ULRICH V. SOLMSSEN
2370
All of the enzymes contained p.a.b., the greater part of it in a firmly bound form. Maximum yields were obtained by autoclaving in 6 N sodium hydroxide at 76 Ib. pressure for one hour. Minimum molecular weights calculated from the p.a.b. content again far exceed accepted values, with the exception of the yeast polypeptidase. Here the calculated value of 1.050,OOO is not unreasonably far from figure 700,000 given by Johnson.’ Yeast, however, is an excellent source of p.a.b., with samples containing up to 1307 p.a.b. per g. of dry yeast or about 250-3007 per g. of protein. There has been no increase in p.a.b. content in the purification of the peptidase over the average found in the total protein of the yeast. The question immediately arises as to whether these enzyme preparations contain foreign proteins. hTorthropl0
has pointed out that crystallization is not necessarily an index of purity. All of the preparations did contain “bound“ p.a.b. and many of them “bound” biotin in varying amounts. However, nothing is known about the molecular size of these p.a.b.- or biotin-containing impurities. Any foreign material present must be extremely high in p.a.b. if it is t o carry all the p.a.b. present. For instance, an impurity amounting t o 1% of the urease preparation would have to contain 0.21% p.a.b. Analysis of materials for their content of various biologically active substances presents itself as a sensitive index of purity. All of the preparations tested here would he regarded as containing some impurity by this criterion.
The authors are indebted to Miss Florence Fox for assistance in making the biotin determinations.
TABLE I1
Summary Six crystalline and three non-crystalline enzyme
P.A.B. CONTENT OF ENZYME
Enzyme Catalase, beef liver Concanavalin A’ Rennin Urease Yeast polypeptidase Phosphorylase Rabbit muscle extract No. 1 Rabbit muscle extract No. 2
----y
P. A. B. per g.-
H20 1 hr., 15 Ib. 4.3 0.25 1.0 1.9 6.6 2 1 7
NaOH Enzyme6 2 N HCI 1 hr., per mole 1 Er.. 15 lb. 75 Ib. P. A. B., g. 17 19 7,200,000 9.2 22 6,200,000 7 I .a 19 7,200,000 11.6 21 6,500,000 120 130 1,050,000 3.6 13 10,500,000 5 . 0 . 5 . 4 (0 N ) 23 5,500,000
1.7
6 . 0 , . 5 . 7 (6
a See footnote to Table I. figures in column 4.
A;)
5N
25
Vol. 65
6,050,000
Calculated from the p.a.b.
(10) Northrop, “Crystalline bnzymrs,” Columbia University Press, New York, N. Y.,1939.
preparations and one crystalline protein not an enzyme have been analyzed for their content of biotin and p-aminobenzoic acid (p. a. b.). The biotin ranged from 0.32 to 11.3 micrograms per gram and the p. a. b. from 13 to 130 micrograms per gram. Assuming 1 mole of biotin or p. a. b. per mole of enzyme, the minimum molecular weights are far beyond the figures assigned to these enzymes. It is concluded that the biotin and p. a. b. are contained in impurities in the crystalline proteins rather than forming an integral part of the enzyme. Determination of these vitamins in crystalline proteins may be a useful means of detecting impurities. MADISON,WISCONSIN
RECEIVED AUGUST24, 1943
- _-.- -._ _____ ._ -- [cO?QWNICATION NO.
15 FROM
THE
LABORATORIES OF HOFFMANN-LA ROCHE,INC.]
The Synthesis of Estrogenic Indene Derivatives and Remarks on the Configuration of Stilbestrol1 BY ULRICH V. SOLMSSEN The drscovery of a great number of estrogenic &done5 and 5-benzo~y-3-hydrindol.~In spite compounds not ohemically related to the natural of these negative results, this Laboratory underestrogens, has demonstrated an astonishing un- took to prepare 2-(P-hydroxyphenyl)-3-ethyl-6specificity of estrogenic activity. hydroxy-indene (V), the structure of which is Stilbestrol (I), (4,4’-dihydroxy-a,@-diethyl-closely related to that of 4,4‘-dihydroxy-a,@-distilbene) and its hydrogenation product hexestrol ethyl-stilbene (I) and 4,4’-dihydroxy-a-methyl(E),(4,4‘-dihydroxy-y,&diphenyl-hexane)first 8-ethyl-stilbene, if the structure formula for stildescribed by Dodds and ~ o - w o r k e r s ,and ~ ~ ~since ,~ bestrol and its derivatives is written according to synthesized following various routes, have ac- the British authors. quired wide interest as clinical substitutes for the natural estrogens. Among the many compounds tested by various investigators, the following three indene derivatives reported were found to be inactive: 6- and 7-hydroxy-hydrindene,6 4- and 6-hydroxy-why(1) Presented at the Meeting of the American Chemical Society, Pittsburgh, Pa., September 6-10, 1943. (2) Dodds, Goldberg, Lawson and Robinson, Nature. 141, 247 (1038). (3) Dodds, Goldberg, Lawson and Robinson, Pioc. Roy. SOC. (London), 187, 114 (1939). (4) Campbell, Dodds and Lawson, Ndurc, 14P, 1121 (1938). ( 5 ) D o d d s and Lawson, PIOC Roy Snc (London). 116,222 (1988)
__
I K = H I1 R = CHa I11 R = COCHa IV R COCH~CHZ . .
