Norden laboratory feline kidney (NLFK) and human cervical carcinoma HeLa cells were cultured in Dulbecco`s modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 1% non-essential amino acids, and 1% penicillin-streptomycin (Gibco, Thermo Fischer Scientific, Waltham, MA) at 37°C with 5% CO₂. CPV type 2 was derived from an infectious plasmid clone p265 (Parrish, 1991) by transfection of NLFK as previously described (Parrish, 1991; Parker et al., 2001). The viruses were grown in NLFK cells, a virus-containing medium was collected, and the virus was concentrated by ultrafiltration.

Experiments with importin α were conducted with importin-α-GFP transfected cells. Importin-α-GFP plasmid was a generous gift from Enrico Gratton (Irvine, CA, United States). Transfections were performed with TransIT 2020 transfection reagent (Mirus Bio LLC, Madison, WI, United States).

CRM1 inhibitor leptomycin B (LMB) was obtained from Abcam (Cambridge, United Kingdom) and it was used at a concentration of 100 ng/ml (Maroto et al., 2004). Cells were treated with LMB for 10 h. Cdk1 (RO-3306) and caspase 3 (Caspase-3 Inhibitor Z-DEVD-FMK) inhibitors were obtained from SelleckChem (Munich, Germany) and Bio-Techne (Minneapolis, MN), and they were used at a final concentration of 10 µM and 50 μM, respectively. Cells were treated with the Cdk1 inhibitor for 7 h and with the caspase 3 inhibitor for 16 h. Rabbit Ab against CPV VP2 capsid protein and mouse monoclonal antibody (MAb) which recognizes the intact capsids (Weichert et al., 1998) was used at a concentration 4–10 μg/ml. The anti-NS1 MAb was a generous gift from Dr. Caroline Astell (University of British Columbia, Vancouver, Canada) (Yeung et al., 1991) (6.7 μg/ml). Full CPV capsids have their 5′-end of the DNA (Wang and Parrish, 1999) and N-terminal end of VP2 (Paradiso et al., 1982; Parrish et al., 1991)exposed outside the capsid. Here full capsids were detected with a rabbit antibody (rAb) against a peptide from the VP2 N-terminal domain (residues 2–19: CDGAVQPDGGQPAVRNER) (Casal et al., 1995) (2.7 μg/ml). Commercial antibodies obtained from Abcam (Cambridge, United Kingdom) against indicated proteins were: CRM1 (rabbit polyclonal Antibody, rAb, ab24189), γ-H2AX (recombinant Alexa Fluor^® 647 H2A.X anti-gamma phospho S139 rabbit monoclonal antibody, rMAb, ab195189 or anti-gamma H2A.X (phospho S139) antibody, rAb, ab2893), cyclin B1 (recombinant Alexa Fluor^® 555 Cyclin B1 rMAb, ab214381 or anti-Cyclin B1, ab2949), Cdk1 (recombinant Anti-Cdk1 rMAb, 133327)importin β (mMAb, KPNB1, ab2811), Ran (rAb, ab53774), and lamin B1 (anti- (rAb, ab16048). The primary antibodies were followed by goat anti-mouse or anti-rabbit Alexa448, Alexa546, Alexa 633, or Alexa 647 conjugated secondary Abs (Thermo Fisher Scientific, Waltham, MA). Commercial antibodies were used according to concentrations recommended by the manufacturer.

