Nucleophilicversusgeneral base catalysis in phosphyl (PV) group transfer: application to alpha;-chymotrypsin action

摘要

1976 515 Nucleophilic versus General Base Catalysis in Phosphyl (Pv) Group transfer : Application to a-Chymotrypsin Action By Kenneth T. Douglas and Andrew Williams,' University Chemical Laboratories, Canterbury, Kent The hydrolysis of aryl dimethylphosphinates is shown to be catalysed via a nucleophilic pathway by a series of bases including imidazole which is less efficient than phosphate dianion. The more reactive nucleophilic pathway is allowed because dimethylphosphinate is less sterically hindered than diphenyl phosphinate where general base- catalysis predominates. Acylation of a-chymotrypsin by 4-nitrophenyl diphenylphosphinate, a bona fide general base mechanism, has a low solvent deuterium oxide isotope effect not characteristic of such a mechanism. STUDYof the factors controlling the choice between nucleophilic and general base routes in catalysed solvolyses of Pvesters is, as yet, at an embryonic stage compared with the situation for carboxylic esters.Important contributions have been suggested to include the nature of the leaving group: the nucleophile,lI2 strain in the substrate,s and steric Recently, it was proposed that the imidazole-catalysed hydrolysis of aryl diphenylphosphinates (I) involved general base activated attack of water because of the high steric constraints which would be imposed on a transition-state for the nucleophilic mechanism (I)? We report a study of the hydrolysis of a series of ten aryl substituted S. A. Khan and A. J. Kirby, J. Chein. SOC.(B),1970, 1172.G. 0. Dudek and F. H. Westheimer, J. Anzev. CJtena. SOC., 1969, $1,2641. dimethylphospliinates (11)in which much of the steric hindrance is lifted and show that this ' steric relief ' does N alter the mechanism from general base to nucleophilic attack. We also use the effect of large steric require- ments which can ' force ' the reaction to take a general A. Williams and R. A. Naylor, J. Chmz. SOC.(B),1971, 1967. H. J. Brass and M. L. Bender, J, Amer. Chem. SOL, 1972,94, 7421. 516 base pathway to investigate the still vexing nucleophilic- general base ambiguity in chymotrypsin catalysis. EXPERIMENTAL MaleriaZs.-The dimethylphosphinate esters, (CH3),-P(0)OAr, were prepared by non-aqueous Schotten-Bauman proccdures from the phenols and dimethylphosphinoyl chloride; the route to the latter common precursor involved steps detailed in the literature reaction (l).Bisdimethyl-cm, (IY) phosphinyl disulphide (111) was prepared by the method of Parshall Q and was recrystallised from ethanol as white needles, n1.p. 230' (lit.,6 227'). Dimethylphosphinoyl chloride (IV) was prepared by the method of Pollart and Harwood as a yellow solid, m.p. 66", b.p. 128-132" at 45 Tom (lit.,* m.p. 66", b.p. 110-113deg; at 36 Torr). Aryl dimethylphosphinates were prepared by the follow- ing general procedure (applied to the 4-nitrophenyl ester) : equimolar quantities of 4-nitrophenol, pyridine, and di- methylphosphinoyl chloride were stirred together overnight at ambient temperature using dichloromethane as solvent.The white precipitate of pyridinium chloride was removed and the filtrate dried (MgSO,). Evaporation of solvent gave a red mobile oil which on trituration with dry ether gave a solid. Purification was effected after three re-crystallisations from dichloromethane. Liquid esters were purified by distillation undcr reduced pressure. The 