Vol. 72

The Mechanism of Ficin-Catalysed Reactions

By B. R. HAMMOND AND H. GUTFREUNDNational In8titute for Research in Dairying, Shinfield, near Reading, Berks

(Received 27 October 1958)

During the last few years a number of hydrolyticand acyl-transfer enzymes have been studied indetail and marked similarities in their reactionmechanisms have been found. Cholinesterase(Wilson & Calib, 1956), chymotrypsin (Gutfreund& Sturtevant, 1956) and trypsin (Gutfreund,1955a) have been shown to react with their sub-strates by a three-step mechanism involving aninitial rapid adsorption of the substrate and asubsequent chemical reaction with an active siteconsisting of a histidine imidazole and a serinehydroxyl group. More recent studies on phospho-glucomutase (Koshland & Erwin, 1957) and onthrombin (Gladner & Laki, 1958) revealed thatthese enzymes also have serine hydroxyl groups asa component of their active sites. Gutfreund(1955b) and Bernhard & Gutfreund (1956) proposeda catalytic mechanism for ficin which is almostidentical with that proposed for papain-catalysedreactions by Kimmel & Smith (1957). This mech-anism again involves an initial adsorption processand a subsequent chemical reaction with thecatalytic groups. The catalytic site of ficin containsa sulphydryl group which is acylated during thereaction between enzyme and substrate. Thispaper describes experiments which were carriedout to extend the studies of the properties andreactivity of the active site of ficin, which werebegun by Bernhard & Gutfreund (1956).

EXPERIMENTAL

Method8The constant-pH titration procedure as adapted byHammond & Gutfreund (1955) was used to follow the rateof enzyme-catalysed ester hydrolysis reactions. Standardconditions for the assay of ficin activity are: temperature,25±0.050; 15 ml. of reaction mixture of the followingcomposition: benzoyl-L-arginine ethyl ester, 13-7 mm;NaCl, M; ethylenediaminetetra-acetic acid (EDTA; mM).Appropriate volumes of enzyme are included in 15 ml. ofreaction mixture. The rate of hydrolysis can be expressedas moles of acid produced/sec./mole of enzyme.For the measurement of amide hydrolysis it was found

that the photometric ninhydrin method of Moore & Stein(1954) gave sufficient sensitivity and accuracy to allow thedetermination of initial rates when only small quantities ofammonia had been produced.

MaterialsN-Benzoyl-L-arginine ethyl ester (BAEE) and N-

benzoyl-L-arginine amide (BAA) were prepared from L-arginine monohydrochloride (British Drug Houses Ltd.)after the method of Bergmann, Fruton & Pollock (1939).BAA gave the following figures on analysis [the figures in

parentheses are those calculated for the formula BAA,HC1,H20 (M, 332)]: N = 20-4% (21-1), Cl = 9.8% (10-7),H20 = 5-8% (5-7).BAEE is a hygroscopic substance associated with

varying amounts of water. Solutions of this substratewere analysed by trypsin hydrolysis. Assays were generally95% of the value estimated from the known weight ofsubstrate and a molecular weight of 343, based on theformula BAEE,HC1. The chief impurity was water.

N-Benzoylglycine methyl ester (hippuric methyl ester)was prepared from hippuric acid (British Drug Houses Ltd.,London) by the ion-exchange method. IR-120 (H) resin(British Drug Houses Ltd., London) was freed from colour-ing matter by refluxing with methanol before use. Driedhippuric acid (5 g.) was suspended in 200 ml. of drymethanol and refluxed for 12 hr. with 5 g. of the resin.Finally the resin was filtered off and the methanol removedunder vacuum. The ester was recrystallized from aqueousmethanol in approx. 80% yield. This method is unsuitablefor positively charged amino acid derivatives since theresin is positively charged.

Hippuric amide was prepared by saturating a solutioncontaining 5 g. of hippuric methyl ester in 200 ml. of drymethanol with ammonia gas under anhydrous conditions.After storage of the resulting solution at room temperaturefor 2 days methanol was removed under vacuum untilcrystallization commenced. The amide was recrystallizedfrom aqueous methanol with approx. 75% yield.

