7
843 Biochem. J. (1980) 187, 843-849 Printed in Great Britain Mechanistic Studies of Carboxypeptidase Y from Saccharomyces cerevisiae pH AND pD PROFILES AND INACTIVATION AT LOW pH (pD) VALUES Wanda T. CHANG and Kenneth T. DOUGLAS* Department of Chemistry, Duquesne University, Pittsburgh, PA 15219, U.SX. (Received I November 1979) Steady-state kinetics of carboxypeptidase Y, a proteinase from yeast, were studied by using the reaction of 4-nitrophenyl trimethylacetate as a probe. The pH profile of kc.t is sigmoidal in H20-based buffers for the carboxypeptidase Y-catalysed hydrolysis of this ester (kc.t referring to the rate of deacylation of trimethylacetyl-carboxypeptidase Y). The corresponding pD profile in 2H20 is doubly sigmoidal, with inflexions at pD - 3.8 and -6.8. The ionization of pKD P. 3.8 is caused by a rapid inactivation in 2H20 media by a process that is only slowly reversed on transfer to pH 7.00 phosphate buffer in H20. The corresponding inactivation in H20-based buffers of low pH is considerably slower (-30-fold), follows a first-order rate-dependence and is very strongly pH- dependent, indicating some form of co-operative change in enzyme tertiary structure. The yeast proteinase carboxypeptidase Y (EC 3.4.16.1) is a glycoenzyme, isolatable in distinct molecular forms, differing in their catalytic properties and in the amounts of attached sugars (Margolis et al., 1978). It exhibits apparent similarities to both the serine and cysteine proteinases, on the one hand, and to the pancreatic metallo-carboxypeptidases, on the other. Carboxypeptidase Y is a relatively non-specific single-chain exopeptidase with no metal- ion-dependence (Hayashi et al., 1973, 1975) and no intrinsic endopeptidase activity (Hermodsen et al., 1972; Hayashi et al., 1973; Lee & Riordan, 1978), capable of cleaving even proline residues from proteins. [ts- high thermal stability, apparent resistance to pH extrema and activity in 6M-urea or in sodium dodecyl sulphate media have led to its increasing use in protein sequencing and structural studies (Hayashi, 1977; Liberatore et al., 1976; Martin et al., 1977). With aryl trimethylacetates, 'burst kinetics' ([S]0 > [El0) have been observed and studied in detail by using stopped-flow techniques (Douglas et al., 1976). The results were quantitatively consistent with formation of a physical Michaelis complex ES (dissociation constant K,) followed by a covalent change (rate constant k,) leading to an acyl-enzyme, ES', and the phenolate component of the products. Deacylation (rate constant k3) of trimethylacetyl- * Present address: Department of Chemistry, Uni- versity of Essex, Colchester C04 3SQ, U.K. Vol. 187 carboxypeptidase Y then leads to free enzyme and trimethylacetate ion (eqn. 1): E+S ES k ES'+P1 - 3-> E+P2 K,2 (1) Although it was recognized that biphasic kinetics of the type observed merely indicate a change in rate-determining step, there is some additional support for the acyl-enzyme mechanism. Thus values of kc.t for a series of aryl-substituted trimethylacetates with carboxypeptidase Y are inde- pendent of the leaving group (Douglas et al., 1976), in contrast with the rate constants for attack of OH- ion and imidazole on these esters (Douglas et al., 1977). Under single-turnover conditions ([ElO > [Slo), by using a stopped-flow technique with 4-nitrophenyl cinnamate as substrate, the 4-nitro- phenoxide ion product is released previously to, and faster than, the cinnamate ion, as predicted from eqn. (1) (Douglas et al., 1975). In addition, an intermediate (as yet unisolated) was detected kinetic- ally in the enzymic hydrolysis of 4-methoxyphenyl cinnamate, presumably cinnamoyl-carboxypeptid- ase Y (Douglas et al.,- 1975). Although such observations are reminiscent of a-chymotrypsin, deacylation of trimethylacetyl-a- chymotrypsin depends, crucially, on the state of ionization of a functional group of pK,,,P = 6.7 (Bender et al., 1962). Deacylation of trimethyl- 0306-3275/80/060843-07$01.50/1 X) 1980 The Biochemical Society

Mechanistic Studies of Carboxypeptidase Y from Saccharomyces

  • Upload
    trantu

  • View
    216

  • Download
    0

Embed Size (px)

Citation preview

843Biochem. J. (1980) 187, 843-849Printed in Great Britain

Mechanistic Studies of Carboxypeptidase Y from Saccharomyces cerevisiae

pH AND pD PROFILES AND INACTIVATION AT LOW pH (pD) VALUES

Wanda T. CHANG and Kenneth T. DOUGLAS*Department ofChemistry, Duquesne University, Pittsburgh, PA 15219, U.SX.

