STEROID EXCRETION IN A CASE OF ADRENOCORTICAL CARCINOMA
III. THE ISOLATION OF 45-PREGNENEDIOL-3(6), 17(8)-ONE-20 AND OF 17a-METHYL-45-D-HOMOANDROSTENEDIOL-3(8), 17a(a)-ONE-17*
BY H. HIRSCHMANN AND FRIEDA B. HIRSCHMANN
(From the Department of Medicine, School of Medicine, Western Reserve University, and the Lakeside Hospital, Cleveland)
(Received for publication, August 19, 1946)
The examination of the non-ketonic fraction from the urine of a boy with an adrenocortical tumor has led to the isolation of two substances which are structurally related to dehydroisoandrosterone, a 45-androstene- triol-3(6),16,17 and 45-androstenediol-3(6),17(a) (1, 2). In analogy to the metabolic interrelationships of the estrogens it was suggested that dehydroisoandrosterone and androstenediol are mutually interconvertible in the body, while the triol was considered as a metabolite of dehydroisoan- drosterone. Since then several reports from other laboratories have ap- peared which lend some support to this concept. It has been shown by Ruzicka, Prelog, and Wieland (3) that the structural analogy between an- drostenetriol and estriol extends to the configuration of the glycol grouping. Marrian and Butler (4) have reported the presence of androstenetriol in normal urine, and Mason and Kepler (5) have obtained it from patients with adrenal tumors. The presence of both the triol and dehydroisoan- drosterone in the same urine is, therefore, no uncommon phenomenon. Androstenediol also has been found in other cases of neoplasm of the adrenal cortex (5, 6). Miller and Dorfman (7) have demonstrated the metabolic conversion of this substance into dehydroisoandrosterone in the guinea pig, while evidence for the reverse reaction has been secured in man by Mason and Kepler (8). In view of this situation it seemed of interest to determine the amounts of dehydroisoandrosterone in our urine extracts and to search for other metabolites of this substance in the ketonic fraction.
A solution of the ketonic compounds in benzene yielded a small crop of a crystalline precipitate which was separated. The remainder of the ma- terial was chromatographed on alumina. This fractionation yielded @-3- chloro-45-androstenone-17, dehydroisoandrosterone, and a compound which had not previously been obtained from natural sources. It melts at 293- 299° and is only sparingly soluble in the common organic solvents. The analyses of this substance and of its acetyl derivative indicated the com- position C21H32O3 for the parent compound. It forms a monoacetate and
*This investigation has been supported by grants from the Commonwealth Fund, the Price Mckinney, Jr., Fund for Leukemia Research, and from the American Cancer Society on the recommendation of the Committee on Growth of the National Research Council.
a monooxime. 1 oxygen atom, therefore, is present as a hydroxyl group and another as a carbonyl group. The function of the 3rd as a tertiary hydroxyl group was established by oxidative degradation. The acetate rapidly decolorized an equimolar amount of bromine at room temperature. While this does not prove that the ketone contains an ethylenic double bond, the reaction is consistent with this interpretation. The presence of an olefinic linkage is also in harmony with the analytical data if they are
70
60
Molecular Extinction Coefficient (8)
50
3
1
40
30
1
20
2
10
3
240
2.60
280
300
320
340
Wave length in mu.
considered in terms of a tetracyclic structure. Treatment of the bromin- ated acetate at room temperature with an amount of chromic acid equiva- lent to 1.5 atoms of oxygen followed by debromination yielded about 80 per cent of acidic material. The analysis of the main reaction product indicated the composition C23H34O5, which differs from that of the starting compound only in the presence of 1 additional oxygen atom. The acetoxy and the carboxyl groups account for 4 of the oxygen atoms. Lack of
material prevented the characterization of the 5th oxygen by conventional methods, but its function as a keto group was demonstrated conclusively by spectrographic examination which revealed a low maximum at 287 mu (Fig. 1). The oxidation of an acetate C23H3404 to an acetoxy keto acid with the same number of carbon and hydrogen atoms can be explained only if the starting compound contains a tertiary hydroxyl group adjacent to a keto group. Both groups must be attached to a ring which is opened on
1
2
3
4
5
6
0.300
0.200
Extinction
100
420
450 470 490 520 550
590
Wave length in mu.
oxidation. This reaction converts the tertiary hydroxyl into a keto group and the original keto group into a carboxyl group.
The acetate of the isolated compound is chromogenic in the von Bittó- Zimmermann reaction. Under the conditions of Holtorff and Koch (9), an intense purple color develops which starts to change to brown at the end of about 45 minutes. Spectrographic measurements taken after 105 minutes are presented in Fig. 2 (Curve 1). A positive test is regarded as
indicative of an active methylene (10)1 such as is present in the CO·CH2 grouping. Investigations in the steroid field have shown, however, that those ketones which contain this structure in a central ring produce no or only a very feeble color in this reaction (12, 13, 9). Although it is possible that a neighboring hydroxyl group may influence the reactivity of the keto methylene grouping,2 it seemed justified to search for the structure of the isolated compound primarily among those steroids and steroid deriva- tives that contain a keto group in a terminal ring. Few of the compounds that remain for further consideration are likely to occur as urinary con-
CH3
Į C=0
OH
RO
3
Ia Ib
R=H R=Ac
CH3
OH =0
CH3 c=0 COOH
Br2-Cr03
Zn
”
0
RO
Br2-Cr03
RO
RO
IIa R=Ac IIb R=H
Zn
Ma
R=Ac R=H
a R=Ac
b R=H
СНЕСН t- Am OK
OR
0Ac
C-CH3
С=СН
Y
OH
C=CH
İl 0
RO
Ac0
HO
KOH VIIa R=R=Ac
$NH2+H20+HgC12
H25
VI +
AC20
V
KOH
VIIb R=Ac; R’=H
pyridine
VIIc R=R’=H
stituents. These include a steroid rearrangement product first synthe- sized by Ruzicka and Meldahl (14) which has since been identified (15-17) as 17a-methyl-45-D-homoandrostenediol-3(8),17a(a)-one-17 (IIb). Its
1 This generalization of Reissert (10) appears not to be fully justified since von Bittó (11) lists a number of positive reactions with a, 6-unsaturated and with aromatic aldehydes which do not contain an active methylene group.
