1506
Vol. 70
COMMUNICATIONS TO THE IZDLTOK
ficant activity in the fluoride assay system, and the F enzyme has none in the hydroxylamine assay system. In the absence of hydroxylamine, Reaction i can be coupled with Reaction 3 (catalyzed by HIV CoA carboxyla~e'~) to yield HMG CoA. The action of the HMG CoA cleavage enzyme14 converts this product quantitatively to equimolar amounts of acetyl coA4 and acetoacetate (Reaction 4). The data in Table I1 demonstrate that the H
OPTICAL ROTATORY DISPERSION STUDIES. X.' DETERMINATION OF ABSOLUTE CONFIGURATION OF or-HALOCYCLOHEXANONES'
Sir: The determination of the absolute configuration of carbonyl compounds by rotatory dispersion
measurements',~involves a comparison of the rotatory dispersion curve of the unknown substance with those of reference ketones (or aldehydes) which have been sterically related to D-glyceraldehyde. In connection with an extensive study5 T A B L E 11 of the effect of a-substitution upon the rotatory dispersion curves of ketones, there has been made C A R B O N DIOXIDIl PIXATIOS 13Y THE COLPI,V,l7 .ICTIC)S O F !I ENZYME A N D 0-I-IYDROXYISOVALERPL Co.1 CARBOXYLASE an observation which appears to offer a means of establishing the absolute configuration of cycloThe test system contained 590 pnioles of potassium bihexanones without requiriiag a r e f e r e ~ c ecompound carbonate, 5 pmoles of Versenc, 50 ptnoles of magr.esiu:n chloride, 10 pmoles of glutathione, 20 pmoles of ATP, 2 of known abso?ute conjiguration. pmoles of HIV CoA, and, where indicated, H enzyme (2.3 While introduction of an equatorid halogen atoiii mg. of protcin), F enzyme (2.0 mg. of protein), a n d a pig (chlorine, broiiiine but ?Lot fluorine) in the aheart fraction containing HIV CoA carboxylase but free of II and F enzymes ( 5 . 2 ing. of protein), in a final volume position of a cyclohexanone does not seem to affect the sign of the rotatory dispersion curve of the of 3.0 ml. An excess of HMG Co.4 cleavage enzyme was present in each experiment. Iiicubatiori was for 60 minutes halogen-free parent k e t ~ n ethis , ~ is not necessarily a t 38". H I V CoA was omitted in control experiments. the case with axial bromine or chlorine substituents. In the latter instance, the stereochernistry pMoles acetoEnzymes a d d e d
H enzyme F enzyme H enzyrnc
acetate f o r m e d l
+ Carboxylase + Carboxylase
0.60 0 0.07 0.02
Carboxylase
enzyme is active in this system whereas the F enzyme is without effect. It is concluded, therefore, that the H enzyme and the carbon dioxideactivating enzyme are probably identical, but that the F enzyme plays no role in carbon dioxide activation for HIV co.4 carboxylation. In accord with these findings, the coupling of Reactions 1, 3 and 4 in 3 heart extract free of pyrophosphatase has been found to furnish approximately equimolar amounts of pyrophosphate and acetoacetate. Since the results presented establish two enzymatic steps in the carboxylation of HIV GoA$, it seemed of particular interest to determine which of these might he lacliirig in biotin deficiency."j Enzyme extracts prepared from the livers of biotindeficient rats possess no H I V CoA carboxylase activity, but have thc same content of H enzyme, F enzyme, and HMG CoA cleavage enzyme as do those preparcd from norinal 1 i ~ e r s . l The ~ H enzyme has recently been purified GSO-fold from pig heart extracts and found to require the presence of Zn++ ions for iiiaxiinal activity. (13) l3. K. 13achharvdt, W. C. Robinson and M. J. Coon, J. Bioi. Chem., 219, 539 (1956). (14) B. K. Bachhawat, W. G. Ilohinsoil aud M. J . Coon, ibid., 216, 727 (1955). (15) Estimated b y a modilicdtion of l h e method of S. S. 'Barkulis a n d A. L. Lehninger. J. Biol. Chewi.. 190, 339 (1951). (10) C. W. E, Plaut a n d H. A. L a r d y , J. Biol. Chcm., 186, 705 (1950).
(17) Unpublished experiments carried o u t b y t h e a u t h o r s in collaboration with Dr. J. F. Woessner a n d Dr. H e n r y A. Lardy.
