6151 complexes, where ligands containing delocalized T orbitals usually cause much larger rate enhancements (cf. Table IV). For copper(II), we cannot unequivocably state whether this factor is due to increased labilization of remaining water molecules in the complex (axial or equatorial) or a facilitation of inversion. It is quite possible, in fact, that Cuz+, Cu(gly)+, and Cu(bipy) 2+ undergo substitution reactions via different mechanisms; the two former perhaps by axial substitution and inversion and the latter by direct equatorial substitution. In conclusion, we see that the stability of this mixed copper(I1) complex may be explained in terms of relatively simple interactions and statistical arguments. Further studies on analogous systems are being undertaken in our laboratories to determine the relative
importance of the structural and electronic features of the bound and attacking ligands. Clearly more work is necessary before we are able to account for the positive value of A log K for the 2,2’-bipyridyl-Cuz+pyrocatechol system.
Acknowledgments. We are grateful to Professor Donald B. McCormick (Cornel1 University, Ithaca, N. Y.) for helpful comments and his interest in this work. This work was supported by the Petroleum Research Fund through Grant No. 2982B (R. F. P.), the NIH through Research Grant No. AM-08721 (given to Professor D. B. McCormick), and by a research grant from the Schweizerischen Nationalfonds zur Forderung der wissenschaftlichen Forschung (H. S.).
Cobalt (111) -Promoted Hydrolysis of Glycine Amides. Intramolecular and Intermolecular Hydrolysis Following the Base Hydrolysis of the cis-[Co( en) ,Br ( glyNR,R,] ’+Ions D. A. Buckingham, D. M. Foster, and A. M. Sargeson Contribution from the Research School of Chemistry, Australian National University, Canberra 2600, Australia. Received March 19, 1970 Abstract: Base hydrolysis of the cis-[Co(en)zBr(glyNRlRz)]2+ ions over the pH range 9-14 results in two paths for the production of [C~(en)~(gly)]z+.Following loss of Br- (kOH= 260 i 20 M-l sec-’, p = 1.0, 25”), competition for the five-coordinate intermediate by solvent and amide carbonyl oxygen results in cis-[Co(en)z(OH)(glyNRIRz)]~+ and [C0(en)~(glyNR~R~)]3+ species in the ratios 54: 46 (R1 = RZ = H), 34: 66 (R1 = CH3; R2 = H), and 18 : 82 (R1= Rz = CH3). The two paths have been isolated, and I80-tracerresults support the product distribution (R1= Rz = H)and demonstrate intra- and intermolecular hydrolysis in the hydroxoamide and chelated amide, respectively. Hydrolysis in the cis-[C~(en)~(OH)(glyNH~)]~f ion is faster by at least a factor of 10 than loss of Br- at pH 9 and 13, requiring a rate at least 107, and possibly more than loll, times faster than hydrolysis for uncoordinated glycine amide. Stereochemical studies show 80% retention of configuration in the hydroxoamide path and 75% retention in the chelated amide path. The significance of the results is discussed in relation to the “carbonyl” and “hydroxide” mechanisms entertained for hydrolytic enzymes.
P
revious studies have discussed the base hydrolysis of cis-[C~(en)~X(glyOR)]~+ (X = C1, Br)1,2 and /3,-[Co(trien)Cl(gly0C2H5>1’+1, in terms of two competing processes: intermolecular hydrolysis of the chelated ester and intramolecular attack of coordinated OH- at the carbonyl center of the monodentate ester. Both paths arose from competition for the five-coordinate deprotonated intermediate formed on loss of halide ion2S3 Evidence for these paths came entirely from l*O-tracer results, and the interpretation remained partly equivocal in that both reactions were fast compared to hydrolysis of coordinated halide and could not be observed independently.
