The leader proteins of Theiler’s virus and Boone cardiovirus use a combination of short linear motifs (SLiMs) to target RSK kinases to the nuclear pore complex

Figures

• 

  Fig 1. LWT and LF48A, but not LM60A localize at the nuclear envelope in infected cells. Representative confocal microscopy images of live HeLaS1-10 cells infected with S11-L viruses for 14 h at an MOI of 5 PFU/cell. Note that in the context of viral infection, mutant LM60V was used and not LM60A for historical reasons. Arrows point at L protein localization at the nuclear envelope (n = 2 biological replicates). Scale bar: 10 µm.

• 

  Fig 2. 20 amino acid peptides of TMEV and BCV L are sufficient to target GFP to the NPC. (A) Mapping of the Cardiovirus L protein region sufficient for GFP-fusion localization to the NPC. Tested constructs are shown (solid line: region included in the constructs; dotted line: deleted region; red squares mark the F-to-A mutation in the RSK binding motif right of the constructs; + denotes nuclear envelope localization and − denotes absence of nuclear envelope localization; framed are sequences defined as sufficient to target GFP to the NPC. A more detailed description of expression plasmids is presented in Table S1. (B) Alignment of representative cardiovirus L proteins with outline of the Zn finger, the DDVF-like SLiM involved in RSK binding, and the identified NPC localization SLiM (framed). SAFV: Saffold virus, a human virus related to TMEV. (C) High-resolution confocal microscopy images of live HeLa cells co-transfected with constructs expressing NUP50-mCherry, and GFP-L variants of TMEV L. Right panels show a 10-fold magnification of the white rectangle area shown in the merge image. Scale bar: 10 µm.

• 

  Fig 3. Alanine scan mutagenesis to define TMEV and BCV L NPC-targeting motifs. (A and B) Alanine-scan mutagenesis of GFP-L57–76 of TMEV (A) and GFP-L62–81 of BCV (B). Graph showing the percentage (mean and SD) of cells with GFP localization at the nuclear envelope (mCherry-Lamin co-localization) in four independently infected cultures (five images were counted per well). * denotes significance in a t-test, where mutants were compared to the corresponding WT peptide. (C) Sequence alignment showing key amino acids required for TMEV L toxicity, TMEV and BCV L NPC localization, and the corresponding peptide in SAFV L. Red residues strongly (>50%) affect NPC localization and blue residues more slightly affect NPC localization, yet significantly.

• 

  Fig 4. Conserved methionines are important for nucleocytoplasmic traffic disruption. (A) Western blot confirming replication (3D polymerase detection) of mutant viruses that do not trigger NCTD (NUP62?: protein tentatively identified as NUP62; ns: non-specific band detected by the anti-3D antibody; the upward shift in NUP98 denotes hyperphosphorylation). Proteins were collected 10 h after infection of HeLa cells expressing GFP-NES and RFP-NLS with 2.5 PFU/cell of indicated viruses. (B) Live cell imaging of cells analyzed in A, just before protein harvest for western blotting (n = 1).

• 

  Fig 5. Structural predictions indicate that cardiovirus L proteins could interact with RAE1 and NUP98 through a similar interface as ORF6 from SARS-CoV-2. (A) Sequence alignment of protein motifs predicted to or shown to bind RAE1/NUP98. Sequences are from M (VSV), ORF10 (KSHV or MuHV4), ORF6 (SARS-CoV-1/2), and L (TMEV, SAFV, and BCV). (B) AlphaFold3 predictions of interfaces between TMEV and BCV full-length L or 20-mer peptides with RAE1 and NUP98GLEBS regions used for crystallography of the RAE1-NUP98-ORF6 complex (PDB: 7VPH). Interface predicted template modeling scores (ipTM) >0.8 are considered accurate high-quality predictions. (C) Structural alignment of SARS-CoV-2 C-terminal tail with RAE1 and NUP98GLEBS (PDB: 7VPH) and the AlphaFold predictions of the L 20-mer peptides of TMEV and BCV, showing similar insertion of the methionine (black arrow) in the hydrophobic pocket of RAE1. (D) Multiple BCV models showing either M64 or M78 insertion in the hydrophobic pocket.

• 

  Fig 6. TMEV and BCV L compete with SARS-CoV-2 ORF6 for RAE1 interaction. (A) Western blots showing the detection of endogenous RAE1 and NUP98 after immunoprecipitation of FLAG (FLAG-L or FLAG-ORF6) from 293T cells transfected for 24 h with plasmids expressing the indicated FLAG-tagged construct. Immunoprecipitation (IP) is shown on the top panel and total cell lysates (Input) are shown on the bottom panel. (B and C) Representative western blots (n = 3) showing the competition for RAE1 co-immunoprecipitation, between BCV L and SARS-CoV-2 ORF6. 293T cells were co-transfected for 48 h with constructs expressing BCV FLAG-L and GFP fused to the 20-mer NPC targeting peptide of SARS-CoV-2 ORF6 (ORF642–61) (B) or with constructs expressing FLAG-ORF6 and GFP fused to the BCV L62–81 NPC targeting peptide (C). FLAG-tagged constructs and competitors were transfected at different ratios: 8:2 and 6:4. Competitors contained either the WT or the indicated mutant NPC targeting peptide. (D) High-resolution microscopy images showing examples of cells displaying clear competition for NPC targeting, between SARS-CoV-2 ORF6 and TMEV L. Left four panels: HeLa cells were co-transfected for 36 h with 25 ng of…

• 

  Fig 7. Cardiovirus L proteins disturb nucleocytoplasmic traffic to a bigger extent than ORF6 (SARS-CoV-2) and ORF10 (KSHV). Confocal microscopy of HeLa cells expressing GFP-NES and RFP-NLS transfected with FLAG-L (TMEV or BCV), FLAG-ORF6, and FLAG-ORF10. Twenty-four hours post transfection, cells were stained with an anti-FLAG antibody. Arrows point at cells expressing FLAG tagged proteins. n = 2. Scale bar: 10 µm.