Structural transitions in the scaffolding and coat proteins of P22 virus during assembly and disassembly.

An in vitro system for investigating the assembly of the Salmonella phage P22 has been exploited to elucidate the structural basis of recognition between scaffolding protein (gp8) and coat protein (gp5) subunits of the viral procapsid. Raman spectroscopy and circular dichroism have been employed to examine structural thermostabilities of both gp8 and gp5 in native procapsids, and to characterize structural changes accompanying scaffolding exit, procapsid expansion, and shell disassembly. It is found that the secondary structure of the isolated gp8 subunit is rich in alpha-helix (approximately 40%), is highly thermolabile, and is characterized by noncooperative unfolding (Tm approximately 49 degrees C). Conversely, the procapsid-bound gp8 subunit exhibits stabilization of its alpha-helical secondary structure, characterized by cooperative unfolding. Because cooperative unfolding of gp8 coincides with exit from the procapsid, the present results suggest that unfolding and release are coupled processes. Structural differences between procapsid-free and procapsid-bound gp8 subunits are also apparent in Raman markers which monitor environments of tyrosine and tryptophan side chains. Temperature-resolved Raman spectroscopy of the empty procapsid shell reveals three distinct structural transitions for the gp5 subunits. The first, which occurs between 50 and 65 degrees C, is attributed to shell expansion and results in an increase in beta-strand secondary structure. The two higher temperature transitions, occurring within intervals of 70-80 and 80-95 degrees C, respectively, are attributed to partial unfolding of the shell subunit and subsequent shell disassembly. The same gp5 structure transitions are detected for procapsids which contain scaffolding protein. On the basis of the observed thermodynamic coupling between gp8 unfolding and its release from the procapsid, we propose a model for P22 procapsid assembly. Implications of the model for in vivo assembly of dsDNA viruses are discussed.