PROGRESS IN THE TOTAL SYNTHESIS OF STEROIDS I. V. ToRGov Institute of the Chemistry of Natural Products, Academy ofSciences, Moscow, U.S.S.R. The total synthesis of steroids is undoubtedly one of the most important achievements in organic chemistry. It is not so long since 1939 when Bach­ mann, Cole and Wilds1 carried out the first total synthesis of the simplest of the sex hormones, equilenin, but in this short period we have witnessed a tremendous development in the chemistry of steroids in general and of their total synthesis, in particular. Allow me to call to mind the principalland­ marks: in 1948-the first synthesis of oestrone by Miescher2 ; 1950-the second, by Johnson3; 1952-synthesis of cholesterol, and of the sex and adrenocortical hormones by Woodward4 and by Robinson5 ; 1952-1953- preparation of the cortical hormones by Sarett6; 1953-1954-modification of Woodward's synthesis by the Monsanto group 7; production of keto­ etianic acid by Wilds8 and oflanosterol by Woodward and Barton9; 1955- 56-preparation of epiandrosterone, testosterone and some 11-oxidized steroids by Johnson10, the first synthesis of aldosterone by Wettstein11 ; 1957-58-syntheses of oestrone by Johnson and Christiansen12 and by Johnson and Walker13 ; three syntheses of aldosterone (by Reichstein14, by the Dutch group15 and by Johnson16), synthesis ofvitamin D by lnhoffen17 ; 1959-60-second modification ofthe synthesis ofaldosterone by Wettstein18. four more routes to oestrone (Hughes and Smith19, Torgov and Anan­ chenko20, Banerjee21 and Velluz22), synthesis of testosterone by Johnson23 and of androsterone by Velluz22 and finally of the sapogenins (and steroid alkaloids) by Sondheimer24-a veritable torreut of syntheses! Finally, in 1961-1962 we have the production of pregnane derivatives by Nagata25, preparation of conessine by Barton26 and by Johnson27 and the first synthesis of a cardiac aglycone by Sondheimer28, We thus see that both the pace and scale of work are ever increasing and chemists have now arrived at the threshold of industrial synthetic methods of steroid production. This, despite the fact that such a multi-stage, stereo­ specific synthesis has many pitfalls in its path and the failure of any stage may Iead to collapse of the whole projected route. To what are these gigantic strides due? First of all they result from the general advance in organic chemistry. The number of specifically directed reactions is constantly increasing, and most of them are capable of being carried out at low temperatures and in the absence of "drastic" reagents. It is difficult to imagine the synthesis of even the simplest of steroid com­ pounds without the Arndt-Eistert reaction, without selective hydrogenation techniques, or without stereospecific reduction. Thus, in the Johnson­ Christiansen synthesis12 (Figure 1) hydrogenation under various conditions 525 I. V. TORGOV solved the problern of ring B formation and of trans-fusion of the C and D rings. Reduction by metal hydrides [LiAlH4, NaBH4, AlH(O-t-C4H 9)s] is now being used so widely that at present there is hardly any synthesis in which it does not play an important part. C~300CH3 CHL --- .C ,-COOCH3 COOCH3 Several Reformatsk~ OOCH 3 steps ~ : : reaction H ~ COOH ~ 1 CH2COOCH3 OH 2 3 CH 30 CH 30 ~ CHCOOCH3 H2. Pd/C HCIO,..AcOH; 76'/. 0 CH3 0 (1)NaBH4 ~' (2) Py·HCI ;/' I ' ' HO ~ Figure I. Johnson-Chr1stiansen synthesis of oestrone Reduction of styrenoid compounds by alkali metals in liquid ammonia has been a decisive factor, determining the success of the numerous syntheses by Johnson10, Nagata's syntheses and those of Velluz22• Oppenauer oxidation and chromic anhydride oxidation in pyridine, extensively used in the synthesis of cortisone and corticosterone by Sarett6, in the syntheses of aldosterone11, and of the sapogenins, have contributed as much to these successes as has periodic acid oxidation utilized by Wood­ ward4, Sarett6 and Nagata. Stereospecific hydroxylation with the aid of osmium tetroxide, or of io­ dine in the presence of silver acetate, was also of great assistance in the same 526 PROGRESS IN THE TOTAL SYNTHESIS OF STEROIDS syntheses. Special mention must be made of Fried and Sabo's method, 9 11 allowing the 11-oxygen function to be incorporated in ß < > compounds, a 7 22 technique made use ofby the Monsanto group (Figure 2), and by Velluz • (Figure 3). I J50CH,I ' CH3;::/' I • PhNHCH3 0 0 (1) H2,Pd;lOO% 96% (2) HCOOC2H5,CtisONa 86"fo 0 Woodward's CHOH ketone (2) KOH;-23% HO~·CH3 0 NBS,~,H+ (1) Cr03,Py; 99•.4 84% , , j< (2lZn,AcOH;54% Br 1 0 CH3 CHO . Piperidin~~ AcOH ~ 65%C2steps) : : Cortisone Figure 2. Synthesis of cortisone by the Monsanto group Incorporation of the vinyl group by the Norman reaction was felicitously employed in the syntheses of oestrone by J ohnson, Rohins