March 20. 1953
ORGANOPHOSPHORUS COMPOUNDS WITH ANTICHOLINESTERASE ACTIVITY
1379
taining group ( k . the , sulfonyl group), as in the benzoylpropionate ion, is somewhat more effective change from hippurate ion to benzenesulfonylgly- in combining with anti-BzP antibodies than are the cinate, causes a very large decrease in the strength phenylvalerate and the phenylpropionate (hydroof combination. Tbis large effect may be entirely cinnamate) ions which contain one methylene group steric and due to the Ixger space requirement of more or less than the butyrate. The importance of the van der Waals interaction two oxygens. In the absence of the carbonyl group, there seems of the benzene ring is emphasized by a comparison to be some correlation between the combining power of the combination of anti-BzP antibody with of the hapten and the distance between the benzene valerate and with phenylbutyrate, A F r e l for the group and the carboxylate ion. Phenylbutyrate combination with the aliphatic compound is a t ion, in which the benzene and carboxylate have the least 700 cal. greater. same distance between them as in the homologous NEWYORK,N. Y.
[CONTRIBUTIOX FROM VENEREALDISEASEEXPERIMENTAL LABORATORY, U. S.PUBLIC HEALTH SERVICE, SCHOOL OF PUBLIC HEALTH, UNIVERSITY O F NORTH CAROLINA]
The Preparation of Some Organophosphorus Compounds Possessing Anticholinesterase Activity' BY LEOND. FREEDMAN, HENRYTAUBER, G. 0. DOAKBND HAROLD J. MACNUSON RECEIVED OCTOBER 20, 1952 Fifty phosphonic and phosphinic acids previously prepared in this Laboratory have been tested for anticholinesterase activity. The enzyme used in the present study was human plasma cholinesterase. Fourteen of the compounds were found to be active at a concentration of 0.003 M. The results obtained suggested the synthesis of other organophosphorus compounds. Since several disubstituted phosphonic and phosphinic acids were required, the diazo synthesis was extended to the preparation of such compounds. The reaction conditions were similar to those described previously. Reasonable yields were obtained in all cases. Several esters of phosphonic and phosphinic acids were also prepared. Most of the esters were more active than the free acids. TABLEI Recent communications2 from this Laboratory have described the preparation of 50 phosphonic ANTICHOLI\ ESrERASE ACTIVITYOF PREVIOUSLY DESCRIBED and phosphinic acids. Although these compounds COMPOUNDS ISQ~ were prepared primarily for testing against a variety Compound mole/l. of microorganisms, the usefulness of certain organic (o - B ~ C ~ H ~ ) C G H B P O & 6 X phosphorus compounds as anticholinesterase ( o-BrcsHa)nPO?H 1 x 10-4 agents4 prompted us to investigate the value of (o-C1C+j&)&'02H 5 x 10-4 aromatic phosphonic and phosphinic acids as .(m-BrCsH&POzH 2 x 10-3 enzyme inhibitors. (V Z - C ~ C ~ H ~ ) ~ P O ~ H 2 x 10-3 The enzyme used in the present study was (m-ClCsH4)CsHsPOaH 3 x 10-3 human plasma cholinesterase prepared from Cohn (m-BrCGHa)(m-CH3h'HC6H4)POiH 4 x 10-3 Fraction IV-6. It possessed an activity 80 times ( m-CHaXHCsHp)zPOzH 4 x 10-3 that of whole plasma protein. All the phosphorus (o-B~C~H~)C~H~POLH 4 x 10-3 compounds were tested first a t a concentration of o-B~C~H~PO~H~ 4 x 10-3 0.003 M ; the majority caused little or no inhibition o-CICGH~PO~HZ 4 x 10-3 and were not further studied. A few, however, m-BrCcH4P03H~ 5 x 10-3 possessed significant anticholinesterase activity a t o-NHJC~H~POJH~ 8X this level and were further tested to determine the m-CH3(C H ~ ~ N H C ~ H I P O ~ H ~ 9 x 10-3 concentration required for 50% inhibition (160). a The 150 values in this and succeeding tables were obThe results obtained are shown in Table I. It is tained from graphs in which % inhibition was plotted seen that 11 of the 14 active compounds were against the