J. CHEM. SOC. DALTON TRANS. 1991 1651A Thermodynamic and Spectroscopic Study of theComplexes of the Undecapeptide Substance P, of itsN-Terminal Fragment and of Model Pentapeptidescontaining Two Prolyl Residues with Copper IonsLeslie D. Pettit,*ee Wojciech Bal! Michel Bataille,= Claude Cardon,= Henryk Kozlowski,bMarie Leseine-Del~tanche,~ Simon Pyburn8 and Andrea Scozzafavaa School of Chemistry, The University of Leeds, Leeds LS2 9JT, UKInstitute of Chemistry, University of Wrocla w, ul. Joliot- Curie 74,50383 Wroclaw, PolandLaboratoire de Chimie Biologique, L.A.2 17, Universite des Sciences et Techniques de Lille, 59655Laboratory of Inorganic and Bioinorganic Chemistry, University of Florence, Via G. Capponi 7,50 72 7Villeneuve dlscq, Cedex, FranceFirenze, ItalyEight pentapeptides have been synthesised which either are models of the N-terminal pentapeptide frag-ment of Substance P (Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met-NH,) or assist in understandingits co-ordinating ability; G ly- Pro- G ly- Pro- G ly, G ly- Pro- G ly- Pro-G lu, G ly- Pro- G ly- Pro- G I n, G ly- Pro-Lys- Pro-Gly, Arg- Pro-G ly- Pro- Gly, Arg- Pro- Lys- Pro-G In (Substance P,J, Gly- Pro- Pro-G ly- Gly andGlu- Pro- Pro-Gly-Gly. A potentiometric and spectroscopic study of the complexes formed with H +and Cu2+ and a potentiometric study of the complexes with Substance P have been made.Theresults demonstrate the profound effect which the prolyl residue can have, when incorporated in apeptide chain, on the formation of copper(ii)-peptide complexes.It acts as a break-point t o co-ordination and encourages the formation of folded peptide chains, through p turns, resulting in large,but very stable, chelate rings. The co-ordination behaviour of Substance P is almost identical to thatof the N-terminal fragment, Substance P,+, with chelation through the N-terminal amino N and the&-amino N of the Lys residue to form a large chelate ring of high stability, the peptide being forcedinto a bent conformation by the prolyl residue. With Substance P and its analogues, bondingbetween Cull and deprotonated peptide nitrogens is absent below pH 10 but with the penta-peptides containing the Pro-Pro unit co-ordination to the peptide N of a Gly residue takes placesurprisingly easily (starting at pH 7) to form a large chelate ring.Substance P is a peptide containing 11 amino acid residuesin the sequences Arg-Pro-Lys-Pro-Gln-Gln-Phe-Phe-Gly-Leu-Met-NH2."f2 It is a member of the group of tachykinins, afamily of biogenic peptides sharing the common C-terminalsequence Phe-Xaa-Gly-Leu-Met-NH, (where Xaa is an aminoacid residue) and to some extent common biological function^.^It is present in various parts of the central nervous system and inthe gastrointestinal tract of mammals.Its actions are believed toinclude the transmission of pain information, vasodilation,smooth muscle contraction, modulation of signal transmissionsin the striato-nigral pathway, evocation of local inflammatoryresponses and possibly also action as a messenger between theimmune system and theAt the present time three classes of tachykinin receptors arerecognised, conforming to three known mammalian tachy-kinins: Substance P (SP), and neurokinin A and B.Binding tothese receptors is quite selective, in contrast to the cross-reactivity observed for non-mammalian tachykinin-like pep-tides, which still share a common C-terminal sequence.6 TheC-terminal part of the SP molecule is recognised by thereceptor, the SP,-, fragment being a minimal active species inmost tests, especially outside the central nervous ~ y s t e m . ~ . ~However some neuronal test systems show the absoluterequirement for the N-terminal part 8,9 and enzymatic cleavaget Abbreviations used for amino acid residues are those recommendedby IUPAC-IUB.'of SP in biological systems results, among others, in theliberation of the peptide fragments SP1-4, SP,-,, (pyro-GluS)-SP,-, and SP,-,Numerous studies have been performed to elucidate therelationship between the structure and activity of SP.As a resultof the incomplete characterisation of a receptor site severalapplied approaches have been followed, including the testingof chemically modified SP analogues in various receptorsystems,".' ' conformational studies of the active forms invarious solvents and the~retical.'