4-metlioxyphenyl ester gave a dark brown intractable oil which gave crystals after chromatography in dichloro-methane through a column of alumina (Cainag MFC; 100-200 mesh; Hopkins and Williams). All the esters readily decomposed as is reflected in the elemental analyses (carried out by Rlr. G. M. Powell of this laboratory with a Hewlett-Packard 185-CHN analyser).The solid species were recrystalliscd just before use, liquid esters were used directly after distillation. Physical properties of the esters are given in Table 1 and TABLE1 Physical properties of substratcs Dimethylphosphinate hlnm 3f.p. or b.p. ("C) 3,S-Dinitrophenyl 410 72-78 0 4-Nitrophen yl 400 97-101 0 3-Nitrophcny l 300 (340) 8 81-84 4-Bthoxycarbonylplienyl 300 210-220 (16) b 4-Chlorophen yl 300 (281) 182-184 (27) b 3-Chlorophcn yl 296 168-164 (17)4-Acctylphenyl 345 (300) e 219-228 (30) b Plien yl 290 (300) 164-166 (27) b 4-Me thylphcn yl 298 160-170 (16)4-Methoxyphenyl 300 60-71 With decomposition. * Pressure in Tom. c Wavelengthfor imidazolc buffers. d bsol;Yavclength for hydroxide buffers. analytical details in Supplementary Publication No.SUP 21624 (3 pp.).* Good lH n.m.r. spectra and integrations * For details of Supplementary Publications see Notice to .Authors No. 7 in J.C.S. Perkin 11, 1975, Index issue. Items less than I0 pp. arc supplied as full-sizc copies. J.C.S. Perkin I1 were obtained in all cases using deuterio-acetone or -chloro- form as solvent; they exhibited 1 : 1 doublets at T ca. 8 for PCH, groups (split by 31P) and complicated aromatic multiplets. The n.m.r. spectrum of the 4-nitrophenyl ester in acetone had an unresolved envelope at -54 p.p.m. relative to H,PO,, shifting to -57 p.p.m. on addition of water. 1.r. data are recorded in SUP 21624. 4-Nitrophenyl diphenylphosphinate and benzoate were from another study h7 and the dimethylphosphate ester was a gift from Dr.R. C. Woodcock. Buffer constituents were of analytical quality or were recrystallised or redistilled from reagent grade materials. Dioxan was purged of peroxides by passage of the AnalaR material through a column of alumina. Reagent grade imidazolc was dissolved in methanol, decolourised with charcoal, and recrystallised twice from dry benzene. Thionyl chloride, methyl bromidc, and thiophosphoryl chloride were commercial products and used without further purification. Double distilled water was used tliroughou t and a-chymotrypsin was purchased from Bochringer. Deuterium osidc was from Ryvaii Chemical Conlpany and 180-cnrichcd water from Prochcm . Methods.--Reactions were followed spectrophotometric- ally using Beckman-DBG or Unicam SP 800 instruments.A typical procedure for the non-enzymatic hydrolyses involved adding a portion (10--GO p1) of a stock solution of substrate (ca. 10) in dioxan to 2.6 ml of buffer equili- brated to the correct teinpcrature in a quartz cell in the thermostatted cell compartment of the instrument. The progress curve of the absorbancc versus time at a fixed wavelength was recorded on a Servoscribe potentiometric recorder. A similar procedure was cmployed for the enzyme reactions except that thc substrate was run with thc buffer to check background hydrolysis bcforc the portion of enzyme was added. Repetitive scanning of the spectrum during reaction for cach substrate in the appropriatc buffer was used to determine the best wavelength for kinetic study (Table 1).First-order rate constants were determined from logarithmic plots using infinity