Oropouche virus (strain BeAn 19991) was kindly provided by Prof. Luiz Tadeu Moraes Figueiredo (University of São Paulo, Ribeirão Preto Medical School, Brazil) and propagated in Vero cells to produce the virus stock. The titer of the virus stock was determined by cytopathic effect induction on Vero cells and expressed as 50% infectious dose (TCDI₅₀). Conditioned medium from control Vero cells (mock) were used as negative control.

Cultures were prepared as previously described (Mendes et al., 2018; Fernandes et al., 2019). Briefly, cortical tissue (anterior third of middle temporal gyrus) was obtained from patients submitted to an anterior temporal lobectomy with amygdalo-hippocampectomy for the treatment of refractory temporal lobe epilepsy. No lesion features were detected in both resonance exams and anatomopathological assessments. A fragment of cortex was collected at the surgical room immediately after resection and transported to the laboratory immersed in ice-cold oxygenated medium [50% v/v Hank’s balanced salt solution in Neurobasal A (Gibco)] supplemented with 10 mM Hepes and glucose 3 mg/ml. Tissue immersed in ice-cold oxygenated HBSS was sliced at 200 μm in a VT1000s automatic vibratome (Leica), and then transferred to 24-wells plate (one slice/well) with 600 μL of pre-warmed culture medium. The slice culture medium contains Neurobasal A (Gibco) medium supplemented with 1% Glutamax (Gibco), 1% Penicillin/Streptomycin (Gibco), 2% B27 (Gibco) and 0.25μg/mL Fungizona (Sigma). At day in vitro 1 (DIV1), 333μL were removed and 133μL of medium was added, and brain slices were cultivated in free-floating conditions until further treatments. One third of the culture medium was exchanged for fresh medium every 24 h. This procedure was approved by the Ethics Committee of the Ribeirão Preto Medical School (protocol HCRP #17578/15).

At days in vitro (DIV) 1 or 2, the medium was removed, and the slices were exposed to 3 × 10⁶ TCDI₅₀ of OROV or an equivalent volume of mock medium. Infection was performed for 2 h at 37°C, 5% CO₂. The inoculum was removed, the tissue was washed five times with PBS, and once with Neurobasal A medium (this medium was stored and referred to as 0 hours post-infection, hpi). The slices were maintained in fresh medium at 37°C and 5% CO₂ until processing for subsequent analysis.

Brain slices (200 μm thick) were fixed in 4% paraformaldehyde for 24 h at 4°C, then incubated for 1 h with 0.3% Triton X-100 in PBS, washed with wash solution (0.2% Triton X-100 plus 1% BSA in PBS) and incubated with 0.1 M glycine for 1 h. The slices were blocked with 2% BSA for 2 h and incubated with the one of the following primary antibodies diluted in 0.2% Triton-X 100 and 2% BSA [mouse polyclonal anti-OROV (1:300; in house, provided by Prof. Luiz Tadeu Moraes Figueiredo), anti- Iba1 (1:1,000, Abcam, ab178846), anti-NeuN (1:1,000; Abcam; ab177487) or anti-GFAP (1:1,000; Merck, ab5804)] for 24 h at 4°C. Slices were washed three times with wash solution and incubated with one of the following secondary antibodies diluted in 2% BSA: Alexa Fluor 488 (Life Technologies; A-11001) or Alexa Fluor 568 (Life Technologies; A-11011). Finally, slices were stained with DAPI diluted in 2% BSA for 1 h, washed with wash solution, and incubated for 5 min with 70% ethanol followed by treatment with 0.1% Sudan Black (Merck) in 70% ethanol for 5 min. After washing (twice) with 70% ethanol, slices were mounted on slides coated with Fluoromont (Thermo Fisher). Stained tissues were analyzed on a Leica TCS SP5 laser scanning (Leica Microsystems) or Axio Observer LSM 780 (Carl Zeiss) confocal microscopes. Post-acquisition image processing and analyses were carried out using Image J/Fiji (Schindelin et al., 2012).