~-'~ These studies suggest thatthe active conformation of SP contains an @-helix beginning atabout the fourth residue and including most of the C-terminalregion. The N-terminal tetrapeptide of SP, apart from itsdistinct biological function, shows a unique solution feature.Inaqueous solution there exists an equilibrium between two forms,differing by a cis or trans configuration of the Lys-Pro peptidebond, the cis species being stabilised by ionic interaction of thelateral Lys-NHt group with the terminal carboxyl.' 5-16Substance P is able to release histamine from rat peritonealmast cells in a concentration-dependent manner l 7 and it isknown that histamine release in the early phase of theinflammatory response coincides with an increase in thecopper concentration in blood p l a ~ m a . ~ , ' ~ Copper activationof various neuropeptides, such as enkephalins,I8 ThyrotropinReleasing Factor (TRF) l9 and Luteinizing hormone Releas-ing hormone (LHRH),*' has been discussed. Taking thisinto account, it is appropriate to consider a possible link be-tween the two above physiological processes although ther1652 J.CHEM. SOC. DALTON TRANS. 1991is no known link between the activity of Substance P and copper.We have undertaken an investigation of the co-ordinationabilities of the SP fragments and their analogues towardscopper(I1) ions, with special attention to the role of the prolineresidues in the conformation of the ligand chain since it isknown that this residue can act as a 'break point' to metal ionc ~ - o r d i n a t i o n . ~ ~ - ~ ~ Proline, as the only naturally occurringamino acid containing a secondary nitrogen atom, is unable toform a N--Cu bond and therefore it induces a break in thenormal mode of co-ordination achieved by regular peptidessuch as tetraalanir~e.~' In the latter case co-ordination consistsof the deprotonation and binding of the terminal amino groupat low pH, followed by a stepwise deprotonation and binding,ofsuccessive peptide nitrogens as the pH is raised until the square-planar geometry of 4N donors is produced with the tetrapeptideheld in a tight, circular conformation.The presence of the Proresidue in the chain leads to very specific modes of bindingincluding the formation of dimers or of large chelate rings.21i22Since SP contains two Pro residues (in the second and fourthpositions) these could have a profound effect on its con-formation and ability to co-ordinate to Cu".In this paper we report the synthesis of five pentapeptideswhich are analogues of the SP N-terminal sequence: i.e.Gly-Pro-Gly-Pro-Gly, Gly-Pro-Gly-Pro-Glu, Gly-Pro-Gly-Pro-Gln, Gly-Pro-Lys-Pro-Gly and Arg-Pro-Gly-Pro-Gly andof the SP1-5 fragment, Arg-Pro-Lys-Pro-Gln, together withtwo pentapeptides which have the two Pro residues adjacentin positions 2 and 3 rather than separated by Gly or Lys:Gly-Pro-Pro-Gly-Gly and Glu-Pro-Pro-Gly-Gly. Complexesformed by these pentapeptides and by the undecapeptide,Substance P, have been studied potentiometrically and spectro-scopically using electronic absorption, circular dichroism (CD)and electron spin resonance (ESR) spectroscopy.A preliminaryaccount of the results has been presented recently.26ExperimentalPeptide Syntheses.