768 After cooliiig to 20' the precipitate was filtered OE and the filtrate was evaporat'ed in mcuo. The residue solidified on standing aiid WYRSpurified as indicated in Table 11.
Acknowledgment.-The aut1iot.s \\-ish to thank Ilrs. I). 1'. ,Tacohiis, T. Ily prevent c d iroiii ioriiziiiy i t t t he enzyiiic~ suhst rat (> addi tioti coinplcs, thus sriggcst irig ihat- it is iiiasked through i r i l v i w t i o i i n-ii 11 t l i v w t c ~fiuictioii of the suhtratc,.Y csstci'at ic, sit(%\\.as also showii t'o be rcwt.ivcs totvards phosphoiyl, carbatiiyl, arid (1
'
h
i
( I ) (a) Published as p a r t 111 of tlie series "Studies un t h e Chemical Baais for Cholinomimetic and Cholinolytic Activity." For par t 11. see B. Belleau and J. Puranen. J . &fed. C h e m . . 6, 326 (1963). (b) This investigation was supported by t h e National Kfwtircll Council of Canada and represents ii Dintion of the thesis submitted bs- (;, 1,acasse in partial fulfillinPnt of t h r , reijllirPiiients for tile h1.S~.cieeree, Unisersity of Ottawa. I n , . C h e m . Soc.. 78, 202 (195tiI. (2) I . I3 \\-iison a n d I:. ('abib r, T r u n s . Faraday Soc., 56, 1477 (IStiOh i:i, I{. AI. Kropka a n d I i\-c suifaccl txxausc t h(, 1 carlioti of p’-iiicihyl-XC‘h dow not f’roiii hcliaviiig ab ci uorixial substrat i> tric (’elit or is of‘ thc r)-configurat ioii hut docis n.licii the iiicthyl is in ilie I,-cotifiguratioti : ( c ) t h e tctrahcdral cliaractcr of C-2 of flit dioxolaiic, fixes the orimtatioii of tlw uwthyl group in sucli a I\ ay t hat tlio oiicntatioli of ilie cquivalerit iriethyl group of c.iihyliir-t)ouiid XCh iiiay hc t*ctprotl1lcedi i i : i i i appropi-iatcb h i c i eoisonier of I1 : id) the pr(w1iiw ot the two eiliclrcal ouygc.ns i i t I1 should alloii 101 li? drogcri lioiidilig wit Ii t hc i.iisyiiic.’s activt .it cs as rwcltI> as tlie two siiiiilarly placed oxygciis of LICli. Tlie only liiiiitatioii of thci selection of tlic l,:2-d10\0latie ring for these btudies is the impossibility of p1.0iiiotirig binding through n uucleophilic attack (pi’(’iiidamlc riitrogeir) at C-2, a situitt iori with tlic susceptibility of the L\Cl~ additioii complcs forlll:Ltiotl. I€(>\\\ no posit ive c~vidciicc~ a- yvt that ilic carlmiyl group of ACh is charigcd t o i l i v t ( ~ i i x h c c 1 t d c~otthgiu.cziioit i t i the c~otiiplcs I\ ttli (%
(20) t .