A recent study has demonstrated that N,O chelated glycine amides in [ C O N ~ ( ~ ~ ~ N R ~ R (R1, , ) ]Rz ~ += H, CH,) base hydrolyze more slowly by a factor of -lo5 than the corresponding chelated esters. This rate difference allows hydrolysis in the chelated amide to be observed following loss of Br- in cis-[Co(en)zBr(glyNR1R2)Iz+,and also allows that intramolecular hydrolysis by coordinated OH- might be observed as a separate reaction. Such studies have direct relevance to the metal ion catalyzed hydrolysis of the amide bond in amino acid amides and peptides. In this paper the results of a kinetic, stereochemical, and l*Otracer study on the base hydrolysis of the cis-[Co(en)sBr(glyNR1Rz)]2+ions are reported,4a
(1) Abbreviations used in this article are as follows: en = ethylenediamine; trien = triethylenetetramine; N4 = trien or (en)%;gly0 = N-bound monodentate glycinate anion; glyOR = glycine alkyl esters: glyNRz = glycine amides; gly = chelated glycinate anion, (2) D. A. Buckingham, D. M. Foster, and A. M. Sargeson, J . Amer.
(4) D. A . Buckingham, C. E. Davis, D. M. Foster, and A. M. Sargeson J . Amer. Chem. Soc., 92, 5571 (1970). (4a) NOTEADDEDIN PROOF.In a recent article [ S . C. Chan and F. K. Chan, Aust. J . Chem., 23, 1175 (1970)], it is reported that base hydrolysis “definitely” results in a stable product at 0” of cis-[Co(en)zCI(glyNH~)]z+ [Co(en)~OH(glyNHn)]2+,which only on acidification forms the N,O-
Chem. SOC.,91,4102 (1969). (3) D. A. Buckingham, D. M. Foster, L. G. Marzilli, and A. M. Sargeson, Znorg. Chem., 9, 11 (1970).
The present results do not subchelated amide, [Co(en)z(glyNH~)]~+. stantiate their proposals.
Buckingham, Foster, Sargeson 1 Co(III)-Promoted Hydrolysis of Glycine Amides
6152 aliquots were withdrawn periodically. From these samples [Co(en)g1y](HgI4) was recovered, and the l80content of the glycine Analar reagents were used throughout without further purificadetermined as previously described.@ tion. Glycine N-methyl and N-dimethyl amides were prepared utilMeasurement of Oxygen Exchange in l*O-Labeled [ C ~ ( e n ) ~ g l y ] ~ ~ izing 5,5-dimethyl-2-cyclohexene-l,3-dione(dimedone) as an Nin Labeled Produced cia Base Hydrolysis of [C~(en)?(glyNH),]~+ protecting g r o ~ p . ~ ,N-(5,5-Dimethyl-2-cyclohexen-l-on-3-yl)~ ( 5 g) in enriched Solvent. [Co(en)2(glyNH~)](N03)2C104~Hz04 glycine ethyl esterE was treated with either 40% aqueous methylwater (50 ml, 1.3 atom% H2180) was hydrolyzed at pH 9.0 and amine or dimethylamine. The (N-dimedone)glycineamide product 25" for 24 hr, by pH-stat titration against 30% NaOH. On addiin each case was filtered from solution and recrystallized from This product (4.5 g) tion of excess NaI, [ C ~ ( e n ) ~ g l yprecipitated. ]I~ ethanol at 70"; the dimedone protecting group was then removed was washed and dried as above, converted to the chloride salt by with bromine' to form the glycine N-methyl and N-dimethyl amide shaking with excess AgC1, made up to 50 ml with 0.1 M HCIO,, hydrobromide salts, respectively. These salts were recrystallized p = 1.0 (NaClOJ, and thermostated at 25"; 10-ml aliquots were from methanol at 30", washed with ether, and dried in an evacuated periodically withdrawn and treated as described above. desiccator. Anal. Calcd for NHzCHzCONHCHa.HBr: C, Resolution of cis-[Co(en)~Br(glyNH2)1(ClOr)iand Measurement 21.30; H , 5.37; N, 16.58. Found: C, 20.95; H, 5.39; N,16.79. of Optical Retention in the Product of the Hgz+-Induced Hydrolysis Calcd for NH2CH2CON(CH3)2.HBr: C, 26.22; H , 6.06; N, in Acid Solution. cis-[C~(en)~Br(glyNH~)](CIO,)~ (11.3 g) was 15.31. Found: C,25.97; H , 5.76; N, 15.51. dissolved in hot, dilute acetic acid (20 ml, p H 3) and ammonium Pmr spectra were recorded on a Varian 100-mHz spectrometer, d-bromocamphorsulfonate (NH4-(+)-BCS, 13.9 g) added. Ethanol and visible spectra on a Cary 14 spectrophotometer. Spectro(75 ml) and acetone (75 ml) were added, and on scratching and photometric rates were obtained using a Cary 14 (I-cm cell) or a cooling in an ice bath, ( -)jss-[Co(en)2Br(glyNHn)l-(+)-(BCS)r Durrum-Gibbs stopped-flow reactor (2-cm cell). Some cobalt slowly crystallized. On reduction of the solution volume, and estimations were made using a Techtron AA4 atomic absorption further addition of acetone and ethanol, three further fractions spectrophotometer. a x values for optically active complexes were of similar activity were obtained, The four fractions were cornmeasured at 25 with a Perkin-Elmer P22 spectropolarimeter, using bined and recrystallized from hot dilute acetic acid by cooling and a 1-dm cell. The following radiometer apparatus was used in the adding ethanol (5.5 g). A 0.1 % solution in 0.01 M HClO4 gave measurement of buffer pH and in pH-stat titrations: TTA3 elecajsg +0.001 O, ajpe+0.073", whence [ a ] j 8 ~ -0 and [a];46+73 '. trode assembly, ABU 1 autoburet, TTT 1 titrator, SBRZtitrigraph, A m l . Calcd for ( -)58y-[Co(en)zBr(glyNHz)]-(+)-(CloHlaBr04S)z: and pHA scale expander. In the pH-stat titrations, the titrant C, 32.75; H, 5.29; N, 8.82. Found: C, 32.84; H, 5.60; N, (NaOH) was added under a nitrogen atmosphere to the continuously 8.56. The diastereoisomer was converted to the bromide salt by stirred solution. Separation of reaction products was achieved trituration with excess NaBr in water (3 ml) and was recrystallized using Bio-Rad Analytical Dowex 5OW-X2 (200-400 mesh) cationfrom hot dilute hydrobromic acid by addition of NaBr and cooling. exchange resin. The I 8 0 content of COl recovered from the laThe product (2.0 g) was washed with methanol and acetone and airbeled compounds was determined using an Atlas M-86 mass specdried. A 0.1% solution in 0.01 M HCIOl gave ajsg-0.116" and trometer. aj46 +0.018", whence - 116" and [a]jdB+18". A n d . Calcd Preparation of Complexes. cis-[Co(en)~Br(glyNH~)]Br~, cisfor (-)58y-[Co(en)zBr(glyNH2)]Brl: C,14.62; H, 4.50; N, 17.05. [Co(en)?(gIyNHCH3)]Br2,and cis-[C~(en)~Br(glyN(CH~)~)]Br~ were Found: C, 14.85; H,4.64; N, 17.17. prepared from rrar~s-[Co(en)~Br~]Br .HBr and the appropriate The filtrate remaining after removal of the (+t):146 diastereoisomer glycine amide hydrobromide using the method described by Alexwas reduced to dryness and dissolved in methanol (30 ml). On ander and Busch for the preparation of analogous glycine ester crystallized; addition of excess NaBr, ( +)5s~-[Co(en)zBr(glyNHz)]Brz complexes.s it was recrystallized from hot dilute hydrobromic acid by addition The product complexes were washed with acetone and air-dried. of NaBr and cooling The product (1.6 g) was washed and dried Ana/. Calcd for [C~(en)~Br(glyNH~)]Br~: C, 14.62; H, 4.50; as before. A 0.1 % solution in 0.01 M HC104 gave aZ8g+0.118' N, 17.05. Found: C, 14.43; H , 4.71; N, 16.84. Calcd for and ajq6-0.011", whence [a]38g118" and [a1546 -11". A u d . [C0(en)~Br(glyNHCH3)]Br~:C , 16.08; H, 4.63; N, 16.07. Calcd for ( +)j8s-[Co(en)zBr(glyNHz)]Br2: C, 14.62; H , 4.50; Found: C, 16.36; H , 4.48; N, 16.38. Calcd for [Co(en)2BrN, 17.05. Found: C, 14.95; H , 4.50; N, 17.00. (glyN(CH3)2)]Brz: C, 17.88; H, 4.88; N, 15.65. Found: C, A partial resolution of cis-[Co(en)~Br(glyNH~)]Br~ was also 17.90; H , 5.08; N, 15.25. The amide and N-methylamide comachieved by chromatography on Biorad Cellex-P weakly acid cationplexes were converted to their perchlorate salts by dissolution in hot