May 20, 1960
COMMUNICATIONS TO THE EDITOR
2643
ON T H E MECHANISM OF THE COBAMIDE precisely to the experimentally obtained trace COENZYME DEPENDENT ISOMERIZATION OF (Fig. 2a,b). METHYLMALONYL CoA TO SUCCINYL CoA While the P3I multiplet pattern clearly pointed to the non-equivalence of the methylenic protons, Sir: It has been established that the isomerization of final corroboration was effected by returning to the proton resonance pattern. The coupling methylmalonyl CoA' to succinyl CoA is a key step constant (10.7 cps.) obtained from the P3I trace in the metabolism of propionate2 and that a cobwas found to correspond with acceptable precision amide coenzyme3 is an obligatory co-catalyst for to the mean of the experimentally measured cou- this isomerization. 4 , 5 The reaction mechanism pling constants from the H1 methylenic trace, on the now has been investigated using 2-C1*-methylassumption of two similar, but chemical-shifted, malonyl C O A . ~With this substance, migration of methylenic multiplets with slightly different cou- the carboxyl group to the beta methyl carbon would pling constants (JP,cH~ = 10.7 cps., JP,c'H~ = 10.3 yield 2-C14-succinylCoA, whereas movement of the cps.). The detailed pattern of each of the multi- thiolester group would yield 3-C1k i c c i n y l CoA. plets could not, however, be clearly established beHO~CCH~~H~COSC~A cause of some overlap of fine structure. In order to obviate this diffculty, the magnetic anisotropy C02H migration of benzene6 was utilized in an ancillary fashion to increase the chemical shift separation of the two CHs I overlapping multiplets. HO~C$HCOSCOA Several solutions of the phosphothionate ester of varying concentration in benzene were prepared. COSCoA migration As expected, the difference in relative shieldings of the non-equivalent methylenic protons showed a HO~C~H~CH~COSC~A concentration dependence, and degeneracies were In the experiment described in Table I, the 2-C14removed a t a sample concentration of 35 volume CoA was isomerized by incu70in benzene solution (Fig. l b ) . The emergent methylmalonyl pattern became that of two similar multiplets, the bation with an enzyme preparation from Profine structure of each corresponding to the pre- pionibacterium shermanii. The acyl-CoA4 comdicted second order perturbation treatment of the pounds thus formed were converted to their acid amide derivatives by treatment with concentrated energy levels for this type of system. The non-equivalence of the two ethoxy groups ammonium hydroxide and, after the addition of in the molecule therefore has been unequivocally 550 mg. of carrier succinic acid amide, the C14established by both the H1 and P31 spectra. I n succinic acid amide was crystallized to constant order better to evaluate the role of restricted in- specific activity' and was degraded to determine ternal rotation in the phenomenon, studies of the the distribution of the isotope. To carry out the temperature dependence of the spectrum were degradation (1) the succinic acid amide was conconducted with the induction probe as modified by a verted to p-alanine and carbon dioxide by a Hofvariable temperature sample cavity. The critical mann degradation.8 The carbon dioxide is den.m.r. parameters remained invariant between rived from carbon l of the succinyl CoA. (2) 300-500 OK., indicating a rather striking stability The p-alanine was converted to acrylic acid by of structure associated with the non-equivalent exhaustive methylation with dimethyl sulfate.g methylenic groups. This can be explained in The acrylic acid was reduced to propionic acid terms of preferred ethoxy group orientations about which was degraded stepwise to carbon dioxide P-O(R) single bonds. However, on