COMMUNICATIONS TO THE EDITOR
July 20, 1956
exposes a tertiary hydroxyl.12 Pyrolytic loss of acetic acid furnishes unsaturated pyro compounds which in certain cases can be hydrogenatedg and isomerized. Thus pyro-a-oxodelphinine isomerizes to isopyro-a-oxodelphinine (A9*l4). Saponification of its benzoyl group gives isopyro-cuoxodelphonine (IT), exposing a secondary hydroxyl vicinal to the free tertiary hydroxyl as shown by subsequent oxidation to an unsaturated keto acid C24H33NOs (111).13 Isomerization of the latter to the keto- y-lactone
OM e-
OMe
OMe
(IV) (uCHC1a 1783, 1706 cm.-l) can be explained by lactonization upon the bond. Placing the benzoyloxy group a t C- 18 and the tertiary hydroxyl a t C- 19 satisfies these observations. Dihydroisopyro-a-oxodelphonine (V) on oxidation gives a ketoacid, C2&&08 (VI) bNujol,(six-membered ketone), 1700 cm.-l; methyl ester: vNujol (COnMe), 1743 cm.-l; (six-membered ketone), 1705 cm.-l] which can be cyclized to a /3-diketone,14C24H33N07 (VII) (vKBr 1766, 1720 cm.-l). This may be reopened to the original acid. On catalytic reduction VI1 affords a hydroxy pentanone. (vNujOl 3409, 1746 cm.-l).
OMe
OMe-
3543
TABLE 1“ Porphyrin
VC-CI
Free tetra-( p-chlorophenyl) porphine Vanadyl salt Copper salt Cobalt salt Nickel salt Zinc salt a Measured as KBr discs on a P. E. Model Spectrophotometer equipped with KBr prism.
(cm. -1)
491 .O 498.3 502.0 503.5 504.5 497 5 12c Infrared
gested that the C-C1 stretching frequency in these compounds might occur a t 941 em.-’, the reference3 upon which this suggestion was based indicates to us that this type of absorption would be expected to occur below 650 cm.-l, and the higher value quoted by these authors is undoubtedly the first overtone of the stretching vibration. The structural formula of tetra-(p-chloropheny1)porphine (Fig. 1) indicates that the C-C1 linkages CI
OMe
For aconitine, which has two more hydroxyls than delphinine but bears an N-ethyl group, we envision the same basic skeleton since much of its chemistry parallels that of delphinine.
Cl
Fig. 1.
are relatively distant from the center of the porphine ring. Furthermore, the suggestion of these authors2 that “the four benzene rings have partial rotation which cannot bring them 60’ of being coplanar with the resonating porphine system’’ would indicate that the benzene rings themselves THEROCKEFELLER INSTITUTE WALTERA . JACOBS could not participate in resonance contributing A-EW YORK21, IK,Y. S. WILLIAMPELLETIERforms of the porphine ring. Therefore, one might RECEIVED J U N E 11, 1956 conclude that the C-C1 bond should not be affected by substituting a metal into the center of the moleFREQUENCY SHIFT OF CARBON-CHLORINE BOND cule in place of the two hydrogen atoms of the free VIBRATION I N METAL DERIVATIVES OF TETRA- porphyrins. For these reasons, the observed (P-CHLOROPHENYL )-PORPHINE frequency shifts of the C-CI vibration seem someSir : what contrary to expectation and may offer direct In a study of the infrared spectra of tetra-($- experimental evidence that the benzene rings are in chloropheny1)-porphine and its metal derivatives, fact in conjugation with the porphine nucleus. we find (Table I) that the absorption band due The higher vibrational frequencies of the metal to stretching of the C-C1 linkage will shift position derivatives seem indicative of more double bond depending upon whether the free porphine or a character of the C-C1 linkage in these cases and, particular metal complex is under consideration. with this in mind, a great number of resonance Although Thomas and Martel12 recently sug- structures can be postulated of which Fig. 2 may (1) Prepared by method of P. Rothemund a n d A. R. Menotti, THIS be taken as an example. J O U R N A L , 63, 268 (1941), and TO, 1809 (1918). The vanadyl salt The hypothesis that the C-C1 bond character is was prepared by method of J. T. Horeczy, et a l . , Anal. Chem., 87, 1899 altered by the introduction of a metal into the por(1955). (13) W. A. Jacobs and Y .Sato, J . B i d . Chem., 180, 479 (1949). (14) Neither the lycoctonine skeleton of Przybylska and Marion [Canad.J . Chem., 34,185 (1956)l nor the related skeleton,derived from veatchine by the biogenetic pathway used by Valenta and Wiesner [Chem. and Ind., 354 (1956) 1 t o relate lycoctonine t o atisine, accommodate all of our data.
( 2 ) D. W. Thomas and (1956).
A.
E. Martell, THISJOURNAL, 18, 1338
(3) L. J. Bellamy, “The Infrared Spectra of Complex Molecules.” John Wiley and Sons, Inc., New York, N. Y.,1952, p. 271.
