18 Peptides and Proteins By P. M. HARDY Department of Chemistry University of Exeter Stocker Road Exeter EX4 400 1 Introduction Peptides were last covered in Annual Reports for 1969 as part of a chapter entitled 'Amino-acids and Peptides'. The aim of the present Report is to outline the advances made in the subsequent five years in the field of peptides with a very selective discussion of some aspects of proteins. Emphasis is laid principally on methods of structure determination and synthesis rather than on the sequences elucidated or the syntheses achieved. The space available has precluded any coverage of conformational studies cyclic peptides or peptides containing non-protein amino-acids. 2 Structure Elucidation Methods.-Enzymic Cleavage. Several novel enzymes have been characterized whose specificities are of interest to the chemist involved in sequencing work.Two of these pyrrolidonyl peptidase' and dipeptidyl aminopeptidase,2 have begun to be used in structural work. The former removes pyrrolidonecarboxylic acid (pyroglutamic acid) from the N-terminus of peptides thereby revealing a free amino-group and unblocking this end of the molecule for subsequent Edman degradation.' The latter cleaves peptides in a stepwise fashion from the N-terminus liberating dipeptides. If the amino-terminal amino-acid residue is removed by one cycle of the Edman degradation subsequent treatment with this enzyme gives a series of overlapping dipeptides. Comparison of the two series of dipeptide sequences should enable the original sequence to be established.The N-terminal pentapeptide segment of promelittin was established in this way.3 Ambiguities however can obviously arise with certain sequences. Application to the A-chain of insulin (21 residues) gave two possible primary ~tructures,~ indicating that this interesting approach may be limited to peptides of very moderate size. ' A. Szewczuk and J. Kwiatowska European J. Biochem. 1970 15 92. R. A. Valyulis and V. M. Stepanov Biokhimiya 1971 36 866. ' G. Kreil European J. Biochem. 1973 33 558. R. J. Rowlands and H. Lindley Biochem. J. 1972 126 685. 497 498 P.M. Hardy Two proteases as yet untested in sequencing applications have great potential in this field. Myxobacter AL-1 protease I1 attacks peptides or proteins only at the amino-side of lysine residues.It can thus provide peptide fragments which overlap those produced by trypsin which if the lysine side-chains are appro- priately masked are cleaved only at arginine residues.' A protease from Staphylo-coccus aureus specifically cleaves at the C-terminal side of aspartyl and glutamyl peptide bonds. Under certain conditions fission can be limited to glutamyl bonds. This enzyme may incidentally assist in the assignment of glutamine and asparagine residues to which it is not sensitive.6 A peptidoglutaminase from B. circulans has been isolated which attacks only the y-amide bonds of peptide- bound glutamine;' this may prove useful in combination with the S. aureus protease. Edman Degradation.Peptide sequencing by the automated Edman method first described in 1967,8 has rapidly gained ground at the expense of the manual technique since commercial machines for this purpose 'sequenators' became available in 1969. However the chemist cannot be entirely replaced by the technician. In general only the first thirty to fifty N-terminal amino-acid residues are directly accessible using this approach. In favourable cases more can be achieved e.g. the first 66 of the 88 residues of bovine parathyroid hormone could be identified.' The complete sequence of P-lactoglobulin has been deter- mined by sequence analysis of the intact protein and its tryptic peptides but this work involved modification of the classical Edman routine." The chief limitation of the automated method is its reliance on the insolubility of the residual peptide during the extraction of the liberated thiazolinones; short peptides are too soluble to be successfully retained during this procedure.Several ways of overcoming this limitation have been found to be useful. The use of buffers based on dimethylbenzylamine enables the less polar solvent benzene to replace ethyl acetate in the extraction step as its salts with organic acids are now sufficiently soluble in the hydrocarbon.' Alternatively the peptide itself can be rendered more hydrophilic. Tryptic peptides which remain the fragments most widely used in sequencing have either arginine or lysine at their C-terminus. The polarity of the arginine side-chain confers more resis- tance to extraction than does that of lysine.However if the first step of the degradation is carried out using a highly sulphonated isothiocyanate such as (1) or (2) the E-amino-group of the lysine can be modified to confer suitable ex- traction resistance even on a peptide as short as five amino-acid residues.I2 M. Wingard G. Matsueda and R. S. Wolfe J. Bacteriol. 1972 112 940. J. Houmard and G. R. Drapeau Proc. Nut. Acad. Sci. U.S.A. 1972 69 3506. ' M. Kikuchi H. Hayashida E. Nakano and K. Sakaguchi Biochemistry 1971,10 1222. P. Edman and C. Begg European J. Biochem. 