A genetic interaction between the core and NS3 proteins of hepatitis C virus is essential for production of infectious virus

By analogy to other members of the Flaviviridae family, the hepatitis C virus (HCV) core protein is presumed to oligomerize to form the viral nucleocapsid, which encloses the single-stranded RNA genome. Core protein is directed to lipid droplets (LDs) by domain 2 (D2) of the protein, and this process is critical for virus production. Domain 1 (D1) of core is also important for infectious particle morphogenesis, although its precise contribution to this process is poorly understood. In this study, we mutated amino acids 64 to 75 within D1 of core and examined the ability of these mutants to produce infectious virus. We found that residues 64 to 66 are critical for generation of infectious progeny, whereas 67 to 75 were dispensable for this process. Further investigation of the defective 64 to 66 mutant (termed JFH1(T)-64-66) revealed it to be incapable of producing infectious intracellular virions, suggesting a fault during HCV assembly. Furthermore, isopycnic gradient analyses revealed that JFH1(T)-64-66 assembled dense intracellular species of core, presumably representing nucleocapsids. Thus, amino acids 64 to 66 are seemingly not involved in core oligomerization/nucleocapsid assembly. Passaging of JFH1(T)-64-66 led to the emergence of a single compensatory mutation (K1302R) within the helicase domain of NS3 that completely rescued its ability to produce infectious virus. Importantly, the same NS3 mutation abrogated virus production in the context of wild-type core protein. Together, our results suggest that residues 64 to 66 of core D1 form a highly specific interaction with the NS3 helicase that is essential for the generation of infectious HCV particles at a stage downstream of nucleocapsid assembly.     

Biophysical characterization of hepatitis C virus core protein: implications for interactions within the virus and host

A primary function of the hepatitis C virus (HCV) core protein is to package the viral genome within a nucleocapsid. In addition, core protein has been shown to interact with more than a dozen cellular proteins, and these interactions have been suggested to play critical roles in HCV pathogenesis. A more complete knowledge of the biophysical properties of the core protein may help to clarify its role in HCV pathogenesis and nucleocapsid assembly and provide a basis for the development of novel anti-HCV therapies. Here we report that recombinant mature core protein exists as a large multimer in solution under physiological conditions. Far-UV circular dichroism (CD) experiments showed that the mature core protein contains stable secondary structure. Studies with truncated core protein demonstrated that the C-terminal region of the core protein is critical for its folding and oligomerization. Intrinsic fluorescence spectroscopy and near-UV CD analysis indicated that the tryptophan-rich region (residues 76-113) is largely solvent-exposed and not likely responsible for multimerization of the mature core protein in vitro.     

Interfering with hepatitis C virus assembly in vitro using affinity peptides directed towards core protein

Viral assembly is a crucial key step in the life cycle of every virus. In the case of Hepatitis C virus (HCV), the core protein is the only structural protein to interact directly with the viral genomic RNA. Purified recombinant core protein is able to self-assemble in vitro into nucleocapsid-like particles upon addition of a structured RNA, providing a robust assay with which to study HCV assembly. Inhibition of self-assembly of the C170 core protein (first 170 amino acids) was tested using short peptides derived from the HCV core, from HCV NS5A protein, and from diverse proteins (p21 and p73) known to interact with HCV core protein. Interestingly, peptides derived from the core were the best inhibitors. These peptides are derived from regions of the core predicted to be involved in the interaction between core subunits during viral assembly. We also demonstrated that a peptide derived from the C-terminal end of NS5A protein moderately inhibits the assembly process.     

Cytosolic Phospholipase A2 Gamma Is Involved in Hepatitis C Virus Replication and Assembly

Similar to other positive-sense, single-stranded RNA viruses, hepatitis C virus (HCV) replicates its genome in a remodeled intracellular membranous structure known as the membranous web (MW). To date, the process of MW formation remains unclear. It is generally acknowledged that HCV nonstructural protein 4B (NS4B) can induce MW formation through interaction with the cytosolic endoplasmic reticulum (ER) membrane. Many host proteins, such as phosphatidylinositol 4-kinase IIIα (PI4KIIIα), have been identified as critical factors required for this process. We now report a new factor, the cytosolic phospholipase A2 gamma (PLA2G4C), which contributes to MW formation, HCV replication, and assembly. The PLA2G4C gene was identified as a host gene with upregulated expression upon HCV infection. Knockdown of PLA2G4C in HCV-infected cells or HCV replicon-containing cells by small interfering RNA (siRNA) significantly suppressed HCV replication and assembly. In addition, the chemical inhibitor methyl arachidonyl fluorophosphonate (MAFP), which specifically inhibits PLA2, reduced HCV replication and assembly. Electron microscopy demonstrated that MW structure formation was defective after PLA2G4C knockdown in HCV replicon-containing cells. Further analysis by immunostaining and immunoprecipitation assays indicated that PLA2G4C colocalized with the HCV proteins NS4B and NS5A in cells infected with JFH-1 and interacted with NS4B. In addition, PLA2G4C was able to transport the HCV nonstructural proteins from replication sites to lipid droplets, the site for HCV assembly. These data suggest that PLA2G4C plays an important role in the HCV life cycle and might represent a potential target for anti-HCV therapy.     