(Received I November 1979)

Steady-state kinetics of carboxypeptidase Y, a proteinase from yeast, were studied byusing the reaction of 4-nitrophenyl trimethylacetate as a probe. The pH profile of kc.t issigmoidal in H20-based buffers for the carboxypeptidase Y-catalysed hydrolysis of thisester (kc.t referring to the rate of deacylation of trimethylacetyl-carboxypeptidase Y).The corresponding pD profile in 2H20 is doubly sigmoidal, with inflexions at pD - 3.8and -6.8. The ionization of pKD P. 3.8 is caused by a rapid inactivation in 2H20media by a process that is only slowly reversed on transfer to pH 7.00 phosphate bufferin H20. The corresponding inactivation in H20-based buffers of low pH is considerablyslower (-30-fold), follows a first-order rate-dependence and is very strongly pH-dependent, indicating some form of co-operative change in enzyme tertiary structure.

The yeast proteinase carboxypeptidase Y (EC3.4.16.1) is a glycoenzyme, isolatable in distinctmolecular forms, differing in their catalytic propertiesand in the amounts of attached sugars (Margolis etal., 1978). It exhibits apparent similarities to boththe serine and cysteine proteinases, on the one hand,and to the pancreatic metallo-carboxypeptidases, onthe other. Carboxypeptidase Y is a relativelynon-specific single-chain exopeptidase with no metal-ion-dependence (Hayashi et al., 1973, 1975) andno intrinsic endopeptidase activity (Hermodsenet al., 1972; Hayashi et al., 1973; Lee & Riordan,1978), capable of cleaving even proline residuesfrom proteins. [ts- high thermal stability, apparentresistance to pH extrema and activity in 6M-ureaor in sodium dodecyl sulphate media have led to itsincreasing use in protein sequencing and structuralstudies (Hayashi, 1977; Liberatore et al., 1976;Martin et al., 1977).

With aryl trimethylacetates, 'burst kinetics'([S]0> [El0) have been observed and studied in detailby using stopped-flow techniques (Douglas et al.,1976). The results were quantitatively consistentwith formation of a physical Michaelis complex ES(dissociation constant K,) followed by a covalentchange (rate constant k,) leading to an acyl-enzyme,ES', and the phenolate component of the products.Deacylation (rate constant k3) of trimethylacetyl-

* Present address: Department of Chemistry, Uni-versity of Essex, Colchester C04 3SQ, U.K.

Vol. 187

carboxypeptidase Y then leads to free enzyme andtrimethylacetate ion (eqn. 1):

E+S ES k ES'+P1 - 3-> E+P2K,2

(1)

Although it was recognized that biphasic kineticsof the type observed merely indicate a change inrate-determining step, there is some additionalsupport for the acyl-enzyme mechanism. Thusvalues of kc.t for a series of aryl-substitutedtrimethylacetates with carboxypeptidase Y are inde-pendent of the leaving group (Douglas et al., 1976),in contrast with the rate constants for attack ofOH- ion and imidazole on these esters (Douglaset al., 1977). Under single-turnover conditions([ElO > [Slo), by using a stopped-flow technique with4-nitrophenyl cinnamate as substrate, the 4-nitro-phenoxide ion product is released previously to, andfaster than, the cinnamate ion, as predicted fromeqn. (1) (Douglas et al., 1975). In addition, anintermediate (as yet unisolated) was detected kinetic-ally in the enzymic hydrolysis of 4-methoxyphenylcinnamate, presumably cinnamoyl-carboxypeptid-ase Y (Douglas et al.,- 1975).

Although such observations are reminiscent ofa-chymotrypsin, deacylation of trimethylacetyl-a-chymotrypsin depends, crucially, on the state ofionization of a functional group of pK,,,P = 6.7(Bender et al., 1962). Deacylation of trimethyl-

0306-3275/80/060843-07$01.50/1 X) 1980 The Biochemical Society

W. T. CHANG AND K. T. DOUGLAS

acetyl-carboxypeptidase Y, in contrast, shows anon-crucial sigmoidal pH-dependence, withpKapp -5.1. The rate constant at low pH is 41% ofthe maximal rate constant in alkaline media(Douglas et al., 1976). The group of pKapp. 5.1may be a carboxy group or a perturbed histidineresidue.

