenylalanine, 1.13(l). HPLC: >98% pure, retention time 18.43 min (see HPLC methods for conditions). T-cell proliferation studies. For the T-cell clone proliferation assay, the human T-cell clone RP-9, specific for the minimal determinant YSYFPSVI (residues 593600 of tetanus toxin), was used to test the biological activity of the peptides as previously described (14). Briefly, 1.5 X lo4 washed clone cells and 2 X lo4 y-irradiated (5000 rad) autologous Epstein-Barr virus-transformed B cells were cultured in 190 ~1 of RPM1 1640 medium supplemented with 2 mM glutamine, 50 pg/ml gentamycin, and 10% fetal bovine serum (CSL, Australia) and 10 ~1 of peptide solution in 96-well microtiter trays (Nunc, Denmark) for 72 h in a 37°C 5% CO, incu-

172

VALERIO

ET

AL.

bator. Wells were pulsed with 0.5 &i tritiated thymidine (Amersham, UK; sp act 40-60 Ci/mmol) for the final 6 h of incubation before harvesting onto glass fiber filter mats (Skatron, U.S.A.). Incorporated radioactivity was determined by liquid scintillation counting (Betaplate, LKB-Pharmacia, Finland) and results of triplicate wells are expressed as mean counts per minute.

BOC-NH

HO-Pin

t

RESULTS

AND

DISCUSSION

HMB has been used in Fmoc peptide synthesis protocols as a detachable handle between the peptide and the solid support (11). The HMB handle is stable to sidechain deprotection conditions (TFA) but may be cleaved with amines (commonly ammonia) or dilute sodium hydroxide to generate peptides with a C-terminal amide or acid, respectively. Because an extra step is required for peptide cleavage, the HMB handle has been superseded by a number of new acid-labile handles which cleave concurrently during side-chain deprotection to give peptide amides (813). However, the HMB handle has a number of advantages when utilized in conjunction with the multipin peptide synthesis method. First, the stability of the HMB handle to TFA means that it is possible to subject peptides to sidechain deprotection conditions and simply remove residual reagents and by-products using an appropriate washing schedule while still retaining the peptides on the pins. This is further facilitated by the 8 X 12 modular format of the multipin method, which would allow simultaneous processing of multiple numbers of peptides. Second, following side-chain deprotection, cleavage of peptides from pins into solution could be achieved by treatment with dilute amine or base. Simple processing should then give solutions that are nontoxic in biological assays. The recently developed diketopiperazine-forming handle (Fig. 1) incorporates a &Ala spacer between the handle and the peptide. The P-Ala spacer was originally included to minimize the effect of the diketopiperazine moiety on peptide structure. In this instance, the placement of a &Ala residue at the C-terminus of HMBlinked peptides was also necessary to ensure comparable cleavage rates regardless of the peptide sequence. This is an important consideration in the screening of large numbers of peptides, where concentrations should be identical to allow comparison of specific binding or activity. Consequently, to allow direct comparison of the effectiveness of the HMB handle with the diketopiperazine-forming handle, a derivative incorporating a /3-alanine residue was prepared in a form which would allow direct coupling to pins according to the strategy in Fig. 2. This approach was favored over individual coupling of the HMB and P-Ala residues since the coupling of protected amino acids to the hydroxyl group of HMB on pins often gives highly variable loadings (11).

Base or Buffer -

FIG.

1.

Mechanism

H2N

of diketopiperazine

formation

The carboxylic acid function of 4-(hydroxymethyl)benzoic acid 3 was protected as the phenacyl ester by reaction with phenacyl bromide in the presence of triethylamine. HMB-OPac 4 was condensed with Fmoc-P-Ala using DCC in the presence of a catalytic amount of DMAP to yield the fully protected Fmoc+Ala-HMB-OPac 5. Reduction of 5 with zinc/acetic acid afforded the protected handle Fmoc-/3-Ala-HMB-OH 6 in an overall yield of 96%. Pins functionalized with the diketopiperazine-forming handle were prepared according to the strategy in Fig. 3. Previously, the diketopiperazine-forming handle was based on an ester of serine (6) and also glycolic acid (7). As part of a series of esters that we are currently evaluating for use in peptide synthesis, the lactate ester was arbitrarily chosen for comparative purposes in this study. We have shown that the lactate ester has stability similar to that of the other esters and is equally efficient in the diketopiperazine-forming reaction (9). The protected handle Fmoc-Pro-Lac-OH 9 was prepared in an 64% overall yield using a strategy identical to the preparation of 6. Sequential addition of 9, BocLys(Fmoc)-OH, and Fmoc-P-Ala-OH to pins gave BocLys(Fmoc-p-Ala)-Pro-Lac functionalized pins which yield peptides with a diketopiperazine at the C-terminus under the appropriate deprotection and cleavage conditions. For Fmoc-fi-Ala-HMB-OH 6 to be useful for multipin peptide synthesis, it should couple efficiently, be stable to the conditions of peptide synthesis, and be cleaved under mild conditions to yield peptides which may be used directly or after minimal processing in biological assays. Coupling of Fmoc-&Ala-HMB-OH 6 to both resins and pins using DCC activation in the presence of HOBt proceeded at a rate similar to those of conventional Fmoc-protected amino acids under identical coupling conditions. Loadings of 40-75 nmol of the handle per pin were generally used for peptide synthesis.

