Cryo-EM structures of Thogoto virus polymerase reveal unique RNA transcription and replication mechanisms among orthomyxoviruses

Abstract

Influenza viruses and thogotoviruses account for most recognized orthomyxoviruses. Thogotoviruses, exemplified by Thogoto virus (THOV), are capable of infecting humans using ticks as vectors. THOV transcribes mRNA without the extraneous 5' end sequences derived from cap-snatching in influenza virus mRNA. Here, we report cryo-EM structures to characterize THOV polymerase RNA synthesis initiation and elongation. The structures demonstrate that THOV RNA transcription and replication are able to start with short dinucleotide primers and that the polymerase cap-snatching machinery is likely non-functional. Triggered by RNA synthesis, asymmetric THOV polymerase dimers can form without the involvement of host factors. We confirm that, distinctive from influenza viruses, THOV-polymerase RNA synthesis is weakly dependent of the host factors ANP32A/B/E in human cells. This study demonstrates varied mechanisms in RNA synthesis and host factor utilization among orthomyxoviruses, providing insights into the mechanisms behind thogotoviruses' broad-infectivity range.

Fig. 1. In vivo and in vitro RNA synthesis by THOV polymerase.

a High-throughput sequencing analysis of the 5′ end regions of mRNA synthesized by mini-replicons of influenza A virus (IAV) and two thogotoviruses – Thogoto virus (THOV) and Dhori virus (DHOV). Histograms show the lengths and frequencies of non-templated (heterologous) sequences found in mRNA transcribed by DHOV, THOV and IAV mini-replicon systems. b Size-exclusion chromatography of THOV polymerase complex and the associated Coomassie-blue stained SDS-PAGE gel showing purified THOVPol containing three polymerase subunits. c Schematics showing sequences and features of 3′ vRNA template, capped (top panel) and uncapped primers (bottom panel). A representative result from at least 3 independent experiments is shown. d In vitro transcriptional and replicative primer extension assays confirming the RNA synthesis activities of the purified THOVPol. Capped (lane 1–6) and uncapped (lane 7–12) primers were tested. Capped/uncapped 3-nt expected elongation products are observed in the presence of CTP alone (lane 5 and 11, indicated by green arrows); Capped/uncapped 10-nt expected elongation products are observed when both ATP and CTP are present (lane 6 and 12, indicated by red arrows). The RNA products from the elongation reactions exhibit heterogeneity (extra bands in lane 5–6 and 11–12) likely due to premature termination (bands smaller than the expected products) and run-off products (bands larger than the expected products). A representative result from at least 3 independent experiments is shown. Source Data are provided as a Source Data file.

Fig. 2. Structures of THOVPol in transcription pre-initiation and initiation conformations.

a Schematic representation of the domain organizations in THOVPol subunits. The color scheme for each component of the polymerase complex is used in b–d. b–d Three different views of THOVPol structures in pre-initiation and initiation conformations. 5′ vRNAs are colored magenta; 3′ vRNAs are colored violet; and GMPCPP is colored orange. Boxed areas in d indicate magnified features shown in e–g. e Details of 5′ vRNA binding within the 5′ promoter binding site. Residues interacting with RNA are shown and labeled, black dashed lines indicate hydrogen bonds and electrostatic interactions. Conventional basepairs within the 5′ vRNA are indicated by hydrogen bonds between bases. f Sequence complementarity between nucleotides 11–17 of the 5′ vRNA and nucleotides 10–16 of the 3′ vRNA in the RNA duplex formed in the initiation conformation. Hydrogen bonds between conventional basepairs are indicated by dash lines. g Close-up views of nucleotides 1–10 of 3′ vRNA and their interaction with polymerase active site residues and GMPCPP. Residues interacting with RNA are shown and labeled, black dashed lines indicate hydrogen bonds and electrostatic interactions.

Fig. 3. Structures of the monomeric THOV polymerase complexes formed during transcriptional RNA synthesis.

a Three different views of elongation THOVPol (THOVPol-EL) in cartoon representation. b The 10-basepair RNA duplex formed between RNA product and 3′ vRNA in polymerase active site cavity. The template exit channel in THOVPol-EL is circled and the trajectory of template exit is indicated by a black arrow; the product exit channel in THOVPol-EL is circled and a red arrow marks the product RNA exit trajectory. c Two structural motifs, PB2^208-221 and PB1^665-680, are found to stack on the template nucleotide U1 and the 5′ cap-nucleotide. d The stretched PB2^43-53 loop allows interactions with the phosphate-ribose backbone of the product RNA. Dash lines indicate hydrogen bonds and electrostatic interactions. The VDW contacts are represented as black dashed curves. e Three different views of the receiving THOVPol (THOVPol-RE) in cartoon representation. The product RNA is colored orange, and the cap-1 structure (m7GpppAm) is shown as spheres.

