Zika Virus Induces Degradation of the Numb Protein Required through Embryonic Neurogenesis

Abstract

Zika virus (ZIKV) is a mosquito-borne flavivirus and causes an infection associated with congenital Zika syndrome and Guillain-Barre syndrome. The mechanism of ZIKV-mediated neuropathogenesis is not well understood. In this study, we discovered that ZIKV induces degradation of the Numb protein, which plays a crucial role in neurogenesis by allowing asymmetric cell division during embryonic development. Our data show that ZIKV reduced the Numb protein level in a time- and dose-dependent manner. However, ZIKV infection appears to have minimal effect on the Numb transcript. Treatment of ZIKV-infected cells with a proteasome inhibitor restores the Numb protein level, which suggests the involvement of the ubiquitin-proteasome pathway. In addition, ZIKV infection shortens the half-life of the Numb protein. Among the ZIKV proteins, the capsid protein significantly reduces the Numb protein level. Immunoprecipitation of the Numb protein co-precipitates the capsid protein, indicating the interaction between these two proteins. These results provide insights into the ZIKV-cell interaction that might contribute to its impact on neurogenesis.

Keywords: ZIKV; Zika virus; the Numb protein; the capsid protein.

Figure 1

ZIKV infection reduces the Numb protein level. (A) Infection of Vero cells with ZIKV PRVABC59 (PR) strain leads to a lower Numb protein level. The cells were inoculated with an MOI of 10 of the ZIKV PR strain and harvested for Western blotting 40 h post-infection (hpi). Relative levels of Numb are shown below the images after normalization with tubulin. (B) Infection with ZIKV MR766 strain also induces a reduction in the Numb protein level. Vero cells were infected at an MOI of 10 and harvested at 40 hpi for WB. Relative levels of the Numb protein are shown below the images after normalization with GAPDH. (C) Infection of HeLa cells with ZIKV PR strain reduces the Numb protein. The cells were inoculated at an MOI of 10 and harvested for WB 36 hpi. (D) Numb protein level decreased in SK-N-SH cells with ZIKV infection. The cells were inoculated at an MOI of 10 and harvested for WB 72 hpi. Because of the variable progress in ZIKV proliferation in the different cell lines, the cells were harvested at different time points.

Figure 2

ZIKV reduces the Numb protein in a temporal and dose-dependent manner. (A) Temporal reduction in Numb by ZIKV infection. The cells were inoculated with ZIKV PR strain at an MOI of 1 and harvested for WB 20 and 40 hpi. Relative levels of the Numb protein are shown below the images. (B) Dose-dependent reduction in the Numb protein level by ZIKV PR strain. The cells were harvested for WB 24 hpi. Relative levels of the Numb protein are shown below the images. (C) ZIKV RNA levels in the infected cells detected by RT-qPCR. Error bars represent the standard errors of the means of three repeated experiments. Asterisks denote significant differences in RNA level from an MOI of 0.1 (** p < 0.01; *** p < 0.001).

Figure 3

Numb was decreased by ZIKV infection through the ubiquitin–proteasome pathway. (A) ZIKV infection has a minimum effect on Numb mRNA level detected by RT-qPCR. The relative Numb mRNA level is shown compared with the mock-infected Vero cells. The cells were infected with the ZIKV PR strain at an MOI of 1 and harvested at 30 hpi. (B) ZIKV RNA level in the infected cells. A standard curve was included for calculating the absolute RNA level. Total ZIKV RNA is shown as copies in Log10 per well of a 24-well plate. (C) MG132 treatment restores the Numb protein levels in ZIKV-infected cells. Vero cells were infected with ZIKV at an MOI of 1 and 30 hpi and treated with MG132 or NH4Cl at a final concentration of 10 µM and 10 mM, respectively, for 6 h. The solvent DMSO was included as a control in the first two lanes. (D) ZIKV infection shortens the Numb protein half-life. The addition of cycloheximide to the Vero cells infected with ZIKV at an MOI of 10 for 24 h and treated with MG132 for 6 h. Mock-infected cells were treated similarly and included for control. (E) ZIKV infection increases the level of Numb protein polyubiquitination. ZIKV-infected Vero cells were harvested for immunoprecipitation with Numb antibody, followed by Western blotting with antibodies against ubiquitin (Ub) and Numb. Western blotting of the IP input of the cell lysate was also performed as a control. Representative data of three independent experiments are shown.

Figure 4

ZIKV capsid protein induces Numb reduction. (A) Screening of the ZIKV proteins for inducing Numb reduction. Co-transfection of HEK293 cells was performed with the plasmids encoding the individual ZIKV proteins and MYC-Numb. At 48 h post-transfection, the cells were lysed for WB analysis. Relative Numb levels are indicated below the images. E.V.: empty vector. (B) ZIKV capsid is present in the Numb co-IP precipitates. The IP input was included in WB for control. (C) The capsid protein is present in the MYC co-IP precipitates. HEK293 cells were co-transfected with MYC-Numb and YFP-C plasmids. E.V.: empty vector. Representative data of three independent experiments are shown.

Figure 5

Numb knockdown has minimal effect on ZIKV replication. (A) Numb knockdown in Vero cells has no significant effect on ZIKV replication compared with the control shRNA. The cells were transduced with the recombinant retrovirus containing control shRNA (C-shRNA) or shRNA against Numb (shNumb) three times, followed by inoculation with ZIKV at an MOI of 1, and harvested at 48 hpi. (B) ZIKV titers from infected Vero cells at different time points. Vero cells were infected by ZIKV at an MOI of 0.01 and harvested at different time points. (C) Numb knockdown has minimal effect on cell viability. NS: no significant difference.