B e ~ g i i i ~ ~ awl iiii
1
b l i i . i i o ~ i i ,i b i d
7. 48i ( 1 9 5 1 1
Soveinber, 1964
MECHANISM OF ENZYMATIC COMPLEX FORMATION
771
sents the optimum configuration for maximum inhibitory power among the analog series most closely related to ACh itself. This isomer (VI) possesses an affinity for AChE which is approximately 10 times greater than that of ACh itself; however, the best inhibitor of the series is the cis-trichloromethyl derivative (XI), a compound which is rather distantly related to ACh. Thus, AChE displays a high degree of absolute stereospecificity, but possesses moderate relative stereospecificity arid little structural specificity towards the dioxolane series of inhibitors. A most striking feature is the influence of the C-2 methyl substituent on affinity for the enzyme, a phenomenon which is discussed below. Optimal cholinomimetic activity is associated with the cis-L-( +) derivative (VI), but is relatively independent of affinity for AChE since the trichloromethyl compound (XI) is weakly active as a cholinomimetic. Similarly to AChE, however, the receptor is found to display a high degree of absolute stereospecificity but liniited structural specificity towards this series of stimulants. As would have been expected, the absolute configuration about C-4 of the best inhibitors or stimulants is the same as the configuration of the @-carbon of u-8-methyl-ACh, the stereoisomer acting as a substrate for the enzyme and also as the active cholinomimetic on muscarinic receptors. C-5 of inhibitor VI may therefore not particiTABLE I INHIBITION CONSTANTS FOR AChE OF QUATERNARY INHIBITORSpate nor interfere to any appreciable degree in the O F THE DIOXOLANE AND TETRAHYDROFERAN SERIESAND RELA- binding phenomenon. However, the C-2 niethyl TIVE POTENCIES AS CHOLINOMIMETICS ON THE GUINEA PIGILEUM substituent, when in the configuration fixed by inhibitor VI, has a dramatic influence on both the affinity of the dioxolane ring for AChE and muscarinic potency. Indeed the cholinergic receptor and AChE display identical patterns of stereospecificity towards RelaRelative tive this class of quaternary ions and a similar predilection affiniConfigurpotenKi X for a properly oriented methyl group TThich is strucCornpd. RI RI 105 ,M ation ties cies turally equivalent to that of the ester methyl of Iv H H DL 45.0 100 0.1 v CHI H DL-C~S 7.5 600 100 ACh. These results suggest that the effects of the C-2 VI CHa H L- ( +)-cis 4.1 1100 625 methyl group of VI on affinity may be attributed to a VI1 H CHI 67.0 D-( -)-Cis 67 6 faithful reproduction of the natural steric orientation VI11 DL-trans 200 20 H C Ha 23.0 CHa CHs IX DL 60.0 75 0.5 of the equivalent methyl group of enzyme- or receptorCHa X CHa 85.0 53 0.1 D-( -) bound ACh. That this is probably the case is shown XI CCls H DL-cis 1.1 4100 1 below. Finally, the influence of the C-2 methyl H CCla XI1 610 0.1 DL-trans 7.4 (O>CH,GMe3 r? group of inhibitor VI on affinity and potency is conXI11 DL 49.0 91 ... ditioned to a high degree by the presence of the two 7 . 17 0 265 XIV I H i < - )-CHINhlei DL-CZ'S oxygen atoms of the dioxolane ring, since the tetrahydrofuran analog XIV (deoxy-DL-muscarine) is about Formylcholine .. . Km = 350 12 5 ACh ... K, = 45 100 100 one third as active as the dioxolane V. Whereas the cis-methyl group increases affinity by a factor of 3 in the tetrahydrofuran series, it increases it by a factor where relative cholinomimetic activities reported earlier are included for easy reference. It mas estabof 6 in the dioxolane series. A perfect fit of the lished that all of these compounds behave as commethyl group on AChE is thus conditioned by interpetitive inhibitors of the enzyme. actions with two oxygen atoms as in the case for ACh itself. On the basis of these results, the conclusion emerges Results and Discussion that both AChE and the muscarinic receptor display A. Stereospecificity of AChE and Muscarinic Chosimilar absolute and relative stereospecificities and that linergic Receptors.-Table I shows that AChE displays the configuration of inhibitor VI must bear some imboth relative and absolute stereospecificity towards portant relationship to the conformation of enzymethe 1,3-dioxoIane series of inhibitors. This stereoand receptor-bound ACh. specificity attains its highest degree towards one optical
the enzyme and, in fact, recent deuterium isotope effect studiesg with AChE would seem to favor the view that the carbonyl of h C h remains essentially trigonal in the addition complex with the enzyme. I n spite of some residual uncertainty on this point, the fact remains that regardless of the resistance of C-2 of the dioxolane ring towards nucleophilic attack, the orientation of the methyl group a t that position ought to bear some relationship to the orientation of the equivalent niethyl group in enzyme-bound ACh. I n order that the role of the oxygen atom in position 1 of the dioxolane may be evaluated in relation to complex formation, some tetrahydrofuran analogs of the l13-dioxolanes were tested as inhibitors of AChE. Finally, the early reports of Fourneau, et U Z . , ~ on ~ the high cholinomimetic activity of quaternary salts of the 1,3-dioxolane series, also provided an important incentive for the initiation of the present studies especially because of the possibility of establishing novel relationships between the enzyme and its receptor counterpart. Using erythrocyte AChE, the affinity constants of a series of 1,3-dioxolane quaternary inhibitors were measured. The synthesis and proofs of configuration of their various isomers and analogs have been reported previously.1a~22The results are assemble in Table I
form of cis-2-methyl-4-trimethylammoniummethyl-1,3- Interpretation of Structure-Activity Relationships.Table I shows that the effect of the C-2 substituents dioxolane iodide (VI), the L-(+) isomer of which repreof the dioxolane series on affinity is erratic. If the cis (21) E. Fourneau, D.Bovet, F. Bovet, and G. Montezin, Bull. soc. chim. configuration is optimal for binding, then a lower affinity bid., 26, 516 (1944). of the trans isomer (1-111) than for the parent desmethyl (22) D. J. Triggle and B. Belleau, Can. J . Chem., 40, 1201 (1962).
