J. J. J

the pregnancy urine of women afflicted with corti- cal tumors. In the description of our investigation of the carbinol fraction of mares' pregnancy ur...
18 downloads 0 Views 345KB Size
July, 1938

KETONESFROM MARES)PREGNANCY URINE

1559

[CONTRIBUTION FROM THE SCHOOL OF CHEMISTRY AND PHYSICS OF THE PENNSYLVANIA STATE COLLEGE]

Sterols. XXXVI. Ketones from Mares’ Pregnancy Urine BY RUSSELLE. MARKER,ELMERJ . LAWSON, EUGENEL. WITTLE AND HARRYM. CROOKS

As part of an extensive investigation of urines semicarbazone mixture from 30,000 gallons (114,from various animal sources, we have been study- 000 liters) of mares) pregnancy urine yielded a ing during the past two years the steroid content water solution (A) containing Heard’s waterof mares’ pregnancy urine. We already have re- soluble semicarbazone, an ethereal solution (B) ported the presence of pregnanediol, aRo-preg- containing chiefly non-steroidal semicarbazones, nanediol,‘ uranetriol,2 pregnanetriol,2 and urane- and a solid crude semicarbazone mixture. The diol3 in the carbinol fraction from mares’ preg- latter on systematic fractional crystallization nancy urine. The presence of a pregnanetriol from alcohol yielded solid semicarbazones I, 11, which is probably identical with our pregnanetriol, I11 and IV and a mother liquor 2. Each fraction was reported earlier by Haslewood, Marrian and was hydrolyzed with alcoholic sulfuric acid, and More recently these authors also have further purifications were effected by high-vacuum reported6 the presence of pregnanetriol-3,17,20 in distillation, half-succinate separation, and fracthe pregnancy urine of women afflicted with corti- tional crystallization. cal tumors. In the description of our FLOWSHEET investigation of the carbinol fraction of Non-phenolic fraction mares’ pregnancy urine we reported the from mares’ pregnancy urine separation of a ketonic fraction from this after removal of triols. source. Since Heards has announced the Girard’s isolation of a number of ketones from reagent mares’ pregnancy urine, we feel obliged to , I 1 report the progress we have made in the $ $ Ketonic fraction course of a similar study. We are able to Non-ketonic fraction I confirm the presence in mares’ urine of the ketone, m. p. 252’) reported by Heard SemicarbaSee Paper XXXVII and in addition we have found a number zine HCl of other steroids present. While a direct I I I comparison of our compound with Heard’s J. 9 $ Ether washings Insol. semicarbaWater-soluble ketone has not been made, we believe semicarbazones B zones that the two are identical in view of their A apparently similar properties. However, Fractionation molecular weight determinations of our from alcohol substance (mol. wt. 672) gave a value twice that of a steroid, indicating that J. J. J. SemicarbaSemicarba- Semicarba- Sernica!bait may be a combination of two sterol Filtrates zone I11 z zone IV zone I1 zone I molecules through their oxygen atoms. We were surprised to find that all of the semiThis would account for the inactivity of these atoms. We are not planning further work on carbazones yielded both hydroxylated and hythis compound since we feel that Dr. Heard is droxyl-free ketones. This occurrence of hydroxylfree ketones in mares’ pregnancy urine is in conentitled to priority in its discovery. The accompanying diagram shows the scheme trast to the almost complete absence of such keof isolation of our various ketones. The crude tones in human pregnancy urine.T Possibly the course of reduction of progesterone and other (1) Marker, Kamm, Crooks, Oakwood, Lawson and Wittle, THIS hormones in the mare follows a somewhat different JOURNAL, 69, 2297 (1937). (2) Marker, Kamm, Crooks, Oakwood, Wittle and Lawson, path, or else reduction may be more complete in ibid., 60,210 (1938). man. (3) Marker, Rohrmann and Wittle, ibid., 60, 1561 (1938). (4) Haslewood, Marrian and Smith, Biochem. J . , 88, 1316 (1934). From the most insoluble semicarbazone frac(5) Marrian and Butler, J . Biol. Chern., 119, LXVI (1937).

.G

(6) Heard, THISJOURNAL, 60, 493 (1938).

(7) Marker, Kamm and McGrew, ;bid., 68, 616 (1937)