COMMUNICATIONS TO THE EDITOR Ci
9 \
CI
0
Fig. 2
phine ring system finds additional support in a recent comparison of the paramagnetic resonance absorption spectra of copper tetraphenylporphyriri and the 9-chloro derivative. This study suggests that the unpaired electron of the copper atoms can move out to the peripheral C1- atoms Tr ’i a T orbitals of the conjugated resonating system. ‘ (4) D J E Ingram, J. E Bennett, P George and J \ I Goldstein, TIUSJ O U R N A L , 78, 3545 ( 1 9 5 0 )
DEPARTNEXT OF CHEMISTRY J bl GOLDSTEIX UNIVERSITY OF PENXSVLVANIA IT 11.S I C S A B B PHILADELPHIA 4,PEYWYLVAYIA J F HAZEL RECEIVED M A Y9, 1956
Vol. 7 8
the interconversion of L-xylulose and D-xylulose. One enzyme, requiring triphosphopyridine nucleotide (TPN) as coenzyme, uses only xylulose and xylitol as substrates. The second enzyme, requiring diphosphopyridine nucleotide (DPN i, has D-xylulose and xylitol as its preferred substrates. The TPX-dependent enzyme catalyzes the dehydrogenation of xylitol but has practically no action on erythritol, ribitol, L-arabitol, D-sorbitol, D-talitOl,*D-gulitol,s or D-mannitol Oxidation of reduced T P N is effected by L-xylulose, whereas D-xylulose, ~-erythrulose,~ D-ribulose,lo D-fructose, D-sorbose,8 and L-sorbose are essentially inactive. Since the L-xylulose-xylitol enzyme has a marked preference for T P N over DPN, i t appears to be unique among polyol dehydrogenases (ketose reductases) thus far reported, all of which are I)PN-dependent. The DPK-dependent enzyme catalyzes the oxidation and reduction of xylitol and D-xylulose, respectively. The D P N system acts more slowly on ribitol and D-ribulose but has little or no action on erythritol, L-arabitol, D-talitol, D-gulitol, D-mannitol, L-xylulose, D-fructose, D-sorbose, or L-sorbose.“ Figure 1 presents evidence for the existence of two distinct xylitol dehydrogenases. The T P N -
w
AN ENZYMATIC PATHWAY FROM L-XYLULOSE TO
D-XYLULOSE’
Sir: The detection of L-xylulose in the urine of “nonpentosuric” h u m a n ~ , ~guinea , ~ ’ ~ pigs,3 and rats,4 together with the discovery of the presence, in guinea pig liver mitochondria, of a highly active enzymatic system for the reduction of xylulose to xylit01,~indicates that this ketopentose may be a normal metabolite. Recent work6,’ has shown that D-xylulose-5-phosphate,rather than D-ribulose&phosphate, is the ketopentose substrate of transketolase (TK) in the 6-phosphogluconate pathway of carbohydrate metabolism. We wish to describe studies on new mitochondrial enzymes which link metabolically the enantiomorphic forms of xylulose. A stable, soluble enzyme preparation has been obtained by a butanol treatment of the insoluble portion of ruptured guinea pig liver mitochondria. I t contains two enzymes which would make possible ( 1 ) Supported in p a r t b y a grant from t h e Sational Science Foundation (2) 0. Tousier, R Rl. Hutcheson and V. H . Reynolds. Tins J O G R XAL, 76, 5005 (1954). (3) 0. Touster, R. hl. Hutcheson a n d L . Rice, J . B i d Chenz., 215, 677 (1955). (1) S. Futterman and J . H . Roe, i b i d . , 2 1 6 , 257 (1955). ( 5 ) 0. Touster, V. H. Reynolds and R . 11. Hutcheson, J . B i d . L‘hem., in press. (6) P. A. Srere, J. R . Cooper, V. Klybas and E. Racker, AYCII. Biorhcm. Biofihys., 69, 833 (19%). ( 7 ) I1 1, €lcu..:.cr, .J lIiiriiiiL ~ n 1’d. 2 . Smymiotis, THISJ D u K N A I . , 18. CjY;: (19;tQ.
TIME
I\
MINUTES.
Fig. 1.--Comparison of the two mitochondrial xylitol dehydrogenases: flask contents, 19.7 p M , xylitol, 0.39 pb1, coenzyme, 0.3 ml. enzyme, in 3.0 nil. of solution 0.05 .If to “tris” buffer (pH 8.0) and 0.008 IIf t o MgC12; temperature, 23.5’. Dialysis mas against water overnight. Heat treatment of enzyme was a t 50” for 20 min. “Washed residue” indicates enzyme prepared from insoluble mitochondrial residue which had been washed three times with \vatcr befoie butanol treatment.
dependent enzyme, unlike the DPN-enzyme, loses all or most of its activity by dialysis against water or by preliminary washing of the mitochondrial residue. Table I shows that both dehydrogenases (8) Kindly supplied b y Dr. S . K. Richtmyer. (9) Kindly supplied b y Dr. G. C . hfueller. (10) Kindly supplied, a s t h e 0-nitrophenylhydrazone, by Dr. B. I,. Horecker. (11) T h e extract catalyzes t h e reduction of D P N in t h e presence of sorbitol. This reaction is probably due t o a third enzyme, since Lsorhose and i>-fructuse. t h e expected dehydrogenatiim I>rodiicts, are inactii c, The p r u d u c t mag he ii Y-ketolmlyiil.