1967 1 80. M. Tanaka M. Haniu G. Matsueda K. T. Yasunobu S. Mayhew and V. Massey Biochemisrry 197 1 10 3041. lo G. Braunitzer R. Chen B. Schrank and A. Stangl 2.physiol. Chem. 1972,353 832.'I M. A. Hermodson L. H. Ericsson K. Titani H. Neurath and K. A. Walsh Bio-chemistry 1972 11 4493. G. Braunitzer B. Schrank A. Ruhfus S. Peterson and U. Peterson 2.physiol. Chem. 1971 352 1730; G. Braunitzer B. Schrank S. Peterson and U. Peterson ibid. 1973 354 1563. Peptides and Proteins Aminoethylation of peptides with C-terminal cysteine residues effects a similar change in properties,' while peptides which lack suitably placed side-chain functional groups can be after their initial conversion into an N-phenylthio- carbamyl derivative coupled through their C-terminal carboxy-group to amino- methanesulphonic acid ' or 2-aminonaphthalene- 1,5-disulphonic acid. ' If a water-soluble carbodi-imide is used to effect reaction at pH 4 amino-groups other than that of the reagent are protected from reaction by protonation.Side-chain carboxy-groups however will always be tagged by additional solubilizing groups. Another successful strategy involves total insolubilization of the peptide by attaching its C-terminus to a polymer. An automated apparatus using this principle has been described.16 However while attaching the C-terminus to the resin amino-groups in side-chains can also form links to the polymer. Thus on degradation the thiazolinones corresponding to aspartic and glutamic acid residues may be missing. A more serious problem which occurs with aspartic acid is formation of the cyclic imide after activation of the P-carboxy-group during coupling to the resin. Edman degradation is then completely blocked when the first aspartimide residue is reached.16 This can be avoided either by coupling a C-terminal lysine residue to the resin through its &-amino-group using p-phenylene di-isothiocyanate' or by use of N-ethyl-N'-(3-dimethyl- aminopropy1)carbodi-imide'* as the coupling reagent.The C-terminal carboxy- group forms an oxazolinone which is susceptible to aminolysis by a resin-bound amino-group but the side-chain carboxy-groups are deactivated by isomerization to N-acyl-ureas. This method has been successfully tested on 20-cycle runs; the modified derivatives of aspartic and glutamic acids liberated can be hydro- . lysed to the parent phenylthiohydantoin for recognition.'8 The only specific chemical cleavage method widely used in protein-fragmenta- tion studies remains the cyanogen-bromide-induced fission at methionine C- terminal homoserine being generated.Homoserine lactone readily prepared from the parent acid by treatment with trifluoroacetic acid has been used to attach cyanogen-bromide-generated peptides to solid-phase supports as it will l3 J. K. Inman J. E. Hannon and E. Appella Biochim. Biophys. Acta 1972 46 2075. G. Crombie J. A. Foster and C. Franzblau Biochem. Biophys. Res. Comm. 1973,52 1228. J. A. Foster E. Bruenger C. L. Hu K. Albertson and C. Franzblau Biochem. Biophys. Res. Comm. 1973 53 70. l6 R. A. Laurson European J. Biochem. 1971 20 89. R. A. Laurson M. J. Horn and A. G. Bonner F.E.B.S. Letters 1972 21 67. I8 A. Previero J. Derancourt and M. A.Coletti-Previero F.E.B.S. Letters 1973 33 135. 500 P. M. Hardy acylate resin amino-groups. The peptide amino-groups need not be blocked if the resin amino-groups are present in large excess. A fragment of rabbit- muscle actin containing 13 residues which had resisted conventional methods of primary structure determination was successfully sequenced using this method. l9 The second and largest protein to have been totally sequenced by the automated Edman method is calf-skin collagen al-CB7. Correct placing of the 268 residues required studies on 19 peptides prepared by tryptic chymotryptic and thermo- lytic cleavage as well as on the intact protein.” It is clear that with large proteins although the use of sequenators may speed up realization of the desired objective the sequencing of the amino-acids is still in general dependent on the attainment of suitable peptides which provide overlaps of breakage points.Mass Spectrometry. Although the sequencing of peptides by mass-spectrometric means is in general only successful on compounds containing up to ten amino- acid residues and offers no advantages in economy of material over conventional sequencing methods work over the past few years has succeeded in considerably refining it as an analytical tool. It offers a single operation (as opposed to a repetitive subtractive method like the Edman degradation) although in general raw peptides must be derivatized prior to insertion into the instrument for analysis. End-group protection provides few problems but permethylation of the amide NH groups (to destroy their tendency to reduce volatility by formation of hydrogen bonds) can also attack a-CH groups and more seriously the side- chains of cysteine methionine and histidine.This opposes the effects of N-methylation because salts are formed and it also complicates the spectra. These side-reactions can be circumvented by conversion of the susceptible amino-acids into less sensitive structures. Methionine can be protected as its sulphoxide,21 and cysteine desulphurized with Raney nickel to alanine (normally using deuterium-saturated catalyst to enable the product to be distinguished from other alanine residues).22 Histidine may be cleaved with diethyl pyro- carbonate to a 1,2-bis(ethoxycarbonylamido)ethylenederivative (Scheme l) I 1 OH C0,Et C0,Et Reagents i diethyl pyrocarbonate; ii H,O Scheme 1 19 M.J. Horn and R. A. Laurson F.E.B.S. Letters 1973 36 285. 20 P. P. Fietzek F. W. Rexrodt K. E. Hopper and K. Kiihn European J. Biochem. 1973 38 396. 21 P. Roepstorff K. Norris S. Severinson and K. Brunfeldt F.E.B.S. Letters 1970 9 235. 21 Yu. A. Ovchinnikov A. A. Kiryushin V. A. Gorlenko and B. V. Rozynov Zhur. obshchei Khim. 1971,41 660. Peprides und Proteins 50 1 which does not form a quaternary salt.23 Arginine residues must also be modified otherwise the mass spectrum contains sequence information only from the N-terminus to the amino-acid preceding the first arginine residue. Hydrazine may be used to convert arginine into ~rnithine,'~ or else P-diketones to incorporate the guanido-group into an N6-pyrimidylornithine derivative." However proper treatment of any peptide presupposes knowledge of its amino-acid content and detracts from the simplicity of the method.A re-evaluation of permethylation in which peptides were treated with CD,I for a short period (in the presence of methylsulphinyl carbanion) and then CH,I for a longer period showed that the optimum time for methylation has in general been over-estimated as only deuteriomethyl groups were incorporated. Exposure for one minute is sufficient in most cases to give complete substitution of amide NH groups and under these conditions side-chains are largely unaffe~ted.