exchange resin.'o The complex (0.05 g) in water (100 ml) was dilute HC104, followed by addition of excess NaC104 and cooling sorbed onto the Hi-form resin (70 X 2 cm) and eluted with 0.05 M in an ice bath. They were washed with ethanol and air-dried. HC1. Eluate aliquots (6 ml) were estimated for Co by atomic abAtla/. Calcd for [C~(en)~Br(glyNH~)](ClO~)~: C, 13.54; H , sorption spectroscopy, and the ajgOvalues measured. Sixteen 4.17; N. 15.80. Found: C, 13.60; H , 4.49; N , 15.98. Calcd aliquots showed optical activity; the first samples were inactive, the C, 15.39; H, 4.43; N. 15.39. for [C0(en)~Br(glyNHCH~)](Cl0~)2: next eight (+):,88r and the last eight ( - ) 5 8 g . Maximum specific Found: C, 15.55; H , 4.53; N , 15.12. The following absorption rotations obtained were [a]j8e +SI" and -64", representing 68"< maxima and absorptivities were obtained in dilute acetic acid (pH and 54% of the activity of the optically pure forms prepared above. 545 i 2 nm (E 84 & I); 5) at 25": cis-[C~(en)~Br(glyNH~)](CIO~)~, ( -)jsg-[Co(en)zBr(glyNH2)]Br2(0.0519 g, [M]w - 572") was discis-[(C0(en)~Br(glyNHCH~)](C10~)~, 545 & 2 nm ( E 87 & 1); cissolved in a 0.305 MHgZf-1.0 MHCIOl solution (25 ml) and allowed [C~(en)~Br(glyN(CH~)~)]Br~, 545 & 2 nm (E 86 C 1). These to stand at 25" for 90 min. A 5-ml aliquot, diluted to 10 ml, gave values remained unchanged in 1 M NaC104 (pH 5). u ~ -0.350", ~ ~ , whence [MIjss -1662". The solution was diluted Measurement of Oxygen Exchange in lBO-Labeled [ C ~ ( e n ) ~ g l y ] ~ +with water and sorbed onto an H+-form resin, from which it eluted Produced cia Base Hydrolysis of cis-[Co(en)rBr(glyNH2)12+in as a homogeneous 3+ band ([C~(en)~(glyNH~)]~') using 2 A4 HCI. Labeled Solvent. cis-[Co(en)?Br(glyNHr)]Br2 (8 g) in enriched The eluate was taken to dryness and redissolved in 50 ml of water. water (80 ml, 2.0 atom % l8O)was hydrolyzed at pH 9.5 and 25' for The Co concentration was estimated spectrophotometrically (648; 3 hr, by pH-stat titration against 30% NaOH. The solution was M , and the op98 for [ C ~ ( e n ) ~ ( g l y N H ~a), ) ][Co] ~ + = 1.84 X then taken to pH 7 with 12 M HC1 and reduced to ea. 50 ml on a tical activity measured spectropolarimetrically, a j g g -0.309". rotary evaporator at ca. 30". Excess NaI and methanol were added, whence [MIssg - 1679". The solution was then hydrolyzed at pH and [Co(er~)~gly]I~ precipitated, after acidification to pH -3, addi9.0 for 8 hr at 25" by pH-stat titration against 0.2 A4 NaOH, tion of Na2S2O4,and warming t o remove 13-. This material was quenched to pH 2 with 6 M HCI, and sorbed and eluted from an collected, washed with cold NaI solution, methznol, and ether, and H+-form resin, using 2 M HC1. The homogeneous 2+ eluate band recrystallized by dissolution in hot water and addition of NaI. The ([Co(en)2gIyl2+)was taken to dryness and redissolved in 50 mi of final product (2 g) was washed as before and dried in an evacuated ~ + , ~=~ 1.63 X IO-' M. water Using edBi 98 for [ C ~ ( e n ) ~ g l y ][Co] desiccator. The dried material was shaken with excess AgCl in Also, a589-0.253'; hence [MIj53- 1550". water (20 ml) for 5 min and the precipitate of AgCl and AgI reProduct Analysis by Ion-Exchange Chromatography. (A) Separamoved. The filtrate was made up to 50 ml, 0.1 M HCIOd, @ = tion of the Products of Base Hydrolysis. The cis-[Co(en)zBr1.0 (NaC104). This solution was thermostated at 25" and 5-ml
Experimental Section
Halpern, Aust. J . Chem., 18, 417 (1965). (6) B. Halpern and L. B. James, ibid., 17, 1282 (1964). (7) B. Halpern and L. B. James, Nurure (London), 202, 592 (1964). (8) M. D. Alexander and D. H. Busch, Inorg. Chem., 5, 602 (1966). ( 5 ) B.
