Viruses have long been known as sources of human disease, but have now also become important tools of biological research used in DNA cloning, targeted gene therapy, and phage display of novel proteins. Although viruses are extremely diverse in their life cycles and infectious mechanisms, most share a common structure, consisting of an inner core of condensed nucleic acid enclosed within a spherical protein capsid. The structures of these capsids are based upon icosahedral symmetry, as illustrated in these images of viruses.
In the case of herpesviruses and adenoviruses, as well as the double-stranded DNA bacteriophages such as P22, the initial product of the viral assembly pathway is not an infectious virion but a closed shell that does not contain DNA. These precursor shells, or procapsids, include proteins not found in the mature virion, but essential for their production. These proteins are termed "scaffolding proteins".
During the assembly pathway of bacteriophage P22, 300 molecules of a 33 kD scaffolding protein coassemble with the 420 coat protein subunits to form a double-shelled structure with the scaffolding inside. The procapsid also includes a dodecameric portal complex at one vertex, which serves as the channel through which the DNA enters. In addition, there are from 10-20 molecules each of three pilot proteins, needed for injection of DNA into the host cell.
Upon the commencement of DNA packaging, all 300 scaffolding molecules are released intact from within the shell, presumably through the channels present at the fivefold and sixfold centers. The DNA is pumped into the capsid in a process that involves two virally-encoded proteins and ATP. During these events the capsid undergoes conformational changes that result in expansion, angularization, and closure of the channels. See the P22 procapsid structure at 19 ™ resolution .
The arrangement of the scaffolding subunits within the procapsid is not known; it may not match the icosahedral symmetry of the coat shell. No structures of scaffolding proteins from any virus have been solved.

The Role of Scaffolding Protein in Phage P22 Capsid Assembly and Maturation
My work focuses on both understanding the roles played by the scaffolding subunits in the complex processes of virus assembly and maturation, and identifying regions of the molecule required for specific functions.
Scaffolding release can be reproduced in vitro in the absence of DNA by low concentrations of guanidine hydrochloride. This release from within the capsid was reversible, permitting study of scaffolding interactions with the assembled coat lattice. The rapidity of scaffolding reentry suggested that the subunits could rebind to specific sites within the capsid. Procapsids contained two classes of scaffolding subunits, which may represent binding to different sites within the lattice; thus, all scaffolding proteins do not make equivalent interactions with the coat protein. These sites became lost or inaccessible upon capsid expansion and maturation.
Image reconstruction from cryo-electron micrographs of procapsids containing scaffolding suggest that although the scaffolding subunits interact with the capsid lattice, the scaffolding core is not arranged with the same icosahedral symmetry as the outer shell. This work was done in collaboration with Drs. Chiu and Prasad and their colleagues at Baylor College of Medicine. Pam Thuman-Commike , a graduate student at Baylor, has a page which discusses image reconstructions of phage P22 structures in more detail.

A set of new missense mutants at four sites in the scaffolding protein were isolated and characterized. Their locations are shown on a map of the bacteriophage P22 scaffolding sequence . These mutants did not prevent capsid assembly under restrictive conditions. Two mutants were defective in incorporation of the portal complex which serves as the channel through which DNA is packaged. These mutations may identify a region of the protein required for interaction with the portal. Two mutants in a different region of the sequence were impaired in scaffolding release both in vivo and in vitro. These mutations may identify a new domain required for scaffolding release. Both these mutations resulted in severe destabilization of part of the protein to thermal denaturation. Scaffolding release appeared to be required for capsid expansion; in turn, scaffolding release seemed to depend upon the presence of a portal. This may help to order the pathway of events in phage maturation.
A scaffolding region required for binding to coat protein was identified by functional assays of proteolytic fragments, and found to be the C-terminal 20-30 residues. The structural organization of the purified wild-type and mutant proteins were also probed by protein denaturation techniques. The unfolding of scaffolding protein was not a simple two-state mechanism as for a typical globular protein, but a complex process invoving the sequential denaturation of multiple domains. The mutant amino acid substitutions selectively destabilized particular domains, allowing them to be assigned to known functions. The scaffolding protein was notably unstable, to the extent that some domains are probably largely unstructured at physiological temperatures. The less stable domains include the regions involved in coat binding, portal insertion and scaffolding release, suggesting that many critical scaffolding functions may require a high degree of conformational flexibility.

Based on these results I propose a model for assembly in which a terminal region of the scaffolding protein induces conformational changes in the coat protein leading to efficient polymerization. The release of scaffolding may not be a passive result of changes within the coat lattice but involve active conformational changes within the scaffolding subunits in response to signals associated with DNA packaging. Docking of the DNA packaging complex at the portal could cause a signal to be propagated throughout the scaffolding core resulting in scaffolding release, after which the coat lattice is freed to expand to its mature conformation.

Greene, B., and King, J. (1994). Binding of scaffolding subunits within the P22 procapsid lattice. Virology 205, 188-197.

Greene, B. and King, J. (1996). Scaffolding mutants identifying domains required for P22 procapsid assembly and maturation. Submitted to Virology.

Thuman-Commike, P. A, Greene, B., Jokana, J., Prasad, B. V. V., King, J., Prevelige, P. E., and Chiu, W. (1996). Three-dimensional structure of scaffolding-containing phage P22 procapsids by electron cryo-microscopy. Submitted to J. Mol. Biol.