~~~~~ If this short treatment is used arginine is the only residue requiring prior modi- ficat ion.Alternative methods to conventional electron-impact ionization (El) show promise for use in structure determination. Application of a field desorption (FD) technique gives a great reduction in fragmentation due to thermal de- composition and unprotected peptides show a molecular or a quasi-molecular (M + 1) ion. Arginine does not need to be protected ; even a nonapeptide con- taining an arginine residue at each end gives an intense molecular ion. Spectra can be obtained with as little as 10 ng adsorbed on the field anode.27 The use of chemical ionization (CI) allows the mass spectra of free oligopeptides to be obtained with sample temperatures at least 150 "Cbelow those required for EL The tetrapeptide Leu-Leu-Val-Tyr for instance gives an abundant (M + 1)' peak using CI at 150°C but none below 340°C with EI.The mass spectra obtained by CI also contain abundant sequencing peaks ; FD seems less satis- factory in this respect.28 The sequencing peaks produced by CI are also of relatively uniform intensity allowing the use of a smaller sample ;penta-alanine for example needs a 200 nmol sample for EI but only 2 nmol for CI.29 One facet of mass spectrometric sequencing that is being developed is the analysis of peptide mixtures. Here this technique may have clear advantages over stepwise degradation. Experiments with model peptides have established the applicability of this approach to sequencing enzymic digests of proteins which have simply been subjected to gel filtration to ensure that the material loaded into the instrument contains only peptides with ten residues or less.23 J. F. G. Vliegenthart and L. Dorland Biochem. J. 1970 117 31P. '' H. R. Morris R. J. Dickinson and D. H. Williams Biochem. Biophys. Res. Comm. 1973 51 247. 25 P. A. Leclercq L. C. Smith and M. M. Desiderio Biochem. Biophys. Res. Comm. 1971 45 937. 26 H. R. Morris F.E.B.S. Letters 1972 22 257. 27 H. U. Winkler and H.-D. Beckey Org. Muss. Spectrometry 1972 6 655; Biochern. Biophys. Res. Comm. 1972 46 391. 2B M. A. Baldwin and F. W. McLafferty Org. Mass. Spectrometry 1973 7 1353. 29 W. R. Gray L. H. Wojcik and J. H. Futrell Biochem. Biophys. Res. Comm. 1970 41 1111. 502 P.M. Hardy Larger peptides must be digested further with other enzymes before attempting ~equencing.~' Using a computer-assisted high-resolution instrument a mixture of four components (Gly Gly-Ala-Leu Val-Gly-Gly and MePhe-Gly) was correctly ~equenced.~ However such expensive and sophisticated hardware is not essential.A mixture of three rather larger peptides from a tryptic digest of pepsin (Val-Gly-Leu-Ala-Pro-Val-Ala, Ala-Asn- Asn-Lys and Gln-Tyr-Tyr-Thr- Val-Phe-Asp-Arg) was apart from the Asp-Arg terminus correctly sequenced using a relatively low-resolution in~trument.~' Without prior knowledge of the sequence as determined by classical methodology eighty per cent of the sequence of the 249-residue enzyme ribitol dehydrogenase from Klebsiellu uerogenes has been determined by m.s.of partially fractionated peptide mixtures.32 Partial- vaporization experiments are essential for mixture work whatever the instrument the change in relative abundance of peaks at varying temperatures serving to distinguish which peaks arise from the same oligopeptide. Sequence ambiguities may arise however due to the formation of 'non-sequence' ion peaks; these may usually be resolved by comparing the spectra of the acetylated and per- methylated peptides with those of the deuterio-acetylated and perdeuterio- methylated mixture. 33 The amide bonds of peptides are not essential to m.s. sequencing. The amines produced by reduction of the carbonyl groups of the peptide bonds show suitable sequencing ions and can be readily separated by g.1.c.Such an approach has been utilized in a computer-assisted g.1.c.-m.s. determination of the structure of a 20-residue fragment from the C-terminus of actin which was liberated by cleavage with cyanogen bromide. A partial acid hydrolysate of the peptide was esterified acetylated and then reduced with lithium aluminium hydride. The resulting polyamino-alcohols were 0-trimethylsilylated separated by g.l.c. and analysed by m.s. Twenty-seven peptides none larger than a tetrapeptide were identified. The single histidine-containing peptide required direct insertion into the ion source before a suitable spectrum could be obtained. The fragment peptides could be fitted together in two ways but identification of tryptophan as being N-terminal resolved the ambiguity.Digestion of the intact peptide with trypsin cleaved it at the single lysine residue exposing glutamine as the new N-terminus and hence establishing the position of the sole amidated side-chain carboxy- The same structure was deduced independently from a conventional sequencing Structures Determined.-The past few years have proved a particularly fruitful period for the isolation and determination of the primary structure of a variety 30 H. R. Morris D. H. Williams and R. P. Ambler Biochem. J. 1971 125 189. 31 F. W. McLafferty R. Venkataraghavan and P. Irving Biochem. Biophys. Res. Comm. 1970 39 274. 32 H. R. Morris D. H. Williams G. G. Midwinter and B. S. Hartley Biochem. J. 1974 141 701. 33 H. W. Wipf P. Irving M. McCamish R. Venkataraghavan and F.W. McLafferty J. Amer. Chem. SOC.,1973,95 3369. 34 H. Nau J. A. Kelley and K. Biemann J. Amer. Chem. SOC.,1973,95 7162. 