Three-Dimensional Arrangement of Sugar Residues along Helical

Biomacromolecules 2005, 6, 2334-2342

2334

Three-Dimensional Arrangement of Sugar Residues along Helical Polypeptide Backbone. 2. Synthesis of Periodic N-Glycosylated Peptides by Polymerization of Tripeptide Active Esters Containing r,r-Disubstituted Amino Acid Akinori Takasu,* Seiji Horikoshi, and Tadamichi Hirabayashi Deapartment of Environmental Technology and Urban Planning, Graduate School of Engineering, Nagoya Institute of Technology, Gokiso-cho, Showa-ku, Nagoya 466-8555, Japan Received April 11, 2005; Revised Manuscript Received May 5, 2005

New type of N-glycosylated peptides having periodic sequence of -[X-Gln(β-D-GlcNAc)-Aib]- [X ) L-Glu(OMe), L-Lys(Ac), L-Ala; Aib ) R-aminoisobutyric acid] were synthesized by polymerization of glycosylated tripeptides with an active ester methods using Cl-+H3N-L-Glu(OMe)-Gln[β-D-GlcNAc(Ac)3]Aib-ONp (Np)p-nitrophenyl) (13a), Cl-+H3N-L-Lys(Ac)-Gln[β-D-GlcNAc(Ac)3]-Aib-ONp (13b), and Cl-+H3N-L-Ala-Gln[β-D-GlcNAc(Ac)3]-Aib-ONp (13c) as the monomers. Polymerizable glycosylated tripeptides were prepared by stepwise N,N-dicyclohexylcarbodiimide (DCC)/1-hydroxybenzotriazole (HOBt) method. Polymerizations of 13a-c were initiated by triethylamine and proceeded in DMSO at 50 °C for 5 days in the presence of 1-hydroxy-7-azabenzotriazole (HOAt) as the activator (conversions were 25-75%). The glycopeptides were deacetylated by hydrazine monohydrate in methanol to afford periodic glycopeptides 14 (12-27 residues) without racemization (yield, 35-89%). CD spectra in methanol, trifluoroethanol, and water of deacetylated glycopolymers 14a, 14b, and 14c showed double minima (206 and 222 nm) of negative Cotton effect indicating that N-glycoside (N-acetyl-D-glucosamine) was arranged three-dimensionally along the R-helical peptides in water as well as in organic protic solvents. The helix content depends on the solvent, peptide sequence, and spacer between peptide backbone and sugar. Interaction of the glycopeptides with wheat germ agglutinin (WGA) lectin was investigated by fluorescence measurement. Introduction Although the sugar branch has been regarded as an accessory modulator, recent progress in glycobiology has showed that cell surface carbohydrates of glycoproteins play essential roles in various biological recognition processes.1 Despite the large number of natural glycoproteins, the types of covalent bonds between the protein and the saccharide part show relatively little variation.2 Most common are N-glycoproteins, in which the side chain amide function is usually joined to an N-acetyl-D-glucosamine residue through a β-N-glycosidic linkage and a second important type of the connection is O-glycosidic linkage including Ser-O-glycoside and Thr-O-glycoside (mucine type).3 The N-glycoside residue in glycoproteins play an important role in fermentation, embryology, immunology, transportation in cell, aging, and cancer.1 Recent interest is focused on the relationship of congenital disorders of N-glycosylation (CDG) and Alzheimer disease and alcohol addiction.4,5 Although the recognition process is essentially based on carbohydrate-protein6 and carbohydrate-carbohydrate7 interactions, individual interactions are generally low. Some glycopolymers in which saccharide residues are incorporated to polymer backbones induce enhancement of binding affinity * To whom correspondence should be addressed. Telphone: +81-52735-7159. Fax: +81-52-735-5342. E-mail: [email protected].

toward proteins ascribed to multivalent recognition, i.e., “cluster effect”.6 Even the glycopolymers in which the spaces between the saccharides are random showed a strong recognition.6 Therefore, if the three-dimensional arrangement, i.e., interval and direction of the pendant carbohydrate could be regulated, remarkable enhancement of the biding ability would be expected. However, there had been few reports of glycopeptides with definite geometrical patterns until now. Aoi et al.8 applied a dendrimer skeleton and Matsuura et al.9b proposed new strategies to prepare periodic glycosylated oligonucleotide (20-mers) for the three-dimensional arrangement of saccharide. Recently we also reported a threedimensional arrangement of sugar residue along the R-helical peptide,10 in which the sugar was linked to peptide backbone via O-glycoside linkage (mucine-type glycoprotein or glycopeptide model)10 and the R-helical conformation was supported by FT-IR and CD measurements and semiempirical molecular orbital calculations.10 The polypeptide chain, being rigid and regular such as the R-helix and β-sheet, is considered to be an excellent framework to support sugar residue, keeping a specific distance and orientation between the neighboring ones. Furthermore, they could be regulated by the peptide conformation. In the present paper, we describe the three-dimensional arrangement of N-glycoside residues along the R-helical polypeptide backbone by the polymerization of R,R disub-

10.1021/bm0502563 CCC: $30.25 © 2005 American Chemical Society Published on Web 06/23/2005