For immunofluorescence, NLFK and HeLa cells were cultured on glass coverslips and infected with CPV. Cells were fixed at 24 h post infection (hpi), and 30 hpi in 4% paraformaldehyde for 12 min and permeabilized with 0.1% Triton X-100 in phosphate-buffered saline supplemented with 1% bovine serum albumin. Immunolabeling with primary antibodies was followed by anti-mouse or anti-rabbit secondary antibodies. Nuclei were stained with Pro-Long Diamond anti-fade media with DAPI or Prolong Glass antifade mountant with NucBlue (Thermo Fisher Scientific). Imaging of immunolabeled samples was performed with Olympus FluoView FV1000 (Olympus, Tokyo, Japan), Nikon A1R (Nikon, Tokyo, Japan), and Leica TCS SP8 FALCON (Leica microsystems, Mannheim, Germany) laser scanning confocal microscopes. For Olympus microscopic parameters were the following: DAPI was excited with a 405 nm diode laser and a 450/50 nm band-pass filter was used to detect the fluorescence. Alexa 488 and Alexa 633 were excited with 488 nm argon and 633 nm He-Ne lasers, and the fluorescence was collected with a 500–555 nm slit-based emission filter and 647 nm long-pass filter, respectively. For Nikon, the following microscopic parameters were used. DAPI was excited with a 405 nm diode laser collected with a 513/30 nm band-pass filter. Alexa 488 was excited with a 488-nm argon laser, and fluorescence was collected with a 515/30 nm band-pass filter. Alexa 546 was exited with a 561 nm sapphire laser and detected with a 595/50 band-pass filter, and Alexa 633 was excited with a 642 nm diode laser and detected with a 660 nm long-pass filter. Microscopy images with Leica TCS SP8 FALCON were acquired as follows: DAPI and Nucblue were excited with a 405-nm diode laser and the fluorescence was monitored between 415–495 nm. Alexa 488, Alexa 546 or Alexa 555, and Alexa 647 were excited with 499 nm, 557 nm, and 653 nm wavelengths of pulsed white light laser (80 MHz). The emission detection range was 505–561 nm for Alexa 488 and 568–648 nm for Alexa 546/555 and 663–776 for Alexa 647. The image size varied from 512 × 512 to 1700 × 1700 pixels with a pixel size between 50–100 nm in the x- and y-directions. For the microscopy image stack, the step size of the images was between 120–333 nm in the z-direction. In the large field microscopy images used for the classification of infected cells, the image size was either 1024 × 1024 with a pixel size of approximately 210 nm or 4700 × 4700 with a pixel size of approximately 415 nm. For the images used for quantitative image analyses, the laser powers were fixed.

To analyze the nuclear and cytoplasmic intensities as well as the distributions of fluorescent labels concerning the nuclear border, the nuclei were segmented using Otsu’s method (Otsu, 1979) for the DNA stain channel. The part of the image that was not within the segmented nuclei was assumed to be within the cytoplasm. This causes the cell exterior to be included in the cytoplasm calculations, but it does not affect the calculated total intensities since the label signals were very weak outside cells. To calculate the intensity of labels concerning the nuclear border, the Euclidean distance to the nuclear border was determined for each pixel and the mean pixel intensity was calculated for ranges of increasing distances. In the analyses of cells with nuclear or cytoplasmic cyclin localization, the cells were grouped so that if the total cyclin signal was higher in the nucleus than in the cytoplasm the cell was classified as belonging to the nuclear cyclin group. Otherwise, it was classified as having cytoplasmic cyclin. In every analysis, the statistics were calculated over all the imaged cells. Student’s t-test, Games-Howell test, and chi-square test were used to evaluate the statistical significance.

NLFK cells stably expressing enhanced green fluorescent protein (EGFP) were used for fluorescence loss in photobleaching (FLIP) experiments. Non-infected and infected cells were incubated in the presence or absence of Cdk1 and caspase 3 inhibitors for 7 and 16 h, respectively, before the FLIP analysis at 24 hpi. The cells were imaged at 37°C and 5% CO2 with Nikon A1R laser scanning confocal microscope. Before the experiments, cells were treated for 15 min with 1 μg/ml Hoechst 33342 (Thermo Fisher Scientific) to visualize the nuclei of the cells and identify infected cells by prominent marginalization of host chromatin characteristic of CPV infection. A circular area of 3 µm in diameter in the cytoplasm was repeatedly photobleached with a 488 nm laser line using full laser power. Images of the whole cell and a neighboring reference cell were taken before and between the photobleaching pulses for 175 s. To analyze NE permeability to EGFP, fluorescence intensities from the nucleus of the photobleached cell and the reference cell were extracted. The values were normalized to pre-bleach intensities and corrected for the loss of fluorescence resulting from the imaging procedure with fluorescence intensities from the reference cell. Student’s t-test was used to analyze the statistical significance of the differences in the averaged nuclear fluorescence half-lives between samples.

