Host ANP32A mediates the assembly of the influenza virus replicase

Abstract

Aquatic birds represent a vast reservoir from which new pandemic influenza A viruses can emerge¹. Influenza viruses contain a negative-sense segmented RNA genome that is transcribed and replicated by the viral heterotrimeric RNA polymerase (FluPol) in the context of viral ribonucleoprotein complexes^2,3. RNA polymerases of avian influenza A viruses (FluPolA) replicate viral RNA inefficiently in human cells because of species-specific differences in acidic nuclear phosphoprotein 32 (ANP32), a family of essential host proteins for FluPol activity⁴. Host-adaptive mutations, particularly a glutamic-acid-to-lysine mutation at amino acid residue 627 (E627K) in the 627 domain of the PB2 subunit, enable avian FluPolA to overcome this restriction and efficiently replicate viral RNA in the presence of human ANP32 proteins. However, the molecular mechanisms of genome replication and the interplay with ANP32 proteins remain largely unknown. Here we report cryo-electron microscopy structures of influenza C virus polymerase (FluPolC) in complex with human and chicken ANP32A. In both structures, two FluPolC molecules form an asymmetric dimer bridged by the N-terminal leucine-rich repeat domain of ANP32A. The C-terminal low-complexity acidic region of ANP32A inserts between the two juxtaposed PB2 627 domains of the asymmetric FluPolA dimer, suggesting a mechanism for how the adaptive PB2(E627K) mutation enables the replication of viral RNA in mammalian hosts. We propose that this complex represents a replication platform for the viral RNA genome, in which one of the FluPol molecules acts as a replicase while the other initiates the assembly of the nascent replication product into a viral ribonucleoprotein complex.

Extended Data Fig. 1. FluPol_C activity depends on ANP32A and alignment of ANP32 proteins.

a, b, Luciferase reporter gene activities reflecting FluPol_C activity in control (a) and dKO (b) eHAP cells in the presence or absence of overexpressed huANP32A, huANP32B or chANP32A. Data are presented as mean values ± s.e.m. n = 3 biologically independent samples from n = 3 independent experiments. Ordinary one-way ANOVA with Dunnett’s post hoc test for multiple comparisons. P < 0.05 is considered significant to reject the null hypothesis. c, Sequence alignment of huANP32A, huANP32B, chANP32A, chANP32B. Residues involved in hydrogen bonding interactions with FluPol_C are indicated in orange. The chANP32A avian-specific 33 amino acid insertion is highlighted in cyan. The SUMO interaction motif (SIM) sequence is indicated by black triangles. The figure was prepared with Espript 3.0.

Extended Data Fig. 2. Data collection, processing and analysis scheme.

a, b, Flowchart for the processing and the classification of the FluPol_C-huANP32A complex (a) and FluPol_C-chANP32A complex (b).

Extended Data Fig. 3. Single-particle cryo-EM analysis of FluPol_C-huANP32A and FluPol_C-chANP32A complexes.

a, e, Representative micrograph of FluPol_C-huANP32A (a) and FluPol_C-chANP32A (e) embedded in vitreous ice. Scale bar, 200 Å. b, f, Representative 2D class averages of FluPol_C-huANP32A (b) and FluPol_C-chANP32A (f). c, d, Data analysis for FluPol_C-huANP32A Subclass1 (c) and Subclass2 (d). 3D reconstruction locally filtered and coloured according to RELION local resolution (left panel). FSC curve indicating overall map resolution and model-to-map FSC (middle panel). Curves are shown for phase randomisation, unmasked, masked and phase-randomisation-corrected masked maps. Angular distribution of particle projections with the cryo-EM map shown in grey (right panel). g-j, Data analysis for FluPol_C-chANP32A Subclass1 (g), Subclass2 (h), Subclass3 (i) and Subclass4 (j). 3D reconstruction locally filtered and coloured according to RELION local resolution (top panel). FSC curve indicating overall map resolution and the model-to-map FSC (middle panel). Curves are shown for phase randomisation, unmasked, masked and phase-randomisation-corrected masked maps. Angular distribution of particle projections with the cryo-EM map shown in grey (bottom panel).

Extended Data Fig. 4. Comparison of FluPol^R and FluPol^E structures with the transcriptase and apo conformations of FluPol.

a-d, Comparison of structures of human influenza A/NT/60/68 (H3N2) bound to vRNA and capped RNA in the transcriptase conformation (PDB: 6RR7) (a) and human influenza C/Johannesburg/1/66 in the apo conformation (PDB: 5D98) (b) with structures of FluPol^R (c) and FluPol^E (d) in the FluPol_C-chANP32A complex. e-h, Comparison of the PB2 domain arrangements in the complexes shown in a-d.

Extended Data Fig. 5. Close-up view of the interaction of 5’ and 3’ vRNA termini with FluPol^R.

a, c, Close-up view of the 3’ vRNA pointing towards the active site in the FluPol_C-huANP32A (a) and FluPol_C-chANP32A (c) structures. b, d, Close-up view of the 3’ vRNA binding in a groove located between P3_CTD and the PB1_thumb and PB2_N1 subdomains in the FluPol_C-huANP32A (b) and FluPol_C-chANP32A (d) structures.

