`
`813
`
`HYCRON, an Allylic Anchor for High-Efficiency Solid Phase
`Synthesis of Protected Peptides and Glycopeptides
`
`Oliver Seitz and Horst Kunz*
`Institut fu¨ r Organische Chemie der Universita¨t Mainz, J.-J. Becher-Weg 18-20, D-55099 Mainz, Germany
`
`Received April 23, 1996X
`
`The recently developed allylic HYCRON anchor1 exhibits excellent properties for the solid phase
`synthesis of protected peptides and glycopeptides. Model reactions with analogous low molecular
`weight compounds assessed the acid- and base-stability of the polar and flexible HYCRON linkage.
`The new anchor is available in a two-step synthesis and allows the use of both the Boc- and the
`Fmoc-strategy, which can even be combined within one synthesis. Protected glycopeptides are
`released under almost neutral conditions, taking advantage of the Pd(0)-catalyzed allyl transfer to
`a weakly basic nucleophile such as N-methylaniline. The highly efficient synthesis of O-RGalNAc-
`(TN)-peptides of the MUC-1 repeating unit is described. Acid- and base-stability of the allyl ester
`linkage enabled the synthesis of an O-glucosylated peptide by first removing a threonine tert-butyl
`group on the solid phase and subsequently glycosylating the liberated resin-bound hydroxyl
`component.
`
`Introduction
`
`Proteins and peptides exist in a variety of conjugated
`forms, the most important being glycoconjugates.2 N-
`and O-Glycosidically bound carbohydrates affect the
`conformation of proteins and regulate their activity and
`biological half-lives.3 Glycoproteins participate in bio-
`logical recognition processes such as cell adhesion, regu-
`lation of cell growth, and cell differentiation. During
`tumor-transformation, the glycosylation pattern of gly-
`coproteins, such as mucins, changes markedly.4 Mucins
`are excessively O-glycosylated proteins with a carbohy-
`drate portion reaching 50-80%.5 Mucins expressed from
`various epithelial cell types in soluble and membrane-
`associated forms typically consist of several serine- and
`threonine-rich repeating units which serve as carbohy-
`drate scaffolds. During tumor progression the balance
`between the mucin-mediated adhesion6 and antiadhesion
`processes is no longer under control. Simultaneously,
`alterations of the glycosylation pattern occur, because
`certain glycosyltransferases are expressed in lower con-
`centrations.7 These two phenomena seem to be causally
`related. For instance, TN-antigen structures (Ser/Thr-
`(RGalNAc)) were demonstrated to be tumor-associated
`
`X Abstract published in Advance ACS Abstracts, January 1, 1997.
`(1) Abbreviations: AA, amino acid; Ac, acetyl; Acm, acetamido-
`methyl; Bn, benzyl; Boc, tert-butyloxycarbonyl; DCC, N,N ¢-dicyclo-
`hexylcarbodiimide; DIC, N,N ¢-diisopropylcarbodiimide; DIPEA, diiso-
`propylethylamine; DKP, diketopiperazine; EtSMe, ethyl methyl sulfide;
`Fmoc, 9-fluorenylmethoxycarbonyl; Gal, D-galactopyranose; GalNAc,
`2-acetamido-2-deoxy-D-galactopyranose; Glc, D-Glucopyranose; GPC,
`gel permeation chromatography; HOBt, 1-hydroxybenzotriazole; HON-
`Su, N-hydroxysuccinimide; HYCRAM, hydroxycrotonoyl (aminometh-
`yl)polystyrene; HYCRON, hydroxycrotyl-oligoethylene glycol-n-al-
`kanoyl; MPLC, medium performance liquid chromatography; Mtr,
`4-methoxy-2,3,6-trimethylbenzenesulfonyl; Muc, mucine; NMM, N-
`methylmorpholine; Pac, phenacyl; PS, polystyrene; tBu, tert-butyl; Trt,
`trityl; TBTU, 2-O-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium
`tetrafluoroborate; Z, Benzyloxycarbonyl.
`(2) (a) Montreuil, J. Adv. Carbohydr. Chem. Biochem. 1980, 37, 157
`(b) Lis, H.; Sharon, N. Eur. J. Biochem. 1993, 218, 1.
`(3) Rudd, P. M.; Woods, R. J.; Wormald, M. R.; Opdenakker, G.;
`Downing, A. K.; Campbell, I. D.; Derek, R. A. Biochim. Biophys. Acta.
`1995, 1248, 1.
`(4) Brockhausen, I.; Yang, J.-M.; Burchell, J.; Whitehouse, C.;
`Taylor-Papadimitriou, J. Eur. J. Biochem. 1995, 233, 607.
`(5) Hilkens, J.; Ligtenberg, M. J. L.;Vos, H. L.; Litvinov, S. L. Trends
`Biochem. Sci. 1992, 17, 359.
`(6) Berg, E. L.; McEvoy, L. M.; Berlin, C.; Bargatze, R. F.; Butcher,
`E. C. Nature 1993, 366, 695.