Ficin was obtained as the dried latex from L. Light andCo. Ltd., Colnbrook, Bucks. In this form the enzyme wasfound to be stable for periods of over 1 year, judged by theconstant specific activity obtained with different prepara-tions during this time interval. The preparation was donein a cold room at 4°. Crude latex (20 g.) was suspended in200 ml. of 0-01 N-HCI (resulting pH 4-0) and stirred for2 hr. The resulting suspension was then dialysed withstirring against 51. of mM-EDTA (pH 4-4) for 15 hr. Theinsoluble material was removed by centrifuging and theresultant enzyme solution [(1) of Table 1] was purified byammonium sulphate fractionation, details of which areshown in Table 1. After each precipitation the suspensionof precipitate was dialysed against mM-EDTA (pH 4-4) andthe volumes indicated are those of the material redissolvedin this manner. Except for pilot experiments fraction (8)was used for kinetic measurements.

349

B. R. HAMMOND ANTD H. GUTFREUNDThe protein concentrations given in Table 1 were deter-

mined by measurement of extinction at 279 m,u after a50-fold dilution of the enzyme solution. A calibrationcurve relating extinction to Kjeldahl nitrogen enabled theeasy evaluation of protein concentration, by assumingthat the protein contained 16.0% of Kjeldahl nitrogen.An extinction of 1 corresponded to protein concentration of2-2 %. The specific enzymic activities listed in column fourhave been converted into maximum velocities of hydrolysisfor 0-1 ml. of a 1% protein solution, with the value of Kmfor BAEE determined later.

In order to obtain ficin preparations which were fullyactive without the addition of a reducing substance, it wasfound necessary to work as rapidly as possible at lowtemperature (40). Such preparations were required forwork with organic mercury inhibitors and for kineticinvestigations in alkaline solution.

RESULTS

Fully active ficin preparations in the absence of theactivator show stability in the range pH 3-5-9-0.This is based on the immediate recovery of fullactivity at the pH optimum after storage at theextremes of pH for times of up to 60 min. at 25°.Fully active preparations show no enhancement ofactivity on the addition of 2 m -2:3-dimercapto-propanol when first prepared, but slowly loseactivity on storage at 4°. After 6 weeks theactivity is approximately one-half of its initialvalue. It is likely that such a change is caused by

I959oxidation of thiol groups to the disulphide form,possibly catalysed by trace-metal impurities in themanner suggested by Lamfrom & Nielson (1957).This oxidation was found to be completely reversedby the addition of dimercaptopropanol.The addition of mercury compounds to ficin

solutions (fully active without reducing substances)destroys enzymic activity. The activity is, however,completely restored by 2 mM-2:3-dimercaptopro-panol. This indicates that the loss of enzymicactivity is not due to denaturation of the protein.By titration of ficin activity with methylmercurichydroxide it was found that 1 mole of this inhibitorcompletely inactivated 26 kg. (or 1 mole) ofenzyme(Bernhard & Gutfreund, 1956). These resultssuggest strongly that a thiol group of the enzyme isessential for its activity and that there is only onegroup/enzyme molecule.

Variation of substrate and substrate concentrationFicin-catalysed hydrolyses of BAEE, BAA and

hippuric methyl ester were found to obey accuratelyMichaelis-Menten kinetics. The observed velocityof hydrolysis was found to be proportional to theenzyme concentration over a 25-fold range.The main limitation of the accuracy of the

Michaelis parameter determinations with Eadie(1942) plots was the maximum solubility of sub-strate relative to KIM. Table 2 lists the values of the

Table 1. Assays of N-benzoyl-L-arginine ethyl ester relating to ficin preparationand description of successive operations

Assay number and description ofsuccessive operations

(1) Original solution (195 ml.)(2) 0-5 Saturated ammonium sulphate

precipitate (158 ml.)(3) Supernatant (300 ml.)(4) 0-3 Saturated ammonium sulphate

precipitate (120 ml.)(5) Supernatant (158 ml.)(6) 0-5 Saturated sodium chloride

precipitate (100 ml.)(7) Supernatant (116 ml.)(8) 0 5 Saturated sodium chloride

precipitate (100 ml.)(9) Supernatant (98 ml.)

Protein(%)2-281-90

0-471-42

0-690*81

0-750-49

0-31

Total amountof protein

(g.)4-443-00

1-411-71

1-090-81

0-870-49

0-30

108 x Specificesterase activity(moles/sec./0.1 ml.of 1% protein)

4.95-5

0-849*02-39.75.09-7

8*3

Table 2. Michaelis parameters for the ficin-catalysed hydrolyses of N-benzoyl-L-arginine ethyl ester,N-benzoyl-L-arginine amide, hippuric methyl ester and hippuric amide

Results were obtained at 250 and pH 6-06 in the presence of mM-EDTA buffer and m-NaCl. For BAEE under theseconditions V.,, = 9-6 x 10-' moles/sec./0-1 ml. of 1% enzyme solution. Hippuric Hippuric

Substrate BAEE BAA methyl ester amide102 x Concentration range (M) 0-5-5-0 1-0-8-0 0-25-3-25 0-40-3-0102 Km (m) 2-5±0-3 48±0-5 4-8±1O0 13±3V.n,a (relative values) 1-0+0-1 0-9±0-1 1-4±0-2 0-04±0-01

350

MECHANISM OF FICIN-CATALYSED REACTIONS

Michaelis parameters of the various substrates atpH 6-06 and 250. The table also includes the rangeof substrate concentrations used for each determi-nation.