Extended Data Fig. 6. Effect of FluPol^R-FluPol^E dimer interface mutations on FluPol_A activity.

a, b, Effect of mutations at the FluPol^R-FluPol^E dimer interface on FluPol_A activity in viral minigenome assays (a) and cRNA encapsidation by FluPol_A (b). Data are presented as mean values ± s.e.m. n = 3 biologically independent samples from n = 3 independent experiments. Ordinary one-way ANOVA with Dunnett’s post hoc test for multiple comparisons. P < 0.05 is considered significant to reject the null hypothesis. Western blot analyses were repeated from n = 3 independent experiments with similar results. For gel source data, see Supplementary Fig. 2.

Extended Data Fig. 7. Effect of FluPol_A mutations at the FluPol_A-ANP32A interface on FluPol_A activity and interaction with huANP32A.

a, b, Effect of FluPol_A mutations at the FluPol_A-ANP32A interface on FluPol_A activity in viral minigenome assays on (a) and FluPol_AANP32A interaction (b). Data are presented as mean values ± s.e.m. n = 3 biologically independent samples from n = 3 independent experiments. Ordinary one-way ANOVA with Dunnett’s post hoc test for multiple comparisons. P < 0.05 is considered significant to reject the null hypothesis. Western blot analyses were repeated from n = 3 independent experiments with similar results. For gel source data, see Supplementary Fig. 2.

Extended Data Fig. 8. Structural comparison of PB2₆₂₇ domains of FluPol_A and FluPol_C.

Structures of the PB2₆₂₇ domains from crystal structures of FluPol from influenza C/Johannesburg/1/1966 (a, PDB ID: 5D98) and A/NT/60/1968 (H3N2) (b, PDB ID: 6QNW) viruses are aligned and shown in cartoon mode. Residues discussed in this study are highlighted in stick mode and coloured in orange.

Fig. 1. Structures of dimers of FluPol_C heterotrimers with or without bound ANP32A.

a-c, Cryo-EM structures of dimers of FluPol_C heterotrimers bound to huANP32A (a), chANP32A (b), and without ANP32A (c).

Fig. 2. FluPol_C-FluPol_C and ANP32A-FluPol_C interaction interfaces.

a, FluPol_C-FluPol_C dimer interface with interacting regions in FluPol^R and FluPol^E highlighted on the molecular surface, by peeling apart the two molecules each by 30°. The letters denote the close-up views shown in panels b-d, Close-up views of the FluPol_C-FluPol_C interaction interface. Dashed lines indicate hydrogen bonds. e, ANP32A-FluPol_C surface interface with interacting regions in FluPol_C highlighted. The letters denote the close-up views shown in panels f-h, Close-up views of the ANP32A-FluPol_C interaction interface. Dashed lines indicate hydrogen bonds.

Fig. 3. Interaction of ANP32_LCAR with FluPol_C and the effect of ANP32A on FluPol_A activity.

a, Schematic of huANP32A and chANP32A highlighting the 33 amino acid (33 aa) insertion in chANP32A and sequence differences. b, Close-up view of the cryo-EM density attributed to chANP32A (grey, threshold 0.934) with the chANP32A_LRR domain represented in cartoon (orange) and the positions of PB2₆₂₇ of FluPol^R (blue) and FluPol^E (green). Regions potentially corresponding to the N-terminal part of ANP32A_LCAR and part of the avian 33 aa insertion are highlighted in cyan and yellow, respectively. V614 and K649 in FluPol_C corresponds to K/Q591 and K/E627 in FluPol_A, respectively. c, Same view as shown in (b), with the PB2₆₂₇ of FluPol^R and FluPol^E shown in surface representation. d, Effect of chANP32A and huANP32A, wild type or mutant (huANP32A176-183), on the activity of FluPol_A with mammalian-adapted PB2 K627 or avian-signature PB2 E627 in a viral RNP reconstitution assay in mammalian HEK 293T cells. Data of primer extension analysis of viral RNA levels with quantitation (upper panels) and western blot analysis of ANP32A levels (lower panel) with molecular weight markers is shown. Data are presented as mean values ± s.e.m. n = 3 biologically independent samples from n = 3 independent experiments. Ordinary one-way ANOVA with Dunnett’s post hoc test for multiple comparisons. P < 0.05 is considered significant to reject the null hypothesis. Western blot analyses were repeated from n = 3 independent experiments with similar results. For gel source data, see Supplementary Fig. 2.

Fig. 4. Functional implications of the FluPol_C-ANP32A complex.

a, Relative positions of the RNA product exit channel in FluPol^R and the 5’ RNA binding site in FluPol^E in the chANP32A-FluPol_C complex. The positions of the RNA product exit channel in FluPol^R and the 5’ RNA binding site in FluPol^E were determined by superposing FluPol^R and FluPol^E with the structure of FluPol_A bound to capped RNA and vRNA template (PDB 6RR7). b, Model for the role of FluPol-ANP32A complex in the replication of the influenza virus RNA genome and its assembly into RNP. FluPol in the context of vRNP or cRNP is flexible (1) but is stabilised in a replicase FluPol^R conformation upon binding of a newly synthesised FluPol in the presence of ANP32A (2). FluPol^R initiates replication in a primer independent manner (3) with a trans-activating FluPol involved in cRNA to vRNA replication by promoting cRNA template realignment (3’). As the 5’ end of the nascent replication product is released from the polymerisation active site of FluPol^R, it is captured in the 5’ RNA binding pocket of the encapsidating FluPol^E bound to FluPol^R (4), initiating the encapsidation of the nascent RNA with NP (5). Nascent vRNA or cRNA assemble into vRNP or cRNP, respectively (6), and are released upon FluPol^R termination. FluPol^E becomes the resident polymerase of the newly produced vRNP or cRNP.