ChIP-seq and ChIP-qPCR were performed according to Mäntylä et al. (Mäntylä et al., 2016). ChIP-seq sequence reads were re-aligned with the updated cat reference genome (Felis_catus 9). Analysis was limited to enrichment in regions surrounding transcription start sites (TSS, TSS between -2000 and 1,500 bp). GO biological process term enrichment analyses were performed with http://geneontology.org/. The ChIP-seq data reported in this paper have been previously deposited in the NCBI Gene Expression Omnibus (GEO) database under accession number GSE77785.

As we have described earlier, the CRM1-mediated nuclear export does not fully explain the nuclear egress of all CPV progeny capsids (Mattola et al., 2022). Here, we analyzed the nuclear exit of full (DNA-containing) capsids in the presence of CRM1 inhibitor LMB. The full capsids with exposed N-terminal end of VP2 (Paradiso et al., 1982; Parrish et al., 1991) were detected with an antibody against the VP2 N-terminal domain.

Our confocal microscopy imaging demonstrated that the nuclear egress of full CPV capsids into the cytoplasm was slightly decreased at 24 hpi in the presence of LMB in comparison to the untreated infected NLFK cells (Figure 1A). Quantitative analysis of the cytoplasmic-to-nucleus ratio showed a moderate but non-significant LMB-induced reduction of cytoplasmic capsids at 24 hpi in comparison to non-treated cells, and a significant decrease at 30 hpi (Figure 1B). However, at both time points, nuclear escape of capsids was detected also in the presence of LMB. The analysis of the cellular distribution of viral capsids in untreated and LMB-treated infected cells at 24 and 30 hpi indicated that the number of cytoplasmic capsids was the highest close to the NE and decreased towards the plasma membrane. In the nucleus, the majority of capsids accumulated close to the NE in the absence and presence of LMB both at 24 and 30 hpi. However, in the LMB-treated cells, specifically at 30 hpi, the capsids were more closely associated with the NE than in non-treated cells. Full capsids were not detected in infected cells earlier at 14 hpi.

Altogether, the LMB treatment cannot fully inhibit the nuclear egress of capsids, thereby suggesting that CRM1-independent pathways are involved in the nuclear egress of full capsids.

The nuclear egress of capsids could be explained by a partial loss of the nuclear permeability barrier. In non-infected cells, NE permeability is increased at the early stages of mitosis (Terasaki et al., 2001; Beaudouin et al., 2002; Burke and Ellenberg, 2002) in apoptosis (Coleman and Olson, 2002; Kopeina et al., 2018; Lindenboim et al., 2020), and during nuclear envelope ruptures (Denais et al., 2016). In mitosis, NE breakdown is regulated by kinases such as Cdk1 (Santamaría et al., 2007). During the apoptotic processes, caspases promote an increase in NE permeability by cleaving nuclear pore complexes and the nuclear lamina Kopeina et al., 2018; Lindenboim et al., 2020). Moreover, Cdk1 is also linked to the regulation of apoptosis (Castedo et al., 2002) and caspase 3 has a potential role in mitosis (Swe and Sit, 2000; Hsu et al., 2006; Hashimoto et al., 2011; Lee et al., 2011).

