Ebola virus VP24 targets a unique NLS binding site on karyopherin alpha 5 to selectively compete with nuclear import of phosphorylated STAT1

Abstract

During antiviral defense, interferon (IFN) signaling triggers nuclear transport of tyrosine-phosphorylated STAT1 (PY-STAT1), which occurs via a subset of karyopherin alpha (KPNA) nuclear transporters. Many viruses, including Ebola virus, actively antagonize STAT1 signaling to counteract the antiviral effects of IFN. Ebola virus VP24 protein (eVP24) binds KPNA to inhibit PY-STAT1 nuclear transport and render cells refractory to IFNs. We describe the structure of human KPNA5 C terminus in complex with eVP24. In the complex, eVP24 recognizes a unique nonclassical nuclear localization signal (NLS) binding site on KPNA5 that is necessary for efficient PY-STAT1 nuclear transport. eVP24 binds KPNA5 with very high affinity to effectively compete with and inhibit PY-STAT1 nuclear transport. In contrast, eVP24 binding does not affect the transport of classical NLS cargo. Thus, eVP24 counters cell-intrinsic innate immunity by selectively targeting PY-STAT1 nuclear import while leaving the transport of other cargo that may be required for viral replication unaffected.

Figure 1. eVP24 binds an ncNLS site on KPNA5 near C-terminal ARMs 8, 9, and 10 with high affinity and specificity

(A) Co-immunoprecipitation experiments with FLAG antibody were performed from lysates of 293T cells co-transfected with plasmids for Flag-KPNA5 of various lengths as indicated in the figure and HA-eVP24. Western blots were performed for Flag, HA, and β-tubulin. WCL, whole cell lysate. (B) Raw ITC data and corresponding binding isotherms for MBP-tagged eVP24 (residues 11-237) binding to KPNA5 66-509 (left) and KPNA5C (right). Representative data are shown from at least 2 independent experiments. Control experiments show that MBP does not bind KPNA5 or KPNA5C. (C) Asymmetric unit of eVP24/KPNA5C complex contains 3 complexes with 1:1 stoichiometry. eVP24 is shown in blue (mol C) and dark gray (mol A and B); KPNA5C is shown in orange (mol F) and light gray (mol D and E). Only eVP24 residues 16-231 and KPNA5C residues 332-506 were modeled in the final refined structure. (D) Representative 1:1 complex structure between eVP24/KPNA5C reveals multiple structural interaction elements between the two molecules. (E) Alignment of a KPNA structure (yellow; PDB ID: 1BK5) with the C-terminal ARM repeats from the eVP24/KPNA5C complex structure (RMS = 0.986). (F) The three helix construction of the ARM repeats and the 10 ARM repeats of KPNA proteins are shown. cNLS and ncNLS sites are marked. Also see Figure S1 and S5.

Figure 2. The eVP24 and KPNA5 interface reveals extensive interactions

Extensive hydrogen bonding and hydrophobic interactions are observed between eVP24 and KPNA5 (A) ARM 8, (B) ARM 9, and (C) ARM 10. LigPlot⁺ diagrams (top) showing protein-protein interactions between eVP24 and KPNA5. Protein side chains are shown as ball-and-sticks. Hydrogen bonds are shown as green dotted lines. Spoked arcs represent non-bonded contacts. PyMOL representation of the highlighted protein-protein interactions is shown on the bottom. A total of 2,100 Ų is covered between the two molecules. (D) Intrinsic and context dependent α-helical propensities of KPNA5 ARM helices H1, H2, and H3. The ordinate shows the value of f_α (0 ≤ f_α ≤ 1) at 298 K. Also see Figure S2.

Figure 3. Multiple residues from KPNA5 ARMs 8-10 are necessary and sufficient for eVP24 binding

(A) Summary of eVP24 mutants, which were grouped based on the three main residue clusters. (B) Summary of KPNA5 mutants used in the study grouped based on their location on the ARM repeats 8-10. Co-IP experiments, precipitating with anti-HA antibody, were performed on lysates from 293T cells co-transfected with plasmids for Flag-KPNA5 and HA-eVP24 (C) single residue or (D) multiple residue eVP24 mutants as indicated. Co-IPs with Flag antibody were carried out on lysates co-transfected with HA-eVP24 WT and Flag-KPNA5 (E-F) single residue KPNA5 mutants or (G) multiple residue KPNA5 mutants as indicated. Western blots were performed for Flag, HA, and β-tubulin. WCL, whole cell lysate. E, pCAGGS empty vector control. Also see Figure S3 and S4.

