Nov. 5, 1961
COJIMUNICATIOXS TO THE EDITOR
4483
substance, a metabolite of 18-hydroxycorticos- anhydride in pyridine gave a t r i a ~ e t o x yderivative '~ ter~ne.~ We wish to report the correct structure of (11) RnocA= 1.1in methylcyclohexane-formamide, the aldosterone metabolite as the 301,5p isomer of = 1.3 in methylcyclohexane-methanol: water 4 : 1, tetrahydroaldosterone (I). 5.71, 5.77p, which like aldosterone15 readily Tetrahydroaldosterone (301,5p) was prepared lost one acetoxy group when treated with dilute using a ~ o l u b l eenzyme ~~~ fraction. Male rat acetic acid. livers were homogenized in acetone a t -15O, the Oxidation of tetrahydroaldosterone (I) with homogenate was filtered and the residue dried a t periodic acid yielded the theoretical amount of room temperature in vacuum. The acetone powder formaldehyde16 and a monohydroxy y-lactone was suspended in buffer solution8 and the filtrate (111), m.p. 252-254', 2.88, 5.64 M , C, 72.46; incubated with d-aldosterone-21 monoacetate9 in H , 8.46. This lactone (111) also was obtained from the presence of TPNH'O under nitrogen a t 37". aldosterone etiolactone (IV) l7 using the enzyme Extraction and chromatography of the incubation preparation described above. Acetylation yielded mixture revealed only one substance (I), other than the monoacetoxy dervative14 (V), RAn = 0.9 in unchanged substrate, which reduced blue tetrazo- methylcyclohexane-formamide, 5.60, 5.78 p . lium." There was a single band of simple contour a t 8.08 p Tetrahydroaldosterone (I) was isolated from which was characteristic of equatorial 3-acetoxy extracts of combined incubates by chromatography steroids.I8 The 301,5/3 configuration was assigned on paper (ethylene dichloride-ethylene glycol, to lacetone I11 by excluding the other possible &ido = 0.312 and on Celite (toluene-ethyl acetate 3-equatorial isomer. Hydrogenation of aldosterone 9:1-methano1:water 1:1, R A l d o = 0.6) as a white etiolactone (IV) over platinum oxide in acetic acid amorphous solid, melting range 107-114,01a [ a ] " ~ gave the known 3P,501, lactone VI,1791g m.p. 24150' (CHC13 (C = 0.9), 2'c;:X: 2.78, 5.88 (weak) 243O, which depressed the melting point of lactone p ; C, 69.00; H, 8.88. Acetylation with acetic 111. The infrared spectra of the pair of isomeric lactones (111 and VII) and of their acetoxy derivaOR CHZOR tives (V and VII) differed in the fingerprint region. @O I I & o- . ,R o /O\ Only lactone VI1 formed an insoluble digitonide. The metabolite of aldosterone (3.8 mg.), isolated as described4 from the urine (17 day pool) of a patient with cirrhosis of the liver, was compared with tetrahydroaldosterone (3cr,5p). Three derivaRO-' tives (11, 111, V) of the isolated and the synthetic I, R = H 111, R = H steroid were prepared. Both ketols as well as their 11, R = AC V, R = A c corresponding derivatives had identical infrared spectra and chromatographic running rates. Both ketols also were shown to be identical by double isotope techniques. Following the administration of d-aldosterone-7-H3, the major radioactive moiety was isolated from the glucuronide fraction of urine and mixed with tetrahydroaldosterone (3a,5P)-4-C14,and derivatives 11, I11 and V of the IV VI, R:H doubly labeled mixture were prepared. The H3: VII, R = AC C1* ratios of the keto1 and its derivatives agreed 5%. within ( 5 ) The separation of the metabolite of 18-hydroxycorticosterone
+
from the aldosterone metabolite and the confirmation of its structure b y synthesis of the y-lactone obtained by periodate oxidation will be reported separately, S. TJlick and K. Kusch. In certain states of hyperaldosteronism such as cirrhosis of the liver and maligant hypertension, the secretion of 18-hydroxycorticosterone as measured by the excretion of this metabolite was greater than 4.0 mg. per day, exceeding that of aldosterone. (6) G. M. Tomkins, J . Bid. Chem., 225, 13 (1957). (7) J. T. August, Biochcm. Biophys. Acto, 48, 203 (1961). (8) p H 7.4 containing these concentrations of salts in m. moles per liter: NaCl 110, KCI 4.6, MgSOl 1.0, NaxHPOl 22. (9) We wish to express our appreciation to Dr. C. H. Sullivan, Ciba Pharmaceutical Products, for providing the d-aldosterone-21monoacetate used in this work. (10) Generated from T P N (triphosphopyridine nucleotide) by the coupled oxidation of glucose-6-phosphate in the presence of glucose-6phosphate dehydrogenase. (11) When cortisol was incubated with the same enzyme preparation, under the same conditions, the only product was 3a,17a,llp,21-tetrahydroxy-20-ketopregnane. (12) H values refer to the distance migrated by the sample on paper chromatogram relative t o the reference steroid. Abbreviations: Aldo 0 aldosterone, DOCA = desoxycorticosterone acetate, An
4-androstene-3,17-dione.