Journal of the American Cheniical Society
92:21
(9) D. A. Buckingham, D. M. Foster, and A. M. Sargeson, J . Amer. Chem. Soc., 90, 6032 (1968). (10) Y. Yoshikawa and K. Yamasaki, Inorg. Nucl. Chem. Lett., 4, 697 (1968). (11) I. IC=O moieties; that oxygen resistant to exchange in 0.1 M H+ at 25' was attributed to the Co-0 sec-' at 25" to the oxygen, and that with k carbonyl oxygen atom. The same analysis here requires the path proceeding via the chelated amide inU H ? termediate to result in exclusive >C=O label and the path via the hydroxoamide intermediate to result in A similar result is implied by the results for path B, label exclusively in the Co-0 position. The former and it is suggested that only the cis-hydroxoamide path therefore involves intermolecular attack of solvent results in [C~(en)~gly]~+. This analysis is consistent OH- at the cobalt(II1) activated carbonyl carbon withwith the slightly larger retention value found for the out opening of the chelate ring, and the latter path hydroxoamide path, and we tentatively propose that involves intramolecular attack of coordinated OHthe small amounts (2-5z)of brown material found at the carbonyl carbon of the monodentate amide. on the ion-exchange column may result from subseThis analysis is shown in the proposed mechanism quent reactions of the tran~-[Co(en),(OH)(glyNH~)]~+ (Scheme I) for decay of the five-coordinate deprotonated ion. In support of this aspect of the mechanism it intermediate of A configuration. is found that the related trans-[Co(en),(OH)(glyO)]+ It is of interest to note that the contribution of the ion does not rapidly19 generate [ C ~ ( e n ) ~ g l y ] ~and +, two paths to hydrolysis is similar for glycine isopropyl trans-cis isomerization in [Co(en),(OH)NH,I2+ is a ester and glycine amide, -50 % each, implying similar very slow process.26 The property of an inbuilt nucompetition between water and carbonyl oxygen in cleophile to compete with the solvent or added anions the intermediate. This correspondence may be exfor coordination is a sensitive method for investigating tended to include the monodentate glycinate anion23 the stereochemical and electrophilic properties of reand monodentate ethanolamine; l8 however, the coractive coordinately unsaturated intermediates. This relation collapses when the amide nitrogen atom is aspect of the present study is being further investigated substituted, Table 11. and results pertaining to these problems will be reIt is also pertinent that the amide nitrogen atom ported in a subsequent paper. does not compete effectively with the carbonyl oxygen The proposed mechanism for hydrolysis is similar for the vacated site in the five-coordinate intermediate. to that favored for hydrolysis of the [Co(en)zX(glyBoth products would involve five-membered chelate OR)I2+ ions, X = C1, Bre2 In the latter study possible rings, and the N,N-bound isomer is certainly stable mechanisms involving hydroxide attack on the monoto subsequent reaction under the conditions, once dentate ester either before or after Br- removal could formed.13 Also, both the N- and 0-bound isomers not be rigorously excluded, since it was not possible of monodentate formamide in [ C O ( N H ~ ) ~ ( N H ~ C H O ) ] ~to+ isolate the two paths. In the present instance such are known, although they are prepared from neutral possibilities are excluded. Similarly, the possibility of nonaqueous The present result is not sur~ duality of mechanism, with synergic S N displacement prising when it is considered (1) that carbonyl oxygen of B r by carbonyl oxygen, accompanying S N ~ C entry B is more basic than amide nitrogen in organic amides of water is unlikely, since this would imply either and ( 2 ) that the N,N product expected from amide 100% retention or inversion of configuration in the nitrogen competition is deprotonated at this nitrogen, [ C ~ ( e n ) ~ g l y product ]~+ formed via the former path. and the most basic center in this complex ion is the Relative