tion (I) we have isolated two substances: a ke- tractable. Semicarbazoiies B and Z yield a mixtone which corresponds in properties to Heard's ture of neutral and hydroxylated ketones. The ketone, C:lgHzeOY,in. p. %52', but has a molecular hydroxyketone mixture from semicarbazone Z weight of 672, : i d cLIld-pregiianedionp. We havc r i o t s r i o t precipitate digitonill, arid therefore no also some indication id the presence of a t least ,i(lo-pre,orienol-:i( 6:)one-20 is present. Distillaone other ketone in this hydrolyzed semicarbarntie tion of the neutral aiid hydroxylated ketones from mixture (I). The preseiice of ullo-pregnanedione B gives chiefly ketones subliming below 125', and and pregnanedione in mares' pregnancy urine possessing characteristic odors. Apparently allends further support to the hypothesis that most io steroidal ketones are present in fraction progesterone is the source of most of the steroidal I3. Because of the small amounts of ketones prescarbinols and ketones, such as pregnanediol, do-pregnanediol, pregnarietriol-3(Lu),4(8),20(~1),~ent in mares' pregnancy urine, a more extensive dlo-pregnanol-3(a$-one-20, pregnanol-&(a)-one- investigation o f 150,ooO gallons (570,000 liters) o f 20, pregaanediol-3(B),20(cr),Y pregnenediol-3@),- uriiie is now in progress, and the results of this P ~ ( L Y ) pregnanedione, ,~ and allo-pregnano1-8(B1 - study will be reported shortly. one-20. Experimental Part Semicarhazones I1 aiid 111 proved to contaiii Isolation of Ketonic Fractions.--The non-phenolic fracthe same components. From the hydrolyzed semicarbazoiies there wa5 obtained a new hy-- tion from 30,000 gallons (114,000 liters) of mares' pregdroxyketone, uranol- 1I-one-3, the known preg- riaucy urine from which the uranetriol and pregnane trio1 had ixen removed was dissolved in about 2.5 liters of alcoIiaiiedione, and uElo-pregnanol-3(B)-one-20. The h u l and heated for one half hour with 400 g. of Girard's structure of uranol-11-one-3 is proved by its coni- reagent. Ice and water were added and the non-ketonic position and oxidation to uranedioiie I t did iiot iiiaterial estracted thoroughly with ether. The aqueous form a half-succinate ester, in agreement with the layer was acidified with an excess of hydrochloric acid and usual postulated inertness of a hydroxyl group a t warmed for one hour on a steam-bath The ketonic matter was estracted with ether and the ether extract washed CII. It gave a Zimmermann test showing the and concentrated. The residual oil was treated again presence of a carbonyl group at Cd. The isola- with Girard's reagent to remove a small amount of inert tion of this substance lends support to the hy- material which was carried through mechanically. About pothesis that uranedioiie and the uranediols 30 g. of ketonic material was obtained. ?'he ketone mixture was dissolved in 300 cc. of alcohol ( U , u: arid S , a ) may be related to some hormone of arid boiled for one hour with 50 g. of semicarbazide hydroas yet unknown functioii,"' although it is possible chloride and A0 g. of sodium acetate. The mixture was that these substances may be derived from the distilled to remove almost all the alcohol and the resulting cortical hormone. The isolation of dlo-preg- paste W d S stirred with water, filtered, and washed with nallol-3(P)-one-%Ofroiri iiiares pregnancy urine xa1c.r and with ether. The water washings were set aside is interesting in s-iew of the fact that it was first as waicarbazone fraction A , and the ether washings as sviiiicarbawne fraction B. The solid semicarbazone mixisolated from corpus luteum extracts." The turt', amounting to about 15-20 g., was fractionated from structure of our substance was proved by the alcohol. This gave semicarbazone solution Z, and semicomposition, melting point, precipitatioa with carhazones 1, 11, 111, and IV Semicarbazone solutions A digitonin, and oxidation to allo-pregnanedione. and Z were evaporated to dryness to yield semicarbawnes This is the first 3(B)-hydroxyketone to be isolated A and 2 . These six semicarbazone fractions were hydrolyzed separately by boiling for one hour with fifty parts of from urine, although its occurrence had been alcohol and fifteen parts of 1 : 2 dilute sulfuric acid. Each foreseen in the light of a proposed course of the hydrolysate was diluted with water, extracted with ether, thc ether layer washed with water, and then evaporated to reduction of progesterone. We have not as yet completed our investigation dryness. With the exception of the ketones from semiof the ketones from semicarbazones A, B, and Z, carhazone I, the residues were converted into the acid succinates in order to separate hydroxylated ketones froni since the mixtures have proved to be rather In- OH-frce ketones. ( 8 ) Marker, Kamm, n'ittle Oakwood and Lawbon, r H I S J o ~ ' R \ 60,1067 (1938). 10) Marker and Rohrmann, ihtd , 60, 1565 (1938)

AI

,

(10) Marker, Lawson, Rohrmnn and Wittle. rbtd., 60,1666 (1938) [ t i ) U'intersteiner and Allen J B:ol Chem , 107, 321 (1934) (121 Hartmartn and Wettstein, Hela Chtm A c l a , 17, 1365 (19341 113) Butenandt, Westphal and I-lohlney, K fihvProi C h i n 237, S4 (1034) [ ! 4 l Slotta Kn\chig and I rli B., 67, LLiO 11!'J41

allo-Pregnanedione and Ketone I from Semicarbazone I. --Hydrolysis of semicarbazone I gave 1.4 g. of solid ketones which could not be separated readily by crystallization, Accordingly, the total ketonic material was distilled in a high vacuum and fractions collected a t bath temperature up to 110, 110-140 and 140-180°. The first and third fractions were too small to be investigated f i r -