2 No enhancing effect has been noted if the keto group is in the 6 position, as Hol- torff and Koch (9) have listed a 5-hydroxy-6-ketosteroid among the non-chromogenic compounds. It is presumed that this substance which is designated as “A5-3-acet- oxy-5-hydroxy-6-ketocholenic ethyl ester” does not contain a double bond between carbon atoms 5 and 6.
properties as reported by Shoppee and Prins (17) are in satisfactory agree- ment with those of the isolated product.3 These workers described three different modifications with transition points at 260° and 290° and a melting point of 302-305°, but failed to confirm the observations of all other in- vestigators who recorded melting points varying from 277° to 282° (14, 20- 23). We, therefore, prepared a reference specimen by the method of Shoppee and Prins which is summarized in the accompanying diagram (IVb — > V → VI → VIIa → IIb). It utilizes reactions first described by Stavely (21, 24) and by Ruzicka and Hofmann (25). The final product agreed in every respect with the isolated substance. The comparison in- cluded the melting points and mixed melting points of the free compound, of the acetate (IIa), of the oxime, and of the oxidation product (IIIa), as well as the behavior of the acetate in the dinitrobenzene reaction (Fig. 2). It is concluded, therefore, that the isolated compound is 17a-methyl- 45-D-homoandrostenediol-3(3),17a(a)-one-17.
The investigations of Shoppee and Prins leave no doubt as to the chemi- cal identity of their product with the lower melting preparations obtained elsewhere. Since it seemed possible that the presence or absence of minor amounts of contaminants may determine the crystal form, we have pre- pared this D-homosteroid also by a second procedure. Treatment of 45-pregnenediol-3(6),17(a)-one-20 3-monoacetate (VIIb)4 with methanolic potassium hydroxide again yielded the higher melting product, whereas Stavely (22), who first described this reaction, reported a melting point of 280°. The shorter procedure of Stavely, therefore, can also be used for preparing the higher melting form of the methylhomoandrostenediolone.
Since 17a-methyl-45-D-homoandrostenediol-3(3), 17a(a)-one-17 does not contain the steroid nucleus but is readily prepared from compounds pos- sessing this structure, it seemed probable that the isolated D-homosteroid might have formed from a 45-pregnenediol-3(3),17-one-20 during the fractionation procedure. This concept received strong support by the isolation of a second substance with the composition C21H32O3, which was identified as 45-pregnenediol-3(3), 17(8)-one-20 (Ia). The presence of a reactive hydroxyl group in this compound was ascertained by the formation of a monoacetate, while a keto group was demonstrated by spectrographic analysis (Fig. 1). Since the position of the maximum (294 mu) was closer to the visible region than the generalizations of Rice (26) and of Lowry and Lishmund (27) had led us to expect, we also investigated the spectrum of the 17-epimer. It showed a maximum at 295 mu (Fig. 1). The isolated
3 This fact was not known to us when the structure of the isolated compound (18) was discussed at the Thirty-seventh annual meeting of the American Society of Biological Chemists in Atlantic City, since the melting point was erroneously re- ported in Chemical Abstracts (19) as 275-277°.
We are indebted to Dr. H. E. Stavely for this preparation.
acetate (Ib) was brominated, oxidized with chromic acid, and debromi- nated. The main reaction product was found in the neutral fraction. It reacted with m-dinitrobenzene and alkali like a 17-ketosteroid (Fig. 2) and was identified as dehydroisoandrosterone acetate (IVa) by melting point, mixed melting point, and by conversion into the benzoate. Since the starting compound is not a 17-ketosteroid (Fig. 2), it follows that its keto group must be present in a side chain attached to C-17. This degradation of a compound C21H32O3 establishes it as a 45-pregnenediol-3(6), 17-one-20 but leaves the configuration at C-17 undetermined. The two stereoisomers with this structure (Ia and VIIc) have been prepared by partial synthesis and show widely differing properties (22-24, 17, 28-30). Comparison of the melting points and of the optical rotation leaves no doubt that the isolated compound belongs to the 8-17-hydroxy series and thus possesses the same spatial arrangement at C-17 as all 17-hydroxy steroids isolated from the adrenal cortex. Through the kind cooperation of Professor Reichstein it was possible to confirm the identity of our product by mixed melting point determination with synthetic samples of the diolone and its monoacetate. A direct comparison with authentic material was partic- ularly desirable, as the wide ranges of fusion which we consistently ob- served with our preparations were in marked contrast to the sharp melting points found by Reichstein and his coworkers (29, 30). However, the synthetic preparations when tested under our conditions (open capillary) also melted over a wide interval. While the reason for this discrepancy remains obscure, it seems justified to conclude that this behavior may be due to factors other than the presence of contaminants.
The two stereoisomeric forms of 45-pregnenediol-3(8), 17-one-20 acetate can readily be distinguished by means of the von Bittó-Zimmermann re- action. If the hydroxyl group at C-17 is in the @ position, the compound yields a reddish brown pigment with a broad maximum near 490 mu. The 17-epimer develops a transitory purple color which changes to brown. The time relationships of this color shift as well as the absorption char- acteristics of the final reaction product, which shows no maximum in the the visible region (Fig. 2), are in striking agreement with corresponding data on 17a-methyl-45-D-homoandrostenedicl-3(8), 17a(a)-one-17 acetate. This close accord between two substances with different chromogenic groups is explained most readily if it is assumed that 45-pregnenediol- 3(6), 17(a)-one-20 rearranges to the homosteroid prior to condensation with dinitrobenzene. These observations suggest that the von Bittó- Zimmermann reaction may prove to be of value in ascertaining the con- figuration of other 17-hydroxy steroids with an aceto group at this carbon atom.