35 M. Elzinga and J. H. Collins Cold Spring Harbor Symp. Quant. Biol. 1973 37 1. Peptides and Proteins 503 of peptide and protein hormones particularly those secreted by the pituitary gland. Among the proteins of medical interest so characterized are human growth hormone (HGH),36,38 ovine pr~lactin,~’ human placental lactogen (HPL),38 and ox parathyroid hormone.39 HGH and HPL are identical over 85% of their sequences. Porcine hypothalami have yielded a decapeptide (3) that controls the secretion of growth hormone from the anterior pituitary and the same source Val-His-Leu-Ser- Ala-Gly-Glu-Lys-Glu- Ala 12 3 4 5 6 7 8 9 10 (3) Glp-His-Trp-Ser-Tyr-GI y-Leu-Arg-Pro-Gly-NH 12 3 4 5 6 7 8 9 10 has yielded another peptide (4)of this size which stimulates the release of both luteinizing hormone (LH) and follicle-stimulating hormone.40 Curiously enough a tetrapeptide analogue of this latter peptide (Glp-Tyr-Arg-Trp-NH,) in which the amino-acids are scrambled from their natural order shows some LH-releasing power whereas its isomer Glp-Trp-Tyr-Arg-NH in which the residues are in the same relative order as in the decapeptide parent is inactive.41 The elusive substance P first described in brain extracts as long ago as 1931 has at last been isolated in a pure form from bovine hypothalami and its sequence (5)determined.42 Four of its five C-terminal amino-acids are identical with those Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met-NH2 1 2 3 4 5 6 7 8 9 10 11 (5) of physalaemin (from amphibian skin) and eledoisin (from cephalopod salivary glands) ;all three compounds stimulate smooth muscle and exert a hypotensive effect.Two peptides which inhibit the release ofgrowth hormone and melanocyte- stimulating hormone from the pituitary have been identified as the pentapeptide 36 C. H. Li and J. S. Dixon Arch. Biochem. Biophys. 1971 146 233. 37 C. H. Li J. S. Dixon T.-B. Lo K. D. Schmidt and Y. A. Pankov Arch. Biochem. Biophys. 1970 141 705. 38 H. D. Niall Nature New Bid 1971 230 90. 39 H. B. Brewer and R. Ronan Proc. Nut. Acad. Sci. U.S.A. 1970,67 1862; H. D. Niall H. D. Keutmann R. Sauer M. Hogan B.Dawson G. Aurbach and J. Potts Z. physiol. Chem. 1971 351 1506. 40 A. B. Schally Y. Baba R. M. G. Nair and C. D. Bennett J. Bid. Cfiem. 1971 246 6647. 4’ J.-K. Chang H. Sievertsson C. Bogentoft B. L. Currie K. Folkers and C. Y. Bowers Biochem. Biophys. Res. Comm. 1971,44 409. 42 M. M. Chang S. E. Leeman and H. D. Niall Nurure New Biol. 1971 232 86. 504 P.M. Hardy Y 'i' Ala-Gly-Cys-Lys-Asn-Phe-Phe-Trp-Lys-Thr-Phe-Thr-S~r-Cys 12 3 4 5 6 7 8 91011121314 (6) Pro-His-Phe-Arg-Gly-NH 43 and the disulphide-bridged tetradecapeptide (6) re~pectively.~~ The latter has been named somatostatin. The sinus gland near the eye of crustaceans resembles in function the hypo- thalamus of vertebrates. The blanching hormone of the shrimp Pandalus borealis which concentrates the red-pigmented areas has been isolated from the eyestalks in sufficient quantity (1OOpg) to be established as an octapeptide (7).4s The controversial compound scotophobin the first compound involved in the Glp-Leu-Asn- Phe-Ser-Pro-Gly-Trp-NH 12345678 Ser-Asp-Asn-Asn-Gln-Gln-Gly-Lys-Ser-Ala-Gln-Gln-Gly-Gly-Tyr-NH~ 1 2 3 4 5 6 7 8 9101112131415 chemical transfer of memory to which a structure has been assigned is thought to be a pentadecapeptide (8).This substance was isolated from the brains of rats trained to avoid the dark and is said to produce the same effect if injected into untrained animals. Material of this sequence synthesized by the solid-phase method gave a product apparently active in the behavioural bi~assay.~' How- ever this work has been criticized in almost all its aspects.47 A hexapeptide (9) 5-oxoPro-Ala-Gly -Tyr-Ser-Ly s 123456 (9) ameletin also isolated from the brains of rats is claimed similarly to transfer a lack of fright-response to noises.48 Hormones controlling the gastrointestinal tract continue to be isolated.That with the rather ungainly title 'gastric inhibitory polypeptide' contains 43 residues and as its name suggests it is a potent inhibitor of gastric acid secretion. It shows some homology to glucagon and rather less to secretin in its 43 R. M. G. Nair A. J. Kostin and A. V. Schally Biochem. Biophys. Res. Comm. 1972 47 1420. 44 R. Burgus N. Ling M. Butcher and R. Guillemin Proc. Nat. Acad. Sci. U.S.A. 1973 70 684; Biochem.Biophys. Res. Comm. 1973,50 127. A5 P. Fernlund Biochem. Biophys. Acta 1974 371 304. 46 G. Ungar M. M. Desiderio and W. Parr Nature 1972 238 198. 47 W. W. Stewart Nature 1972 238 202. 48 H. Lackner and N. Tiemann Naturwiss. 1974 61 217. Peptides and Proteins 505 N-terminal27 residue^."^ The rather smaller peptide motilin (22 residues) which stimulates gastric motor activity has been obtained from hog intestinal mucosa. Its structure is quite distinct from all previously described gastrointestinal pep tide^.^ Since the discovery of the insulin single-peptide-chain precursor proinsulin in 1968 there has been considerable work devoted to the characterization of other prohormones. Precursors of this type are not limited to disulphide-bridged materials.Proglucagon for example contains an extra eight amino-acid residues at its C-terminus,’ while promelittin contains its additional octapeptide unit at the N-termin~s.