values of absorbance measured after at least six half lives. Mass spectral studies on the product 3-nitrophenol wcre carried out by Dr. R. B. Turner on an AEI-MS902 high resolution mass spectrograph. N.1ii.r. spectra were recorded with a Perkin-Elmer K10 machinc. RESULTS Repetitive scanning of the spectrum of the hydrolysing esters gave good isosbestic wavelengths indicative of clean 1 :1 stoicheiometry and these arc given in SUP 21624. In the case of the 4-nitrophenyl estcrs the quantity of phenol released is estimated using the absorption coefficient at 400 nm and confirmed the 1 :1 stoicheiometry.The rates for buffer and lyate catalysed reactions exhibited first-ordel- kinetics up to ca. 90 of the reaction. Hydroxide Ion Catalysis.-Thc pseudo-first-order ratc constants for release of 4-nitrophenol from the corresponding dimethylphosphinate ester were proportional to hydroxidc ion concentration up to 10 (see SUP 31624). The bi-molecular rate constants for hydroxidc attack at the othcr phenyl esters were obtained using 1O-3~r-NaOH(see Tablc 2). G. W.Parshall, OY~.Syntlz., 19136, 45, 102. K. A. Pollart and H. J. Harwood, J. Org. Chem., 1962, 27, 4446.' A. IYjlliaiiis and G. Salvadori, J. Clirm. SOC.(B),1971, 2401. Imidazole-cafalysed IIjdroZyses.-Plots of pseuclo-first-order rate constants for release of 4-nitrophenol from the corresponding dimethylphosphinate ester in imidazole buffers versus total buffer concentration are linear. The effect of concentration change at different pH values TABLE2 Alkaline hydrolysis of substituted phenyl dimethyl- phospliinates a Phenpl substituent a-b koH/l mol-l s-l 3,B-Dini tro 4-Nitro 1.42 1.24 18.8 16.0 Nitro 0.71 6.77 4-Acetyl4-Ethoxycarl)oiiyl4-Chloro 0.87 0.68 0.23 6.63 6.3 2.20 3-Ctloro 0.3'7 2.48 H 0.00 0.80 4-Methyl4-Methoxy -0.17 -0.27 0.61 0.73 Q 26", ionic strength made up to 0.1~with NaC1, 10dioxan, tolerance in the rate constants is amp;50,amp;.b From G. euro;3. Barlin and D. D. Perrin, Qicart. Reu., 1968, 20, 76. (slope/total buffer colicelitration vemts fraction of base species) indicates that catalysis is via the base species (Figure 1).Values of kimidszole (hi,) for other aryl esters were obtained from experiments using 0.9M-imidazole buffers (fr. base = 0.9) and the data are collected in Table 3. TABLE3 Imidazole-catalyscd hydrolysis of substituted phenyl dimethylphosphinates Phenyl substituents 104him/l mol-' s-' 4-Nitro 32.2 3-Nitro 6.72 4-Acetyl 6.78 4-Ethox ycarbonyl 4.09 4-ChdOro 0.947 3-Chloro 1.36 H 0.232 b69.8', ionic strength made up to 0.1~ with NaCI, 10 dioxan. *Arrhenius parameters AH,,t9.3 kcal mol'l and AS2amp;* -40 cal mol" K-1 may be calculated from values at 25 and 39.9' (4.94 x and 1.46 x 10-1 mol-' s-l respectively). C Tolerance in rate constants is f6.Other Buffer Sfiecies.-Other buffers (phosphate dianion, carbonate, Tris, pyridine, and collidine) catalyse the release of 4-nitrophenol from the dimethylphosphinate ester to a greater or lesser extent; the parameters (Table 4) are derived from concentration studies at a single pH assuming only the basic species to be effective. The effect of water as a. 'catalyst' is determined from intercepts of rate constant versus buffer concentration at constant pH and allowing for the small contribution of hydroxide catalysis (water concentration is taken as 55.5M to derive the bi- molecular constant). Analysis of phosphate catalysis at different pH values (as in Figure 1) indicates that the H,PO,-species has little catalytic power compared with the HPOd2-species.The intercept at fr: base = 0 has a value equal to its error: k=po,r-= 3.56 x 1 rno1-I s-l and k~,po,-0.68 x 1 rno1-I s-l (at 39.9', ionic strength = 0.1).