Assembly of human papillomavirus type-16 virus-like particles: multifactorial study of assembly and competing aggregation

Pentameric capsomeres of human papillomavirus capsid protein L1 expressed in Escherichia coli self-assemble into virus-like particles (VLPs) in vitro. A multifactorial experimental design was used to explore a wide range of solution conditions to optimize the assembly process. The degree of assembly was measured using an enzyme-linked immunosorbent assay, and a high-throughput turbidity assay was developed to monitor competing aggregation. The presence of zinc ions in the assembly buffer greatly increased the incidence of aggregation and had to be excluded from the experiment for meaningful analysis. Assembly of VLPs was optimal at a pH of about 6.5, calcium and sodium ions had no measurable effect, and dithiothreitol and glutathione inhibited assembly. Tryptophan fluorescence spectroscopy demonstrated that an increase in urea concentration reduced the rate of VLP formation but had no effect on the final concentration of assembled VLPs. This study demonstrates the use of the hanging-drop vapor-diffusion crystallization method to screen for conditions that promote aggregation and the use of tryptophan fluorescence spectroscopy for real-time monitoring of the assembly process.     

Carboxyl terminus of severe acute respiratory syndrome coronavirus nucleocapsid protein: self-association analysis and nucleic acid binding characterization

Coronavirus nucleocapsid (N) protein envelops the genomic RNA to form long helical nucleocapsid during virion assembly. Since N protein oligomerization is usually a crucial step in this process, characterization of such an oligomerization will help in the understanding of the possible mechanisms for nucleocapsid formation. The N protein of severe acute respiratory syndrome coronavirus (SARS-CoV) was recently discovered to self-associate by its carboxyl terminus. In this study, to further address the detailed understanding of the association feature of this C-terminus, its oligomerization was systematically investigated by size exclusion chromatography and chemical cross-linking assays. Our results clearly indicated that the C-terminal domain of SARS-CoV N protein could form not only dimers but also trimers, tetramers, and hexamers. Further analyses against six deletion mutants showed that residues 343-402 were necessary and sufficient for this C-terminus oligomerization. Although this segment contains many charged residues, differences in ionic strength have no effects on its oligomerization, indicating the absence of electrostatic force in SARS-CoV N protein C-terminus self-association. Gel shift assay results revealed that the SARS-CoV N protein C-terminus is also able to associate with nucleic acids and residues 363-382 are the responsible interaction partner, demonstrating that this fragment might involve genomic RNA binding sites. The fact that nucleic acid binding could promote the SARS-CoV N protein C-terminus to form high-order oligomers implies that the oligomeric SARS-CoV N protein probably combines with the viral genomic RNA in triggering long nucleocapsid formation.     

Conformational features of truncated hepatitis C virus core protein in virus-like particles

HCVc 120 is a truncated protein from the hepatitis C virus (HCV) core protein that interacts with itself to form nucleocapsid-like particles. We present here the infrared and Raman spectra of oligomeric HCVc 120 protein in order to obtain insights into its secondary structure as well as the environment surrounding some protein side chains. When compared with its monomer form, oligomeric HCVc 120 protein shows an increase in beta-sheet structure. Tryptophan residues have been found to be solvent exposed in the oligomeric form, and they likely do not significantly participate in the protein assembly. However, the beta-sheet content in oligomeric HCVc 120 protein suggests that this structural motif cannot be excluded in nucleocapsid formation, as shown recently in other viruses.     

