Structural basis of paramyxo- and pneumovirus polymerase inhibition by non-nucleoside small-molecule antivirals

Abstract

Small-molecule antivirals can be used as chemical probes to stabilize transitory conformational stages of viral target proteins, facilitating structural analyses. Here, we evaluate allosteric pneumo- and paramyxovirus polymerase inhibitors that have the potential to serve as chemical probes and aid the structural characterization of short-lived intermediate conformations of the polymerase complex. Of multiple inhibitor classes evaluated, we discuss in-depth distinct scaffolds that were selected based on well-understood structure-activity relationships, insight into resistance profiles, biochemical characterization of the mechanism of action, and photoaffinity-based target mapping. Each class is thought to block structural rearrangements of polymerase domains albeit target sites and docking poses are distinct. This review highlights validated druggable targets in the paramyxo- and pneumovirus polymerase proteins and discusses discrete structural stages of the polymerase complexes required for bioactivity.

Keywords: chemical probe; measles virus; mononegavirales; parainfluenza virus; paramyxovirus; pneumovirus; polymerase inhibitor; polymerase structure; respiratory syncytial virus; viral polymerase.

Fig 1

Architecture of the MeV L protein. (A) 2D schematic of the organization of the five conserved MeV L protein domains. RNA-dependent RNA polymerase (RdRP) domain in light blue (residues 1–912), polyribonucleotidyltransferase (PRNTase) domain in forest green (residues 913–1393), connector domain in yellow (residues 1429–1693), methyltransferase (MTase) domain in orange (residues 1722–2028), and the C-terminal domain (CTD) in red (residues 2029–2183). (B) Homology model of MeV L protein (Edmonston strain, GenBank: QOT47606.1) in its elongation-3 stage that is based on the cryoEM structure of parainfluenza virus 5 (PDB: 6v86), generated using the Swiss model server. Domains are colored as in the 2D schematic. (C) Conserved regions of the RdRP domain: CR I in deep blue (residues 217–409), CR II in pale yellow (residues 495–596), and CR III in deep teal (residues 653–869) with the GDN shown as spheres. (D) Conserved regions of the PRNTase domain: CR IV in pale yellow (residues 928–1093) and CR V in sky blue (residues 1129–1376). (E) Conserved region of the MTase domain: VI in marine blue (residues 1753–1832). Spatial orientation of CRs in C–E is conserved from the homology model on the left in B. (F) MeV L central cavity showing the GDN (residues 772–775) in gray as spheres, the putative priming loop in black (residues 1200–1233), and the putative intrusion loop in magenta (residues 1279–1302).

Fig 2

Channels and loop re-arrangements of the mononegavirus polymerase. (A) NTP entry, template entry, template exit, and RNA product exit channels extending from the central cavity (black dashed circle) of a MeV L homology model (PDB: 6v86). (B) RSV L with promoter bound (PDB: 8snx) representing a pre-initiation state for de novo RNA synthesis. RNA (orange). (C) In a pre-initiation state, the priming and intrusion loops are extended away from the active site allowing space for the incoming RNA template. (D) Upon transitioning into an initiation state, the priming loop is positioned within proximity of the GDN and the intrusion loop is extended into the PRNTase domain. (E) Entering an elongation-1 stage, the priming loop moves away from the active site and into the PRNTase domain while the intrusion loop moves slightly closer to the GDN. (F) As the polymerase enters an elongation-2 stage, the priming loop extends deeper into the PRNTase domain and the intrusion loop continues to move closer to the GDN. (G) In the elongation-3 stage, the priming loop is retracted even further into the PRNTase domain, and the intrusion loop begins to move away from the GDN toward the PRNTase domain. (C–G) Different MeV L homology models that were generated based on the coordinates of RSV (C; PDB: 8snx), VSV (D; PDB: 5a22), HMPV (E; PDB: 6u5o), NDV (F; PDB: 7you), and PIV5 (G; PDB: 6v86), respectively, and aligned to show loop arrangements within the central cavity in relation to the GDN. To improve clarity, only residues 1217–1233 of the priming loop are shown. RdRP (light blue), PRNTase (forest green), connector (yellow), MTase (orange), CTD (red), GDN (gray spheres), priming loop (black), intrusion loop (magenta), HR motif (yellow sticks). Arrows show loop movement.