* Deuterium Oxide Solvent Isotope Effect.-For imidazole catalysis of 4-nitrophenyl dimethylpliosphinate hydrolysis kamp;/kk = 1.7 (see SUP 21624). Site of Fzssiow.-When 3-nitrophenyl dimethylphos-phinate ester is hydrolysed in enriched water there was no incorporation of the enriched oxygen into the product 3-nitrophenol as compared with the appropriate control experiments (SUP 21624) indicating P-0 cleavage as the major reaction. This experiment eliminates reactions occurring via attack on nucleophiles on the aromatic nucleus. 0.5 1.0 Fraction base FIGURE1 Imidazole-catalysed hydrolysis of 4-nitrophenyldimethylphosphinate at different pH.Intercept on fraction base = 1.0 axis is 1.77 x lo-. Conditions: 39.9'; ionic strength 1.0~ In any case the reactivity is higher than would be expected for such a mechanism.* TABLE4 Bimolecular rate constants for catalytic hydrolysis of 4-nitrophenyl dimethylphosphinate a,b Catalyst pKaa k,,t/l niol-1 S-1 Hydroxide 16.7 16 Carbonate 10.33 0.143 Tris(hydroxymethy1)aminomethane 8.10 9100 Imidazole 6.96 4.94 x 10-4 Phosphate dianion 7.21 1.06 x 10-2 sym-Collidine 7.48 1.28 x 10-6 Pyridine 6.17 lo-* Water -1.7 1.6 x Phosphate monoanion 2.14 (0.68 x 10-8~ 25", ionic strength made up to 0.1~with NaCl, 107; dioxan, tolerance in the rate constants f6.6 13ransted law holds for the oxygen nucleophiles: loglokoat = 0.41pK8 -6.09 (correlation coefficient Y 0.99). c 39.9". d pK, values from W. P. Jencks and J. Regenstein in 'Handbookof Biochemistry ', ed. H. A. Sober, The Chemical Rubber Co., Cleveland, 1970, 2nd edn., section 5-187. SoZvoZysis in Dimethyl SuZphoxide.4-Nitrophenyl di-methylphosphinate spontaneously lost 4-nitrophenol in A. J. Kirby and W. P. Jencks, J.Amev. Chem. SOL, 1966'87, 3209. 518 DMSO (Koch-Light; puriss) at 26" with hob8 = 6.28 x s-l (hkinetic 436, hisosbamp;ic 300 nm). The corresponding diphenylpliosphinate ester was unreactivc even at 70". Exchange of a-Hydrogen Atoms.-The lability of the hydrogen atoms on the phosphinoyl methyl groups was shown to be negligible by measuring the lH n.m.r.spectrum of the 4-nitrophenyl ester in D,O in the presence and absence of NaOD. The aromatic envelope integrated with respect to the phosphinoyl methyl doublet gave the ratio 0.67 (expected for no exchange, 0.67). If an elimination-addition (EA)type reaction (2) occurs then at least 1/6 of the methyl protons would be exchanged and the minimum value for the ratio would be 0.8. The error limits on this technique allow us to exclude as a major path the EA mechanism which can therefore be omitted from the discussion. DISCUSSION NucZeo+hiZic Catalysis.-Although a high correlation coefficient is not expected for a wide variety of bases in a general base process" the high degree of dispersion exhibited by the Brransted type plot (Figure 2) is in- compatible with such a route and more representative of a nuclcophilic me~hanisrn.l,~,*~~ Where a continuity of OIi -I 0 5 10 115 PK, FIGURE 2 Brmstcd type Plot of reactivity of basic reagents with 4-nitrophenyl dimethylphosphinate ; line is theoretical (see Table 4) ; data from Table 4; arrows refer to upper and lower limits structural type exists, as in the attack of oxygen species (HO-, C032-, HP0,2-, H,O), a good correlation holds (see Table 4 and Figure 2).The rate constant for Tris 0 F. Covitz and F. H. Westheimer, J. Amer. Chewa. SOC.,1963, 85, 1173. 