To study further the possible association between capsid egress and NE leakage, we investigated NE permeability at a late stage of infection (24 hpi). We used FLIP to determine the transport rate of freely diffusible EGFP through the NE (Hoffmann-rohrer et al., 2003; Lippincott-Schwartz et al., 2003). The results show that the repeated cytoplasmic EGFP bleaching led to a faster depletion of nuclear fluorescence of infected cells at 24 hpi in comparison to non-infected cells (Figures 2A,B). The fluorescence half-life in the nucleus was significantly (p < 0.001) decreased from 77 ± 4 s to 39 ± 3 s (mean ± the standard error of the mean). This implies that CPV infection is accompanied by an increase in NE permeability. We next studied the potential role played by the key effector proteins of mitosis and apoptosis, Cdk1 and caspase 3, using FLIP analysis for NE leakiness in the presence of Cdk1 and caspase 3 inhibitors. The analysis showed that the infection-induced increase in NE permeability was partially reversed upon exposure to the inhibitors (Figure 2B), leading to significantly increased fluorescence half-life in the nucleus from 39 ± 3 s to 55 ± 5 s (p < 0.01) and 66 ± 6 s (p < 0.001) for Cdk1 inhibitor and caspase 3 inhibitor, respectively. Notably, inhibition of Cdk1 or caspase 3 failed to fully restore NE permeability to the level of non-infected cells. This suggests that neither mitosis nor apoptosis alone is responsible for NE leakage. In non-infected cells treated with inhibitors, NE permeability remained similar to the untreated cells (Supplementary Figure S1).

Together, our results suggest that Cdk1- and caspase 3 -mediated mitotic and apoptotic pathways may affect NE permeability in infection, and thereby most likely influence the nuclear egress of progeny capsids.

As parvovirus infection progresses, it induces cellular DNA damage and evokes the DDR (Adeyemi et al., 2010; Luo et al., 2011a; Luo et al., 2011b). The localization of DDR can be detected by staining the phosphorylated histone variant, γ-H2AX, formed from H2AX in a response to double-strand break formation (Kinner et al., 2008). DNA damage induces chromatin decondensation, which leads to an enhancement of nuclear molecular diffusion (dos Santos et al., 2021). Here, we examined the effect of DNA damage-induced chromatin changes on the nuclear egress of CPV capsids.

Earlier MVM and CPV studies have demonstrated that at late times of infection the pattern of capsid localization progresses through characteristic stages. These stages include localization of progeny capsids in the nuclear replication center, accumulation of capsids to the nuclear periphery sometimes accompanied by the presence of cytoplasmic capsids, and egress phase with major cytoplasmic localization of egressed capsids. The intracellular distribution of capsids varies between individual cells, thereby reflecting cell-specific variation in the progression of infection (Ruiz et al., 2011; Mihaylov et al., 2014; Mäntylä et al., 2016). To analyze the nuclear distribution of γ-H2AX at 24 hpi, we divided infected cells into two classes: a replication center class showing capsid accumulation into the central nuclear region and an egress class showing capsid localization both in the nuclear periphery and in the cytoplasm.

Confocal images showed that in the replication center class γ-H2AX was mostly distributed homogenously throughout the nucleus with a nucleolar exclusion, whereas in the egress class γ-H2AX was concentrated close to the NE and around the nucleoli (Figure 3A). Host cell chromatin was distributed both near the NE and nucleoli (Figs 3A and S2). Distance analyses showed that γ-H2AX was distributed throughout the nucleoplasm both in cells with nuclear accumulation of capsids and in cells with egressed capsids, however, in the egress class cells there was a clear increase in γ-H2AX intensity toward the NE (Figure 3B). It is important to note that an increase in the cytosolic amount of capsid label (Figure 3C) correlated with the perinuclear localization of damage induces chromatin decondensation γ-H2AX. As shown for herpesvirus, transport through the host chromatin can be a rate-limiting step in the nuclear egress of capsids (Aho et al., 2021). Our results imply that cellular DNA damage and DDR-related events could enhance the nuclear exit of CPV capsids.