Three-Dimensional Arrangement of N-Glycoside

stituted amino acid containing tripeptide active esters.11 In contrast to the tedious stepwise solid-state method, polymerizations of glycosylated tripeptides will realize construction of three- or twodimensional carbohydrate ligands in a few steps. The R-aminoisobutyric acid (Aib) residue is one of the R,R disubstituted amino acids and well-known to be a strong inducer that forms a 310- or R-helix depending on peptide sequence, chain-length, or environment.12-14 The 310helix is just slightly more elongated than the R-helix.12 Synthesis of a series of N-glycosylated peptides containing Aib and the conformational investigation would address the effect of N-glycoside on the biological activities of the peptides with defined conformational preference as well as molecular design of a new type of three-dimensional carbohydrate ligand. Experimental Section Materials. Z-Glu(OH)-OBn (1), Di-tert-butyl carbonate (Boc2O), L-alanine, γ-methyl-L-glutamate [H-Glu(OMe)OH], and 1 M HCl in dioxane solution were purchased from Kokusan Chemical Works Ltd. (Tokyo, Japan). Aib, 1-hydroxybenzotriazole monohydrate (HOBt-H2O), and 1-hydroxy-7-azabenzotriazole (HOAt) were obtained from Tokyo Kasei Co. (Tokyo, Japan). N,N-Dicyclohexylcarbodiimide (DCC), 10% palladium-carbon (Pd-C), formic acid, pnitrophenol (NpOH), triethylamine (TEA), formamide, N,Ndimethylformamide (DMF), diethyl ether (Et2O), 2-propanol, 2,2,2-trifluoroethanol (TFE), N-ethoxycarbonyl-2-ethoxy-1,2dihydroquinone (EEDQ), and N-methylmorphorin (NMM) were purchased from Nacalai Tesque (Kyoto, Japan). Chloroform, dichloromethane, and dimethyl sulfoxide (DMSO) were distilled from calcium hydride. Methanol, ethanol, acetonitrile, and water used were purified by distillation. FITC-labeled WGA lectin was purchased from Sigma. Measurements. FT IR spectra were recorded in KBr disks using a JASCO FT/IR-430 spectrometer. 1H and 13C NMR spectra were measured at 27 °C using a Bruker DPX200 spectrometer (200 MHz for 1H NMR). All chemical shifts were expressed as δ downfield from tetramethylsilane (TMS). Number average molecular weights (Mn) and the polydispersity indexes (Mw/Mn) of polymers were estimated by size exclusion chromatography (SEC) calibrated with polystyrene and poly(ethylene oxide) standards using a system of Tosoh HLC 803D with a Tosoh RI-8000 detector and Tosoh TSK-GEL R5000-HXL columns [eluent, DMF+ LiBr (0.05 wt %); flow rate, 1.0 mL/min; temperature, 40 °C] and a system of JASCO model PU-980 with JASCO 830-RI and Amersham Pharmacia Biotech Superdex Peptide HR 10/30 (eluent, 0.05 M K2HPO4 aq.; flow rate, 0.5 mL/ min; temperature, 27 °C), respectively. CD and UV absorption spectra were simultaneously recorded using a JASCO J-820 spectrometer in methanol, water, and TFE at 27 °C. The path length of the quartz cell was 1.0 mm, and the concentration of peptide was 0.5-0.6 mM. These solvents were purified by distillation before use. AFM measurements were carried out on a Nanoscope III (Digital Instrument) in air. The glycopeptide 14a methanol or aqueous solution (0.5-0.6 mM) was deposit on a mica plate and methanol or

Biomacromolecules, Vol. 6, No. 4, 2005 2335

water was evaporated under reduced pressure. Matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectra were measured on a Voyager DE-PRO (Applied Biosystems) using as a matrix reagent. To generate sodiumcationized ions ([M+Na]+), NaI was used as a cationization salt. Preparation of Z-Gln[β-D-GlcNAc(Ac)3]-OBn (3). To a CH2Cl2 solution of 1 (1.49 g, 4.03 mmol) and 215 (1.4 g, 4.03 mmol) in CH2Cl2 (33 mL) was added EEDQ16 (0.57 g, 4.43 mmol) at 0 °C. The reaction mixture was stirred at 27 °C for 17 h. After the reaction, the solution was concentrated and the residue was purified by reprecipitation (CH2Cl2/ EtOH) to give 3 (0.95 g, 34% yield); Rf ) 0.5-0.7 (methanol). 1H NMR (δ, CDCl3) 1.90 (3H, s, NHCOCH3), 2.05, 2.06, 2.07 (9H, 3s, OCOCH3), 2.13-2.30 (4H, CHCH2CH2CO), 3.67-3.78 (1H, m, H-5), 4.02-4.22 (2H, m, H-2 and 6a), 4.28 (1H, dd, 4.2, 12.4 Hz, H-6b), 4.354.44 (1H, m, CHCH2CH2CO), 4.96-5.16 (7H, m, H-1β, H-3, H-4, CH2Ph, CH2Ph), 5.66 (1H, d, J ) 7.8 Hz, CONH), 5.90 (1H, d, J ) 8.0 Hz, CONH), 7.04 (1H, d, J ) 8.0 Hz, CONH), 7.35 (10H, br, aromatic). FT IR (cm-1, KBr) 3315 (νN-H), 2923 (νC-H), 1743 (νCdO, ester), 1664 (νCdO, amide I), 1534 (δN-H, amide II), 1241 and 1049 (νC-O), 740 (νC-O, phenyl). Preparation of Boc-Gln[β-D-GlcNAc(Ac)3]-Aib-OBn (7). To a solution of 5 (824 mg, 1.43 mmol) and HOBtH2O (241 mg, 1.58 mmol) in DMF (5.5 mL) was added DCC (325 mg, 1.58 mmol) at 0 °C. After addition of TsOH-HAib-OBn (6)17 (523 mg, 1.43 mmol), the reaction mixture was neutralized by addition of NMM (193 µL, 1.76 mmol) and stirred at 27 °C for 93 h. After the reaction, the solution was concentrated and the residue was redissolved in ethyl acetate. Dicyclohexylurea was removed by filtration and the filtrate was washed with saturated NaCl, 5% KHSO4, saturated NaCl, 5% NaHCO3, and saturated NaCl aqueous solutions successively and dried over MgSO4. The organic layer was evaporated to give 7 (0.67 g, 62% yield); Rf ) 0.38 (ethyl acetate). 1H NMR (δ, CDCl3) 1.42 (9H, s, CH3), 1.50, 1.58 (6H, s, CH3), 1.94 (3H, s, NHCOCH3), 2.05, 2.06, 2.08 (9H, 3s, OCOCH3), 1.96-2.22 (4H, m, CHCH2CH2N), 3.70-3.78 (1H, m, H-5), 4.03-4.29 (4H, m, H-2, 6a, 6b, CH), 5.02-5.19 (5H, H-1, 3, 4, CH2Ph), 5.43 (1H, d, J ) 8.4 Hz, CONH), 6.04 (1H, d, J ) 9.0 Hz, CONH), 7.34 (5H, br, aromatic), 7.98-8.01 (2H, NHCO). FT IR (cm-1, KBr) 3438 (νN-H), 2982 (νC-H), 1747 (νCdO, ester), 1665 (νCdO, amide I), 1540 (δN-H, amide II), 1242 and 1047 (νC-O), 750 (νC-O, phenyl). Preparation of Boc-Glu(OMe)-Gln[β-D-GlcNAc(Ac)3]Aib-OBn (10a), Boc-Lys(Ac)-Gln[β-D-GlcNAc(Ac)3]-AibOBn (10b), and Boc-Ala-Gln[β-D-GlcNAc(Ac)3]-Aib-OBn (10c). As shown in Scheme 1, protected tripeptide 10a was prepared by a similar procedure to that for dipeptide 7 (79% yield); Rf ) 0.51 (ethyl acetate). MS (MALDI-TOF) (m/z) calcd for [M+Na]+, 916.92; found 917.41. 1H NMR (δ, CDCl3) 1.43 (9H, s, CH3), 1.50, 1.56 (6H, s, CH3), 1.93 (3H, s, NHCOCH3), 2.04, 2.05, 2.09 (9H, 3s, OCOCH3), 1.592.48 (8H, CH2CH2COOCH3, CH2CH2CONH), 3.69 (3H, s, OCH3), 3.72-3.78 (1H, m, H-5), 4.03-4.31 (3H, m, H-2, 6a, 6b), 4.35-4.46 (1H, CHCH2CH2CONH), 4.65-4.74 (1H,