`
`and T-antigen structures (Ser/Thr(R-3-((cid:226)Gal)-GalNAc))
`even tumor-specific in breast tissue.8
`Oligonucleotides9 and oligopeptides10 are readily avail-
`able via solid phase synthesis. The properties of the
`anchor group positioned between the oligomer to be
`synthesized and the polymeric support are crucial for the
`success of a solid phase synthesis. There is an increasing
`demand for linkage groups which provide an additional
`orthogonal stability and can, thus, accommodate a wider
`range of reaction conditions. This particularly holds true
`for the synthesis of glycopeptides and for the growing
`field of combinatorial chemistry.11 The synthesis of
`glycopeptides and phosphopeptides as well as peptides
`with acid- and base-labile protecting groups, which are
`to be employed in fragment condensations,12 requires
`anchor groups which allow the release of the target
`molecule under conditions that leave the labile structures
`unaffected.
`Most of the commonly used anchors for the solid phase
`synthesis of protected peptides and glycopeptides are
`either acid-labile13 or base-labile. Nearly all of the acid-
`labile anchor groups belong to the benzyl ester type14 and
`produce long-lived, stable carbocations during their
`cleavage. These may cause undesired alkylations of
`nucleophilic moieties.15 Furthermore, acid-labile protect-
`ing groups such as the tert-butyloxycarbonyl(Boc)-group
`
`(7) Vavasseur, F.; Dole, K.; Yang, J.; Matta, K. L.; Myerscough, N.;
`Corfield, A.; Paraskeva, C.; Brockhausen, I. Eur. J. Biochem. 1994,
`222, 415.
`(8) Karsten, U.; Papsdorf, G.; Pauly, A.; Vojtesek, B.; Moll, R.; Lane,
`E. B.; Clausen, H.; Stosiek, P.; Kasper, M. Differentiation 1993, 54,
`55.
`(9) (a) Narang, S. A. Tetrahedron 1983, 39, 3. (b) Bannwarth, W.
`Chimia 1987, 41, 302.
`(10) (a) Merrifield, R. B. J. Am. Chem. Soc. 1963, 85, 2149. (b) Fields,
`G. B.; Noble, R. L. Int. J. Pept. Protein Res. 1990, 35, 161. (c) Williams,
`P. L.; Albericio, F.; Giralt, E. Tetrahedron 1993, 347, 11065.
`(11) a) Jung, G.; Beck-Sickinger, A. G. Angew. Chem., Int. Ed. Engl.
`1992, 31, 375. (b) Gallop, M. A.; Barrett, R. W.; Dower, W. J.; Fodor,
`S. P. A.; Gordon, E. M. J. Med. Chem. 1994, 37, 1233. (c) Gallop, M.
`A.; Barrett, R. W.; Dower, W. J.; Fodor, S. P. A.; Gordon, E. M. J. Med.
`Chem. 1994, 37, 1385.
`(12) Benz, H. Synthesis 1993, 337.
`(13) (a) Paulsen, H.; Merz, G.; Weichert, U. Angew. Chem., Int. Ed.
`Engl. 1988, 27, 1365. (b) Lu¨ ning, B.; Norberg, T.; Tejbrant, J.
`Glycoconjugate J. 1989, 6, 5. (c) Laczko, I.; Hollosi, M.; Urge, L. E.;
`Ugen, K.; Weiner, D. B.; Mautsch, H. H.; Thurin, J.; Otvos, L.
`Biochemistry 1990, 4282. (d) Mucin-type glycopeptides: Meldal, M.;
`Mouritsen, S.; Bock, K. Am Chem. Soc. Symp. Ser. 1993, 519, 19.
`
`S0022-3263(96)00743-8 CCC: $14.00 © 1997 American Chemical Society
`
`Illumina Ex. 1104
`IPR Petition - USP 10,435,742
`
`
`
`814 J. Org. Chem., Vol. 62, No. 4, 1997
`
`Seitz and Kunz
`
`Scheme 1
`
`Scheme 2
`
`cannot be removed on solid phase without detaching the
`peptide. Base-labile linkages such as in Kaiser’s oxime
`resin are not compatible with the widely used fluorenyl-
`methoxycarbonyl(Fmoc)-strategy. Most of the existing
`photolabile linkers can only be cleaved in low yields if
`longer peptide sequences are synthesized. Photolabile
`anchors with high cleavability are usually base-sensi-
`tive16 and not fully compatible with Fmoc chemistry.
`Linkers of the general allyl-type17 are of particular
`value, because they are removable under almost neutral
`conditions and are orthogonally stable to the commonly
`used acid- and base-labile protecting groups.18 Allylic
`anchoring offers the possibility of applying both the Boc-
`and the Fmoc-group for temporary N-terminal protection
`in solid phase peptide synthesis. The strategies may
`even be combined in one synthesis. The cleavage of the
`allylic linkage is achieved by a palladium(0)-catalyzed
`transfer reaction19 of the allyl group to a nucleophile,
`which acts as an allyl scavenger. Nucleophiles such as
`morpholine,20 the less basic N-methylaniline,21 or dime-
`done22 irreversibly trap the allylic moiety.