Other studies directed towards detection ofacyl-enzyme intermediacy have been reported (Baiet al., 1975). For example, in the carboxypeptidaseY-catalysed cleavages of N-acetyl-L-tryptophanderivatives it was stated that values of kcat werenearly constant, indicating that k3 was rate-deter-mining (Bai et al., 1975). In addition, studies of thesolvent deuterium isotope effect have been reported(Bai et al., 1975). These were performed at a singlepH (pD) and for derivatives whose Km and kcatvalues may well be complex, as eqn. (1) yieldskcat = k2k3/(k2 + k3) and Km = K8k3/(k2 + k3) onkinetic analysis. In addition, the inhibition patternexhibited by L-phenylalanine apparently changes ongoing from H20 to 2H20 media in studies ofN-acetyl-L-phenylalanyl ethyl ester.

Consequently, we have studied the solventdeuterium kinetic isotope effect [over a wide pH(pD) range] on the carboxypeptidase Y-catalysedhydrolysis of 4-nitrophenyl trimethylacetate, forwhich detailed knowledge of the microscopic rateconstants in H20 had already been obtained.

Materials

Compressed baker's yeast was obtained fromAnheuser-Busch (Pittsburgh, PA, U.S.A.), DEAE-cellulose DE-52 from Whatman (Maidstone, Kent,U.K) and Sephadex A-50 from Pharmacia FineChemicals (Uppsala, Sweden). N-Acetyl-L-tyrosineethyl ester (Ac-Tyr-OEt) and N-benzyloxycar-bonyl-L-glutamyl-L-tyrosine (Cbz-Glu-Tyr) werefrom Sigma Chemical Co. (St. Louis, MO, U.S.A.).4-Nitrophenyl trimethylacetate, from a previousstudy, was recrystallized from benzene. Acetonitrile(Pesticide grade; Fisher Chemical Co., Silver Spring,MD, U.S.A.) was further purified by molecular-sieve(Linde 4A) treatment, followed by distillation from asmall quantity of P205 and finally distillation offCaH2. 2H20 (99.8 atom% 2H; Aldrich ChemicalCo., Milwaukee, WI, U.S.A.) was either used asreceived or redistilled off KMnO4 for further purifi-cation. Buffer solutions were prepared from reagent-grade materials.

Carboxypeptidase Y was purified by the pro-cedure of Hayashi et al. (1973), with slightmodifications. Sodium dodecyl sulphate/polyacryl-amide-gel rod electrophoresis was performed with aPharmacia GE-II apparatus, by the method ofWeber & Osborn (1969). Protein samples (approx.1 mg/ml) were boiled with 0.1% sodium dodecyl

sulphate/1% dithiothreitol (Sigma) for 10min on awater bath; less drastic treatment does not com-pletely unfold carboxypeptidase Y, leading to mul-tiple bands on sodium dodecyl sulphate/polyacryl-amide-gel electrophoresis. Reservoir buffer was0.1M-potassium phosphate buffer, pH7.3, contain-ing 0.5% sodium dodecyl sulphate. Electrophoresiswas performed for 4h at constant current (1.8mA/rod).

Methods

Absorption coefficients on H20 and 2H20Around neutrality the absorption change of the

hydrolysis reaction of 4-nitrophenyl trimethyl-acetate is very pH-sensitive, as the PKa of 4-nitrophenol is approx. 7. Consequently, for anygiven pH or pD, the absorption change wasmeasured. Complete hydrolysis of substrate in10mM-NaOH (30min) was performed. The absorb-ances of samples of this hydrolysate in the appro-priate buffer medium allowed calculation of theabsorption change. Above pH (or pD) -6.4, kineticsof 4-nitrophenyl ester hydrolysis were studied at400nm, and below pH 6.4 at 340 nm.

Enzyme concentrations and assaysApproximate enzyme concentrations were

measured by using A"8 = 15 and a molecularweight of 61000 (Aibara et al., 1971; Margoliset al., 1978). Accurate functional enzyme concen-trations were determined by active-site titration asdescribed elsewhere (Margolis et al., 1978).