( 6 ) Miyaaaka, Pharm. SOL. Japan, SS, 407 (1939).
THESYNTHESIS OF ESTROGENIC INDENE DERIVATIVES
Dec.. 1943
2371
and Bogert'~ prepared l-methy1-2-phenyl-3-ethylindene (XVIII) from sodium ethyl- phenyl-cyanoacetate and a-phenylethyl bromide by a route different from ours. No biological activity of XVIII was reported and most likely it is inactive V VI VI1 VI11
CHs
R = H R = CHI R COCHs R COCH&!H,
"-0 (yx I
f
In view of the high estrogenic activity of 4,4'dihydroxy-y ,a-diphenyl-hexane (hexestrol) (IX), the preparation of the corresponding 2-(p-hydroxy-phenyl)-3-ethyl-6-hydroxy-indane(X) was undertaken.
H
XVIII
d XIX
due to the absence of phenolic hydroxyl groups in the molecule. These investigators proposed to HnC apply their method to prepare the corresponding methoxy- and phenolic hydroxy compounds likely to possess estrogenic activity, but so far no further results have been published. More recently SalzersJoreported the synthesis of 2-(p-hydroxyphenyl)-3-methyl-4-(or 6) -hyIX droxy-indene (XIX) from p-methoxyphenylacetone and m-methoxybenzyl chloride by a CHe----f\-OR third method different from both the one described in t h i s paper and the one devised by Plentl and Bogert.' Salzers states that it was found impossible to prepare 2-(p-hydroxyphenyl)-3ethyl-6-hydroxy-indene (V) by this method since X R = H only the methyl derivative proved capable of ring XI R CHa closure. Compound XIX was reported to be estrogeniThe synthesis of the desired indene derivatives has callv active indoses of 0.2 y g and 0.3-0.5 in rats been camed out accordina to the scheme (route of administration OHC\ /OCHI unreported). Salzer reports that an oily hydrogenation COONa product of the methyl de-C HaCd rivative XIX proved inac\COOH tive in doses as high as HnCO/ 200 y . This low activity is XI11 XIV XI1 strikindv different from the I one rebdrted below for the /OCHa similar indane derivative
+
0 - "O/ocH8 ; ;"H a
HtCO XVII VI1
7v+vI VI11 Y!
+
+
Xt-XI
HsC
XVI
The experimental work reported in this paper was begun in 1938 and completed in 1939, and publication of our results at the presenttirne is prompted Only by two publications by Other investigators on similar indene derivatives. Plentl
(XI The sodium salt of pmethoxy-phenylacetic acid (XII) was treated with m - methoxy -benzaldehyde (XIII), resulting in mmethoxy - a - (p - methoxyphenyl)-cinnamicacid, m. p. 169O," (XIV). As by-products, 3,4'-dimethoxy-stilbene, m. p. 108-109°, and m-methoxy-a- (p-methoxyphenyl) - cinnamic anhydride, m. p. 119-120°, were (7) Plentl and Bogert, TEIS
JOURNAL, 63, 989 (1941). (8) The numbering system used by these authors does not conform with the Chemical Abstracts and has therefore been changed here. (9) Solrer, U. S. Patent 2,281,956. (IO) Salzer, 2. 9hysiol. Chem., 974, 39 (1942). (11) ~ i melting i points are uncorrected.