Effect of pHThe stability of ficin under different conditions of

pH and activator concentration has already beenmentioned. Fig. 1 shows the pH-dependence of theobserved rates of hydrolysis of four substrates.For the esterase runs at low pH, where the

products were incompletely ionized, a small per-centage of the product did not produce hydrogenions. Such hydrolysis rates were corrected byusing pK values for benzoyl-L-arginine or benzoyl-L-glycine.

Only esterase runs were attempted in alkalinesolution because of the difficulties of working in acarbon dioxide-free atmosphere. All measurements

10090

_ 80 --° 70>- 60

:' 50u 40

30 -

20

3 5 4-0 4-5 5 0 5-5 6-0 6 5 7-0pH

Fig. 1. Variation of observed rates of ficin-catalysedhydrolysis with pH for the substrates 0, BAEE, *,hippuric methyl ester, A, BAA and A, hippuric amide.Substrate, 0-02M; NaCl, 0-O1M. Results are expressed aspercentages of their respective rates at pH 6-5, so thatthey can be presented as one graph.

Table 3. pH-Dependence of Km value for the actionof ficin on N-benzoyl-L-arginine ethyl ester andN-benzoyl-L-arginine amideResults wereobtained at 25°in mm-EDTAand0-1m-NaCl.

Substrate pH 102 Km (M)BAEE 8-43 2-1

6-33 2-56-04 2-86-04 2-46-04 2-66-04 2-76-04 2-35*34 2-04-63 2-34-63 2-34-63 2-24.04 3-04.04 3-3

BAA 6-30 5-26-08 4-45-15 4-14.33 4.5

at alkaline pH were performed under an atmos-phere of carbon dioxide-free nitrogen. A blank runwas performed which furnished a correction for thespontaneous hydrolysis of the substrate, withpossibly a small contribution from absorption ofcarbon dioxide. Such corrections were only a fewper cent of the measured enzymic rates.The full line in Fig. 1 is the theoretical curve for

the ionization of two groups of the enzyme(pKB 4 40 and pKA 8.46) and indicates how closelythe experimental points fit an ionization curve.Km measurements at 250 for the substrates

BAEE and BAA at the pH specified are given inTable 3. This is to indicate the accuracy of themeasurements. There may be a slight increase inKm at pH 4 04, but in view of the reproducibilityof the measurements this is barely significant.The variation of ficin activity with pH must be

interpreted in terms of ionizations of two groups onthe enzyme, as the substrates used do not changetheir states of ionization in this pH range.The relevant equation for the variation of

enzymic activity in terms of the ionization of botha basic and an acidic group is

V° [H+] KA= 1+ K +

VO is the rate of hydrolysis at the optimum pH andVi is the rate of hydrolysis at other pH values. Thesubscripts of the ionization constants, KA and KB ,indicate whether the ionizing group is a Bronstedacid or base at the pH optimum. The KA term isdominant at low concentrations of hydrogen ionand the KB term is dominant at high concentra-tions of hydrogen ion.

Fig. 2 shows the data of Fig. 1 for BAEE andhippuric methyl ester plotted in terms of the aboveequation. In acid solution the graph of V0/V'versus [H+] has a slope 1/KB and in alkaline solu-tion the graph of VO/V' versus 1/[H+] has a slopeKA. The two pK values for each substrate at 250are shown on the graph.

Influence of ionic 8trengthFicin action has been found to depend markedly

upon salt concentration. Fig. 3 shows the variationof the observed velocity of hydrolysis at low ionicstrengths for the substrates BAEE, BAA andhippuric methyl ester. Different salts were used toinvestigate whether the effect was due to ionicstrength.

It was not possible to work at ionic strengthsbelow 0 01 M (the substrate concentration em-ployed) because the arginine substrates themselvesmake a contribution to the ionic strength. It wasnot possible to decide whether such an effect wasdue to changes in one of the Michaelis parametersor in both.

