Differential Disruption of Nucleocytoplasmic Trafficking Pathways by Rhinovirus 2A Proteases

Abstract

The RNA rhinoviruses (RV) encode 2A proteases (2A^pro) that contribute essential polyprotein processing and host cell shutoff functions during infection, including the cleavage of Phe/Gly-containing nucleoporin proteins (Nups) within nuclear pore complexes (NPC). Within the 3 RV species, multiple divergent genotypes encode diverse 2A^pro sequences that act differentially on specific Nups. Since only subsets of Phe/Gly motifs, particularly those within Nup62, Nup98, and Nup153, are recognized by transport receptors (karyopherins) when trafficking large molecular cargos through the NPC, the processing preferences of individual 2A^pro predict RV genotype-specific targeting of NPC pathways and cargos. To test this idea, transformed HeLa cell lines were created with fluorescent cargos (mCherry) for the importin α/β, transportin 1, and transportin 3 import pathways and the Crm1-mediated export pathway. Live-cell imaging of single cells expressing recombinant RV 2A^pro (A16, A45, B04, B14, B52, C02, and C15) showed disruption of each pathway with measurably different efficiencies and reaction rates. The B04 and B52 proteases preferentially targeted Nups in the import pathways, while B04 and C15 proteases were more effective against the export pathway. Virus-type-specific trends were also observed during infection of cells with A16, B04, B14, and B52 viruses or their chimeras, as measured by NF-κB (p65/Rel) translocation into the nucleus and the rates of virus-associated cytopathic effects. This study provides new tools for evaluating the host cell response to RV infections in real time and suggests that differential 2A^pro activities explain, in part, strain-dependent host responses and diverse RV disease phenotypes.IMPORTANCE Genetic variation among human rhinovirus types includes unexpected diversity in the genes encoding viral proteases (2A^pro) that help these viruses achieve antihost responses. When the enzyme activities of 7 different 2A^pro were measured comparatively in transformed cells programed with fluorescent reporter systems and by quantitative cell imaging, the cellular substrates, particularly in the nuclear pore complex, used by these proteases were indeed attacked at different rates and with different affinities. The importance of this finding is that it provides a mechanistic explanation for how different types (strains) of rhinoviruses may elicit different cell responses that directly or indirectly lead to distinct disease phenotypes.

Keywords: 2A; live-cell imaging; nuclear export; nuclear localization; nuclear trafficking; protease; rhinovirus.

FIG 1

Cleavage of NPC Nups by 2A^pro. Reaction mixtures containing isolated HeLa cell nuclei (6 × 10⁶) and recombinant 2A^pro (0.2 nmol) were incubated at 35°C for 4 h. Aliquots were fractionated by SDS-PAGE and then visualized after Western analyses (MAb414).

FIG 2

mCherry cell lines. (A) Expressed mCherry fusion proteins linked to NLS or NES are described in Materials and Methods. (B) Transduced HeLa cells correctly localized mCherry signals to the nuclei or cytoplasm or, if there was no fusion fragment (none), to both. DIC, differential inference contrast. (C) Protein localization signals and the transport receptor responsible for shuttling the mCherry fusion proteins. Image capture was at ×10 magnification.

FIG 3

Transcript design and cleavage verification. (A) GFP-containing RNA transcripts. The arrows indicate a 2A^pro self-cleavage site between the viral capsid 1D protein fragment or GFP and 2A itself. (B) Transduced mCherry reporter cells were transfected with tGFP-2A and then lysed 12 h p.t. The integrity of mCherry was verified in Western assays relative to tubulin. The right-hand lanes have unmodified mCherry markers from control cells. (C) tGFP-2A was translated in reticulocyte lysates. Protein detection after SDS-PAGE fractionation was by Western analyses. (D) Proteins expressed in HeLa cells after transfection with tGFP-2A were analyzed similarly to panel C. Cell collection was at 8 h p.t. except for control samples (tGFP and A16*, 16 h p.t.), which required longer incubation for GFP detection. Relative band intensities by densitometry (Norm) were normalized to A16.