and Walker13 and by Torgov and Ananchenko20 (Figure 4). An interesting route to highly unsaturated steroid systemswas elaborated by Zavialov16a (Figure 5). The Wittig reaction though as yet not used very often, was employed three 17 times by Inhoffen in his synthesis of vitamin D 3 (Figure 6). An unexpectedly important r6le has been played by certain " nameless" reactions. Thus, the very simple method of protecting the carbonyl group by means of ethylene glycol greatly facilitated, andin some cases determined the success of, the synthesis of ß 4-3-ketosteroids and ofsapogenins. The same holds true for the tosylation reaction. As is weil known, the tosyloxy-group, 527 I. V. TORGOV very similar to halogen, can take part in a variety of exchange reactions, a property which has aided in the solution of many complicated synthetic problems as, for example, the formation of ring D in the synthesis of cortico­ sterone and of cortisone by Sarett's method (Figure 7), injohnson's synthesis of aldosterone16 (Figure 8), and in Barton's synthesis of conessine (Figur .. 9). Johnson's OtrCH,a;~,h OVCH, OH method 1- reduction;75'/, CHjO ~ 0 CH3 OBz (CH20Hh.H• 0 OH HOY'HOH --:-:-H-=-o,-,-:-Hr+- ~ 2 (1) Oxidation (2) Reduction Figure 3. Velluz's synthesis of cortisone One can expect that in the very near futureextensive use will be made of radical reactions, hydroboration, etc. Thus, the organic ehernist has at his disposal an ever-increasing variety of specific and stereospecific reactions which serves as a mighty tool in the materialization of his ideas, as evidenced by the syntheses mentioned above. In the total synthesis of steroids chemists have had to cope with certain specific problems dealing with formation of the steroid skeleton and the incorporation of functional groups. The introduction ofthe angular methyl group was solved by Robinson and Johnson and is, in general, reduced to a Michael reaction and alkylation of a 528 PROGRESS IN THE TOTAL SYNTHESIS OF STEROIDS ketone. Recently Stork has proposed, and verified on bicyclic systems, a novel method for incorporating the 18-CH3 group, based on a Favorsky rearrangemen t of chloroketones (Figure 10). The chloroketones were pre­ pared by condensing the ketones with chloronitriles. ~0 0~ ~ CH 0 3 ~ Johnson-Walker Sl!veral Stl!pS '?'I HO ~ Figure 4. Syntheses of oestrone by Johnson, Rohins and Walker, and by Torgov and Ananchenko c"""""'' /"-.. t ot:O 0JoXJ 24'/.(2 s.teps) CH~ oN,oH- 0 0 0 a•/.(2 steps) Figure 5. Zavialov synthesis Trans-fusion of rings C and D was solved in two major ways, namely, catalytic hydrogenation of ß 14<15>-compounds (Miescher, Johnson and Christiansen, Hughes and Smith, Torgov and Ananchenko) or isomeriza­ tion of n-homosteroid ketones of the type of cis-decalone to ketones of the type of trans-decalone (Woodward, Johnson) cJ. Figure 11. 529 I. V. TORGOV CH~H3 CHl H 8 17 o-C"H9 CH~PPh3 (1)H2,Pt 58% --- AcO AcO w-0 CH3 C H 8 17 CH2:CHCH = PPh3 (1) 03; ~ 100'1. 40'/. (2) LiAlH4 : BO'I. (3) MnC z; 34'{, I CH r2 hv -7'1. HCH 3 CH3 o~ 6CH 2 OH HO/ F~gure 6. Inhoffen's synthesis of vitamin D 3 22% (3 steps) Figure 7. Part of Sarett's cor tico~tterone synthesis 530 PROGRESS IN THE TOTAL SYNTHESIS OF STEROIDS _-COCH 3 43"/.(3 steps) Figure 8. Part of J ohnson's aldosterone synthesis HO H N ~ ~ HCOOH (2)liAIH4 N 2 (CH~N Figure 9. Barton's synthesis of conessine CH3CHC!CN ""'- t-BuOK;~ Q COOC H 2 5 Figure 10. Stork's method for introducing angular methyl group 531 I. V. TORGOV Trans-fusion of rings B and C was accomplished by hydrogenation of ß. 9 <11>-compounds, by use of the Fried-Sabo method (Woodward, Velluz) or by means of Birch reduction (Johnson and Walker synthesis). For the ß. 8 <9>-compounds as yet only the Birch method gives good results (syntheses by Johnson, by Hughes and Smith, by Torgov and Ananchenko); but this too is not completely stereospecific and yields some cis- tagether with the trans-compound (Nagata). Finally, in Sarett's synthesis isomerization of the cis-decalone to the trans-decalone system took place under conditions of Oppenauer oxidation6 (Figure 12). As for the incorporation of functional groups, with very few exceptions, this is at present readily accomplished. Introduction of the oxygen function at C-18 ceased tobe a difficult prob­ lern after the syntheses of aldosteronein 1955-58 [by Wettstein, by Reich­ stein, by J ohnson, by the Organon group as weil as the brilliant work by Barton and by the Swiss and French chemists (Eger-Vellus) who used radical reactions for this purpose30] ( Figure 13). Another difficult problem, the incorporation of an oxygen function at C-19, is nearing its practical solution. Thus, Barton, applying his method of nitrite photolysis to 6-hydroxy-steroids synthesized 19-carbonyl derivatives31 (Figure 14) .