Vol. 72 351

B. R. HAMMOND AND H. GUTFREUND

At high ionic strengths, however, the results ofTable 4 suggest that any change in the observedvelocity of hydrolysis is due to changes in V.and that Km is almost constant. Fig. 4 presents theresults of an investigation on the effects of differentsalts on the hydrolysis of BAEE at high ionicstrengths.

5

4

3

2

06

5

4

2

0

Fig. 2. Determination of the dissociation constants, KAand KB, of groups on the enzyme from the data of Fig. 1and Table 3 for the pH dependence of the hydrolysis ofBAEE (0) and hippuric methyl ester (0).

0-760q75

.0

0 0.70

0-650

I959Influence of propan-2-ol

The effect of propan-2-ol on the ficin-catalysedhydrolysis of BAEE, BAA and hippuric methylester was studied at pH 6-3 and 25-00 with mM-EDTA as buffer and in the absence ofadded sodiumchloride. The maximum volume concentration ofpropan-2-ol used was 33% (approx. 4-3M) and nodenaturation of enzyme occurred under these con-ditions, as judged by the complete recovery ofenzymic activity on dilution.The results obtained for the inhibition of ficin

action by propan-2-ol do not fit the usual equationsfor competitive or non-competitive inhibition. Thevariation of the Michaelis parameters for BAEE,BAA and hippuric methyl ester with propan-2-olconcentration is given in Table 5.For all substrates the presence of propan-2-ol

causes a marked increase in Km. The influence ofpropan-2-ol concentration on V,.. for thecharged substrates BAEE and BAA is also verysimilar. For the uncharged substrate hippuricmethyl ester, however, there does not appear to beany significant change in Vm with increasing

Table 4. Effect of sodium chloride concentration onKm and V.. for the action of ficin on N-benzoyl-L-arginine ethyl esterResults were obtained at 25° and pH 6-04 in mm-EDTA.

In the presence of m-NaCl, V.,,. = 9-6 x 10-8 M/sec./0-1 ml.of % ficin.

V.,, (relative)[NaCl] (M) 102Km (M) values

0-05 2-5 1-00-50 2-5 1.11-00 2-7 1-1-70 2-8 1-5

8

-1

- ,go0

0

0

W5 0-10 0-1 5-I1

0-20 0-25

Fig. 3. Variation of observed velocity of ficin-catalysedhydrolysis of BAEE, BAA and hippuric methyl ester atlow ionic strengths; pH 6-04, mn&-EDTA, 0-01-sub-strate. BAEE-NaCl, 0; BAA-NaCl, i\; hippuricmethyl ester-NaCl, 0; BAEE-Na2SO4, 0.

,0

0. 1-0 2-0 3-0

Fig. 4. Effect of different salts on the ficin-catalysedhydrolysis of BAEE; pH 6-04, mM-EDTA, O-O1M-substrate. NaCl, 0; KCI, 0; NH4C1, 0; KBr, *;KI, x; NaNO8,AA; KNOs, A; KSCN, +.

352

C)

1

MECHANISM OF FICIN-CATALYSED REACTIONSTable 5. Effect of propan-2-ol concentration on Kmand V,,,, for the action of ficin on N-benzoyl-L-arginine ethyl ester, N-benzoyl-L-arginine amideand hipuric methyl e8ter at specified pH and 250mM-EDTA buffer was used, in the absence of added

sodium chloride.

Substrateand pH

BAEE, 6-04

BAA, 6-30

Hippuric methylester, 6-34

Ooncn. of V.,..propan-2-ol (relative)

(m) 102 Km (M) (%)

0 2X5 1000X87 4-6 761-74 6-5 542-63 6-7 290 5 1001-31 11 682-62 45 230 5 1000-87 9 1002-62 25 100

0~~~~

10 (M)

9

8

x7

§6-

4

20 1 2 3 4

102[SO] (M)Fig. 5. Effect of propan-2-ol concentration (0, 2-62mr;0, 0.87m) on the ficin-catalysed hydrolysis of hippuricmethyl ester at pH 6-04 in mi-EDTA. A, No propan-2-ol.

propan-2-ol concentration. The evidence for this isshown in Fig. 5, where Eadie (1942) plots for thedetermination of the Michaelis parameters atspecified propan-2-ol concentrations are shown.The slope of an Eadie plot is 1IVi.. and theordinate intercept is Km/Vma..