FIG 4

2A^pro disruption of nuclear transport. mCherry-M9 NLS (A) and mCherry PKI NES (B) cells were transfected with GFP or A16 GFP-2A pCITE transcripts. At 10 h p.t., the cells were fixed and stained with DAPI or treated with 6 nM leptomycin B for 30 min and then fixed and stained with DAPI (B, right column). Image capture was at ×10 magnification.

FIG 5

Single-cell fluorescent profiles. Live-cell imaging of transduced HeLa cells transfected with 2A pCITE transcripts was as described in Materials and Methods. The fluorescence intensities for nuclear and cytoplasmic mCherry, nuclear DAPI, and whole-cell GFP were measured in single cells at 15-min intervals for 8 h, beginning 2 h p.t. (A) Example of recorded fluorescent signals in a control nontransfected mCherry-M9 NLS cell. (B and C) Examples of recorded fluorescent signals in A16 2A^pro-transfected mCherry-M9 NLS cells. (D) Example of recorded fluorescent signals in an A16 2A^pro-transfected mCherry-PKI NES cell.

FIG 6

2A^pro transport disruption in single cells. ΔratioN and ΔratioC were measured at several time points (3.5, 4, 5, 6, 7, and 8 h p.t.) and normalized to cellular GFP fluorescence intensities in >200 transfected cells for each 2A^pro as described in Materials and Methods. The results from one experiment with the mCherry-M9 NLS cells are presented as scatter plots for 4 and 6 h p.t. Each point represents a single cell. The gray points are outliers (the upper and lower 5%), and the black points represent the median 90% of analyzed cells. The mean GFP normalized Δratios of the median 90% of cells were calculated and are shown below the graphs. The differences (P values) between the means (t test; PRISM) for each pair of proteases were <0.0001, except as indicated.

FIG 7

Comparative 2A^pro disruption efficiencies. Relative ΔratioN or ΔratioC values induced in GFP-2A-transfected cells were calculated from the mean GFP normalized values measured for the median 90% of cells at 3.5, 4, 5, 6, 7, and 8 h p.t. for replicate experiments performed with each mCherry reporter cell line. The highest measured mean Δratio in a given experiment was set equal to 1 for calculations of relative values. The relative Δratios induced by each 2A^pro and plotted for each mCherry reporter and its cognate trafficking pathway are the mean values from 4 replicate experiments (M9 NLS, transportin 1; RS NLS, transportin 3; SV40 NLS, importin α/β) or 3 replicate experiments (PKI NES, Crm1). The error bars indicate standard deviations.

FIG 8

Reporter relocalization during infection. (A) Dual transformed PKI NES and YFP SV40 cells were infected with A16 virus (MOI = 10). The cell fields were imaged every 15 min for 14 h (times are shown in each image). An example of single-cell progression is shown. (B and C) mCherry-p65/RelA cells were infected with A16 virus with image capture as in panel A (B) or treated with TNF-α (10 ng/ml) with image capture at 1-min intervals for 20 min (C).

FIG 9

Transport disruption during virus infection. (A to D) Dual-transformed PKI NES with YFP SV-40 NLS cells were infected with A16, B04, B14, or B52 (A, B, and D) or chimeric B14-based (C) rhinoviruses. Individual cells (100) in captured cell field images (15-min intervals for 14 h) were visually scored for cytoplasmic/nuclear mCherry signals (A), YFP signals (B), or CPE (D) at each time point (see Materials and Methods). The values were averaged (n = 2 separate experiments) and plotted. (E and F) Similar to panels A and D, mCherry-p65/RelA cells were infected with the indicated viruses. The captured images (100 cells per RV infection; n = 2 separate experiments) were evaluated for signal relocalization (E) and CPE (F).