Titrimetric measurement of the Ac-Tyr-OEt ac-tivity of carboxypeptidase Y was performed at 25 0Cat pH 8.00 (10mM-Ac-Tyr-OEt in 0.1 M-KCI) as astandard assay. One unit of Ac-Tyr-OEt activity isl,pmol of Ac-Tyr-OEt hydrolysed/min.The peptidase activity could be studied either by

the colorimetric ninhydrin method of Hayashi et al.(1973) (for the determination of liberated tyro-sine) or, quantitatively, by the decrease in absorb-ance at 224nm (Kuhn et al., 1974), determined to bethe best wavelength by repetitive spectral scanningof a reacting mixture.

In the ninhydrin method, enzyme solution (0.1 ml)and substrate (0.1 ml of 3 mM-Cbz-Glu-Tyr) wereadded to 0.8 ml of 50mM-sodium acetate buffer,pH5.5, in a test tube. After incubation at 250C for15 min, reaction was terminated by addition ofninhydrin reagent and heating in a boiling-waterbath. Blanks with substrate and enzyme absent inturn were performed.

Kinetic proceduresSpectrophotometric studies were performed with a

GCA-McPherson 707-K double-beam spectro-photometer, whose cuvette chamber was thermo-

1980

844

pH (pD) PROFILES OF CARBOXYPEPTIDASE Y ACTIVITY

statically controlled at 25.00 + 0.020 C by means ofa Haake E52 thermoregulator pump.The Radiometer recording pH-stat and titration

system used consisted of a PHM 64 pH-meter(0.001 pH readability) with a TTT60 Titrator andREC 61 chart recorder. The glass reaction vesselwas kept at 25.00 + 0.020C by means of a Thermo-mix model 1420. A type ABU 12 autoburette(0.25 ml capacity) delivered titrant (50 mM-NaOH).All pH-stat reactions were carried out under astream of scrubbed C02-free dry N2. The pH-meterwas standardized against Fisher standard buffers;titrant concentration was checked against standardacid.

Generally, pH values measured before and afterreaction differed by <0.04 pH unit; post-reactionvalues were used in calculations. As the maximumamount of acetonitrile was low (never greater than4%, v/v) it was assumed (Kezdy & Bender, 1962)that this had a negligible effect on the electrodereading.

Studies with 2H20The apparent pH values of 1.2mM-HCl in H20

and 1.2mM-2HCl in 2H20 were measured, and therelationship pD =pHmeas. + 0.353 was establishedfor the pH-meter used throughout.

For studies of inactivation in 2H20, carboxy-peptidase Y was incubated in buffers of various pDvalues at ice-bath temperature for 15min, in thepresence and absence of substrate. The (remaining)activity of inc-ubated enzyme was measured in anappropriate buffer.

lected in Table 1 and shown in Fig. 2. ThepD-dependence of kcat is doubly sigmoidal, withpKD values of -3.8 and -6.8 and a limiting valueof Pat at high pD of 8.15 x 10-3s- . The con-cat. Do .5x1- -.Tecn

3 I l

rs n fi

1,0 2 - 0I.-, ~ 0 0

x 00

3 4 5 6 7 8 9 10pH

Fig. 1. Steady-state pH profile (k3) for the carboxy-peptidase Y-catalysed hydrolysis of 4-nitrophenyl tri-

methylacetate at 25.00 C at I 0.1Each value of k3 was determined from the zero-orderslopes at two different enzyme concentrations. Allmeasurements were made with a single enzymepreparation from either Anheuser-Busch yeast (0)or Fleischmann yeast (OT). Points are experimental;the continuous line is theoretical, obtained by usingthe Henderson-Hasselbach equation with pKapp.=5.35 and limiting values of k3 in acid of 1.00 x10-2s-I and in base 2.50x 10-2s-1. Data for theFleischmann-yeast enzyme were taken from Douglaset al. (1966). See the text for experimental details.

Results

pH and pD profiles of kcat for 4-nitrophenyltrimethylacetateBy using the procedure previously detailed

(Douglas et al., 1976), values of kcat. (=k3) for'Anheuser-Busch carboxypeptidase Y' were ob-tained for the pH range 3-8. Each point repre-sented, in effect, an active-site titration and kcatwas taken as the zero-order slope rate (steady state)divided by the initial burst (proportional to thefunctioning enzyme concentration). The data arepresented in Fig. 1, along with the values of k3 froma previously published study (Douglas et al., 1976)for 'Fleischmann carboxypeptidase Y', i.e. theenzyme prepared in the same manner, but fromFleischmann's baker's yeast. There is no significantdifference in deacylation rate or its sensitivitytowards pH changes for these preparations, inagreement with the previous study.A similar experiment, but with enzyme from