Conditions which lead to enlargement of Ring D in 17-hydroxy-20-keto- steroids include treatment with acid (31), alkali (22, 31, 32), and alumina
(24, 17). Since 45-pregnenediol-3(6), 17(6)-one-20 was subjected to the action of all three factors, the step responsible for the suspected conversion into 17a-methyl-45-D-homoandrostenediol-3(3),17a(a)-one-17 is not read- ily determined. Treatment of 8-17-hydroxyprogesterone with a boiling mixture of glacial acetic and concentrated hydrochloric acids effects some isomerization to 17a-methyl-44-homoandrostenol-17a(6)-dione-3,17 (31, 17). Although these reaction conditions are not identical with those prevailing in the acid hydrolysis of urine, it seems noteworthy that this rearrangement product differs from the isolated D-homosteroid in the configuration at C-17a. Since 8-17-hydroxyprogesterone remains largely unchanged when methanolic solutions are heated with moderate amounts of potassium hydroxide (31), it seems quite unlikely that the partition of the urinary extract into neutral and acidic material could have led to the formation of significant amounts of D-homosteroid. The effect of aluminum oxide on 8-17-hydroxy-20-ketosteroids has not been adequately studied. It is ap- parent, however, that they are far less sensitive than their 17-epimers (17). Von Euw and Reichstein (31), who isolated 17a-methyl-44-D-homoan- drostenol-17a(a)-dione-3,17 from adrenal extracts, considered alumina as a factor in its formation from 8-17-hydroxyprogesterone but were unable to exclude other causes. We obtained the highly insoluble 17a-methyl- 45-D-homoandrostenediol-3(8),17a(a)-one-17, chiefly from fractions that had been in protracted contact with large amounts of aluminum oxide, while the benzene-insoluble ketones which had not undergone such pro- longed treatment contained mostly 45-pregnenediol-3(8),17(8)-one-20. We suspect, therefore, that the action of alumina is largely responsible for the presence of the D-homosteroid in our fractions. When 8-3-acetoxy- 45-pregnenol-17(6)-one-20 was kept in contact with alumina for several days, no starting compound could be recovered. The eluates when tested in the von Bittó-Zimmermann reaction showed the chromogenic properties of 17a-methyl-45-D-homoandrostenediol-3(3), 17a(a)-one-17 monoacetate. However, the isolation of this substance from the eluates could not be ac- complished on the small scale on which it had to be attempted. Definite evidence, therefore, is still lacking that the pregnenediolone rearranges to the D-homosteroid under conditions that prevailed during the isolation. Nevertheless such a conversion has been assumed in estimating the titer of pregnenediolone as greater than 11.7 mg. per liter of urine.
Neither 45-pregnenediol-3(6),17(6)-one-20 nor 17a-methyl-45-D-homo- androstenediol-3(3),17a(a)-one-17 has been described before as a uri- nary constituent. In view of the numerous studies which have been carried out on the ketonic fraction of normal urine, it may be justified to conclude that the concentration of pregnenediolone in normal urine is substantially lower than that encountered by us. It would appear, therefore, that 45-pregnenediol-3(6), 17(6)-one-20 originates at least in part in the adrenal
cortex. Two closely related substances, 6-17-hydroxyprogesterone (33, 31) and allopregnanediol-3(3),17(6)-one-20 (34-36), have been isolated from this source. Both may be metabolites of 45-pregnenediol-3(8), 17- (6)-one-20. Reasons which may justify this assumption have been stated in a previous communication (2). Biogenetic relationships may also exist with a pregnanediol-3(a),17-one-20 (37) and a pregnanetriol-3(a),17,20 (38, 5) which have been isolated from the urine of patients with hyper- plasia or tumors of the adrenal (5, 37, 38) or hypofunction of the testis (37).
One of the most consistent features in the steroid excretion of patients with adrenocortical tumors (5) is the very high titer of dehydroisoandroste- rone (IVb). The present case conforms to this pattern, as the yields of dehydroisoandrosterone and of its chloride (a-3-chloro-45-androstenone-17) correspond to a total of 321 mg. of dehydroisoandrosterone per liter of urine. Dehydroisoandrosterone has not been found in adrenal extracts and its precursor in this gland has remained obscure. In general it has been proposed that pregnane derivatives with oxygen substituents at C-17 and at C-20 are metabolized into 17-ketosteroids (39-41). Such compounds have been isolated both from adrenal (42) and from urinary (37, 38) ex- tracts, but as none contain a double bond in the 5-6 position the theory was not directly applicable to the formation of dehydroisoandrosterone. The isolation of 45-pregnenediol-3(6),17(8)-one-20 appears to fill this gap. It should be pointed out, however, that the metabolic conversion of 17- hydroxy-20-ketosteroids into 17-ketosteroids is not readily reconciled with the observation that adrenal hyperactivity is frequently associated with a subnormal output of urinary 17-ketosteroids (43). This apparent anomaly could be explained if it is assumed that the reaction between 17-ketosteroids and 17-hydroxy-20-ketosteroids can proceed in both directions. The reduced excretion of 17-ketosteroids during the resistance phase of the adaptation syndrome would then reflect an increased rate of utilization of 17-ketosteroids in the synthesis of the pregnane compounds of the adrenal cortex. It is conceivable that the dehydroisoandrosterone needed to initiate such a synthesis may be formed from cholesterol (44, 45) without passing through an intermediate containing 21 carbon atoms.
