Role of the C-terminal tryptophan residue for the structure-function of the alphavirus capsid protein

The Semliki Forest virus capsid protein is a multifunctional protein which packages genomic RNA into nucleocapsid structures and binds to viral spike protein during budding. In addition, the capsid protein has an autoproteolytic activity whereby the C-terminal tryptophan is used as the substrate for cotranslational cleavage of the viral structure polyprotein. The autoproteolytic domain of the capsid protein has a chymotrypsin-like fold but has two additional short beta-strands which place the tryptophan into the active site. Here, we have substituted the C-terminal tryptophan of Semliki Forest virus capsid protein for alanine, arginine and phenylalanine and analysed the effects on different functions of the C protein such as nucleocapsid formation, spike binding and autoproteolytic activity. We found that (i) tryptophan is a better substrate for the autoproteolytic activity, (ii) the wild-type tryptophan is the only residue that allows efficient viral growth and (iii) an aromatic residue is important for correct initial folding and stability of the protein.     

Proteolytic refolding of the HIV-1 capsid protein amino-terminus facilitates viral core assembly

After budding, the human immunodeficiency virus (HIV) must 'mature' into an infectious viral particle. Viral maturation requires proteolytic processing of the Gag polyprotein at the matrix-capsid junction, which liberates the capsid (CA) domain to condense from the spherical protein coat of the immature virus into the conical core of the mature virus. We propose that upon proteolysis, the amino-terminal end of the capsid refolds into a beta-hairpin/helix structure that is stabilized by formation of a salt bridge between the processed amino-terminus (Pro1) and a highly conserved aspartate residue (Asp51). The refolded amino-terminus then creates a new CA-CA interface that is essential for assembling the condensed conical core. Consistent with this model, we found that recombinant capsid proteins with as few as four matrix residues fused to their amino-termini formed spheres in vitro, but that removing these residues refolded the capsid amino-terminus and redirected protein assembly from spheres to cylinders. Moreover, point mutations throughout the putative CA-CA interface blocked capsid assembly in vitro, core assembly in vivo and viral infectivity. Disruption of the conserved amino-terminal capsid salt bridge also abolished the infectivity of Moloney murine leukemia viral particles, suggesting that lenti- and oncoviruses mature via analogous pathways.     

Self-association of herpes simplex virus type 1 ICP35 is via coiled-coil interactions and promotes stable interaction with the major capsid protein

The ordered copolymerization of viral proteins to form the herpes simplex virus (HSV) capsid occurs within the nucleus of the infected cell and is a complex process involving the products of at least six viral genes. In common with capsid assembly in double-stranded DNA bacteriophages, HSV capsid assembly proceeds via the assembly of an outer capsid shell around an interior scaffold. This capsid intermediate matures through loss of the scaffold and packaging of the viral genomic DNA. The interior of the HSV capsid intermediate contains the viral protease and assembly protein which compose the scaffold. Proteolytic processing of these proteins is essential for and accompanies capsid maturation. The assembly protein (ICP35) is the primary component of the scaffold, and previous studies have demonstrated it to be capable of intermolecular association with itself and with the major capsid protein, VP5. We have defined structural elements within ICP35 which are responsible for intermolecular self-association and for interaction with VP5. Yeast (Saccharomyces cerevisiae) two-hybrid assays and far-Western studies with purified recombinant ICP35 mapped a core self-association domain between Ser165 and His219. Site-directed mutations in this domain implicate a putative coiled coil in ICP35 self-association. This coiled-coil motif is highly conserved within the assembly proteins of other alpha herpesviruses. In the two-hybrid assay the core self-association domain was sufficient to mediate stable self-association only in the presence of additional structural elements in either N- or C-terminal flanking regions. These regions also contain conserved sequences which exhibit a high propensity for alpha helicity and may contribute to self-association by forming additional short coiled coils. Our data supports a model in which ICP35 molecules have an extended conformation and associate in parallel orientation through homomeric coiled-coil interactions. In additional two-hybrid experiments we evaluated ICP35 mutants for association with VP5. We discovered that in addition to the C-terminal 25 amino acids of ICP35, previously shown to be required for VP5 binding, an additional upstream region was required. This region is between Ser165 and His234 and contains the core self-association domain. Site-directed mutations and construction of chimeric molecules in which the self-association domain of ICP35 was replaced by the GCN4 leucine zipper indicated that this region contributes to VP5 binding through mediating self-association of ICP35 and not through direct binding interactions. Our results suggest that self-association of ICP35 strongly promotes stable association with VP5 in vivo and are consistent with capsid formation proceeding via formation of stable subassemblies of ICP35 and VP5 which subsequently assemble into capsid intermediates in the nucleus.     

