INACTIVATION OF VIRUSES 375 



between the different curves, which all showed a maximum near 2600 A and 

 more or less pronounced minima between 2300 and 2400 A. 



Very similar action spectra have later been established for influenza virus 

 (Hollaender and Oliphant, 1944), for the coliphages Tl and T2 (Fluke and 

 Pollard, 1949; Zelle and Hollaender, 1954), and for a megatherium phage 

 (Franklin et al, 1953). 



There are, however, exceptions to this rule. Duggar and Hollaender 

 (1934a,b) and Hollaender and Duggar (1936) showed that TMV was most 

 efficiently killed by the shortest wavelengths used (2200 A) and that the 

 efficiency decreased sharply with increasing wavelength, except for a possible 

 minor peak between 2500 and 2600 A. Rous' sarcoma virus was found later 

 to exhibit the same unusual pattern (Hollaender and Oliphant, 1944). 



With these two exceptions, the known action spectra are characterized by 

 a maximum which more or less coincides with the absorption maximum for 

 nucleic acids or nucleoproteins, at about 2600 A; a strong suggestion that, in 

 the range between 2000 and 3000 A, the principal photo-labile component of 

 a virus particle is its nucleic acid. This important point has been tested by 

 studying the quantum efficiency or quantum yield, cf), of the inactivation 

 process at different wavelengths. The quantity (f) may be defined as the 

 number of virus particles inactivated per quantum absorbed in still viable 

 particles.^ 



Oster and McLaren (1950) fomid that TMV is inactivated exponentially 

 by UV and that the quantum yield, for A = 2537 A, is about 4 X 10~^. 

 This means that, out of 23,000 absorbed quanta, only one is effective, and 

 that this one alone is responsible for inactivation. The ionic yield for simple 



^ In a primary ionization the locally released energy is of the order of 100 ev and the 

 killing efficiency of such an event is, as we have seen, high. The UV quanta responsible 

 for inactivation have energies ranging from about 3-8 to 5-5 ev. A priori, nothing can be 

 said about their efficiency. To estimate this, Zelle and Hollaender (1954) used the expres- 

 sion <j> = l/(DgAj^), where ^ j, is the absorption coefficient of the phage (in cm.- particle). 

 Dg can be determined with as much precision as needed if monochromatic UV of suffi- 

 cient intensity is available, whereas reliable estimates of ^ j, are difficult to obtain. In the 

 first place, the total number of virus particles per milliliter, not only the number of 

 viable particles, should be loiowTi; secondly, purification of the virus must be such that 

 impurities do not contribute significantly to the absorption as read in the spectrophoto- 

 meter; and, finally, correction for scattering must in some way be made. 



Luria et al. (1951) found that the number of viable, or plaque-forming T2 particles, 

 varied between 40 and 100 % of the total number, as determined by direct counts on 

 electron micrographs (using the method of Backus and Williams, 1950). On the average, 

 50-60 % of the counted particles formed placjues. In the case of TMV and the coliiDhages, 

 for which (j> has been calculated, adequate purification is possible. Correction for scatter- 

 ing is more problematic, since it rests on the validity of an extrapolation to the UV region 

 from readings made in the range between 3200 and 4000 A, in which virtually no absorp- 

 tion takes place (Luria et al., 1951; Zelle and Hollaender, 1954). 