2336

Biomacromolecules, Vol. 6, No. 4, 2005

Takasu et al.

Scheme 1. Synthesis of N-Glycosylated Tripeptide Containing Aib Residea

a Reagents and conditions: (a) EEDQ, CH Cl , r.t., 112 h, 60%; (b) H /Pd-C, mehtanol/water (9/1, v/v), 40°C, 12 h, 95%; (c) Boc O, water/1,4-dioxane 2 2 2 2 (1/1, v/v), r.t., 60 h, 86%; (d) DCC/HOBt, DMF, r.t., 93 h, 62%; (e) HCOOH, r.t., 18 h, 88%; (f) DCC/HOBt, CH2Cl2 (for 10a) or DMF(for 10b and 10c), r.t., 45 h, 10a: 79%, 10b: 51%, 10c: 73%; (g) H2/Pd-C, methanol/water (9/1, v/v), r.t., 18 h, 11a: 97%, 11b: 95%, 11c: 98%; (h) NpOH/DCC, CH2Cl2, r.t., 36 h, 12a: 79%, 12b: 99%, 12c: 68%; (i) HCl, 1,4-dioxane, r.t., 8 h, 13a: 99%, 13b: 73%, 13c: 78%.

Scheme 2. Polymerization of Glycosylated Tripeptide 13 by Active Ester Method

m, CHCH2CH2CONH), 5.02-5.20 (5H, H-1,3,4,-CH2Ph), 5.28 (1H, brd, 6.8 Hz, CONH), 6.18 (1H, brd, 8.8 Hz, CONH), 7.15 (1H, brd, 7.4 Hz, CONH), 7.34 (5H, br, aromatic), 7.92-7.96 (2H, br, CONH). FT IR (cm-1, KBr) 3315 (νN-H), 2931 (νC-H), 1742 (νCdO, ester), 1660 (νCdO, amide I), 1555 (δN-H, amide II), 1371 (νC-H), 1243 and 1048 (νC-O). For 10b (51% yield); Rf ) 0.85 (methanol). MS (MALDI-TOF) (m/z) calcd for [M+Na]+, 943.98; found

944.57. 1H NMR (δ, CDCl3) 1.44 (9H, s, CH3), 1.51, 1.54 (6H, s, CH3), 1.94, 2.00 (6H, 2s, NHCOCH3), 2.04, 2.05, 2.09 (9H, 3s, OCOCH3), 1.63-2.42 (10H, m, CH2CH2CH2CH2NHCOCH3, CH2CH2CONH), 3.17-3.22 (2H, br, CH2NHCOCH3), 3.71-3.82 (1H, br, H-5), 3.97-4.33 (4H, m, H-2, 6a, 6b, CHCH2CH2CONH), 4.38-4.48 (1H, m, CHCH2CH2CONH), 4.99-5.38 (5H, H-1,3,4, -CH2Ph), 6.21 (1H, br, CONH), 6.42 (1H, brd, 9.2 Hz, CONH), 7.18 (1H, brd,