10 M. L. Bender, ' Mechanisms of Homogeneous Catalysis- J.C.S. Perkin I1 catalysis is greater than that of carbonate by ca.lo3 although the latter is the stronger base by ca. 2 pK, units. Imidazole and Tris are respectively some lo3-and 106-fold more efficient than collidine although the pK, values are similar. Such comparisons show either a change in mechanism from, presumably, general base catalysis for collidine to nucleophilic catalysis or constancy of mechanism with high steric demand in the transition state, a situation more marked in nucleophilic than in general base cataly~is.~~J*J~ The higher reactivity of collidine compared with pyridine, although the former is the more sterically hindered, implies that general base catalysis is the predominant mechanism for these species. Hydroxide attack exhibiting a IS-dependence (Table 2, Figure 3) is a unique observation because only a 0 3,5-(NO2$ I 0 0.5 1-0 c-FIGURE Hammett O-dependence of reactivity of hydroxide3 ion with substituted phenyl dimethylphosphinates ; values of O-from G.B. Barlin and D. D. Pemn, Quart.Rev., 1966, aO,76. Line is theoretical Pog,,,k = 0.930-+ 0.06 (Y = 0.984); data from Table 2 correlation is observed in all the recorded cases of hydroxide attack at PV.8,1s The data are consistent with considerable bond breaking (Ad-P) in the transition state of the rate-limiting step. Either a stepwise mechanism with the breakdown of an intermediate rate limiting or a synchronous process occurs. The absence of oxygen-18 incorporation into unchanged ester in the partial hydrolysis of Pv estersl6 indicates that the oxygens of the intermediate reaction (3)Jare not equiva- lerk-or that the process is synchronous reaction (4).Support for a non-s~chronous in PhoS-phinate hydrolysis has been claimed in a recent report by Haake et-d.17 who observed an induction period in the 13 L. W. Deady and J. A. Zoltiewiecz,J. Org. Chem.. 1972, 87, 603. 14 S.L.Johnson, Adv. Phys. Org. Chem., 1967,5, 237. 15 A. Williams and K. T. Douglas, J.C.S. Perkin 11, 1972,1464 and references therein. 16 (a)D. Samuel and B. L. Silver, Adv. Phys. Ovg. Chem., 1965, 8, 177; (b) P. C. Haake, C. E. Diebert, and R. S. Marmor,Tetrahedron Letters, 1968, 6247. Fier-H.A.Turley, and C.P.Cook,D.R.Haake,C.P.(a) 1' man, J. Amer. Chem. SOC.,1972, 94, 9260;Phosphorus, Elsevier, Amsterdam, 1967, pp.31 7-322. (b) R. D. Cook, C. E. H. J. Brass, J. 0. Edwards, and M. Biallas, J. Amer. Chem. Deibert, W. Schwarz, P. C. Turley, and P. C. Haake, ibid., 1973, Soc., 1970, 92, 4676. 95, 8088. From Protons to Proteins,' Wiley, New Yprk, 1971, pp. 176 et seq. 11 A. J. ?by and S. G. Warren, Organic Chemistry of 1' base hydrolysis of methyl di-isopropylphosphinate and a very high dependence of the rate constant for alkaline hydrolysis on a* (p* ca.8). 0-bsol;bsol;0 H6 + ,P -0Ar RV R k3C k-3 I kcHO-,F'-OAr-products I' bsol;R R (3) (Y) 0 HO---,y---OAr 8-ll 8, R 'I (4) Species (V; R = CH,) corresponds to the most efficient mode of attack for the esters because (a) nucleophilic attack directly produces the permutational isomer of minimal energy; any isomerisation will be disfavoured since it must produce a higher energy isomer ; (b) the isomer with the aryloxy group apical preparatory to departure (asdemanded by microscopic reversibility) is directly formed; (c) the face attacked might be expected to provide the