Structural and Functional Basis of the Fidelity of Nucleotide Selection by Flavivirus RNA-Dependent RNA Polymerases

Abstract

Viral RNA-dependent RNA polymerases (RdRps) play a central role not only in viral replication, but also in the genetic evolution of viral RNAs. After binding to an RNA template and selecting 5'-triphosphate ribonucleosides, viral RdRps synthesize an RNA copy according to Watson-Crick base-pairing rules. The copy process sometimes deviates from both the base-pairing rules specified by the template and the natural ribose selectivity and, thus, the process is error-prone due to the intrinsic (in)fidelity of viral RdRps. These enzymes share a number of conserved amino-acid sequence strings, called motifs A-G, which can be defined from a structural and functional point-of-view. A co-relation is gradually emerging between mutations in these motifs and viral genome evolution or observed mutation rates. Here, we review our current knowledge on these motifs and their role on the structural and mechanistic basis of the fidelity of nucleotide selection and RNA synthesis by Flavivirus RdRps.

Keywords: Flavivirus; RNA genome; RNA polymerase; RNA synthesis; RNA virus; active site; fidelity; mutation; nucleotide inhibitor; selectivity.

Figure 1

The mechanism of RNA synthesis by a viral de novo initiating RdRp. Although primer synthesis can be template-independent in flaviviruses, the current general model for de novo RNA synthesis by viral RdRp including Flavivirus RdRp is best described with three distinct phases, that of template binding into a productive mode (upper scheme), followed by primer synthesis (here, a di-ribonucleotide, middle scheme) and primer elongation (lower scheme). The constants k₊₁ to k₊₅ and k₋₁ to k₋₅ represent the forward and backward kinetic constants of each individual reversible reaction, respectively. The constants k₊₁ to k₊₅ and k₋₁ to k₋₅ for each phase are not equivalent, as they do not represent the same reactions. T, template; E₀, free enzyme; E, enzyme bound to the template under a productive mode for nucleotide binding; *E, catalytically competent enzyme for nucleotide incorporation; NTP, correct (i.e., template-complementary) ribonucleotide 5′-triphosphate; R_n, RNA made of the primer of n nucleosides annealed to the template. Only the dinucleotide primer synthesis step is shown, primer synthesis may comprise more steps; PPi: pyrophosphate.

Figure 2

The NTP-mediated nucleotide excision pathway by HCV RdRp [30]. A mismatched 3′-end nucleotide is excised by a pyrophosphorolysis-like reaction, in which the pyrophosphate donor comes from an undefined NTP of the intracellular nucleotide pool. E, enzyme; R_nN_m: primer annealed to the template with a 3′-terminal mismatch; R_n: perfectly matched doubled-stranded RNA; *E: catalytically competent enzyme; NTP: correct (i.e., template-complementary) NTP; N_mp₄N: diribonucleoside 5′-5′ tetraphosphate made of the 3′-base of mismatched nucleoside N_m linked by a 5′-tetraphosphate-5′ bridge to a nucleoside coming from the intracellular nucleotide pool.

Figure 3

Flavivirus RdRp structures. (A) Structure of the ZIKV RdRp domain (PDB ID: 5U0C) in ribbon representations above and surface representations below. Back, front, and top views are shown. Important structural and functional features are colored (palm subdomain, gray; fingers subdomain, dark red (index finger, dark green and ring finger, gold); thumb subdomain, medium blue (priming loop, light blue)) and/or labeled. The side chains of the active site residues D335 (motif A) and D666 (motif C) are shown in sticks. (B) Structures of ZIKV full-length NS5 (PDB ID: 5U0B), JEV NS5 (PDB ID: 4K6M), and DENV3 NS5 (PDB ID: 4V0Q) containing the RdRp and methyltransferase (MTase in light brown) domains in surface representation using the same color code as in (A). Back views are rotated by ca. 20° to the right in relation to the representation in (A) in order to show the NTP entry, which is partially obstructed by the MTase domain.

Figure 4

Conservation of sequence and structural motifs A to G within Flavivirus RdRps (Weblogo) and selected positive-strand RNA virus RdRps (alignments). Motifs are given as they appear; N- to C-terminus in the primary sequence of the RdRps: G, F, A, B, C, D, E. Selected sequences of positive-strand RNA virus RdRps are from: ZIKV (PDB ID: 5U0C), JEV (PDB ID: 4K6M), DENV1 (NCBI NP_722465), DENV2 (PDB ID: 5K5M), DENV3 (PDB ID: 5CCV), DENV4 (NCBI NP_740325), YFV (NCBI NP_041726), WNV (NCBI NP_041724), St. Louis encephalitis virus (SLEV, NCBI YP_001008348), HCV (PDB ID: 4WTA), bovine viral diarrhea virus (BVDV, PDB ID: 1S49), PV (PDB ID: 3OL6), Coxsackie virus B3 (CVB3, PDB ID: 4ZPD), and bacteriophage Q-β (PDB ID: 3AVT). The alignment was based on the structural alignment of the RdRps provided by Olve Peersen for the present series of reviews. Manual adjustments were done within motifs F and D. Numbers correspond to the first and last residue in the given motif. The alignment representation was generated using Jalview [71], which applies ClustalX color codes. The secondary structure of ZIKV RdRp (PDB ID: 5U0C) is given above the sequence alignment. For the Weblogo part, the RefSeq protein database of NCBI was searched for “NS5 Flaviviridae”; 123 sequences were recovered of which repeated sequences were removed. Of the resulting 82 Flavivirus sequences, sequences with unknown residues were removed. Finally, the amino acid conservation within the motifs of 79 unique Flavivirus RdRp sequences was analyzed using Jalview [71] and represented with Weblogo [70]. The color codes given byWeblogo consider the chemical characteristics of each amino acid.