8.4 Hz, CONH), 7.33 (5H, br, aromatic), 7.94-8.12 (3H, br, CONH). FT IR (cm-1, KBr) 3315 (νN-H), 2931 (νC-H), 1742 (νCdO, ester), 1660 (νCdO, amide I), 1555 (δN-H, amide II), 1371 (δC-H), 1243 and 1048 (νC-O). For 10c (73% yield); Rf ) 0.50 (ethyl acetate). MS (MALDI-TOF) (m/z) calcd for [M+Na]+, 844.86; found 845.30. 1H NMR (δ, CDCl3) 1.33 (1H, d, 6.6 Hz, CH3CH), 1.44 (9H, s, CH3), 1.50, 1.55 (6H, s, CH3), 1.93 (3H, s, NHCOCH3), 2.04, 2.05, 2.09 (9H, 3s, OCOCH3), 1.60-2.20 (4H, br, CH2CH2CONH), 3.713.80 (1H, br, H-5), 4.06-4.30 (4H, m, H-2, 6a, 6b, CHCH2CH2CONH), 4.40 (1H, q, CH3CH), 5.05-5.19 (6H, H-1,3,4, -CH2Ph, NHCO), 6.23 (1H, d, 8.8 Hz, CONH), 7.00 (1H, d, 8.0 Hz, CONH), 8.04 (1H, d, 10.0 Hz, CONH), 7.32 (5H, br, aromatic), 8.10 (1H, s, CONH). FT IR (cm-1, KBr) 3325 (νN-H), 2930 (νC-H), 1746 (νCdO, ester), 1654 (νCdO, amide I), 1572 (δN-H, Boc), 1368 (δC-H), 1245 and 1047 (νC-O). Preparation of Boc-Glu(OMe)-Gln[β-D-GlcNAc(Ac)3]Aib-ONp (12a) and Boc-Lys(Ac)-Gln[β-D-GlcNAc(Ac)3]Aib-ONp (12b). p-Nitrophenyl esterification1,18 of 11 was established in a good yield (79-99%). For 12a (79% yield); 1H NMR (δ, CDCl ) 1.44 (9H, s, CH ), 1.68 (6H, s, CH ), 3 3 3 1.97 (3H, s, NHCOCH3), 2.05, 2.07, 2.10 (9H, 3s, OCOCH3), 1.90-2.48 (8H, m, CH2CH2CO, CH2CH2CONH), 3.70 (3H, s, OCH3), 3.72-3.80 (1H, m, H-5), 4.07-4.25 (4H, m, H-2, 6a, 6b, CHCH2CH2CONH), 4.46-4.57 (1H, m, CHCH2CH2CO), 5.09-5.16 (3H, m, H-1β, 3, 4), 5.34 (1H, brd, 6.4 Hz, CONH), 6.18 (1H, d, 9.1 Hz, CONH), 7.27-7.30 (3H, aromatic and CONH), 8.10 (1H, d, 7.9 Hz, CONH), 8.248.28 (3H, m, aromatic and CONH). For 12b (99% yield); Rf ) 0.66 (ethyl acetate). MS (MALDI-TOF) (m/z) calcd for [M+Na]+, 974.96; found 975.53. 1H NMR (δ, DMSO) 1.61 (9H, s, CH3), 1.74 (6H, s, CH3), 1.94, 2.02 (6H, 2s, NHCOCH3), 2.15, 2.20, 2.23 (9H, 3s, OCOCH3), 1.79-2.52 (10H, CH2CH2COOCH3, CH2CH2CH2CH2NHCOCH3), 3.974.59 (6H, m, H-2, 5, 6a, 6b, CHCH2CH2CONH, CHCH2CH2CH2CH2NHCOCH3), 5.04 (1H, t, 9.7 Hz, H-4), 5.265.48 (2H, H-1 and 3), 5.78 (1H, d, 8.0 Hz, CONH), 6.81 (1H, brd, 9.0 Hz, CONH), 7.58 (2H, d, 9.0 Hz, aromatic), 7.91, 8.06 (2H, br, CONH), 8.53 (2H, d, 9.2 Hz, aromatic), 8.63 (1H, brd, 9.6 Hz, CONH), 8.85 (1H, br, CONH). FT IR (cm-1, KBr) 3326 (νN-H), 2932 (νC-H), 1751 (νCdO, ester), 1655 (νCdO, amide I), 1526 (δN-H, amide II), 1368 (νC-H), 1243 and 1046 (νC-O). For 12c (68% yield); Rf ) 0.50 (ethyl acetate). MS (MALDI-TOF) (m/z) calcd for [M+Na]+, 875.83; found 876.29. Preparation of HCl-H-Glu(OMe)-Gln[β-D-GlcNAc(Ac)3]-Aib-ONp (13a). In a tube, 12a (115 mg, 0.13 mmol) was dissolved in 1,4-dioxane (3.2 mL) and 4 N HCl/1,4dioxane (0.5 mL) was added to the solution at 0 °C. The mixture was stirred at 27 °C for 8 h. Excess diethyl ether was added to the solution. The precipitate was washed with diethyl ether and dried to give 13a (105 mg, 99%). Polymerization of HCl-H-Glu(OMe)-Gln[β-D-GlcNAc(Ac)3]-Aib-ONp (13a) and HCl-H-Lys(Ac)-Gln[β-DGlcNAc(Ac)3]-Aib-ONp (13b). Tripeptide 13a (136 mg, 0.16 mmol) and HOAt (4.3 mg, 0.032 mmol) were dissolved in dimethyl sulfoxide (82 µL), and the solution was vigorously stirred. Triethylamine (34 µL, 0.25 mmol) was added.


Polymerization was allowed to proceed at 50 °C for 5 days. To the solution was added excess water to precipitate the polypeptide. The precipitate was washed with water and diethyl ether successively and dried to give pale yellow powder poly(13a) (81 mg, yield, 75%). 1H NMR (δ, DMSO) 1.34-1.41 (6H, br, CH3), 1.73 (3H, s, NHCOCH3), 1.90, 1.95, 1.97 (9H, COCH3), 1.82-2.31 (8H, br, CH2CH2CO, CH2CH2CONH), 3.55 (4H, OCH3, H-5), 3.72-4.01 (4H, m, H-2, 6a, 6b, CHCH2CH2CONH), 4.08-4.25 (1H, br, CHCH2CH2CO), 4.81 (1H, br, H-4), 5.05-5.19 (2H, br, H-1 and 3), 7.60-8.19, 8.20-8.73 (5H, br, NHCO). FT IR (cm-1, KBr) 3335 (νN-H), 2939 (νC-H), 1746 (νCdO, ester), 1661 (νCdO, amide I), 1539 (δN-H, amide II), 1239 and 1046 (νC-O). For poly(13b), FT IR (cm-1, KBr) 3309 (νN-H), 2941 (νC-H), 1747 (νCdO, ester), 1659 (νCdO, amide I), 1541 (δN-H, amide II), 1242 and 1045 (νC-O). Deacetylation of Poly(13). Deprotection of acetylated glycopeptide poly(13a) was performed according to previous reports.10 To an ice-cooled solution of poly(13a) (13 mg, 0.101 mmol) in methanol (6.0 mL) was added 391 µL (8.06 mmol) of hydrazine monohydrate. After mixing at 27 °C for 6 h, 1.18 mL (16.11 mmol) of acetone was added to the solution to quench hydrazine with cooling at 0 °C in an icebath. The mixture was evaporated, and then the product was purified by repeated reprecipitations from water to ethanol. After drying in vacuo, periodic glycopeptide 14a was isolated in 53% yield (30 mg). 1H NMR (D2O) δ 1.37-1.46 (6H, br, CH3), 1.98 (3H, s, NHCOCH3), 2.04-2.66 (8H, br, CH2CH2CO, CH2CH2CONH), 3.40-3.90 (5H, br, H-2, 5, 6a, 6b, CHCH2CH2CO), 3.67 (3H, s, OCH3), 4.00-4.30 (1H, CHCH2CH2CO), 5.07 (1H, brd, 9.4 Hz, H-1β). FT IR (cm-1, KBr) 3425 (νN-H), 2928 (νC-H), 1733 (νCdO, ester), 1655 (νCdO, amide I), 1540 (δN-H, amide II), 1078 (νC-O). For 14b, 1H NMR (D2O) δ 1.20-1.54 (6H, br, CH3), 1.90, 1.93 (6H, 2s, NHCOCH3), 1.55-2.51 (10H, br, CH2CH2CO, CH2CH2CH2CH2NHCO), 3.09 (2H, CH2NHCOCH3), 3.294.29 (7H, H-2, 3, 4, 5, 6a, 6b, CHCH2CH2CH2CH2NHCO), 5.02 (1H, brd, 9.4 Hz, H-1β). FT IR (cm-1, KBr) 3317 (νN-H), 2936 (νC-H), 1659 (νCdO, amide I), 1542 (δN-H, amide II), 1049 (νC-O). For 14c, 1H NMR (D2O) δ 1.10 (3H, d, 6.2 Hz, CH3CH), 1.40 (6H, br, CH3), 1.93 (3H, s, NHCOCH3), 2.00-2.40 (4H, br, CH2CH2CO), 3.35-4.27 (8H, H-2, 3, 4, 5, 6a, 6b, CHCH3, CHCH2CH2CONH), 5.03 (1H, d, 9.0 Hz, H-1β). FT IR (cm-1, KBr) 3296 (νN-H), 2937 (νC-H), 1656 (νCdO, amide I), 1540 (δN-H, amide II), 1052 (νC-O). Recognition of Glycopeptides by WGA Lectin. Fluorescence spectra (at 520 nm) of FITC-labeled WGA lectin were recorded on a JASCO FP-777 spectrometer with excitation at 490 nm in the range of 3-20 µM GlcNAc residue concentration. The solutions were contained in 10 mm quartz cells maintained at 27 °C. The concentration of WGA in 4-(2-hydroxyethyl)-1-piperazineethansulfonic acid (HEPES) buffer (5mM, pH 7.2) was 2 µM. Results and Discussion In our strategy, polymerizable glycosylated tripeptides were synthesized in solution, preferentially by a step-by-