~ The ingenious use of ‘SC-labelled amino-acids allowed sequencing experiments on the latter compound to be carried out on very small amounts (-50 pg) of material3 Evidence has been obtained for pr~gastrin,~’ proparathyroid hormone,’ and pro~alcitonin,~~ and it seems likely that the formation of prohormones may be a general phenomenon. The most active growth area in primary structure determination over the past five years has been that of the snake-venom toxins; only in the cytochromes has there been a comparable number of individual compounds being sequenced annually.The most lethal of the various components are the ‘long’ toxins which exert a curariform effect. They contain 71-74 amino-acid residues and have four disulphide bridges. The ‘short’ toxins contain 60-62 residues but the cystine cross-links occur in the same positions as the deleted fragment is at the C-terminus. With a couple of dozen sequences now established only two residues besides the half-cystine residues are common to all toxins although considerable homology exists between those of similar length and physiological The structures of four scorpion neurotoxins have also been elucidated ; these resemble the ‘short’ snake toxins both in size and in their possession of four disulphide bridge^.'^ 3 Peptide Synthesis Protecting Groups.-cc-Group Protection.Attempts to extend the range of amino-protecting groups available which can be selectively removed continue to centre largely on the urethane type. The aa-dimethyl-3,s-dimethoxybenzyloxy-carbonyl group (10) is somewhat more labile to acid than the widely used t-butoxycarbonyl group but it can also be removed photochemically. Quantitative 49 J. C. Brown and J. A. Dryburgh Cunad. J. Chem. 1971,49 867. J. C. Brown M. A. Cook and J. A. Dryburgh Canad. J. Chem. 1973 51 533; €3. Schubert and J. C. Brown Cunad. J. Biochem. 1974 52 7. 5’ H. S. Tager and D. F. Steiner Proc. Nut. Acad. Sci. U.S.A. 1973 70 2321. 52 R. A. Gregory and H. J. Tracy Lancet 1972 ii 797. ’’ J. F. Habener B. Kemper J. T. Potts and A. Rich Endocrinology 1973 92 219.54 B. A. Roos K. Okama and L. J. Deftos Biochem. Biophys. Res. Comm. 1974 60 1 134. See ‘Amino-acids Peptides and Proteins’ ed. G. T. Young (Vol. 4) and R. C. Sheppard (Vols. 5 6) (Specialist Periodical Reports) The Chemical Society London 1973 1974 1975 Vols. 4-6 Ch. 2 under ‘Toxins’. 56 D. R. Babin D. D. Watts S. M. Goos and R.V. Mlejnek Arch. Biochem. Biophys. 1974 164. 694. 506 P.M. Hardy (10) deprotection occurs if 6-millimolar solutions are passed through a spiral made of one metre of quartz tubing irradiated with a one kilowatt high-pressure mercury lamp; trytophan seems to survive this treatment.57 The 1,l-dimethyl-2- propynyloxycarbonyl group (1 1)can be removed by hydrogenolysis in the presence Me 0 I II HCEC-C-0-C-I Me (1 1) of sulphur-containing amino-acid~.~ * Two base-sensitive groups the 9-fluorenyl- methoxycarbonyls9 and the cyano-t-butoxycarbony160 moieties have been examined.The former undergoes solvolysis in liquid ammonia (Scheme 2) Polymer +- A CO + RNH CH -0,CNHR Reagents i NH,; ii H,O Scheme 2 and the latter 8-elimination at pH 10 in aqueous solution. The 4-picolyloxy- carbonyl group (12) is stable in strong acids owing to protonation but it can be removed by zinc dust in 50 ”/ acetic acid or catalytic hydrogenation.61 The basic ‘handle’ in this group may aid purification of peptide derivatives. 57 C. Birr W. Lochinger G. Stahnke and P. Lang Annalen 1972 763 162. 58 G. L. Southard B. R. Zabarowsky and J. M..Pettee J. Amer.Chem. SOC.,1971 93 3302. ’’ L. A. Carpino and G. Y.Han J. Org. Chem. 1972 37 3404. 6o E. Wunsch and R.Soangenberg Chem. Ber. 1971 104 2427. 61 D. F. Veber S.F. Brady and R. Hirschmann Proceedings of the Third American Peptide Symposium held at Boston Massachusetts 1972 ‘Chemistry and Biology of Peptides’ ed. J. Meienhafer Ann Arbor Science Publishers Inc. Ann Arbor Michigan 1972 p. 315. Peptides and Proteins 0 NT CH2-0-C-II At the present time the most utilized amino-protecting groups as far as published syntheses of natural peptides and their analogues are concerned are those removable by acidic conditions or hydrogenolysis ;base-sensitive OT photolabile groups have not in general proved so applicable. Protecting groups cleavable by electrolysis and by solvolysis with organic solvents are now being developed although they have yet to be widely used.Two groups of the latter type have been reported. The 2-bromo-l,l-dimethylethoxycarbonylcan be removed simply by heating in ethanol (Scheme 3);62 the vinyloxycarbonyl group must be treated with bromine in dichloromethane to potentiate it. Subsequent treatment in methanol causes cleavage presumably by a mechanism similar to that outlined in Scheme 3.63 A comparative study of the electrolysis of 2- CH2cBr .q//I H02CCHRNH-C.\ -+ H0,CCHRAH <'&Me I 0-CMe ,O Me 'H Br-Reagents i EtOH Scheme 3 2,2-,and 2,2,2-halogenoethoxycarbonylprotecting-groups has established in simple models that if the half-wave potentials of two members of the series differ by at least 0.3 V selective reduction should be p~ssible.~~,~~ The use of metal complexes in connection with amino-protection is a further area beginning to be explored.The allyloxycarbonyl group can be cleaved from amino-acids with nickel carbonyl in the presence of tetramethylethylenediamine but peptides have yet to be examined.66 Treatment of pentacarbonyl[methoxy- (phenyl)carbene]chromium with amino-esters gives the corresponding [amino- (pheny1)carbenelpentacarbonylchromium complexes (13) which are stable to alkali enabling the ester group to be saponified and the peptide chain built up. The group can be removed from the yellow complex by trifluoroacetic The only other method of note using a non-urethane protecting group involves 62 T.Ohnishi H. Sugano and M. Miyoshi Bull. Chem. SOC.Japan 1972,45,2603. 63 R. A. Olofson and Y. S. Yamamoto U.S.P.3 711 458 (CI. 260/112.5; CO7cg COSh) (Chem. Abs. 1973,78 98 024). 64 E. Kasafirek Tetrahedron Letrers 1972 202 I. 65 M. F. Semmelhack and G. E. Heinsohn J. Amer. Chem. SOC.,1972 94 5139. 66 E. J. Corey and J. W. Suggs J. Org. Chem. 1973,38 3223. " K. Weiss and E. 0.Fischer Chem. Ber. 1973 106 1277. 