logarithm of the molar' concentration of the halogen derivatives, The two most active phos- compound. phonic acids and the three most active phosphinic acids contained a halogen atom in ortho position. ticular interest to prepare o-fluoro and o-iodo The m-halo derivatives were less active; and all the substituted phosphonic and phosphinic acids for comparison with the corresponding o-bromo and p-substituted compounds were inactive. These results suggested the synthesis of other o-chloro derivatives. I t seemed advisable also to other compounds with ortho substituents organophosphorus compounds. It was of par- prepare since every o-substituted compound we had tested (1) Presented before the Organic Division of the American Chemical possessed significant anticholinesterase activity. Society in Atlantic City, N. J., September, 1952. Several disubstituted phosphonic and phosphinic (2) (a) G . 0. Doak and L. D. Freedman, THIS JOURNAL, 73, 5658 acids were synthesized to determine the effect of a (1951); (b) G . 0. Doak and L. D. Freedman, ihid., 74, 753 (1952); (c) L. D . Freedman and G . 0. Doak, ihid., 74, 2884 (1952); (d) second ring substituent on the activity of the G . 0. Doak and L. D . Freedman, ibid.. 76, 683 (1953). halogen-substituted acids. (3) Some of these results have been reported, J. D . Thayer, H. J. Because the inhibition studies were made a t pH Magnuson and &I. S. Gravatt, Antihiofics and Chemofherapy, in press. 7 , the phosphinic acids were present entirely as (4) Cf.G. B . Koelle and A. Gilman, Pharmacol. Reu., 1, 166 (1948).
L. D. FREEDMAN, H. TAUHER, C:. 0.I ~ A AND K H. J. MAGNUSW
Vol. 75
TABLE I1 DIARYLPHOSPHINIC ACIDS
ARYLPHOSPHOSIC AND Yield,"
hl.p.,b OC.
70
Phosphorus, Yo Found
Cdlcd.
NPutral equivalentc Calcd. Found
I&O, mole11
4 x 10-4 10.71 142.0 143.0 67.4 2! 4 x 10-3 14.94 4 x 10-3 91.1 95.3 16.02 6 X IW3 122.5 123.4 12.27 7 x 10-2 150.0 149.0 10.12 1 x 10-2 103.8 104.0 14.66 >1 x 1 0 - 2 88.0 87.7 16.99 >1 x 10 ? 125.5 121.3 12.00 > l x 10 118.8 118.8 12.91 >1 x I o - ' 125,ii 125.0 12.03 1 x 10 34s. 1 339 0 9.01 1 x 10 3356, I1 359.5 8.54 I y 1') 1 2778.2 27(j ;I 10.87 1 od 2*;+, ? :!.TI 'i ' Y 10 3 11.90 11' 7 IO-l ; :i')) 7 7.58 > 1 >< 10 4(jfj.(l Jfi'.; {I 12 "1 >:j(Io 6 . (i.5 0.48 > l X 10377 1 3i2.7 2 4 , 0' Lkc. 270-275 8.22 7.99 Reactiol co:itiitio:is w e r e choseil which a The yield data, in most cases, are the results ohtained in a single experinwit. were expected to give a reasonable yield of both phosphonic and phosphinic acids; sf. ref. 2d. Larger yields of either acid The Melting points were taken as previously described ; c j . ref. 2d. could probably be obtained under other conditions, indicator used for the phosphonic acids was thymolphthalein; the indicator used for the phosphinic acids was phenolphthalein. This yield was cbtained by the use of ethyl acetate as the solvent, cuprous bromide as the catalyst, and phosphorus tribromide as the source of phosphorus. e Previously prepared by A. Michaelis, Ann., 293,193 (1896); m.p. 172'. f Calcd. : C, 41.60; H , 3.19. Found: C, 41.90; H, 3.39. 0 Potentiometric titration ( c f . ref. 6 ) of this compound showed the second dissociation c o n r a n t of the phosphono group was too small ( p K , 9.17) t o permit the use of thymolphthalein as the indicaPreviously prepared by A . Burger, personal communication. This compound was isolated and purified by procetor. Previously prepared by G. dure A as previously described; cf, ref. 2a. i 3l.p. varied slightly with the rate of heating. Kosolapoff, T I I I S J ~ U R X69, A I ,2020 , (1937): m.p. 191-197'. Calcd.: S, 4.67; HzO, 6.00. Found: K, 4.69; loss at 100 , 6.20. ln This yield mas cbtained by t h e use of isopropyl acetate as the solvent, cuprous bromide as the catalyst, and phosphorus tribrornide as the source of