Fig 3

Overview of viral mRNA capping and methylation. (A) The N^ε2 of the highly conserved histidine of the HR motif nucleophilically attacks the α-phosphate of 5′-pppN-RNA and forms a covalent interaction. (B) Within the PRNTase active site GDP, generated from GTP through the guanosine, 5′-triphosphatase activity of L protein is transferred to pN-RNA when the β-phosphate of GDP nucleophilically attacks the α-phosphate of the pN-RNA that is covalently linked to histidine. (C) With GpppN-RNA now in the S-adenosyl-l-methionine (SAM)-dependent MTase core, the 5′-cap is methylated at the adenosine-2′-O and guanine-N⁷ positions, (D) yielding a bioactive mRNA product, m⁷GpppNm-RNA. RdRP (light blue), PRNTase (forest green), MTase (orange), GDN (yellow star), intrusion loop (black), nascent RNA (brown), GDP (dark blue), SAM and methyl groups (red).

Fig 4

GHP-88309, ERDRP-0519, JNJ-8003, and AVG-233 compound structures. Structures of (A) GHP-88309, (B) JNJ-8003, (C) AVG-233, and (D) ERDRP-0519.

Fig 5

Spatial arrangement of resistance mutations. (A) Resistance sites are tightly clustered around GHP-88309. (B) Location of ERDRP-0519 resistance sites. (C) The majority of JNJ-8003 resistance mutations cluster around the binding site. (D) AVG-233 resistance mutations are observed within proximity of the compound. For GHP-88309, ERDRP-0519, and AVG-233, the original docking pose of each compound is shown. The cryo-EM structure (PDB: 8fu3) is shown for JNJ-8003 and RSV L-P (residues 1–1461), residues 1462–2165 have been modeled. Resistance mutation (red sphere), compound (marine blue), RdRP (light blue), PRNTase (forest green), connector (yellow), MTase (orange), CTD (red), GDN (gray spheres), priming loop (black), intrusion loop (magenta), HR motif (yellow sticks).

Fig 6

In silico docking of GHP-88309 in MeV L. (A) Original top-scoring docking pose of GHP-88309 in a MeV L homology model based on PIV5 P-L (PDB: 6v86). Resistance mutation (red spheres), predicted interaction residue (yellow spheres), resistance residue that is predicted to interact (cyan). (B) Top-scoring docking pose of GHP-88309 in a MeV L homology model based on VSV P-L (PDB: 5a22), representing the polymerase in an initiation stage which is the predicted, targeted conformation. Predicted interaction residue (red spheres), priming loop (black), intrusion loop (magenta), HR motif (yellow sticks). (C) 2-D ligand map of new predicted interactions. (A and B) Compound (marine blue), GDN (gray spheres), RdRP (light blue), PRNTase (forest green), connector (yellow). Docking poses and ligand maps were generated using the molecular operating environment (MOE) software package.

Fig 7

Docking of ERDRP-0519 into MeV L. (A) Original top-scoring docking pose of ERDRP-0519 in a MeV L homology model based on PIV5 P-L (PDB: 6v86). Resistance mutation (red spheres), predicted interaction residue (yellow spheres). (B) ERDRP-0519 in a top-scoring pose in a MeV L homology model based on the RSV P-L-RNA structure (PDB: 8snx) to represent the hypothesized targeted pre-initiation conformation. Predicted interaction residue (red spheres), priming loop (black), intrusion loop (magenta), HR motif (yellow sticks). (C) New predicted interactions are shown in a 2-D ligand map. (A and B) Compound (marine blue), GDN (gray spheres), RdRP (light blue), PRNTase (forest green). Docking poses and ligand maps were generated using the molecular operating environment (MOE) software package.