The role of nucleocapsid of HIV-1 in virus assembly

The role of the nucleocapsid protein of HIV-1 Gag in virus assembly was investigated using Gag truncation mutants, a nucleocapsid deletion mutant, and point mutations in the nucleocapsid region of Gag, in transfected COS cells, and in stable T-cell lines. Consistent with previous investigations, a truncation containing only the matrix and capsid regions of Gag was unable to assemble efficiently into particles; also, the pelletable material released was lighter than the density of wild-type HIV-1. A deletion mutant lacking p7 nucleocapsid but containing the C-terminal p6 protein was also inefficient in particle release and released lighter particles, while a truncation containing only the first zinc finger of p7 could assemble more efficiently into virions. These results clearly show that p7 is indispensable for virus assembly and release. Some point mutations in the N-terminal basic domain and in the basic linker region between the two zinc fingers, which had been previously shown to have reduced RNA binding in vitro [Schmalzbauer, E., Strack, B., Dannull, J., Guehmann, S., and Moelling, K. (1996). J. Virol. 70: 771-777], were shown to reduce virus assembly dramatically when expressed in full-length viral clones. A fusion protein consisting of matrix and capsid fused to a heterologous viral protein known to have nonspecific RNA binding activity [Ribas, J. C., Fujimura, T., and Wickner, R. B. (1994) J. Biol. Chem. 269: 28420-28428] released pelletable material slightly more efficiently than matrix and capsid alone, and these particles had density higher than matrix and capsid alone. These results demonstrate the essential role of HIV-1 nucleocapsid in the virus assembly process and show that the positively charged N terminus of p7 is critical for this role.     

A proline-rich motif (PPPY) in the Gag polyprotein of Mason-Pfizer monkey virus plays a maturation-independent role in virion release

Virus assembly represents one of the last steps in the retrovirus life cycle. During this process, Gag polyproteins assemble at specific sites within the cell to form viral capsids and induce membrane extrusion (viral budding) either as assembly progresses (type C virus) or following formation of a complete capsid (type B and type D viruses). Finally, the membrane must undergo a fusion event to pinch off the particle in order to release a complete enveloped virion. Structural elements within the MA region of the Gag polyprotein define the route taken to the plasma membrane and direct the process of virus budding. Results presented here suggest that a distinct region of Gag is necessary for virus release. The pp24 and pp16 proteins of the type D retrovirus Mason-Pfizer monkey virus (M-PMV) are phosphoproteins that are encoded in the gag gene of the virus. The pp16 protein is a C-terminally located cleavage product of pp24 and contains a proline-rich motif (PPPY) that is conserved among the Gag proteins of a wide variety of retroviruses. By performing a functional analysis of this coding region with deletion mutants, we have shown that the pp16 protein is dispensable for capsid assembly but essential for virion release. Moreover, additional experiments indicated that the virus release function of pp16 was abolished by the deletion of only the PPPY motif and could be restored when this motif alone was reinserted into a Gag polyprotein lacking the entire pp16 domain. Single-amino-acid substitutions for any of the residues within this motif confer a similar virion release-defective phenotype. It is unlikely that the function of the proline-rich motif is simply to inhibit premature activation of protease, since the PPPY deletion blocked virion release in the context of a protease-defective provirus. These results demonstrate that in type D retroviruses a PPPY motif plays a key role in a late stage of virus budding that is independent of and occurs prior to virion maturation.     