phosphorus. '' Recrystallized from 3 N hydrochloric acid. Calcd.: N, 6.75. Found: N, 6.80. P M.p. varied greatly with the rate of heating. c This compound was isolated and purified by procedure B as previously described; rf. ref. 2a. This yield was obtained by the use of ethyl acetate as the solvent, cuprous bromide as thc catalyst, arid phosphorus trichloride 3 s the source of phosphorus. This compound was dried in vacuo a t 100' t o remove solvent of crystallization. Calcd.: N, 5.90. Found: S, 5.77. Calcd.: N, 5.58; H20, 7.17. Found: N, 5.43; loss PreviRecrystallized from aqueous alcohol. L-0 phosphinic acid was obtained from this reaction. at 100°, 7.23. Y Recrystallized from 6 iV hydrochloric acid. ously reported by G . 11.Kosolapoff, ihid., 71, 369 (19 Calcd.: S,6.01. Found: S ,%5.99."" Calcd.: S ,7. 22d 37d 20d 32d 15" 86 41' 2;d 20d, 32' 31'8" 28 13d
219-222 175.5-179 202-203' 193-196 Ilec. >230 >20OP 146- 149 199-203 221.5-224 226.5-227.5 206-307.5 23"-23 3 2:34-2.'38 220.-222 2iis -271
10.91 15.33 16.47 12.64 10.33 14.93 17.59 12.34 13.01 15.31 9.00 8.70 11. 1 3 1 2 . I!) 7 ,(i7
'.
5.
u(
singly charged a t ~ i o n s and , ~ the phosphonic *ids hydroxide, iodide ion can be detected in the resultas mixtures of tlic ions XrP03H- and XrP0.1=.6 ing solution and the presence of a phenol can be Since the receptors of the enzyme might be more demonstrated by the ferric chloride test. For this accessible to a neutral molecule than to a nega- reason the o-iodo derivatives could not be purified tively charged one,' we prepared esters of sevcral of in the usual way, one step of which requires the our active compounds. precipitation of the acid from alkaline solution. Table I1 lists the new phosphonic and plies- The methods used for purifying the o-iodo acids phinic acids prepared. n'ith the exception of 2- are given in the Experimental section. Also shown in Table I1 are the concentrations chloro-kmiinohenzenephosphonic acid, which was prepared by c:it:ilytic reduction of the correspond- necessary for the 50% inhibition of the plasma ing nitro derivative, ?' the compounds listed were enzyme employed. Although most of the comsynthesized from the appropriate diazonium fluo- pounds listed possessed some activity, the results borates.?" In general, the disubstituted phos- with the disubstituted phosphonic acids were disphonic and phosphinic acids were obtained by the appointing. In every compound tested a second same procedures used for the monosubstituted substituent in addition to the ortho group markedly reduces the inhibitory effect. The considerable compounds. I n the presence of copper salts, both o-iodoben- activity of the o-:net'ioxy and o-carboxy derivatives zenephosphonic acid and (0-iodophenyl 1 -phenvl- demonstrates that a halogcn i; nqt essentisl for phosphinic acid were unstable a t high pII's. The appreciable anticholinesterase activity. Table I11 shows the esters prepared and tested. instability is due, presumably, to replacement of the iodine by a hydroxyl group. T ~ L I ifS ,the The activity of tetraethyl pyrophosphate is given crude iodo compound is dissolved in 10'6 sodium for comparison. As we had hoped, most of the esters were more active than the free acids. The (3) H. H. J a n e , unpublished results. activity of o-bromobenzenephosphonic acid, in (B) H. H. JaffC, L. D. Freedman and C . 0. Drrxk. paper presented hefore t h e American Chemical Society. Atlantic C i t y . N J , September, particular, was greatly enhanced by esterification. 1032. dUdri(ige and Davison8 have pointed out that ( 7 ) A. Hood a n d W. Lange, THISJow~xar.,7 9 , 4956 (1!150), for there are enzymes in mammalian tissues which can example, have reported t h a t t h e acidic half-esters of fluorophosphoric acid a r e much less active t h a n t h e corresprinding neutral ebters.