Rates of Inter- and Intramolecular Hydrolysis. carbonyl oxygen. l 3 These results, and other studies The significance of the coordinated nucleophile in carried out in these laboratories, 2s suggest that amide facilitating hydrolysis of monodentate glycine esters nitrogen competition in aqueous solution occurs only has previously been demonstrated for NHz- l 3 and under strongly basic conditions where some amide anOH-.2v3 A similar large rate enhancement for NHRion exists, and where the kinetically preferred 0-bonded and OH- lysis at a saturated carbon center has been product is susceptible to base-catalyzed dissociation. observed. l8 The present study extends these results The stereochemical results require that some raceto neighboring-group effects in amide hydrolysis. mization accompanies both paths, with 75 f 2 % The bimolecular rate constant for base hydrolysis retention of configuration via the chelated amide of monodentate Co(II1)-bound glycine amide in [(NH&(path A), and 80 f 2 % retention via the hydroxoamide Co(glyNH2)]3+ is unknown. However, if the influence (path B). Similar values were obtained at pH 9 and of the cobalt(II1) center is similar to the 80-fold en13, suggesting pH independence over this range. Other hancement found for N-bound monodentate glycine results on closely related ions20 have been interpreted ethyl ester, l 3 then the former value may be estimated in terms of a single kinetically significant deprotonated ~ activation of the carbonyl at -0.2 M-1 ~ e c - l . ~Direct reactant and five-coordinate intermediate, and the fact center in the chelated amide complex [Co(en)z(glythat in the present example only cis coordination by ",)I3+ results in significant additional activation. Hythe amide carbonyl group can obtain the retention drolysis presumably occurs by bimolecular attack of value implies that the five-coordinate intermediate canOH- at the carbonyl carbon with a rate constant of not be of a symmetrical form involving deprotonated 25 M-' ~ e c - ' ; ~this represents a rate enhancement glycine amide. of lo4over the uncoordinated molecule.
-
(24) R. J. Balahura and R. B, Jordan, J . Amer. Chem. SOC., 92, 1533 ( 1970).
(25) D. M. Foster, unpublished results,
(26) D. F. Martin and M. L. Tobe, J . Chem. Soc., 1388 (1962). (27) The bimolecular rate constant for glycinamide hydrolysis at 25", p = 0.1, is 2.2 X 10-3 M-1 sec-1: H. L. Conley and R. B. Martin, J . Phys. Chem., 69, 2914 (1965).
Buckingham, Foster, Sargeson
/ Co(Ill)-Promoted Hydrolysis of GIycine Amides
6158
The present study demonstrates that amide hydrolysis KH 0 / \: The product analysis results demonstrate that this must be at least as fast as removal of Br- at pH 9 (2.6 X sec-I) and pH 13 (26 sec-I), and the failure to observe any [Co(en)z(OH)(glyNHz)12+ at the isosbestic point for Br- removal suggests that internal hydrolysis is considerably faster than this. The similar [C~(en)~(OH)(glyO>]+ ion, formed in -50% yield on hydrolysis of cis-[Co(en)2Br(glyO)],+ is easily distinguished from [Co(en),3+ at the isosbestic point gIy]*+ and [C~(en)~(glyNH,)] for bromide removal. 2 1 Also, the cis-[Co(en),X(gly“,)Iz+ and cis-[C~(en)~X(gIyO)]+ ions (X = C1, Br) have almost identical spectra between 350 and 650 nm, and this correspondence is likely to extend to the hydroxo ions, X = OH-. On the basis of these similarities, we estimate that the inability to detect any hydroxoamide species implies that hydrolysis in this ion is faster by at least a factor of 10 than loss of Br-. If hydrolysis were pH independent, as would be implied by rate-determining attack of bound hyIv droxide,I8 then the results in 1 M NaOH require the bony1 function. Other discussionz8 suggests that dirate constant for intramolecular amide hydrolysis, k, valent metal ions compare favorably with Co(II1) in to be at least 2.6 X lo3sec-’, which at pH 9 corresponds promoting hydrolysis in model systems, and we believe to rate enhancements over the chelated and uncoorthat there is no