NS5B

RNA Dependent RNA Polymerase from Hepatitis C Virus

In 2019, there were approximately 290,000 deaths due to a Hepatitis C infection. The majority of these deaths were largely due to the long-term adverse effects such as cirrhosis and liver cancer. Currently in the world, about 58 million people are chronically infected with the Hepatitis C virus and there are about 1.5 million new infections annually. In the United States, data estimates that 2.4 million people are infected with Hepatitis C. Due to the low percentage of diagnoses and the frequency of acute infection, it is believed that the number of Hepatitis C infections can range anywhere from 2.5 million to 4.7 million infected individuals. Nearly all of RNA viruses depend on specific activities performed by RNA dependent RNA polymerases (RdRp). These activities importantly include the copying of the RNA sequence and facilitating the process of transcription. Therefore, if a vaccine was created to target the RNA dependent RNA polymerase and inhibit the function of this protein, then the viral genome could not be replicated. If replication halts, the infection cannot spread to other cells. Due to the viral dependence on RdRps, it is essential to understand the structure for further antiviral drug research. Sofosbuvir is an inhibitor of NS5B used for Hepatitis C treatment.

NS5B is the RNA dependent RNA polymerase of Hepatitis C virus. NS5B, like other RNA dependent RNA polymerases, is error prone. This viral RNA replicase is of approximately a million times lower fidelity than a replicative prokayrotic or eukaryotic DNA polymerase. This is due in part to the fact that NS5B contains no exonuclease or proofreading domain. In NS5B, two divalent cations coordinated by carboxyl groups (as seen in DNA polymerases) catalyze the polymerization of monomers of RNA triphosphates to extend a primer strand, that may have initiated de novo. In the case of NS5B the residues that coordinate divalent cations (Mg2+ or Mn2+ in vitro) are the three active site aspartates (220, 318 and 319) seen here (PDB entry 2hai).

Though Hepatitis C virus is of the Flaviviridae family the structure of NS5B is similar to the polymerase of bacteriophage Φ 6. The similarity to the bacteriophage polymerase is due to NS5B containing a fully encircled active site. Like many template-dependent nucleotide polymerases, NS5B can be visualized similar to a right hand. NS5B contains several domains, fingers in blue, palm in magenta, thumb in green and a c-terminal domain in yellow. Specific to Hepatitis C virus family, the NS5B protein has a 21-24 residue hydrophobic membrane anchor at its C-termini that allows involvement with the replication complex that is located in membrane bound vesicles derived from the endoplasmic reticulum. The fingers, palm, and thumb region of the NS5B contain 531 residues and are composed of 21 alpha helices and 18 beta barrels. The fingers region of NS5B can be divided into two distinct sections: the palm-proximal region and the distal region. The palm-proximal region is folded as a bundle of alpha helices and the distal region is composed of a singular beta barrel. These regions are termed the fingertips. The fingers domain interacts with the major groove of the template RNA and aids in recognition and binding. The fingers domain of NS5B are composed of a set of 4 beta strands that can establish contact with the thumb region closing the active site. The thumb region of HCV is structurally larger in comparison to other RdRps. It is composed mainly of alpha helices, 7 of them, and two beta hairpins. It contains amino acids 370-531. One is a beta-hairpin in thumb domain that is proposed to move upon formation of exiting double-stranded RNA. The core of the thumb is composed of two repeating helix motifs. The thumb region of NS5B also contains a “priming loop” that extends from the thumb and provides stability during initiation. The palm domain contains the active site aspartates and there are several contacts between the fingers and thumbs domain that give the active site an encircled structure.

DNA was modeled into the NS5B model (PDB entries 1rtd and 2hai, respectively) by aligning of palm domain of NS5B and the palm domain of HIV reverse transcriptase, which was co-crystallized in complex with DNA and an incoming dTTP. Then removing the protein portion HIV RT model while leaving the DNA where it fell into the proposed NS5B binding cleft. Looking closely at the active site the catalytic Mg2+ ions are modeled in green, these would be coordinated by the three aspartic acid carboxylates, (D220, D318 and D319). A beta-hairpin (residues 440-455) in the thumb domain has been shifted to accommodate DNA, the hairpin is modeled into the minor groove, a possible binding site, particularly in the larger minor goove of dsRNA. There are noticeable steric clashes between the modeled DNA and the random coil at the end of the c-terminal domain. This domain is a linker that attaches to the membrane anchor of NS5B.