Figure 4. eVP24 and PY-STAT1 share an overlapping binding site on KPNA5

293T cells were treated with human IFNβ (1000U/mL) for 30 minutes. Co-IPs with HA or Flag antibody were performed as indicated on the 293T cell lysates transfected with (A) HA-tagged eVP24 or Nipah virus V protein (NipV), (B) KPNA5 truncation mutants and (C-D) KPNA5 single and multiple residue mutants from the eVP24/KPNA5C structural interface. Western blots were performed for PY-STAT1, STAT1, and Flag or HA. WCL, whole cell lysate, E, pCAGGS empty plasmid transfection. Also see Figure S1.

Figure 5. Recognition of a shared ncNLS binding site on KPNA by eVP24 is important for inhibiting PY-STAT1 nuclear localization and ISG induction

(A-B) eVP24/KPNA5 interface mutants R137A and cluster 3 mut fail to inhibit IFN-mediated STAT1 nuclear translocation. Vero cells were mock-untreated or treated with IFN-β for 30 minutes to induce STAT1 nuclear localization ((E) vector + IFN-β). Ectopic expression of eVP24 WT or mutants differentially affects PY-STAT1 translocation to the nucleus after IFN-β treatment. White arrows highlight IFN-β treated cells that also express eVP24 WT or eVP24 mutants. Representative data from one of two independent experiments is shown. (C) The ability of eVP24 to inhibit induction of ISG54 was assessed. 293T cells were co-transfected with an ISG54 firefly luciferase reporter, a constitutively expressed Renilla luciferase plasmid and the indicated eVP24 expression plasmids. The values represent the mean and SEM of six samples, and statistical significance was assessed by a one-way ANOVA comparing individual mutants to the corresponding eVP24 wt transfection, where ***p<0.001, **p<0.01 and *p<0.05. E, pCAGGS empty plasmid transfection. (D) eVP24 competes with PY-STAT1 for binding to KPNA5. FLAG-KPNA5 was used to co-IP PY-STAT1 in the presence or absence of eVP24 or mutants with attenuated KPNA5 binding. Two concentrations (2 and 4 μg) of eVP24 plasmids were tested. E, pCAGGS empty plasmid transfection.

Figure 6. Model of eVP24 inhibition of KPNA mediated STAT1 signaling

(A) Schematic diagram illustrating the domain organization of KPNA containing an N-terminal importin β binding (IBB) domain followed by 10 ARM repeats, each of which is comprised of 3 α-helices (H1, H2, and H3), and a short C-terminus. eVP24 and STAT1 bind overlapping sites in ARMS 8-10 (highlighted in red, blue, and green). Overlapping binding site between eVP24 and KPNA5 is shown as a solid black line and the additional regions potentially important for PY-STAT1 binding are shown as a black dotted line. (B) Relative locations of cNLS and ncNLS sites based on available data, including the current study. (C) Co-IP experiments with Flag antibody were performed on lysates of 293T cells co-transfected with either Flag-KPNA1 or Flag-KPNA5 and Myc-DBC1 and concentrations of 2 and 4 μg of HA-eVP24 as indicated. Western blots were performed on precipitated (IP) material and on whole cell lysates (WCL) for HA, Myc and Flag tags. (D) Model of KPNA (PDB ID: 1BK5) in cylinder representation. The major and minor nuclear localization signals (NLS) span the inner surface of ARMs 2-4 and ARMs 6-8, respectively. The ncNLS used by PY-STAT1 or the KPNA binding site of eVP24 is independent of the cNLS sites. Therefore, KPNA loaded with PY-STAT1 ± cNLS cargo can translocate into the nucleus. In contrast, eVP24 binding, via a portion of the region used by PY-STAT1 via ARMs 8-10, inhibits PY-STAT1 nuclear translocation, but not the transport of cNLS containing cargo. Also see Figure S6, S7, and Table S1.