the basis of and methylamine by the Schmidt reaction as modichemical reactivity, the author prefers, and pro- fied by Phares.lo Carbons 1, 2 , and 3 of the proposes to elaborate elsewhere, an explanation based pionic acid thus correspond to carbons 4, 3 and 2 , on unequal P-O(R) bond orders in some molecular respectively, of the succinyl CoA. The data sumsystems containing p,d pi bonds. The net effect marized in Table I show that 80% of the isotope can be likened to a resonance stabilization of the was in carbon 3 of succinyl CoL4.11 It is thus molecule in which the anti-bonding electrons of (1) Coenzyme A is abbreviated as CoA. (2) hl. Flavin and S. Ochoa, J . Biol. Chem., 229, 965 (1957); K . E. only one of the oxygen atoms contribute to canS a i c k and H . G. Wood, Proc. Null. A c a d . Sci. U . S . , 4 6 , 28 (19GO). onical structures involving a conjugated P=S (3) H. Weissbach, J. Toohey and H. A. Barker, i b i d . , 4 5 , 521 (1959). bond. (4) E. R. S t a d t m a n , P. Overath, H. Eggerer and F. Lynen, Biochem. A final feature may be noted. The molecular B i o p h y s . Research Comm., 2, 1 (1960). ( 5 ) J. R. Stern and D. L. Friedman, i b i d . , 2, 82 (1960). asymmetry as demonstrated by the nuclear reso(6) The 2-C~4-methylmalonyl-CoA was prepared by t h e enzyme nance evidence can lead to the prediction of pos- catalyzed transcarboxylation reaction: methylmalonyl CoA + 2sible optical activity for this compound. C"-propionyl CoA % 2-C"-methylmalonyl CoA + propionyl CoA (4). The author thanks R. Schumm and E. J. Prosen (7) The succinic acid amide contained no 2-C"-methylmalonic acid for their cooperation in making available the puri- amide. (8) "Organic Syntheses," Coll. Vol. 11, John Wiley and Sons, New fied sample used in this study, York, N. Y., 1943, p. 19.
t
.1
KATIONAL BUREAUOF STAXDARDS (9) R. Willstatter. C h e m . Ber., 35, 584 (1902). CHEMISTRY DIVISION HAROLD FINEGOLD (10) E. F. Phares, Archiu. Biochem. Biophys., 33, 173 (1951). WASHIXGTON 25, D. C. (11) T h e small amount of isotope found in t h e number 2 carbon a t o m probably is due t o t h e enzyme (CoA-transphorase) catalyzed equiliRECEIVED MARCH7, 1960 _____ bration of 3-Cl4-succinyl CoA with its symmetrical hydrolysis product, (6) A. A. Bothner-By and R . E. Glick, J . Chem. Phys., 26, 1651 (19571.
2,3-CL4-succinate,which is produced i n trace amounts under the experimental conditions. Evidence t h a t such equilibration can account
Vol. 82
COMMUNICATIONS TO THE EDITOR
2644
TABLE I ISOTOPEDISTRIBUTIONIN SUCCINYL CoA DERIVEDFROX 2-c'"METHYLMALONYL COA Carbon
Cpm./patom
COSCoA 0 CHz 14 CHI 54 4 COOH 0 Ten ml. of reaction mixture containing 0.4 pniole of 2C14-methylmalonyl CoA (1.9 X 106 cpm.), 0.035 pmole of dimethylbenzimidazole-cobamide-coenzyme (supplied by H. A. Barker), 2.6 units of avidin,'* 200 pmoles ot histidineHCl buffer (PH 6.2), and an isomerase preparation from P . shermanii (700 pg. protein) were incubated a t 25' for 10 minutes. The mixture was then treated with concentrated NHnOH t o convert the acyl CoA compounds t o their acid amide derivatives. This and three similar reaction mixtures containing a total of 1.8 pmoles of 2-c14 D,L-niethplmalonyl CoA (7.3 X lo5 cpm.) were pooled and the succinic acid amide was isolated and degraded as described in the text. 1 2 3
established that isomerization involves rearrangement of the thiolester group. Although the data do not eliminate the possibility that isomerization occurs by an intermolecular transfer of the thiolester group, a much more reasonable mechanism is suggested by analogy to the intramolecular rearrangement of phenylneopentyl radical discovered by Urry and Kharasch.I3 In the proposed mechanism (Fig. 1) fOOH HC-CH?
