floating colonies with a white aerial mycelium, almost forming a pellicle. A similar layer of growth covered the bottom of the flask. On the fifth clay a heavy wrinkled white mat of mycelium had formed and on the eighth day there were some submerged areas containing spores. Ten liters of culture medium, after eight days incubation, was freed of mycelium by decantation and filtered through several layers of cheese cloth. The resulting clear greenishvellow solution was divided into three portions and ex&acted thoroughly by repeatedly ihaking with benzene, u s k g a total of about 5-1. of solvent. The benzene extracts were washed with water, coinbiiieti, and coiicentrated on L: steam-bath to a volume of ahout 300 1711. The last of the benzene was distilled off under reduced pressure in an all-glass apparatus. The slightly yellow crystalline residue was triturated with two 25-mI. portions of petroleum ether to wash cut fatty material. The crystalline mixture of antibiotics, d t e r washing, weighed 1.1 g. The mixture was dissolved in 100 mi. of benzene, warmed with decolorizing carbon and filtered. The filtrate was concentrated to a volume of about 1 U ml., diluted with petroleum ether (11.p. 30-BO")until a turbidity developed, and cooled in a refrigerator. This procedure effected precipitation of most of the gliotoxin, which was separated by filtration. The crystallirie precipitate was taker] up in 40 ml. of benzene and the gliotoxin fraction reprecipitated 131- dilution with petroleum ether and cooling, as before. The crude gliotoxin after the second precipitation melted a t 180190'; after several crystallizations from ethanol it w a s 01)tained in substantially pure condition, m . p . 190-192'. The mother liquors from the isolation of gliotoxin were evaporated to dryness under reduced pressure arid the residue was recrystallized twice from ettanol. This gave 300 mg. of a pure product, m.p. 163-163 , for which analytical data are shown in Table I. Acetyl determination gave the value l l . l ~ ov ,s . 11 .7yo calculated for one acetyl group. Methoxvl analysis showed 2.0 anti l.8%, which is much lower than the value 8.3Y0 required for one methoxyl group. The crystals of gliotoxin monoacetate are quite different from those of gliotoxin1° and are readily identified under the microscope. Dr. John H. Andreen has kindly prepared for us a crystallographic description of gliotoxin monoace(10) J R Johnson, W.F. Brtice and J D. Dutcher, THIS JOIJKNAL, 68, 2006 (1913).
[CONTRIBUTION FROM T H E
tate, crystallized from benzene (Fig. 1). The crystals are orthorhonlbic and in their most common habit possess the forms: basal pinacoid, 001; macrodome, 101; and brachydome, 011. I n some crystals the basal pinacoid, 001, is very small or entirely absent. The interfacial angles 0ll:Ofl aud 101: ICli are 113 i .'1 The optic axial plaue is 100, with -/, the acute bisectrix, parallel to "c". The sign of double refraction is positive. I n sodium light the optic axial angles are 2T7 = 9 f l o , and 2E = 11.5 zk 1". For sodium light the refractive indices are: a . 1.6110: J , 1.0120, and j (calcd.), 1.80 i 0.05. Methanolysis of Gliotoxin Monoacetate .-In a small flak 500 mn. of the new antibiotic, m.p. 162-183'. vias placed with i 0 0 mg. of p-toluenesulfonic acid and 30 ml. of methanol. The mixture was heated and methanol was distilled off slowly at the rate of about 15 ml. per hour. The distillate mas collected in a receiver cooled in ice. More methanol was added from time to time to maintain a volume of 20-30 ml. of liquid in the flask. After ten hours heating, the reaction mixture was reduced to a volume of 15 ml., 50 ml. of chloroform added, and the organic solvent washed with 100 ml. of water. The aqueous layer was extracted again with 50 ml. of chloroform. The combined chloroform extracts were washed with 50 ml. of water and then evaporated to dryness. Benzene (50 ml .'I was added and evaporated to dryness to remove all of the chloroform. After two crystallizations from methanol the residue formed pale yellow crystals, m.p. 188189"; weight 150 mg. A mixture with authentic gliotoxin showed no depression of the melting point. The distillate (150 ml.) obtained during the ten hours heating with methanol was treated with 5 ml. of N sodium hydroxide solution and refluxed for two hours to saponify methyl acetate. Titration of an aliquot indicated that 76.4% of the theoretical amount of acetic acid had been formed. The neutralized solution was evaporated to dryness and the resulting sodium acetate was identified as the S-henzylthiouronium salt, according to the procedure of Shriller and Fuson." The derivative melted a t 130-132" and showed no depression when mixed with an authentic specimen of the acetate derivative.
