Fig 1.
Design and characterization of Coxsackievirus B3 with catalytically-inactive 2A protease (2Apro).
(A) Schematic representation of the wild-type (wt) Coxsackievirus B3 (CVB3) genome and its engineered derivative viruses with inactivated 2Apro (2Amut). The 2Apro cleavage site was replaced by a Theiler’s murine encephalomyelitis virus (TMEV)-derived IRES sequence, a T2A peptide, or a sequence corresponding to the cleavage site of 3C/3 CD protease (3Cpro) (3 CDcs). Specific cleavage sites of the viral proteases are indicated with blue (2Apro) and green (3Cpro) triangles. (B) Growth kinetics of recombinant 2Apro wt viruses (3 CDcs-2Awt, T2A-2Awt and IRES-2Awt) in HeLa-R19 cells infected at a multiplicity of infection (MOI) of 5, as measured by TCID50 assay (upper panel) or qPCR (lower panel). (C) Growth kinetics of the different 2Awt and 2Amut viruses in HeLa-R19 cells. Experiment was performed as in panel (B). (D) HeLa-R19 or HEK293-T cells were transfected with equal amounts of in vitro transcribed RNA of Renilla-luciferase (RLuc)-3 CDcs-2Awt or -2Amut, and luciferase levels were determined at the indicated times post infection. (E) Growth curves of 2Awt and 2Amut viruses obtained from infected HeLa-R19, BGM, Vero E6 and A549 cells. Data represent the mean ± SEM of one representative experiment (B-D) or three independent experiments (E). The illustration in (A) was created in BioRender. Schipper, J. (2025) https://BioRender.com/7bqvkpe. Statistical significance was assessed by multiple two-tailed unpaired t-tests with multiple comparisons corrections using the Bonferroni-Dunn method (* p < 0.05; ** p < 0.01; *** p < 0.001).
Fig 2.
2Apro activity is essential for CVB3 replication and pathogenesis in vivo.
(A) Schematic representation of the experimental procedure. SJL mice (n = 4 per group) were infected intraperitoneally with equivalent viral doses (either 1.0*107 TCID50/ml or 5.0*105 TCID50/ml) of CVB3, CVB3–2Awt or CVB3–2Amut. Mice were sacrified three days post-infection for pancreas isolation and processing. (B) Viral RNA levels and infectious viral titers in the pancreas of infected mice were determined by RT-qPCR and end-point titration, respectively. (C) Pancreatic histopatology was assessed using a standardized scoring system ranging from 0 to 4 for necrosis (0 = not affected, 1 = < 10% of tissue affected, 2 = 11–25% of tissue affected, 3 = 26–50% of tissue affected, 4 = more than 50% of tissue affected), and from 0 to 3 for inflammation and edema (0 = none, 1 = mild, 2 = moderate, 3 = severe). Analyses were conducted without knowledge of the experimental groups. Statistical signifance was determined by Mann-Whitney U test. Differences were considered significant if p < 0.05. The illustration in (A) was created in BioRender. Schipper, J. (2025) https://BioRender.com/0oodpvz.
Fig 3.
2Apro plays a critical role in distrupting nucleocytoplasmic trafficking in CVB3 infected cells.
(A) Western blot analysis for Nup98, FG-repeat-containing nucleoporins, CVB3 2C protein and β-actin. HeLa-R19 cells, (mock-)infected with CVB3, 3 CDcs-2Awt, or 3 CDccs-2Amut at an MOI of 10, were lysed at 2, 4, 6, 8 and 10 hpi. Black triangles denote intact proteins, white triangles indicate cleavage products, asterisk refers to a non-specific band which has been observed before with this antibody [18]. For densitometric analysis of all western blot data, see S1 Data. (B) Nucleocytoplasmic trafficking disorder (NCTD) is abolished in 3 CDcs-2Amut infected cells. HeLa-3xNLS-GFP were infected at MOI of 10 with CVB3, 3 CDcs-2Awt, and 3 CDcs-2Amut. Cells were fixed with 4% paraformaldehyde at 2, 4, 6, 8, 10 hpi and analyzed by IF analysis for dsRNA as a marker of infection, for Nup98 as marker of the NPC, for hnRNPK and 3xNLS-GFP localization as indicators of NCTD. Scale bar: 25 μm.
Fig 4.
2Apro, but not 3Cpro, selectively inhibits host translation and suppresses stress granule formation.
(A) Western-blot analysis of eIF4GI, DAP5, PABP, PKR, p-PKR and p-eIF2α. Cell lysates correspond to the same experiment shown in Fig 3A. Black triangles denote intact proteins, white triangles indicate cleavage products. (B) THRONCAT in vivo metabolic labeling of HeLa-R19 cells infected with 3 CDcs-2Awt and 3 CDcs-2Amut. Cells treated with sodium arsenite served as a positive control for translation arrest. Lysates were run on SDS-PAGE gels and subsequently transfered to nitrocellulose membranes for analysis. (C) Representative images (left) and quantification (right) of SG formation in HeLa-R19 cells infected at an MOI of 10 with CVB3, 3 CDcs-2Awt or 3 CDcs-2Amut. Cells were fixed at 4, 6, 8, 10 hpi and analyzed by IF analysis for dsRNA as a marker of infection and eIF3 and G3BP2 as SG markers. Magnified regions are shown in white boxes. Scale bar: 25 μm. Quantification of small blue and large purple SG formation is shown on the right. At least 1000 cells were analyzed for each virus and time point. Small SG positivity is given as a percentage of total cells, since most cells are not yet positive for dsRNA at early timepoints. (D) Western blot analysis of G3BP1, CVB3 2C protein, and β-actin. HeLa-R19 cells, either mock-infected or infected with recombinant CVB3s at an MOI of 10, were lysed at 6, 8, 10, 12 and 16 hpi. Black triangles denote intact proteins, empty triangles indicate cleavage products.