Identification of residues in the hepatitis C virus core protein that are critical for capsid assembly in a cell-free system

Significant advances have been made in understanding hepatitis C virus (HCV) replication through development of replicon systems. However, neither replicon systems nor standard cell culture systems support significant assembly of HCV capsids, leaving a large gap in our knowledge of HCV virion formation. Recently, we established a cell-free system in which over 60% of full-length HCV core protein synthesized de novo in cell extracts assembles into HCV capsids by biochemical and morphological criteria. Here we used mutational analysis to identify residues in HCV core that are important for capsid assembly in this highly reproducible cell-free system. We found that basic residues present in two clusters within the N-terminal 68 amino acids of HCV core played a critical role, while the uncharged linker domain between them was not. Furthermore, the aspartate at position 111, the region spanning amino acids 82 to 102, and three serines that are thought to be sites of phosphorylation do not appear to be critical for HCV capsid formation in this system. Mutation of prolines important for targeting of core to lipid droplets also failed to alter HCV capsid assembly in the cell-free system. In addition, wild-type HCV core did not rescue assembly-defective mutants. These data constitute the first systematic and quantitative analysis of the roles of specific residues and domains of HCV core in capsid formation.     

The Replacement of 10 Non-Conserved Residues in the Core Protein of JFH-1 Hepatitis C Virus Improves Its Assembly and Secretion

Hepatitis C virus (HCV) assembly is still poorly understood. It is thought that trafficking of the HCV core protein to the lipid droplet (LD) surface is essential for its multimerization and association with newly synthesized HCV RNA to form the viral nucleocapsid. We carried out a mapping analysis of several complete HCV genomes of all genotypes, and found that the genotype 2 JFH-1 core protein contained 10 residues different from those of other genotypes. The replacement of these 10 residues of the JFH-1 strain sequence with the most conserved residues deduced from sequence alignments greatly increased virus production. Confocal microscopy of the modified JFH-1 strain in cell culture showed that the mutated JFH-1 core protein, C10M, was present mostly at the endoplasmic reticulum (ER) membrane, but not at the surface of the LDs, even though its trafficking to these organelles was possible. The non-structural 5A protein of HCV was also redirected to ER membranes and colocalized with the C10M core protein. Using a Semliki forest virus vector to overproduce core protein, we demonstrated that the C10M core protein was able to form HCV-like particles, unlike the native JFH-1 core protein. Thus, the substitution of a few selected residues in the JFH-1 core protein modified the subcellular distribution and assembly properties of the protein. These findings suggest that the early steps of HCV assembly occur at the ER membrane rather than at the LD surface. The C10M-JFH-1 strain will be a valuable tool for further studies of HCV morphogenesis.     

Hepatitis C virus core protein is a dimeric alpha-helical protein exhibiting membrane protein features

The building block of hepatitis C virus (HCV) nucleocapsid, the core protein, together with viral RNA, is composed of different domains involved in RNA binding and homo-oligomerization. The HCV core protein 1-169 (C(HCV)169) and its N-terminal region from positions 1 to 117 (C(HCV)117) were expressed in Escherichia coli and purified to homogeneity suitable for biochemical and biophysical characterizations. The overall conformation and the oligomeric properties of the resulting proteins C(HCV)169 and C(HCV)117 were investigated by using analytical centrifugation, circular dichroism, intrinsic fluorescence measurements, and limited proteolysis. Altogether, our results show that core protein (C(HCV)169) behaves as a membranous protein and forms heterogeneous soluble micelle-like aggregates of high molecular weight in the absence of detergent. In contrast, it behaves, in the presence of mild detergent, as a soluble, well-folded, noncovalent dimer. Similar to findings observed for core proteins of HCV-related flaviviruses, the HCV core protein is essentially composed of alpha-helices (50%). In contrast, C(HCV)117 is soluble and monodispersed in the absence of detergent but is unfolded. It appears that the folding of the highly basic domain from positions 2 to 117 (2-117 domain) depends on the presence of the 117-169 hydrophobic domain, which contains the structural determinants ensuring the binding of core with cellular membranes. Finally, our findings provide valuable information for further investigations on isolated core protein, as well as for attempts to reconstitute nucleocapsid particles in vitro.     