In non-infected cells, the G2/M activation is accompanied by cytoplasmic interaction of cyclin B1 and Cdk1, which leads to nuclear translocation of the cyclin B1-Cdk1 complex (Hagting et al., 1999; Yang et al., 2001; Porter and Donoghue, 2003; Miyazaki and Arai, 2007). Nuclear accumulation of cyclin B1 is also detected in DNA damage-induced apoptosis at the G2/M checkpoint (Porter et al., 2003). In parvovirus-infected cells, virus-induced DDR results in cytoplasmic retention of cyclins during G2 arrest (Morita et al., 2001). Notably, in CPV infection the number of cells in the G2/M phase is relatively high at the late stages of infection (Nykky et al., 2010). To study the possible involvement of the G2/M checkpoint activation in infected cells, we analyzed the intracellular distribution of cyclin B1. Specifically, we were interested in observing the nuclear localization of cyclin B1 in cells with infection-induced DDR.

Our confocal images showed that at 24 hpi the main localization of cyclin B was either in the nucleus or in the cytoplasm. Approximately 30% (n = 9) of CPV-infected NLFK cells (n = 26) showed nuclear localization of cyclin B1. Viral infection was confirmed by the viral NS1 protein staining and by marginalized host cell chromatin, and DDR activation was detected by γH2A.X staining (Figure 4A). Quantitative analysis showed that in the presence of nuclear cyclin B1 the amount of viral NS1 was significantly increased in comparison to cells with cytoplasmic cyclin B1 (Figure 4B), whereas the amounts of γH2A.X in the presence of nuclear or cytoplasmic cyclin B1 were relatively similar (Figure 4C). Comparison between distributions of chromatin and γH2A.X as a function of the distance from the NE demonstrated that both in cells with cytoplasmic or nuclear cyclin B1, both chromatin (Fig S3A) and γH2A.X were accumulated close to the NE (Figures 4D,E), however, the amount of γH2A.X was increased in the presence of nuclear cyclin B1 (Fig S3B). Further distance analyses demonstrated that in cells with cytoplasmic cyclin B1 NS1 was localized in the nucleoplasmic area surrounded by the perinuclear γH2A.X (Figure 4D). In contrast, in cells with nuclear cyclin B1, NS1 was concentrated towards the γH2A.X layer (Figure 4E). Finally, the distribution analyses of cyclin B1 and NS1 demonstrated that in cells with mainly a cytoplasmic localization of cyclin B1, there was a nuclear fraction of cyclin B1 accumulated close to the NE, whereas NS1 was mostly localized into the center of the nucleus (Figure 4F). In cells with nuclear accumulation of cyclin B1, it was distributed throughout the nucleoplasm, however, with a slight increase toward the center of the nucleus. At the same time, NS1 remained in the nucleoplasm close to the perinuclear γH2A.X (Figure 4G).

Together, our experiments demonstrated that at 24 hpi infected cells showed differential localization of cyclin B1, both in the nucleus and in the cytoplasm. One-third of the infected cells had a nuclear accumulation of cyclin B1. These cells showed an increase in the amount of NS1, which suggests that these two phenomena might be connected. Nuclear localization of cyclin B1 implies that the infection-imposed G2/M block, represented by a cytoplasmic localization of cyclin B1, is not fully maintained at the late stages of infection. Alternatively, nuclear accumulation of cyclin B1 could be an indicator of cyclin B1–dependent apoptosis (Porter et al., 2003).

In non-infected interphase cells, cyclin B1 shuttles between the nucleus and the cytoplasm, and it is accumulated into the nucleus only after activation of the G2/M checkpoint (Lindqvist et al., 2007; Gavet and Pines, 2010; Müllers et al., 2014). Nuclear accumulation of cyclin B1 is also a key regulator in the cellular decision to undergo apoptosis in response to DNA damage (Porter and Donoghue, 2003).