Virus maturation by budding

Enveloped viruses mature by budding at cellular membranes. It has been generally thought that this process is driven by interactions between the viral transmembrane proteins and the internal virion components (core, capsid, or nucleocapsid). This model was particularly applicable to alphaviruses, which require both spike proteins and a nucleocapsid for budding. However, genetic studies have clearly shown that the retrovirus core protein, i.e., the Gag protein, is able to form enveloped particles by itself. Also, budding of negative-strand RNA viruses (rhabdoviruses, orthomyxoviruses, and paramyxoviruses) seems to be accomplished mainly by internal components, most probably the matrix protein, since the spike proteins are not absolutely required for budding of these viruses either. In contrast, budding of coronavirus particles can occur in the absence of the nucleocapsid and appears to require two membrane proteins only. Biochemical and structural data suggest that the proteins, which play a key role in budding, drive this process by forming a three-dimensional (cage-like) protein lattice at the surface of or within the membrane. Similarly, recent electron microscopic studies revealed that the alphavirus spike proteins are also engaged in extensive lateral interactions, forming a dense protein shell at the outer surface of the viral envelope. On the basis of these data, we propose that the budding of enveloped viruses in general is governed by lateral interactions between peripheral or integral membrane proteins. This new concept also provides answers to the question of how viral and cellular membrane proteins are sorted during budding. In addition, it has implications for the mechanism by which the virion is uncoated during virus entry.     

The RER-localized rotavirus intracellular receptor: a truncated purified soluble form is multivalent and binds virus particles

A budding event transfers the immature, single-shelled rotavirus particle (SSP) across the RER membrane prior to assembly of mature virions in the ER lumen. Budding is triggered by the interaction of the SSP with a viral receptor glycoprotein (NS28) which is located in the RER membrane. We have expressed the cytoplasmic domain of the NS28 receptor as a glutathione S-transferase fusion protein to generate a soluble polypeptide that in turn can be cleaved to yield a carboxy-terminal receptor domain. The soluble terminal domain (delta 1-85 NS28) has been purified to homogeneity and retains SSP-binding activity when immobilized on a solid matrix. Integral membrane status therefore is not an essential prerequisite for ligand binding. The Kd for the interaction between immobilized delta 1-85 NS28 and purified particles is 4.6 x 10(-11) M, a value indistinguishable from the value obtained for the full-length and membrane-anchored receptor. Cross-linking with the bifunctional reagent dimethylsuberimidate indicates that delta 1-85 NS28 is a tetramer. When delta 1-85 NS28 is added to a monodisperse suspension of purified virus, the particles aggregate, indicating that the receptor is multivalent. The rotavirus intracellular receptor therefore provides a model for the detailed analysis of the early events that trigger the budding of cytoplasmically located particles across cell membranes.     

Two-helper RNA system for production of recombinant Semliki forest virus particles

Alphavirus expression systems based on suicidal virus particles carrying recombinant replicons have proven to be a very efficient way to deliver genes for heterologous protein expression. However, present strategies for production of such particles have biosafety limitations due to the generation, by RNA recombination, of replication-proficient viruses (RPVs). Here we describe a new packaging system for Semliki Forest virus (SFV) based on a the use of a two-helper system in which the capsid and spike proteins of the C-p62-6K-E1 polyprotein are expressed from two independent RNA molecules. The capsid gene contains a translational enhancer and therefore that sequence was also engineered in front of the spike sequence p62-6K-E1. A sequence coding for the foot-and-mouth disease virus 2A autoprotease was inserted in frame between the capsid translational enhancer and the spike genes. This allows production of the spike proteins at high levels with cotranslational removal of the enhancer sequence and normal biosynthesis of the spike complex. The autoprotease activity of the capsid protein was abolished by mutation, further increasing the biosafety of the system. Cotransfection of cells with both helper RNAs and an SFV vector replicon carrying the LacZ gene led to production of recombinant particles with titers of up to 8 x 10(8) particles per 10(6) cells. Extensive analysis failed to demonstrate the presence of any RPVs, emphasizing the high biosafety of the system based on two-helper RNAs.     