2338


Takasu et al.

Figure 1. (a) Repeating unit of a new type of periodic glycopeptide and (b) schematic representation of helical glycopeptide.

step approach, starting from a triacetylated N-acetyl-Dglucosamine (GlcNAc) substituted L-glutamine, Z-Gln [β-DGlcNAc(Ac)3]-OBn (3) (Z, benzyloxycarbonyl; Bn, benzyl) (Scheme 1). The glycosylated glutamine 3 was prepared by the condensation reaction of protected L-glutamic acid, Z-LGlu(OH)-OBn (1), and triacetylated GlcNAc amine 215 using EEDQ as the coupling reagent16 (yield, 60%), in which the coupling efficiency was higher than that using the DCC/ HOBt method (yield, 14%). The glycosylamine 2 was prepared by hydrogenation of β-glycosyl azide.15 The β-selective glycosidation was confirmed by 1H NMR resonance ascribed to anomeric proton of 3 (H-1β, at 5.10 ppm, J1,2 ) 9.2 Hz).15 The sugar-substituted L-glutamine 5, which was prepared by hydrogenation of 315 and subsequent N-terminal protection, was coupled with H-Aib-OBn to afford a new glycodipeptide 711 in 62% yield using DCC/ HOBt coupling strategy. Deprotection of Boc, followed by coupling with Boc-L-Glu(OMe)-OH (9a), Boc-L-Lys(Ac)OH (9b), and Boc-L-Ala-OH (9c) by the DCC/HOBt method gave terminally protected, tripeptide, Boc-X-Gln[β-DGlcNAc(Ac)3]-Aib-OBn [10a: X)Boc-L-GluOMe, 10b: X)Boc-L-Lys(Ac), 10c: X)Boc-L-Ala] (yield, 51-79%). To obtain glycopeptide, we tried an active ester method according to our previous report dealing with the synthesis of the O-glycoslated peptide.10 p-Nitrophenyl ester is one of the most important active esters.19 Therefore, we prepared HCl-H-L-Glu(OMe)-L-Gln[β-D-GlcNAc(Ac) 3 ]-AibONp (13a),11 HCl-H-L-Lys(Ac)-L-Gln[β-D-GlcNAc(Ac)3]Aib-ONp (13b), and HCl-H-L-Ala-L-Gln[β-D-GlcNAc(Ac)3]-Aib-ONp (13c) (yield, 73-99%).

Polymerizations of 13a were carried out in DMSO at 50 °C for 5 days using TEA as the bases. The reaction mixture became viscous for ca. 6 h after addition of TEA. After the polycondensation, poly(13a) was purified by reprecipitation using the DMF/Et2O system. The results are summarized in Table 1. The structure was confirmed by IR and NMR measurements. Addition of HOAt improved the yield. Although polymerization of 13a gave corresponding poly(13a) with Mn of 2.56 × 103 (45% yield) without HOAt, the TEA/HOAt system afforded the corresponding peptide with Mn of 2.18 × 103 (75% yield, run 2). The result was ascribed to activation of the C-terminal carbonyl of 13a by HOAt, and similar activation was reported by Yamamoto et al. in polycondensation of octapeptide active esters (X-GlyTyr-Ser-Ala-Gly-Tyr-Lys-ONp, X ) Ala, Thr).20 The conformation was investigated by FTIR and CD spectroscopies, which are the most extensively used tools to elucidate secondary structure of peptides. FTIR spectrum of poly(13a) (run 1 in Table 1, Mn ) 2.6 × 103) showed the absorptions at 1661 and 1539 cm-1 (Table 2), which are assigned to amide I and II ascribed to R-/310-helical polypeptide,12 respectively. As shown in Figure 2, CD spectra of acetylated GlcNAc-substituted periodic glycopeptide, poly(13a), in methanol showed the two negative maxima (208 and 227 nm) and the intensities are very close, indicating an R-helical conformation.21 The result is quite different from that (single minimum at 206 nm) of O-glycosylated peptide1 having a similar sequence (Figure 2). This is due to the spacer between the sugar residue and the peptide backbone. We expected that the sugar moiety in O-glycosylated peptide



Table 1. Polymerization of Glycosylated Tripeptide 13 by Active Ester Methoda monomer run 1 2 3 4 5 6

