Structure of the zinc-binding domain of an essential component of the hepatitis C virus replicase. Nature 2005; 435: 374–379. (Reprinted by permission of Nature; http://www.nature.com.), , .
Hepatitis C virus (HCV) is a human pathogen affecting nearly 3% of the world's population. Chronic infections can lead to cirrhosis and liver cancer. The RNA replication machine of HCV is a multi-subunit membrane-associated complex. The nonstructural protein NS5A is an active component of HCV replicase, as well as a pivotal regulator of replication and a modulator of cellular processes ranging from innate immunity to dysregulated cell growth. NS5A is a large phosphoprotein (56-58 kd) with an amphipathic α-helix at its amino terminus that promotes membrane association. After this helix region, NS5A is organized into 3 domains. The N-terminal domain (domain I) coordinates a single zinc atom per protein molecule. Mutations disrupting either the membrane anchor or zinc binding of NS5A are lethal for RNA replication. However, probing the role of NS5A in replication has been hampered by a lack of structural information about this multifunctional protein. Here we report the structure of NS5A domain I at 2.5-Å resolution, which contains a novel fold, a new zinc-coordination motif, and a disulfide bond. We use molecular surface analysis to suggest the location of protein-, RNA-, and membrane-interaction sites.
The structures of the hepatitis C virus (HCV) nonstructural protein 3-4A (NS3-4A) serine protease/RNA helicase and of the NS5B RNA-dependent RNA polymerase were resolved more than 5 years ago.1 The function and biochemical properties of these rather classical viral enzymes were known before their structures became available. By contrast, structural information is sorely needed to elucidate the functions of the more difficult HCV proteins, such as the p7 polypeptide, the NS2-3 autoprotease as well as the NS4B and NS5A proteins. In this context, the structure of the N-terminal domain of NS5A reported by Tellinghuisen, Marcotrigiano, and Rice in the May 19, 2005 issue of Nature2 represents the result of an arduous effort and a major breakthrough in the field.
HCV NS5A has attracted considerable interest because of its potential role in modulating the interferon response.3 In addition, numerous other functions and a plethora of interaction partners have been postulated for NS5A.4, 5 However, only few of these potential interactions have been validated in a meaningful experimental context involving active HCV RNA replication,6, 7 and surprisingly little effort has been devoted to the basic biochemical characterization of this protein.
HCV NS5A is a 447-amino-acid phosphoprotein (Fig. 1). It is found in a basally phosphorylated form of 56 kd and in a hyperphosphorylated form of 58 kd. The centrally located serine residues Ser 225, Ser 229, and Ser 232 (corresponding to Ser 2197, Ser 2201, and Ser 2204 of the HCV polyprotein) are important for NS5A hyperphosphorylation (Fig. 1). However, whether these residues are actually phosphorylated or affect the NS5A phosphorylation state indirectly is unknown, because only few phosphoacceptor sites have been mapped experimentally8, 9 (Fig. 1). A number of kinases capable of phosphorylating NS5A have been identified. However, which cellular kinase generates the different phosphoforms of NS5A during the viral life cycle is unknown.
Interestingly, adaptive mutations have been found to cluster in the central region of NS5A in the context of selectable HCV replicons (Fig. 1), suggesting that NS5A is involved, either directly or by interaction with cellular proteins and pathways, in the viral replication process. Adaptive mutations often affect serine residues required for hyperphosphorylation, suggesting that the phosphorylation state of NS5A modulates the efficiency of HCV RNA replication.6, 10, 11 The finding of adaptive mutations in NS5A, evidence for physical and functional interactions of NS5A with other HCV nonstructural proteins,12, 13 and the results of site-directed mutagenesis14, 15 point toward an essential role of NS5A in HCV RNA replication. However, its function remains elusive.
A characteristic feature of HCV is that the nonstructural proteins form a membrane-associated replication complex together with replicating viral RNA, altered membranes, and additional as yet largely unidentified host cell components.16 In this context, an N-terminal amphipathic α-helix mediates membrane association of NS5A.14, 15, 17 This helix exhibits a hydrophobic, tryptophan-rich side embedded in the cytosolic leaflet of the membrane bilayer, whereas the polar, charged side is exposed to the cytosol. Thus, NS5A is a monotopic protein with an in-plane amphipathic α-helix as membrane anchor. Recent structure–function analyses showed that this helix displays fully conserved polar residues at the membrane surface, which define a platform probably involved in specific protein–protein interactions essential for the formation of a functional HCV replication complex.15 The identification and structural characterization of the in-plane membrane anchor of NS5A facilitated the rational design and expression of a recombinant soluble form of NS5A. Limited proteolysis of recombinant NS5A has recently allowed specification of a domain organization of this protein predicted by comparative sequence analyses18 (Fig. 1).