( 8 ) \V S A!drirlgc a n 1 A S . I)avisr,n, Binchem. J , 61,62 tlrJ32>:.
ORGANOPHOSPHORUS COMPOUNDS WITH ANTICHOLINESTERASE ACTIVITY
March 20, 1953
1381
TABLE I11
ESTERS Compound
Yield,"
3' 5
OC.
B.P..
Mm.
Carbon, % Calcd. Found
Hydrogen, % Calcd. Found
TEPP~ 2.0 53.11 53.22 4.76 (o-BrC6H1)C6H5P0zCH(CH1)~ 50 183-191 (~ - B ~ C ~ H I ) Z P O Z C Z H ; ~ ~ 38 186-194 0.1 41.61 41.48 3.24 51.44 4.34 .1 51.71 ( O - B ~ C ~ H ~ ) C ~ H ~ P ~ ~ C ~74H : 157-165 o - B ~ C ~ H I POCZH~'? O( 75 115-121 .1 40.97 40.91 4.81 ( O - B ~ C ~ H ~ ) C ~ H ~ P O ? C H ~ 55 160-168 .1 50.18 49.47 3.89 a These figures must be considered only approximate, since they are the results of a single run. phate, Monsanto, 40% pure. Calcd.: P, 7.67. Found: P, 7.66. M.p. 115-118'.
hydrolyze certain phosphate esters. Therefore, it was important to learn whether the enzyme used in our tests could split esters of phosphonic or phosphinic acids. The Warburg manometric technique was used to determine the rate of hydrolysis of 0.001 X ethyl bis- (o-bromophenyl)-phosphinate in sodium bicarbonate buffer a t pH 7.0.8 I n one hour there was no evidence of hydrolysis; and eve? after 18 hours the amount of carbon dioxide evolved indicated that less than 4% of the ester had hydrolyzed. It seems safe to conclude, therefore, that this ester does not hydrolyze under the experimental conditions used for determining anticholinesterase activity.
Experimental o-Iodobenzenephosphonic Acid.-This compound was prepared from o-iodobenzenediazonium fluoborate by the general method described previously.2a After the reaction mixture was steam distilled, the residual liquid was filtered. A small amount of phosphinic acid remained on the filter but could not be readily purified. The filtrate was evaporated on the steam-bath to incipient crystallization, cooled and filtered. The resulting crude phosphonic acid was recrystallized from dilute hydrochloric acid. o-Carboxybenzenephosphonic Acid.-This compound was prepared from o-carbomethoxybenzenediazonium fluoborate. N o phosphinic acid was obtained. The phosphonic acid was isolated and purified by procedure A as previously described.Zd 2,5-Dichlorobenzenephosphonic Acid and Bis-(Z,S-dichloropheay1)-phosphinic Acid.-These compounds were separated and purified by the procedure used for the pbromo derivatives. A previous paper28 from this Laboratory described an unidentified phosphonic acid isolated from the reaction of o-nitrobenzenediazonium fluoborate and phosphorus trichloride. This phosphonic acid has now been shown to be 2,5-dichlorobenzenephosphonicacid by analysis and mixed m.p. with an authentic sample. The replacement of a nitro group in a diazonium salt by a chlorine atom has been previously observeda; the origin of the second chlorine atom is more difficult to understand. (o-Iodopheny1)-phenylphosphinic Acid.-o-Iodobenzenediazonium fluoborate (31.8 g., 0.1 mole) was suspended in 125 ml. of ethyl acetate and 13.5 ml. of phenyldichlorophosphineI0 and 0.8 g. of cuprous bromide" were'added. The apparatus used has been described.% A short lag period (about 30 minutes) was followed by a vigorous reaction. When the evolution of gases was complete, 25 ml. of water was added and the mixture was steam distilled until approximately 500 ml. of distillate had been collected. The residual liquid in the flask was transferred t o a beaker and cooled. The precipitated phosphinic acid was removed by filtration and recrystallized several times from 95% alcohol. The Esters.-The conditions used for the preparation of the esters were similar to those described below for diethyl (9) Cf. J. F. Bunnett and R. E. Zahler, Chcm. Rrus., 49, 273 (1951). (10) Kindly furnished by the Victor Chemical Works. (11) The amount of catalyst used in this preparation was only 40% of the amount specified in previous papers.* We have found that such a decrease moderates the violence of the reaction without lowering the yield.