`Solid phase peptide synthesis with allylic anchoring
`was introduced using hydroxycrotonic acid linkers ((cid:226)-
`HYCRAM 1, Scheme 1). A range of complex glycopep-
`tides have been successfully synthesized on the (cid:226)-HY-
`CRAM-support.23 While Boc-strategy provided glyco-
`peptides in high yield, the application of the Fmoc-
`strategy showed losses of some peptide during the
`synthesis.24 Moreover, complete release of the peptide
`was sometimes difficult to achieve.
`
`(14) (a) Wang, S. S. J. Am. Chem. Soc. 1973, 95, 1328. (b) Mitchell,
`A. R.; Erickson, B. W.; Ryabster, M. N.; Hodges, R. S.; Merrifield, R.
`B. J. Am. Chem. Soc. 1976, 98, 7357. (c) Flo¨rsheimer, A.; Riniker, B.
`Peptides 1990. Proceedings of the 21st European Peptides Symposium;
`Giralt, E., Andreu, D., Eds.; ESCOM: Leiden, 1991; p 131. (d) Mergler,
`M.; Tanner, R.; Gosteli, J.; Grogg, P. Tetrahedron Lett. 1988, 29, 4005.
`(e) Mergler, M.; Nyfeller, R.; Tanner, R.; Gosteli, J.; Grogg, P.
`Tetrahedron Lett. 1988, 29, 4009. (f) Albericio, F.; Barany, G. Tetra-
`hedron Lett. 1991, 32, 1015.
`(15) (a) Atherton, E.; Cameron, L. R.; Sheppard, R. C. Tetrahedron
`1988, 44, 843. (b) Albericio, F.; Kneib-Cordonier, N.; Biancalana, S.;
`Geva, L.; Masada, R. I.; Hudsen, D.; Barany, G. J. Org. Chem. 1990,
`55, 3730.
`(16) Tjoeng, F. S.; Tam, J. P.; Merrifield, R. B. Int. J. Pept. Protein
`Res. 1979, 14, 262.
`(17) Kunz, H.; Dombo, B. (ORPEGEN GmbH), DE-A P 3720269.3,
`1987; USA. 4 929 671, 1990.
`(18) (a) Kunz, H.; Dombo, B. Angew. Chem., Int. Ed. Engl. 1988,
`27, 711. (b) Kunz, H.; Dombo, B.; Kosch, W. In Peptides 1988; (Jung,
`G., Bayer, E., Eds.; W. de Gruyter: Berlin, 1989, 154. (c) Blankemeyer-
`Menge, B.; Frank, R. Tetrahedron Lett. 1988, 29, 5871. (d) Guibe´, F.;
`Dangles, O.; Balavoine, G.; Loffet, A. Tetrahedron Lett. 1989, 30, 2641.
`(e) Zhang, X.; Jones, R. A. Tetrahedron Lett. 1996, 22, 3789.?
`(19) (a) Trost, B. M. Acc. Chem. Res. 1980, 13, 385. (b) Trost, B. M.
`Pure Appl. Chem. 1980, 53, 2357.
`(20) (a) Kunz, H.; Waldmann, H. Angew. Chem., Int. Ed. Engl., 1984,
`23, 71. (b) Friedrich-Bochnitschek, S.; Waldmann, H.; Kunz, H. J. Org.
`Chem. 1989, 54, 751.
`(21) Ciommer, M.; Kunz, H. Synlett 1991, 593.
`(22) Kunz, H.; Unverzagt, C. Angew. Chem., Int. Ed. Engl. 1984,
`23, 436.
`(23) Kosch, W.; Ma¨rz, J.; Kunz, H. React. Polym. 1994, 22, 181 and
`references therein.
`(24) Kosch, W. Dissertation, Universita¨t Mainz, 1992.
`
`In the design of a new allylic anchoring group, a more
`flexible spacer was inserted between the anchor and the
`polymer in order to facilitate an efficient access of the
`Pd(0) complex during the detachment reaction. The
`success achieved with craft copolymers25 of polyethylene
`glycol and polystyrene inspired the incorporation of a
`polar spacer of the ethylene glycol type. In addition, in
`this HYCRON anchor26 2 (Scheme 1), the R,(cid:226)-unsatur-
`ated carbonyl structure was avoided, since it could be
`responsible for the occasionally low yields in Fmoc
`chemistry with (cid:226)-HYCRAM 1. As in (cid:226)-HYCRAM, (cid:226)-ala-
`nine is used as a standard amino acid, since it simplifies
`the determination of the peptide loading.27
`
`Results and Discussion
`
`Synthesis of a Low Molecular Weight Model. The
`first step in synthesis of an anchor derivative 4, which
`allows the coupling of the starting amino acid via
`nucleophilic esterification consists in the sodium glycolate
`catalyzed 1,4-addition of triethylene glycol to tert-butyl
`acrylate (Scheme 2). The subsequent reaction of the
`adduct 3 with 1,4-dibromo-2-butene is carried out under
`phase transfer conditions, which avoids the retro-Michael
`reaction leading to unseparable mixtures. This elimina-
`tion is initiated by abstraction of an R-proton, which is
`unlikely to occur in the resin-bound molecule, because
`proton abstraction would take place at the amide function
`disabling further deprotonation. Treatment of 4a with
`TFA removes the tert-butyl group.