Fig 8

JNJ-8003 bound to RSV L and AVG-233 docked into RSV L. (A) The cryo-EM structure of RSV L-P with JNJ-8003 (PDB: 8fu3). Resistance mutation (red spheres), predicted interaction residue (yellow spheres), resistance residue that is predicted to interact (cyan). (B) Original top-scoring docking pose of AVG-233 in RSV L. The RdRP and PRNTase domains are from the apo structure of RSV L-P (PDB: 6pzk), and the connector, MTase, and CTD were modeled based on VSV L-P (PDB: 5a22) and RABV L-P (PDB: 6ueb). Resistance mutation (red spheres) predicted interaction residue (yellow spheres). (C) The top-scoring docking pose of AVG-233 in an RSV L homology model based on the VSV P-L coordinates (PDB: 5a22), representing the hypothesized targeted initiation confirmation. Predicted interaction residue (red spheres), priming loop (black), intrusion loop (magenta), HR motif (yellow sticks). (D) 2-D ligand interaction map of new predicted interactions. (A–C) Compound (marine blue), GDN (gray spheres), RdRP (light blue), PRNTase (forest green), connector (yellow). Docking poses and ligand maps were generated using the molecular operating environment (MOE) software package.

Fig 9

GHP-88309, ERDRP-0519, JNJ-8003, and AVG-233 sequence identity analysis of resistance sites. Resistance profile alignments of (A) GHP-88309, (B) ERDRP-0519, (C) JNJ-8003, and (D) AVG-233. High sequence identity (highlighted red and bold). Partial sequence identity (highlighted yellow and bold). Sequences alignments: MeV (NP_056924), HPIV3 (AXU38775), NiV (ACT32616), HeV (APT69531), HPIV5 (YP138518), RSV (YP_009518860), HMPV (AAQ67700), RABV (ABN11300), VSV (UTK57347), EBOV (AHX24663), Marburg virus; MARV (ABA87130).

Fig 10

GHP-88309, ERDRP-0519, JNJ-8003, and AVG-233 sequence identity analysis of predicted interacting residues. (A) MeV L residues originally predicted to interact with GHP-88309 and (B) new predicted interactions based on docking into MeV L in an initiation conformation. (C) Residues of MeV L originally predicted to interact with ERDRP-0519 and (D) new predicted interactions based on docking into MeV L in a pre-initiation conformation. (E) RSV L residues that were identified to interact with JNJ-8003 in the cryo-EM structure. (F) Original predicted RSV L residue interactions with AVG-233 and (G) new predicted interactions based on docking into RSV L in an initiation conformation. Residues that are both a site of resistance and prediction interaction or are both an original and new predicted site of interaction for a given compound are not shown twice. High sequence identity (highlighted red and bold). Partial sequence identity (highlighted yellow and bold). Sequence alignments: MeV (NP_056924), HPIV3 (AXU38775), NiV (ACT32616), HeV (APT69531), HPIV5 (YP138518), RSV (YP_009518860), HMPV (AAQ67700), RABV (ABN11300), VSV (UTK57347), EBOV (AHX24663), Marburg virus; MARV (ABA87130).

Fig 11

Distinct binding sites of GHP-88309, ERDRP-0519, AVG-233, and JNJ-8003 in the mononegavirus polymerase. (A) In silico predicted binding sites of GHP-88309, ERDRP-0519, and AVG-233 and the structure-informed target site of JNJ-8003 in a polymerase model generated using MeV L-based PIV5 coordinates (PDB: 6v86). Close-up view of compounds predicted to target the polymerase when the priming and intrusion loops are extended away from the active site, either in a pre-initiation or elongation-1/3 stage (B), and in an initiation (C) conformation, modeled based on RSV P-L-RNA (PDB: 8snx) and VSV P-L (PDB: 5a22), respectively. GHP-88309 (blue), ERDRP-0519 (orange), AVG-233 (red), JNJ-8003 (green), RdRP (light blue), PRNTase (forest green), connector (yellow), MTase (orange), CTD (red), GDN (gray spheres), priming loop (black), intrusion loop (magenta).