( 2 3 ) .:I .I. Colin mid .J. T. Edsoll. "l'roteins, Aimnoacids I L I Ll'wLide*,' ~ Ileinliold Publishing Corp., New T o r k , N.P., 19.13, Chapter 9.
MECHANISM OF ENZYMATIC COMPLEX FORMATION
November, 1964
Waals forces.17 It is significant that only inhibitor VI should allow the application of these forces to the C-2 methyl group. This observation confirms the conclusion that the steric orientation of that group must correspond closely to a natural and special predilection of the enzyme surface for the geometry of inhibitor VI. Since a lock-and-key type of fit applies to the C-2 methyl of VI and that such fits are generally believed to be a privilege of natural substrates, the conclusioii is also confirmed that inhibitor VI may reproduce closely the preferred conformation assumed by ACh when bound onto the enzyme. If this were the case, it would be expected that the ester methyl group of ACh must also allow the application of distancespecific van der Waals forces in the Xichaelis complex with the enzyme. This could be verified through a comparison of the affinity constants of formylcholine and ,4Ch. After appropriate corrections for the spontaneous hydrolysis of formylcholine, a value of 1.2 kcal. for the free-energy contribution to binding of the acetyl methyl group mas obtained. The methyl groups of both ACh and inhibitor VI therefore lead to the formation of lock-and-key fits with the esteratic portion of the enzyme. We conclude on that basis that the conformation of enzyme-bound ACh most probably corresponds to the geoinetry of the unique inhibitor VI. It now becomes possible to rationalize structureactivity relationships at the level of the muscarinic cholinergic receptor. I n agreement with the finding that VI interacts with AChE in an ACh-like manner, it mould be expected that the same should apply to other natural bioreceptors that are specific for -4Ch. The fact that 1'1 largely surpasses ACh in potency a t the cholinergic receptor level substantiates these expectations. However, all those inhibitors whose affinities are conditioned by nonspecific accommodative perturbations do not interact in an ACh-like maimer with the receptor and therefore cannot act as efficient inducers of the specific receptor perturbation that is required for the initiatioii of a physiological stimulus. Moreover, those molecules whose affinities for the protein surface do not benefit from the operation of van der Waals attractions or hydrophobic forces should be less effective as stimulants. This is the case, for instance, for the unsuhstituted dioxolane I V and its 2,2-dimethyl analog (IX). The low affinity and low potency of the latter is ascribable to the unfavorable AF, for a gemdimethyl group (Table 111),a phenomeNvr X Compd.
10 -8
TABLE111" Y 4 x 10 -3
log ivA/AvW
Apt, cal.
30.8 4.9 -0.70 7.35 2 7 -0.44 350 0.13 1.i5 0 93 2280 SI1 0 63 6 65 1 02 2340 a -Vw = solubility in water a t 23" (mole fraction), -V.k = solubility in ethanol a t 23" (mole fraction), AFt = free energy of transfer for the 2-substituent(a) = free energy of transfer of substituted compound - free energy of transfer for compound IV.