DEPARTMENT O F BIOLOGICAL CIILMISTRY MEDICAL SCHOOL R. K. BACHIUWAT UNIVERSITYOF MICHIGAN M. J. COON ANN ARBOR,MICHIGAN RECEIVED JANIJARY
81, 1957
of the
o s ij I -c-c-
chronlophore as a whole becomes
dominant and controls the sign o j the Cotton eJect. The gross shape of the dispersion curve can be predicted by examining a model of the appropriate ketone in the following nianner : Place the cyclohexanone ring in such a way that the carbonyl carbon atom is the "head" of the chair and look along the 0-C bond as shown by the arrow in I. As demonstrated in Table I , if
Axial a-haloketone
Cotton effect Halogenfree a-Halohetoue ketone
3a-llroiiioandrostan-?-otlc- 176-01 propionate Positive 7a-Bromocholestane-~13,ja-diol-6-one3-acetate h-egative 7ol-Broniochoiestatie-Xil,.ja-diol-0-one 3,S-diacetate Negative tj~-Bromocholestan-.'{3-~~l-7-one acet3te Negatiic Ba-Rromoergostan-3~-oi-ll-oucacetate I'osi t ive 15ol-Bromoergostan-~~-O~-ll-one acetate I'ositive 12~-Chloro-ll-ketotigogeninacetate Positive 12a,S3-Dibrorno-ll-ketoti#ugeajn acetate Positive I r e t h y 1 1 I p - b r o m t ~ - 3 aacetoxp-12-ket[,cliol~~natePositive Sa-Bromolriedelin Kegativc la-Bromofriedelin Negative
Positive l'osi ti v e Positive Positive Positive h-egative Negative Segative Negative Negative l'usitive
the axial a-chluriiie or broiiiiiie atom lies to the right (ca. 105' angle') of the observer (I, X = Cl or Ur), the single Colton effect ,will be positiue; if (1) Paper I S , C. Djcrassi and W. l i l y n e . Chcirtirfry a n d I % d u s t v y , 988 (195G). (2) Supported b y G r a n t No. CY-PO19 f r o m t h e National Cancer I n s t i t u t e , National Institutes of Health, U. S. Public H e a l t h Service. (3) C. Djerassi, R. Riniker a n d B. Riniker, T I ~ SJouRNAr., 7 8 , 0363 (1950). (-1) C. Djerassi, e l ai., to be published. T h e rotatory dispersion curves of t h e ketones listed i n Table I will be reproduced in t h a t article. ( 5 ) For example, cholestan-3-one, Sa-bromocholestanone a n d 4 a broniocholestanone all exhibit a positive single Cotton effect curve.' ( 0 ) F o r nomenclature in rotatory dispersion work see C. Djerassi a n d W. Blyne, Proc. C h r ? ~Soc,, . 5 5 (1957). (7) It. C. Cookson, J. Chcin. SOL.,181 (1954).
March 20, 1957
COMMUNICATIONS TO THE EDITOR
the chlorine or bromine is to the left (I, X’ = C1 or Br) the effect will be n e g a t k 8
I 0
1507
hibits a positive Cotton effect (peak6 a t 295 mp, [ a ] +498’) as does the halogen-free 1l-ketone,l3 while the axial llp-isomer shows a strong negative Cotton effect (trough6a t 337.5 mp, [a] - 1020”). (13) C. Djerassi and W . Closson, THISJOURNAL, 78,3761 (1956).