156 I

URANEDIOLFROM MARES'PREGNANCY URINE

July, 1935

ther. Considerable residue remained in the distillation vessel. The fraction subliming at 110-140° was crystallized repeatedly from dilute acetone and yielded allopregnanedione, m. p. 196-200°, which gave no depression in me!ting point when mixed with authentic allo-pregnanedione. Anal. Calcd. for CZI&,O~:C, 79.7; H, 10.2. Found: C, 79.3; H, 10.3. The non-distillable residue in the sublimation flask was crystallized from benzene-pentane, and yielded a crude ketone (ketone I) which melted a t 220'. This ketone approximated the properties of Heard's ketone (m. p. 252') upon further crystallization. Anal. Found: C, 76.06; H, 9.85; mol. wt. (several determinations), 672. The semicarbazone prepared from this ketone melted at 300". allo-Pregnanol-3( p)-one-20.-A preliminary investigation of the hydroxyketone from semicarbazones I1 and I11 showed that one substance was present in both fractions, which were therefore combined. This hydroxy ketone was sublimed in a high vacuum a t a bath temperature of 185', and the sublimate crystallized from carbon tetrachloride to give a ketone melting a t 193'. This ketone depressed the melting point of allo-pregnanol-20(~)-one-3(m. p. 195'), showed no unsaturation to bromine in acetic acid and gave a hpavy precipitation with an alcoholic solution of digitonin. A portion oxidized with chromic acid in the usual manner gave allo-pregnanedione, m. p. 197-199". and showed no depression with an authentic sample of allo-pregnanedione, m . p . 200'. Anal. Calcd. for C21H3202: C, 79.7; H, 10.2. Found: C, 79.7; H, 10.1. Uranol-11-one-3, Pregnanedione, and Ketone I1 from Semicarbazones I1 and 111.-The ketones from semicarbazones I1 and I11 were distilled separately and fractions collected a t bath temperatures of 135-160', 160-200", leaving considerable amounts of glassy residue. The fraction 135-160' was crystallized from diluted acetone to

(CONTRIBUTION FROM

THE

yield a solid melting over the range 100-135". The material in the mother liquors was then crystallized repeatedly from dilute acetone, sublimed in a high vacuum, and then crystallized from ether-pentane to give pregnanedione, m. p. 1 1 8 O , which did not depress the melting point of an authentic sample. Anal. Calcd. for G&zOt: C, 79.7; H, 10.2. Found: C, 79.5; H, 10.7. The fraction subliming a t 160-200" was a mixture which could be separated only with difficulty. Repeated crystallization from dilute alcohol and dilute acetone yielded a ketone which melted a t 165", and depressed the melting points of uranedione (m. p. 180') and allo-pregnanedione (m. p. 200"). Anal. Calcd. for CzlH&: C, 79.2; H, 10.8. Found: C, 79.2; H, 10.8. It gave a gelatinous semicarbazone melting a t 250" with decomposition. Oxidation of a sample with chromic acid in the customary fashion yielded uranidione, m. p. 175-176', which showed no depression with an authentic sample of uranedione. Anal. Calcd. for CZIHJZO~: C, 79.7; H, 10.2. Found: C, 79.7; H, 10.5. The glassy residues from sublimation of the neutral ketones up to 200' were combined and crystallized from benzene-pentane to yield about 100 mg. of a ketone melting a t 115-120". I t did not precipitate with digitonin, Anal. Found: C, 74.3; H,9.65; mol. wt., 462.

We wish to thank Dr. Oliver Kamm and Parke, Davis and Company for their generous help and assistance in various phases of this work.

Summary The ketonic fraction of mares' pregnancy urine yielded pregnanedione, allo-pregnanedione, allopregnanol-3(P)-one-20, and uranol-11-one-3 in addition to the ketone reported by Heard. STATECOLLEGE, PENNA.

RECEIVED MARCH16, 1938

SCHOOL OF CHEMISTRY AND PHYSICS OF THE PENNSYLVANIA STATECOLLEGE 1

Sterols. XXXVII. Uranediol from Mares' Pregnancy Urine BY RUSSELL E. MARKER, EWALD ROHRMANN AND EUGENE L. WITTLE In paper XXIX' of this series the isolation of uranetriol from mares' pregnancy urine was reported. The urane compounds were shown to differ in structure from pregnane compounds in the configuration a t CS. In the pregnane series the hydrogen a t CSis trans to the methyl group a t Cl0, while in the urane series the hydrogen atom is considered to be cis to the methyl group. Clemmensen reduction of uranetrione gave the parent hydrocarbon, urane, which is different from either (1) Marker, et al., Tnrs

JOUHNAI..

60, 1061 (1938).

pregnane or allo-pregnane. Evidence was presented to show that the hydroxyl groups in uranetriol are located a t the 3-, 11- and 20-positions. The presence of uranediol in mares' pregnancy urine was indicated by the fact that oxidation of a mixture of crude diols gave a diketone which on Clemmensen reduction yielded the hydrocarbon urane. We have now succeeded in isolating uranediol from the digitonin precipitated fraction of the ketone-free carbinol fraction of mares' pregnancy