E6AP ubiquitin ligase mediates ubiquitylation and degradation of hepatitis C virus core protein

Hepatitis C virus (HCV) core protein is a major component of viral nucleocapsid and a multifunctional protein involved in viral pathogenesis and hepatocarcinogenesis. We previously showed that the HCV core protein is degraded through the ubiquitin-proteasome pathway. However, the molecular machinery for core ubiquitylation is unknown. Using tandem affinity purification, we identified the ubiquitin ligase E6AP as an HCV core-binding protein. E6AP was found to bind to the core protein in vitro and in vivo and promote its degradation in hepatic and nonhepatic cells. Knockdown of endogenous E6AP by RNA interference increased the HCV core protein level. In vitro and in vivo ubiquitylation assays showed that E6AP promotes ubiquitylation of the core protein. Exogenous expression of E6AP decreased intracellular core protein levels and supernatant HCV infectivity titers in the HCV JFH1-infected Huh-7 cells. Furthermore, knockdown of endogenous E6AP by RNA interference increased intracellular core protein levels and supernatant HCV infectivity titers in the HCV JFH1-infected cells. Taken together, our results provide evidence that E6AP mediates ubiquitylation and degradation of HCV core protein. We propose that the E6AP-mediated ubiquitin-proteasome pathway may affect the production of HCV particles through controlling the amounts of viral nucleocapsid protein.     

Structural and functional characterization of nonstructural protein 2 for its role in hepatitis C virus assembly

The hepatitis C virus (HCV) is a flavivirus replicating in the cytoplasm of infected cells. The HCV genome is a single-stranded RNA encoding a polyprotein that is cleaved by cellular and viral proteases into 10 different products. While the structural proteins core protein, envelope protein 1 (E1) and E2 build up the virus particle, most nonstructural (NS) proteins are required for RNA replication. One of the least studied proteins is NS2, which is composed of a C-terminal cytosolic protease domain and a highly hydrophobic N-terminal domain. It is assumed that the latter is composed of three trans-membrane segments (TMS) that tightly attach NS2 to intracellular membranes. Taking advantage of a system to study HCV assembly in a hepatoma cell line, in this study we performed a detailed characterization of NS2 with respect to its role for virus particle assembly. In agreement with an earlier report ( Jones, C. T., Murray, C. L., Eastman, D. K., Tassello, J., and Rice, C. M. (2007) J. Virol. 81, 8374-8383 ), we demonstrate that the protease domain, but not its enzymatic activity, is required for infectious virus production. We also show that serine residue 168 in NS2, implicated in the phosphorylation and stability of this protein, is dispensable for virion formation. In addition, we determined the NMR structure of the first TMS of NS2 and show that the N-terminal segment (amino acids 3-11) forms a putative flexible helical element connected to a stable alpha-helix (amino acids 12-21) that includes an absolutely conserved helix side in genotype 1b. By using this structure as well as the amino acid conservation as a guide for a functional study, we determined the contribution of individual amino acid residues in TMS1 for HCV assembly. We identified several residues that are critical for virion formation, most notably a central glycine residue at position 10 of TMS1. Finally, we demonstrate that mutations in NS2 blocking HCV assembly can be rescued by trans-complementation.     

Rous sarcoma virus Gag protein-oligonucleotide interaction suggests a critical role for protein dimer formation in assembly

The structural protein Gag is the only viral product required for retrovirus assembly. Purified Gag proteins or fragments of Gag are able in vitro to spontaneously form particles resembling immature virions, but this process requires nucleic acid, as well as the nucleocapsid domain of Gag. To examine the role of nucleic acid in the assembly in vitro, we used a purified, slightly truncated version of the Rous sarcoma virus Gag protein, Delta MBD Delta PR, and DNA oligonucleotides composed of the simple repeating sequence GT. Apparent binding constants were determined for oligonucleotides of different lengths, and from these values the binding site size of the protein on the DNA was calculated. The ability of the oligonucleotides to promote assembly in vitro was assessed with a quantitative assay based on electron microscopy. We found that excess zinc or magnesium ion inhibited the formation of virus-like particles without interfering with protein-DNA binding, implying that interaction with nucleic acid is necessary but not sufficient for assembly in vitro. The binding site size of the Delta MBD Delta PR protein, purified in the presence of EDTA to remove zinc ions at the two cysteine-histidine motifs, was estimated to be 11 nucleotides (nt). This value decreased to 8 nt when the protein was purified in the presence of low concentrations of zinc ions. The minimum length of DNA oligonucleotide that promoted efficient assembly in vitro was 22 nt for the zinc-free form of the protein and 16 nt for the zinc-bound form. To account for this striking 1:2 ratio between binding site size and oligonucleotide length requirement, we propose a model in which the role of nucleic acid in assembly is to promote formation of a species of Gag dimer, which itself is a critical intermediate in the polymerizaton of Gag to form the protein shell of the immature virion.     