DEPARTMENT OF CHEMISTRY WAYNESTATE UNIVERSITY DETROIT, MICHIGAN X’
I
CARLDJERASSI
POSTGRADUATE MEDICAL SCHOOL
LONDON W. 12, ENGLAND RECEIVED FEBRUARY 14, 1957
W. KLYNE
Since the axially-oriented bromine or chlorine substituent in a-halocyclohexanones is often the more stable the presently described method MULTILAYER MEMBRANE ELECTRODES’ for the determination of absolute configuratiom of such ketones should be rather widely applicable. Sir: The steps to be taken are simple and particularly We have prepared a new membrane which is if the first one can be omitted, the entire sequence permeable almost exclusively to certain ions in the can be carried out on a micro-scale without using presence of other ions of the same charge. These up any material: (a) if both a and a’ positions membranes are multilayers, where transport of can be substituted and the exact location of the the potential-determining ion is transverse to the halogen atom does not follow from its mode of axis of orientation. preparation (e.g., epoxide opening followed by oxiMembranes were constructed by cracking a dation), then this must be established chemicallystandard slide in half, so as to insure a nearly perusually by dehydrohalogenation ; (b) the axial fect fit, because the effective membrane pore orientationlo of the a-halocyclohexanone must be width was 2.5 x cm. Each glass half-slide confirmed by ultraviolet7 and infraredg,l1 measure- was coated with 50 Y-type monolayers of barium ments; (c) the rotatory dispersion curve must be stearate, the two halves fitted together and cemeasured. It is then possible t o determine in the mented in place between slides having holes in their manner illustrated above (cf. I) which one of the centers, arms were attached for solutions and the two absolute configurations leads to the observed potential measured between saturated calomel (positive or negative) Cotton effect. electrodes. Attention should be called to two further apThe resistance of this membrane was 5.8 X lo7 plications of rotatory dispersion measurements ohms. Accordingly, a highly sensitive Keithley among a-halocyclohesanones. If the absolute Model 200 B vacuum-tube voltmeter accurate to configuration of the parent ketone is known but =t 1% was used with proper shielding. A conthe location ( a or a’)of the axial halogen atom is stant asymmetry potential of 1-3 mv. for identical not, then the rotatory dispersion curve will provide barium chloride solutions was read; constant and the answer. Conversely, if information concerning stable potential readings were obtained rapidly. the absolute configuration of the ketone and the Preliminary results obtained with this Multilocation of the halogen substituent is available, layer Membrane Electrode are given. Experibut spectroscopic determination of the orientation mentally determined potentials were corrected of the halogen atom is obscured by additional for asymmetry. Calculated e.m.f. values were obcarbonyl groups, a decision may be reached by tained using the tabulated data of ConWaf; a t means of the rotatory dispersion curve. An in- higher concentrations, mean activity coefficients teresting example is provided by the two epimeric were used. For dilute (< 0.02 M ) solution pairs methyl 3a-acetoxy-ll-bromo-12-ketocholanates1z whose concentration ratio was less than 2:1, calwhich possess virtually identical rotations ( [ y ] ~culated e.1n.f. values were reasonably accurate. +44’ and $-4o”) a t the conventional sodium D Multilayer Membrane Electrodes are shown to line. However, the rotatory dispersion curves* be reversible to barium ions, even in the presence disclose that once rotation measurements are con- of high concentrations of sodium ions, .A. real tinued in the ultraviolet range, a marked change is measure of this specificity cannot be obtained frorn noted. I n agreement with the postulates made in potential measurements, but is being deterniincd this paper, the equatorial 11a-bromo isomer ex- using transport, exchange, and other methods. ( 8 ) This effect may be so pronounced that it may carry over into the The ability of Multilayer Membrane Electrodes visible range of the spectrum. Consequently, a t times even comparito concentrate a specific ionic species without coniSOUS of molecular rotation differences (based only on [ P ] D values) with pletely immobilizing that species is responsible for steroid reference compounds may be used as has been done in the its specific function. The glass electrode is an exfriedelin series (E. J. Corey and J. J. Ursprung, THISJOURNAL, 78, 5041 (1956)). ample. This ideal or nearly ideal degree of speci(9) E. 5. Corey and H. J. Burke, i b i d . , 77, 5418 (1955), and earlier ficity results from its highly oriented or semi-crystalpapers. See, however, W. D. Kumler and A. C . Huitric, ibid., 78, line state. Pores, in the usual sense, are not pres3369 (1956). ent. Transport probably takes place by ex(10) I t appears (ref. 4) that even if the equatorial orientation is the more stable one it might still be possible to use this method if the correchange between adjacent barium ions. Nultisponding gem-dihalo ketone is available since one of the halogen layer Membrane Electrodes specific for other catatoms must now be axial. (11) R. N . Jones, D. A. Ramsay, F. Herling and K. Dobriner, THIS 2828 (1952). (12) Kindly provided by Dr. T. F. Gallagher, Sloan-Kettering Institute for Cancer Research. JOURNAL, 74,
(1) We wish t o thank tbe Office of Naval Research and the U. S. Public Health Service for the support given this work. (2) B. E. Conway, “Electrochemical Data,” Elsevler Publishing Co., N e w York, N. Y.,1952, p. 102.