-
(11) Shriner and Fuson, "The Systematic Identification of Organic Compounds," John Wiley and Sons, Inc., Kew York, N. Y.
ITHACA, PI;. Y .
RESEARCH LABORATORIES
OF
MERCK& CO.,
Approaches to the Total Synthesis of Adrenal Steroids. VII. Attachment of Ring D. Part A BY L. H. SARETT, LY.F.
JOHNS,
'INC.]
A New Method for the
R. E. BEYLER,R. M. LUKES,G. I. Poos AND G. E. ARTH
RECEIVED DECEMBER 22, 1952 The first phase of a new method for attaching ring D t o 1-ketopolyhydrophenanthrenesis described. A three-carbon system corresponding t o C-17, C-20 and C-21 of the pregnane skeleton is added. Thus the condensation of methallyl iodide with 2~,4b-dimethyl-7-ethylenedio~y-1,2,3,4,4aa,4b,5,6,7,8,10,1Oa~-dodecahydrophenanthrene-4~-ol-l-one ( I ) gives the 2Pmethyl-2-methallyl derivative 11. Application of the same reaction t o the Z-methy1-1,4-diketone IV gives chiefly the epimeric 2a-methyl-2-methallyl compound V. In partial ozonolyses and partial hydroxylations with osmium tetroxide the side chain double bond is affected preferentially. The structures of dilactols related to these I-keto-2-methallylcompounds are discussed.
The gain of five carbon atoms is required to threne-4fi-ol-l-one (I) with methallyl iodide in convert the tricyclic methyl hydroxyketone' I the presence of potassium t-butoxide gave a good into a pregnane derivative. Three of these, which yield of a single crystalline 2-methyl-2-methallyl represent C-17, C-20 and C-21 of the pregnane derivative I I . 2 The corresponding diketone I11 skeleton, may be conveniently introduced by a resulted from oxidation of this alcohol with the condensation with an allylic halide. I t is with this chromium trioxide-pyridine complex. When the condensation and attendant stereochemistry that methyl diketone IV was methallylated, the major the present paper is concerned. product consisted of a n iqomeric methyl methThe reaction of 2a,.lb-dimethyl-7-ethylenedioxy(2) The ceaulta of structure assignments given below are antici1,2,3,4,4ace,4b,5,6,7,8,10, lOap - dodecahydrophenan - pated here for the sake of clarity. (1) R M Lukes, G I Poos, R E B r y l r r rx' I' bhfcll, ?'&e J t r t f n m A t , $ 0 , l?O? ( l O i 7 )
Tohnq end
I
W
(3) 0. I. Poos, G. E. Artb, R 75 Rayler (1063).
f , r i + d * . i ~T , I , 112
anti I .
17 Sprptt T a r *
May 5, 195.3
TOTAL SYNTHESIS OF ADRENAL STEROIDS: ATTACHMENT OF RINGD
21 13
thesis of the missing 2-methyl-2-methallyl hydroxyketone VI was attempted. The unmethvlated hydroxyketone VIi gave CH3 the 2a-methallyl' derivaI tive VI11 in satisfactory C=CHs yield. Treatment of this compound with methyl iodide resulted in a mixture from which a few per cent. of a new methyl methallyl hydroxyketone VI could be isolated. Its oxidation afforded a compound V identical with I11 the major product of CHJ methallylation of the diI C==CHz ketone IV. This result showed conclusively that I CH? this product was the exO\A,/ pected 2-methyl-2-methallyl diketone V.8 With the establish/v/: p o ment of the C-2 position +