EXPERIMENTAL5
Fractionation of Ketones-The separation of the neutral ether-soluble ketones from 4.2 liters of urine from a boy with adrenocortical carcinoma has been described in preceding communications (1, 2). This material (6.1 gm.) was dissolved in 8 cc. of benzene. A crystalline precipitate formed on standing at room temperature. It was separated after several days and washed with a total of 10 cc. of benzene. The purification of these benzene-insoluble ketones (Ketones B, 83.6 mg.) is described below. The
“All melting points reported are corrected
supernatant and washings (Ketones A) were passed through a column (190 × 36 mm.) of alumina (prepared according to Brockmann by Merck and
| Fraction No. (A) | Eluant | Eluate | ||
|---|---|---|---|---|
| Volume | Composition* | Weight | Description or compounds isolated | |
| cc. | ng. | |||
| 1- 4 | 1150 | Petroleum ether + benzene (20%) | 9 | Oils |
| 5 | 250 | Petroleum ether + benzene (50%) | 35 | @-3-Chloroandrostenone- 17 |
| 6-10 | 1500 | Petroleum ether + benzene (50%) | 103 | «-3-Chloroandrostenone- 17 |
| 11 | 300 | Petroleum ether + benzene (50%) | 21 | a-3-Chloroandrostenone- 17 |
| 12 | 150 | Benzene | 43 | a-3-Chloroandrostenone- 17 |
| 13-19 | 1800 | . | 395 | Dehydroisoandrosterone |
| 20-25 | 1850 | “ | 436 | .. |
| 26-33 | 3600 | .. | 411) | “ |
| 34 | 250 | .4 + ether (5%) | 75 | “ |
| 35-39 | 1350 | . + " (5-20%) | 560 | « |
| 40-46 | 2200 | . + (20-50%) | 662 | “ |
| 47-48 | 650 | “ + " (50%) | 113 | Oils |
| 49-52 | 1200 | Ether | 85 | . |
| 53-57 | 1700 | .. + acetone (2-10%) | 137 | . |
| 58 | 300 | “ + “ (50%) | 134 | 17a-Methyl-45-D-homo- androstenediol-3(8), 17a- (a)-one-17 |
| 59 | 300 | . + ‹‹ (50%) | 132 | 17a-Methyl-45-D-homo- androstenediol-3(3), 17a- (a)-one-17 |
| 60 | 300 | " + .. (50%) | 78 | Oil |
| 61 | 600 | 6 + . (50%) | 104 | Partly crystalline |
| 62 | 300 | ‹‹ + ‹‹ (50%) | 126 | “ “ |
| 63 | 300 | Acetone | 141 | Oil |
| 64 | 300 | ‹‹ | 207 | Partly crystalline |
| 65-66 | 600 | .. | 146 | Oils |
| 67-68 | 550 | ‹‹ + methanol (5%) | 310 | . |
| 69-71 | 1800 | Methanol | 626 | .. |
* Whenever a change of eluant is indicated, the preceding fraction was obtained with a final 150 cc. of the new solvent mixture. This was done to correct for the holdup of the column.
Company, Rahway, New Jersey). The chromatographic separation of the adsorbate into 71 fractions required 2 weeks. A condensed summary of the results is given in Table I.
Isolation of a-3-Chloro-45-androstenone-17-The presence of this com- pound in Fractions A-5 to A-12 was indicated by a positive Beilstein test. It was isolated by recrystallization from methanol. Fraction A-5 yielded 7.9 mg. of crystals melting at 156-158°. The melting point remained un- changed after admixture with a synthetic specimen of @-3-chloro-45-an- drostenone-17 melting at 155-157° which had been prepared from dehy- droisoandrosterone by the method of Wallis and Fernholz (46). The yield from Fractions A-6 to A-11 was 17.0 mg. (m.p. 155-157º) and from Fraction A-12 3.9 mg. (m.p. 154-156°). These amounts are small com- pared with the yield of dehydroisoandrosterone which (in conjugated or free form) gives rise to the chloride during acid hydrolysis (47, 48). No attempt has been made, therefore, to isolate the additional amounts present in the mother liquors.
The first two mother liquors (74.3 mg.) of Fractions A-6 to A-10 gave a red color when tested with 90 per cent trichloroacetic acid, and showed an absorption maximum at 235 mu (E) mg/cc. = 20.1) which may indicate the presence of 22 mg. of 43,5-androstadienone-17 (49). This has not yet been verified by isolation.
Isolation of Dehydroisoandrosterone-The Dirscherl color reaction (50) indicated the presence of dehydroisoandrosterone in Fractions A-13 to A-46. Five batches of this material (Fractions A-13 to A-19, A-20 to A-25, A-26 to A-34, A-35 to A-39, and A-40 to A-46) were kept separate, ben- zoylated, and subjected to a systematic fractional crystallization from ethyl acetate. Most preparations obtained in this manner (1298 mg.) gave a clear melt between 256° and 258°, with a fusion interval of 4-8º. In addition 505 mg. were obtained which melted at or above 242-254°. These products were identified as dehydroisoandrosterone benzoate by mixed melting point determinations with a synthetic specimen and by hydrolysis to the free alcohol which showed a transition point at 139.5° and melted at 147.5-149°. The corresponding constants for a synthetic reference specimen were 141° and 150-151º, respectively. A mixture of both prep- arations melted at 149-150°.
The late eluates in particular gave evidence for the presence of other crystalline compounds. These have accumulated in the mother liquors but have not yet been investigated.
Isolation of 17a-Methyl-A5-D-homoandrostenediol-3(8),17a(a)-one-17 (IIb) -Fraction A-59 was repeatedly recrystallized from methanol and from ethanol. 13 mg. of heavy hexagonal plates were obtained which melted at 293-299° with decomposition.
Analysis-Sample dried at 110° in vacuo
C21H3203. Calculated, C 75.86, H 9.70; found, C 75.81, H 9.54
Admixture with a sample of synthetic 17a-methyl-45-D-homoandrostene- diol-3(8),17a(a)-one-17 melting at 296-302° did not depress the melting point. The mother liquors of Fractions A-59 and A-58 yielded 7.9 mg. of crystals melting at 289-295°.
17a-Methyl-45-D-homoandrostenediol-3(8),17a(a)-one-17 3-Monoacetate (IIa)-A solution of 7 mg. of the isolated D-homosteroid in 1 cc. of pyridine and 0.8 cc. of acetic anhydride was kept at room temperature for 17 hours. The excess of anhydride was hydrolyzed by the gradual addition of 0.2 cc. of water. The mixture was distributed between ether and water. The ether phase was washed repeatedly with dilute hydrochloric acid, with sodium carbonate solution, and with water and taken to dryness. The residue (8.0 mg.) upon recrystallization from ethanol yielded elongated platelets melting at 273-278° with decomposition.