Anheuser-Busch yeast only, was carried out todetermine the pD-dependence of kc.t in 2H20 for4-nitrophenyl trimethylacetate. The results are col-

Table 1. Steady-state pD-profile data for carboxy-peptidase Y-catalysed hydrolysis of 4-nitrophenyl

trimethylacetateExperimental details are given in the text. kcat.values were obtained from the zero-order (steady-state) slope at two different enzyme concentrations.pD values were from the relationship pD =pHmeas. + 0.3 5.

pD8.658.397.757.697.086.816.445.525.084.794.544.253.853.673.623.182.93

103 x kcat. (s-5)8.838.678.808.428.236.505.895.505.325.144.733.902.732.141.850.47

Not detected

Vol. 187

845

W. T. CHANG AND K. T. DOUGLAS

-e

xIn0

8

6

4

2

I I I I I I

,-0

/" X

3 4 5 6 7 8pD

I--

4..

6

Fig. 2. Steady-state pD profile (kcat) for the carboxy-peptidase Y-catalysed hydrolysis of 4-nitrophenyl tri-

methylacetate at 25.00C at I 0.1Points are experimental (data from Table 1). Thecontinuous line is theoretical for the independentsuccessive ionizations of acids with pKafp values of3.82 and 6.85 and k, = 5.4 x 1O-3 s- and k2 =8.75 x 10-3s-1, as described by eqn. (1). The brokenline is for a single ionization of pKapp = 5.2 with k,,mat high pH of 8.75 x 1O-' s-. Experimental detailsare given in the text.

100 II

80 0

60

40 (pK.°)=4.1

20-

0

2 3 4 5 6 7pD

Fig. 3. Activity regained by carboxypeptidase Y ontransfer to pH7.00 H20-based phosphate buffer after

incubationfor 15min in buffers of variouspD valuesThe rate assay used was the kc.t. value in the4-nitrophenyl trimethylacetate reaction. Points areexperimental; the continuous line is theoretical foran acid of pKapp =4.1. Experimental details aregiven in the text. For pD values above 4 substratewas present during the incubation, whereas atpD < 4 it was absent.

tinuous line in Fig. 2 was calculated for the kin(scheme of eqn. (2):

K,EH2 EH-

|k,

K l k2

Ik2

etic

(2)

with pK, = 3.82, pK2= 6.85, k, = 5.4 x 10-3 -1 andk2= 8.75 x 10-3s-. For comparison, the brokenline shows the pH profile expected if only a singleionization (pKpp. = 5.2) were involved, with klim athigh pH of 8.75 x 10-3s-. The fit to this latter caseis so obviously poor that such a scheme can bedisregarded.

Inactivation ofcarboxypeptidase Y at lowpDThe doubly sigmoidal pD profile (Fig. 2) is

surprising in view of the singly sigmoidal pH profile(Fig. 1) and the short time required to obtain kcatvalues (usually only 8-10s after addition of car-boxypeptidase Y to the 2H20 media before chartrecording was commenced). Consequently carboxy-peptidase Y was incubated in an ice bath for 15 minin buffers of various pD values, samples wereremoved and k¢,t. values determined in the (burstkinetics) reaction with 4-nitrophenyl trimethyl-acetate in pH 7.00 H20-based phosphate buffer at250C. The results are shown in Fig. 3 and indicatethat on transfer back to pH 7.00 in H20-based bufferthe level of activity regained follows an ionization ofpKD - 4.1, which corresponds well to the lower-pD

Table 2. Reversibility of pD3 deactivation of carboxy-peptidase Y on transfer to pH17.00 H20-based

phosphate buffer (10.1) at 00CExperimental details are given in the text. Thecontrol enzyme had not been exposed to 2H20media. The treated carboxypeptidase Y wasincubated at pD3.0 for 1Omin before transfer to thepH 7.00 H20-based phosphate medium.

Time at 0°Cat pH 7.00 (min)

20406080100120140200

Activity (%)

Control pD3-treated

105

95

100

75

0

0

0

1019.5

limb of kcat in 2H20 (Fig. 2). Further experimentsshowed the inactivation at low pD (pD 3.0) to be veryfast (requiring only 12s to reach a stable activityvalue of 17%) and slowly (Table 2) reversible ontransfer to pH 7.00 H20-based phosphate buffer.

Bursts, comparable with those at higher pH, werestill obtained at higher pD values, indicating k2> k3in slightly alkaline 2H20 media.