In the paper under discussion, Tellinghuisen et al. report the x-ray structure at 2.5 Å resolution of the relatively highly conserved domain I of NS5A, which immediately follows the membrane-anchoring amphipathic α-helix. Based on the 3-dimensional structure, they divided domain I into an N-terminal subdomain IA and a C-terminal subdomain IB (Fig. 2A). Subdomain IA consists of an N-terminal extended loop lying adjacent to a 3-stranded anti-parallel β-sheet (B1-B3) with an α-helix, designated as H2 to allow numbering of the N-terminal membrane-anchoring amphipathic α-helix as H1, at the C-terminus of the third β-strand. These elements constitute the structural scaffold for a 4-cysteine zinc coordination site previously predicted based on biochemical studies by the same authors.18 This site is composed of the 4 fully conserved cysteine residues Cys 39, Cys 57, Cys 59, and Cys 80 that have previously been shown to be absolutely required for HCV RNA replication18 (Fig. 2B). Based on the location of the zinc coordination site and their previous biochemical data, the authors suggest that this essential zinc has a structural rather than a functional role in NS5A. Subdomain IB consists of 2 anti-parallel β-sheets (B4-B7 and B8-B9) surrounded by extensive random coil structures. No similar structures have thus far been reported, indicating that this protein represents a novel fold. A surprising finding was the presence of a disulfide bond near the C-terminus of subdomain IB, connecting the conserved cysteine residues Cys 142 and Cys 190 (Fig. 2A). Previous site-directed mutagenesis indicated that the disulfide bond formed between Cys 142 and Cys 190 is not essential for HCV RNA replication.18 Thus, further studies will have to show whether this disulfide bond is formed in cells during HCV RNA replication and may contribute to a function of NS5A.
Interestingly, the structure reveals a dimer created by contacts near the N-terminal ends of the molecules (Fig. 2C–D). High amino acid conservation of the molecular surface involved in dimer formation strongly suggests that oligomerization of NS5A actually occurs in an infected cell. Because subdomain IA has an almost exclusively basic surface whereas subdomain IB is predominantly acidic, the NS5A dimer exhibits a strikingly asymmetric charge distribution (Fig. 2C). Based on analyses of the electrostatic surface potential, the authors proposed a model of how this dimer may interact with proteins, membranes, and RNA within the viral replication complex. Because only 5 amino acid residues are missing between the structure of the H1 membrane-anchoring amphipathic α-helix resolved by nuclear magnetic resonance (NMR) spectroscopy and the domain I structure resolved by x-ray crystallography, the N-terminus of domain I is likely located close to the membrane. Consistent with this notion is the basic nature of the surface of domain I close to H1, favoring contacts with the negatively charged phospholipid head groups on the membrane surface. This interaction would position the groove generated by the “claw-like” dimer facing away from the membrane, where it could interact with RNA (Fig. 2D). Indeed, the dimensions of this groove are sufficient to accommodate either single- or double-stranded RNA. In addition, the deep, highly basic portion of the groove could favor contacts with RNA, whereas the “arms” extending out past this groove are more acidic, and might serve to prevent RNA from exiting the groove. Finally, a conserved surface would be placed outside of the “arms” of the dimer, where it would be available for protein–protein interactions.
Taken together, Tellinghuisen et al. provide the first molecular view of HCV NS5A and report a number of exciting findings, including a novel protein fold, a new zinc coordination motif, a dimerization interface, and a putative protein–protein interaction surface. Furthermore, analyses of the molecular surface together with the recently reported NMR structure of the N-terminal membrane anchor domain of NS5A allowed the authors to tentatively position the molecule with respect to the membrane and to propose a potential RNA-binding groove exposed toward the cytosol. Each of these findings has immediate implications for our understanding of the functional architecture of the HCV replication complex and represent a long-awaited foundation for further studies on the role of NS5A in the viral life cycle.
What then is the function of NS5A? A number of exciting possibilities can now be tested. For example, multiple NS5A dimers may form a 2-dimensional array on intracellular membranes, thereby creating a long basic cleft that could bind either double- or single-stranded viral RNA through electrostatic interactions. Such a “basic railway” would allow the free sliding of RNA, whereas the acidic domains would protect it from degradation by cellular RNAses or from recognition by double-stranded RNA–induced antiviral defense mechanisms. One could imagine that the NS3 RNA helicase interacts with NS5A at one end of the “basic railway” to dissociate the 2 RNA strands and direct the positive strand to the ribosome for polyprotein translation or to core protein multimers for encapsidation while the negative strand would serve the RNA-dependent RNA polymerase as template for the production of progeny positive-strand RNA. At the opposite end of the “basic railway,” positive-strand RNA would become accessible as a template for the synthesis of new negative-strand RNA by NS5B. According to this model, NS5A would tether the viral RNA onto intracellular membranes and coordinate its different fates during HCV replication. Given that numerous core protein molecules are needed to form a nucleocapsid for incorporation of 1 genomic RNA, a large excess of nonstructural proteins are produced in an infected cell. Thus, one can easily conceive that these proteins have additional functions as arrays or lattices on membranes. In fact, it has been shown that only a small proportion of HCV nonstructural proteins expressed in cells harboring HCV replicons are actively engaged in RNA replication,19, 20 and viral RNA is protected not only by membranes but also by a protease-sensitive component.21
In conclusion, the structure reported by Tellinghuisen et al. provides a wealth of novel information on NS5A and a rational framework for further studies addressing the function of this enigmatic HCV protein. Clearly, the next steps will be a structure of the entire NS5A protein and a picture of how this fits into the HCV replication complex. Given the current pace of HCV research, answers to these questions are likely to come soon.