-The pentapeptides were synthesised bystandard liquid-phase methods using, as starting materials:Bu'OCO-Arg(NO,), Bu'OCO-Gly, Bu'OCO-Lys(Z), Bu'OCO-Gln, Bu'OCO-Glu-OCH,Ph, HCl-Gly-OMe, HClOPro-OMeand HCl~Glu(OCH,Ph),.Glutamine benzyl ester hydrochloride.To Bu'OCO-Gln (3 g)in diethyl ether, with a few drops of ethanol, was added slowly anethereal solution of freshly prepared diazotoluene. Addition wascontinued until the vigorous reaction subsided and the reddishbrown colour of the diazotoluene persisted. After evaporation ofthe solvent, the Bu'OCO group was removed using 2 mol dm3HCL in ethanoic acid. Glutamine benzyl ester hydrochloridewas obtained as a white powder by addition of diethyl etherGeneral procedure.Coupling and deprotection steps were fol-lowed by TLC (Merck G60, eluent CHC1,-MeOH-MeC0,H85 : 15 : 5). Free pentapeptides were purified by chromatographyon Whatman 3 (eluent butanol-pyridine-acetic acid-water30: 20: 6 : 24).Coupling steps. C-Protected derivatives [in CHCl, ordimethylformamide (dmf) solution J were neutralised withtriethylamine. The Bu'OCO-amino acid was dissolved in thissolution (at 0 "C) followed by addition of dicyclohexylcarbo-diimide (dcci) (Merck) and 1-hydroxybenzotriazole (Aldrich) in10% excess and the solutions kept at 0 "C for 3 h. After standingovernight at room temperature, the dicyclohexylurea wasfiltered and the CHC1, solution washed with NaHCO, solution(5%, twice), water, HCl (l%, twice) and again with water.Deprotecting steps.After drying over MgSO, and evaporationof the solvent, protecting groups were removed as follows.Bu'OCO. The protected derivative was dissolved in a few cm3of dioxane and a cold solution (4 "C) of 4 mol dm-3 HCl indioxane added. After 30 min TLC plates showed completecleavage. The solvent was removed by evaporation and the oily(90%).material triturated with diethyl ether to give the hydrochlorideof the C-protected (or free) peptide.Methyl ester. The protected derivative was dissolved inmethanol, cooled to 0 "C and 1.1 equivalents of aqueous NaOHadded. When cleavage was complete, 10% HCl was added to pH3 and the methanol evaporated. The N-protected peptide wasextracted with ethyl acetate (three times) and dried overMgS04. It was recrystallised from ethyl acetate-hexane.Benzyl, nitro and benzyZuxycarbonyl(2) groups.The protectedpeptides, in ethanoic acid solution, were hydrogenated using10% Pd on charcoal (100 mg catalyst per 0.002 rnol peptide).The extent of cleavage was followed by uptake of H, and bychromatography on Whatman 3.As an example, the synthesis of Arg-Pro-Lys-Pro-Gln(Substance Pl-5) is outlined in Scheme 1.Purification of Pentapeptides.-The chloride ion of thecationic form was exchanged using AGlXS (Bio-Rad) in theacetate form. The resulting zwitterionic peptides were purifiedby gel filtration (Sephadex G-15; eluent water) followed byfreeze drying. All the pentapeptides gave a single spot by paperchromatography.The absence of acetate ions in the products was confirmed by400 MHz NMR spectroscopy and purity of the peptides waschecked by amino acid analysis, elemental analysis (wheresufficient samples were available) and fast atom bombardment(FAB) mass spectrometry. The pentapeptides containing theGln residue could not be checked by amino acid analysisbecause the amido group of the Gln side chain was removedduring acid cleavage of the peptide.Substance P was a commercial sample (Bachem), provided asthe acetate salt.The ratio of Substance P to acetate wasdetermined by potentiometric titration and the ratio checked ineach titration performed. A ratio of 1 : 2.246( f 0.009) was foundover seven titrations. The purity of all the ligands studied wasconfirmed by pH-metric titration. The calculation of proton-ation constants gave highly consistent results with low standarddeviations (better than 10 = 0.01) and random distributions ofresiduals, without any general trends, reflected in low values forx2.Analytical results are given in Table 1.Potentiometric studies.