`The carboxylic acid 5 is activated with DCC28 and
`N-hydroxysuccinimide29 (HONSu) and reacted with (cid:226)-Ala-
`NH-Bn to give 6 (Scheme 3). A slight excess of the acid
`compound must be used in order to prevent the concomi-
`tant allylation of HONSu. Compound 6 can be consid-
`ered a low molecular weight model of an (aminomethyl)-
`polystyrene (H2N-PS) resin functionalized with the
`standard amino acid (cid:226)-alanine and the anchor-bromide.
`The reaction of the cesium salt of Z-protected alanine
`
`(25) (a) Rapp, W.; Zhang, L.; Ha¨bich, R.; Bayer, E. In Peptides 1988,
`Jung, G.; Bayer, E., Eds.; de Gruyter: Berlin 1989, 199. (b) Bayer, E.
`Angew. Chem., Int. Ed. Engl. 1991, 30, 113.
`(26) Seitz, O.; Kunz, H. Angew. Chem., Int. Ed. Engl. 1995, 34, 807.
`(27) (a) Atherton, E.; Clive, D. I.; Sheppard, R. C. J. Am. Chem. Soc.
`1975, 97, 6584. (b) Albericio, F.; Barany, G. Int. J. Pept. Protein Res.
`1985, 26, 92.
`(28) Sheehan, J. C.; Hess, G. P. J. Am. Chem. Soc. 1955, 77, 1067.
`(29) Weygand, F.; Hoffmann, D.; Wu¨ nsch, E. Z. Naturforsch. 1966,
`21b, 426.
`
`
`
`A Novel Allylic Anchor in Solid Phase Synthesis
`
`J. Org. Chem., Vol. 62, No. 4, 1997 815
`
`Table 1. Nucleophilic Esterification of Fmoc-Amino Acids with Anchor-Bromides 4a,b and Removal of the C-Terminal
`Protecting Group (see also Scheme 4)
`Pro
`Gly
`(R ) tBu),
`(R ) tBu),
`% yield
`% yield
`9a, 65
`10a, 81
`9b, 88
`10b, 85
`
`Val
`(R ) tBu),
`% yield
`11a, 67
`11b, 57
`
`Thr(tBu)
`(R ) Pac),
`% yield
`12a, 83
`12b, 97
`
`ester formation
`R removal
`
`Ala
`(R ) tBu),
`% yield
`8a, 72
`8b, 84
`
`Scheme 3
`
`Scheme 4
`
`with bromide 6 provides the model compound 7. This
`nucleophilic esterification is known to proceed without
`any detectable racemization.30 Unfortunately, the chro-
`matographic behavior of 6 and 7 is nearly identical
`resulting in a small contamination with 6 even after two-
`fold MPLC and preparative HPLC. However, the mate-
`rial is useful for stability tests, since the reactions of both
`6 and 7 can be monitored by analytical HPLC. In order
`to assess the stability of the HYCRON system toward
`the conditions of N-Boc- and N-Fmoc removal, compound
`7 is treated with dichloromethane/TFA (1:1) and mor-
`pholine, respectively. Several reaction cycles are simu-
`lated by an exposure of two days. HPLC analysis shows
`that formation of new substances occurs in an amount
`of less then 4%. This is a promising result for the
`application of HYCRON-anchoring in solid phase peptide
`synthesis.
`Synthesis of Anchor Conjugates and Resin Load-
`ing. The starting amino acids should be loaded onto the
`polymer via their anchor conjugates, since attempts to
`couple amino acid cesium salts to a PS-resin functional-
`ized with (cid:226)-alanine and the bromo-anchor 5 resulted in
`low loading yields. Using the linkage agent 4a, anchor
`conjugates of Fmoc-amino acids can be synthesized. For
`conjugate synthesis of amino acids with acid-labile side
`chain protecting groups, the tert-butyl group of 4a is
`exchanged (Scheme 2). The phenacyl (Pac) ester 4b
`provides orthogonally stable protection to most side chain
`protecting groups in SPPS.
`Usually Fmoc-amino acids are alkylated via the cesium
`salt method. This reaction is less favorable for the
`HYCRON-system, because the rate of the alkylation
`reaction is slow and Fmoc-cleavage competes in the basic
`environment. In contrast, this nucleophilic esterification
`with the anchor bromides 4a and 4b proceeds mildly
`under phase transfer conditions using the base NaHCO3
`(Scheme 4). Reactants need not to be applied in excess.
`Conjugates 8a-12a are easily accessible (Table 1) using
`this method. TFA liberates the carboxylic function in the
`case of tert-butyl protection to give 8b-11b, and zinc
`powder in acetic acid reductively cleaves the Pac-group
`from 12a to form 12b.
`Conjugates consisting of starting amino acid, anchoring
`group, and standard amino acid are synthesized via two
`possible routes, which differ in the sequence of condensa-
`
`Scheme 5
`
`tions. For synthesis of conjugate 16b containing Boc-
`Ala, the anchor-bromide 5 is condensed with (cid:226)-Ala-OPac.
