G. ROBERT GREENBERG 



available, every possible source of intermediates should be 

 explored. The correlation by Krebs and Henseleit (19) of 

 citrulline from watermelon seeds with the urea cycle is a classical 

 example. Several of the purine precursors have been shown to 

 accumulate in bacterial systems under various conditions. But 

 one of the most important tools in these studies is the obtaining of 

 information on pathways and intermediates by the use of micro- 

 biological mutants. Such studies have been admirably carried 

 out with the tryptophan-synthesizing system (31) and aroma- 

 tization reactions (6). The use of antimetabolites can give 

 similar results. These tools are not for the confirmed micro- 

 biologist alone but are readily available for multi-enzyme studies. 



Whether fractionation precedes isolation of intermediate 

 compounds or vice versa, it is clear that progress in the resolution 

 of a MES is greatly dependent on the isolation and identification 

 of intermediate compounds. It seems reasonable to predict that 

 a plot of the activity in the field of glycolysis versus time would 

 show spurts of activity after an intermediate had been isolated 

 and made available. When an intermediate is available, the 

 fractionation of the ensuing reactions in a stepwise manner is 

 greatly facilitated. 



Frequently the investigator on the basis of organic-chemical 

 or biochemical precedent or by pure guess resorts to the syn- 

 thesis of possible intermediate compounds. The author has 

 adopted the principle that (7) only the most straightforward 

 syntheses should be attempted and (2) a compound should be 

 prepared and tested only when some objective data point to it. 

 In the latter category one may include the results of balance 

 studies. Such syntheses can take one far astray. For example, 

 before it was known that purines were synthesized as their ribo- 

 tide derivatives, the author spent many months synthesizing the 

 free base, 5-amino-4-imidazolecarboxamide, as a likely precur- 

 sor of hypoxanthine in pigeon liver on the basis of the findings 

 of Shive et al. (28) in bacteria. Actually this compound is not 

 synthesized, per se, although it is utilized (2). Likewise later we 

 prepared and gathered a large number of three-carbon amino- 



540 