Analysis-Sample dried at 80° in vacuo
C23 II3404. Calculated, C 73.76, H 9.15; found, C 73.96, H 9.46
There was no depression of the melting point on admixture with a syn- thetic specimen melting at 271-279°. The following melting points have been reported for this compound: 270-272° (14), 276-278° (30, 17), 277- 279° (17).
17a-Methyl-45-D-homoandrostenediol-3(8),17a(a)-one-17 Oxime-6.8 mg. of 17a-methyl-45-D-homoandrostenediol-3(3),17a(@)-one-17 (m.p. 295°), 40 mg. of hydroxylamine acetate, and 1.5 cc. of 90 per cent ethanol were heated under a reflux for 2 hours. The mixture was concentrated, diluted with an equal volume of water, brought to a boil, and placed in a refrigerator. The oxime was separated by centrifugation and washed with 30 per cent cold ethanol and repeatedly with water. Upon recrys- tallization from dilute ethanol rectangular plates were obtained which melted at 269-272° with decomposition.
Analysis-Sample dried at 110° in vacuo
C21H33NO3. Calculated, N 4.03; found, N 4.41
A mixture with synthetic 17a-methyl-45-D-homoandrostenediol-3(8), 17a(a)- one-17 oxime (m.p.266-274°, with decomposition) melted at 266-271º. Melt- ing points were found to be reproducible only within about ±4°. The follow- ing have been recorded in the literature: 243-244° (14), 245-246° (21), 239º (20), and 263-265° (51).
Oxidation of B-3-Acetoxy-17a-methyl-A5-D-homoandrostenol-17a(a)-one-17 (IIa); 8-3-Acetoxy-17-methyl-45-etiobilienone-17-carboxylic Acid-16 (IIIa) -2.6 mg. of bromine in 0.16 cc. of glacial acetic acid were added to a solution of 5.9 mg. of 17a-methyl-45-D-homoandrostenediol-3(8),17a(a)- one-17 monoacetate (isolated from Ketones B) in 2.8 cc. of the same solvent.
The mixture, which became almost colorless, was treated with 1.6 mg. of chromium trioxide in 0.12 cc. of 90 per cent acetic acid. The reaction was allowed to proceed for 4.5 hours at room temperature (25° ± 2º), when the excess of oxidant was reduced with 0.2 cc. of methanol. The solution was treated with 28 mg. of zinc dust and heated on a steam bath with stirring for 12 minutes. The insoluble material was removed by centrifugation and washed with ethanol (3 X 0.3 cc.) and with ether. The supernatant and washings were distributed between ether and water. The ether phase was thoroughly extracted with 6 per cent sodium carbonate solution. These extracts, which were promptly acidified with concentrated hydro- chloric acid, yielded a crystalline precipitate (2.3 mg.) which was separated and washed free of salt. The mother liquor and washings were extracted with ether. The ether phase, which was washed with water, concentrated, and freed of acetic acid in vacuo, yielded 2.8 mg. of a less pure product. Upon recrystallization from dilute methanol 2.3 mg. of colorless needles were obtained which melted at 152.5-153.5°. For analysis this product was combined with an identical preparation obtained by the oxidation of 2.4 mg. of the monoacetate (IIa) derived from Ketones A.
Analysis-Sample dried at 110° in vacuo
C23H34O5. Calculated, C 70.74, H 8.78; found, C 71.07, H 8.68
A mixture of this product with a specimen of the acetoxy keto acid (IIIa) (m.p. 152-153.5°), prepared in the same manner from synthetic starting material, melted at 152.5-153.5°.
An unsuccessful attempt was made to effect the opening of Ring D also by treating 1.1 mg. of the isolated diolone (IIb) in 1.5 cc. of methanol at room temperature with 1.5 mg. of periodic acid in 0.15 cc. of water. After 24 hours 1.1 mg. of neutral material were recovered. Its melting point after recrystallization was 284-294°.
Preparation of 17a-Methyl-A5-D-homoandrostenediol-3(8),17a(a)-one-17. Procedure 1-45-Pregnenyne-20-diol-3(6), 17(a) (V) melting at 242-245° was prepared from dehydroisoandrosterone (IVb) by the method of Stavely (21). The processing of the mother liquors was greatly facilitated by the removal of unchanged starting material with the aid of Girard’s reagent (25). The diol was converted into the diacetate (VI) according to the procedure of Ruzicka and Hofmann (25). The hydration of the triple bond according to Stavely (24) has been described by Shoppee and Prins (17). Reduction of the reaction time to 8 hours (32) was found not to be detrimental to the yield of substance VIIa (m.p. 190-193º). Hydrolysis of this substance was carried out according to Shoppee and Prins (17). The reaction product (IIb) which was recrystallized from ethanol melted at 296-302° with decomposition. The melting point was reproducible only within ±3º.
Procedure 2-9 mg. of 45-pregnenediol-3(3),17(a)-one-20 3-monoacetate (VIIb)4 and 1.5 cc. of 3 per cent methanolic potassium hydroxide solution were heated under a reflux for 15 minutes and treated with 0.4 cc. of water. The mixture was brought to a boil and placed in an ice chest. The crystal- line precipitate was collected, washed with water, and recrystallized from ethanol. This preparation melted at 292-297º.
Preparation of 3-3-Hydroxy-17-methyl-45-etiobilienone-17-carboxylic Acid- 16 (IIIb)-The preparation of this substance seemed desirable for the further characterization of the keto acid obtained in the degradation of substance IIa. 11.8 mg. of synthetic 8-3-acetoxy-17-methyl-45-etiobi- lienone-17-carboxylic acid-16 were dissolved in 2.4 cc. of 3 per cent aqueous sodium hydroxide solution and kept at room temperature for 20 hours. An ether extract of the acidified solution yielded 10.4 mg. of a crystalline residue. The acid, which crystallized in fine needles from dilute acetone, melted at 208-210°.
Analysis-Sample dried at 110° in vacuo
C21H3204. Calculated, C 72.38, H 9.26; found, C 71.92, H 9.39
5.9 mg. of the free acid were treated for 2 hours with 56 mg. of hydroxyl- amine acetate and 7 mg. of potassium acetate in 2 cc. of boiling 90 per cent ethanol. The oxime was precipitated by addition of water. It crystal- lized in needles from 95 per cent ethanol and melted at 271-275° with decomposition.