`Amide 14 is employed in an alkylation of the Boc-Ala
`cesium salt followed by reductive removal of the Pac-
`group (Scheme 5). Conjugate 16a is synthesized by
`coupling 8b to (cid:226)-Ala-OPac and subsequent reductive
`cleavage of the Pac-ester 13.
`In order to attach the amino acid-anchor-conjugates
`to the polymer, Boc-(cid:226)-Ala loaded (aminomethyl)polysty-
`rene (H2N-PS) is treated with TFA and neutralized with
`diisopropylethylamine (DIEA). The conjugates 8b-12b
`are coupled to this resin by the DIC/HOBt31 method
`(Scheme 6). The same reaction conditions are applied
`for binding the conjugates 16a and 16b to unmodified
`H2N-PS resin. Any amino groups that have not been
`acylated are capped with acetic anhydride/pyridine.
`Table 2 shows the loadings and yields achieved by these
`reactions.
`In the previously described synthesis of anchor conju-
`gates the allylic ester is formed by allylation of the
`carboxylic acid. However, nucleophilic esterifications can
`be accompanied by alkylations of other nucleophilic
`structures. For instance, the esterification of Fmoc-Cys-
`(Acm) under the above mentioned conditions proceeds
`
`(30) Wang, S.-S.; Gisin, B. F.; Winter, D. P.; Makofske, R.; Kulesha,
`I. D.; Tsougraki, C.; Meienhofer, J. J. Org. Chem. 1977, 42, 1286.
`
`(31) (a) Sarantakis, D.; Teichmann, J.; Lien, E. L.; Fenichel, R. L.
`Biochem. Biophys. Res.Commun. 1976, 73, 336. (b) Ko¨nig, W.; Geiger,
`R. Chem. Ber. 1970, 103, 788.
`
`
`
`816 J. Org. Chem., Vol. 62, No. 4, 1997
`
`Seitz and Kunz
`
`AA
`
`Ala
`
`Table 2. Attachment of Amino Acid-Anchor-Conjugates
`to the Polystyrene Resin (see also Scheme 6)
`loading/
`PG
`anchor-conjugate
`(mmol of AA/g), yielda
`8b
`17a, 0.52, 89%
`Fmoc
`16a
`17b, 0.71, 89%
`Fmoc
`16b
`17c, 0.67, 75%
`Boc
`9b
`18, 0.50b, 80%
`Gly
`Fmoc
`10b
`19, 0.52, 61%
`Pro
`Fmoc
`11b
`20, 0.41, 75%
`Val
`Fmoc
`12b
`21, 0.40c, 66%
`Thr(tBu)
`Fmoc
`(cid:2)
`a Yield ) lAA/l(cid:226)-Ala (cid:2) 100 for resins 17a and 18-21, lAA/lNH2
`100 17b,c (l ) loading, determined by amino acid analysis).
`b Content of Fmoc group (photometrically). c Original loadings and
`yields are higher, and determination by amino acid analysis gives
`values that are too low.
`
`Scheme 6
`
`with a concomitant allylation of the thiol function. The
`side reactions are avoided if the carboxylic group is
`reacted with the anchor-alcohol. In contrast to the amino
`acid loadings with HYCRON-bromides as substrates, this
`procedure can also be performed on the solid support.
`Acetoxylation of the anchor-bromide 4a and removal of
`the tBu-group produces the protected form 23 of the
`allylic anchor-alcohol (Scheme 7). Using DIC/HOBt,
`compound 23 is coupled to (cid:226)-Ala-NH-PS, and 24 is
`deacetylated by means of aqueous NaOH/dioxane. The
`attachment of the starting amino acid Fmoc-Cys(Acm)
`is achieved by activating the C-terminus with 2,6-
`dichlorobenzoyl chloride32 in situ and then coupling to
`the resin-bound allylic alcohol in the presence of pyridine.
`This reaction takes place without affecting the thiol
`function as was demonstrated by the esterification of
`compound 26 in solution. It should be taken into account
`that the application of this more convenient method
`results in a higher propensity of racemization.
`Synthesis of Tri- and Tetrapeptides. In order to
`investigate the properties of the HYCRON-anchor re-
`garding stability and cleavability, protected tri- and
`tetrapeptides 28-31 have been synthesized (Scheme 8).
`The coupling reactions were accomplished with a three-
`to four-fold excess of the N-protected amino acids acti-
`vated by DIC and HOBt. Temporary N-Boc-groups
`(Scheme 8, PG ) Boc) were removed by treating the
`peptide-resin with dichloromethane/TFA (1:1). Reactions
`with DMF/morpholine (4:3) for 2 h deprotected the amino
`groups if the Fmoc-strategy was applied (Scheme 8, PG
`) Fmoc). Removal of the N-Fmoc-groups is usually
`complete in less than 45 min. However, in these syn-
`theses, the cleavage times were extended in order to
`
`(32) Sieber, P. Tetrahedron Lett. 1987, 28, 6147.
`
`Scheme 7
`
`Scheme 8
`
`facilitate the detection of undesired anchor cleavage
`(Table 3). The final release of the peptides was achieved
`by treating the peptide-resin with a catalytic amount of
`the palladium(0)-catalyst and a ten- to fifteen-fold excess
`of the scavenger nucleophile under exclusion of oxygen.
