Cell, Vol. 69, 577-576.
May 15, 1992, Copyright
0 1992 by Cell Press
Unpacking the Incoming Influenza Virus Ari Helenius Department Yale School New Haven.
of Cell Biology of Medicine Connecticut 06510
The role of virus particles is to transport the viral genome and accessory proteins from infected cells and organisms to the noninfected. In the virion, the genome is present in a transport configuration; it is condensed, inactive, and protected from the extracellular environment. When the virus enters a new cell, the genome must be uncoated to allow replication and transcription. Many of the macromolecular interactions, which are specifically established to assemble all the viral components, to condense the viral RNA or DNA, and to hold the particle together, must be reversed. For many viruses, the uncoating process constitutes something of a paradox. Why does a structure that is extraordinarily stable outside the cell suddenly fall apart when it reaches a new host cell, and why do incoming capsids disassemble, when progeny virus particles in the same cell only a few hours later remain fully intact? There are two possible general solutions to the assembly-disassembly paradox. The virus particles, after leaving the cell of origin, may undergo some alteration that primes them for uncoating upon entry into the next cell. Alternatively, the cell may change during infection, so that it can only support uncoating during a limited time at the beginning of the infection. In the former case, the incoming viruses would be structurally different from the newly assembled ones. In the latter, the cells could, for instance, provide an uncoating factor that would be inactivated when the structural components of the virus start to be synthesized (see Wengler, 1987). Although these are crucial questions for understanding the replication cycle of viruses, they have, with a few exceptions, not been adequately addressed. The interactions that determine the composition and stability of influenza A viruses are shown schematically in the figure. The eight viral RNA molecules are individually packaged into helical viral ribonucleoprotein particles (vRNPs) covered by numerous copies of nucleoprotein (NP). The vRNPs are associated with the matrix protein M,, which also interacts with the cytoplasmic tails of the spike glycoproteins and/or the viral membrane (see figure and Lamb, 1989). This hierarchy of interactions is established during a stepwise assembly process in the infected cell. When the virus enters the cytoplasm of a new host cell (by a hemagglutinin [HA]-mediated, low pH-activated membrane fusion reaction in a late endosome [Matlin et al., 19821) dissociation of spike glycoproteins, M,, and the vRNPs is observed (Martin and Helenius, 1991 a). Thereafter, the vRNPs rapidly enter the nucleus and initiate replication and transcription. The topic of the paper by Pinto et al. (1992) may at first glance have little to do with the regulation of influenza
Minireview
nucleocapsid disassembly. The paper provides electrophysiological evidence that influenza virus MP, a tetramerit transmembrane protein (see figure), serves as a channel for monovalent cations. However, aside from providing proof for the existence of virus-encoded ion channels, the paper supplies the latest chapter in an exciting story that has revealed how an animal virus may be primed to undergo its disassembly program upon entry into a host cell. The MP gene was identified by Lamb and coworkers as an open reading frame in the seventh RNA of the virus, which also codes for MI. It was subsequently found to give rise to a spliced mRNA and a type Ill membrane protein (97 aa) with only eight N-terminal residues in common with MI (Lamb et al., 1985). The protein is translated in the endoplasmic reticulum with 24 N-terminal residues in the lumen, 54 residues in the cytosol, and a single membranespanning segment. It was found to be a disulfide-bonded homotetramer (Holsinger and Lamb, 1991; Sugrue and Hay, 1991) that is abundantly expressed in the plasma membrane and the organelles of the secretory pathway. It is also present in mature virus particles, but only at a level of 4-16 tetramers per virion (Zebedee and Lamb, 1988). The first clue as to the function of MP came from a pharmacological observationrecombination and sequencing studies linked the inhibitory action of amantadine (l-aminoadamantane hydrochloride) at low concentrations exclusively to MP. Amantadine is an anti-influenza drug, mainly used for prophylactic purposes in risk groups during influenza A epidemics. Point mutations in MP that render human viruses resistant to amantadine were found to be clustered within the transmembrane domain of this small protein (Hay et al., 1985; Lamb et al., 1985). Since amantadine was already known to block an early step in the replication cycle, these findings also revealed that the MP in the virus particles must somehow play a role during influenza entry into the cell.