Analysis-Sample dried at 110° in vacuo
C21H33NO4. Calculated, C 69.39, H 9.15; found, C 69.38, H 9.08
Fractionation of Ketones B-The benzene-insoluble ketones were re- crystallized from ethanol. The prismatic crystals showed a constant melting point at 254-269°. At the time this was judged to be a mixture which could not be separated by this method. The crystals were combined, therefore, with their mother liquors and acetylated at room temperature with 2 cc. of acetic anhydride and 4 cc. of pyridine. The resulting acetates were dissolved in 5 cc. of benzene and passed through a column (8.4 × 0.8 cm.) of Brockmann’s aluminum oxide (preparation of Merck, Darmstadt). The adsorbed material was subjected to fractional elution as indicated in Table II.
Isolation of B-3-Acetoxy-45-pregnenol-17(6)-one-20 (Ib)-This compound was isolated from Fractions B-2 to B-10 by recrystallization from methanol, acetone, or mixtures of benzene and petroleum ether. Fractions B-3 and B-4 yielded 6.4 mg. of elongated plates melting at 219-231º and Frac- tions B-5 to B-10 furnished 6.5 mg. (m.p. 219-230°). 3.2 mg. of this substance melting at 219-227º were obtained from Fraction B-2 after
rechromatographing and recrystallization. The combined mother liquors yielded 9.1 mg. of platelets melting at 220-229º.
Analysis-Sample dried at 80° in vacuo
C23 H3404. Calculated, C 73.76, H 9.15; found, C 74.24, H 9.19 Rotation-[a]p =- 39° (c = 4 mg. per c., in dioxane)
A mixture of the isolated compound (m.p. 220-229°) with synthetic 8-3- acetoxy-45-pregnenol-17(8)-one-20 (m.p. 220-236°) melted at 220-229°.
| Fraction No. (B) | Eluant | Eluate | ||
|---|---|---|---|---|
| Volume | Composition | Weight | Description or compounds isolated | |
| cc. | mg. | |||
| 1 | 20 | Petroleum ether + benzene (50%) | 16.2 | Oil |
| 2 | 30 | Petroleum ether + benzene (50%) | 11.2 | ß-3-Acetoxy-45-pregnenol- 17(8)-one-20 |
| 3- 4 | 60 | Petroleum ether + benzene (50%) | 12.2 | 8-3-Acetoxy-45-pregnenol- 17(8)-one-20 |
| 5- 9 | 190 | Petroleum ether + benzene (50%) | 12.8 | B-3-Acetoxy-A5-pregnenol- 17(B)-one-20 |
| 10 | 50 | Petroleum ether + benzene (50%) | 0.6 | B-3-Acetoxy-45-pregnenol- 17(B)-one-20 |
| 11-12 | 70 | Benzene | 2.8 | Crystalline |
| 13 | 50 | + ether (50%) | 25.8 | B-3-Acetoxy-17a-methyl-45- D-homoandrostenol-17a(a)- one-17 |
| 14-15 | 105 | Ether | 2.5 | Amorphous (?) solids |
| 16-17 | 100 | + acetone | 3.6 | Oils |
| 18 | 60 | Acetone | 0.7 | Oil |
| 19 | 60 | Methanol | 12.4 | Solid |
Another sample of the same mixture resolidified at 233º and melted at 247º. A conversion of finely powdered specimens to higher melting prod- ucts has been described previously (29, 30). It was also noted with syn- thetic and isolated material. The following constants are recorded for the acetate: m.p. 231-233° (29), 234-235° (30); [a] = - 40.9° ± 1°, = - 41.8º ± 2.5° (in dioxane) (30).
45-Pregnenediol-3(3) ,17(B)-one-20 (Ia)-3.3 mg. of potassium carbonate in 0.12 cc. of water were added to a solution of 9.1 mg. of isolated 8-3- acetoxy-45-pregnenol-17(6)-one-20 in 4.7 cc. of methanol. The mixture was kept at room temperature for 48 hours and distributed between ether and water. The ether phase was washed with water and taken to dryness.
The residue, which was recrystallized from ethanol, yielded hexagonal plates melting at 259-272° with decomposition. The melting point could neither be raised nor sharpened by further recrystallization.
Analysis-Sample dried at 110° in vacuo
C21H32O3. Calculated, C 75.86, H 9.70; found, C 75.52, H 9.78
One sample of a mixture with synthetic 45-pregnenediol-3(3), 17(6)-one-20 (m.p. 256-272° with decomposition) melted at 259-277°; another became clear at 272°. The compound has been reported to melt at 270° or 271- 273° with frequent conversion to long needles melting at 286° or 287º (29, 30).
Oxidation of B-3-Acetoxy-45-pregnenol-17(6)-one-20 (Ib)-17.3 mg. of isolated 8-3-acetoxy-45-pregnenol-17(3)-one-20 in 2.5 cc. of glacial acetic acid were treated with 7.3 mg. of bromine in 0.3 cc. of the same sol- vent. After 5 minutes 9.3 mg. of chromium trioxide in 0.6 cc. of 90 per cent acetic acid were added. The mixture, which was kept at room temperature (30-32°) for 23 hours, was treated with 0.3 cc. of methanol and then with 50 mg. of zinc dust as described above. The solid phase was separated from the supernatant fluid, and washed with ethanol (3 X 0.5 cc.) and with ether. These solutions were combined, diluted with ether, washed with water, sodium carbonate, and water, and taken to dry- ness. The crystalline residue weighed 13.8 mg., while the carbonate ex- tracts yielded 1.7 mg. of acidic material which was not investigated further. The neutral fraction was dissolved in 5 cc. of petroleum ether and chromato- graphed on a column (4.5 × 0.6 cm.) of alumina (Merck, Darmstadt). Traces of oil were obtained by washing with petroleum ether (25 cc.). Elution with a 4:1 mixture of petroleum ether and benzene yielded 7.5 mg. of needle-shaped crystals (Fractions d to g, 65 cc.) and 1.3 mg. of an ap- parently less pure product (Fraction h, 80 cc.). The fractions obtained with a 1:1 mixture of these solvents and with benzene did not crystallize. Fractions d to g were recrystallized from methanol. The melting point (165.5-169.5°) of this product (6.3 mg.) was raised to 168-170.5° by further recrystallization. A mixture with synthetic dehydroisoandrosterone ace- tate, melting at 169-170.5°, melted at 168-170.5°.