Fig 5.
Distinct effects of CVB3-2Awt and CVB3-2Amut on cell morphology and death.
(A) HeLa-R19 cells were infected with 3 CDcs-2Awt and 3 CDcs-2Amut at an MOI of 50, fixed at 6, 8, 10, 12 and 16 hpi and, and analyzed by IF analysis. Nuclei were stained with DAPI, dsRNA was detected as marker for infection, and β-tubulin was used as a cytoskeleton marker. Fragmented nuclei are indicated by white triangles. Scale bar: 20 μm. (B) Quantification of infected cells displaying nuclear fragmentation in panel (A).
Fig 6.
2Apro is the key viral protease responsible for the suppression of type I interferon signaling.
(A) HeLa-R19 cells were infected at an MOI of 10 and lysed at 2, 4, 6, 8, 10 and 16 hpi. Total RNA was analyzed by RT-qPCR for IFN-β, IFIT1, viral RNA (vRNA) and actin. IFN-β, IFIT1 and vRNA levels were calculated as fold induction compared to levels in mock-infected cells, after correction for actin mRNA levels, and normalized to the expression levels at 2 hpi. Data represent mean ± SEM of three technical replicates. Statistical significance was calculated by multiple two-tailed unpaired t-tests; * p < 0.05; ** p < 0.01; *** p < 0.001. (B) Western blot analysis for MDA5, RIG-I, MAVS, IF3 and p-IRF3. Cell lysates belong to the same experiment shown in Fig 3A. Black triangles point out intact proteins, white triangles indicate the cleavage products. (C) Hela-R19 cells were pretreated for 16 hrs with IFN-α and subsequently infected with RLuc-3 CDcs-2Awt and -2Amut viruses. Cells were lysed at 6 hpi and luciferase levels were assessed. Data represent the mean ± SEM of three independent experiments normalized to the untreated control. Statistical significance was assessed by multiple two-tailed unpaired t-tests with multiple comparisons corrections using the Bonferroni-Dunn method (* p < 0.05; ** p < 0.01; *** p < 0.001).
Fig 7.
Single-molecule imaging to dissect the role of 2Apro during the earliest phases of CVB3 infection.
(A) Schematic representation of the SunTag-CVB3 viruses. 5xSunTag cassette flanked at its 3’-side by a 3 CDcs was inserted before the P1 region of CVB3, 3 CDcs-2Awt, or 3 CDcs-2Amut. (B) Schematic representation of the VIRIM assay. Left panel shows active translation. Right panel denotes situation during viral replication. The blue region of the viral genome indicates the SunTag peptide sequences. Note the opposing directions of both processes, and their consequent temporal exclusiveness. (C) Schematic representation of the different phases of early enterovirus infection as defined by Boersma et al. 2020. (D) Pictures of the different phases of viral infection, as observed for representative infections with 3 CDcs-2Awt, and 3 CDcs-2Amut, respectively. The time since start phase 1 is stated in grey. The numer of translating vRNAs in the infected cell is stated in red. Schematic illustration of VIRIM spot number is shown in the top panel. The far right graphs show VIRIM spot count for the infections shown in the representative images on the left. (E) Kaplan-Meier curves depicting the duration of the different phases of infection for CVB3, 3 CDcs-2Awt, and 3 CDcs-2Amut. (F) The average number of spots observed over time during CVB3, 3 CDcs-2Awt, and 3 CDcs-2Amut infection. Shaded areas represent SEM. (G) Violin plots depicting the average rate of increase in the number of translating vRNAs during phase 3 (phase 3 slope) (left), and the average number of translating vRNAs in phase 4 (phase 4 plateau, see also (C)) (right). Scale bar 10 μm. The illustration in (A, B, and D) were (partially) created in BioRender. Schipper, J. (2025) https://BioRender.com/unbzo59.
Fig 8.
The increase in viral translation efficiency during infection depends on the action of 2Apro.
(A) VIRIM spot intensity was measured during phase 1 and phase 4 in the same infected cells. Spot intensities are plotted relative to phase 1 spot intensity. Each line and red dot represents an individual infected cell. Median values are indicated above the graphs. (B) Normalized intensity (AU) of phase 1 translation spots during infection with CVB3, 3 CDcs-2Awt, and 3 CDcs-2Amut. The higher intensity of the 3 CDcs viruses can be explained by the comparatively longer effective transcript length (i.e., the length of vRNA that is decoded by the ribosome before the SunTag containing nascent polypeptide is released from the ribosome) of these viruses due to the replacement of the 2Acs with the 3 CDcs. Statistical significance was assessed by a two-way analysis of variance (ANOVA) with multiple comparisons testing. * p < 0.05; ** p < 0.01; *** p < 0.001.