Identification of basic amino acids at the N-terminal end of the core protein that are crucial for hepatitis C virus infectivity

A major function of the hepatitis C virus (HCV) core protein is the interaction with genomic RNA to form the nucleocapsid, an essential component of the virus particle. Analyses to identify basic amino acid residues of HCV core protein, important for capsid assembly, were initially performed with a cell-free system, which did not indicate the importance of these residues for HCV infectivity. The development of a cell culture system for HCV (HCVcc) allows a more precise analysis of these core protein amino acids during the HCV life cycle. In the present study, we used a mutational analysis in the context of the HCVcc system to determine the role of the basic amino acid residues of the core protein in HCV infectivity. We focused our analysis on basic residues located in two clusters (cluster 1, amino acids [aa]6 to 23; cluster 2, aa 39 to 62) within the N-terminal 62 amino acids of the HCV core protein. Our data indicate that basic residues of the first cluster have little impact on replication and are dispensable for infectivity. Furthermore, only four basic amino acids residues of the second cluster (R50, K51, R59, and R62) were essential for the production of infectious viral particles. Mutation of these residues did not interfere with core protein subcellular localization, core protein-RNA interaction, or core protein oligomerization. Moreover, these mutations had no effect on core protein envelopment by intracellular membranes. Together, these data indicate that R50, K51, R59, and R62 residues play a major role in the formation of infectious viral particles at a post-nucleocapsid assembly step.     

The conserved N-terminal region of Sendai virus nucleocapsid protein NP is required for nucleocapsid assembly.

Sendai virus nucleocapsid protein NP synthesized in the absence of other viral components assembled into nucleocapsid-like particles. They were identical in density and morphology to authentic nucleocapsids but were smaller in size. The reduction in size was probably due to the fact that they contained RNA only 0.5 to 2 kb in length. Nucleocapsid assembly requires NP-NP and NP-RNA interactions. To identify domains on NP protein involved in nucleocapsid formation, 29 NP protein mutants were tested for the ability to assemble. Any deletion between amino acid residues 1 and 399 abolished formation of nucleocapsid-like particles, but mutants within this region exhibited two different phenotypes. Deletions between positions 83 and 384 completely abolished all interactions. Deletions between residues 1 and 82 and between residues 385 and 399, at the N- and C-terminal ends of the region from 1 to 399, resulted in unstructured aggregates of NP protein, indicating only a partial loss of function. Deletions within the C-terminal 124 amino acids were the only ones that did not affect assembly. The results suggest that NP protein can be divided into at least two separate domains which function independently of each other. Domain I (residues 1 to 399) seems to contain all of the structural information necessary for assembly, while domain II (residues 400 to 524) is not involved in nucleocapsid formation.     

In vitro self-assembled HCV core virus-like particles induce a strong antibody immune response in sheep.

The in vitro self-assembly properties of the entire hepatitis C virus core protein (HCcAg) obtained from Pichia pastoris cells and the induction of specific antibody immune response were studied. HCcAg was purified as a low-molecular-weight species by electroelution under denaturing conditions for confirmation of its self-assembly properties. After renaturalization, electron microscopy showed that HCcAg assembled into spherical particles of 30 nm. HCcAg also showed homogeneity and was specifically recognized by serum from a chronic HCV carrier patient. The data indicated that in vitro assembly of HCcAg, into virus-like particles resembling HCV nucleocapsid particles at a mature stage, is an intrinsic quality of this protein. Finally, HCcAg generated a strong antibody immune response in sheep, suggesting its usefulness for stimulating the host immune response against HCV.     

Ultrastructural and immunocytochemical evidences of core-particle formation in the methylotrophic Pichia pastoris yeast when expressing HCV structural proteins (core-E1).

Particulate antigens of the Hepatitis C virus (HCV) are reported for the first time by transmission electron microscopy in Pichia pastoris. The yeast was cloned to express the first 339 NH2-terminal amino acids of the HCV polyprotein (C-E1.339 polypeptide). The C-E1.339 polypeptide covers the putative 191 aa of the core protein (aa 1-191) and 148 aa of the E1 envelope antigen (aa 192-339). Virus-like particles (VLP) with diameters ranging from 20 nm to 30 nm were specifically observed in those cells expressing the HCV polyprotein. The VLP appeared along the membrane of the endoplasmic reticulum, but were fundamentally localized in vacuoles, either free or inside autophagic bodies. Clustered particles, chains of particles, high-density reticular structures, and crystalloid bodies were also detected, the last one being an orderly arrangement of particles with 20 nm diameters. The crystal-associated particles are well differentiated from the intracellular VLP because of their uniform size and shape. We argue that membrane components are retained in the architecture of the VLP, conferring to this particle certain heterogeneity. Both kinds of particles, the VLP formed after treatment with NP-40 and the crystal-associated particles, were core protein-positives. Whether they reflect mature HCV nucleocapsid or intermediary states in the viral nucleocapsid morphogenesis remains unknown. We conclude that, like mammalian cell lines, the P. pastoris yeast could be an appropriate host for the analysis of HCV polyprotein processing and, eventually, virus assembly.     

