STRUCTURAL AND CHEMICAL ARCHITECTURE OF HOST CELLS 125 



is also metabolically stable. The stability of DNA and RNA for actively 

 growing and multiplying bacterial cultures had also been established earlier 

 by a number of laboratories (Fujisawa and Sibatani, 1954; Manson, 1953; 

 Hershey, 1954). 



Of the results presented in Fig. 24 on mammahan cells, surely that con- 

 cerning the stability of RNA is the most surprising. However, we must note 

 that these studies apply to cultures in an exponential phase of growth and 

 that cells in which mitosis has substantially stopped have not yet been 

 subjected to such an analysis. Siminovitch and Graham (1956) have pointed 

 out that even in exponentially growing cultures the initial rates of incorpora- 

 tion of P^^ into RNA and DNA may differ considerably, but the observed 

 stabilities of the nucleic acid suggests the need for explanations other than 

 that of turnover, explanations such as differences in precursor pool size and 

 mechanism of incorporation in each instance. Thus, since most of the so-called 

 turnover data concerning RNA in the tissues of intact animals do not take 

 these possibilities, as well as the turnover of cell population, into account, 

 Siminovitch and Graham (1956) conclude there are no satisfactory data on 

 the in vivo turnover of the nucleic acids. 



As in the case of the proteins, it seems likely that this judgment, while 

 rigorous at the time made, may possibly be too extreme and fails to take into 

 account a good deal of data concerning RNA metabolism under conditions 

 of metabolic stress. Two days of starvation will produce a sharp decline in 

 the RNA content of mammalian liver; this increases again after feeding, 

 although the DNA content remains unchanged (Davidson, 1947). As noted 

 above, starvation provokes the degradation of an RNA particle in E. coli 

 (Dagley and Sykes, 1957) and, indeed, of RNA structures in a variety of 

 microorganisms, followed by resynthesis on feeding. Also, the phenomenon 

 of chromatolysis and regeneration in nervous tissues is correlated with the 

 degradation and resynthesis of ultraviolet-absorbing basophilic structures 

 (Caspersson, 1950). The turnover of RNA in phage-infected bacteria and in a 

 uracil-requiring strain of E. coli strain 15^" has also been mentioned earher. 

 Most recently, it has been shown that the RNA formed by E. coli in the 

 presence of chloramphenicol is rapidly degraded in vivo when the antibiotic 

 is removed from the medium (Neidhardt and Gros, 1957). 



The existence of this data suggests that under conditions of metabolic 

 stress the system of Siminovitch and Graham may produce another type of 

 result. One would wonder if virus infection may constitute such a stress, as 

 indeed it does in the case of DNA turnover in phage-infected bacteria, 

 wherein, as summarized by Kozloff (1953), host DNA is degraded to deoxy- 

 ribotides, which are then resynthesized to form viral DNA. 



However, in apparent contradiction of the possibility that RNA turnover 

 occurs primarily in cells which have stopped growth and multiplication is the 