13a 13a 13b 13b 13c 13c

deacetylationc

yield

mg (mmol)

[M]0

activator

mg (%)

111 (0.129) 136 (0.158) 81 (0.091) 234 (0.264) 76 (0.096) 94 (0.119)

1.2 1.2 1.2 0.8 1.2 1.0

none HOAte HOAte HOAte HOAte HOAte

40 (45) 81 (75) 17 (25) 53 (27) 60 (96f) 74 (97f)

Mnd

×

10-3

Mw/Mnd

2.56 2.18 2.72 2.51 0.97 1.41

2.72 5.30 2.15 1.60 2.53 1.74

yield (%)

Mnd × 10-3

Mw/Mnd

47 53 89 35 38 59

4.6 4.4 5.5 4.4 1.8 1.6

1.2 1.2 1.3 1.2 1.2 1.2

14a 14a 14b 14b 14c 14c

a In DMSO; temp., 50 °C, for 5 days. Feed molar ratio of monomer to amine ) 1.6. b Determined by SEC in DMF relative to polystyrene. c [H2NNH2]0 ) 1.3 mol/L in methanol at 0 °C for 6 h, [H2NNH2]0/[acetyl group]0 ) 26. d Determined by SEC in 0.05 M K2HPO4 aq. relative to PEG. e 3H-1,2,3-Triazolo[4,5-b]pyridin-3-ol (HOAt). f Crude (without reprecipitation).

Table 2. Peak Position of Amide I and II Infrared Absorption Bands in FT-IRa and ∆222 Value in CD Measurementb of Glycopeptide poly(13) IR (cm-1) poly(13a) poly(13b) poly(13b) poly(13c) poly(13c) a

after deacetylation IR

∆222

amide I

amide II

methanol

1661 1659

1539 1541

-1.69 (222) -1.31 (223)

1665 1662

1540 1539

-0.85 (225) -0.78 (225)

14a (n ) 8) 14b (n ) 7) 14b (n ) 9) 14c (n ) 3) 14c (n ) 4)

∆222

amide I

amide II

methanol

TFE

water

1652 1653 1653 1656 1655

1540 1542 1542 1540 1543

-5.43 (223) -3.22(223) -2.68(223) -1.08 (222) -1.43 (222)

-4.69 (224) -1.27 (223)

-0.94(227) -1.13(225) -1.77(224) -0.34 (228) -0.50 (227)

-1.51(226)

Measured by KBr method. b 0.5-0.6 M solution at 27 °C.

disrupts helix formation by steric interference, as it is suited closer to the backbone. Concerning the helix content, the ∆222 value calculated with respect to the amide group is a good measure for the helix content.22 The negative Cotton effect around 222 nm (∆222) of poly(13a) (run 2, Table 1) was -1.69 at 27 °C, corresponding to a 14-16% helix content, although that was smaller than that expected for a long polypeptide chain in a 100% helix conformation (-10.6 to -12.1).22 These results indicated that poly(13a) showed R-helix preference both in solid and solution (methanol). Acetyl protecting groups of poly(13a) consisting of glycosylated glutamic acid could be removed very safely and with no racemization using hydrazine in methanol at 27 °C for 6 h to give periodic glycopeptide 14a (47-53% yield). Complete deprotection was confirmed by IR and 1H NMR measurements. Unexpectedly, methyl ester of Glu(OMe) units in poly(13a) was not deprotected in this experimental condition. It might be due to the helical structure of poly(13a) in which the methyl ester groups were surrounded by acetylated sugar moieties, so that hydrazine could not attack the methyl ester.10 Mns of periodic glycopeptide 14a estimated by SEC using poly(ethylene oxide) standards were 4.4 × 103-4.6 × 103 (ca. 24-27 residues; runs 1and 2 in Table 1). The Mns seem to be close to absolute Mn, because Mn of -[Glu(OMe)-Ser(β-D-GlcNAc)-Aib]n- (n ) 7) estimated from the SEC measurement coincided well with that determined by MALDI-TOF mass measurement in our previous report.10 The result indicated that Mn of poly(13a) (2.18 × 103-2.56 × 103) calculated by SEC in DMF (polystyrene standards) were underestimated. Deacetylation of poly(13a) (n ) 8) induced CD spectra change in methanol, i.e., negative Cotton effect at 222 nm reflecting R-helix content increased remarkably after deacetylation (∆222, from -1.69 to -5.43; helix content, 14-16% to 45-51%; Figure 2B). It seems that the steric hindrance

of acetylated sugar derivative destabilizes the R-helical conformation. The result does not contradict the previous discussion about the spacer between the peptide backbone and the sugar moiety. In general, glycopeptides tend to prefer random coil structures to helical structures in water.23 The results reveal that the Aib unit in glycopolymer 14a seems to act as a strong inducer for the helical conformation. This is the first example of the periodic N-glycosylated peptide having a regular helical backbone as far as we know. In water, solvent interactions with the sugar residue are strong enough to disrupt the H-bonding found in the folded structure (dot line in Figure 2). Therefore, the ∆222 value (-0.94) was lower than that in methanol (-5.43). This is reasonable since it is known that the sugar residue interacts strongly with water molecules through formation of an ordered H-bonded structure, which can counterbalance the loss of amide H-bonding. On the other hand, TFE has a marked potential to induce the formation of R-helical structures in peptides.24 In the measurement using TFE as the solvent, the ∆222 was -4.69 (dashed line in Figure 2), which was comparable to that in methanol. In the CD spectra in water (Figure 2), ∆206 is little larger than ∆222, indicating that N-glycosylated peptides 14 tend to aggregate.25 The aggregation behavior was estimated by AFM analysis of 14a. 0.5 mM MeOH and aqueous solutions were prepared and deposit on mica surface, respectively, whereas ca. 50 nm of globular particles were observed from the methanol solution, in which the particle size distribution was relatively narrow (Figure 3). On the other hand, large particles (>500 nm) were observed from the aqueous solution, although water is a better solvent than methanol for 14a. The results supported that 14a aggregates in water, which coincides with the CD analysis. Polymerizations of 13b and 13c were also carried out in the same procedure in order to evaluate the effect of the sequence on the polymerization behavior and conformation.

2340


Takasu et al.