1980

846

pH (pD) PROFILES OF CARBOXYPEPTIDASE Y ACTIVITY

Inactivation ofcarboxypeptidase Y at lowpHA corresponding inactivation occurs at low pH,

but considerably more slowly, the activity de-creasing with time until it reaches a steady value,which we have called (kc.t)0H. A typical example, atpH 2.86, is shown in Fig. 4. The rates of in-activation in 0.05 M- and 0.10 M-sodium formatebuffer are essentially the same, indicating that theinactivation is not a buffer effect. Inactivation can beanalysed as a first-order kinetic process (see Fig. 4for the fit to the first-order rate equation). BelowpH 2.86 the rate of inactivation became high, andabove pH 3.7 inactivation was extremely slow. Ratesof inactivation were determined in the intermediatepH range, rate constants for inactivation beingobtained from the best fit by non-linear regressionanalysis to a first-order rate equation. Data arecollected in Table 3. The rate of inactivation is

2.4

2.0

_ 1.6

x

° 1.2

0.81

0.41

0 200 400

Time (min)

Fig. 4. Time course of the inactivation of carboxy-peptidase Y in H20-basedformate buffers atpH2.86Values of (kC8t )' were obtained by assaying samplesin the 4-nitrophenyl trimethylacetate reaction inpH 7.00 H20-based phosphate buffer. Points areexperimental (0, 0.1 M-formate buffer; *, 50mM-formate buffer). The continuous line istheoretical for a first-order reaction with a rateconstant (kinact.) of 1.98 x 10-2min-' and a limitingvalue of (kt )'H of 0.422 x 10-1 s-1. See the text forexperimental details.

Vol. 187

Table 3. Rates of inactivation of carboxypeptidase Ytowards 4-nitrophenyl trimethylacetate in H20-based

buffers at 00CValues of klnact. and (kcat.)o, were obtained by least-squares regression analysis of the data to fit afirst-order rate equation. The control (pH 7.00)retained activity at 2.45 x 10-2S-1 (kcat. for 4-nitrophenyl trimethylacetate) throughout theexperiments. ki,nact. values were determined byincubation of a sample of carboxypeptidase Yin the appropriate buffer in an ice bath. Sampleswere withdrawn at suitable times and the remainingenzyme activity was assayed by determiningkcat. for 4-nitrophenyl trimethylacetate at pH7.00.Experimental details are given in the text.

pH2.252.863.003.273.303.503.663.904.40

kinact. (h-')ti < 30s1.190.4080.2570.08450.05270.00825Not determinedNot determined

102 x (kcat.).' (s-')0.1170.4220.3730.9090.9870.5 10*0.253*2.30t2.39t

*Only small changes in the activity were observedover 24 h, requiring a large extrapolation to obtain(kcat.)o, and kinact., with consequent loss of precision.For this reason these points are excluded from Fig. 5.

t This was the value of (kcat.)' measured after 72 h.t This was the value of (kcat.)' measured after 48 h.

(A

.8

x

116

pHFig. 5. pH profile of limiting values of (ka8.), reachedafter incubation of carboxypeptidase Y in various

H20-based buffersThe (k,.t )' values were obtained by non-linearregression analysis by using the first-order rateequation as described in the text. The line describesthe ionization of pK1PP = 3.4. See the text forexperimental details.

a

_ *

I l-

847

W. T. CHANG AND K. T. DOUGLAS

extremely pH-sensitive for this small pH range, andcan be described by the linear relationship (eqn. 3):

log (kinact.) = 7.14 (+0.1 1)-2.46 (±0.34) pH (3)

In general, the activity remaining did not reach zerobut levelled off (e.g. see Fig. 4). The 'equilibrium'values of (kCt )'H after incubation, (kcat ) , in variousH20-based buffers were obtained from the fit to thefirst-order equation (kcat.)i = (kcat.)' e-ke and areplotted versus pH in Fig. 5. Although the scatter israther high, for visualization purposes we havedrawn an ionization curve for an acid of pKapp. 3.4through the data.

Incubation in pH 3.0 H20-based buffer alsodestroys activity towards Cbz-Glu-Tyr. Althoughonly a few points were determined for the timecourse with Ac-Tyr-OEt, the overall course wassimilar to that with the kcat. of 4-nitrophenyltrimethylacetate as the assay for activity. Incu-bation of carboxypeptidase Y in buffers of pH8.0,pD 8.4, pH 3.0 and pD 3.0 gave the relative peptidaseactivities of 1.0, 1.0, 0.48 and 0 respectively.