least stereoelectronic repulsion to the entering ligand.A consequence of this route of attack is that inversion should occur at phosphorus and this is the case.18 Attack at phosphorus in cyclic phosphinates in the same manner as in acyclic ones would yield a ' diequatorial ' ring (VI) and an equatorial-axial ring (VII) is preferred ; permutational isomerisation of this species to put the leaving group axial leads to retention of configuration.OH G-O-V A OH (Vr1 cvn) The intermediate (V) would be unlikely to give rise to a 0-relationship (breakdown rate limiting) because hydroxide ion is presumably a much poorer leaving group than aryl oxide. Thus the evidence at present points to a synchronous process for the aryl dimethylphosphinates. Attack of the heavily solvated hydroxide ion would have stringent steric requirements and this is presumably the reason for the large difference in character between the reaction of diphenyl- and dimethyl-phosphinates with this ion. The a-relationship (Figure 4) for imidazole-catalysed hydrolysis of dimethylphosphinates is consistent with either a stepwise or a synchronous mechanism analogous to (3) and (4) respectively involving nucleophilic attack by imidazole at Pv.The deuterium oxide solvent (a)M. Green and R. F. Hudson, J. Ckem. Soc., 1963, 3883; (b) M. J. Gallagher and I. D. Jenkins, Topics Stereochem., 1968, 8, 31, 82. AT. L. Bender and F. C. Wedler, J. Amcv. Chem. SOC., 1972, 94,2101. 2o G. Aksnes, Acfa Chem. Scand., 1960,14, 1476, 1626. a1 (a) L. Ginjaar and S. Blasse-Vel, Rec. Tvau. chim., 1966, 8!5, 694; (6) ref. 11, p. 302. isotope effect of 1.7 on the 4-nitrophenyl ester reaction is rather high for nucleophilic attack; Johnson l4 suggests a value 2 for general base catalysis and 0.8-1.9 for nucleophilic routes. There is a precedent for a high value of KH/KD (ca. 1.8) in the imidazole catalysed hydrolysis of 4-nitrophenyl diethy1pho~phate.l~ It was suggested that -5c 0 0.5 1.0 c-FIGURE4 Hammett O-dependence of reactivity of imidazole with substituted phenyl dimethylphosphinates ; (I-values as for Figure 3; line is theoretical gog,,k = 1.600--4.49 (v = 0.986) ; data from Table 3 this might be due to concerted proton transfer (VIII) but this cannot be so for the dimethylphosphinates because the rate law implied by (VIII) involves H+ and imidazole and there is no conjugate acid term (Figure 1).A similar scheme (IX) involving considerable water participation could give the higher ratio and this is supported by the known high tendency of the phos- phinoyl oxygen atom to act as a hydrogen-bond acceptor.20 Such an ordered transition state could explain the very high negative entropy of activation for imidazole catalysis (AS$ = -40 cal mol-l K-l).In general, bimolecular nucleophilic processes on PV have AS between -10 and -30 cal mol-l K-1.21 Although N-dimethylphosphinoylimidazole was not directly observed, its probable reactivity, by comparison with the known activity of phosphorylimidazoles to solvolysis 22 and the higher inherent reactivity of 2a (a) J. Baddiley, J. G. Buchanan, and R. Letters, J. Chm. SOC.,1966, 2812; (b) T. Wagner-Jauregg and B. H. Hackley, J. Amev. Chem. SOC., 1963, 75, 2126; (c) F. Cramer, H. Schaller, and H. A. Staab, Chem. Bey., 1961,94, 1612; (d)L. N. Nikolenko and E. V. Degterev, Zhuv. obshchei Khim., 1967, 87, 1360; (e) B.Atkinson and A. L. Green, Trans. Faraduy SOC., 1967, 68, 1334; (fl R. L. Blakeley, F. Kerst, and F. H. Westheimer, J. Amev. Chem. SOC., 