Localization of recombinant UL13 and UL13-△NLSs mutated protein in transfected DEFs
We have proved that UL13 protein of DEV also can enter into nucleus. PSORT II Prediction revealed amino acids 4 to 7 (Arg-Arg-Arg-Arg) and 90 to 96 (Pro-Gly-Lthoys-Arg-Lys-Thr-Lys), as basic amino acid patches, as putative NLSs (Figure S1). To examine the function of these predicted NLSs, we constructed mammalian expression plasmids for UL13-GFP fusion proteins or proteins with NLS1 or/and NLS2 deleted (Fig. 1A) and transfected them into DEFs. Compared with that the fluorescence of UL13-GFP fusion proteins located both in nucleus and cytoplasm (Fig. 1B b, g, l), when NLS1 or NLS2 was deleted, the majority of the fluorescence displayed a cytoplasmic distribution in DEFs (Fig. 1B c, h, m, d, i, n). Furthermore, the UL13-GFP fusion protein, with NLS1 and NLS2 both deleted, exhibited a completely cytoplasmic distribution (Fig. 1B e, j, o). We also analyse the nuclear compartmentalization of UL13/UL13-△NLSs. As shown in Fig. 1C, the nuclear/cytoplasmic fluorescence ratio of the UL13-△NLS1&NLS2 protein was significantly decreased compared to the UL13 protein (P༜0.01), and both the UL13-△NLS1 and UL13-△NLS2 proteins were also significantly decreased (P༜0.001). The decrease in the nuclear/cytoplasmic fluorescence ratio was higher for UL13-△NLS1&NLS2 than for UL13-△NLS1 and UL13-△NLS2. All these results suggested that both NLS1 and NLS2 are responsible for the nuclear localization of DEV UL13 and work together to enhance this functionality. We also separated the cytoplasmic and nuclear proteins expressed from pEGFP-N1-UL13 or pEGFP-N1-UL13-△NLS1&△NLS2 plasmids in DEF cells, and analysed the quantity of UL13 protein distributed in the cytoplasm and nucleus via western blotting (Fig. 1D). Compared with the distribution of the UL13 protein expressed in the pEGFP-N1-UL13 group, the ratio of UL13 protein in the nucleus in the pEGFP-N1-UL13-△NLS1 &△NLS2 group was significantly reduced. We used β-actin and Lamin A/C as controls for these cytoplasmic/nuclear fractions.
Localization of the NLS-GFP fusion protein in transfected DEFs
To further examine whether NLS1 and NLS2 can import heterologous proteins into the nucleus, sequences encoding NLS1 and NLS2 were fused to GFP to obtain NLS1-GFP, NLS2-GFP, and NLS1&NLS2-GFP recombinant plasmids and transfected into DEFs (Figure 2A). Fluorescence assays showed a greater proportion of the NLS1-GFP and NLS2-GFP fusion proteins in the nucleus than in the cytoplasm (Figure 2B b, f, j, c, g, k); in contrast, NLS1&NLS2-GFP showed a predominantly nuclear distribution (Figure 2B d, h, l), and GFP alone showed a homogeneous distribution (Figure 2B a, e, i). We also analysed differences in the nuclear and cytoplasmic fluorescence ratios between GFP and NLSs-GFP and found that the nuclear/cytoplasmic fluorescence ratios for NLS1-GFP, NLS2-GFP and NLS1&NLS2-GFP were significantly increased compared to GFP (P<0.001) (Figure 2C), with that of NLS1&NLS2-GFP being particularly increased. We also separated the cytoplasmic and nuclear proteins expressed from NLS1-GFP or NLS2-GFP plasmids in DEF cells, and analysed the quantity of UL13 protein distributed in the cytoplasm and nucleus via western blotting (Figure 2D). Compared with the distribution of the UL13 protein expressed in the nucleus, the ratio of UL13 protein in the cytoplasm in the NLS1-GFP or NLS2-GFP group was significantly reduced.
Our results confirmed that both NLS1 and NLS2 of DEV UL13 act as nuclear localization signals. Each could mediate protein entry into the nucleus, and their combination increased the function of the nuclear localization signal.
The effects of nuclear import inhibitors on the location of DEV UL13
NLS1 and NLS2 of DEV UL13 are predicted to be pat4 and pat7 monopartite prototypical nuclear localization signals, respectively. It has been reported that ivermectin blocks importin α and β interaction and inhibits the nuclear transport facilitated by the prototypical NLS-mediated mechanism, with no effect on proteins containing NLSs recognized by alternative nuclear import pathways or other nuclear import pathways. Therefore, we tested whether the nuclear import of the DEV UL13 protein requires the interaction of importin α and β by treating DEF cells transfected with pEGFP-N1-UL13 recombinant plasmids with/without ivermectin. As shown in Figure 3A, less nuclear UL13 protein was found in DEFs treated with ivermectin than DEFs without ivermectin treatment at 16h post-transfection (Figure 3A a, g, m, b, h, n). The nuclear distribution of NLS1&NLS2-GFP with ivermectin treatment was also reduced compared to the no treatment group (Figure 3A c, i, o, d, j, p), whereas there was no change in the distribution of GFP alone with/without ivermectin treatment (Figure 3A e, k, q, f, l, r). We also analysed differences in the nuclear and cytoplasmic fluorescence ratio of NLS1&NLS2-GFP and UL13 in DEF cells with or without ivermectin treatment and found that the ratio of UL13 declined significantly(P<0.001), with the mean ratio of the NLS1&NLS2-GFP group reduced from 0.8 to 0.4 after ivermectin treatment. Statistical analysis showed that the nuclear accumulation of NLS1&NLS2-GFP and UL13 was significantly impaired by ivermectin (Figure 3B) (P<0.01). These results suggested that the nuclear import of DEV UL13 occurs by the classical transport process that requires the interaction of importin α and β and is mediated by the NLSs of DEV UL13.
Entry nucleus of UL13 protein has no effect on DEV replication in cell culture
To examine the effect of entry nucleus of UL13 protein on viral proliferation, we constructed DEV CHv-UL13ΔNLS, which was a mutant of both NLS1 and NLS2 of DEV UL13 being deleted, and DEV CHv-UL13ΔNLS R (a revertant) (Figure 4A). By RFLP analysis and sequencing, we determined that the DEV CHv-UL13ΔNLS and DEV CHv-UL13ΔNLS R recombinant viruses were mutated only at their appropriate target site (Figure 4B).
A plaque assay indicated that the DEV CHv-UL13ΔNLS could grow in DEF, and the difference of plaque size between DEV CHv-UL13ΔNLS and DEV CHv has no significance (Figure 5A). The growth curve of DEV CHv-UL13ΔNLS exhibited the same trend of DEV CHv in both the supernatant and cells during the infection period (Figure 5B). The phenotype of DEV CHv-ΔUL13R exhibited similar