Discussion

The difference between the doubly sigmoidal kcat(k3) profile in 2H20 and the singly sigmoidal k3 pro-file in H20 for the carboxypeptidase Y-catalysedhydrolysis of 4-nitrophenyl trimethylacetate can beexplained by the inactivation at low pH and pD. Therate of inactivation at low pH is sufficiently slow tobe undetected on the time scale of the experiments tomeasure the k3-pH profile. However, deactivation atlow pD occurs in a few seconds, i.e. faster than thetime taken to measure kcat in 2H20. The limb ofpKDp - 3.8 refers to the inactivation process inacidic 2H2O media. The higher pKD t, (-6.8) pre-sumably corresponds to the ionization of pKHp =5.35 observed in H2O. However, the shift in pK.expected of an acid ofpK' -5 on transfer from H20to 2H20 is only of the order of +0.5 pK unit (Bell &Kuhn, 1963). It is likely that kcat in 2H20 is stillequal or close to k3, in view of the retention ofquantitative bursts. The unusually large shift (of-+1.5 pK units) for the enzyme may indicate achange in environment for the feature responsible forthe ionization with pKHpp.*5.3, or an entirelydifferent ionization may be involved. It has beenreported (Bai et al., 1975) that, in the carboxy-peptidase Y-catalysed hydrolysis of N-acetyl-L-phenylalanyl ethyl ester, inhibition by L-phenyl-alanine was non-competitive in 2H20 (pD 7.0), butcompetitive in H20 (pH 6.5). This change in inhibi-tion type may indicate a difference in environment(e.g. a minor conformational change or a con-figurational alteration of the active site), such as thatdiscussed above.

There is a report (Bai et al., 1975) of kcat (H/2H)

equal to 2.33 and 1.34 for carboxypeptidaseY-catalysed hydrolysis of N-acetyl-L-phenylalanylethyl ester and benzyloxycarbonyl-L-phenylalanyl-L-leucine respectively in basic media. One cancalculate a kinetic solvent deuterium isotope effecton k3 (H/2H) in basic media from our results of -2.9for hydrolysis of trimethylacetyl-carboxypeptidase Y(by using the alkaline plateaux rate constants forcomparison). For the low-pH plateaux [comparingvalues of 1 x 10-Is-I for k3 in H20 (see Fig. 1) and5.4 x 10-3s -1 for k3 in H20 (see Fig. 2)] a lowervalue of 1.9 is obtained. This change in kineticsolvent isotope effect may reflect the effect of thesame feature as perturbs the value of pKapp.,described above. With such possibilities of con-formational or even mechanistic changes with pH, itis difficult to use kinetic solvent isotope effects onkcat as a mechanistic probe for this enzyme. This istrue even of the simple kinetic system chosen here(with kcat =k3, a single microscopic rate para-meter) and underscores the difficulties of inter-preting kinetic solvent isotope effect data on un-defined kcat values, which may well be composite,for more complex hydrolyses.

The inactivation at low pH probably explains theobservation (Hayashi et al., 1968) that, in thepreparation of carboxypeptidase Y from Orientalyeast, there was a rapid appearance of activity froma crude yeast autolysate at pH 3, but that this wasfollowed by a decrease in activity. Hayashi (1976)has reported that the salt-free pure enzyme losessome 20% of its activity ai being freeze-dried, butthat almost full activity is restored by leaving asolution of the freeze-dried enzyme overnight at 5oCor at -200C. In addition, freeze-dried enzyme whenredissolved requires a long equilibration time (- 24 h)before a stable reading of the A280 can be made(Margolis et al., 1978). These observations mayagain reflect conformational flexibility in it. It is alsoapposite to report that this enzyme has been isolatedin several distinct molecular forms, differing pri-marily in attached carbohydrate. These differencesare reflected in the microscopic pre-steady-state rateparameters for the 4-nitrophenyl trimethylacetatereaction as well as in the c.d. spectra (Margolis etal., 1978).

The rates of inactivation at low pH and low pDare enigmatic; e.g. consider the tremendous differ-ence in the rates of deactivation. In 2H20 at pD 3.0,over 80% of the activity is lost in the first minute; inH20 at pH 2.86 (Fig. 4) the half-life is approx.35 min (=O.7/kinact ). Such an enormous inverseapparent kinetic solvent isotope effect of over 30(2H/H) is untypical of a simple single-step process,and presumably indicates some form of co-operativebreakdown of catalytically viable tertiary structureas an explanation of the inactivation process. Therate law for the dependence of rate of inactivation on

1980

848

pH (pD) PROFILES OF CARBOXYPEPTIDASE Y ACTIVITY 849

pH (eqn. 3) with its high coefficient for pH-dependence again indicates a complex inactivationpathway. Specific acid catalysis would produce acoefficient of-1.0.