MULTIPIN

HOW

\’ VK

A

PEPTIDE

SYNTHESIS

PEPTIDE

173

ANALOGUES

suggesting that it may also be useful in the noncleaved form for ELISA assays (10). Previous work using Pepsyn KB resins (11) established that peptides were cleaved from the HMB handle using 0.1 M NaOH to generate peptide free acids. Subsequent treatment of DNP-fi-Ala-HMB pins with 0.1 M NaOH (1 h) gave 75-80% cleavage as judged by DNP absorbance. Quantitative cleavage was achieved using 0.25 M

OH 0

/TEA

OF

/ EtOAc

Fmoc ‘N

0 P

0

Y 0 -b

I

191

FmOC-BAla-OH/ DMAP/

OH

DCC / DCM

\ Pin

III BLX-~~F~~OH~DCCM~F (After

Pro

PaKOtectlon)

0

FIG.

2.

Preparation

of Fmoc-fl-Ala-HMB-OH

6.

To assess the criteria of stability and cleavage of the handle, pins were derivatized with 6, the Fmoc group was removed, and dinitrofluorobenzene was coupled to yield DNP-PAla-HMB pins. The DNP group is a strong yellow chromophore and loss or cleavage from pins may be conveniently quantitated by measuring the absorbance (405 nm) of solutions with a microtiter plate reader. DNP-P-Ala-HMB pins were subjected to several coupling cycles and a side-chain deprotection. Under these conditions, no loss from the handle was detected. The handle is also stable to neutral aqueous conditions,

FIG. 3. zine-forming

Preparation handle

of pins 1.

functionalized

with

the diketopipera-

174

VALERIO

NaOH (1 h) and this was subsequently employed as the standard procedure to assess cleavage yields. The yield of cleavage in 0.1 M NaOH may be increased with longer reaction times (up to 3 h) but longer exposure is generally not recommended because it may be detrimental to peptide quality. The high pH of sodium hydroxide solution makes it unsuitable for use in biological assays. We have found, however, that 0.1 M sodium hydroxide solutions may be simply neutralized with small volumes of 0.6 M sodium dihydrogen orthophosphate to give solutions compatible with a range of cell and biological assays. The structure of peptides generated by base cleavage of the @-AlaHMB handle is illustrated in Fig. 4. Saturated ammonia in methanol has also been used to generate peptide amides via cleavage from the HMB handle. This reagent is highly volatile and resin cleavages are performed in tightly sealed flasks or bottles, therefore, the pin format is not ideally designed for this type of reaction. Nevertheless, using small screwcapped vials as reaction vessels, we have shown that for smaller numbers of pins (approx 20), preparation of peptide amides with saturated ammonia in methanol is possible (12). However, preparation of large numbers of peptides required for screening studies using this reagent is not a practical proposition at this stage. The HMB handle is susceptible to cleavage with other amines such as hydrazine and we therefore reasoned that methylamine might be a useful alternative to ammonia. We have subsequently found that methylamine, unlike ammonia, may be used in aqueous solution without formation of the corresponding free acid peptides. DNP-/3-Ala-HMB pins were treated with various concentrations (l-40%) of methylamine/water for 1 h. The rate of cleavage in 4% methylamine/water was approximately the same as 0.1 M NaOH with an 80% average yield after 1 h at room temperature. Quantitative cleavage was achieved using a 10% solution after 1 h. Peptides cleaved from the HMB handle with methylamine solution possess an N-methylamide group at the C-terminus (Fig. 4). Unfortunately, traces of methylamine are toxic to

PPm!Im+-Ala-Pni

o.lMM

a*-

A PElTlIE+-Alaa aciac-- . FIG.