1966, 88, 112; (g) B. S. Cooperman and G. L. Lloyd, ibid., 1971, 93, 4883, 4889; (h) W. P. Jencks and M. Gilchrist, ibid. 1966, 87, 3199. phosphinates compared with phosphate^,^^ is consistent with its being an intermediate. The pyrophosphate (X), would have a high solvolysis rate 2,24 compatible with its existence as an intermediate in phosphate catalysis. For general base processes 2-+/O,P-O-PO( CH3)Z (CH31, PO-O-Sbsol; imidazole and phosphate dianion should have a similar activity but as a nucleophile the former is often 103-fold J.C.S. Perkin I1 mean that the steric effect will be unimportant in enzyme mechanisms since the effective co-ordination number of a bound acyl group may be raised increasing its stereo- selectivity.For example thc imidazolyl of histidine-57 acts as a general base in the ageing of or-chymotrypsiri inhibited by bis4-nitrophenyl carbonate l9 whereas imidazole acts as a nucleophile in model ~ystems.~g,~~ The only direct evidence presently available to distinguish general base from nucleophilic catalysis in acylation of chymotrypsin is thc deuterium oxide solvent isotope effect which is (2 for a number of ac~lations.~~ The observation of an isotope effect (1.6) for acylation by Pnitrophenyl diphenylphosphinate which is normally TABLEli Comparison of 4-nitrophenyl phosphinate and phosphate hydrolyscs Substrate A011 ki, kap0.t-llipliciiylphospliinatc f 7.3 1.03 x 10-3 2.05 x 10-4 (4.88 x 10-3e)J imcthylphosphinatc 16 4.94 x 10-4 1.06 x Dimethylphosphatc 8.8 x 3.22 x load 3.47 x 10-0 Unless otherwise statcd thc teinpcraturc is 25"; units in 1 mol-l s-'.Units in s-1. 66".C Vel, Rec. Tvuv. chiiit., 1968, 77, 966. f Valucs from ref. 3. lCR,O 8.88 x 1.16 x d 50". 6 L. Ginjaar and S. Blassc-more reactive than the latter.14 Imidazole and phos- phate catalysed hydrolysis of 4-nitrophenyl diphenyl- phosphinate hydrolyses have similar reactivities in accord with a general base pathway.3 The high reactiv- ity of phosphate compared with imidazole for hydrolysis of the dimethylphosphinate (Table 4) is probably not due to bifunctional action since phosphate lies on the same Brgnsted line as hydroxide and water.The hydrolysis of 4nitrophenyl dimethylphosphate is catalysed to an equal extent by phosphate and imidazole (Table 5) but the deuterium oxide solvent isotope effect (1.4) indicates nucleophilic attack. These observations cast doubt on the use, alone, of the imidazole :phosphate reactivity ratio as a simple mechanistic tool. Steric Hindrance and Catalytic Mechanism.-This work provides direct evidence that lowering steric demand of the substrate leads to a transition from general base to nucleophilic catalysis. The results for methyl bis-(4- nitrophenyl) phosphate also argue this view4 and the higher steric requirements of the diphenyl- compared with the dimethyl-phosphinate are confirmed by solvolysis of the latter ester in moist DMSO presumably via a mechan-ism similar to that proposed by Ratz and others 26 for phosphodichloridate reactions via (XI).Steric inhibition of nucleophilic catalysis at the four- co-ordinate phosphyl centre does not have an analogue in carboxy chemistry due to the lower steric demands of the three-co-ordinate carboxy carbon. For example, even 4-nitrophenyl pivalate hydrolysis is catalysed by imidazole via a nucleophilic pathway. This does not O3 R. F. Hudson and G. E. Moss, J. Ciiern. SOC.,1964, 1040. a4 (a) T. C. Bruice and S. J. Benkovic, Bio-ovg. Chem., 1966, 2, 168; (b) D. M. Brown