Mechanism of in vitro collagen fibril assembly. Kinetic and morphological studies

The kinetics of in vitro fibril assembly of Type I collagen preparations that contain different amounts of covalently cross-linked oligomers was studied with turbidimetry. Fibril formation showed a lag phase with no solution turbidity and a growth phase with a sigmoidal increase in the solution turbidity. The length of the lag phase was inversely related to both the total collagen concentration and the amount of covalently cross-linked oligomers in the solution. Double logarithmic plots of t1/4, the amount of time it takes for 1/4 of the collagen to assemble into fibrils, versus the total collagen concentration were linear but the slope decreased from -0.84 to -2.3 with decreasing amounts of covalently cross-linked oligomers in the samples. Electron microscopy showed the formation of unbanded microfibrils with diameters in the range of 3-15 nm early in the lag phase and larger diameter banded fibrils coexisting with the microfibrils near the end of the lag phase. Centrifugation of the solution at the lag phase prolonged the lag time, presumably by removal of microfibrils, but subsequent growth of the fibrils was unaffected. The results suggest a cooperative nucleation-growth mechanism for the in vitro assembly of collagen fibrils which is consistent with the results of an equilibrium study of the fibril assembly reaction we reported earlier (Na, G. C., Butz, L. J., Bailey, D. G., and Carroll, R. J. (1986) Biochemistry 25, 958-966).     

Conformational Plasticity of Hepatitis C Virus Core Protein Enables RNA-Induced Formation of Nucleocapsid-like Particles

Many of the unanswered questions associated with hepatitis C virus assembly are related to the core protein (HCVcp), which forms an oligomeric nucleocapsid encompassing the viral genome. The structural properties of HCVcp have been difficult to quantify, at least in part because it is an intrinsically disordered protein. We have used single-molecule Förster Resonance Energy Transfer techniques to study the conformational dimensions and dynamics of the HCVcp nucleocapsid domain (HCVncd) at various stages during the RNA-induced formation of nucleocapsid-like particles. Our results indicate that HCVncd is a typical intrinsically disordered protein. When it forms small ribonucleoprotein complexes with various RNA hairpins from the 3′ end of the HCV genome, it compacts but remains intrinsically disordered and conformationally dynamic. Above a critical RNA concentration, these ribonucleoprotein complexes rapidly and cooperatively assemble into large nucleocapsid-like particles, wherein the individual HCVncd subunits become substantially more extended.     

Mutations in the endodomain of Sindbis virus glycoprotein E2 define sequences critical for virus assembly

Envelopment of Sindbis virus at the plasma membrane is a multistep process in which an initial step is the association of the E2 protein via a cytoplasmic endodomain with the preassembled nucleocapsid. Sindbis virus is vectored in nature by blood-sucking insects and grows efficiently in a number of avian and mammalian vertebrate hosts. The assembly of Sindbis virus, therefore, must occur in two very different host cell environments. Mammalian cells contain cholesterol which insect membranes lack. This difference in membrane composition may be critical in determining what requirements are placed on the E2 tail for virus assembly. To examine the interaction between the E2 tail and the nucleocapsid in Sindbis virus, we have produced substitutions and deletions in a region of the E2 tail (E2 amino acids 408 to 415) that is initially integrated into the endoplasmic reticulum. This sequence was identified as being critical for nucleocapsid binding in an in vitro peptide protection assay. The effects of these mutations on virus assembly and function were determined in both vertebrate and invertebrate cells. Amino acid substitutions (at positions E2: 408, 410, 411, and 413) reduced infectious virus production in a position-dependent fashion but were not efficient in disrupting assembly in mammalian cells. Deletions in the E2 endodomain (delta406-407, delta409-411, and delta414-417) resulted in the failure to assemble virions in mammalian cells. Electron microscopy of BHK cells transfected with these mutants revealed assembly of nucleocapsids that failed to attach to membranes. However, introduction of these deletion mutants into insect cells resulted in the assembly of virus-like particles but no assayable infectivity. These data help define protein interactions critical for virus assembly and suggest a fundamental difference between Sindbis virus assembly in mammalian and insect cells.