I \. IX XI
non related to the fact that steric hindrarice prevents the nonpolar chains of the protein froin interacting optimally with all of the CH bonds of the methyl groups, As a consequence, fewer new bonds are created between the niethyl groups and the enzyme or receptor than are
773
destroyed by them. Hence the unfavorable energy of binding for the gem-dimethyl drug. This interpretation is in line with the known detrimental effects of methyl branching on the van der Waals interactions between fatty acid chains. l 7 The low muscarinic potency but high affinity for AChE of the trichloromethyl drugs (XI and XII) is a reflection of the operation of AFt which is coriditioried by nonspecific acconiinodative perturbations of the nonpolar chains of the protein. Such perturbations bear little relationship to the specific conformational changes induced by ACh or the inhibitor VI and account for the inactivity as stimulants of the trichloroniethyl compounds. To suiiiinarize, it can nom be recognized that the binding of quateriiary ions onto AChE or its receptor counterpart can have quite divergent effects on the coriforniation of the protein; either little disruption of pre-existing bonds between the nonpolar chains occurs or more or less profound perturbations of the hydrophobic network take place when nonpolar substituents are brought into contact with the enzyme's active surface. It is probable that this is what forms the basis of structure-activity relationships at the receptor level, stimulant activity being a characteristic of those ions which do iiot induce undesirable accommodative perturbations of the protein's binding surface. These considerations form the basis of a molecular theory of drug action which is proposed iii an accompanying paper.24 In a coinparison of our results in the dioxolarie series of AChE inhibitors with those of Freiss, ef al.," in the muscarine series of isomers, the affinity of m-deoxymuscarine (XIV) for AChE is higher than that of DL-iiiuscarine itself. This suggests that the hydroxyl function of the latter hinders rather than promotes complex formation. Of interest was the observation that similarly to the dioxolane series, a c i s arrangement of the substituents on the ring of DL-muscarine is optimal for high affinity for the enzyine, although this requirement is iiot as critical as in the dioxolane series. Curiously enough, in the DL-muscarine series, the trans configuration appears to be more favorable to binding iiotcd above, the effect than the c i s arraiigenient. of the cis-inet hyl group on affinity of dcoxyiiiuscarine for AChE does not suggest any positive contribution from van der Waals forces to binding since hydrophobic forces alone account for the effect of the methyl group on affinity (about 700 cal./CH3). I t is therefore likely that in both the muscarine and muscaroiie series, affinity for AChE may be largely controlled by hydrophobic forces, a possibility which makes coniparisons with the dioxolane isonicrs rat her meaningless since in that series, the cis isomer (VI) allom the application of van der Waals attractions to the 2methyl substituent, The probable participation of AFt alone in the muscarine series can only serve to mask the inherent stereospecificity of the erizyiiie. Friess, et uZ.,I1 also reported that acetylation of the hydroxyl function of the muscarine isomers uniforinly increases affinity for the enzyme by a factor of 2-3. The magnitude of the effect (the binding energies increasing by about 500 to 700 cal.) strongly suggests that hydrophobic forces are chiefly responsible for the increased affinities of the acetylmuscarines and that no (24) B. Belleau, J . M e d . Chem., 7, 776 (1964).
specific bonds, other thati "dissolution" info t l i ~i i o i i polar iictwork, are formed betivccii t h x ac aiid the eiizyriie's active sites. Thc fact that deoxyriiuscariiie (XIV) possesses a coiiiparat)lv affiiiity strongly supports this view. Tlic part ial iiiverGoii of the optical specificity of receptors towards ilie 111113('arotie enaiitioiiiers has been int erprc3ted i i i aiiotlirr paper.Ia Ho~vevci~, thr receptor and XC'hlC display siiiiilar absolute aiid rchtive stcrcospcvificit i c 3 i 1 o u aid. 11oth the iiiuscariiic mid dioxolaiicl h ( w \ s of isoiii(>i~* Thc t)caring of these coilsiderat ioii- o i i t h r iactorh iiitcweiiitig i i i coniplex foriiiatioii \vit h .l("hI: is t akeri advantage of in the elaboratioii of tliv ~ i i a c r o ~ i i o l c c u l ~ ~ r perturbation theory of drug actioti.24 Role of the Quaternary Nitrogen of Drugs in Complex Formation.