Molecular Virus
Interactions
between
Structural
Components
of Influenza
NP, nucleoprotein; HA, hemagglutinin; NA. neuraminidase. M2, which is depkcted as a protein channel, is a minority component present only at a level of 4-16 tetrameric copies per vlrion.
Cell 578
The main stimulus for the most recent studies have come from an analysis of the transport of HA from the endoplasmic reticulum to the surface of cells infected with an avian influenza strain (Sugrue et al., 1990). It was found that, when micromolar concentrations of amantadine are present, HA emerges on the cell surface in its low-pH conformation. Normally, this is only seen after acid-induced membrane fusion in the endosomes during virus entry. Apparently, the presence of the drug causes the HA molecules to be exposed to low pH en route from the endoplasmic reticulum to the cell surface. Sugrue et al. hypothesized that MOmolecules might be responsible for elevating the otherwise mildly acidic pH in the transQolgi network or the Golgi complex by forming proton channels through which the proton gradient between the lumen of these organelles and the cytosol would be dissipated. The elevation in luminal pH would allow passage of HA without acid conversion. When blocked by amantadine, the MP channels would not be able to perform this function, and the HA molecules would undergo a premature conformational conversion that would prevent the formation of infective virus. Although such an inhibitory mechanism of amantadine probably only applies to virus strains with a particularly high pH threshold for conversion (such as the avian virus used by Sugrue et al.), the observations contributed two important concepts: MP isan ion channel, and amantadine an MP channel blocker. Both of these conclusions are now unambiguously demonstrated by the elegant voltage clamp experiments of Pinto et al. (1992). M, molecules expressed in Xenopus oocytes are shown to form amantadine-sensitive channels for monovalent cations in the plasma membrane. That the channels prove to be acidactivated is quite satisfying, because this explains why cells are relatively undisturbed for hours after infection although they have plenty of M2 in their plasma membrane. It was not possible to measure a H’ conductance for technical reasons, but it is more than likely that protons can also flow through the the MP channels. While the channel activity of MP provides an explanation for the late effects of amantadine, it is not immediately clear what role amantadine and the MP channel could play during virus entry. Entry is, after all, the main step inhibited by amantadine during human flu infections (Skehel et al., 1977). It is known from previous work that micromolar concentrations of amantadine do not block virus binding to cells nor their internalization by receptor-mediated endocytosis. When exposed to low pH, the virus is able to fuse in the presence of amantadine, and it can release its nucleocapsids into the cytosol from late endosomes, where the pH is low enough to trigger the above-mentioned conformational change in HA. Two groups have reported that the dissociation of MI from the vRNPs as well as their transport from the cytosol into the nucleus is inhibited (Bukrinskaya et al., 1982; Martin and Helenius, 1991 b). From this, one can tentatively conclude that the amantadineinduced block involves nucleocapsid uncoating and/or nuclear transport. But why would MP, a membrane protein left behind in the endosomal membrane after fusion, have anything to do
with uncoating and nuclear transport? The explanation is probably found in events that occur during the 25 min lag period that the virus spends in endosomes before fusion takes place. We have proposed that the role of MP in the viral membrane is to allow protons to enter the virus particle while the particle is still in early endosomes (Martin and Helenius, 1991 b). Inside the viral particle, these protons may induce changes in the conformation of the nucleocapsid, thus priming it for dissociation upon penetration into the cytosol. One such change could be the loosening of M,-vRNP interactions, which would result in the observed dissociation of MI from vRNPs and the separation of individual vRNPs (Martin and Helenius, 1991a). Support for this notion comes from detergent solubilization studies that demonstrate that the interaction between MI and vRNPs, which are normally very tight in the virus particle, is disrupted by mildly acidic pH (Zhirnov, 1990). Another alteration could result in the dissociation of glycoproteins and MI, an effect also easily observed when incoming viruses are traced by immunofluorescence (Martin and Helenius, 1991 a). Changes within vRNPs themselves could also occur, allowing their further uncoating inside the nucleus. If correct, this model provides a solution for the assembly-disassembly paradox as far as influenza virus is concerned. It predicts that the incoming nucleocapsids are structurally and functionally different from those later assembled for export in the same cell. The reason is that they have taken an acid bath during entry. Whether this is the full story, time will tell. In the meantime, the work of Pinto et al. provides new assays and new approaches that should allow a more target-oriented development of antiinfluenza drugs, as well as encouragement to those analyzing obscure open reading frames and small proteins in other viruses. References Eukrinskaya, A. G., Vorkunova, J. Gen. Viral. 60, 49-59. Hay, A. J., Wolstenholme, EMBO J. 4, 3021-3024. Holsinger,