Here, we investigated the intracellular localization of capsids in NLFK cells together with cyclin B1. Nuclear retention of cyclin B1 (Figure 5A) in infected cells was further confirmed by total fluorescence intensity analysis. Infected cells showed a significant increase in the amount of nuclear cyclin B1 at 24 hpi compared to non-infected cells. Moreover, a significant increase in the amount of nuclear cyclin B1 was detected at 30 hpi (Figure 5B). This was consistent with the experiments described in Figure 4 and supports the model that upon infection a part of the infected cells enter mitosis. Nuclear amounts of nuclear capsid label did not change between 24 and 30 hpi, which suggests that egressing capsids were balanced by increased production of progeny viral capsids as verified in Figure 5C. Increasing nuclear cyclin B1 retention at 30 hpi was accompanied by capsid accumulation at the nuclear periphery (Figure 5A) and significant acceleration of capsid egress into the cytoplasm in comparison to 24 hpi (Figure 5C). Distribution analysis of capsids and cyclin B1 revealed that at 24 hpi capsids were more clearly retained in the nucleus whereas at 30 hpi the capsid distribution was more even throughout the nucleus and the cytoplasm. Cyclin B1 was mainly localized into the nucleus in infected cells with a small variation in distribution (Figure 5D).

Taken together, capsids escape from the nucleus in late infection when cyclin B1 is accumulated in the nucleus. The nuclear accumulation of cyclin B1 supports our overall model suggesting that G2/M checkpoint activation and associated apoptotic events enhance viral capsid egress out of the nucleus.

Nuclear accumulation of cyclin B1 during the G2/M checkpoint activation is accompanied by nuclear localization of Cdk1(Gavet and Pines, 2010; Morgan, 2007; Schmitt et al., 2007). To further investigate the association between Cdk1 distribution and nuclear egress of capsids, we analyzed infected NLFK cells at 24 hpi. We observed that the amount of nuclear Cdk1 was increased in infected cells at the same time as the nuclear egress of viral capsids (Figure 6A). Quantitative analyses of Cdk1 distribution confirmed the infection-induced significant nuclear accumulation of Cdk1 at late infection (Figure 6B). We next sought to analyze how the inhibition of Cdk1 affects the nuclear egress of capsids. In these analyses, we used a classification based on capsid localization into cytoplasmic vesicles (capsids in endosomal vesicles), replication center (capsids located in the nucleus), nuclear periphery (capsids in the nucleoplasm and cytoplasm), and cytoplasm (empty nucleus after capsid egress) (Figure 6C). The intracellular distribution of capsids was monitored in the presence of a selective Cdk1 inhibitor RO3306, which reversibly arrests human cells at the G₂/M border (Vassilev, 2006). Our findings showed that the inhibition of Cdk1 led to a clear increase in the amount of capsid label retained in the nucleus, and to a slight increase in capsid label found near the NE. Importantly, a significant decrease in the amount of capsid label translocated into the cytoplasm was detected (Figure 6D).

Altogether, this indicated that Cdk1 is located in the nucleus in infected cells and Cdk1 activity has a role in the nuclear egress of CPV capsids. The strong correlation between Cdk1 inhibition and a decrease in the cytoplasmic amounts of capsid label is consistent with our results above showing that G2/M checkpoint-mediated NE leakage is required for the nuclear egress of capsids in late infection.

Early apoptosis leads to nucleoporin cleavage, an increase in passive diffusion through nuclear pore complexes, and nucleocytoplasmic relocalization of active nuclear transport factors, importin α, importin β, and Ran, between the nucleus and the cytoplasm. The cytoplasmic distribution of importin α and β, the nucleoplasmic localization of Ran in non-apoptotic interphase cells are converted to nuclear importins and cytoplasmic Ran in apoptotic cells (Ferrando-May et al., 2001; Ferrando-May, 2005; Wilde and Zheng, 2009). CPV-induced DNA damage modulates cell cycle progression and induces apoptosis (Luo and Qiu, 2013; Zhao et al., 2016). Apoptosis is known as the major cell death pathway at the final stages of CPV infection (Gupta et al., 2016; Nykky et al., 2010), however, its role in the nuclear egress of capsids remains to be determined.

