RNA AND CONTROL OF CELLULAR PROCESSES 103 



in the production of RNA, l)iit tlie picture is marred bv the existence 

 of "self-reproducing" RNA. The prime examples are plant and ani- 

 mal viruses, of which tobacco mosaic virus and recently poliomy- 

 elitis virus have been tlie most thoroughly studied. It was proved 

 that the virus itself carries -the necessary information for its repro- 

 duction and that RNA controls the composition of newly formed 

 virus proteins (see above). Does the RNA inside the infected cell 

 reproduce itself? If this is so, it may be that such a self-reproducing 

 RNA is a special case, different from that of the usual RNA which 

 is devoid of this abilitv. On the other hand, the infective RNA may 

 convey its information to the cellular DNA, i.e., induce in the cell 

 the production of the corresponding DNA which is used for further 

 RNA production. Such a sequence of events could be admitted if 

 something like Stent's scheme (Stent, 1958) of DNA reproduction 

 is accepted. According to this scheme, DNA first produces a com- 

 plementarv RNA which then reproduces another DNA molecule. 

 This model would allow not only the transfer of information from 

 DNA to RNA, but also in the opposite direction. 



Not enough is known of the cellular site of virus RNA reproduc- 

 tion. If the basic steps of reproduction occur inside the nucleus, as 

 has been shown for some DNA viruses ( Morgan et al., 1954; Kellen- 

 berger et al, 1959 ) , and indicated for tobacco mosaic virus (Schramm 

 and Rottger, 1959), the implication of DNA would become more 

 probable. If, on the contrary, it could be demonstrated that the in- 

 fectious virus never penetrates the nucleus, a self-reproduction of 

 RNA should be admitted. Whichever be true, I feel that the self- 

 reproducing RNA is a special case, occurring onl\' in viruses and 

 possibly in isolated instances of so-called cytoplasmic inheritance. 



At the present time, the bulk of evidence indicates that the prime 

 carrier of genetic information is DNA. If one considers the imme- 

 diate function of the DNA molecule, the possibilitv cannot yet be 

 entirely excluded that, instead of RNA, a protein is the first product 

 of gene action. The gene could serve as a model for the formation 

 of a specific protein which would act as a template for RNA. This 

 RNA could then remake the protein in the cytoplasm. Simultane- 

 ous RNA and protein synthesis need not be necessary, since RNA 

 synthesis could proceed with previously formed protein templates. 

 According to this hvpothesis, RNA formed in the absence of tem- 

 plate proteins should be biologicalh' inactive. In fact, it has been 

 observed that RNA formed in the presence of chloramphenicol has 