The template strand is seen entering through a gap in the fingers domain. An incoming dTTP that would extend the primer strand lines up with the NS5B active site and duplex DNA exits the enzyme through the large central hole in the closed active site formed by the unusual contacts between the fingers and thumb domains.

Empirically determined sites of protein-ssRNA interactions. The highlighted peptide segments were each identified without x-ray crystallography or NMR. These RNA binding peptides were identified by cross linking single stranded RNA to NS5B followed by a tryptic digest of the protein, then purification of the RNA bound peptide segments by affinity (for the RNA) chromatography. The segments of peptide that stuck to the column meaning they had been cross linked to RNA were then analyzed with MALDI mass spectrometry. It is interesting that all of the contacts were in the fingers domain. This could be in part due to the fact that single stranded RNA was cross linked to the enzyme, the fingers domain is thought to bind templating ssRNA while other regions of the polymerase would bind duplex RNA.

This image is a depiction of each of the protein products of NS5B genomic translation. The proteins coded for by the hepatitis C virus (HCV) genome all associate with the ER membrane. The proteins are translated as one large poly-protein that is enzymatically cleaved by both host and viral proteases. Of the several proteins that are coded for by HCV, three functions have been identified as relevant drug targets, these are the NS3 protease and helicase domains and the NS5B RNA dependent RNA polymerase. Currently the therapy for HCV is interferon therapy often in combination with ribavirin. This therapy however is inadequate; it is not effective in each genotype of HCV, it is not well tolerated, and is expensive. For these reasons many academic and industrial laboratories have been working on developing novel inhibitors of NS5B.

Much of the x-ray crystallography work that has gone into NS5B to date (Ligands are from: 1C2P, 1CSJ, 1GX5, 1GX6, 1NB4, 1NB6, 1NB7, 1HNU, 1NHV, 1OS5, 1QUV, 1YUY, 1YV2, 1YVF, 1YVX, 1YVZ, 1Z4U, 2AWZ, 2AX0, 2AX1, 2BRK, 2BRL, 2D3U, 2D3Z, 2D41, 2GIQ, 2GIR, 2HWH, 2HWI, 2I1R, 2IJN, 2JC0, 2JC1, 3BR9, 3BSA, 3BSC, 3CIZ, 3CJ0, 3CJ2, 3CJ3, 3CJ4, 3CJ5). Superpositions in the scene by Rould MA, and Villanueva NL. Each of the forty three structures included in the superposition contain at least one ligand, the ligands range from ions and small molecules to nucleotides and non-nucleoside analogue inhibitors. Positioning the structure in the familiar orientation with the domains colored as above, it is obvious that there are two primary areas where ligands are clustering. Several nucleotides, oligonucleotides and non-nucleoside analogue inhibitors can be found near the active site. There is another site that is about 30-35Å from the active site, an allosteric site, where again ligands are clustering. There is a rGTP binding site here that seems to activate the enzyme and a nearby site where inhibitors bind and disrupt the activity of the enzyme.

The inhibitors that bind near the active site are thought to work by disrupting the primer grip site causing an inability of the enzyme to efficiently hold and extend a growing primer strand. The inhibitors that bind near the allosteric site work through a poorly understood mechanism, several explanations for their ability to inhibit the enzyme have been proposed. The commonly proposed explanations are that these inhibitors lock the enzyme in an inactive conformation, inhibit binding of the rGTP that binds in the area and has been shown to stimulate activity, or that this interface is critical in protein-protein interactions and disruption of these interactions by the inhibitors disrupts an oligomerization of several NS5B protomers thus making each less efficient.

3D structures of RNA polymerase

RNA polymerase

Additional Resources

For additional information, see: Viral Infections