It is difficult to evaluate the Michaelis para-meters accurately in this case because the solu-bility ofhippuric methyl ester is low compared withthe value of Km to be determined. A decrease inV,,,= with increasing propan-2-ol concentrationwould require an increase in slope of the Eadieplots. The experimental results do not support this.The lines drawn through the experimental pointsin Fig. 5 have a constant slope which would implythat VmX. is unaffected by propan-2-ol for thissubstrate. The degree to which the experimentalresults support this proposal may be gauged fromthe Figure. This relative invariance of V..for hippuric methyl ester may be compared withthe change in V. observed with BAEE andBAA at corresponding propan-2-ol concentrations(Table 5). There is, however, an increase in Km forhippuric methyl ester and this is illustrated by theincrease in the ordinate intercept in Fig. 5 withincreasing propan-2-ol concentration.

Effect of temperatureThe variation of the Michaelis parameters Km

and Vx., in the temperature range 10-400, hasbeen investigated for the substrates BAEE, BAAand hippuric methyl ester. The ester substrateswere studied at two pH values; the amide wasstudied in the region of the pH optimum only. Theresults axe set out in Table 6.

Table 6. Variation of the MichaeUis parameters with temqperature and the derived Arrhenius activation energymM-EDTA was used as buffer throughout. Different convenient amounts of enzyme were used for each substrate:

BAEE experiments, 13-3 x 10-8 moles of enzyme; BAA experiments, 115 x 10-8 moles of enzyme; hippuric methyl ester(HME) experiments, 5-06 x 10-8 moles of enzyme.

Substrate Temp. Vl Eaand pH (o K) (arbitrary units) (kcal./mole) 102 Km (M)BAEE, 6-25 285-0 11-6 1.9

33-285-66-6

19*658-89.0

17-738-65-29-1

24-09-1

25-862-4

12-8

14-0

10-4

11.0

11-6

2-53-3

3.94X86-5

4.34-85.5

Bioch. 1959, 72

BAEE, 4-60

ITE, 6-25

HME, 4-25

BAA, 6-25

298*0312-5285-0298-0312-5287-7297-9312*8287-7297-9312-8284-2298-6313-7

23

Vol. 72 353

B. R. HAMMOND AND H. GUTFREUND

Fig. 6 is a graph of the data for hippuric methylester to determine the activation energy, EB, bymeans of the usual Arrhenius equation

log k = 233RT+ comtant.

This relationship was found to be linear for allsubstrates in the temperature range 10-400, withinexperimental error.

DISCUSSION

The results of detailed kinetic investigations of theficin-catalysed hydrolysis of a number of esters andamides give information about the physicochemicalproperties of those groups on the enzyme moleculewhich can be called the catalytic groups. Fromthese properties one can make deductions about theprobable identity of these groups and about themechanism of the reactions which they undergowith the substrates.A number of different methods have been used

to determine the dissociation constants of groupsessential for the reactivity of enzyme-substratecompounds. Our previous studies on trypsin(Gutfreund, 1955a) and chymotrypsin (Hammond& Gutfreund, 1955) encouraged us to use the simplemethod of interpreting plots of reaction rateagainst pH as titration curves to get a first pictureof the situation. Fig. 1 shows that one acidic andone basic group are involved in the reactivity officin. The pK values for the two groups areevaluated from the plots in Fig. 2. Comparisons ofpK values obtained for groups on protein surfaceswith known constants for free amino acids have tobe treated with caution. For the characterizationof active groups one must consider both the generalvariations of pK values with charge configurationand dielectric properties of the surroundings and

33104 (1/T)

Fig. 6. Analysis of data for temperature-dependence of theficin-catalysed hydrolysis of hippuric methyl ester (seeTable 6); mM-EDTA; M-NaCl; *, Km at pH 6-25; A,VM,. at pH 6x25 and 0, V,,. at pH 4x25.

the special circumstances which are responsible forthe peculiar reactivity of groups on the active site;this point is elaborated below. The acidic group ismost likely to be a carboxyl. The basic groupcannot be identified from its pK alone, but it willbe shown below that there is good evidence for apositively charged group on the active site. Itcan therefore be concluded that an NH3+ group isinvolved. None of the substrates used has groupswhich change their ionization over the pH rangeexamined here.

It has been shown previously (Bernhard &Gutfreund, 1956) that a sulphydryl group isrequired for the activity of the enzyme. This hasbeen confirmed during the present studies bytitration with methylmercuric hydroxide and sub-sequent reactivation by dimercaptopropanol.