chemical reason to prevent many didinated substrates of 3 IO7 and 3 IO”, respectively. valent metal ions behaving by mechanisms similar to Two factors are striking if this mechanism holds: (1) those delineated here. Thus, provided amide nitrogen the rate of hydrolysis will be maintained at biological can be prevented from coordinating, and provided pH’s, since the complex will still exist essentially coordinated water is sufficiently acidic to exist in the in the hydroxo form (pK, 6); and (2) that despite hydroxo form, the present experiments suggest that the greatly reduced basic character of coordinated OHdivalent metal ions or aquo complexes containing one (-IO8), it is apparently a much more efficient nucleophile or more coordinated water molecules might well hythan solvent OH- molecule for molecule (ca. 310Q). drolyze amide or ester linkages cia an intramolecular 0 II process. This conclusion may have relevance to the action of divalent-metal-ion catalysis in the various hydrolytic enzymes. For instance in Zn2+-activated bovine carboxypeptidase A, an enzyme of current topical interest, two pechanisms consistent with crystallographic s t ~ d i e s *and ~ , ~the ~ kinetic studies3’ are (1) the Zn-carbonyl mechanism,32 in which a coor\ I *-ego dinated water molecule is displaced by the carbonyl oxygen of the peptide, resulting in activation of the Alternatively, if attack of coordinated OH- were carbonyl group toward nucleophilic attack; and (2) not rate controlling, a rate law first order in base is the Zn-hydroxide mechanism, in which Zn2+ increases likely. This situation, which obtains for intramolecular the “availability” of bound OH-. Davis33 has prolysis of glycine ethyl ester by coordinated NH2-,I3 posed a similar Zn-hydroxide mechanism for the zinc would result from rate-determining loss of NH, from metalloenzyme carbonic anhydrase. The present study the deprotonated amino-diol intermediate (see I IV). offers a direct appraisal of the relative efficacies of This mechanism would require an overall rate conthe metal-ion promoted “carbonyl” and “hydroxide” stant of at least l o 4 M-’ sec-’, which is larger by a mechanisms, and suggests that the latter path is far more factor of -lo3 than that for hydrolysis in the Co(II1) efficient. chelated ion, and larger by a factor of lo’ than the Acknowledgment. The authors wish to thank Dr. second-order rate constant for hydrolysis in the organic F. Bergerson and Mr. G. Turner of the CSIRO Division molecule. of Plant Industry for the isotope ratio measurements. Concluding Remarks. The remarkable facility of cobalt(II1) in inducing hydrolysis of amino acid esters (28) D. A. Buckingham, manuscript in preparation. and amides is sufficiently dramatic to have relevance to (29) W. N. Lipscomb, M. L. Ludwig, J. A. Hartsuck, T. A . Steitz, H . Muirhead, J. E. Coppola, G. N. Reeke, and T.A . Quicho, Fed. Proc., metal-ion-catalyzed hydrolysis of these and similar Fed. Amer. SOC.Exp. Biol., 26, 385 (1967); T. A. Steitz, M . L. Ludwig, substrates in biological systems. The present results F. A , Quicho, and W. N. Lipscomb, J . Biol. Chem., 242, 4662 (1967). (30) W. N. Lipscomb, Accounts Chem. Res., 3 , 81 (1970). demonstrate that the cobalt(II1)-induced intramolecular (31) B. L. Vallee and J. F. Riordan, Annu. Rea. Biochem., 38, 733 hydrolysis reaction is at least lo7, and possibly more (1969); Brookhaven Symp. Biol., 21, 91 (1968). than lo”, times faster than the uncatalyzed reaction, (33) I. M. Klotz in “The Mechanism of Enzyme Action,” W. D. McElroy and B. Glass, Ed., Johns Hopkins University Press, Baltimore, and that this mechanism provides a pathway for hyMd., 1954, p 257. drolysis which is decidedly more efficient than that (33) R. P. Davis in “The Enzymes,” Vol. 5, 2nd ed, P. D. Boyer, Ed., provided by direct metal-ion polarization of the carAcademic Press, New York, N. Y., 1961, p 545. uia coordinated OH- is rapid.
-
-
Journal of the American Chemical Society
92.21 1 October 21, 1970