-The inr-estigatious of Wilson6 o n the role of the S-methyl groups of ACh on affinity for AChE have revealed that each of the first t\vo methyl gYOUpS iiicreascs affiriitj7 for the eiieyme by a factor of i ,n-hereas the third methyl group is without effect on biiidiiig, although its presence causcs a marked increase in thc ciitropy of the complex. This facto1 of i oil the affinity correspoiids to a frw-eiicigv contributioii to hindirig of 1.3 kcal.jniethyl group, a value which was tentatively ratioiialized by Renihnrd' and also n'ilsoii" 0'1 the basis of the operatioii of vaii der Waals attract ionb. Hon ever, usiiig Salein's ralculatioiis as a basis," the coritributioiis of vaii der Kauls forces to tlic biiiditig of a single inethyl group caiitiot PX cl 600 cal., lcaviiig ail excess of 700 cal. uriaccouiitc or If Oil(' talws into accouiit tlie fact that hydiophohic f'oi~cesaloiie should contribute about 730 cal. of driviiig force for t)iiidiiig, oiic arrives at a iiia1\itii~iii1total cticlrgy of ititoraction of 1.3x BcaI./iiiethyl g l ~ ) u p a, , loiig a': the .lChE surface is largely hydrophobic i i i charactc~r. I ~ h d c n c ef o i - this view has I t w i i discu>sed a h o \ ~ ~Tlic . for(w cotit ributiiig to tlw hiiiding of the S-iiiethyl groups arid t h e w t e r iiietliyl giuup of A4Cli or tlw 2iiirthyl group of inhibitor TI arc3 of aii idciit Iiiagnitudc. The operation of liydropliohic foi in the adsorptioii of the X-iii(.tIiyl groups cleal'ly clstahlishes that the enviroiinicvit of t Iic aiiioiiic sitc of ACliE is a140 highly liydropliohic. This allo\vs thcs prcdictioii that groups bulkier t f i a i i S-iiietliyl would also induce noiispecific accoiiiiiiodative perturbatioiiof tlie ncltn-orli of nonpolar chaiiit at t tic periphery of the anionic hinding sitc. It folloni that bulky substituents on the iiitrogeii of qiiateriiary halts n ill cause dcformatioii of AChK o r tlic receptor protviti couiiterpart. 111 this latter case, tlic effect ivoiild \ w reflected in a reduction of stiiiiulatiiig activity iipoti coriiplex formation. The absence of an effect on hidiiig iii the case of the third inethyl group and the draiiiatic effect 011 the mitropp of the coiiiplex produced by the saiiie group are most interesting. Clearly, this iiiethyl group has a profound effect on the stnictiw of the proteiii and a large increase. in the flexibility of sollie parts of the proteiii must occur as a result of its presence. In oui' opinion, this effect can best bc rationalized as follows. When in the aqueous soh-eiit the T-methyl group exert 5 hydrophobic iiiteractions aiid thus repels \vat el inolecule.. . The free energy of this iiitcractiori ainouiit.: to a l m t 730 cal. Since the samp iiiethyl group docs riot coritiibutc positively to affiiiity wlirri on thc1 cwzyLii(1 ~
MECHANISM OF ENZYMATIC COMPLEX FORMATION
November, 1964
gradual decrease in affinity constants for AChE produced by separation of the positive nitrogen from the phenyl ring through the insertion of methylene groups is probably the simple reflection of a gradual decrease in the ability of the benzene ring to form r-bonds with the enzyme's active sites. It is likely that the distance between the positive nitrogen and the r-electron cloud is critical for optimal interaction with the enzyme. (d) Should the substituent carbon atoms of quaternary nitrogen share the positive charge, the Y-C bonds would have increased polarizability and hence the bond length should be different from that in the corresponding tertiary amines. However, this is not the case; both types of amine derivatives have nearly identical K-C bond lengths.29 It is most likely that the results of Thoinas are relevant to the operation of hydrophobic forces arid n-interactions with the enzyme's active sites.
Experimental30 Materials.-Except for compounds X I and XII, the 1,3dioxolane inhibitors (Table I ) have been described previously.'a,22 Compounds XI11 and XIV were prepared according to the literature.31 ~~-cts-2-Trichloromethyl-4trimethylammoniummethyl1,3-dioxolane Iodide (XI).-Five grams of ~~-cis-2-trichloromethyl-4-tosyloxymethyl-l,3-dioxolane**was treated with excess anhydrous dimethylamine in 50 nil. of dimethyl sulfoxide. After standing overnight, followed by heating for 1 hr. a t loo", the solvent was evaporated in vacuo, the residue was mixed with saturated aqueous potassium carbonate, and the amine was extracted with ether. The ether was dried and evaporated, and the residue was distilled in vacuo to give n~-cis-2-trichloromethyl-4-dirnethylaminomethyl-l,t-dioxolane,b.p. 115' (10 mm.) (85'7, yield). -4nal. Calcd. for C,H12C13N02: C, 33.80; H, 4.82. Found: C, 33.64; H, 4.96. The methiodide X I was obtained in the usual manner and was recrystallized for ethanol, m.p. 243-245'. Anal. Calcd. for CsH1,ClJNOZ: C, 24.58; H, 3.84. Found: C, 24.41; H, 3.70. ~~-trans-2-Trichloromethyl-4-trimethylammoniumme thyl1,3-dioxolane Iodide (XII).-The previous:y described DLtr~ns-2-trichloromethyl-4-tosyloxymethyl-l,3-dioxolane~~ was carried through the same procedures described above in the case of the preparation of XI. In this manner, the trans-2trichloromethyl-4-dimethglaminomethyl-1,3-dioxolane, b.p. 110115' (12 mm.), was obtained in 757, yield. (29) Tables of Interatomic Distances and Configuration in Molecules and Ions, Special Publication KO.11, T h e Chemical Society (London), Burlington House, London, 1958, p p . M156, h2175, M115. (30) Microanalyses b y Midwest Microlab. Indianapolis, Ind. Melting points were determined on a Kofler hot stage and are corrected. (31) A. C. Cope and E. E. Schweizer. J . A m . Chem. Soc., 81, 4577 (1959).
