159 



Terpenoids and Steroids 



experiments-'- -' -- ~-^ have shown that this "extra" carbon atom 

 is not derived from acetate, but is furnished by formate and, 

 more efficiently, by methionine. 



Progressing along the biosynthetic route from squalene to 

 ergosterol (and cholesterol), it is obvious that lanosterol must 

 lose the two methyl groups at C-4 and one at C-14. These are 

 probably removed oxidatively, and eventually some of the inter- 

 mediates may be isolated. Zymosterol has been considered as 

 an intermediate in the biosynthesis of cholesterol; but while it 

 occurs together with ergosterol in yeasts, it has been found-** that 

 squalene, but not zymosterol, is converted to ergosterol by yeast 

 homogenates. 



The biogenesis of the interesting diterpenoids gibberellic acid, 

 rosenonolactone and trichothecin has been studied. In the case 

 of gibbereUic acid-* studies with CH:^C"OOH and with C-2-la- 



CH3COOH 



beled m-evalonate gave the labeling pattern shown. A precursor 

 was inferred, and the followdng deductions made: (a) The 

 methyl carbon atom attached to ring A is derived specifically 

 from position 2 of mevalonic acid lactone, (b) The carboxyl 

 carbon atom is derived specifically from position 9 of the pre- 

 cursor, (c) The phyllocladene ring system of gibberellic acid 

 is formed by migration of C-6 to C-18. 



Rosenonolactone, rosololactone and trichothecin are all pro- 

 duced by the fungus Trichothecium roseum. 



2" George J. Alexander, Allen M. Gold and Erwin Schwenk, ibid. 79 

 2967 (1957). 



21 William G. Dauben and John H. Richards, ibid. 78 5329 (1956). 



2- William G. Dauben, Yoshio Ban and John H. Richards, ibid. 79 

 968 (1957). 



2^ William G. Dauben, Gerhard J. Fonken and George A. Boswell, 

 ibid. 79 1000 (1957). 



2* A. J. Birch, R. W. Rickards and Herchel Smith, Proc. Chem. 

 Soc, 192 (1958). 