180,
mole/l.
5 x 10-9 5.03 1x 3.42 3 X 4.54 1X 4.96 1X 3.96 8 X loT5 * Tetraethyl pyrophos-
o-bromobenzenephosphate. All the esters in Table 111 are liquids a t room temperature except for ethyl bis-(o-bromophenyl)-phosphinate. Diethyl o-Bromobenzenephosphonate.-o-Bromobenzenephosphonic acid (22.8 g.) and 44.0 g. (about a 10% molar excess) of phosphorus pentachloride were heated under reflux for 30 minutes. After the phosphorus oxychloride and excess phosphorus pentachloride had been removed in vacuo (water-pump), the phosphonyl chloride was dissolved in 50 ml. of dry carbon tetrachloride. To this solution was added 56 ml. of absolute ethanol. The mixture was refluxed for one hour, and the carbon tetrachloride and excess alcohol were stripped off a t the water-pump. Distillation of the residue at 0.1 mm. gave the pure ester. Inhibition Studies. Enzyme Solution.-0.9 mg. of the enzymel2 was dissolved in 1.0 ml. of 0.6% sodium chloride solution. A trace of insoluble material was removed by filtration. Substrate Solution.-O.O16 M acetylcholine bromide in 0.1 M phosphate buffer of pH 7.0. Inhibitor Solutions.-The free acids were dissolved in dilute sodium hydroxide solution and adjusted to PH 7.0 with dilute hydrochloric acid. The esters were dissolved in the corresponding alcohol and then diluted with water. The final concentration of alcohol was less than 4%. I t was ascertained that this amount of alcohol had no effect on the activity of the enzyme. Procedure.-1.0 ml. of 0.1 Mphosphate buffer at PH 7.0 was mixed with 0.5 ml. of enzyme solution. The volume was made up to 3.0 ml. with inhibitor solution and distilled water. This mixture was kept a t 23' for 20 minutes; then 1.0 ml. of the acetylcholine bromide solution was added. After an additional 30 minutes an aliquot of the mixture was analyzed for residual substrate by the method of Hestrin13 employing the Klett-Summerson photoelectric colorimeter. The activity of the enzyme in the absence of the inhibitor was always determined simultaneously. In the case of o-aminobenzenephosphonic acid, the colorimetric method failed because the compound reacted with the alkaline hydroxylamine reagent. Therefore, the IVarburg manometric method was used for measuring the hydrolysis of acetylcholine by the enzyme. The inhibitor, dissolved in 3.6 ml. of 0.01 M sodium bicarbonate, W d S placed in the main compartment of the XVarb jrg flask. After this solution was ga'ssed with 95% nitrogen-,, /o carbon dioxide, 0.2 ml. of enzyme solution (which contained 0.28 mg. of enzyme in 0.01 M sodium bicarbonate) was tipped into the main compartment. The reaction WAS allowed to proceed for 20 minutes a t 23' and then 0.2 ml. of 0.08 M acetylcholine bromide in 0.01 sodium bicarbonate was added. Thirty minutes later the carbon dioxide production was compared with that in the control vessel which contained no inhibitor.
Acknowledgments.-The authors wish to thank Miss Sadie Herndon for performing the analyses necessary for this research and Mrs. Carolina Mockler and Mr. Edward L. Petit for skilled technical assistance. Appreciation is due also to Dr. Arthur Roe and Dr. John Montgomery for a gift of o-fluoroaniline. CHAPELHILL, N. C. (12) Prepared from Cohn Fraction IV-6 by Cutter Laboratorier. (13) S. Hestrin, J . B i d . Chem., 180, 249 (1949).