775
Anal. Calcd. for CTH&laN02: C, 33.80; H, 4.82. Found:
C, 33.95; H, 4.66. The methiodide XI1 was recrystallized from ethanol, m.p. 195'. Anal. Calcd. for CsHlsC131NOz:C, 24.58: H , 3.84. Found: C, 24.68; H, 3.78. Formylcholine Bromide.-A mixture of acetic anhvdride (10 ml. j in excess formic acid (98%) was prepared and 15 g. of 2-bromoethanol was added. After several hours a t room temperature, the 2-bromoethyl formate was isolated in the usual manner. The ester thus obtained was treated with excess anhydrous trimethylamine in benzene a t 100" in a pressure bottle. After 10 hr., the solid was collected and recrystallized from absolute ethanol, m.p. 147-149'. Anal. Calcd. for C6HI4BrNO2:Br, 37.73; sapon. equiv., 212. Found: Br, 37.85; sapon. equiv., 205. Methods.-Bovine erythrocyte acetylcholinesterase from a commercial source (Kutritional Biochemicals Corp.) was used. The rates of acetylcholine brorn.de hydrolysis were measured with an automatic titrator using a pH stat equipped with a recorder (Copenhagen Radiometer, Ole Dich recorder and syringe attachment). The reaction medium consisted of glass-distilled water, 0.15 If in NaCl and 0.015 -14. in MgClz; the temperature was kept a t 25 + 0.1' and the pH a t 7.4. A COz-free nitrogen atmosphere v a s maintained throughout. The value of K , for ACh was consistently 4.5 f 0.1 x 10-4 M in agreement with published data. The determination of inhibition constants was accomplished in the usual manner using four different inhibitor concentrations and 4 X 10-3 it1 ACh. The Ki values were calculated using the relationship Ki = 0.09 X Cj0 where C ~ repreO sents the concentration of inhibitor required to produce 50% inhibition of the enzyme; this relationship is derived from the following equation.32
Using an inhibitor concentration to produce 25% inhibition, initial velocities for four ACh concentrations inferior to the K , were determined and conventional Lineweaver-Burk plots33 were constructed. All curves thus obtained had a common extrapolated intercept, thus establishing that all the inhibitors of Table I act competitively. Three separate determinations of K , were made. Solubility Determinations.-The procedure advocated by Cohn and Edsall23 was applied. Excess solute IX, XI, and XI1 was added to distilled water and to absolute ethanol. The mixtures were allowed to equilibrate a t room temperature (23") by shaking for 20 hr. and then centrifuged. An aliquot of the clear supernatants was drawn and the respective densities of the solutions measured. The concentrations thus obtained are given in Table 111. The calculated values for the free energy of transfers can only be approximate since the assumption is implicit that the activity coefficients are close to unity. However, due to the low solubilities of IX, XI, and XII, the approximation should be a good one. On the other hand, the higher solubility of IV probably introduces larger deviations from ideal behavior. (32) J. A. Cohen and R. A. Oosterbaan, "Handbuch der Experimentellen Pharmakologie," Val. XV, G. B. Koelle, Sub-Ed., Springer-Verlag, Berlin, 1963, Chapter 7. ( 3 3 ) H. Lineweaver a n d D. Burk. J. A m . Chem. Soc.. 6 6 , 658 (1934).