THE SYNTHESIS OF THE DIPROPYLCY CLOPROPENIUM ION
Sir: Although the properties of the triphenylcyclopropenium ion1 strongly support the idea that the cyclopropenyl cation is a fundamental aromatic system, i t seemed desirable to prepare a derivative free of phenyl groups. The properties of the diphenylcyclopropenyl c a t i ~ n ~ suggested .~ that a simple alkylated cation might well be stable, and we wish to report that this is indeed the case. Reaction of dipropylcyclopropene carboxylic acid ( I ) 4 with acetyl perchlorate in acetic anhydride,3-5and then ether precipitation, led to dipropylcyclopropenium perchlorate (11),m.p. 4 3 0 ' (dec.), which was recrystallized from ethyl acetate and ether. Found: C, 48.7; H, 6.6. The compound is insoluble in ether, hexane, or other nonpolar solvents, but is soluble in acetone and other polar organic solvents and in 1 N HC1. It gives a positive perchlorate test with potassium nitrate, and the infrared spectrum also reveals the presence of perchlorate ion. I n strong acid the compound has only end absorption in the ultraviolet. CH3CHzCH2 Tr_7-C'lI~CHd2HHj
H
f00 H HS-CH,
xCOZH Y
c104
H I1
The n.rn.r. spectrum, in SOY0 aqueous sulfuric acid, is as expected, with a sharp band a t -4.15 p.p.m., shifted to very low field for the proton on a positive carbon, and the characteristic pattern of two I equivalent propyl groups also strongly shifted, with SCoA SCoA a 4 proton triplet centered a t +3.12, a four proton Fig. 1 sextuplet a t +4.43, and a six proton triplet a t the role of the cobamide coenzyme is to produce a +5.30 p.p.m. relative to a benzene capillary; radical by the one electron oxidation of methyl- the absence of any i u r t l a : i"!-b ~ d was s estabmalonyl-CoA. The radical then could isomerize lished by dilution of the solvent, with consequent by the mechanism of Urry and Kharasch and the shift of its absorption. succinic acid derivative could be stabilized by Although the cation is soluble and stable in 1 N acid, its solution in 0.1 N acid or water becomes interaction with reduced cobamide coenzyme. In view of the fact that the cobalt in cobamide turbid. This reaction of the carbonium ion with coenzyme has a valence of +3 i t is of further sig- water gives a complex mixture of products, but nificance for the proposed mechanism, that COf 3 a t least some of the components are cyclopropene is known to produce radicals by reaction with derivatives, as evidenced by their characteristic a variety of organic compounds.14 double-bond stretching absorption in the infrared M AX-PLANCK-INSTITUT H. EGCERER a t 5 . 3 ~ . From this it can be estimated that the FOR ZELLCHEMIE P. OVERATH p K a of the cation is greater than 0 and probably is F. LYNEN less than 1. For comparison, the diphenylcycloMUNCHEN,GERMANY NATIONAL HEARTINSTITUTE BETHESDA, MARYLAND E. R. STADTMAN propenyl cation has a p K a of 0.3.* That these two are so close is certainly a chance coincidence, RECEIVED APRIL6, 1960 for randomization of the carbon atoms was obtained by showing that, under otherwise identical conditions, the isomerization of unlabeled rnetbylmalonyl CoA in the presence of added trace amounts of 1,4-C"succinate led to the formation of C14-succinyl CoA n,hich was isolated as the succinic amide derivative. (12) Avidin was added t o prevent the decarboxylation of rnethylmalonyl CoA t o propionyl CoA ('4). (13) W. H. Urry and M . S Kharasch, THIS J O U R N A L , 66, 1438
(1944).
(14) A. F. Trotman-Dickenson, "Free Radicals," John Wiley and Sons, New York, N. Y.,1959,p. 10.
(1) R.Breslow and C. Yuan, THISJ O U R N A L , 6 0 , 5991 (1958). (2) T h e diphenylcyclopropenyl cation has been prepared by Dr. Joyce Lockhart, unpublished work, by the reaction of phenylrhlorocarbene with phenylacetylene. (3) Prepared independently by Dr. Donald Farnum by the use of acetyl perchlorate on the appropriate acid. We wish t o thank Dr. Farnum for making his results available prior to publication. ( 4 ) I. A. Dyakonov, cl al., Zhrrr. Obshchei Khitn., 29,3848 (1959). (5) A similar reaction with acetyl Euoroborate, t o form the tropylium ion, has been reported by M. Dewar and C. Ganellin, J . Ciiem. Soc.,
2438 (1958).