and N. K. Hamer, J. Chem. Soc., 1960, 1166; (c) D. S. Samuel and B. Silver, ibid., 1961, 4321.26 (a) R. Ramp;tz and 0. J. Sweeting, J. Org. Chem., 1963, 28, 1612; (b) M. A. Ruveda, E. N. Zerba, and E. M. de Moutier Aldao, Tctvahcdvon, 1972, 28, 5011. constrained to liydrolyse via a general base mechanisni indicates that the isotope data are not as conclusive as was previously thought .27 AL 6 7 a 9 10 pH (0) FIGURE6 pH(D)-Dependence of h,/km for 4-nitroplicnj-ldiphenylphosphinate and a-chymotrypsin. Conditions as in Table 6; x ,deuterium oxide solvcnt; 0,water solvent Binding of Substrates to Chymotry@sin.-The bell-shaped pH profiles (ko/Km)for the non-specific substrates 28 (a) J. R. Corfield, N. J. De'Ath, and S. Trippett, Chew. Comrtt., 1970, 1502; (b) S. E. Cremer and B. C. Trivedi, J. Amcr. Chem.Soc., 1969, 91, 7200; (c) I. F. Hudson and C. Brown, Accounts Cltem. Res., 1972, 5, 204. 27 C. D. Hubbard and J. F. Kirsch, Biochemistry, 1972, 11, 2483. 1976 (Table 6 and Figure 5) agree with those for other non-specific substrates. It is proposed that the reactivity of TABLE6 Limiting values for kinetic parameters of substrates of a-chymotrypsin ko-lKrn/Substrate lmol-I s-l pK, pK, 4-Nitrophenyl diphenylphosphinate 2 800 (H,O 6.7 9.6 PNitrophenyl diphenylphosphinate 1 800 (DO 6.8 9.6 4-Nitrophenyl dimethylphosphinate 91 6.8 4-Nitrophenyl bcnzoate 603 7.1 8.g6 026'. ionic strength made up to 0.1~with NaCl, 10 acetonitrilc. bThe tolerance in this value is rather high(f0.6) because only one point was measured in the high pH region.As we are only interestcd in k,,/Km(llrn) this is of little consequence. acylation of chymotrypsin involves a considerable binding component; presumably the aromatic functions bind in the ' tosyl-hole ' which accepts aromatic side chains and dioxan.2B The greater number of binding possibilities for the diphenyl as opposed to the dimethyl 2.9 (a)M.L. Bender, G. E. Clement, F. J. Kezdy, and H. d'A. Heck, J. Amev. Chenz. SOC.,1964, 86, 3688; (b) M.L. Bender and F.J. Kezdy, Biochemistvy, 1962, 1,1097. 29 (a)T. A. Steitz, R. Henderson, and I). 11.Blow, J. Mol. Biol., 1969, 40,337; (b) Ii. P. Bell, J. E. Critchlow, and M. I. Page, J.C.S. Pevkin 11,1974, 66. substrate (three instead of one aromatic group) is reflected by the higher reactivity of the former substrate in contrast to the normal order (Table 5).Model building studies with 4-nitrophenyl acetate and a model of chymotrypsin built with Kendrew models (Cambridge Repetition Engineers) using crystallographic co-ordinates, supplied by Dr. D. $1. Blow before public- ation?() indicate that if the 4-nitrophenyl group is placed in the ' tosyl-hole ' the nucleophilic participation of histidine-57 is impossible. The situation is not so simple however because 4-nitrophenyl pivalate and acetate acylate elastase, where the ' tosyl-hole ' is blocked for aromatic binding by a side chain of valine-216 and of threonine-226,S1 react at approximately the same rate as for chyinotrypsin.s2 K. T. D. thanks the Government of Northern Ireland for a Scholarship. 6/1571 Received, 8th August, 1976) 3O J. J. Birktoft, B. W. Matthew, and D. 31. Blow, Biochem. Biophys. Res. Comm., 1969, 36,1. 81 B. S. Hartley and D. M. Shotton, ' The Enzymes,' eds. P. 1).Boyer, euro;3. A. Lardy, and K. Myrback, Academic Press, New York, 1971, 3rd edn., p. 323. 32 (a) M. L. Bender and T. H. Marshall, J. Anzer. Chem. SOC., 1968,90,201; (b) M. L. Bender and K. Nakamura, ibid., 1962,84, 2677.