 |
Mechanism of Retrovirus Infection
Summary: James Cunningham and his colleagues have identified proteins on the surface membrane of cells that are required for attachment and infection of retroviruses. They are studying how retroviruses use these proteins to gain entry to the host cell.
Retroviruses are ~100-nM particles that contain an RNA genome that
encodes for the proteins that form the particle and direct the
replication and transfer of the RNA to new cells. These viruses are
released from infected cells by budding from the plasma membrane.
During budding, the core of the nascent virion containing the viral RNA
is covered by a portion of the host cell membrane enriched in
virus-encoded "envelope" proteins. After release, retroviruses are
competent to fuse with the membrane of uninfected cells, thereby
delivering the virion core into the cell cytoplasm. Virus fusion and
infection are the result of a conformational change in the envelope
protein that is triggered by binding to host proteins, termed
receptors, expressed on the target cell membrane. Therefore, receptor
expression is a major determinant of retroviral tropism and
pathogenesis. Previously, my colleagues and I identified the receptors
for several members of a large family of oncogenic retroviruses, the
mammalian leukemia viruses (MLVs). Our focus is on elucidating the
molecular mechanism by which these proteins trigger MLV infection.
Role of Receptor Binding in Infection
Retroviral envelope proteins are trimers of heterodimers composed of a
surface subunit (SU) that binds to the receptor and a transmembrane
subunit (TM) that mediates fusion of the virus to the host cell
membrane. Initial hypotheses for the mechanism of cell entry by
retroviruses supposed, by analogy to the hemagglutinin envelope protein
of influenza virus, that the SU sits as a clamp on the TM, repressing
its membrane fusion activity. In this scheme, receptor binding would
dissociate the three SU subunits, thereby releasing TM from a
kinetically trapped, metastable conformation. Recent studies from our
lab, however, require a modification of this view. We showed that the
amino-terminal half of MLV SU forms a discrete receptor-binding domain
(RBD) that binds to the receptor with 1:1 stoichiometry and is
connected by a short, proline-rich, "hinge" to a conserved domain
formed by the remainder of SU. Contrary to expectation, addition of
soluble, monomeric RBD to the culture medium restored infection by
mutant retroviruses in which membrane fusion is uncoupled from receptor
binding. This observation excludes a simple model in which receptor
binding activates the TM fusion mechanism by disrupting the quaternary
structure of SU.
How receptor binding to the virus envelope triggers fusion is not
yet understood. To address this issue, we recently obtained an atomic
resolution structure of RBD from feline leukemia virus (with Deborah
Fass, Weizmann Institute) for comparison to the previously obtained
structure of RBD from Friend MLV, which binds to a distinct receptor.
Informed by this comparison, we identified a pocket on one face of each
RBD that mediates receptor binding and a conserved histidine residue on
the opposite face that is critical for the postbinding activation step.
Our experiments strongly suggest that receptor binding induces a
conformational change in RBD that triggers fusion through a direct
interaction with the conserved portion of SU. Currently, our efforts
are directed at determining the structure of RBD bound to receptor. In
addition, we are testing the hypothesis that receptor-bound RBD
activates an intrinsic thiol reductase activity of the viral envelope
that is critical for the fusion mechanism.
Cooperative Retrovirus Infection Underlies Pathogenesis
In natural infections in animal hosts, genetic variation caused by
mutation and recombination creates a population of related
retroviruses. Over time, the composition of the virus population
changes in response to selective pressures created during replication
in different tissues and by the immune response to infection. Previous
studies indicate that the appearance of specific viruses in the
population that can repeatedly infect the same cell is a key step in
the pathogenesis of virus-induced leukemia and with immunodeficiency.
We have demonstrated that infection by these viruses is markedly
enhanced in cells previously infected by a nonpathogenic virus that
uses a distinct receptor. This phenomenon depends on expression of the
receptor and envelope protein of the nonpathogenic virus, but
unexpectedly, not the receptor for the pathogenic virus. Recent
experiments indicate that repeated rounds of infection are mediated by
a direct interaction between the non-RBD portion of the pathogenic
virus and the RBD of the inciting virus. This observation reveals a new
and robust pathway for virus spread in natural infection. In addition,
cooperative infection among retroviruses may provide a previously
unappreciated mechanism for virus evolution.
Infection by Avian Leukosis Viruses
In collaboration with John Young and his colleagues (University of
Wisconsin), we studied the requirements for infection by avian leukosis
viruses (ALVs). These viruses do not contain a thiol reductase motif
and utilize receptors that are structurally distinct from MLV
receptors. These studies reveal that ALV, but not MLV, infection is
blocked by inhibitors of endosomal trafficking and acidification. The
effect of acid pH on receptor-mediated ALV infection and
envelope-induced syncytia formation was measured. Additional studies
were also performed in which the inactivation of ALV infection by
preincubation in acid pH and/or the soluble ectodomain of the receptor
were correlated with the conformational change in TM. These experiments
indicate that both receptor binding and acid pH are required to trigger
the conformational change in TM and to inactivate the virus. Our
findings differ strikingly from the results of similar studies of
classically pH-dependent viruses, such as influenza and vesicular
stomatitis virus, in which acid pH alone is sufficient to trigger the
fusion mechanism irreversibly. Receptor binding is required for
trafficking of these viruses to endosomes, but, unlike ALV, is not
required for priming of envelope glycoproteins for activation by acid
pH. We are developing new assays to characterize the changes in the ALV
envelope that are induced by receptor binding and by acid pH.
Previous studies have indicated that infection by all retroviruses
depends on binding of SU to receptor and by activation of a highly
conserved mechanism of membrane fusion mediated by conformational
changes in TM. Our studies indicate that the molecular mechanisms by
which receptor binding is coupled to fusion are markedly different for
ALV and for MLV. A detailed examination of the importance of these
mechanisms for infection by other retroviruses has not been
performed.
Application to Retroviral Targeting
We exploited our observation that MLV infection can be supported by
purified RBD to develop a method for targeting infection by retroviral
vectors to specific cells. We prepared Friend MLV in which the hormone,
erythropoietin, was inserted into the envelope protein in place of RBD.
Viruses expressing the chimeric envelope protein bound specifically to
cells that expressed the erythropoietin receptor, but were defective
for infection. However, addition of purified RBD to the culture medium
increased infection by this virus by more than 100,000-fold. In
collaboration with Gary Gilliland (HHMI,
Harvard Medical School), we are now adapting this protocol to achieve
targeted infection of the erythroid lineage of hematopoietic cells by
retroviral vectors expressing the chimeric envelope protein.
Last updated October 01, 2002
|
|
|
|
