III. IU('TKKIOPHA(iE DNA AND BACTKRIAL UNA 123 



short, they were unable to demonstrate any iieterogeneity in tlie original 

 unbroken material. These results were in contrast to those obtained with 

 shear-broken molecules, which could be demonstrated to show a certain 

 variability in size. Thus, it seemed likely that whatever the molecular 

 weight of the unbroken material was, it was uniform and not as indi- 

 cated by the autoradiography up to that time. 



This situation was finally resolved by the use of autoradiography 

 and column fractionation by Rubenstein, Thomas, and Hershey (1961). 

 The general scheme of these experiments is shown in Fig. 4. 



Phage were labeled with P^-, then mixed w^ith a large excess of cold 

 carrier phage. The mixture was extracted with phenol following a pro- 

 cedure known to liberate unbroken molecules. The resulting DNA 

 preparation was loaded onto the chromatographic column and the excess 

 P^- from the growth medium was washed out with the loading solution. 

 This avoids pelleting the labeled phage which probably results in some 

 radiation damage to the DNA. The DNA is then eluted and, as seen in 

 Fig. 4, the radioactive DNA and the carrier DNA elute in an over- 

 lapping peak, which means that as far as chromatographic properties are 

 concerned, the labeled and unlabeled molecules are the same. A single 

 fraction is then selected for gentle mixing with the nuclear emulsion. 

 The resulting population of star sizes shown in Fig. 4E is the same as 

 the population of star sizes that arise from intact phage particles shown 

 in Fig. 4B. Thus, there is a single molecule of DNA which comprises all 

 of the phosphorus of the phage. 



By knowing the specific activity of the P^- in the growth medium, it 

 is possil)le to calculate that the minimum molecular weight of this mole- 

 cule is 130 X 10'\ while the phosphorus content per plaque-forming unit 

 sets an upper limit of 160 X lO*'. This conclusion is supported by the 

 results of Davison et al. (1961), who showed that drojjlets containing 

 phage particles or DNA molecules containing equal amounts of P''- had 

 an equal number of star-forming units. This molecule is not broken 

 down by heating to 70°C, proteases, or column treatment. The kinetics 

 of breakage (Burgi and Hershey, 1961 ) indicate that there are not a 

 few preferential weak points along the molecule which are more sus- 

 ceptible to shear. 



If the mixed DNA preparation is subjected to stirring before loading 

 onto the column, both the lal)eled and unlabeled molecules are eluted at 

 lower salt concentrations (Fig. 4D). When a sample of this material is 

 examined in the nuclear emulsion, the star population is seen to have 

 a lower mean value (Fig. 4F). By an examination of a number of dif- 

 ferent fractions of shear-broken preparations, it was concluded that the 

 molecule breaks successively into approximately halves and then into 