N. K., and Pushkarskaya.
A. J., Skehel,
L. J., and Lamb,
N. L. (1982).
J. J., and Smith, M. H. (1985).
Ft. A. (1991).
Virology
183, 32-43
Lamb, R. A. (1989). In The Influenza Viruses (The Viruses, R. M. Krug, ed. (New York: Plenum Press), pp. 1-87.
Vol. 6)
Lamb, R. A., Zebedee, 627-633.
Cell 40,
S. L., and Richardson,
C. D. (1985).
Martin,
K., and Helenius,
A. (1991a).
Martin,
K.. and Helenius,
A. (1991 b). Cell 67, 117-130.
Matlin, K. S., Reggio, Biol. 97, 601-613.
H., Helenius,
Pinto, L. H., Holsinger. 528.
J. Virol. 265, 232-244. A., and Simons,
L. J., and Lamb,
Skehel, J. J., Hay, A. J.. and Armstrong, 97-l 10. Sugrue,
R. J., and Hay, A. J. (1991).
Sugrue, R. J.. Bahdur. G., Zambon, A. R., and Hay, A. J. (1990). EMBO Wengler.
G. (1987).
Zebedee,
S. L., and Lamb,
Zhirnov,
0. P. (1990).
Arch.
K. (1982).
R. A. (1992).
Cell 69, 517-
J. A. (1977). J. Gen. Virol. 38, Virology
780, 617-624.
M. C.. Hall-Smith. J. 9, 3469-3476.
M., Douglas,
Virol. 94, 1-14. R. A. (1988).
Virology
J. Cell
J. Virol. 62, 2762-2772.
776, 274-279.
May 15, 1992, Copyright
0 1992 by Cell Press
Unpacking the Incoming Influenza Virus Ari Helenius Department Yale School New Haven.
of Cell Biology of Medicine Connecticut 06510
The role of virus particles is to transport the viral genome and accessory proteins from infected cells and organisms to the noninfected. In the virion, the genome is present in a transport configuration; it is condensed, inactive, and protected from the extracellular environment. When the virus enters a new cell, the genome must be uncoated to allow replication and transcription. Many of the macromolecular interactions, which are specifically established to assemble all the viral components, to condense the viral RNA or DNA, and to hold the particle together, must be reversed. For many viruses, the uncoating process constitutes something of a paradox. Why does a structure that is extraordinarily stable outside the cell suddenly fall apart when it reaches a new host cell, and why do incoming capsids disassemble, when progeny virus particles in the same cell only a few hours later remain fully intact? There are two possible general solutions to the assembly-disassembly paradox. The virus particles, after leaving the cell of origin, may undergo some alteration that primes them for uncoating upon entry into the next cell. Alternatively, the cell may change during infection, so that it can only support uncoating during a limited time at the beginning of the infection. In the former case, the incoming viruses would be structurally different from the newly assembled ones. In the latter, the cells could, for instance, provide an uncoating factor that would be inactivated when the structural components of the virus start to be synthesized (see Wengler, 1987). Although these are crucial questions for understanding the replication cycle of viruses, they have, with a few exceptions, not been adequately addressed. The interactions that determine the composition and stability of influenza A viruses are shown schematically in the figure. The eight viral RNA molecules are individually packaged into helical viral ribonucleoprotein particles (vRNPs) covered by numerous copies of nucleoprotein (NP). The vRNPs are associated with the matrix protein M,, which also interacts with the cytoplasmic tails of the spike glycoproteins and/or the viral membrane (see figure and Lamb, 1989). This hierarchy of interactions is established during a stepwise assembly process in the infected cell. When the virus enters the cytoplasm of a new host cell (by a hemagglutinin [HA]-mediated, low pH-activated membrane fusion reaction in a late endosome [Matlin et al., 19821) dissociation of spike glycoproteins, M,, and the vRNPs is observed (Martin and Helenius, 1991 a). Thereafter, the vRNPs rapidly enter the nucleus and initiate replication and transcription. The topic of the paper by Pinto et al. (1992) may at first glance have little to do with the regulation of influenza
Minireview