The dangers of 2H20 solvent perturbations inprotein chemistry have long lurked as a caveat tokinetic solvent isotope studies, to protein n.m.r. in2H20 and, more recently, to neutron-scatteringstudies of biological materials. Although the H20-2H20 'solvent' change can alter protein conforma-tion for ribonuclease (Hermans & Scheraga, 1959;French & Hammes, 1965) as well as possiblyaltering the degree of association of some oligo-meric proteins (Henderson et al., 1970) and affect-ing proline racemase (Cardinale & Abeles, 1968),this major kinetic difference between the acidicinactivation of carboxypeptidase Y in H20 and2H20 contrasts sharply with the minor effects of the2H20-H20 change for proteins in general. Forexample, a-chymotrypsin has been carefully studiedin 2H20 media and no unusual effects were noted.Changes in the magnitudes of rate constants or pKvalues are in line with straightforward physico-chemical expectation (Bender & Hamilton, 1962).The results with carboxypeptidase Y underscore theneed for careful control experiments with proteins in2H20 media.

We are pleased to acknowledge financial support of thisresearch by the Health Research and Services Founda-tion of the United Way in Pittsburgh, and Mr. JohnTorkington for programming assistance.

References

Aibara, S., Hayashi, R. & Hata, T. (1971) Agric. Biol.Chem. 35, 658-666

Bai, Y., Hayashi, R. & Hata, T. (1975) J. Biochem.(Tokyo) 78, 617-626

Bell, R. P. & Kuhn, A. T. (1963) Trans. Faraday Soc. 59,1789-1793

Bender, M. L. & Hamilton, G. A. (1962) J. Am. Chem.Soc. 84, 2570-2576

Bender, M. L., Schonbaum, G. R. & Zerner, B. (1962)J.Am. Chem. Soc. 84, 2562-2570

Cardinale, G. J. & Abeles, R. H. (1968) Biochemistry 7,3970-3978

Douglas, K. T., Nakagawa, Y. & Kaiser, E. T. (1975)Abstr. Nat. Meet. Am. Chem. Soc. 170th BIOL 62

Douglas, K. T., Nakagawa, Y. & Kaiser, E. T. (1976) J.Am. Chem. Soc. 98, 8231-8236

Douglas, K. T., Nakagawa, Y. & Kaiser, E. T. (1977) J.Org. Chem. 42, 3677-3681

French, T. C. & Hammes, G. G. (1965) J. Am. Chem.Soc. 87, 4669-467.3

Hayashi, R. (1976) Methods Enzymol. 45B, 568-587Hayashi, R. (1977) Methods Enzymol. 47E, 84-93Hayashi, R., Oka, Y., Doi, E. & Hata, T. (1968) Agric.

Biol. Chem. 32, 367-373 -

Hayashi, R., Moore, S. & Stein, W. H. (1973) J. Biol.Chem. 248, 2296-2302

Hayashi, R., Bai, Y. & Hata, T. (1975) J. Biochem.(Tokyo) 77, 1313-1318

Henderson, R. F., Henderson, T. R. & Woodfin, B. M.(1970) J. Biol. Chem. 245, 3733-3737

Hermans, J. & Scheraga, H. (1959) Biochim. Biophys.Acta 36, 534-535

Hermodsen, M. A., Kuhn, R. W., Walsh, K. A., Neurath,H., Eriksen, N. & Benditt, E. P. (1972) Biochemistry11, 2934-2938

Kezdy, F. J. & Bender, M. L. (1962) Biochemistry 1,1097-1106

Kuhn, R. W., Walsh, K. A. & Neurath, H. (1974)Biochemistry 13, 3871-3877

Lee, H.-M. & Riordan, J. F. (1978) Biochem. Biophys.Res. Commun. 85, 1135-1142

Liberatore, F. A., Mclsaac, J. E., Jr. & Royer, G. P.(1976) FEBS Lett. 68, 45-48

Margolis, H. C., Nakagawa, Y., Douglas, K. T. & Kaiser,E. T. (1978)J. Biol. Chem. 253, 7891-7897

Martin, B., Svendsen, I. & Ottesen, M. (1977) CarlsbergRes. Commun. 42, 99- 102

Weber, K. & Osborn, M. (1969) J. Biol. Chem. 244,4406-4412

Vol. 187