Fig. 3 shows that the effect of variation of ionicstrength (I) in the observed velocity of hydrolysisat low values of I is different for charged and un-charged substrates. The line drawn through theexperimental points for BAEE and BAA in Fig. 3has the theoretical slope for a uni-univalent-likecharge interaction with the well-known Debye-Huckel-Bronsted equation. As the arginine sub-strates have unit positive charge at pH 6-0 theseresults suggest that there is a positively chargedgroup at the active site of the enzyme. There is asimilar enhancement of observed velocity ofhydrolysis for charged and uncharged substrates atconcentrations of sodium chloride greater thanapproximately 015M.

Propan-2-ol is a suitable solvent for studies ofdielectric effects on enzyme systema. Although alarge variety of complex changes can occur withaltered solvent composition, the data presented inFigs. 5 and 7 are readily interpreted as a dielectriceffect on the charge interaction in the enzyme-substrate compound. Propan-2-ol molecules are nodoubt adsorbed on the enzyme by van der Waals'forces and hydrogen bonds. This would hinder theadsorption of the substrate at the active site andthe formation of the Michaelis-Menten complex.It is possible to explain the increased Km values inthe presence of propan-2-ol in this way. On amolar basis propan-2-ol is a better inhibitor thanethanol, presumably because of its greater van derWaals' attraction to the protein. The invariance ofV... with propan-2-ol concentration for theuncharged substrate hippuric methyl ester com-pared with the decrease observed for the chargedsubstrates BAEE and BAA su-gosts that the last-mentioned effect is most likely due to the repulsionbetween the two positive charges on the substrateand active site. The variation of V, for BAEEhydrolysis with propan-2-ol concentration ispresented in Fig. 7 in terms of the theoreticalrelationship between the log of the rate constant

354 I959

MECHANISM OF FICIN-CATALYSED REACTIONSand the reciprocal of the dielectric constant of themedium (Scatchard, 1932). A linear relationship isobtained and its slope enables the calculation of thedistance of closest approach of the arginine sidechain and the active site, assuming that the inter-action is between two unit positive charges. Thevalue of 131 so obtained is small but of a reason-able magnitude for the approximations involved.Two important conclusions can be drawn from

the investigations of the effect of temperature onthe ficin-catalysed hydrolysis of esters and amides.First it can be seen in Table 6 that the differencebetween the energy of activation at a pH near thepK ofthe acidic group and at pH 6 25 is small. Onecan derive from this that the heat of ionization ofthe acidic group is small (Gutfreund, 1955 a), whichwould confirm that a carboxyl ionization is in-volved. The second point of interest is demon-strated by the fact that the energies of activationfor the ficin-catalysed hydrolysis of esters andamides are very similar. [For the trypsin-catalysedhydrolysis of the same pair of ester and amide sub-strates considerably different energies of activationwere found (Schwert & Eisenberg, 1949; Butler,1941).] This confirms an earlier conclusion fromsimilarity of rate of hydrolysis (Bernhard & Gut-freund, 1956) that the rate-determining step for thetwo substrates is the same. Since esters and amideshave such widely different characteristics towardshydrolytic attack as demonstrated, for example,by the oxygen-exchange experiments of Bender,Gringer & Kemp (1954), it can only be concludedthat the rate-determining step is the decompositionof an intermediate common to the reaction of ficinwith BAEE and BAA. It seems reasonable tosuggest that this common intermediate is anenzyme acylated by the substrate at the sulphydryl

12 13 14102/6

Fig. 7. Variation of log V,. with di1for the ficin-catalysed hydrolysis ofin mm-EDTA (see Table 5).

group. The reaction path of ficin (E) with BAA orBAEE (AB) can then be written:

k, k2 k3E +AB EAB -+ EA+ B -+ E +A+B

k-Iwhere EAB is a rapidly formed loose enzyme-substrate compound; the formation of the acylenzyme EA and its decomposition to free enzymeand acid can be regarded as irreversible whileproduct concentrations are low. From the dis-cussion above it is concluded that the step charac-terized by k.3 is rate-determining and correspondsto the turnover number or Vm..a of the enzyme.The step characterized by k2 (the acylation of theenzyme) must be slower for an amide substratethan for an ester. It was stated above that thesimilarity of overall rate of hydrolysis and energiesof activation of the reactions with BAA and BAEEindicates the existence of a common intermediate.The coincidence of the dependence of the rate ofreaction of the two substrates on pH, temperatureand ionic strength is, however, not sufficientlyexact to exclude the possibility that for BAA k2 andk3 are of the same order of magnitude. Somecomments must be made on the relative rates ofhydrolysis of hippuric methyl ester and hippuricamide. The fact that Vm.. for hippuric methylester is somewhat larger than for BAEE indicatesthat the decomposition of hippuryl acylatedenzyme is somewhat faster than that ofthe benzoyl-L-arginyl acylated enzyme. The very much slowerrate of hydrolysis of hippuric amide indicates thatfor this substrate the formation of the acyl enzymeis rate-determining. The low solubility of the un-charged substrates and the large Km for hippuricamide make accurate calculation of constants forthis substrate difficult.One of us (Gutfreund, 1955b) has attempted to