Figure 2. CD (top) and UV (bottom) spectra of poly(13a) (before deacetylation) and 14a (after deacetylation) at 27 °C in methanol (solid line), water (dot line), and TFE (dashed line).

Figure 3. AFM image of periodic glycopeptides 14a (24 residues, n ) 8) in methanol.

Under same condition, poly(13b)s having Mn’s of 2.72 × 103 and 2.51 × 103 were obtained (runs 3 and 4 in Table 1). On the other hand, polymerization of 13c afforded oligo-

(13c) with a shorter chain length (Mn ) 0.97 × 103-1.41 × 103), because solubility of 13c in DMSO is low and the polymerization proceeded heterogeneously (runs 5 and 6 in Table 1). Poly(13b) and poly(13c) were also deacetylated using hydrazine in methanol at 27 °C for 6 h to give periodic glycopeptide 14b and 14c (35-89% yield). Complete deprotection was confirmed by IR, 1H NMR. As shown in Figure 4, we could observe double minima in both of the CD spectra indicating R-helical conformation in solution. The ∆222 values in some solvents are summarized in Table 2. The R-helix content depends on the solvent and sequence. The contents of 14bs (n ) 7 and 9) in methanol are 2730% (∆222 ) -3.22) and 22-25% (∆222 ) -2.68), and those in water were 9-11% ((∆222 ) -1.13) and 15-17% (∆222 ) -1.77). The helical content of 14c (n ) 4, Mn ) 1.8 × 103) in water was 4-5% (∆222 ) -0.50). The low content was due to the low molecular weight. Among the N-glycosylpeptides 14s, helix content of 14b containing Lys(Ac) was the highest in water, whereas the R-helical conformation of 14a containing Glu(OMe) was the most stable in MeOH and TFE.



Figure 4. CD and UV spectra of 14b (top) and 14c (bottom) in methanol (left side) and water (right side) at 27 °C. Table 3. Interaction of Glycopeptides with FITC-WGA Lectin

GlcNAc 14a 14b 14c O-glycosylated peptidee

MW

Ka [M-1]a

221 4400b 4400b 1840b 3600b

6.9 × 102c (6.8 × 102d) 6.6 × 104 5.5 × 104 9.3 × 104 7.4 × 104

a Evaluated from fluorescence intensity of FITC-WGA at 520 nm depending on sugar residue concentration (at 27 °C) using Hill plot. b Molecular weight (MW) is calculated from Mnestimated by SEC measurements. c Reference 27. d Reference 6g. e -[Glu(OMe)-Ser(b-DGlcNAc)-Aib]n- (n ) 7) synthesized in ref 10.

Interaction of glycopeptide with a lectin was evaluated by fluorescence spectroscopy using FITC-labeled WGA lectin, according to previous reports.6e,g,9,26 WGA is one of the well-studied plant lectins and specifically recognizes GlcNAc, its β-1,4-oligomers, and N-acetyl neuraminic acid.27 Changes of fluorescence intensity of FITC-labeled WGA lectin at 520 nm (Ex ) 490 nm) are plotted as a function of GlcNAc concentration ([Sugar]) with eq 1 (Hill plot6e,g,9,26).

The association constants (Kas) calculated from the plot are summarized in Table 3. The Kas of 14a-14c and Oglycosylated polypeptide are 5.5 × 104-9.3 × 104 M-1, which were higher than that of GlcNAc but lower than that of poly(acrylamide)6e having pendant GlcNAc (Ka: order of 107 × 108 M-1). The distance of GlcNAc-binding site in WGA lectin is reported to be 31 Å. As the inter-sugar distance of this glycopeptides containing Aib is estimated to be ca. 10-20 Å,1 it is expected that all the saccharides cannot bind to the binding site log(Y/I - Y) ) n log[sugar] + n log Ka (Y ) ∆F/∆Fmax) (1) Conclusion In this paper, we describe the synthesis of a new type of N-glycosylated periodic peptide 14 (12-27 residues) having a GlcNAc residue. Conformational analysis of the glycopeptide 14 by CD, FT-IR spectroscopies indicated that it has

2342


an R-helical conformation in solid and protic solvent (methanol and water) in which three-dimensional regular arrangement of N-glycoside along the R-helical peptide was established. The helix content changed depending on the solvent, spacer between sugar, and the peptide sequence. The periodic glycopeptides were recognized by WGA lectin and the Kas were higer than that of GlcNAc. This fundamental data would provide a new guideline to regulate threedimensional arrangement of carbohydrate ligand and would provide a significant information concerning conformation of glycoproteins and glycopeptides model. Acknowledgment. The authors express their sincere gratitude to Professor Y. Inai and Mr. S. Tahara, Nagoya Institute of Technology for their fruitful discussion and help in measuring the MALDI-TOF mass. This work was funded by the Ministry of Education, Science and Culture of Japan (Grant-in-Aid for Development Scientific Research, No. 16750095). References and Notes (1) (a) Kobata, A. Eur. J. Biochem. 1992, 209, 483. (b) Dwek, R. A. Chem. ReV. 1996, 96, 683. (2) A review of glycoproteins: Montreuil, J. In ComprehensiVe Biochemistry; Neuberger, A., van Deenen, L. L. M. Eds.; Elsevier: Amsterdam 1982; 19B11, p. 1. (3) Johansen, P. G.; Marshall, R. D.; Neuberger, A. Biochem. J. 1961, 78, 518. (4) Wang, J.-Z.; Grundke-Iqbal, I.; Iqbal, K. Nat. Med. 1996, 2, 871. (5) Landberg, E.; Pahlsson, P.; Lundblad, A.; Arnetorp, A.; Jeppsson, J. O. Biochem. Biophys. Res. Commun. 1995, 210, 267. (6) (a) Neoglycoconjugates: Preparation and Applications; Lee, Y. C., Lee, R. T., Eds.; Academic Press: San Diego, CA, 1994. (b) Kobayashi, A.; Akaike, T.; Kobayashi, K.; Sumitomo, H. Makromol. Chem. Rapid Commun. 1986, 7, 645. (c) Roy, R.; Tropper, F. C. J. Chem. Soc., Chem. Commun. 1988, 1058. (d) Manning, D. D.; Hu, X.; Beck, P.; Kiessling, L. L. J. Am. Chem. Soc. 1997, 119, 3161. (e) Nishimura, S.-I.; Furuike, T.; Matsuoka, K.; Maruyama, K.; Nagata, K.; Nishi, N.; Tokura, S. Macromolecules 1994, 27, 4876. (f) Wulff, G.; Schmid, J.; Venhoff, T. Macromol. Chem. Phys. 1996, 197, 259. (g) Yamada, K.; Minoda, M.; Miyamoto, T. Macromolecules 1999, 32, 3553. (7) (a) Kojima, N.; Hakomori, S. J. Biol. Chem. 1991, 266, 17552. (b) Matsuura, K.; Oda, R.; Kitakouji, H.; Kiso, M.; Kitajima, K.; Kobayashi, K. Biomacromolecules 2004, 5, 937. (8) (a) Aoi, K.; Ito, K.; Okada, M. Macromolecules 1995, 28, 5391. (b) Aoi, K.; Tsutsumiuchi, K.; Yamamoto A.; Okada, M. Tetrahedron 1997, 53, 15415. (9) (a) Hasegawa, T.; Kondoh, S.; Matsuura, K.; Kobayashi, K. Macromolecules 1999, 32, 6595. (b) Matsuura, K.; Hibino, M.; Yamada, Y.; Kobayashi, K. J. Am. Chem. Sci. 2001, 11, 1281.