-
(13) Melting points (uncorrected) were determined on a micro hot stage. Only the hydroxyl and carbonyl regions of the Infrared are reported here. Elemental a n a l y s ~ swere performed by Sfhnsrzkopf Microandytical Lsboretoriar.
(14) The acetate number was determined with "-labeled acetic anhydride. T h e number of moles of steroid present in the sample was determined from its Cl4 content. T h e c.p.m. per mole for tetrahydroaldosterone and its derivative was determined from the specific activity of aldosterone-4-Cl4 used as a substrate in the incubation. (15) E. A. Ham, R. E. Harman, N. G. Brink, and L. H. Sarett, J . A m . Chem. Soc., 77, 1637 (1955). (16) Measured colorimetrically with chromotropic acid: D. A. MacFayden, J . B i d . Ckem., 158, 107 (1945). (17) S. A. Simpson, J. F. Tait, A. Wettstein, R. Neher, J. v. Euw, 0. Schindler, and T. Reichstein, Helu. Chim. Acta, S7, 1200 (1954). (18) R. N. Jones and F. Herling, J . A m . Chum. SOL,78, 1152 (1956). (19) M. M. Pechet, R. H. Hesse and H. Kohler, ibid., 83, 5251 (1960). VETERANS L4DMINISTRATION HOSPITAL
STANLEY ULICK
BRONX68, NEW YORK KATHRYN KUSCH DEPARTMENT OF MEDICINE J. T. AUGUST STANFORD UNIVERSITY SCHOOL OF MEDICINE PALO ALTO, CALIFORNIA RECEIVED SEPTEMBER 26, 1961
VINYL AZENE CHEMISTRY: FORMATION OF AZACYCLOPROPENE
Sir: As a continuation of our studies of the chemistry of the monovalent nitrogen speciea, ozenei,llathe
COMMUNICATIONS TO THE EDITOR
4484
investigation of the decomposition products of vinyl azides was undertaken. Presumably the vapor phase pyrolysis1c of a-azidostyrene (I)
I
I1
I11
Yol. s3
H, 4.87; N, 29.181. The n.m.r. spectrum (CCL) showed the non-equivalent methylene hydrogens as two single sharp lines a t 5.68 and 4.687 and the phenyl hydrogens as a multiplet in the region 2.77. The data confirm the structure of I. BELLTELEPHONE LABORATORIES GERALD SMOLINSKY MURRAYHILL,NEWJERSEY RECEIVED SEPTEMBER 8, 1961
could lead to the formation of styryl azene (11). Of the several conceivable reaction paths open to THE PHOTOLYSIS OF 9-AMINOPYRIDINES AND 9-PYRIDONES such an intermediate, i t was thought that a likely possibility might be ring closure to give an aza- Sir: A recent report‘ from these laboratories described cyclopropene. Quite gratifyingly, the pyrolysis of I resulted in a 65% yield of 2-phenylazirine (111). the irradiation of 2-aminopyridine hydrochloride That I11 arises by cyclization of I1 appears quite with sunlight through Pyrex glass to give a photoplausible since any reaction of the double bond of I isomer with very interesting chemical and physical with the azide moiety would more likely involve the properties, which was considered to possess the terminal nitrogen atom of the azide leading t o a valence-bond tautomeric “Dewar” structure Is. triazole in a manner analogous to tetrazole forma- An analogous photo product (Ib) was obtained from tion from imino azide derivativesGa The ready the hydrochloride of 2-amino-5-chloropyridine. formation of this novel unsaturated ring system is Reduction of either I a or I b gave the same hydroessentially analogous to the formation of cyclo- genation product, considered to be 11. In spite of the evidence which strongly supported these propenes from alkenylcarbenes.4 Although azacyclopropenes have been postulated structures, we now wish to report that the photoas intermediates in the Neber reaction: only one products are, in fact, dimers (IIIa and b). such compound, 2-(2,4-dinitrophenyl)-3-methyl-2- Conclusive evidence in favor of these revised azirine, has been isolated and its structure estab- structures was obtained by alkaline hydrolysis of lished.6 Thus the formation of I11 from I is not the tetrahydro derivative IVa (previously cononly interesting in itself, but represents the first sidered to be 11) to give a product (V) identical in example of what appears to be a useful method for every respect with the tetrahydro derivative of the preparing azacyclopropenes-work is now in prog- photodimer (VIa) of 2 - p ~ r i d o n e . ~Since n.m.r. studies on IIIb conclusively exclude the possibility ress along these lines. 