Virus-specific interaction between the human cytomegalovirus major capsid protein and the C terminus of the assembly protein precursor

We previously identified a minimal 12-amino-acid domain in the C terminus of the herpes simplex virus type 1 (HSV-1) scaffolding protein which is required for interaction with the HSV-1 major capsid protein. An alpha-helical structure which maximizes the hydropathicity of the minimal domain is required for the interaction. To address whether cytomegalovirus (CMV) utilizes the same strategy for capsid assembly, several glutathione S-transferase fusion proteins to the C terminus of the CMV assembly protein precursor were produced and purified from bacterial cells. The study showed that the glutathione S-transferase fusion containing 16 amino acids near the C-terminal end was sufficient to interact with the major capsid protein. Interestingly, no cross-interaction between HSV-1 and CMV could be detected. Mutation analysis revealed that a three-amino-acid region at the N-terminal side of the central Phe residue of the CMV interaction domain played a role in determining the viral specificity of the interaction. When this region was converted so as to correspond to that of HSV-1, the CMV assembly protein domain lost its ability to interact with the CMV major capsid protein but gained full interaction with the HSV-1 major capsid protein. To address whether the minimal interaction domain of the CMV assembly protein forms an alpha-helical structure similar to that in HSV-1, peptide competition experiments were carried out. The results showed that a cyclic peptide derived from the interaction domain with a constrained (alpha-helical structure competed for interaction with the major capsid protein much more efficiently than the unconstrained linear peptide. In contrast, a cyclic peptide containing an Ala substitution for the critical Phe residue did not compete for the interaction at all. The results of this study suggest that (i) CMV may have developed a strategy similar to that of HSV-1 for capsid assembly; (ii) the minimal interaction motif in the CMV assembly protein requires an alpha-helix for efficient interaction with the major capsid protein; and (iii) the Phe residue in the CMV minimal interaction domain is critical for interaction with the major capsid protein.     

Coronavirus particle assembly: primary structure requirements of the membrane protein

Coronavirus-like particles morphologically similar to normal virions are assembled when genes encoding the viral membrane proteins M and E are coexpressed in eukaryotic cells. Using this envelope assembly assay, we have studied the primary sequence requirements for particle formation of the mouse hepatitis virus (MHV) M protein, the major protein of the coronavirion membrane. Our results show that each of the different domains of the protein is important. Mutations (deletions, insertions, point mutations) in the luminal domain, the transmembrane domains, the amphiphilic domain, or the carboxy-terminal domain had effects on the assembly of M into enveloped particles. Strikingly, the extreme carboxy-terminal residue is crucial. Deletion of this single residue abolished particle assembly almost completely; most substitutions were strongly inhibitory. Site-directed mutations in the carboxy terminus of M were also incorporated into the MHV genome by targeted recombination. The results supported a critical role for this domain of M in viral assembly, although the M carboxy terminus was more tolerant of alteration in the complete virion than in virus-like particles, likely because of the stabilization of virions by additional intermolecular interactions. Interestingly, glycosylation of M appeared not essential for assembly. Mutations in the luminal domain that abolished the normal O glycosylation of the protein or created an N-glycosylated form had no effect. Mutant M proteins unable to form virus-like particles were found to inhibit the budding of assembly-competent M in a concentration-dependent manner. However, assembly-competent M was able to rescue assembly-incompetent M when the latter was present in low amounts. These observations support the existence of interactions between M molecules that are thought to be the driving force in coronavirus envelope assembly.     

Antiviral agent blocks breathing of the common cold virus

A dynamic capsid is critical to the events that shape the viral life cycle; events such as cell attachment, cell entry, and nucleic acid release demand a highly mobile viral surface. Protein mass mapping of the common cold virus, human rhinovirus 14 (HRV14), revealed both viral structural dynamics and the inhibition of such dynamics with an antiviral agent, WIN 52084. Viral capsid digestion fragments resulting from proteolytic time-course experiments provided structural information in good agreement with the HRV14 three-dimensional crystal structure. As expected, initial digestion fragments included peptides from the capsid protein VP1. This observation was expected because VP1 is the most external viral protein. Initial digestion fragments also included peptides belonging to VP4, the most internal capsid protein. The mass spectral results together with x-ray crystallography data provide information consistent with a "breathing" model of the viral capsid. Whereas the crystal structure of HRV14 shows VP4 to be the most internal capsid protein, mass spectral results show VP4 fragments to be among the first digestion fragments observed. Taken together this information demonstrates that VP4 is transiently exposed to the viral surface via viral breathing. Comparative digests of HRV14 in the presence and absence of WIN 52084 revealed a dramatic inhibition of digestion. These results indicate that the binding of the antiviral agent not only causes local conformational changes in the drug binding pocket but actually stabilizes the entire viral capsid against enzymatic degradation. Viral capsid mass mapping provides a fast and sensitive method for probing viral structural dynamics as well as providing a means for investigating antiviral drug efficacy.     

