The General Assembly of Retroviruses

By Jason Socrates Bardi

Under the electron microscope, different retroviruses often look different because of the shapes of their capsids—protein shells that surround the retroviral RNA on the inside of the virus.

Human immunodeficiency virus (HIV) has a cone-shaped capsid, for instance, whereas the Rous sarcoma virus capsid is shaped more like a sphere. Some retroviruses have capsids that are shaped like rods. Even different HIV subtypes can have distinct appearances under the electron microscope.

But these differences among retrovirus capsids belie an underlying similarity among them.

Even though the shapes of different capsid shells vary, they may assemble using one common mechanism, reports a team of scientists led by investigator Mark Yeager at The Scripps Research Institute (TSRI) and investigator Wesley Sundquist at the University of Utah. Yeager, Sundquist, and their colleagues recently published a paper in the European Molecular Biology Organization Journal that proposes a general model for retroviral capsid assembly.

This research is significant because understanding how the virus matures may be important for finding targets for intervening in this process. Stopping the maturation of HIV has already been proven as a therapeutic strategy through protease inhibitors, which process the Gag precursor proteins. Inhibitors that block the formation of the hexameric capsid lattices might prove an effective complement to existing antiretrovirals.

Capsid Catch Can

Retroviruses contain a dozen or so genes, and these enable the various stages of the virus lifecycle—from the initial entry into a new cell to the replication and formation of new virus particles late in the lifecycle.

When retroviruses form infectious "virions," they do so by first expressing their RNA and protein components and then assembling these molecular components on the inside of an infected cell. These components will bind to the cell membrane and assemble there, leading to the budding off of an immature virion. Inside the immature virion are long Gag polyproteins, which are precursors of the structural proteins nucleocapsid, capsid, and matrix.

After budding, the virion must still "mature" into an active, infectious particle by using a protease to chop the precursor Gag proteins into their component pieces. Once these structural proteins are free, they can self-assemble into the structures that give a retrovirus its classic shape and appearance. Capsid proteins, for instance, assemble into the protein capsid shell that surrounds the RNA of the retrovirus and appears as a cone in HIV and as a sphere in Rous sarcoma virus.

The capsid proteins from several different retroviruses have now been solved, and their 3-D structures are virtually identical. However despite these similarities, different retroviruses appear quite different under the electron microscope. Somehow the retroviruses achieve completely different forms from the same ingredients—like two different chefs who both cook with the same rice, fish, and vegetables but make, respectively, a dish of fried rice and a dish of sushi.

According to the new general model, capsid proteins achieve their final form by assembling first into honeycomb-like hexameric lattices. The Sundquist lab had previously suggested that all retroviruses use this same lattice to produce their capsids. The characteristic rod, cone, and spherical shapes are generated by the insertion of pentamers of molecules within the hexameric lattice.

When you put a pentamer into a hexameric lattice, you cause that lattice to bend. Enough of these pentamers will close the lattice. This situation is analogous to a soccer ball, which relies on pentagons of leather sewn onto the sides of hexagons to achieve a shape approximating a sphere.

In a retrovirus, a relatively even distibution of pentamers in the hexameric lattice will generate a spherical shape. Similarly, a rod or cone-shaped hexameric lattice can be closed by inserting pentamers at the ends of the lattice in a defined pattern.

Barbie Ganser, a graduate student with Wes Sundquist, visited the Yeager lab and grew 2D crystals of capsid molecules from the Moloney murine leukemia virus (M-MuLV), a virus that forms a spherical capsid. They found that the packing of capsid domains in M-MuLV conforms to a general model that had originally been developed as a way of describing HIV virus assembly. They propose a general model for the assembly of retrovirus capsid molecules with the mature virion.

According to the model, the structures of the capsid protein domains are more or less the same in the various retroviruses. The packing of these domains into hexameric lattices is the same as well, but flexible linker motifs between domains of individual capsid proteins must somehow produce the different shapes and sizes of the final products by controlling the location of pentamers in the hexameric lattice.

To read the article, "Three-dimensional structure of the M-MuLV CA protein on a lipid monolayer: a general model for retroviral capsid assembly" by Barbie K. Ganser, Anchi Cheng, Wesley I. Sundquist, and Mark Yeager, please see The EMBO Journal Vol. 22, pp. 2886–2892, 2003.




A three-dimensional reconstruction of the murine leukemia virus caspid protein was derived by electron microscopy and image analysis of two-dimensional crystals.