2-Phenylazirine (111) [calcd. for CsH7N: C, that i t could be a 3,4-dimer,l the conversion of IV 82.02; H, 6.02; N, 11.96; mol. wt., 117. Found: tovconstitutes convincing proof that the 2-pyridone C, 81.87; H, 6.23; N, 11.64; mol. ~ t . 1261 , ~ is a photodimer (VIa) (and its N-methyl derivative VIb, colorless, thermally unstable, irritating liquid of to which i t has been related by methylation) is b.p. 80” (10 mm.). Its infrared spectrum (CCla) also a 3,6- rather than a 3,4- dimer as previously shows strong C=N absorption a t 5.74 fi6 and its suggested.a This conclusion is consistent with the ultraviolet spectra in cyclohexane and in ethanol observation that N,6-dimethy1-5,6-dihydro-Z-pyrishow maxima a t 239 (13,000) and 242 mp (13,000), done does not dimerize upon irradiation. Of the respectively. The n.m.r. spectrum (CClJ showed four 3,6-structures which can be written for the the methylene hydrogens as a sharp line a t 8.357 2-pyridone dimers, only two (VI and VII) are conand the meta-para and ortho phenyl hydrogens as sistent with the negligible dipole moment previously multiplets in the regions 2.4 and 2. 17, respectively, found for VIb. Of these, VI1 can be ruled out, with intensity ratios of 2 :3 :2, respectively. Boil- since the corresponding structure in the 2-aminoing a solution of I11 in acidic aqueous ethanol pyridine series (VIII) would involve unfavorable resulted in the formation of 2,5-diphenylpyra~ine~proximity of positive charges. We had previously reported that solution of I1 in 30y0 yield. These data are consistent with the (now shown to be IVa) in dilute alkali resulted in assigned structure for 111. a-Azidostyrene was prepared as follows : styrene the rapid separation of a product which, on the dibromide was treated with one mole of sodium basis of all available evidence, was considered to be azide in dimethylformamide and the resulting cis-2-aminocyclobutanecarboxamide (IX) . This bromoazide dehydrobrominated with potassium was not an unreasonable proposal, since I1 would tertiary butylate in benzene. The styryl azide was be expected to undergo a facile hydrolytic ringpurified by chromatographylc [calcd. for CsH7Ns : cleavage to a less strained system. However, we C, 66.19; H, 4.86; N, 28.95. Found: C, 65.96; now have been able to show that this product is actually an extremely stable dihydrate (IVb). (I)(a) G. Smolinsky, J . Am. Chcm. SOC.,84, 4717 (1960); (b) ibid., 88, 2489 (1961); ( c ) J . Org. Cham., 26, 4108 (19131). (2) D. H. R . Barton and L. R . Morgan, Jr., Proc. Ckcm. Soc., 206
(1961).
( 1 ) E. C. Taylor, W. W. Paudler and I . Kuntz, Jr., J . A m . Ckcm. Soc., 88,2967 (1961). (2) E. C. Taylor and W. W. Paudler, Tcftahcdron Lctfars, 26, 1
(3) F.Eloy, J . Org. Chcm., 26,952(1961). (4) G.L.Closs and L. E. Closs, J . Am. Ckcm. Soc., 83,2015 (1961). (5) P. W. Neber and A. Burgard, Ann., 493, 281 (1932); P . W. Neber and G. Huh, ibid., 616, 283 (1935). (6) D. J. Cram and M. J. Hatch, J . A m . Chcm. Soc., 76, 33 (1953). (7) Cryoscopic determination in cyclohexane. (8) Identity of this muterial established by comparison with an authentic sample.
(3) We would like t o express our appreciation to Professor P. deMayo, Department of Chemistry, University of Western Ontario, London, Canada, for revealing to us in advance of publication (Tetrahedron Lcffers, in press) his conclusion that the N-methyl-2-pyridone dimer possesses structure VIb. In point of rime, Professor deMayo’s conclusion on this revised structure preceded our own, and knowledge of his results greatly facilitated our own investigations.
(1960).