As a verification of CPV-induced apoptosis, we analyzed the nuclear distribution of transport factors. Nuclear localization of importin β was used as an infection marker since visualization of infected HeLa cells demonstrated that both importin β and the viral capsid protein VP2 were accumulated into the nucleus at 24 hpi. In contrast, importin β showed mostly perinuclear and cytoplasmic localization in non-infected cells (Fig S4). Moreover, importin α (importin-α-GFP) and importin β was accumulated in the nucleus at 24 hpi in NLFK cells (Figure 7A). Ran was localized both to the nucleus and to the cytoplasm in infected cells, in contrast to the exclusively nuclear localization of Ran in non-infected cells (Figure 7B). Quantitative analyses of nucleus-to-cytoplasmic ratios in infected and non-infected cells confirmed the nuclear accumulation of importins and cytoplasmic localization of Ran in infected cells. The nucleus-to-cytoplasmic ratio for importin α in infected and non-infected cells were 4.20 and 2.04, and for importin β 2.84 and 0.64 (Figure 7C). At the same time, the amount of cytoplasmic Ran was increased with ratios of 3.64 in non-infected cells in comparison to 1.40 in infected samples (Figure 7D).

Taken together, late infection is characterized by an apoptotic alteration of nucleocytoplasmic localization of nuclear transporters, importins, and Ran, suggesting that the apoptotic events are induced in infected cells.

Finally, to study the effects of G2/M checkpoint-related apoptotic events for nuclear egress of capsids, we applied inhibition of caspases in late-infection. Caspases are cysteine-dependent proteases that are actively involved in the execution of apoptosis (McIlwain et al., 2013). One of the downstream caspases, caspase 3, is activated in early apoptosis (Varghese et al., 2003). It is essential for apoptotic events such as the formation of apoptotic bodies, chromatin condensation, and DNA fragmentation (Jänicke et al., 1998; Woo et al., 1998). Caspase 3 has also a role in the regulation of mitosis (Swe and Sit, 2000; Hsu et al., 2006; Hashimoto et al., 2011; Lee et al., 2011). Notably, caspase 3, which is also involved in the induction of NE breakage, has a role in the nuclear import of MVM (Cohen et al., 2011).

To characterize the importance of early apoptosis for the nuclear exit of viral capsids, we treated infected cells with a caspase 3 inhibitor (Z-DEVD-FMK), which is a peptide that prevents the induction of early apoptosis (Garcia-Calvo et al., 1998). Figure 8A shows that the amount of cytoplasmic capsid label was visibly decreased in the presence of caspase 3 inhibitor in comparison to untreated NLFK cells at 24 hpi. The image data analysis using capsid localization classes described above (Figure 6C) confirmed that the nuclear egress of capsids was significantly decreased during caspase 3 inhibition. Simultaneously, capsid retention close to the NE was increased. Interestingly, the early steps of infection shown by capsid localization into cytoplasmic vesicles were reduced in the presence of the inhibitor (Figure 8B). This implies that the inhibition of early apoptosis or mitosis does not only affect the nuclear egress of capsid but might also have an influence on capsid entry.

Together, these results suggest that caspase 3 and early apoptosis have a potential role in enhancing the nuclear egress of viral capsids. However, it remains possible that caspase 3 might also affect the mitotic facilitation of capsid egress.

Our findings suggested that CPV egress is enhanced by the promotion of G2/M checkpoint activation and regulation of apoptosis. To follow up on these observations and to analyze specific virus-induced changes in the epigenetic landscape of cell cycle and death, we surveyed the histone 3 lysine 27 acetylation (H3K27ac) by ChIP–seq. H3K27ac marks enhancer-specific modifications required for the activation of target gene transcription (Heintzman et al., 2009, 2007; Wang et al., 2008).