Takasu et al. (10) Takasu, A.; Houjyou, T.; Inai, Y.; Hirabayashi, T. Three- or TwoDimensional Arrangement of Sugar Residues along a Helical Polypeptide Backbone 1. Biomacromolecules 2002, 3, 775. (11) As a preliminary result: Horikoshi, S.; Takasu, A.; Houjyou, T.; Inai, Y.; Hirabayashi, T. Polym. Prepr. (Am. Chem. Soc. DiV. Polym.) 2002, 43 (2), 1111. (12) Kennedy, D. F.; Crisma, M.; Toniolo, C.; Chapman, D. Biochemistry 1991, 30, 6541. (13) (a) Toniolo, C.; Polese, A.; Formaggio; F.; Crisma, M.; Kamphuis, J. J. Am. Chem. Soc. 1996, 119, 10278. (b) Yorder, G.; Polese, A.; Silva, R. A. G. D.; Formaggio, F.; Crisma, M.; Broxterman, Q. B.; Kamphuis, J.; Toniolo, C.; Keiderling, T. A. J. Am. Chem. Soc. 1997, 118, 22744. (c) Formaggio, F.; Crisma, M.; Rossi, P.; Scrimin, P.; Kaptein, B.; Broxterman, Q. B.; Kamphuis, J. Chem. Eur. J. 2000, 6, 4498. (14) (a) Inai, Y.; Kurokawa, Y.; Hirabayashi, T. Biopolymers 1999, 49, 551. (b) Inai, Y.; Ashitaka, S.; Hirabayashi, T. Polym. J. 1999, 31, 246. (c) Inai, Y.; Kurokawa, Y.; Hirabayashi, T. Macromolecules 1999, 32, 4575. (e) Inai, Y.; Ishida, Y.; Tagawa, K.; Takasu, A.; Hirabayashi, T. J. Am. Chem. Soc. 2002, 124, 2466. (e) Inai, Y.; Komori, H.; Takasu, A.; Hirabayashi, T. Biomacromolecules 2003, 4, 122. (15) (a) Bolton, C. H.; Jeanloz, R. W. J. Org. Chem. 1963, 28, 3228. (b) Cowley, D. E.; Hough, L.; Peach, C. M. Carbohydr. Res. 1971, 19, 231. (16) (a) Kunz, H.; Waldman H. Angew. Chem., Int. Ed. Engl. 1985, 24, 883. (b) Kunz, H. Angew. Chem., Int. Ed. Engl. 1987, 26, 294. (17) Balasubramanian, T. M.; Nancy, C. E.; Kendrick, M.; Taylor, M.; Marshall, G. R.; Hall, J. E.; Vodyanoy, I.; Reusser, F. J. Am. Chem. Soc. 1981, 103, 6127. (18) Paredes, N.; Rodriguez-Galan, A.; Puggali, J. Polymer 1996, 37, 4175. (19) (a) Papaka, R. S.; Urry, D. W. Int. J. Peptide Protein Res. 1978, 11, 97. (b) Papaka, R. S.; Okamoto, K.; Urry, D. W. Int. J. Peptide Protein Res. 1978, 11, 109. (20) Yamamoto, H.; Sakai, Y.; Ohkawa, K. Biomacromolecules 2000, 1, 543. (21) Manning, M.; Woody, R. W. Biopolymers 1991, 31, 569. (22) (a) Sisido, M. Macromolecules 1989, 22, 3280. (b) Scholtz, J. M.; Qian, H.; York, E. J.; Stewart, J. M.; Baldwin, R. L. Biopolymers 1991, 31, 1463. (23) (a) Filira, F.; Biondi, L.; Scolaro, B.; Foffani, M. T.; Mammi, S.; Peggion, E.; Rocchi, R. Int. J. Biol. Macromol. 1990, 12, 41. (b) Aoi, K.; Tsutsumiuchi, K.; Okada, M. Macromolecules 1994, 27, 875. (24) (a) Nelson, J. W.; Kallenbach, N. R. Protein: Struct., Funct., Genet. 1986, 1, 211. (b) Buck, M.; Radford, S. E.; Dobson, C. M. Biochemistry 1993, 32, 669. (b) Hong, D.-P.; Hoshino, M.; Kuboi, R.; Goto, Y. J. Am. Chem. Soc. 1999, 121, 8427. (25) Maeda, H.; Kato, H.; Ikeda, S. Biopolymers 1984, 23, 1333. (26) (a) Lotan, R.; Sharon, N. Biochem. Biophys. Res. Comun. 1973, 55, 1340. (b) Privat, J.-P.; Delmote, F.; Mialonier, G.; Bouchard, P.; Monsigny, M. Eur. J. Biochem. 1974, 47, 5. (27) Nagata, Y.; Burger, M. M. J. Biol. Chem. 1974, 249, 3116.

BM0502563

Three-Dimensional Arrangement of Sugar Residues along Helical

Recommend Documents