`Morpholine was used for the release of N-Boc-protected
`peptides and N-methylaniline for the liberation of N-
`Fmoc protected peptides, since this base does not to
`cleave the Fmoc-group. Detachment yields were dem-
`onstrated to be in the range of 90% on average. Table 3
`shows peptide losses, which occur at the stage of the
`dipeptide-resin due to the formation of diketopiperazines
`(DKP) and the total peptide losses caused by any other
`undesired cleavage of the HYCRON-linkage. It is clear
`that the application of the Boc-strategy (peptide 28) does
`not result in any premature losses at all. Fmoc-cleavage
`with DMF/morpholine (4:3) for 2 h leads to a considerable
`amount of DKP formation (see peptides 29-31). The
`
`
`
`A Novel Allylic Anchor in Solid Phase Synthesis
`
`J. Org. Chem., Vol. 62, No. 4, 1997 817
`
`Table 3. Synthesis of Protected Tri- and Tetrapeptides (see also Scheme 9)
`
`scavenger
`peptide-DKP,a
`losses,
`detachment,
`R2NH
`peptide
`PG removal
`% yield
`% totalb
`% yieldc
`28
`CH2Cl2/TFA, 1:1, 45 min
`0
`0
`morpholine
`89
`29
`DMF/morpholine, 4:3, 2 h
`20
`22
`N-methylaniline
`81
`30
`DMF/morpholine, 4:3, 2 h
`38
`33
`N-methylaniline
`93
`31
`DMF/morpholine, 4:3, 2 h
`33
`34
`N-methylaniline
`96
`a Diketopiperazine formation ) [(lAla/l(cid:226)-Ala)tripep-resin/(lAla/l(cid:226)-Ala)dipep-resin - 1] (cid:2) 100. b ((lAla/l(cid:226)-Ala)pep-resin/(lAla/l(cid:226)-Ala)AA-resin - 1) (cid:2) 100.
`c (1 - [(lstart-AA/l(cid:226)-Ala)a/(lstart-AA/l(cid:226)-Ala)b] (cid:2) 100, (l ) loading, determined by amino acid analysis, a ) after cleavage, AA ) amino acid, b
`) before cleavage).
`
`peptide (n-mer)
`32 (4-mer)
`33 (9-mer)
`
`34 (9-mer)
`
`35 (20-mer)
`
`1.-4.: DIC/HOBt
`5.-6.: TBTU/HOBt/NMM
`7.-8.: DIC/HOBt
`1.-19.: TBTU/HOBt/NMM
`
`93
`
`95d
`
`Table 4. Solid Phase Synthesis of Peptide 33 and O-Glycopeptides 32, 34, and 35
`PG removal (PG, conditions)
`couplings
`detachment, % yielda
`1.-3.: Fmoc, DMF/morpholine 1:1, 50 min
`1.-3.: DIC/HOBt
`96
`1.: Fmoc, DMF/morpholine 1:1, 5 min
`1.-8.: DIC/HOBt
`87
`2.: Boc, CH2Cl2/TFA 1:1, 50 min
`3.-9.: Fmoc, DMF/morpholine 1:1, 50 min
`1.: Fmoc, DMF/morpholine 1:1, 50 min
`2.: Boc, CH2Cl2/TFA 1:1, 50 min
`3.-9.: Fmoc, DMF/morpholine 1:1, 50min
`1.: Fmoc, DMF/morpholine 1:1, 50 min
`2.: Boc, CH2Cl2/TFA 1:1, 50 min
`3.-19.: Fmoc, DMF/morpholine 1:1, 50min
`a (1 - [(lstart-AA/l(cid:226)-Ala)a/(lstart-AA/l(cid:226)-Ala)b] (cid:2) 100. b npeptide/(mstart-AA-resin/lstart-AA) (cid:2) 100. c Based on loading with (cid:226)-Alafminimum yield.
`d Loadings determined photometrically by Fmoc-cleavage (l ) load, a ) after cleavage, b) before cleavage, n ) amount, m ) mass).
`
`total, % yieldb
`77c
`83
`
`95
`
`45d
`
`Scheme 9
`
`total losses are solely caused by DKP formation since they
`are identical the DKP losses within the experimental
`error. No additional losses occur, and it can be concluded
`that the HYCRON-anchor is orthogonally stable to both
`the Boc- and the Fmoc-group. The amount of intramo-
`lecular aminolysis of the resin-bound dipeptides can be
`decreased by reducing the Fmoc-cleavage times.
`Synthesis of O-Glycopeptides. The efficiency of
`HYCRON-anchoring in solid phase glycopeptide synthe-
`sis was demonstrated by the synthesis of the protected
`O-glycotetrapeptide 32 (Scheme 9). Peptide 32 is a
`glycosylated variant of the N-terminal part of peptide T,
`a segment of the HIV-envelope glycoprotein gp120.33
`Starting from resin 21, treatment with DMF/morpholine
`
`(33) (a) Kowalski, M.; Potz, J.; Basiripour, L.; Dorfman, T.; Goh,
`W. C.; Terwilliger, E.; Dayton, A.; Rosen, G.; Haseltine, W.; Sodroski,
`J. Science 1987, 237, 1351. (b) Pert, C. B.; Hill, J. M.; Ruff, M. R.;
`Berman, R. M.; Robey, W. G.; Arthur, L. O.; Ruscetti, F. W.; Farrar,
`W. L. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 9254.