ULRICHV. SOLMSSEN
2372
Vol. 65
isolated. Hydrogenation of (XIV) gave m-meth- rivative (VII), m. p. 118-120', and by treatment oxy-a-(p-methoxy-phenyl)-hydrocinnamic acid with n-propionic anhydride the corresponding diproproxy derivative (VIII), m. p. 88-89', were (XV), m. p. 106'. Ring closure was effected by shaking a benzene obtained, both substances being perfectly stable solution of the hydrocinnamic acid with phos- on exposure to air. phoric anhydride. Two isomeric reaction prodBy hydrogenation of the di-methoxy derivative ucts, 2-(p-methoxy-phenyl)-6-methoxy-indanone- (VI) 2-(p-methoxy-phenyl)-3-ethyl-6-methoxy-in(3) and 2-(p-methoxyphenyl)-4-methoxy-indan- dane (XI) was obtained as an oil, and after deone-@) were obtained and separated by crystalli- methylation 2-(p-hydroxy-phenyl)-3-ethyl-6-hyzation or chromatographic adsorption. The one droxy-indane (X) as a colorless crystallized subisomer has a melting point of 172', the other 96". stance, m. p. 164-165'. Both isomers were subjected to the Grignard reConfiguration of Stilbestrol.-Apart from the action with ethylmagnesium iodide. Whereas further problem of synthesizing and biologically the substance with m. p. 96' reacted readily, the testing these various indene derivatives, this proother isomer was found to be completely inert ject offered another interesting aspect. Stiltoward the Grignard reagent, even under forced bestrol may have the cis- or trans-configurationand conditions.12 Though no structural proof has Dodds and co-workers' have reported the isolation been attempted and therefore the two possible of stilbestrol, m. p. 171" (I), and of +stilbestrol, structural formulas cannot be assigned definitely m. p. 140-142'. These investigators stated that to the two isomers, it is believed that the inert the biologically less active compound probably isomer, m. p. 172O, has the methoxy group in or- has a cis-configuration and the more active stiltho-position to the keto group, preventing the bestrol a trans-configuration "being more closely Grignard reaction due to steric hindrance. There- related stereochemically to estradiol." This hyfore, it has been tentatively assumed that the iso- pothesis has been widely accepted, but in spite of mer, m. p. 96O, used for the next steps in the syn- a number of s t ~ d i e s of' this ~ ~problem ~ ~ ~ ~ ~ ~ thesis has the methoxy group in 6-position as no unequivocal proof for the trans-configuration represented by formula XVI. The crude reac- of stilbestrol has as yet been offered. This is tion product with ethylmagnesium iodide was mainly due to the fact that difficulties were entreated with sulfuric acid in order to complete the countered in preparing $-stilbestrol in a pure form. apparently spontaneous dehydration and 2-@- Consequently no direct and conclusive comparison methoxy-phenyl)-3-ethyl-6-methoxy-indene (VI) between the two isomers was possible. was obtained in peach-colored crystals, m. p. 87Only recently Walton and Brownlee20reported 88', characterized by a very strong blue fluores- a further purification of $-stilbestrol, m. p. 140cence under the ultraviolet lamp. The possi- 142O, of Dodds and co-workers,l yielding stilbility has been considered of the double bond be- bestrol m. p. 171' and apparently pure $-stilbesing formed in 1,2-position rather than 2,3-posi- trol m. p. 1.51'. The latter was converted into tion, but in analogy with all similar 2,3-substi- stilbestrol by heating with alcoholic hydrochloric tuted indene derivatives reported in the literature, acid. Whereas the structure of stilbestrol was the more likely 2,3-position of the double bond established16 by ozonization, the structure of fihas been assumed to be correct. Furthermore stilbestrol remains to be proved. there is physiological evidence in favor of the 2,3The new phenyl-indene derivatives offered anunsaturation. Linnel and Sharma13showed that other approach to this problem. Wiegand and 3,4'-dihydroxy-cu,fl-diethylstilbeneis only active Merke121showed that the absorption curve of 2in doses of 4.5 mgm. This indicates that this phenyl-indene in dioxane is very similar to that compound is 1/4g,w0as active as stilbestrol. The of trans-stilbene, whereas that of cis-stilbene is high activities reported below for compounds V entirely different. and VI1 made the 2,3-unsaturation the more The absorption spectra of stilbestrol (I) and its likely one since a lJ2-unsaturated compound di-propionate (IV) have been reported pre~J~ no reference was made to would correspond to 3,4'-dihydroxy-a,P-diethyl- v i o ~ s l y . ~ However, stilbene, rather than to stilbestrol {I). their significance regarding the configuration of The demethylation of the methoxy groups was stilbestrol. effected by treatment with hydrobromic acid and (14) Von Wessely and Kleedorfer, Nolurwissensckafkn, 17, 567 2 - ( p - hydroxy - phenyl) 3 - ethyl - 6 - hydroxy (1839). (16) Von Wessely, el ol., Monofsb., 78, 127 (1940). indene (V) was obtained as a slightly colored (18) Von Wnsely, Anpew. Chcm., 68, 197 (1940). crystalline substance, m. p. 136'. This substance (17) Von Wesaely and Welleba, Nofu~irsensckoflrn, 28, 780 proved to be unstable, being partly decomposed (1940). with discoloration after brief exposure to air. An (18) Von Wessely and Welleba, Bn., T4, 777 (1941). (19) Giacomello and Bianchi. G a s . ckim.