-Stability constants for complexes ofH + and Cu2+ were calculated from titration curves obtainedusing total volumes of about 1.5 cm3. Alkali was added from a0.1 cm3 micrometer syringe which had been calibrated byboth weight titration and the titration of standardisedmaterials. Experimental details were: peptide concentration0.003 mol drn-,; copper concentration 0.001-0.0028 mol dmP3;ionic strength 0.10 mol dm-, (KNO,); pH range forcomplexation 4-9.7; method, pH-metric titration, calibrated inconcentrations using HC10,;27 number of titrations three perligand; temperature 25 "C; method of calculation SUPER-QUAD.28Calculations were made with the aid of the SUPERQUADcomputer program which allows for the refinement of totalligand concentrations and was able to confirm the purity of thepeptides synthesised and, in particular, the absence of acetate.The commercial sample of Substance P contained an excess ofacetate ions, which could be included in the SUPERQUADcalculations once its concentration had been measured bypH-metric titration.Standard deviations (lo) quoted werecomputed by SUPERQUAD and refer to random errors only.They give, however, a good indication of the importance of theparticular species in the equilibrium.Spectroscopic Studies.-Solutions were of similar concen-trations to those used in the potentiometric studies, usingCu(C10,),*6H20 as the source of copper(r1) ions.Absorptionspectra were recorded on a Beckman UV 5240 spectrometerand CD spectra on automatic recording spectropolarimeters(JASCO-J-20 and Yvon-Jobin Mark 111 dichrograph) in thJ. CHEM. SOC. DALTON TRANS. 1991(70%)1653(ii)Bu'OCO-Arg (NO,) -OH + CI -H,'-Pro-OMe Bu'OCO-LYS (Z) -OH + CI H2+-Pro-OMe(7%) 0 ) I1 Bu'OCO-Arg (NQ) -Pro-OMe(79%) (ii)Bu'OCO-Arg (NO2) -PreOH(75%) ( i ) II Bu'~)CO-LYS (z) -Pro-OMe(90%) (iii)CI H2+-Lys (Z) -Pro-OMe(=%I ( 1 ) I BubCO-Arg (NO2) -Pro-Lys (Z) -Pro-OMetBu'OCO-Arg (NO2) -Pro-Lys (Z) -Pro + CI -H2*-Gln-OCH2Ph(61%) (iv) 1I1Bu'OCO-Arg (NQ) -Pro-Lys (Z) -Pro-Gln-OCH2Ph(91%) ( V )Bu'OCO- Arg- Pro-Ly s-Pro-Gln(90%) (vi)C I -H2+-Arg- Pro-Lys- Pro-GI nScheme 1(u) 10% Pd/C in MeC0,H; ( v i ) 2 mol dm-, HCI in MeC0,H(i) (a) NEt,, (b) Bu'OH4cci in CHCI,; (ii) NaOH in MeOH; (iii) 4 mol dm-3 HCI in dioxane; (iu) (a) NEt,, (b) Bu'OH4cci in dmf;800-200 nm region.ESR spectra were obtained on a JEOLJES-Me-3X spectrometer at 130 K and at 9.13 GHz, FAB massspectra on a VG AutoSpec spectrometer.ResultsStability constants of hydrogen-ion complexes are given inTable 2. Two of the pentapeptides synthesised, as well as SPitself, contain an Arg residue. The guanidine side-chain of thisresidue does not normally co-ordinate to Cu" or similarmetals." Protonation and complex formation by this side-chain was therefore not included in the equilibrium calculations.All ligands are assumed to have the empirical formula HL,charges are omitted for clarity.Stability constants of thecomplexes with Cu" are also given in Table 2 and the speciesdistribution curves for selected pentapeptides and for SubstanceP, calculated for 0.001 mol dmP3, are shown in Fig. 1.Spectroscopic data for the copper(n) complexes are shown inTable 3. With all the ligands in the presence of Cu", slowprecipitation was observed above pH 10.DiscussionFor the six pentapeptides without a Lys residue the protonationrepresented by PHL corresponds to protonation of the a-amino-nitrogen. The values for Gly-Pro-Gly-Pro-Gly, Gly-Pro-Gly-Pro-Gln and Gly-Pro-Pro-Gly-Gly are very close to that forGly-Pro-Gly-Gly (log p = 8.25) 2 1 with Gly-Pro-Gly-Pro-Glua little larger.The constant for Glu-Pro-Pro-Gly-Gly issomewhat lower as a result of the inductive effect of the carboxylgroup of the Glu side-chain. With Arg-Pro-Gly-Pro-Gly theconstant is significantly smaller as a result of the positive chargeon the Arg residue. In the case of peptides containing a Lysresidue (including SP itself) protonation of the a-amino andE-amino-nitrogens will overlap somewhat but protonation ofthe &-amino group of the Lys would be expected to occur athigher pH making it the greater contributor to the firstmacroc~nstant.