nucleocapsid disassembly. The paper provides electrophysiological evidence that influenza virus MP, a tetramerit transmembrane protein (see figure), serves as a channel for monovalent cations. However, aside from providing proof for the existence of virus-encoded ion channels, the paper supplies the latest chapter in an exciting story that has revealed how an animal virus may be primed to undergo its disassembly program upon entry into a host cell. The MP gene was identified by Lamb and coworkers as an open reading frame in the seventh RNA of the virus, which also codes for MI. It was subsequently found to give rise to a spliced mRNA and a type Ill membrane protein (97 aa) with only eight N-terminal residues in common with MI (Lamb et al., 1985). The protein is translated in the endoplasmic reticulum with 24 N-terminal residues in the lumen, 54 residues in the cytosol, and a single membranespanning segment. It was found to be a disulfide-bonded homotetramer (Holsinger and Lamb, 1991; Sugrue and Hay, 1991) that is abundantly expressed in the plasma membrane and the organelles of the secretory pathway. It is also present in mature virus particles, but only at a level of 4-16 tetramers per virion (Zebedee and Lamb, 1988). The first clue as to the function of MP came from a pharmacological observationrecombination and sequencing studies linked the inhibitory action of amantadine (l-aminoadamantane hydrochloride) at low concentrations exclusively to MP. Amantadine is an anti-influenza drug, mainly used for prophylactic purposes in risk groups during influenza A epidemics. Point mutations in MP that render human viruses resistant to amantadine were found to be clustered within the transmembrane domain of this small protein (Hay et al., 1985; Lamb et al., 1985). Since amantadine was already known to block an early step in the replication cycle, these findings also revealed that the MP in the virus particles must somehow play a role during influenza entry into the cell.
Molecular Virus
Interactions
between
Structural
Components
of Influenza
NP, nucleoprotein; HA, hemagglutinin; NA. neuraminidase. M2, which is depkcted as a protein channel, is a minority component present only at a level of 4-16 tetrameric copies per vlrion.
Cell 578
The main stimulus for the most recent studies have come from an analysis of the transport of HA from the endoplasmic reticulum to the surface of cells infected with an avian influenza strain (Sugrue et al., 1990). It was found that, when micromolar concentrations of amantadine are present, HA emerges on the cell surface in its low-pH conformation. Normally, this is only seen after acid-induced membrane fusion in the endosomes during virus entry. Apparently, the presence of the drug causes the HA molecules to be exposed to low pH en route from the endoplasmic reticulum to the cell surface. Sugrue et al. hypothesized that MOmolecules might be responsible for elevating the otherwise mildly acidic pH in the transQolgi network or the Golgi complex by forming proton channels through which the proton gradient between the lumen of these organelles and the cytosol would be dissipated. The elevation in luminal pH would allow passage of HA without acid conversion. When blocked by amantadine, the MP channels would not be able to perform this function, and the HA molecules would undergo a premature conformational conversion that would prevent the formation of infective virus. Although such an inhibitory mechanism of amantadine probably only applies to virus strains with a particularly high pH threshold for conversion (such as the avian virus used by Sugrue et al.), the observations contributed two important concepts: MP isan ion channel, and amantadine an MP channel blocker. Both of these conclusions are now unambiguously demonstrated by the elegant voltage clamp experiments of Pinto et al. (1992). M, molecules expressed in Xenopus oocytes are shown to form amantadine-sensitive channels for monovalent cations in the plasma membrane. That the channels prove to be acidactivated is quite satisfying, because this explains why cells are relatively undisturbed for hours after infection although they have plenty of M2 in their plasma membrane. It was not possible to measure a H’ conductance for technical reasons, but it is more than likely that protons can also flow through the the MP channels. While the channel