4.

Cleavage

-*-3

N-mthylzdde~ of fi-Ala-HMB-linked

peptides

from

pins.

ET

AL. -m-al

10 a-f

IlmFilfa-PXWSaPVi%l

11 a-f

mWPz04er-Val-Ile

12 a-f

!I?lE--I&F-Ilwmm

I3

FIG. 8. Structure of peptide as six variants as follows: C-terminus a b c d e f

acetyl acetyl acetyl amino amino amino

analogues.

Each peptide

a-f was prepared

N-terminus P-Ala-free acid fl-Ala-methylamide P-Ala-diketopiperazine &Ala-free acid /3-Ala-methylamide @-Ala-diketopiperazine

cells and total removal of this reagent is necessary before any subsequent biological assays. Removal is simplified due to the format of the multipin method, which allows peptides to be cleaved directly into the wells of a microtiter plate. The solutions are conveniently dried down by placing the microtiter plates into a desiccator containing phosphorus pentoxide. The desiccator is placed under efficient vacuum (~15 Torr) for 5-6 h. The peptides are reconstituted in the buffer of choice and may be used in assays without further processing. To evaluate the suitability of the P-Ala-HMB handle for the production of peptides on pins, a known T-cell determinant from tetanus toxin was chosen as a target for synthesis (Fig. 5). Using a series of different length peptides,the minimum T-cell stimulatory sequence was previously found to require either I592 or Iso in addition to the core sequence 593YSYFPSV59B (6). All peptides were capped by acetylation and ended with a p-Ala&ketopiperazine. The same sequences were synthesized with the three different C-termini for comparison with the previous work. Figure 6 shows a representative cross-section of the purity of the peptides as determined by HPLC. Of the 24 analogues prepared, 22 were >95% pure by HPLC; the remaining two contained a major peak. Minor impurities could not be attributed to any peptide type with the three different cleavage protocols yielding peptides of comparable quality. This was an excellent result considering that the peptides were not subjected to chromatographic purification. Amino acid analysis of the peptide solutions gave the correct composition and showed on average 20-30 nmol of peptide/well for peptides cleaved from the P-AlaHMB handle and 25-35 nmol/well for peptides cleaved from the diketopiperazine-forming handle. The structure of several of the peptides was also confirmed by FAB-mass spectrometry (Fig. 7).

MULTIPIN

PEPTIDE

SYNTHESIS

OF

PEPTIDE

175

ANALOGUES

2.00

1

1.00 (iii) f

0.50

0.00

F? k +

I” 0.00

c

A

.A

1

I

1.00

2.00 x 10' minutes

”

3.00

I

1.00

2.00 x 10'minutes

3.00

”

6.00

6.00

4.00

2.00

0.00

-2.00 r 0.00

1.00

(i)l $

0.50

n I

FIG.

6.

HPLC

r

r

’

1.00

I

of cleaved

peptides.

I

c

c

2.00 x10' minutes

(i) Peptide

lOa,

I

”

c

3.00

(ii) peptide

lob,

(iii)

peptide

10~.

176

VALERIO

ET

AL.

80

(iii)

Lz d 5 S 'G!2 I d

60

40 (MtNa)+ 20

0 200

(ii)

$

80

3 5 2

60

'ZP I d

40

400

600

1000

800

1200

M/Z

(M+Na)+

400

600

800

1000

12 M

327 (M-H+2Na)+

1019 I

400

600

800

1000 'f%!

FIG.

‘7.

FAB-mass

spectrometry

of cleaved

peptides.

The peptides were used in T-cell proliferation assays, the results of which are summarized in Fig. 8. The full details of the T-cell proliferation results are reported elsewhere (16). The free acid and N-methylamide peptides gave results comparable to those of peptides with the diketopiperazine group at the C-terminus. Importantly, the levels of proliferation for the three different peptide analogues were comparable to that of a control peptide (H-IYSYFPSVI-NH,) prepared by conventional solid-phase peptide synthesis methods.

(i) Peptide

lOa,

(ii) peptide

lob,

(iii)

peptide

10~.

This indicates that for this system, the C-terminus is not important and thereby reinforces the validity of previous T-cell work using peptides with a diketopiperazine at the C-terminus. It is also important to note that the cleaved peptide solutions stimulated T-cell proliferation, indicating their nontoxicity to the cells involved. This suggests that the procedures and buffers used do not introduce significant contaminants and that the method should be compatible with other biological systems.

