260 H. K. SCHACHMAN AND R. C. WILLIAMS 



formed a stoichiometric complex with, water which could be isolated and 

 weighed as such, then the ultracentrifuge would reveal the molecular weight 

 of that material. Seldom, if ever, is this situation realized with macromolecules. 

 Frequently the amount of virus is so small as to preclude measurements 

 of concentration and density; in this case all versions of the classic density 

 method are of little value. Moreover, the virus preparation may be of such 

 doubtful purity that density measurements on the solutions are not meaning- 

 ful. In such circumstances an ultracentrifugal method is the one of choice. 

 Examination of Equation 22 reveals that the sedimentation coefficient can 



become zero only when (1 — Vp) = 0. Accordingly, the sedimentation rate 

 is measured in a series of solutions of increasing density and the resulting 

 data plotted in a manner to permit extrapolation to the value of p corres- 

 ponding to zero sedimentation rate. This value for the density of solution is 



equal to 1/F. Since the equation is restricted to two-component systems, it is 

 not vahd to employ any material such as sucrose to increase the density of the 

 solution (Schachman and Lauffer, 1950). It appears, however, that mixtures 

 of DgO and HgO act as a one-component solvent and therefore they can 

 be used for these experiments. Unfortunately, DgO solutions are not suffi- 

 ciently dense to reduce the sedimentation rate of viruses to zero, and a long 

 extrapolation is necessary. With accurate data, however, this can be per- 

 formed in a satisfactory manner and the method has, therefore, much to 

 offer. It should be noted that there will be exchange of soms of the hydrogens 

 on the virus with deuterium from the heavy water and the measured value 

 corresponds to a deuterated macromolecule. The latter has a partial specific 

 volume about 1.5 % less than that of the virus in ordinary water. In this 

 respect the method will doubtless be improved by the use of HgO^^ since 

 there is much less exchange of oxygen than hydrogen. This ultracentrifugal 

 method has already found wide application with viruses, including swine 

 influenza virus (Sharp et al., 1950), the virus of avian erytliromyeloblastic 

 leukosis (Sharp and Beard, 1954), bushy stunt virus (Cheng and Schachman, 

 unpublished), and poliomyelitis virus (Schwerdt and Schaffer, 1955). In the 

 latter instance only microgram amoimts of virus were available for the whole 

 study. 



Often the partial specific volume of viruses and protein has been assumed 

 because of the absence of relevant experimental data. Although this has been 

 done frequently with no apparent rationale, a real basis does exist for such 

 guesses. It has been fomid for many proteins that the volumes are closely 

 equal to the sum of those of the mdividual amino acid residues (McMeekin 

 and Marshall, 1952). Thus composition data in terms of the amino acids 

 permits the calculation of the specific volume. Similar computations work 

 satisfactorilv for viruses when allowances are made for the nucleic acid. 