`
`(1:1) for 50 min removed the Fmoc-group (Table 4). The
`carbohydrate portion was introduced by coupling the
`preformed glycosyl amino acid (Fmoc-Thr(RAc3GalNAc)-
`OH.34 This, as well as the other coupling reactions, was
`performed using DIC/HOBt. In the presence of O-acetyl
`groups long coupling times, which may be necessary due
`to the bulkiness of the glycoside, may result in an acetyl
`shift to the amino group that is to be acylated. In order
`to accelerate the elongation reaction of the resin-bound
`glycodipeptide, the reagents were employed in a larger
`excess. Release of the O-glycosylated tetrapeptide 32 was
`accomplished in a yield of 96% by palladium(0)-catalyzed
`transfer of the allyl moiety to N-methylaniline. An exact
`overall yield can not be given. Calculations based on the
`initial amino acid load is not reasonable because the
`amino acid analysis of hydroxyamino acids generally
`gives low values resulting in overestimated yields. As-
`
`(34) Kunz, H.; Birnbach, S. Angew. Chem., Int. Ed. Engl. 1986, 25,
`360.
`
`
`
`818 J. Org. Chem., Vol. 62, No. 4, 1997
`
`suming a threonine/(cid:226)-alanine-ratio of 1, which requires
`a quantitative coupling of 12b to (cid:226)-Ala-NH-PS, a mini-
`mum overall yield of 77% can be calculated.
`Taking advantage of the HYCRON-system, glycopep-
`tide structures of the MUC-1 mucine were synthesized.
`Synthesis of the protected forms of the nonapeptide 33,
`its TN-variant 34 and of the TN-eicosapeptide 3535 (Scheme
`9), which spans the entire 20-mer MUC-1 repeating unit,
`starts from resins 18 and 19, respectively. The 20-mer
`is continuously repeated up to 90 times in the MUC-1
`mucin. It is obvious that it would be convenient to couple
`two 20-mers in order to obtain a dimer of the repeating
`unit. If possible, a Pro-Gly bond should be formed in
`fragment condensations, since proline is the amino acid
`least prone to racemization, and the fastest couplings are
`achieved when the C-terminal fragment contains N-
`terminal glycine. For that reason the frame of the
`repeating unit in peptide 35 is slightly moved to C-
`terminal proline. Removal of the Fmoc-groups was again
`accomplished by treatment with DMF/morpholine (1:1)
`for 50 min. N-Boc-amino acids were used for elongation
`of the amino acid-resins (Table 4). The subsequent
`acidolysis of the Boc-group reduces the formation of
`diketopiperazines during cleavage. The strategy was
`then returned to Fmoc-protection. All acylations were
`carried out in DMF. For the synthesis of 33 and 34 DIC/
`HOBt-activation was used except for the couplings of the
`O-glycosyl-threonine and the following aspartic acid
`building block. These acylations were anticipated to be
`slow, and therefore the amino acids were activated with
`TBTU,36 HOBt, and N-methylmorpholine (NMM). Cou-
`plings of the N-terminal amino acids proline and alanine
`were each repeated once. The N-terminal alanine was
`acetylated after Fmoc-removal. TBTU-activation was
`applied for all coupling reactions in the synthesis of TN-
`eicosapeptide 35. Treating the peptide-resins with cata-
`lytic amounts of the palladium(0)-complex in the presence
`of morpholine or N-methylaniline in DMSO/DMF de-
`tached the peptides in yields between 87 and 95%.
`Compounds 33, 34, and 35 were obtained in overall yields
`of 83%, 95% and 45% respectively. The overall yields are
`based on the initial amino acid load determined by amino
`acid analysis or photometrically after Fmoc-cleavage. In
`the case of 34 the overall yield is higher than the
`detachment yield. Of course, this is impossible, but the
`deviation is within the experimental error of the amino
`acid analysis.
`The O-glycononapeptide 34 was deprotected with TFA
`in the presence of ethyl methyl sulfide and anisole as
`scavengers37 (Scheme 10). A sodium methylate-catalyzed
`transesterification38 removed the O-acetyl groups of the
`sugar moiety yielding 37, which was purified by gel
`permeation chromatography and isolated in an overall
`yield of 74% relative to the loaded starting amino acid
`(22 steps).
`Removal of Permanent Protecting Groups and
`O-Glycosylation on Solid Phase. Solid phase synthe-
`sis of O-glycopeptides is possible following two routes.
`The carbohydrate can be introduced by preformed gly-
`
`(35) Dupradeau, F.-Y.; Stroud, M. R.; Boivin, D.; Li, L.; Hakomori,
`S.-E.; Singhal, A. K.; Toyokoni, T. Bioorg. Med. Chem. Lett. 1994, 4,
`1813.