MULTIPIN

PEPTIDE

SYNTHESIS

OF

PEPTIDE

177

ANALOGUES

ACETYLATED

I

r;

6000

8 5000, 5 4000 E 3000 1000, 0. Is

FIG. 8. Peptide-specific T-cell clone proliferation assay. All analogues of YSYFPSVI synthesized on pins (dark columns) stimulated proliferative responses from the tetanus toxin-specific human T-cell clone RP9 when added along with irradiated antigen-presenting cells. Assay controls without peptide (B, T, T + B) did not proliferate. The bulk-synthesized positive control peptide (IYSYFPSVI) was used at 10 and 1 pg/ml. Peptides cleaved from pins were present at a concentration of approximately 10 pg/ml.

In conclusion, these results illustrate the utility of the multipin peptide synthesis method for producing large numbers of peptides which may then be cleaved into solution under mild conditions. The purity of peptides is comparable to that obtained by conventional solidphase peptide synthesis; moreover, this purity is achieved using simple postdeprotection processing steps and does not involve time-consuming chromatography. The quantities cleaved are sufficient for characterization and multiple biological assays. In addition, cleaved peptide solutions are nontoxic to cells and should be applicable to other biological systems. The method we have outlined should enable research programs requiring hundreds or thousands of peptides or peptide analogues to be undertaken in a short time without specialized equipment. This study incorporated a @-alanine spacer for the purpose of comparison with previous work with diketopiperazine terminated peptides. By substituting any of the 20 amino acids for P-alanine, it is also possible to generate peptides with native C-termini (10).

REFERENCES 1. Wolfe, H. R. (1987) Manual to RaMPS Multiple Peptide Synthesis System, DuPont Biotechnology Systems Division, Wilmington, DE. 2. Houghten, R. A. (1985) Proc. Natl. Acad. Sci. USA 82, 51315135. 3. Krchnak, J. (1990) 4. Frank,

The authors thank Chrys Margellis, Alice Wang, David Stanton, and Brian Sutherland for their skilled technical assistance. We also thank Stuart Rodda and David Mutch for conceiving and initiating the T-cell proliferation work which allowed the validation ofthe cleavage procedures. Finally, we thank Dr. H. Mario Geysen, Chief Scientist of Coselco Mimotopes Pty Ltd. for his continuing advice and encouragement during the course of this work.

R., and Doring,

5. Geysen, Schoofs,

R. (1988)

7. Bray, A. M., Maeji, Lett. 31,5811-5814.

Tetrahedron

8. Rink,

H. (1987)

A. M., and Maeji,

Tetrahedron

A. M., Jhingran, in preparation.

11. Atherton, Synthesis

and Geysen,

N. J., and Geysen,

9. Bray, 10. Bray, script

A., and Roubal, 44,6031-6040.

H. M., Rodda, S. J., Mason, T. J., Tribbick, G., and P. G. (1987) J. Immunol. Methods 102,259-274.

6. Maeji, N. J., Bray, A. M., Methods 134,23-33.

12. Data

ACKNOWLEDGMENTS

V., Vagner, J., Novak, J., Suchankova, Anal. Biochem. 189,80-83.

Lett.

28,

H. M. (1990)

J. Immunol.

H. M.

Tetrahedron

3787-3790.

N. J. Unpublished A., Maeji,

(1990)

results.

N. J., and Valerio,

R. M. Manu-

E., and Sheppard, R. C. (1989) Solid Phase Peptide a Practical Approach, Oxford Univ. Press, Oxford.

not presented.

13. Bernatowicz, M. S., Daniels, dron Lett. 30,4645-4648.

S. B., and Koster,

H. (1989)

Tetrahe-

14. Ho, P. C., Mutch, D. A., Winkel, K. D., Saul, A. J., Jones, Doran, T. J., and Rzepczyk, C. M. (1990) Eur. J. Immurwl.

G. L., 20,

477-483. 15. Gammon, Ametani, 16. Mutch, Geysen,

G., Geysen, H. M., Apple, R., Pickett, A., and Sercarz, E. (1991) J. Exp. Med. D. A., Rodda, S. J., Benstead, H. M. (1991) Peptide Research

E., Palmer,

M., Valerio, 4, 132-137.

M.,

1'73,609-617. R. M.,

and