The formation of intramolecular disulfide bridges is required for induction of the Sindbis virus mutant ts23 phenotype

The Sindbis virus envelope protein spike is a hetero-oligomeric complex composed of a trimer of glycoprotein E1-E2 heterodimers. Spike assembly is a multistep process which occurs in the endoplasmic reticulum (ER) and is required for the export of E1 from the ER. PE2 (precursor to E2), however, can transit through the secretory pathway and be expressed at the cell surface in the absence of E1. Although oligomer formation does not appear to be required for the export of PE2, there is evidence that defects in E1 folding can affect PE2 transit from the ER. Temperature-sensitive mutant ts23 of Sindbis virus contains two amino acid substitutions in E1, while PE2 and capsid protein have the wild-type sequence; however, at the nonpermissive temperature, both E1 and PE2 are retained within the ER and can be isolated in protein aggregates with the molecular chaperone GRP78-BiP. We previously demonstrated that the temperature sensitivity for ts23 was lost as oligomer formation took place at the permissive temperature, suggesting that temperature sensitivity is initiated early in the process of viral spike assembly (M. Carleton and D. T. Brown, J. Virol. 70:952-959, 1996). Experiments described herein investigated the defects in envelope protein maturation that occur in ts23-infected cells and which result in retention of both envelope proteins in the ER. The data demonstrate that in ts23-infected cells incubated at the nonpermissive temperature, E1 folding is disrupted early after synthesis, resulting in the rapid incorporation of both E1 and PE2 into disulfide-stabilized aggregates. Furthermore, the aberrant E1 conformation which is responsible for induction of the ts phenotype requires the formation of intramolecular disulfide bridges formed prior to E1 association with PE2 and the completion of E1 folding.     

The transmembrane domains of Sindbis virus envelope glycoproteins induce cell death

Sindbis virus, the prototype alphavirus, kills cells by inducing apoptosis. To investigate potential mechanisms by which Sindbis virus induces apoptosis, we examined whether specific viral gene products were able to induce cell death. Genes encoding the three structural proteins--capsid, the precursor E1 (6K plus E1), and the precursor E2 (P62 or E3 plus E2)--were cotransfected with a beta-galactosidase reporter plasmid in transient-transfection assays in rat prostate adenocarcinoma AT3 cells. Cell death, as determined by measuring the loss of blue cells, was observed in AT3 cells transfected with 6K plus E1 and with P62 but not in cells transfected with capsid. Deletion mutagenesis of P62 indicated that large regions of the cytoplasmic domain and extracellular domain were not essential for the induction of cell death. However, constructs containing the minimal E3 signal sequence fused to the E2 transmembrane domain and the minimal E3 signal sequence fused to the E1 transmembrane domain induced death as efficiently as full-length P62 and 6K plus E1, whereas no cell death was observed after transfection with a control construct containing the E3 signal sequence linked to the transmembrane domain of murine CD4. These data demonstrate that intracellular expression of the transmembrane domains of the Sindbis virus envelope glycoproteins can kill AT3 cells.     