Figure 5

Conformational changes of motifs A, B, and D of PV RdRp and/or HCV RdRp upon reaction (left panels) in comparison to motif conformations of Flavivirus RdRps (right panels). (A) Motif A; left panel: ribbon representation of PV RdRp (PV closed, PDB ID: 3OL7, blue) in complex with template (only the base part of −1 and +1 shown) and product (only P−1 and P+1 shown) represented in sticks (colored according to atom type: C, blue; O, red; P, orange), and two Mg²⁺ ions as blue spheres, just after the incorporation of P+1. The liberated pyrophosphate (PPi) is still present at the catalytic site. PV open (PDB ID: 3OL6, dark green) corresponds to a complex after reaction and translocation. Only the protein is shown. Selected amino acid side chains are shown and colored according to atom type. Conformational changes are indicated by black arrows. Right panel: PV open as in the left panel superimposed to ZIKV RdRp (PDB ID: 5UOC, orchid) and WNV RdRp (PDB ID: 2HFZ, grey). The Mg²⁺ ion (grey) belongs to the WNV RdRp structure. (B) Motif B; left panel: ribbon representation of HCV RdRp (HCV closed, PDB ID: 4WTA, gold) in complex with template (only −1 and +1 nucleobases are visible), product (only P−1 shown), and incoming UDP (P+1) represented in sticks (colored according to atom type: C, gold; O, red; P, orange), and two Mn²⁺ ions as golden spheres. HCV open (PDB ID: 2XI2, black) corresponds to the apo enzyme. Selected amino acid side chains are shown and colored according to atom type. Conformational changes are indicated by black arrows. Right panel: HCV closed, as in the left panel, superimposed to ZIKV RdRp (PDB ID: 5UOC, orchid), WNV RdRp (PDB ID: 2HFZ, grey), DENV3 NS5 (PDB ID: 4V0Q, light blue) and DENV3 NS5 (PDB ID: 5CCV, brown). (C) Motif D; left panel: PV closed (PDB ID: 3OL7, blue) with template, product, and PPi (see explanation in (A). A larger protein stretch than only motif D is shown including the helix before and the loop after motif D. For PV open (PDB ID: 3OL6, dark green) only the protein region is shown that corresponds to motif D as given in Figure 4. Selected amino acid side chains are shown and colored according to atom type. A conformational change of the loop and residue K359 is indicated by a black arrow. Right panel: larger motif-D stretch of PV closed shown with template, product, and PPi as in the left panel, and superimposed to ZIKV RdRp (PDB ID: 5UOC, orchid), WNV RdRp (PDB ID: 2HFZ, grey) and bacteriophage Q-β RdRp (PDB ID: 3AVT, green). Selected amino acid side chains are shown and colored according to atom type.

Figure 6

Motifs F and G of PV RdRp and HCV RdRp complexes (left panels) in comparison to motif conformations of Flavivirus RdRps (right panels). (A) Motif F; left panel: ribbon representation of PV RdRp (PV closed, PDB ID: 3OL7, blue) in complex with the template and product represented in sticks (colored according to atom type: C, blue; O, red; P, orange), and two Mg²⁺ ions as blue spheres, just after the incorporation of P+1. The formed pyrophosphate (PPi) is still present at the catalytic site. Only motif F is shown of HCV RdRp (HCV closed, see explanations in the legend of Figure 5B, PDB ID: 4WTA, gold). Selected amino acid side chains are shown and colored according to atom type. Right panel: PV closed (blue) as in the left panel superimposed to WNV RdRp (PDB ID: 2HFZ, grey, motif F only partial and in helical conformation, 90° to the catalytic conformation), ZIKV RdRp (PDB ID: 5UOC, orchid), ZIKV NS5 (PDB ID: 5U0B, dark red), JEV RdRp (PDB ID: 4MTP, green). The arrow indicates the significant flexibility of motif F in Flavivirus RdRps. (B) Motif G; left panel: as in (A), PV RdRp (PV closed, PDB ID: 3OL7, blue) with the template, product, and PPi. Only motifs G and F are shown of HCV RdRp (HCV closed, see explanations in the legend of Figure 5B, PDB ID: 4WTA, gold). Right panel: PV closed (blue) as in the left panel superimposed to ZIKV RdRp (ZIKV 1 and ZIKV 2, chains A and D, respectively, of PDB ID: 5UOC, orchid and purple, respectively), ZIKV NS5 (PDB ID: 5U0B, dark red), JEV RdRp (PDB ID: 4MTP, green, motif G only partial) and JEV NS5 (PDB ID: 4K6M, dark blue).

Figure 7

Current model of the involvement of catalytic and structural motifs of Flavivirus RdRps in NTP incorporation into RNA and RNA synthesis fidelity. Above the reaction scheme catalytic steps (explained in Figure 1, lower scheme) are given and the involvement of the motifs in each step. Below the putative involvement of the motifs in RNA synthesis fidelity is given based on evidence or assumptions explained in the text. For explanations of the reaction scheme, see the legend of Figure 1.