278 Harold J. Morowitz 



sample size and the assay in the presence of all the inactivated material. 



It may be noted in passing that the consideration of the system in terms of 

 a Gibbsian ensemble may provide some insight into the origin of life or the 

 a priori probabihty of a biologically functional structure arising de novo. 



In considering the information aspects of ionizing radiation, we shall 

 confine ourselves to anhydrous systems and consider only the direct action 

 of radiation. We must then consider the effect of a primary ionization in 

 altering the structure of biological molecules. Present evidence indicates that 

 primary ionizations occur in a random fashion along the track of the fast 

 charged particle. However, the subsequent events are much less clear. It 

 is difficult to make quantitative statements about the probability of the energy 

 being transferred from the site of the original ionization to an energy sink 

 in the material. For purposes of developing the theory, we shall first make 

 the simplest possible assumption that the result of a primary ionization is a 

 bond break, or rearrangement of bonds at the site of the ionization. Many 

 structures are inactivated by a single ionization within the structure. If the 

 previous hypothesis applies, such structures have an information content 

 close to //niax5 sincc L must be unity if any rearrangement destroys the functional 

 integrity of the structure. It should be remembered that //max is the formal 

 upper limit if the calculation is based on atomic specification. It would be 

 possible to start from other points of view, such as monomer specification, 

 functional unit specification, or genetic specification, and arrive at different 

 values of an H function for use in subsequent analysis. 



However, there are many indications that the simple assumption made 

 above is not valid. For tobacco mosaic virus (2), the target volume is about 

 half the total volume of the particle, yet the infectious unit is presumably the 

 RNA which is only six percent of the total volume. Many enzymes show a 

 target volume equal to the physical volume of the molecule (3), yet recent 

 evidence suggests that several amino acids can be removed from the enzyme 

 without loss of activity (4). It is difficult to see why bond rearrangements in 

 these amino acids should lead to loss of function. Some enzymes show a target 

 volume larger than the physical size of the molecule. 



These factors indicate energy transfer from the site of the ionization to 

 an energy sink within the molecule. Recent studies by Gordy (5) and Setlow 

 (6) tend to suggest that sulfur-sulfur bonds are the ionization sinks in protein. 

 If we assume that this is the case and that the energy of a primary ionization 

 is transferred with a high efficiency to these bonds, then we can arrive at a 

 minimum value of the infonuation content of molecules which contain these 

 bonds and are inactivated by a single ionization. Since about one in every 

 400 of the atoms is involved in an S-S bond, we may conclude that MH/400 

 per atom is a crude estimate of the minimum information necessary to specify 

 the structure of M atoms. In this case radiation experiments would enable 

 us to set a lower limit to the information content. 



Consider next a structure which requires several ionizations to cause an 

 inactivation. If there are M atoms in the structure, each ionization may trans- 

 fornn the system from its original state to any one of the MB neighboring 

 states, where B is the average number of ways in which each atom can be bonded 

 to its neighbors. If on the average x hits are required to inactivate the structure, 



