br Materials Methods br Results

Materials & Methods
Results To better understand how ZIKV infection may lead to microcephaly, we developed an assay to infect cultured mouse embryonic NHS-SS-Biotin manufacturer slices and tested the ability of ZIKV and other flaviviruses to infect the neocortex (Fig. 1a). We selected an isolate of DENV serotype 4 (DENV-4), responsible for a lethal fulminant hepatitis (Kudelko et al., 2012) and the encephalitic West Nile virus (WNV) Israel 1998 strain (IS98) (Lucas et al., 2004). Importantly, ZIKV infection was performed using the microcephaly-associated French Polynesian 2013 strain (Pf13) which is over 97% identical at the nucleotide level to the Brazilian strain but phylogenetically more distant from the African lineage (Zhu et al., 2016) (Fig. 1b). E15 mouse embryonic brains were sliced and cultured in modified cortical culture medium (M-CCM) (Baffet et al., 2016) and brain slices were infected for 2h with either virus produced on C6/36 mosquito cells (6.105 FFU). Viral input was then removed and brain slices cultivated for 24 to 48h at 37°C (Fig. 1a). Viral titration revealed productive infection of ZIKV and WNV after 24 and 48h, but not of DENV-4, indicating that the mouse developing neocortex is able to support replication of these two viruses (Fig. 1c). Immuno-staining experiments confirmed that ZIKV and WNV efficiently infect E15 brain slices whereas no cells infected by DENV-4 were detected (Fig. 1d). Interestingly ZIKV and WNV showed very different distributions. While WNV was distributed throughout the developing neocortex with enrichment in the intermediate zone (IZ) and cortical plate (CP), ZIKV was strongly concentrated in the ventricular zone (VZ) (Fig. 1d). Moreover, ZIKV-infected cells exhibited a characteristic RGP cell morphology, with an apical and a basal process running all the way up to the pial surface (Fig. 1e, arrowheads mark basal processes). Viral factories could be observed, most often located in the perinuclear region (Fig. 1e, arrows). To confirm this differential tropism of infection between ZIKV and WNV, we performed cell fate analysis. Quantification of Pax6+ infected cells revealed that the vast majority of ZIKV-infected cells were indeed RGP cells (84.9%±7.6%, n=2612 cells) (Fig. 2a, c). Conversely, only 29.7%±4.8% of WNV-infected cells were Pax6+ RGP cells (n=4370 cells) (Fig. 2a, c). Consistent with these results, only 10.76%±1.67% of ZIKV-infected cells were NeuN+ neurons (n=3281 cells), versus 74.6%±5.6% of WNV-infected cells (n=4287 cells) (Fig. 2b, d). At this stage the ratio between these two cell types was 1.69±0.2 neuron for 1 RGP cell (Fig. 2e), indicating that the strong ZIKV tropism of infection for RGP cells was not merely due to their higher abundance. Next we asked if ZIKV infection affected proliferation and survival of RGP cells. Immuno-staining of mitotic Pax6+ RGP cells using anti-phospho Histone H3 antibody (pH3) revealed a mitotic index of 3.97%±0.16% for noninfected cells (n=10,121 cells), consistent with previously reported measurements (Hu et al., 2013) (Fig. 3a, c). This value was not significantly affected after WNV infection (4.01%±0.39%, n=4422 cells). However, ZIKV infection strongly reduced mitotic index (1.19%±0.33%, n=4115 cells), indicating that ZIKV, but not WNV, impairs cell cycle progression of RGP cells (Fig. 3a, c). Cleaved caspase 3 staining (Cas3) did not reveal increased apoptotic cell death after infection with ZIKV or WNV (Fig. 3b, d). In fact, these viruses both appeared to inhibit apoptosis 24-h post-infection, consistent with the anti-apoptotic activity of flaviviruses in the early stages of infection (Lee et al., 2005).
Discussion In this study, we describe the infection pattern of ZIKV and other closely related flaviviruses in the developing brain. Our results indicate that ZIKV infection exhibits a striking bias towards RGP cells and strongly impairs their proliferation, suggesting a potential route leading to microcephaly. Importantly, we did not observe this preferential infection of RGP cells for neurotropic WNV, while DENV-4 does not infect the neocortex at all. Recent evidence revealed absence of placenta infection by DENV-3 after inoculation into pregnant mice (Miner et al., 2016), supporting the notion that dengue viruses may represent a useful comparison point to identify ZIKV-specific infection routes. How ZIKV enters RGP cells is still unclear but recent observations showed strong expression of the candidate entry receptor AXL in mice and human RGP cells (Nowakowski et al., 2016).