activity of MP provides an explanation for the late effects of amantadine, it is not immediately clear what role amantadine and the MP channel could play during virus entry. Entry is, after all, the main step inhibited by amantadine during human flu infections (Skehel et al., 1977). It is known from previous work that micromolar concentrations of amantadine do not block virus binding to cells nor their internalization by receptor-mediated endocytosis. When exposed to low pH, the virus is able to fuse in the presence of amantadine, and it can release its nucleocapsids into the cytosol from late endosomes, where the pH is low enough to trigger the above-mentioned conformational change in HA. Two groups have reported that the dissociation of MI from the vRNPs as well as their transport from the cytosol into the nucleus is inhibited (Bukrinskaya et al., 1982; Martin and Helenius, 1991 b). From this, one can tentatively conclude that the amantadineinduced block involves nucleocapsid uncoating and/or nuclear transport. But why would MP, a membrane protein left behind in the endosomal membrane after fusion, have anything to do
with uncoating and nuclear transport? The explanation is probably found in events that occur during the 25 min lag period that the virus spends in endosomes before fusion takes place. We have proposed that the role of MP in the viral membrane is to allow protons to enter the virus particle while the particle is still in early endosomes (Martin and Helenius, 1991 b). Inside the viral particle, these protons may induce changes in the conformation of the nucleocapsid, thus priming it for dissociation upon penetration into the cytosol. One such change could be the loosening of M,-vRNP interactions, which would result in the observed dissociation of MI from vRNPs and the separation of individual vRNPs (Martin and Helenius, 1991a). Support for this notion comes from detergent solubilization studies that demonstrate that the interaction between MI and vRNPs, which are normally very tight in the virus particle, is disrupted by mildly acidic pH (Zhirnov, 1990). Another alteration could result in the dissociation of glycoproteins and MI, an effect also easily observed when incoming viruses are traced by immunofluorescence (Martin and Helenius, 1991 a). Changes within vRNPs themselves could also occur, allowing their further uncoating inside the nucleus. If correct, this model provides a solution for the assembly-disassembly paradox as far as influenza virus is concerned. It predicts that the incoming nucleocapsids are structurally and functionally different from those later assembled for export in the same cell. The reason is that they have taken an acid bath during entry. Whether this is the full story, time will tell. In the meantime, the work of Pinto et al. provides new assays and new approaches that should allow a more target-oriented development of antiinfluenza drugs, as well as encouragement to those analyzing obscure open reading frames and small proteins in other viruses. References Eukrinskaya, A. G., Vorkunova, J. Gen. Viral. 60, 49-59. Hay, A. J., Wolstenholme, EMBO J. 4, 3021-3024. Holsinger,
N. K., and Pushkarskaya.
A. J., Skehel,
L. J., and Lamb,
N. L. (1982).
J. J., and Smith, M. H. (1985).
Ft. A. (1991).
Virology
183, 32-43
Lamb, R. A. (1989). In The Influenza Viruses (The Viruses, R. M. Krug, ed. (New York: Plenum Press), pp. 1-87.
Vol. 6)
Lamb, R. A., Zebedee, 627-633.
Cell 40,
S. L., and Richardson,
C. D. (1985).
Martin,
K., and Helenius,
A. (1991a).
Martin,
K.. and Helenius,
A. (1991 b). Cell 67, 117-130.
Matlin, K. S., Reggio, Biol. 97, 601-613.
H., Helenius,
Pinto, L. H., Holsinger. 528.
J. Virol. 265, 232-244. A., and Simons,
L. J., and Lamb,
Skehel, J. J., Hay, A. J.. and Armstrong, 97-l 10. Sugrue,
R. J., and Hay, A. J. (1991).
Sugrue, R. J.. Bahdur. G., Zambon, A. R., and Hay, A. J. (1990). EMBO Wengler.
G. (1987).
Zebedee,
S. L., and Lamb,
Zhirnov,
0. P. (1990).
Arch.
K. (1982).
R. A. (1992).
Cell 69, 517-
J. A. (1977). J. Gen. Virol. 38, Virology
780, 617-624.
M. C.. Hall-Smith. J. 9, 3469-3476.
M., Douglas,
Virol. 94, 1-14. R. A. (1988).
Virology
J. Cell
J. Virol. 62, 2762-2772.
776, 274-279.