508 P.M. Hardy (13) E and Z forms incorporation of the amino-group into a 4,S-diphenyl-A4-oxazolin-2-one ring system by successive reaction with benzoin cyclic carbonate and trifluoroacetic acid. Catalytic reduction or peracid oxidation afford derivatives sensitive to hydrolysis (Scheme 4).68 Ph Ph Ph Ph Ho0 +H,N- H -+ ~ O H 't(0 2 OKN-O Ph Ph Ph ph)=( II HOWOH 0 N-Ph Ph PhCOCOPh PhCH,CH,Ph + CO, HffH * +co + + + H,N-OK" + H,N-0 Reagents i CF,CO,H; ii m-CIC,H,CO3H; iii H +; iv Pd-H Scheme 4 Protecting groups which have fallen into disuse or which never fulfilled their initial promise can ofcourse take on new leases of life if novel methods of removal are found.Electrolytic cleavage of the tosyl group for example now enables it to be cleaved without affecting N-benzyloxycarbonyl groups or S-benzyl N-Trifluoroacetyl groups have been found to be sensitive to reduction by b~rohydride,~' and in contrast to earlier reports the o-nitro- phenylsulphenyl group can be cleaved with ammonium thiocyanate ;the secret here lies in the use of 2-methylindole as a scavenger for the o-nitrophenylsul- phenyl thiocyanate produced.' The N-chloroacetyl group can now be removed by thioureas under mild neutral conditions.It is preferable to use l-piperidine- thiocarboxamide (14) to avoid the further condensation of the 2-iminothiazolidi- J. C. Sheehan and F. S. Guziec J. Amer. Chem. SOC.,1972 94 6561. 69 K. Okamura T. Iwasaki M. Matsuoka and K. Matsumoto Chem.and Ind. 1971,929. 70 F. Weygand and E. Frauendorfer Chem. Ber. 1970 103 2437. 7' E. Wunsch and R. Spangenberg Chem. Ber. 1972 105 740. Peptides and Proteins none liberated if free thiourea is added (Scheme 5).72 The cleavage can also be effected after treatment with 3-nitropyridine-2-thione but liberation of the amino- group in this case requires the use of trifluoroacetic acid (Scheme 6).73 ,NHz CICH,CONH-+ S=C \ 1 - + NH2 l;>Yli +C1 H,N- H 1 Scheme 5 + 0 + CF,C02- H,N- Reagents i aq.NaHCO, 40°C; ii CF,CO,H Scheme 6 Two carboxy-protecting groups have similarly become more practical propo- sitions. The phthalimidomethyl group can be removed with zinc and acetic acid which are conditions leaving benzyl- and t-butyl-based protecting groups undi~turbed.’~ The resurrection of the use of phenyl esters is based on their rapid hydrolysis at pH 10.5 in the presence of hydrogen peroxide in aqueous dioxan or DMF. The apparently racemization-free reaction no doubt involves initial formation of the peptide pere~ter.~~ Several photolabile C-protecting groups ’’ W.Steglich and H. G. Batz Angew. Chem. Internat. Edn. 1971 10 75. 73 K. Undheim and P. E. Fjeldstat J.C.S. Perkin I 1973 829. 74 D. L. Turner and E. Baczynski Chem. and Ind. 1970 1204. ’’ G. W. Kenner and J. H. Seely J. Amer. Chem. SOC.,1972,94 3259. 510 P.M. Hardy -co-o Me -CO-0-CH-CO ..Omph OMe (15) HC -0-COW (15)-(17) have been de~cribed,~~-~~ but they have not so far been seriously evaluated in the synthesis of any sizeable peptide. Another group with potential is the NN'-di-isopropylhydrazide which is stable under both acidic and basic conditions but removable by mild oxidation with for example lead tetra-acetate in pyridine. The reaction is particularly clean as the carboxylic acid is the only non-volatile Side-chain Protection.The protection of side-chain functional groups has made some useful progress over the past few years. Conversion of the guanido-group of arginine into a bis-l-adamantyloxycarbonylderivative (18) provides such a hydrophobic shell for the side-chain that intermediates protected in this way readily dissolve in organic solvents. The urethane is like t-butoxycarbonyl derivatives stable to hydrogenolysis but cleaved by trifluoroacetic acid. Arginine protected in this way was used in the first synthesis of porcine proinsulin- connecting peptide.80 A new protecting group developed for histidine the 1,1,1,3,3,3-hexafluoro-2-(p-chlorophenoxymethoxy)propyl (HF-PA) group (19) is stable to base and hydrogenolysis but it can be removed with acid at room l6 A.Patchornik B. Amit and R. B. Woodward J. Amer. Chem. SOC.,1970,92 6333. 77 J. C. Sheeman and K. Umezawa J. Org. Chem. 1973 38 3771. la J. C. Sheeman R. M. Wilson and A. W. Oxford J. Amer. Chem. SOC.,1971,93 7222. 7g D. H. R. Barton M. Girijavallabhan and P. G. Sammes J.C.S. Perkin I 1972 929. G. Jager and R. Geiger Chem. Ber. 1970 103 1727. Peptides and Proteins 511 7F3 C-0-CH2-0 /I Nim I CF3 I -His-temperature. It can be introduced into N-benzyloxycarbonyl-histidine methyl ester by condensation with hexafluoroacetone followed by acylation with a,p-dichloroanisole.** Solid-phase peptide synthesis requires side-chain-protecting groups to be more stable to the acidic reagents used in the a-N-deprotection cycles than in ordinary peptide synthesis.The sheer number of repetitions of this treatment in a synthesis means that a level of fission acceptable in conventional synthesis may lead to considerable impurities in the polymer-linked stepwise process. It has been suggested that no more than 0.05% cleavage of a side-chain group should occur during the removal of more than 99.95%of the N-a-protecting group if it is to be acceptable for solid-phase work. The use of E-N-benzyloxy-carbonyl protection for lysine leads to a loss of 0.8% of protection per hour the free E-amino-group then being available for chain branching. Replacement with 2- 2,4-,or 3,4-chlorobenzyloxycarbonylprotection however leads to essentially linear products ;final deprotection now requires hydrogen fluoride treatment.82 In a similar way the use of S-4-methylben~yl~~ has advantages over.the S-p-methoxybenzyl group.84 For tryptophan the utility of the N'-formyl group removable by hydrazine in DMF to protect the indole nucleus has been proven in several ~yntheses.~~ Formation of the Peptide Bond.-The four most widely used methods for the synthesis of the peptide bond remain the active ester the azide the mixed anhy-dride and the dicyclohexylcarbodi-imide-mediatedreactions. As far as the latter is concerned the use of additives which suppress racemization caused by the reagent has made it more suitable for fragment condensations. The most efficient additive is 3-hydroxy-4-oxo-3,4-dihydro-1,2,3-benzotriazine (20),but '' H.H. Selztmann and T. M. Chapman Tetrahedron Letters 1974 2637. 