Fig. 4. Formation of an asymmetric THOVPol dimer during transcriptional RNA synthesis.

a–d Four different views of the asymmetric THOVPol dimer formed during transcriptional RNA synthesis. The product RNA is colored orange, and the cap-1 structure (m7GpppAm) is shown as spheres. The template exit channel in THOVPol-EL is circled and the template exit trajectory is indicated with a black arrow. The product exit channel in THOVPol-EL is circled and the product exit trajectory is indicated with a red arrow. e Product RNA (orange) bound to the product-receiving pocket (5′ promoter binding site) formed by PA and PB1. Cap-1 structures comprising the triphosphate group and the 2-O-methyl group at nucleotide A1 located at the 5′ end of the product RNA are indicated and shown as sticks. Sidechains of the RNA interacting residues from the product-receiving pocket are shown as sticks; they form similar interactions as observed for 5′ vRNA binding (see Fig. 2e); hydrogen bonds and salt-bridges formed by them are omitted for clarity. The black dashed lines indicate the interactions from K309^PA of the handle-helix to the triphosphate group and bases (G2 and A10). The conventional basepairs within the bound product RNA are shown as thin gray dashed lines. Note that the triphosphate group engages in intra-molecular interaction forming 2 charged hydrogen bonds with G2 guanine of product RNA. f Multiple sequence alignment of the PA handle-helix regions of different orthomyxoviruses. The highlighted (blue) residues, K305^PA, K306^PA and K309^PA, are conserved in the thogotovirus genus. g Comparison of the 5′ promoter binding sites and the bound RNA molecules in polymerases of THOV and influenza viruses (PDB: 6RR7 (IAV), 4WRT (IBV), 6XZG (ICV), 6KUU (IDV)).

Fig. 5. The interfaces of the asymmetric THOVPol dimer.

a The asymmetric THOVPol dimer is shown with regions involved in dimer formation highlighted in colors. The locations of the three interaction interfaces primarily between the PA of THOVPol-RE and the PB2 of THOVPol-EL are shown by the dashed boxes and their structures are detailed in b–d. b Interactions in interface 1 between THOVPol-RE PB2-627 domain (salmon) and THOVPol-RE PA structural elements (blue). Hydrogen bonds and salt-bridges are shown by dashed lines, amide-π interactions are indicated by wide dashed lines. c Interactions in interface 2 between THOVPol-RE PA handle-helix (blue) and THOVPol-EL PB2 (green), PB1 (gray) and PA (blue) structural elements. d Interface 3 interactions between THOVPol-RE PA (blue) and THOVPol-EL PB2 structural elements (green and lime).

Fig. 6. Asymmetric dimer structures of THOVPol and FluCPol compared.

Up-and-down comparison of THOVPol (a, b) with FluCPol (d, e) (PDB 6XZQ) asymmetric dimer structures. Note that ANP32A binds at the FluCPol dimer interface. The THOVPol and FluCPol are aligned based on a structural alignment between THOVPol-RE and FluCPol_EN. In a and d, Polymerase dimers are shown as topview surface representations; In b and e, Polymerase dimers are shown in cartoon sideviews. Location of the THOVPol-EL PB2-627 domain is indicated in a and b. The location of interface 1 involving the THOVPol-EL PB2-627 domain is shown in the dashed box in a; the configuration of the peripheral domains near the dimer interface 1 in the THOVPol dimer (a) is noticeably different from that in the FluCPol dimer (d). c The structural elements near the THOVPol-RE PA handle-helix are magnified and the structural elements involved in dimer formation are highlighted in dashed red-boxes. f The equivalent FluCPol_EN regions near the dimer interface are magnified and structural elements involved in dimer formation are highlighted in the dashed red-box. g Effects of dimer interface mutations on THOVPol activity as measured by a mini-replicon driven luciferase assay. h Effects of dimer interface mutations on levels of vRNA, mRNA and cRNA in a mini-replicon assay as measured by RT-qPCR. i In a human ANP32A (hANP32A), ANP32B (hANP32B) and ANP32E (hANP32E) triple knock-out (TKO) 293 T cell line, activities of THOVPol and FluAPol were assessed by a mini-replicon driven luciferase assay; effects of various human (h), mouse (m) and tick (t) ANP32 protein (hANP32A, hANP32B, hANP32E, mANP32A, mANP32B, mANP32E, tANP32A and tANP32B) expression in the TKO cells on THOVPol (THOV) and FluAPol (IAV) activities were tested. Activity and expression level data are reported as mean ± SEM from three independent experiments (n = 3). All statistics used one-way ANOVA Dunnett’s multiple comparisons test. (ns, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001). Source Data are provided as a Source Data file.