determine k1 for the reaction of ficin with BAEE bythe 'initial acceleration' method. The value of500 m-1 sec.-l was assigned to k, because it wasshown to be a second-order constant and only theformation of the first enzyme-substrate compoundis a second-order process. It can, however, beshown from the analysis of the three-step processused to describe the reaction of this and otherhydrolytic enzymes (Gutfreund & Sturtevant,1956) that if, at substrate saturation, [EAB] < [EA]the rate of formation of EA (the rate-determiningenzyme-substrate compound) will be

d[dA] = k2[EAB],dt

1-5 16 and if the steady-state concentration of [EAB],which is very small compared with the total

electric constant (s) enzyme concentration, is reached rapidly, then theBAEE IpH 6*04 pseudo-second-order constant quoted above is an

evaluation of k2 [EAB].23-2

VoI. 72 355

B. R. lVEMOND AND H. GUTFREUND

Free active enzyme

Substrate0

±R~

t+ H20 0

-R -

Initial adsorption

c(O,

\\ 0

N R

I± XH

Possible intermediateduring decompositionof acylated enzyme

Acylated enzyme

Scheme 1

We have in this manner obtained good evidencefor the chemical nature of the catalytic site and forthe relative rates of the three distinguishable stepsin the formation and decomposition of the enzyme-substrate compounds of ficin. The reaction mech-anism of steps involved in the interaction, betweenthe three groups CO2-, NH8+ and SH on the enzymesurface and the carboxyl group of different sub-strates, is open to a number of interesting specula-tions. We should like to propose the steps shown inScheme 1 for the reaction between enzyme andsubstrate.

Bender (1957) has described some intramolecular-catalysis reactions which would simulate suchsteps in enzymic reactions. Stockell & Smith(1957), Smith, Chavr6 & Parker (1958) and Kimmel& Smith (1957), who have undertaken a detailedanalysis of the reactions and properties of theclosely similar enzyme papain, have proposed asomewhat different reaction mechanism. Theyassign the pK 8-4 to the catalytic -SH group andassume that the enzyme cannot be acylated by thesubstrate when the sulphydryl group is ionized.For ficin we object to this interpretation on twogrounds. First, slowing down of the acylation bypartial ionization of -SH should not affect theoverall rate since the hydrolysis of the acyl enzymeis rate-determining. Secondly, this mechanism

does not involve the positive group known toparticipate in the reaction. Our interpretation in-volves both ionized -CO2 and NH8+ groups in therate-determining decomposition ofthe acyl enzyme.

Finally it should be pointed out that the relativestability of the acyl enzyme compound accountsfor the efficiency of the ficin- and papain-catalysedacyl transfer to acceptors other than water, as intranspeptidation. This is discussed by one of uselsewhere (Gutfreund, 1957) in quantitative terms.

SUMMARY

1. The kinetics of ficin-catalysed hydrolyses ofbenzoyl-L-arginine ethyl ester, methyl hippurateand of the corresponding amides was studiedunder a variety of conditions of temperature, pH,ionic strength and dielectric constant.

2. From the results of these experiments it canbe concluded that three reactive groups arenecessary for the catalytic action of ficin; these are-SH, NH3+ and C02Th

3. A reaction sequence between enzyme and sub-strate is proposed which involves the rapid forma-tion of a loose enzyme-substrate compound, asubsequent acylation of the enzymic -SH group bythe carbonyl of the substrate and finally the de-composition of the acyl enzyme.

356 I959

Vol. 72 MECHANISM OF FICIN-CATALYSED REACTIONS 3574. Reaction mechanisms for the second and

third step can be formulated by analogy withhomogeneous bifunctional catalysis.One of us (B.R.H.) is indebted to the Ramsay Memorial

Trustees for the award of a Fellowship. We should also liketo record our appreciation of the most helpful and stimu-lating advice we have received from Dr E. L. Smith atvarious times.

