2!)72
COMSIPNIC.\TIONS TO T I I E EDITOR
dation and infrared spectra.i In the original hydrocarbon fraction, the cis isomer is present to the extent of 5.67,, and the frans, to the extent of 10.1%. This is the first time that the formation of a cyclopropane from the deamination of a saturated amine (IV + VI) has been reported. It seemed possible that similar cyclizations might be observable in other systems. For instance] the i3-methyl-2-butyl cation (Y) could be formed by hydrogen migration in the isoamyl cation (111, which in turn could be generated by the deamination of isoamylamine (thus, I I1 + V). If the :3-niethyl-2-butyl cation were implicated in the formation of 1,2-dimethylcyclopropane from the deamination of IV, then the 3-methyl-2-butyl cation formed by this alternative path from isoamylamine might also yield 1,%dimethylcyclopropane (I VI). On the other hand, the deamination of isoamylamine might produce l ,1dimethylcyclopropane ( I 111). Experimentally, the deamination of isoamylamine gave a hydrocarbon fraction which contained 0.5% cis-],2-dimethylcyclopropane and 1%
-
-
-
1-01. sl!
A second explanation is that a discrete bridged methylcarbonium ion (VII) is produced and that this ion then loses a proton to yield the cyclopropanes9 Xttempts a t interpreting the results with isoamylamine and neopentylamine in terms of this mechanism also lead to difficulties, but they do not appear insurmountable. In any event, the present results and future experiments suggested by them should prove illuminating in the study of the nature of the intermediates in the amine-nitrous acid reaction. The author thanks Dr. -4. Kropf for helpf~il discussions. (0) D. J . C r a m and J. E. M c C a r t y , THISJ O U R N A L 79, , 2866 ( 1 9 5 i 1 , recently have discussed t h e question of bridged methylcarhimiurn ions.
LABORATORY O F CHEMISTRY COLLEGE ?\IARC AMHERST,X~SSACHVSETTS RECEIVED FEBRUARY 26, 1960
hIOORE
=\MHERST
s. SILVER
T H E SYNTHESIS OF MALONYL-C14 COENZYME A
Sir:
Nalonyl CoA has been established as one of the key intermediates in the biosynthesis of fatty acids.'.? In order to discover the detailed mechanism of the condensation of malonyl CoA with acyl CoL4 acceptors, it was essential to obtain isotopically labeled malonyl CoA of high purity in good yield. I n a recent communication, T'agelos3 has described a method for the synthesis of malonyl-coenzyme -2, based on the preparation IT' of a mixed anhydride of malonic acid. For the preparation of isotopically labeled malonyl CoA, the method of Vagelos seemed uneconomical since the over-all reported yield was of the order C H 3 L of 3.1 to 5.t370in respect to malonic acid. In our \'I1 laboratory, a synthesis based on the observations trans-isomer, as determined by the procedures of Khorana and co-workers4 employing dicyclooutlined above for the deamination of 3-methyl-2- hexylcarbodiimide (DCC) as a condensing agent amin0butane.j Xo 1,l-dimethylcyclopropane was was used to synthesize the monothiophenyl ester of detected in the deamination of isoamylamine and malonic acid. The latter compound was transno cyclopropanes a t all were detected in either the esterified with coenzyme A according to the method deamination of neopentylamine or in the acetoly- of Wieland and Rueff . 5 Experimental.-1 mM. of malonic acid (1sis of 3-methyl-2-butyl P-toluene~ulfonate.~~~ I t is interesting to speculate on the mode of C14 or 2-C14) and 1 mM. of thiophenol were disformation of these cyclopropanes. At least two solved in 5.0 ml. of N,N'-dimethylformamide a t possible explanations come to mind. First, a 0". To the mixture was added with stirring over carbene-type mechanism may be involved, related a period of 1 hour 500 mg. of DCC in 5 nil. of N,N'to the "intramolecular insertion" reaction of di- dimethylformamide. The mixture was stirred for azoalkanes, recently described by Friedman and three hours at 0". After the addition of 10 ml. of S h e ~ h t e r . However, ~ the conditions of the de- water, the stirring was continued for 15 minutes. amination reaction (acetic acid) do not appear The mixture was filtered with suction and the favorable t o the formation of a diazoalkane from precipitate washed with water. The filtrate was an alkyldiazonium ion. The present observations made slightly acidic and extracted with several on the behavior of isoamylamine and neopentyl- volumes of ether. The ether phase was washed amine are also not readily accommodated in such a with 0.01 M hydrochloric acid and with water. mechanism.* After drying over anhydrous sodium sulfate, the solution was shaken out with activated charcoal (4) Infrared spectra were r u n a t t h e University of Massachusetts, and filtered. The yieId varied from 400 to 600 b y kind permission of Dr. L. Carpino. (6) T h e &/trans ratio is about 1:2 for both isoamylamine a n d 3micromoles of monothiophenylmalonic ester as
f
methyl-2-aminobutane. (6) It is estimated t h a t 0.1% would have been detected in t h e present experiments. (7) L. Friedman a n d H. Shechter, THISJ O U R N A L , 81,5512 (1959). (8) Decomposition of 2,2-dimethylpropanal tosylhydtazone (analogous to neopentylamine) gives a hydrocarbon fraction t h a t is mainly 1,l-dimethylcyclopropane(ref. 7 ) .