~~ The second protonation constant of thesepeptides would then be the macroconstant for the second aminoprotonation, mostly the cc-NH2.The other protonation con-stants refer to carboxyl protonations. Substance P itself has nocarboxyl groups, hence it shows only two protonation constantswith values comparable to those for Arg-Pro-Lys-Pro-Gln.Copper(I1) Complexes.-A number of peptides, particularlywhen they contain the Lys residue, are able to form dimericcomplexes with Cu". In the systems studied there was noevidence of dimerisation in the ESR spectra and models used tofit the potentiometric data which included dimeric species wereeither rejected or gave results which were less satisfactorystatistically than those calculated for simpler systems whichomitted dimers. Spectroscopic results indicate that the majo1654 J.CHEM. SOC. DALTON TRANS. 1991Table 1 Analytical data for the pentapeptides synthesisedAmino acid analysisG1 y-Pro-Gly-Pro-G1 yG1 y-Pro-Gly-Pro-GluArg-Pro-GI y-Pro-GlyGly-Pro-Lys-Pro-GlyG1 y-Pro-Pro-G1 y-GlyGlu-Pro- Pro-G1 y-G1 yFAB-mass spectroscopy *G1 y-Pro-G1 y -Pro-G1 yG1 y-Pro-G1 y-Pro-GluGI y-Pro-G1 y-Pro-GlnG1 y-Pro-L ys-Pro-G1 yArg-Pro-Lys-Pro-GlnGI y-Pro-Pro-Gly-GlyGlu- Pro-Pro-G1 y-G1 yMajor peak383.4 + 1 (H)455.3 + 1454.3 + 1454.4 + 1624.5 + 1383.3 + 1455.3 + 1Elemental analysis (%)FoundGly Pro Others1.0 0.611.0 0.93 Glu0.511.0 1.01 Arg 0.481.0 1.02 Lys 0.461.0 0.641.0 0.97 Glu 0.52Minor peaks383 + 23 (Na) and 383 + 133 (Cs)455 + 23 and 455 + 134 (Cs + H)454 + 23,454 + 133454 + 23,454 + 133 and 454 + 156 (Cs + Na)624 + 23383 + 23,383 + 133455 + 23,455 + 133RequiredC H NGly-Pro-Gly-Pro-Gly 46.8 6.8 16.85Gly-Pro-Gly-Pro-Glu 45.65 6.7 14.05Gly-Pro-Gly-Pro-Gln 45.05 6.85 16.3Gly-Pro-Pro-Gly-Gly 46.15 6.7 16.5Glu-Pro-Pro-Gly-Gly 46.2 6.5 13.8* m/z values; matrix = m-nitrobenzyl alcohol.C H NC16H25N506-1.5H20 46.8 6.85 17.05C19H29Ns0,*2.5H20 45.6 6.85 14.0C19H30N607.3H20 44.9 7.1 16.5cl 6H25N506.2H20 45.85 6.9 16.65C19H29N508*2H20 46.4 6.75 14.1Table 2 Stability constants (log p values) for complexes of H+ and Cu2+ at 25 "C and I = 0.10 mol dm-3 (KNO,)Proton complexes (estimatedstandard deviations 0.01) Copper(I1) complexesHL H2L H3L Cu(HL) CuL CUL, &H-,LGly-Pro-Gly-Pro-Gly (C, ,H2,NSO6)Gly-Pro-Gly-Pro-Glu (C,,H,,N,O,)Gly-Pro-Gly-Pro-Gln (C, 9H30N607)Arg-Pro-Gly-Pro-Gly (C, ,H ,,N,06)Arg-Pro-Lys-Pro-Gln (C,,H,,N, 007)Substance P (C,,H,,N,,O,,S)Gly-Pro-Pro-Gly-Gly (C, 6H25N506)Glu-Pro-Pro-Gly-Gly (C, 9H29N&)Gly-Pro-Lys-Pro-Gly ( C ~ O H ~ , N ~ O ~ )8.368.538.297.4710.249.8510.068.307.9711.6712.93 15.3911.5110.8618.36 21.7717.12 21.1217.2811.5812.17 15.265.88(2) 10.60(2)6.57(5) 10.54(9)5.73(1) 10.16(2)4.83( 1) 8.96( 1)15.45( 1) 8.99( 1)14.30(4) 7.84(2)14.35(9) 7.73(3)5.38(3) - 1.36(3)5.18( 1) - 1.82(2)co-ordination site in all systems studied is the N-terminalnitrogen donor. At pH c9.5 two distinct copper(1r) complexesare observable using spectroscopic techniques.The energy ofthe d-d transitions as well as the charge-transfer bands showthese to be species with 1N and 2N co-ordination respectively.The pentapeptides Gly-Pro-Pro-Gly-Gly and Glu-Pro-Pro-Gly-Gly were the only ones which showed spectroscopicevidence for co-ordination between copper and amide nitrogenatoms, with characteristic charge-transfer bands in the CDspectra in the 315-325 nm region (see Table 3). With all theother peptides either one or two N-terminal amino nitrogenswere involved in the metal-ion binding but the peptidescontaining the Lys residue formed only mono complexes[Cu(HL) and CuL] while