82 B. W. Erickson and R. B. Merrifield J. Amer. Chem. SOC.,1973,95 3757. 83 B. W. Erickson and R. B. Merrifield J. Amer. Chem. SOC.,1973 95 3750. 84 M. Ohno S. Tsukamoto S. Sato and N. Izumiya Bull. Chem. SOC.Japan. 1973 46 3280. D. Yamashiro and C. H. Li J. Org. Chem. 1973,38 2594. 512 P. M. Hardy this can itself undergo a side-reaction through conversion into an acylating agent (21).86 I-Hydroxybenzotriazole (22) has become the most widely used promoter of this type.86 Other additives which have been recommended include ethyl 2-hydroximino-2-cyanoacetate(23),875,7-dichloro-8-hydroxyquinoline,88 and p-chlorobenzenesulphohydroxamic acid (24).89 N-Hydroxy-compounds FN HON=C \ C0,Et 0 have also been found to catalyse the aminolysis of certain active esters notably of the p-nitrophenyl and 2,4,5-trichlorophenyI types in polar solvents such as DMF.Reactions can be complete in a few minutes. Additives with a pK close to that of acetic acid are the most effective although the nature of the solvent plays a crucial role; in THF for example additives actually inhibit aminolysis.’* The assessment of novel types of active ester shows no sign of diminishing in interest. Perhaps the ones most likely to find use are those of 2-hydroxypyridine and 2-mer~aptopyridine.”.~~ These are more reactive than the standard p-nitrophenyl esters but in contrast to these the rate of coupling increases as the solvent polarity decreases.This seems to be characteristic of aminolyses which proceed by intramolecular general base catalysis (and hence are racemization- resistant) and renders esters like these useful in solid-phase synthesis.” An unusual type of active ester is that derived from 5-amino-3-methyl-4-nitroso-1-phenylpyrazole (25). In the presence of dicyclohexylcarbodi-imide it can be mono- R2 I MerlNo MerlNo R1 N-0-CO-CHNHZ N I ‘N NH ’\N NHCOCHNHZ R’ Ph Ph I N\N NCOCHNHZ Reagents i ZNHCHR’C0,H-dicyclohexylcatbodi-imide; ii ZNHCHR’C0,H-dicyclohexylcarbodi-imide Scheme 7 acylated (26)or diacylated. The diacyl-product (27) is a stable crystalline material which undergoes rapid aminolysis on treating with amino-esters (Scheme 7) n6 W.Koenig and R. Geiger Chem. Ber. 1970 103 788 2024 2034. ‘’ M. Itoh Bull. Chem. SOC.Japan 1973 46 2219; 1974,47 471. H. Yajima M. Kurobe and K. Koyama Chem. and Pharm. Bull. (Japan) 1973 21 1612. 89 H. Yajima K. Kitagawa and M. Kutobe Chem. and Pharm. Bull. (Japan) 1973 21 2566. 90 W. Koenig and R. Geiger Chem. Ber. 1973 106 3626. 91 K. Lloyd and G. T. Young J. Chem. SOC.(0,1971 2890. Peptides and Proteins C-CH-NHZ II I 0 R’ presumably by intramolecular catalysis (28) since it is racemization-free. The co-product (26) can be removed by extraction with aqueous sodium carbonate and recovered by a~idification.~~ A further active ester with a ‘safety catch’ has been described. N-Protected amino acids can be esterified with 2,2-diphenyl- 2-hydroxyethanol to give compounds (29) which can be N-deprotected and elaborated at their N-terminus before activating the ester by treatment with trifluoroacetic acid to generate the corresponding vinyl ester (30).94 0 II -C-0-CH,-C-OH Ph I I Ph 0 I1-c-0-CH=C /Ph \ Ph (29) (30) That original cornerstone of racemization-free synthesis the azide coupling has been shown to require careful control if it is in fact to live up to its reputation.Slight racemization may occur even if the added base is not present in excess especially if triethylamine is used. N-Methylmorpholine is more satisfactory but di-isopropylethylamine gives the best result^.'^ The use of the stable non- explosive diphenylphosphorazidate (31)as a coupling reagent involves formation of the acyl azide of the carboxyl component as an intermediate rather than a mixed anhydride ;this racemization-free method is a convenient complement to the classical azide procedure.96 The application of various reagents containing ’’ A.S. Dutta and J. S. Morley J. Chem. SOC.(C),1971 2896. q3 M. Guarneri P. Giori and C. A. Benassi Tetrahedron Letters 1971 665. 94 T. Wieland J. Lewalter and C. Birr Annalen 1970 740,31. 95 P. Sieber M. Brugger and W. Rittel Proceedings of the Tenth European Peptide Symposium held at Abano Terne Italy 1969 ‘Peptides 1969’ ed. E. Scoffone North- Holland Publishing Co. Amsterdam 1971 p. 60. 96 T. Shiori and S. Yamada. Chem. and Pharm. Bull. (Japan) 1974 22 849 855 859. 514 P.M. Hardy phosphorus to peptide synthesis is a considerable feature of the past few years.Diethylphosphoryl cyanide prepared by the action of cyanogen bromide on triethyl phosphite is able also to mediate racemization-free coupling.97 Both the 'oxidation-reduction' and the toluene-p-sulphonic anhydride-hexa- methylphosphortriamide methods of peptide synthesis [see Annual Reports (B) 1969,66 503-5041 involve acyloxyphosphonium salts as the effective acylating agent; a third method which generates this species in a different way is that involving triphenylphosphine and carbon tetrachloride (Scheme 8).98 Racemiza-0 Ph,P + CCI -+ Ph,&CCI Ph,&O-C-R1 II c1-/Cl-+ CHCI R'CONHR2 + Ph,P=O + HhEt C1-Reagents i R'C0,H; ii R2NH2 NEt Scheme 8 tion may occur on linking peptide component^,^^ but if the temperature is kept low and tris(N-methy1piperazino)phosphine is used instead of triphenylphos- phine this can be prevented.loo Amide bonds may also be formed by the action of diphenyl phosphite and pyridine presumably by the route outlined in Scheme H-P R~CONHR~ H-P-OCOR' HO/\OPh Reagents i RlCO,H-pyridine; ii R'NH Scheme 9 9,1°1 and by the use of hexachlorophosphatriazine. Half an equivalent of the latter reagent in the presence of one equivalent of triethylamine or N-methyl- morpholine is effective (Scheme 10).'O2 A particularly interesting novel synthetic method involves conversion of a urethane-protected amino-acid into an aziridin-2-one ; phosgene is the preferred reagent for this although thionyl chloride or phosphoryl chloride may be used " S.Yamada Y.Kasai and T. Shiori Tetrahedron Letters 1973 1595. 