Analysis of the host genome indicated that the H3K27ac signal was increased in infection (24 hpi) in comparison to non-infected NLFK cells. In total, the H3K27ac profile contained 43512 peaks in control cells compared to 54174 in CPV-infected cells (24% increase after infection). Gene Ontology (GO) enrichment analysis of the H3K27ac ChIP-seq demonstrated either gain or loss of acetylation in mitotic and apoptotic categories in comparison to non-infected cells (Figure 8A). Significantly enriched categories with H2K27ac loss were regulation of cell death (GO:0010941), regulation of apoptotic process (GO:004281), and regulation of programmed cell death (GO:0043067), whereas GO categories significantly enriched with acetylation gain were cell cycle process (GO:0022402), cell cycle (GO:0007049) and cellular response to DNA damage stimulus (GO:0006974) (Table S1). Further analyses of acetylation in infected cells showed that H3K27ac was enriched significantly in genes related to mitosis signaling pathways including several GO classes such as regulation of mitotic cell cycle (GO:0007346), mitotic cell cycle (GO:0000278), mitotic cell cycle process (GO:1903047), and regulation of mitotic cell cycle phase transition (GO:1901990) in infected samples. The infection led also to a significant increase in acetylation of transcription start sites region related to apoptosis. These signaling pathways pooled from GO categories were negative regulation of apoptotic process (GO:0043066), positive regulation of apoptotic process (GO:0043065), apoptotic process (GO:0006915), and apoptosis (GO:0042981). In infected cells, the regions with high H3K27ac included 205 mitotic and 251 apoptotic genes and 44 genes which belonged to both categories, whereas in non-infected cells 27 mitotic, 68 apoptotic, and 11 shared genes were acetylated. Moreover, six mitotic, 10 apoptotic, and two shared genes were activated both in the infected and non-infected cells (Figure 9B). Venn´s analysis of ChIPseq-enriched activated genes unique to the late infection indicated a total of 42 genes associated with apoptotic and mitosis signaling (Figure 9B). These 42 genes were analyzed further by using STRING (https://string-db.org/) online analysis tool for identifying protein-protein interaction networks. Next, gene ontology (GO) enrichment analysis for biological processes (BP) was performed. The BP enrichment analysis revealed enrichment of e.g., cell cycle checkpoint, signal transduction involved in G2/M transition checkpoint, signal transduction involved in cell cycle checkpoint, G1/S checkpoint, G1/S transition checkpoint, G2/M checkpoint, G2/M transition checkpoint (GO:0000075); processes participating mitotic cell cycle (GO:1903047); DNA integrity checkpoint signaling (GO:0031570); DNA damage checkpoint signaling (GO:0000077); negative regulation of cell cycle G2/M phase transition (GO:1902750); regulation of cell cycle G2/M phase transition (GO:1902749); and signal transduction in response to DNA damage (GO:0042770). In addition, mitotic events such as mitotic DNA integrity checkpoint signaling (GO:0044774), mitotic DNA damage checkpoint signaling (GO:0044773), intrinsic apoptotic signaling pathway in response to DNA damage (GO:0008630), negative regulation of G2/M transition of the mitotic cell cycle (GO:0010972) were enriched. The analysis had a false discovery rate i.e., p-value p < 0.01 corrected for multiple testing within each category using the Benjamini–Hochberg procedure (Table 2).

From the associated GO terms, the top 15 genes that appeared to be involved in most of these pathways were recognized (Figure 9C, Table S 2, and Table S 2). Interestingly, among these was the Chk1/Chk2(Cds1) pathway, which is known to be activated in response to DNA damage resulting in the conversion and/or maintenance of cyclin B1:Cdk1(Cdc2) complex in its Tyrosine 15 phosphorylated (inactive) state. Activation of Chk1 can be done via ATM, which was also included in the analysis results. Specifically, the dsDNA breaks are considered to be processed in an ATM-dependent manner causing primarily ATR and hence Chk1 activation. Based on our findings, it is thus tempting to speculate that the DNA damage-induced G2/M checkpoint activation in CPV infection might occur via ATM-dependent cyclin B1/Cdk1 inhibition in a Chk1-mediated manner. However, as signaling pathways were not in the scope of this study, this needs to be further investigated in the future.

Taken together, these results support the findings that DNA-damage-induced G2/M checkpoint regulation takes place in late infection. Our ChIP-seq findings are consistent with the model that mitosis and apoptosis are regulated at late infection. Both of these pathways lead to NE leakiness, thereby potentially contributing to the enhancement of the nuclear egress of viral capsids.