the others formed mono and biscomplexes. Hence, in terms of complex formation with copper,the peptides studied can be divided into three significantlydifferent groups depending on the amino acid sequence ofresidues 2 4 , namely Pro-Gly-Pro, Pro-Lys-Pro and Pro-Pro-Gly.Complexes of the three groups will therefore be consideredseparately.Potentiometric studies of the pentapeptides containing thePro-Gly-Pro sequence show the 1N and 2N species to be CuLand CuL, respectively with co-ordination through the amino-Nand the carbonyl oxygen of the neighbouring peptide linkage.With Gly-Pro-Gly-Pro-Gly and Gly-Pro-Gly-Pro-Gln thevalues for log pcuL are similar to that for Gly-Pro-Gly-Gly whilethe value for Arg-Pro-Gly-Pro-Gly is rather lower as a result ofthe positive charge on the Arg residue. With Gly-Pro-Gly-Pro-Glu, however, the CuL complex is of more significance thanwith the others, the stability constant being higher by about0.7 log units (see Table 2).This stabilisation must result fromco-ordination to the lateral carboxyl group of the Glu residuesince this is the only difference between this ligand and theGly5 analogue; this was difficult to accept initially as theresulting chelate ring would be 18- or 19-membered (dependingon whether the carbonyl 0 of the N-terminal Gly residue isco-ordinated). Independent support for such bonding wasobtained from the CD spectrum which had a very intense, broadband at 250 nm. This energy is on the border of the rangeaccepted for NH,-Cu charge-transfer transitions but is typicalof CO,--Cu bands.30 If this is the case, it is yet another strikingexample of the influence of the prolyl residue on the con-formation of peptide molecules.The effect of this increase instability resulting from the inclusion of Glu as the 0-terminaJ. CHEM. SOC. DALTON TRANS. 1991 1655'0°11001u) a,0 a,0. v1.-c0PHFig. 1 Species distribution curves for complexes of 1 : 2 mixtures of Cu"(0.001 mol drn-j) with (a) Gly-Pro-Gly-Pro-Gly, (b) Gly-Pro-Gly-Pro-Glu, (c) Arg-Pro-Lys-Pro-Gln (SP,-5), (d) Substance P, (e) Gly-Pro-Pro-Gly-Gly. Curves: 1, Cu2 +; 2, Cu(HL); 3, CuL; 4, CuL,; 5 , CuH-, Lresidue is evident from a comparison of the species distributioncurves shown in Fig. l(a) and (b).There was no spectroscopic evidence for bonding betweenpeptide nitrogens and Cu" (e.g. no CD charge-transfer bandsin the 300-330 nm region) and no potentiometric evidence forpeptide-nitrogen deprotonation.Values calculated for log p(CuL,) for these peptides also support the suggestion fromspectroscopy of 2N co-ordination through the amino nitrogensof two ligand molecules.The second group of peptides (those containing the Lys3residue, including SP itself) gave potentiometric models whichare very similar to one another but are distinctly different fromthe other ligands studied. Spectroscopic results showed 1N and2N complexes without Cu-N- (peptide) co-ordination. Thecombined potentiometric and spectroscopic results demonstratethat with Gly-Pro-Lys-Pro-Gly and Arg-Pro-Lys-Pro-Gln(SP,-,) the Cu(HL) species would be comparable to the CuLspecies with the other peptides, but with the &-amino group ofthe Lys residue protonated making it a 1N complex.Valuesfor log PCu(HL) corrected for this protonation (i.e. log PCU(HL) -log PHL) are entirely compatible with this suggestion. Thespecies CuL, shown by spectroscopic studies to be a 2Ncomplex, would then involve chelation of a second nitrogendonor from the side-chain amino group of the Lys residue toform a large chelate ring (13 or 14 membered, depending onwhether the carbony10 of the first Gly residue is co-ordinated).Again this chelation would be encouraged by the p turn ofthe Pro residue [comparable to similar chelation found with(Gly-Gly-Pro-Ly~),].