Effects of nucleocapsid mutations on human immunodeficiency virus assembly and RNA encapsidation

The human immunodeficiency virus (HIV) Pr55Gag precursor proteins direct virus particle assembly. While Gag-Gag protein interactions which affect HIV assembly occur in the capsid (CA) domain of Pr55Gag, the nucleocapsid (NC) domain, which functions in viral RNA encapsidation, also appears to participate in virus assembly. In order to dissect the roles of the NC domain and the p6 domain, the C-terminal Gag protein domain, we examined the effects of NC and p6 mutations on virus assembly and RNA encapsidation. In our experimental system, the p6 domain did not appear to affect virus release efficiency but p6 deletions and truncations reduced the specificity of genomic HIV-1 RNA encapsidation. Mutations in the nucleocapsid region reduced particle release, especially when the p2 interdomain peptide or the amino-terminal portion of the NC region was mutated, and NC mutations also reduced both the specificity and the efficiency of HIV-1 RNA encapsidation. These results implicated a linkage between RNA encapsidation and virus particle assembly or release. However, we found that the mutant ApoMTRB, in which the nucleocapsid and p6 domains of HIV-1 Pr55Gag were replaced with the Bacillus subtilis MtrB protein domain, released particles efficiently but packaged no detectable RNA. These results suggest that, for the purposes of virus-like particle assembly and release, NC can be replaced by a protein that does not appear to encapsidate RNA.     

Visualization of a 4-helix bundle in the hepatitis B virus capsid by cryo-electron microscopy

Despite the development of vaccines, the hepatitis B virus remains a major cause of human liver disease. The virion consists of a lipoprotein envelope surrounding an icosahedral capsid composed of dimers of a 183-residue protein, 'core antigen' (HBcAg). Knowledge of its structure is important for the design of antiviral drugs, but it has yet to be determined. Residues 150-183 are known to form a protamine-like domain required for packaging RNA, and residues 1-149 form the 'assembly domain' that polymerizes into capsids and, unusually for a capsid protein, is highly alpha-helical. Density maps calculated from cryo-electron micrographs show that the assembly domain dimer is T-shaped: its stem constitutes the dimer interface and the tips of its arms make the polymerization contacts. By refining the procedures used to calculate the map, we have extended the resolution to 9 A, revealing major elements of secondary structure. In particular, the stem, which protrudes as a spike on the capsid's outer surface, is a 4-helix bundle, formed by the pairing of alpha-helical hairpins from both subunits.     

Incorporation of human immunodeficiency virus type 1 Gag proteins into murine leukemia virus virions

The retroviral Gag polyprotein is necessary and sufficient for assembly and budding of viral particles. However, the exact inter- and intramolecular interactions of the Gag polyproteins during this process are not known. To locate functional domains within Gag, we generated chimeric proviruses between human immunodeficiency virus type 1 (HIV-1) and murine leukemia virus (MuLV). In these chimeric proviruses, the matrix or capsid proteins of MuLV were precisely replaced with the matrix or capsid proteins of HIV-1. Although the chimeric proviruses were unable to efficiently assemble into mature viral particles by themselves, coexpression of wild-type MuLV Gag rescued the HIV proteins into virions. The specificity of the rescue of HIV proteins into MuLV virions shows that specific interactions involving homologous matrix or capsid regions of Gag are necessary for retroviral particle formation.     

Hepatitis B virus capsid: localization of the putative immunodominant loop (residues 78 to 83) on the capsid surface, and implications for the distinction between c and e-antigens

Hepatitis B virus capsid protein comprises a 149 residue "assembly" domain that polymerizes into icosahedral particles, and a 34 residue RNA-binding "protamine" domain. Recently, the capsid structure has been studied to resolutions below 10 A by cryo-electron microscopy, revealing much of its alpha-helical substructure and that it appears to have a novel fold for a capsid protein; however, the resolution is still too low for chain-tracing by conventional criteria. Aiming to establish a fiducial marker to aid in the process of chain-tracing, we have used cryo-microscopy to pinpoint the binding site of a monoclonal antibody that recognizes the peptide from residues 78 to 83. This epitope resides on the outer rim of the 30 A long spikes that protrude from the capsid shell. These spikes are four-helix bundles formed by the pairing of helix-turn-helix motifs from two subunits; by means of a tilting experiment, we have determined that this bundle is right-handed. Variants of the same protein present two clinically important and non-crossreactive antigens: core antigen (HBcAg), which appears early in infection as assembled capsids; and the sentinel e-antigen (HBeAg), a non-particulate form. Knowledge of the binding site of our anti-HBcAg antibody bears on the molecular basis of the distinction between the two antigens, which appears to reflect conformational differences between the assembled and unassembled states of the capsid protein dimer, in addition to epitope masking in capsids.