Quantitative electron microscopy and polyacrylamide gel electrophoresis have been applied to characterize and compare the structures of three viruses of prokaryotes. The molecular weights of blue-green algal viruses LPP-1M and LPP-2 and phage T7 are (in millions) 53.4 ± 0.2, 50.9 ± 0.4, and 44.7 ± 0.2, respectively. These values were derived from the median dry masses of populations of virus particles, measured by an electron microscopic technique. For all three viruses one protein, of around 40,000 daltons, comprises over 50% of the viral protein mass. Approximately 310 copies of this molecule are used in the construction of LPP-1M, 290 copies in LPP-2 and 460 copies in T7. The blue-green algal viruses differ from the bacteriophage in utilizing a second major structural protein in building the viral shell. Most other viral structural proteins are represented by no more than about a half dozen or a dozen copies. There appear to be no proteins represented only once in any of the viruses examined.
Bibliographical noteFunding Information:
protein in LPP-1M and LPP-2 complicates This work was supported by Research Grant speculationc oncerningt he construction of GB 17514f rom the National Science Foundation. these viral heads. Two possibilities would K. W. A. is a predoctoral trainee of the National retain the principle of quasi-equivalence. Science Foundation. The skeleton of the viral shell could be We wish to thank Professor E. Zeitler for many built with one of the two major structural valuable discussions, Dr. G. F. Bahr of the Armed proteins and the secondp rotein located in Forces Institute of Pathology, Washington, DC., the interstices between capsomeresA. lter-for providing electron microscopic facilities and natively, the viral heads may be two con-Dr. II. M. Golomb and Walter Engler of the AFIP centric icosahedrals hells, each shell being debted to A. Tomic for technical help and G. for their courtesy and assistance. We are also in- built with a singlep rotein species. Grofman for assistance with the photograph. Several models for the construction of LPP-IM and LPP-2 are compatiblew ith the first possibility. These models need to ac- count for about 300 copies of the heavy structural protein and 460 copies of the light protein.A viral skeletonb uilt as a T = 4 icosahedronw ith the heavy protein would require 230 structure units (removing 5% for the tail). Four hexamerso f the light protein pluggedi nto the intersticeso n each face account for 460 polypeptides (again removing 5 %). Alternatively, the frame- work of a virus may be a T = 9 icosahedron built with 510 copies of the light protein; 340 copies of the heavy structural protein can be accountedf or by placing 6 trimers in each of the 20 faces. Although the numberso f structure units found for the viruses studied are consistent wivithth e requirementso f quasi-equivalence, the appearanceo f thesev irions in negatively stained preparations does not allow the icosahedralt riangulation number to be de- t,erminedu nequivocally. The capsid faces are one-rained, rather than coarse.T 7, for example, shows no hint of hcxamer and pentamerc lustering which, for the T = 7 icoxahedron,w ould produce a 72capsomer shell. The other qualitative result apparent in Tables 24 is the abundance of proteins present in small numbers of copies. One might guesst hat most of thesea re involved in formation of the tail and fitting the tail into the head. The prospecto f discovering t8hep recisef unctiono f each minor component appears brightest for T7, for which mutants that accumulate partial structures are known. Finally, none of the viruses examined has a protein represented only once in the virion.