The crystals were combined with their mother liquors and hydrolyzed with 8 mg. of sodium hydroxide in 1.2 cc. of aqueous methanol. The crude reaction product was dissolved in 0.84 cc. of warm pyridine containing 5 per cent of benzoyl chloride and was kept at room temperature overnight. The resulting benzoate was only sparingly soluble in acetone and was re- crystallized twice from this solvent. The crystals melted at 255-258º and at 257-260° in mixture with authentic dehydroisoandrosterone ben- zoate melting at 257-260°.
Analysis-C26H3203. Calculated, H 8.22; found, 7.95 (the carbon analysis was lost by accident)6
Effect of Alumina on B-3-Acetoxy-45-pregnenol-17(B)-one-20 (Ib)-2.3 mg. of the pregnenediolone acetate (Ib) were dissolved in 4 cc. of benzene which had been saturated with water. 95 mg. of alumina (Merck, Rahway) were added. The suspension, which was shaken occasionally, was kept at room temperature for 6 days and filtered through cotton. Washing with benzene (16 cc.), ether (20 cc.), and methanol (20 cc.) yielded 1.2, 0.3, and 0.7 mg. of residue, respectively. The first two fractions were colorless and produced a transitory purple color in the dinitrobenzene reaction. The absorption curves of the final pigments showed a shape very similar to that obtained with compound IIa. The two fractions were combined and recrystallized from dilute ethanol and from methanol. Lack of ma- terial prevented further recrystallization of this product, which melted at 247-256°.
Isolation of 3-3-Acetoxy-17a-methyl-A5-D-homoandrostenol-17a(a)-one-17 from Ketones B-Fraction B-13 was recrystallized repeatedly from ethanol. 5.4 mg. of platelets melting at 263-269° were obtained. The mother liquors yielded an additional crop of 1.1 mg. While the melting point could not be raised to that of the product obtained from Ketones A, it showed no depression on admixture of this preparation. Furthermore, the two products gave identical curves in the dinitrobenzene reaction and yielded the same acid on oxidation.
Dinitrobenzene Reaction-The m-dinitrobenzene was purified according to Callow et al. (13). The reaction mixtures were prepared as described by Holtorff and Koch (9). They were kept in glass-stoppered tubes at 25° ± 0.2° for 105 minutes (52) and then diluted with 4 cc. (53) of a 2:1 mixture of 95 per cent ethanol and water (54). Test and blank solution were transferred to 1 cm. Corex cells and read against each other in a grating spectrophotometer (Cenco-Sheard). The band width was 17 mu at the blue end and 12 mu through the remainder of the spectrum. The light transmitted by the blank at 520 mu amounted to 79 to 83 per cent of that passing through an equal layer of water. The acetoxy group produces no color under these conditions, as a test solution prepared from 100 y of pregnanediol diacetate gave transmission values of 100 per cent of the blank at all wave-lengths. In order to compare the behavior of 3-ketosteroids with that of 8-3-acetoxy-17a-methyl-45-D-homoandrostenol-17a(a)-one-17 (mol. wt., 374.5) (IIa), 3-ketoallocholanic acid (mol. wt., 374.5) was tested at a level of 100 y per 0.2 cc. of alcohol. The purple color which formed
6 Analysis of benzoate recovered from the mother liquors (m.p. 252-256°) gave C 79.1; H 8.4 (calculated, C 79.55).
initially started to change to brown after approximately 11 minutes. The final product gave these extinction values: 0.35 (420 mu), 0.28 (450 mu), 0.23 (470 mu), 0.21 (490 mu), 0.16 (520 mu), 0.12 (550 mu), and 0.07 (590 mu).
SUMMARY
The ketonic fraction of the urine of a boy with adrenocortical carcinoma has yielded dehydroisoandrosterone (315 mg. per liter), a-3-chloro-45- androstenone-17 (7 mg. per liter), and two substances not previously ob- tained from natural sources. These have been identified as 17a-methyl-45- D-homoandrostenediol-3(8),17a(a)-one-17 (6.4 mg. per liter) and as 45-pregnenediol-3(3), 17(3)-one-20 (5.3 mg. per liter of urine).
The D-homosteroid is regarded as an artifact formed from the preg- nenediolone in the course of its isolation. The hypothesis has been ad- vanced that 45-pregnenediol-3(3), 17(8)-one-20 and dehydroisoandroster- one are mutually interconvertible in the body.
8-3-Acetoxy-45-pregnenol-17(8)-one-20 and its 17-epimer can be dif- ferentiated by means of the dinitrobenzene color reaction.
The preparation of 8-3-hydroxy-17-methyl-45-etiobilienone-17-carboxylic acid-16 has been described.
The authors want to express their sincere thanks to Dr. F. Harvie of the Department of Pediatrics, University of Pennsylvania, who had ar- ranged for the urine collections; to Professor T. Reichstein, Basel, for samples of 45-pregnenediol-3(3), 17(6)-one-20 and its acetate; to Dr. Homer E. Stavely, The Squibb Institute for Medical Research, for a preparation of 45-pregnenediol-3(3), 17(a)-one-20 acetate; to Dr. E. Schwenk, the Schering Corporation, for supplying us with the dehydroisoan- drosterone used in the preparation of the methylhomoandrostenediolone and of chloroandrostenone; to Dr. Margaret Daus, who performed the dinitrobenzene reactions; and to Mr. W. Saschek, Department of Bio- chemistry, Columbia University, for the microanalyses reported in this paper.