REFERENCES

Bender, M. L. (1957). J. Amer. chem. Soc. 79, 1258.Bender, M. L., Gringer, R. D. & Kemp, K. C. (1954).

J. Amer. chem. Soc. 76, 3350.Bergmann, M., Fruton, J. S. & Pollock, H. (1939). J. biol.

Chem. 127, 643.Bernhard, S. A. & Gutfreund, H. (1956). Biochem. J. 63,

61.Butler, J. A. V. (1941). J. Amer. chem. Soc. 63, 2971.Eadie, G. S. (1942). J. biol. Chem. 146, 85.Gladner, J. A. & Laki, K. (1958). J. Amer. chem. Soc. 80,

1263.

Gutfreund, H. (1955a). Tran8. Faraday Soc. 51, 441.Gutfreund, H. (1955b). Disc. Faraday Soc. 20, 167.Gutfreund, H. (1957). Advanc. Catalys. 9, 284.Gutfreund, H. & Sturtevant, J. M. (1956). Proc. nat. Acad.

Sci., Wash., 42, 719.Hammond, B. R. & Gutfreund, H. (1955). Biochem. J. 61,

187.Kimmel, J. R. & Smith, E. L. (1957). Advanc. Enzymol. 19,

267.Koshland, D. E. & Erwin, M. J. (1957). J. Amer. chem. Soc.

79, 2657.Lamfrom, H. & Nielson, S. 0. (1957). J. Amer. chem. Soc.

79, 1966.Moore, S. & Stein, W. H. (1954). J. biol. Chem. 211, 907.Scatchard, G. (1932). Chem. Rev. 10, 229.Schwert, G. S. & Eisenberg, M. A. (1949). J. biol. Chem.

179, 665.Smith, E. L., Chavre, V. J. & Parker, M. J. (1958). J. biol.

Chem. 230, 283.Stockell, A. & Smith, E. L. (1957). J. biol. Chem. 227, 1.Wilson, I. B. & Calib, E. (1956). J. Amer. chem. Soc. 78,

202.

The Effects of Adenosine Triphosphate on the MechanicalProperties of Surface Films of L-Myosin and Actomyosin

BY D. F. CHEESMAN, M. MARIE KEELER* AND 0. STEN-KNUDSENtDepartment of Phy8iology, Bedford College, University of London

(Received 20 October 1958)

The earliest study of the effects of the adeninenucleotides on the properties of the myofibrillarproteins spread at the air-water interface was madeby Munch-Petersen (1948), who found that whenactomyosin was spread on 0-5 M-potassium chloridesolution containing 0-1 mM-adenosine triphosphate,it occupied an area considerably greater, for agiven surface pressure, than in the absence ofadenosine triphosphate. The effect was later shownto be due, at least in part, to an increase byadenosine triphosphate in the rate of spreading ofthe protein (Lajtha & Rideal, 1952; Cheesman,1952). Similar effects were found for actin-freemyosin by Munch-Petersen, although Cheesmanwas unable to repeat this observation.The acceleration by adenosine triphosphate of

the spreading of actomyosin may be explained bythe dissociation of the proteii by this nucleotideinto its constituent actin and myosin. A similaracceleration is produced by high salt concentra-tions, which also dissociate actomyosin in bulk

solution. For this reason, the results seemed to beof little biochemical interest, especially as noevidence of chemical alteration of the protein byadenosine triphosphate could be obtained fromphase-boundary potential measurements (Chees-man, 1952).Cheesman & Sten-Knudsen (1959) have de-

veloped a procedure for the study of the visco-elastic properties of protein films. It was foundwith this method that when actomyosin or L-myosin was spread on M-potassiiln chloride solu-tion containing 0- 1 mM-adenosine triphosphate, themechanical rigidity of the resulting film was con-siderably lower than that of a similar film in theabsence of adenosine triphosphate at the samesurface pressure with the first protein, and at thesame surface pressure and specific area with thesecond. Two explanations of these findings pre-sented themselves. It was conceivable that acomplex of adenosine triphosphate and denaturedprotein persisted in the spread film, with a con-sequent 'effect upon the mechanical properties ofthe latter. Alternatively, since the protein underthe conditions of the experiments was not fullyspread, the effect could have been due to an inter-action between adenosine triphosphate and native

* Mother Marie de l'Enfant-J6sus, R.S.H.M. Presentaddress: Marymount College, Tarrytown-on-Hudson,N.Y., U.S.A.

t Present address: Universitetets NeurofysiologiskeInstitut, Copenhagen, Denmark.