~~ It is also supported by an X-ray studyof the Cu"-Lys-Tyr complex which suggests bonding from thelateral Lys-NH, group.31 Additional supporting evidence forco-ordination of the lateral Lys amino nitrogen comes fromthe ESR spectra for the 2N complexes which show a small butsystematic trend with complexes of ligands containing the Lys3residue having slightly larger values for All and marginallysmaller values for gll than the others. This could be caused by themore basic second nitrogen atom, as e.g.in the 2N complex withN(amino),N(amide) bonding (cf. Pro-Pro peptides in Table 3)and also by a more symmetrical surrounding to the metal ionwhich could be a result of the high flexibility of the long Lysside-chain. What is more, there was no evidence for normaldeprotonation of the &-amino nitrogen of the lysyl side-chain. Inthe absence of co-ordination this would be expected to occurabove pH 9.Substance P itself behaves in an almost identicalfashion to Arg-Pro-Lys-Pro-Gln (SP1-5), forming the samecomplexed species with almost the same stability constants. Thisis demonstrated by the almost identical species distributioncurves shown in Fig. l(c) and (d). It is therefore reasonableto assume that co-ordination will be the same involving 2N(a- and E-NH,) co-ordination without Cu-N- (peptide) bond-ing but with the peptide backbone locked in a bent conform-ation. Hence, as a ligand to Cu", Substance P behaves in analmost identical fashion to its N-terminal pentapeptide fragment.The 1N and 2N complexes formed by the third group ofpentapeptides coctaining the Pro-Pro-Gly unit were shown byboth potentiometry and spectroscopy to be the CuL andCuH-,L species.The CuL species of Glu-Pro-Pro-Gly-Gly isa little weaker than with Gly-Pro-Pro-Gly-Gly suggesting nosignificant co-ordination through the lateral carboxyl group ofthe Glu residue. The deprotonated complexes, however, mustinvolve chelation through the terminal amino N (and probablythe neighbouring carbonylo) and a deprotonated peptide N ofeither the Gly4 or Gly5 residues, with the formation of a largechelate ring of either 10 or 13 atoms, assuming additional co-ordination through the carbonyl 0 of the N-terminal Glyresidue.The stability of this large ring is unexpectedly high, being onlyabout an order of magnitude less than that for the CuH-,Lcomplex of tetraglycine (log p = -0.4)30 which forms a stablefive-membered chelate ring.It is more stable by an order ofmagnitude than the corresponding complex with Gly-Pro-GlyJ. CHEM. SOC. DALTON TRANS. 1991Table 3 Spectroscopic data far copper(I1) complexes with SubstancePl-5 ,and its 'model' peptapeptidesVisibleSpecies ~,,,"/nmGI y-Pro-Gly -Pro-GluCuL (1N) 725(49)CuL, (2N) 692(63)G1 y-Pro-G1 y-Pro-GlnCuL (1N) 732(35)CuL, (2N) 695(50)Arg-Pro-G1 y -Pro-GlyCuL (1N) 702(41)CuL, (2N) 660(61)Gly -Pro-Lys-Pro-GlyCu(HL) (1N) 720(29)CuL (2N) 662(56)Arg-Pro-Lys-Pro-GlnCu(HL) (1N)CuL (2N)Gly- Pro-Pro-Gly-G1 yCuL (1N) 750(30)CuH-,L (2N) 672(55)Glu-Pro-Pro-GI y-GI yCuL (1N) 745(40)CuH-,L (2N) 680(68)a Approximate absorption756( -0.02)' 160695( -0.25)' 160250 ( + 0.6)250( + 0.42) dse750( -0.03)' 159718( -0.10)' 169253( + 0.08)26q + 0.04)706( - 0.09) '682( -0.08) '262( + 0.19)d753( -0.09)' 159722( -0.15)' 175258( +0.41)d> 740( - 0.19)'260( - 0.2)715(-0.14)' 177270( - 0.1 3)730 (weak) 160645(+0.16)' 167270( + 0.20)260( - 0 .0 4 ) d3 15( - 0.18)228( - 1.90)'750(-0.06)' 171260( - 0.1 8)d226( + 0.67) '680( + 0.08) ' 180325( - 0.07)250( -0.45)d227( + 0.69)2.3252.2822.3322.2772.3332.2692.2702.3302.2762.3102.269coefficients (&/dm3 mol-' cm-') inparentheses. b A ~ in parentheses (dm3 mol-' cm-I). ' d-d Transition.NH,-Cu Charge-transfer transition. C0,--Cu Charge-transfertransition. N--Cu Charge-transfer transition.Intraligand transition.Gly (log p = -2S21) and a corresponding chelated complexwith Gly-Pro-Gly-Pro-Gly could not be detected. This suggeststhat the peptide N of a Gly3 residue in the