BIBLIOGRAPHY
1. Hirschmann, H., J. Biol. Chem., 150, 363 (1943).
2. Hirschmann, H., and Hirschmann, F. B., J. Biol. Chem., 157, 601 (1945).
3. Ruzicka, L., Prelog, V., and Wieland, P., Helv. chim. acta, 28, 1609 (1945).
4. Marrian, G. F., and Butler, G. C., Biochem. J., 38, 322 (1944).
5. Mason, H. L., and Kepler, E. J., J. Biol. Chem., 161, 235 (1945).
6. Schiller, S., Miller, A. M., Dorfman, R. I., Sevringhaus, E. L., and McCullagh, E. P., Endocrinology, 37, 322 (1945).
7. Miller, A. M., and Dorfman, R. I., Endocrinology, 37, 217 (1945).
8. Mason, H. L., and Kepler, E. J., J. Biol. Chem., 160, 255 (1945).
9. Holtorff, A. F., and Koch, F. C., J. Biol. Chem., 135, 377 (1940).
10. Reissert, A., Ber. chem. Ges., 37, 831 (1904).
11. von Bittó, B., Ann. Chem., 269, 377 (1892).
12. Kaziro, K., and Shimada, T., Z. physiol. Chem., 249, 220 (1937).
13. Callow, N. H., Callow, R. K., and Emmens, C. W., Biochem. J., 32, 1312 (1938).
14. Ruzicka, L., and Meldahl, H. F., Helv. chim. acta, 21, 1760 (1938).
15. Ruzicka, L., Gaetzi, K., and Reichstein, T., Helv. chim. acta, 22, 626 (1939).
16. Ruzicka, L., and Meldahl, H. F., Helv. chim. acta, 23, 364 (1940).
17. Shoppee, C. W., and Prins, D. A., Helv. chim. acta, 26, 201 (1943).
18. Hirschmann, H., Federation Proc., 5, 137 (1946).
19. Chem. Abstr., 38, 372 (1944).
20. Ruzicka, L., Goldberg, M. W., and Hunziker, F., Helv. chim. acta, 22, 707 (1939).
21. Stavely, H. E., J. Am. Chem. Soc., 61, 79 (1939).
22. Stavely, H. E., J. Am. Chem. Soc., 62, 489 (1940).
23. Hardegger, E., and Scholz, C., Helv. chim. acta, 28, 1355 (1945).
24. Stavely, H. E., J. Am. Chem. Soc., 63, 3127 (1941).
25. Ruzicka, L., and Hofmann, K., Helv. chim. acta, 20, 1280 (1937).
26. Rice, F. O., Proc. Roy. Soc. London, Series A, 91, 76 (1915); J. Am. Chem. Soc., 42, 727 (1920).
27. Lowry, T. M., and Lishmund, R. E., J. Chem. Soc., 1313 (1935).
28. Goldberg, M. W., Aeschbacher, R., and Hardegger, E., Helv. chim. acta, 26, 680 (1943).
29. Fuchs, H. G., and Reichstein, T., Helv. chim. acta, 24, 804 (1941).
30. Hegner, P., and Reichstein, T., Helv. chim. acta, 24, 828 (1941).
31. von Euw, J., and Reichstein, T., Helv. chim. acta, 24, 879 (1941).
32. Shoppee, C. W., and Prins, D. A., Helv. chim. acta, 26, 185 (1943).
33. Pfiffner, J. J., and North, H. B., J. Biol. Chem., 132, 459 (1940); 139, 855 (1941).
34. Wintersteiner, O., and Pfiffner, J. J., J. Biol. Chem., 116, 291 (1936).
35. Reichstein, T., Helv. chim. acta, 19, 1107 (1936).
36. Reichstein, T., and Gaetzi, K., Helv. chim. acta, 21, 1497 (1938).
37. Lieberman, S., and Dobriner, K., J. Biol. Chem., 161, 269 (1945).
38. Butler, G. C., and Marrian, G. F., J. Biol. Chem., 119, 565 (1937); 124, 237 (1938). 39. Marrian, G. F., Bull. New York Acad. Med., 15, 27 (1939).
40. Marker, R. E., J. Am. Chem. Soc., 60, 1725 (1938).
41. Wintersteiner, O., in Glandular physiology and therapy, Chicago, 327 (1942).
42. Reichstein, T., and Shoppee, C. W., in Harris, R. S., and Thimann, K. V., Vita- mins and hormones, New York, 1, 345 (1943).
43. Selye, H., J. Clin. Endocrinol., 6, 117, 182 (1946).
44. Bloch, K., J. Biol. Chem., 157, 661 (1945).
45. Sayers, G., Sayers, M. A., White, A., and Long, C. N. H., Proc. Soc. Exp. Biol. and Med., 52, 200 (1943). Sayers, G., Sayers, M. A., Fry, E. G., White, A., Long, C. N. H., Yale J. Biol. and Med., 16, 361 (1944).
46. Wallis, E. S., and Fernholz, E., J. Am. Chem. Soc., 59, 764 (1937).
47. Butenandt, A., Dannenbaum, H., Hanisch, G., and Kudszus, H., Z. physiol. Chem., 237, 57 (1935).
48. Venning, E. H., Hoffman, M. M., and Browne, J. S. L., J. Biol. Chem., 146, 369 (1942). Munson, P. L., Gallagher, T. F., and Koch, F. C., J. Biol. Chem., 152, 67 (1944).
49. Burrows, H., Cook, J. W., Roe, E. M. F., and Warren, F. L., Biochem. J., 31, 950 (1937).
50. Dirscherl, W., and Zilliken, F., Naturwissenschaften, 31, 349 (1943); Chem. Abstr., 38, 2355 (1944).
51. Ruzicka, L., and Meldahl, H. F., Helv. chim. acta, 24, 1321 (1941).
52. Nathanson, I. T., and Wilson, H., Endocrinology, 33, 189 (1943).
53. Langstroth, G. O., and Talbot, N. B., J. Biol. Chem., 128, 759 (1939).
54. Pincus, G., and Pearlman, W. H., Endocrinology, 29, 413 (1941).