STRUCTURAL AND CHEMICAL ARCHITECTURE OF HOST CELLS 131 



affect the activity of a single enzyme, as in the case of /3-galactosi(lase pro- 

 duction in E. coli (Lederberg, 1951), and it may be asked how many of these 

 are primary genes and how many fulfill the functions of "other genes" 

 described in Fig. 26. In the case of genes controlling the synthesis of a 

 compound enzyme, such as tyrosine decarboxylase, for which pyridoxal 

 phosphate is the prosthetic group, or catalase containing Fe proto- 

 porphyrin IX, auxotrophs have been obtained unable to synthesize 

 pyridoxine or hemin, respectively. In each case, the organisms nevertheless 

 synthesized the protein portion of the enzyme, demonstrable by exogenous 

 addition of the prosthetic group (Gunsalus, 1952; Beljanski and Beljanski, 

 1957). 



In the early 1940's, Beadle and Tatum and their collaborators undertook 

 to probe the problem of the mode of gene action, and elected to use the mold, 

 Neurospora, for this basic study. In this fungus the alternating sexual and 

 vegetative phases of the life cycle permitted both classic genetic analysis and 

 the study of enzyme production, and the simple nutritional requirements of 

 the wild type enabled the ready detection of auxotrophy among mutant 

 progeny. By 1950, about 1500 different auxotrophic mutants of three 

 different fungi, Neurospora, Aspergillus, and Ophiostoma, possessing a single 

 genetic change had been obtained; of these 92% were fomid to have a 

 requirement for but a single growth factor (Horowitz, 1950). By 1952, only 

 a single mutant of Neurospora (over 500 auxotrophic mutants were known) 

 had been recognized in which the genetic control of metabolic capability was 

 not centered in a nuclear Mendelian gene (Mitchell and Mitchell, 1952). A 

 few more "cytoplasmic" mutants have been recorded since then; some of 

 these in turn have been recognized to be controlled by nuclear genes (Mitchell 

 and Mitchell, 1956). 



Given this wealth of biological material, possessing lesions in pathways 

 leading to the formation of essential metaboUtes, the work was concentrated 

 for the most part during the 1940's on a delineation of many of the pathways 

 themselves, so that it became possible to establish the position of the 

 genetically controlled metabolic lesions. Many of such lesions have been 

 analyzed m detail and results covering the major pathways have been 

 summarized by Vogel and Bonner (1956), and by DeBusk (1956). Difficulties 

 in establishing the precise positions of metabohc lesions have been discussed 

 by Adelberg (1953) and Davis (1955b); these discussions record the variety of 

 biological and chemical techniques which must be employed to define exactly 

 the site of action of the enzyme affected by the gene mutation. In view of 

 these difficulties, it is not surprising that, even as in determining the primary 

 site of virus action m an intact animal, we still have not defined the nature 

 and site of most of the inborn metabohc errors of man described by Garrod 

 over 50 years ago (Wagner and Mitchell, 1955). 