Gly-Pro-Gly unitcannot chelate effectively and that the chelate ring in Gly-Pro-Gly-Gly must be t o the peptide N of Gly4, as suggested byspace-filling models and by the CD spectra of the CuH-,Lcomplex of G l y - P r ~ - G l y - P h e . ~ ~ Models show that thepresence of a Pro-Pro pair creates a very favourable con-formation for the formation of a large chelate ring spanningGlyl t o Gly4 in good agreement with the unexpectedly highstability found. This is demonstrated in the species distri-bution curves for the complexes of Gly-Pro-Pro-Gly-Glyshown in Fig.l(e).AcknowledgementsThis work was partly supported by grants from the PolishAcademy of Sciences (project C.P.B.P. 01.12) and from theBritish Council.References1 Nomenclature and Symbolism for Amino Acids and Peptides, PureAppl. Chem., 1987,56,595.2 M. M. Chang, S. E. Leeman and H. D. Niall, Nature (New Biol.),1971,232,86.3 P. C. Emson, F. J. Diez-Guerra and H. Arai, Neuropeptides and theirPeptidases, ed. A. J. Turner, VCH, Weinheim, New York; Horwood,Chichester, 1987, pp. 87-106.4 B. Pernow, Pharmacol. Rev., 1983,35,85.5 J. E. Morley, N. E. Kay, G. F. Solomon and N. P. Plotnikoff, LifeSci., 1987,41, 527.6 D. Regoli, G. Drapeau, S. Dion and P. d'orleans-Juste, Life Sci.,1987,40,109.7 J. R. Brown, A.B. Hawcock, A. G. Hayes, M. B. Tyers and R. G. Hill,J. Physiol. (London), l983,334,91P.8 M. E. Hall and J. M. Stewart, Fed. Proc., Fed. Am. SOC. Exp. Biol.,1986,45, 187.9 Z. Bar-Shavit, R. Goldman, Y. Stabinski, P. Gottlieb, M. Fridkin,L. Teichberg and S. Blumberg, Biochem. Biophys. Res. Commun.,1980,94, 1445.10 C. Poulos, G. Stravopoulos, J. R. Brown and C. C. Jordan, J. Med.Chem., 1987,30,1512.11 J. C. Foreman, C. C. Jordan, P. Oehme and H. Renner, J. Physiol.(London), 1987,335,449.12 R. Schwyzer, D. Erne and K. Rolka, Helv. Chim. Acta, 1986,69,1789.13 K. Rolka, D. Erne and R. Schwyzer, Helv. Chim. Acta, 1986,69,1798.14 D. Erne, K. Rolka and R. Schwyzer, Helv. Chim. Acta, 1986,69,1807.15 A. Otter and G. Kotovych, J. Magn. Reson., 1987,74,293.16 J. R. J. Sorenson, in Metal Ions in Biological Systems, vol. 14, ed.H. Sigel, Marcel Dekker, New York and Basel, 1982, pp. 77-124.17 C. M. S. Fewtrell, J. C. Foreman, C. C. Jordan, P. Oehme, H. Rennerand J. M. Stewart, J. Physiol. (London), 1982,330, 395.18 P. Sharrock, R. Day, S. Lemaire, S. St-Pierre, H. Mazarguil, J. E.Gairin and R. Haran, Znorg. Chim. Acta, 1982,66, L91.19 G. Formicka-Kozlowska, M. Bezer and L. D. Pettit, J. Inorg.Biochem., 1983,18,335.20 K. Gerega, H. Kozlowski, E. Masiukiewicz, L. D. Pettit, S. Pyburnand B. Rzeszotarska, J. Inorg. Biochem., 1988,33, 11.21 L. D. Pettit, I. Steel, G. Formicka-Kozlowska, T. Tatarowski and M.Bataille, J. Chem. SOC., Dalton Trans., 1985, 535,22 L. D. Pettit, I. Steel, T. Kowalik, H. Kozlowski and M. Bataille,J. Chem. SOC., Dalton Trans., 1985,2101.23 M. Bataille, L. D. Pettit, 1. Steel, H. Kozlowski and T. Tatarowski,J. Znorg. Biochem., 1985,24,211.24 C. Livera, L. D. Pettit, M. Bataille, J. Krembel, W. Bal andJ. Kozlowski, J. Chem. SOC., Dalton Trans., 1988, 1357.25 B. Decock-Le Reverend, L. Andrianarijaona, C. Livera, L. D. Pettit,I. Steel and H. Kozlowski, J. Chem. Soc., Dalton Trans., 1987,2221.26 L. D. Pettit, S. I. Pyburn, W. Bal, H. Kozlowski and M. Bataille,J. Chem. SOC., Chem. Commun., 1990,391.27 M. H. Irving, M. G, Miles and L. D. Pettit, Anal. Chim. Acta, 1967,38,475.28 P. Gans, A. Sabatini and A. Vacca, J. Chem. SOC., Dalton Trans.,1985,1195.29 G. Brooks and L. D. Pettit, J. Chem. SOC., Dalton Trans., 1976,42.30 H. Sigel and R. B. Martin, Chem. Rev., 1982,82,385.31 B. Radomska, M. Kubiak, T. Glowiak, H. Kozlowski and T. Kiss,32 M. Bezer, L. D. Pettit, I. Steel, M. Bataille, S. Djemil andInorg. Chim. Acta, 1989, 159, 11 1.H. Kozlowski, J. Znorg. Biochem., 1984,20, 13.Received 18th January 199 1; Paper 1/00262