Frontiers

Despite being far more intensively studied than any other viral group, retroviruses have lagged behind many other viruses in yielding structural information. The continuing mystery of retroviral structure reflects less a lack of will—or skill—on the part of researchers than on a quirk of nature. Mature virions are so irregular and so labile that we have been unable to apply the tools of structural analysis to good effect. Although efforts to apply standard techniques to the virion will continue to yield some information, the greatest future success in understanding the structure is likely to come from understanding its genesis. What are the protein-protein contacts in immature virions that drive assembly and how are they altered by cleavage? How does the MA domain interact with the membrane of the host cell to guide the Gag protein to the site of budding? Do Gag molecules select preexisting areas of the plasma membrane with specialized structure and lipid composition? Are portions of the MA domain or the mature MA protein inserted into the lipid bilayer or are these interactions only peripheral? How does Gag recognize the genomic RNA that is packaged and how is this interaction altered during the maturation that generates free NC? How are the characteristic shapes of retroviral cores determined? What features of Env on the cell surface are recognized by Gag that lead to its incorporation into virions? What are the structural and functional roles of proteins other than Gag, Pol, and Env in virions?

The technological advances that have allowed the molecular architecture of other viruses to be unraveled will be brought to bear on retroviral structure in the near future. Crystallographic and NMR techniques will allow the three-dimensional structure of the several individual mature Gag proteins to be solved. Comparison of these structures from different retroviruses should provide insight into the features that distinguish viruses (e.g., core shapes). These studies can probably be extended to domains of Gag that are still covalently joined, for example, MA-CA or CA-NC in HIV-1. Cryo-EM of immature cores and mature cores that can be isolated in favorable cases should lead to a picture of how protein subunits are packed in these structures. In vitro assembly systems for virus-like particles, from fragments of Gag, intact Gag, and ultimately Gag and Gag-Pol and viral RNA together no doubt will have important roles in the further unraveling of virion architecture. Molecular genetics will continue to be one of the most important tools for studying the virion. Among the most interesting mutations in the structural proteins will be those that apparently do not perturb structure or assembly, yet render the virion noninfectious. This growing collection of mutations is likely to uncover subtle and perhaps dynamic roles for the Gag proteins, for example, in viral disassembly, reverse transcription, and integration in the newly infected cell. No doubt, some aspects of virion structure that have been refractory in the past, for example, the nature of the interaction between Gag and the lipid membrane, will require that novel approaches be applied or developed.

Because of the intrinsic complexity of biology, it is a challenge in all biological investigations to distinguish idiosyncracies from generalizations. Applied to retroviruses, this means that studies on viruses important for human pathogenesis need to continue to be complemented by parallel experiments in other model systems. Thus, the workhorses of the past, the avian and murine retroviruses, are likely to continue to figure prominently in studies on virion structure. However, it would not be surprising if newly described retroviruses (e.g., those of flies or fish) or retroviruses with novel replication strategies (like the spumaviruses) will turn out to be advantageous for some studies, and thus give unexpected insight into the structure and biology of retroviruses in general.