`(36) Knorr, R.; Trzeciak, A.; Bannwarth, W.; Gillessen, D. Tetrahe-
`dron Lett. 1989, 30, 1927.
`(37) Atherton, E.; Sheppard, R. C.; Wade, J. D. J. Chem. Soc., Chem.
`Commun. 1983, 1060.
`(38) Peters, S.; Bielefeldt, T.; Meldal, M.; Bock, K.; Paulsen, H. J.
`Chem. Soc., Perkin Trans. 1 1992, 1163.
`
`Seitz and Kunz
`
`Scheme 10
`
`cosyl amino acids or by glycosylation of an unprotected
`hydroxyl group on the solid phase. Due to the difficulties
`involved in the latter method, only very little data have
`been published so far. Solid phase glycosylation using
`sugar oxazolines was reported, but unfortunately no
`yields were given.39 A yield of 5% was achieved in a silver
`triflate promoted mannosylation of a serine employing
`the corresponding glycosyl bromide.40
`We had developed a method in which soft electrophiles
`transform carbohydrates carrying an allylic carbamate
`at the anomeric position into potent glycosyl donors.41 In
`order to examine the efficiency of this glycosylation, a
`resin-linked peptide with an deblocked hydroxyl group
`was prepared starting from 20 (Scheme 11). Fmoc-
`removal was achieved with DMF/piperidine (4:3) in 10
`min, and couplings were performed using DIC/HOBt.
`TFA cleaved the tert-butyl ether of the fully protected
`resin-bound tripeptide. The tripeptide-resin was divided
`and one part subjected to the palladium(0)-catalyzed allyl
`transfer to N-methylaniline. Tripeptide 38, a partial
`structure of MUC-1, was obtained in a total yield of 96%.
`Before glycosylation, the remaining resin-linked peptide
`was dried for several days under high vacuum and
`repeatedly washed with dry dichloromethane. The resin-
`bound tripeptide with an unprotected hydroxyl group was
`reacted with 7 equiv of the glucose donor 1-O-(N-
`allylcarbamoyl)-2,3,4,6-tetra-O-acetylglucose and 2.5 equiv
`of the promotor S-methylbis(methylthio)sulfonium hexa-
`chloroantimonate.41 The large excess of the donor rela-
`tive to the promotor was used in order to prevent an
`electrophilic attack of the promotor at the double bond
`of the anchor system. The glycosylation reaction was
`repeated twice. After palladium(0)-catalyzed cleavage,
`the glycosylated Fmoc-tripeptide 39 was isolated in an
`overall yield of 4%, the unchanged Fmoc-tripeptide 38
`in 64%. In the solid phase O-glucosylation of a sterically
`hindered threonine-peptide a yield comparable to the
`O-mannosidation of the serine-peptide described previ-
`ously was achieved.
`
`Conclusion
`
`The allylic HYCRON-anchoring allows the efficient
`solid phase synthesis of peptides and glycopeptides in
`high yields and high purity. The acid- and base-stability
`
`(39) Hollo´si, M.; Kolla´t, E.; Laczko´, I.; Medzihradszky, K. F.; Thurin,
`J.; Otvo¨s, L. Tetrahedron Lett. 1991, 32, 1531.
`(40) Andrews, D. M.; Seale, P. W. Int. J. Pept. Protein Res. 1993,
`42, 165.
`(41) Kunz, H.; Zimmer; J. Tetrahedron Lett. 1993, 34, 2907.
`
`
`
`A Novel Allylic Anchor in Solid Phase Synthesis
`
`J. Org. Chem., Vol. 62, No. 4, 1997 819
`
`Scheme 11
`
`of the HYCRON-anchor makes the application of both
`the Fmoc- and the Boc-strategy possible. They may even
`be combined in one synthesis. Peptides and glycopep-
`tides are liberated under almost neutral conditions via
`the palladium(0)-catalyzed allyl transfer to scavenger
`nucleophiles such as morpholine or the less basic N-
`methylaniline. The Fmoc-group, the Boc-group, tert-
`butyl-esters, tert-butyl ethers, the O-acetyl groups in
`carbohydrates, and particularly O-glycosidic bonds are
`left intact. Therefore, the HYCRON-anchor will be a
`versatile tool in solid phase synthesis of compounds which
`demand high standards of orthogonality and also shows
`properties promising for combinatorial chemistry.
`
`Experimental Section
`
`General.42 Fmoc-amino acids, (aminomethyl)polystyrene,
`and Boc-(cid:226)-Ala-loaded polystyrene were kindly donated by
`ORPEGEN GmbH, Heidelberg, Germany. THF was freshly
`distilled from potassium/benzophenone before use. Dichlo-
`romethane used in the glycosylation reaction was dried over
`P4O10 and freshly distilled. Reactions were carried out at room
`temperature if no specifications are given. Solid phase syn-
`thesis was performed manually using a reaction vessel similiar
`to the Merrifield-reactor.
`Amino acid-analysis (AAA) was carried out by ORPEGEN
`GmbH. Photometric determination of the Fmoc-loadings was
`performed by treating 1 aliquot of the resin with 2.00