98 L. E. Barstow and V. J. Hruby J. Org. Chem. 1971 36 1305. 99 T. Wieland and A. Seeliger Chem. Ber. 1971 104 3992. loo Y. Takeuchi and S.-I. Yamada Chem. and Pharm. Bull. (Japan),1974 22 32 41. lo' N. Yamazaki and F. Higashi Terrahedron Letters 1972 5047. Y. Martinez and F. Winternitz Bull. SOC.chim. France 1972 4707. Peptides and Proteins c1 Cl N C1 \/\/ \p/N\p/C1 0 Cl'i CA IPCl --L I\O-!!-R1 \/ N \p/N C<p\C1 2R'NHCOR' 0/\ c1 I c=o I R' Reagents i 2R1C0,H-base; ii R'NH Scheme 10 0 Reagents i COC1,-2NEt3 -20 "C; ii R2NH Scheme 11 (Scheme11). In the presence of an amino-component the strained three-membered ring is opened to give a peptide without change of configuration.Formation of an aziridinone is markedly dependent on the N-protecting group ;they have been prepared so far only from N-benzyloxycarbonyl-amino-acidsor their para-substituted derivatives. N-Benzyloxycarbonylaziridinoneitself is too unstable to be isolated and proline of course cannot form such a derivative. The coupling of these compounds appears to be particularly effective with hindered com- ponent~.''~ Lactones involving seven-membered rings are also active towards aminolysis. Treatment of the Schiff base derived from an amino-acid and 5-nitrosalicylaldehyde with dicyclohexylcarbodi-imidegenerates such a species ; leucine for example forms 3-isobutyl-4-nitro-2,3-dihydrobenzo[f] [1,4]oxa-zepin-2-one (32).Other hydroxycarbonyl compounds may be used e.g. pentane-2,4-dione and a-formyl-N-hydroxysuccinimide.104Most work in the area of activation through cyclic derivatives of carboxyl components however still centres on the N-carboxy-anhydrides. Methods of preparing N-carboxy- anhydrides from amino-acids with functional groups in their side-chains either free or suitably protected have now been detailed in connection with the use of such compounds in the controlled stepwise synthesis of peptides in aqueous media without isolation of intermediate^."^ A two-phase solvent system Io3 M. Miyoshi Bull. Chem. SOC.Japan 1973 46 212 1489. lo' M. Bodanszky U.S.P. 3 704 246 (Cl. 260-333; C07d) (Chem. Abs. 1973,78 58 801).lo' R. Hirschmann H. Schwam R. G. Strachan E. F. Schoenewaldt H. Barkemeyer S. M. Miller J. B. Conn V. Gorsky D. F. Weber and R. G. Denkewalter J. Amer. Chem. Soc. 1971,93 2746. 516 P.M. Hardy (acetonitrile-water 60 50 v/v) containing sodium carbonate at an apparent pH of 11.5 can be used for this purpose at -15 "C;this mixture has the advantage ofnot requiring such careful control of pH as the original method in homogeneous solution and over-reaction tends to be less because the resulting carbamates are more stable under these conditions. O7 N-Thiocarboxy-anhydrides have been explored as alternative reagents but although in general they tend to give higher yields significant amounts of epimer are formed. In contrast to histidine N-carboxy-anhydride which tends to form an imidazo-tetrahydropyrimidinone (Scheme 12) the corresponding sulphur analogue can be successfully used in '0 1 0 Scheme 12 repetitive peptide synthesis.lo* N-Carboxy-anhydrides can be acylated with o-nitrophenylsulphenyl chloride in anhydrous solvents in the presence of tri- ethylamine to give the relatively stable crjstalline N-o-nitrophenylsulphenyl-N-carboxy-anhydrides which do not tend to polymerize. However these derivatives are unsuitable for use in aqueous media. Good yields of dipeptides occur on treatment with N-trimethylsilylamino-acid trimethylsilyl esters.log N-4,4'-Dimethoxybenzhydryl-N-carboxy-anhydrides can be prepared from N-4,4-dimethoxybenzhydryl-amino-acidsin 30-40 "/ yield using phosgene in THF and have been used in solid-phase synthesis.' lo '06 Y.Iwakura K. Uno M. Oya and R. Katakai Biopolymers 1971 9 1419. lo' R. Katakai M.Oya K. Uno and Y. Iwakura Biopofymers 1971 10 2199; J. Org. Chem. 1972 37 327. lo* R. S. Dewey E. F. Schoenewaldt H. Joshua W. J. Paleveda H. Schwam H. Barkemeyer B. H. Arison D. F. Weber R. G. Strachan J. Milkowski R. G. Denkewalter and R. Hirschmann J. Org. Chem. 1971 36 49. lo9 H. R. Kricheldorf Angew. Chem. Internat. Edn. 1973 12 73 H. R. Kricheldorf and M. Fehrle Chem. Ber. 1974 107 3533. lo J. Halstr~rm and K. Kovacs Proceedings of the Twelfth European Peptide Symposium held at Schloss Reinhardsbrun German Democratic Republic 1972 'Peptides 1972' ed. H. Hanson and H.-D. Jakubke North-Holland Publishing Co.Amsterdam and London 1973 p. 173. Peptides and Proteins Z-Gly-L-Ala + L-Leu-OBzl 1' Z-G~Y-D-and -L-Ala-L-Leu-OBzl 1ii G~Y-D-and -L-Ala-L-Leu Reagents i coupling agent; ii H,-Pd Scheme 13 The degree of racemization of amino-acids can be determined conveniently by their reaction with an optically active N-carboxy-anhydride and separation of the diastereoisomeric products on an amino-acid analyser.' ' A racemization test involving separation in a similar way of glycyl-L- and D-alanyl-L-leucines (Scheme 13) has become a popular method of monitoring coupling methods; 0.1 "/ D-L-isomer can be detected in 5 pmol L-L-peptide.lI2 The use of isotopic dilution for determination of racemate content has extended the range of applica-tion of methods based on oligopeptide diastereoisomers down to the 1.0-0.001 level much below that accessible by other methods.' 'I1 A.V. Barooshian M. J. Lautenschlager J. M. Greenwood and W. G. Harris Analyt. Biochem. 1972,49 602. ' l2 N. Izumiya H. Muraoka and H. Aoyagi Bull. Chem. Soc. Japan 1971,44 3391. 'I3 D. S. Kemp Z. Bernstein and J. Rebek J. Amer. Chem. SOC.,1970 92 4756.