(1) R . 0. B r a d y , Proc. U . S . Kat. A c a d . Sci., 44, 993 (1958). (2) S. J. Wakil, THISJOURNAL,80,6465 (1958). (3) P.R . Vagelos. J . B i d . C h e m . , 2SS, 346 (1960). (4) H. G. Khorana, W. E . Razzell, P. T. Gilham, C. M. Tener and E. H . Pol, THISJ O U R N A L79, , 1002 (19.57). ' 3 ) 1'.Wieland and I*. R u e f f , .4ngrrc C h e m . , 6 5 , 186 (1953).
June 5 , 1960
COJIhlITNIC.\TIONS TO
determined by hydroxamate formation and radioactivity. The purity of the thiophenylmalonic ester was 98Yc as determined by measuring the extinction a t 237 mp ( E = 8.4 X lo3)in ethanol and the yield of hydroxamic acid. The thiophenylmalonate was concentrated under reduced pressure and a slight excess was added slowly with shaking to a solution of 10 p M . of coenzyme 4 in 0.1 X bicarbonate buffer a t pH 8.0. Nitrogen was bubbled through the mixture a t 0" for 3 hours. The mixture was then acidified and extracted several times with ether. In the aqueous phase 7.5 p M . of malonyl-CoA wasobtained which had an hydroxamate :adenine ratio of 0.66. The aqueous phase was then lyophilized or frozen. The aqueous phase which contained some unreacted CoA could be purified by paper Chromatography in 0.1 d l potassium acetate a t pH 4.5:ethanol 1: 1 a t 4". The eluted material was 96% pure based on the ratio of hydroxamic acid to adenine. The yield was 4.5 p M . of malonyl coenzyme A. Chromatography of the hydroxamic acid derivatives obtained from malonyl CoA yielded Rf values of 0.36 in water-saturated l-butanoP and 0.5 in pyridine :2-butanol :water (1 : 1: 1).7 The over-all yield in respect t o malonic acid-C14was 18-27%. When malonyl-C14 CoA was incubated with an enzyme system which synthesized fatty acids from malonyl C o h and acetyl CoA, radioactivity was incorporated into palmitic acid in good yield. Methylmalonyl Co-4 also has been prepared satisfactorily by this procedure.
TIIE
2073
EDITOR ("2
R0&
110s (.Hs
I
o&OH
I
II
CH3
CH3
0
I
I
HO&
o&=o IV
111
0 ,CH3
CHQ 1
&;.+.&5
O b iL O H
0
V
Column chromatography furnished, in addition to small amounts of progesterone and starting 2Op-01 (I), some 15yo of 20~-hydroxy-18-oximino-4pregnen-3-one (11, 20p-OH; m.p. 242-243'; [a]2 4 ~ 154.2; €;;?A = 16,800; XNurol a t 2.85, 3.14, 6.05, 6.18 and 10.75 p ; found: 73.01; H, 9.34; N, 4.29). On the other hand, a like series of reactions with the 20a-hydroxy-4-pregnen-3-one (6) 0. Hayaishi, J. E d . Chem., 216, 125 (1955). (II R = a-OH), via the nitrite (I, R = a-ONO; (7) P. R . Vagelos, THISJOURNAL, 81, 4119 (1959). m.p. 192-198'; [ c Y ] ~ ~ D 103.9; = 19,000; I-ATIONAL INSTITUTE OF NEUROLOGICAL E. G. TRAMS X N " l 0 l a t 6.00, 6.14, 6.20, 6.29, and 12.78 p ; found: DISEASESA N D BLINDNESS R. 0. BRADY N, 4.21), gave the oxime (11, 20a-OH; m.p. BETHESDA 14, MARYLAND RECEIVED APRIL 18, 1960 184-186", [ c Y ] * ~ D 148.8, ~ T t p=~ 17,300, XNuJol a t 2.95, 3.10, 6.04 and 6.19 p ; found: C, 73.18; T H E PHOTOLYSIS OF ORGANIC NITRITES. I. H, 8.86; H , 4.03) in 60% yield by direct crystal18-NITRILOPROGESTERONE : CORRELATION WITH lization from the photolysis mixture. This difference in yield is believed to reflect the more CONESSINE favorable conformation of the active intermediate Sir : The discovery that it is possible to effect specific in the 20a-series vis-&vis the angular methyl intramolecular hydrogen abstraction by the photol- group. Both oximes were converted to a common ysis of a strategically located nitrite ester grouping derivative by destroying assymetry a t C-20: already has led to a simple synthesis of the potent cortical steroid aldosterone. This general method the "a-oxime" (11, 20a-OH) was smoothly dehyopened the way to analogous syntheses of a variety drated to 20a-hydroxy-18-nitrilo-4-pregnen-3-one [ a I z 4 D +136.2; of steroidal structures heretofore not easily ac- (111, 20a-OH; m.p. 237-241'; = 16,400, XNuJa' a t 2.97, 4.46; 6.01 and 6.6 cessible. We wish to describe here the synthesis p ; found: C, 76.99; H , 8.70; N, 4.31) by hot pyriof 18-nitriloprogesterone (IV). 20p-Hydroxy-4-pregnene-3-one (I, R = P-OH) dine-acetic anhydride and then alkaline hydrolysis. in pyridine, upon treatment with nitrosyl chloride, In the same manner, conversion of the "p-oxime" gave the corresponding nitrite ester (I, R = (3- (11, 2Op-OH) gave the isomer (111, 20P-OH; 115.9; E?