Slices were fixed in 4% formaldehyde for 24 h at 4°C, and cryoprotect in 30% sucrose in PBS for 48 h. Slices were placed in a freezing microtome (Leica) at –40°C and cut at 30 μm. In 24-well plates containing 0.1 M PB solution (pH 7.4; Na₂HPO₄ Merck, 28 g; NaH₂PO₄.H₂O Merck, 2.76 g, for 1 L), slices were incubated for 20 min with 10% hydrogen peroxide, washed and blocked with 2% donkey serum in 0.1 M PB for 40 min. Slices were immunolabeled with anti-NeuN antibody (1:1,000; Abcam; ab177487), overnight at 4°C. After washes, incubation with secondary antibody diluted in TBS-Tx 0.05 M was carried out for 2 h. Slices were then washed and incubated for 2 h with ABC Kit (Vector – Vectastain Standard Kit PK4000) diluted 1/200 according to the manufacturer’s instructions. Immunohistochemistry was performed using DAB + Nickel solution. Finally, slices were placed on frosted cut slides (Knittel), dehydrated, and covered with Permount (Fisher Scientific) and observed under a light microscope (Leica).

Virus particles in the supernatant from human brain slices were precipitated using polyethylene glycol (PEG) as described for other Bunyavirus (Hover et al., 2018), with minor modifications. Supernatants were incubated with 20% (PEG 6000, Sigma) diluted in 2.5 M NaCl (PEG/NaCl solution) in the proportion of 1: 4 (PEG/NaCl: supernatant). The mixture was gently homogenized, kept at 4°C for 1 h, and centrifuged at 10,800 g for 30 min at 4°C. The pellet was resuspended in 8 μL of PBS plus 2 μL of PEG, homogenized and kept at 4°C for 20 min, and centrifuged at 3,300 g for 30 min at 4°C. The pellet was resuspended in 50 μL of PBS, which was centrifuged at 11,600 g for 10 min at 4°C. The supernatant was immediately used to viral titration. Serial 10-fold dilutions of the concentrated supernatants were prepared in DMEM medium supplemented with 2% fetal bovine serum (FBS, Thermo Fisher) and added to Vero ATCC CCL81 cells monolayers plated on 96-well plates. After incubation at 37°C for 4 days, the titer was obtained by analyzing cytopathic effect events in each dilution and expressed as 50% infectious dose (TCID₅₀) (Barbosa et al., 2018). The data was normalized by the mass of the slice correspondent to each supernatant.

Tissue was fixed with 3% glutaraldehyde in 0.1 M phosphate buffer for 24 h at 4°C and processed for electron microscopy as described below. Tissue was incubated with 2% osmium tetroxide by 2 h at 4°C, washed in 0.1 M phosphate buffer, and dehydrated in a graded ethanol series. The tissue was imbedded in resin (araldite 6005) and stained with toluidine blue to select the area of interest. The tissue was cut with a diamond knife and the contrast were obtained by incubating the tissue with 0.5% uranyl acetate and lead citrate. After careful drying of the grid, the material was analyzed using a JEOL JEM-100CX II electron microscope (JEOL, MA, United States) at the Multi-User Electronic Microscopy Laboratory of the Department of Cellular, Molecular Biology and Pathogenic Bioagents, Ribeirão Preto Medical School.

Differentiation of SH-SY5Y neuroblastoma cells was carried out as previously described (Lopes et al., 2019), with minor modifications. Cells were plated on coverslips previously coated with poly-D-lysine (Sigma) and maintained in Dulbecco’s Modified Eagle Medium (DMEM) containing 4,500 mg/L of glucose, L-glutamine (Gibco), 10% heat-inactivated fetal bovine serum (Gibco), 100 U/L of penicillin, and 0.1 g/L of streptomycin in a saturated humidity atmosphere, 5% CO₂ at 37°C. After 24 h, the medium was replaced by differentiation medium containing 0.5% FBS, 100 U/L of penicillin and 0.1 g/L of streptomycin, and 10 μM of retinoic acid (RA; Sigma) in DMEM for 4 days. After this period, cells were incubated with differentiation medium supplemented with 50 ng/mL of brain-derived neurotrophic factor (BDNF, Sigma) for another 4 days. The differentiation medium was refreshed every 48 h. After differentiation, cells were incubated with the OROV inoculum for 1 h at 4°C (MOI = 1, multiplicity of infection) in DMEM supplemented with 2% FBS. The inoculum was removed, cell monolayer was washed and maintained in DMEM medium at 37°C and 5% CO₂ until analysis.

The levels of cytokines in the conditioned culture medium were determined by ELISA (BD OptiEIA, BD Biosciences). The results were expressed as picograms of each cytokine analyzed and normalized by the content of total protein in each slice (determined by the Bradford assay).

Extracts from mock and infected slices were prepared in RIPA buffer (50 mM Tris, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100 and 0.1% SDS, pH 7.5, supplemented with protease and phosphatase inhibitor) using a motorized pestle. Homogenates were centrifuged at 4°C for 15 min at 16,000 × g and supernatants collected. Protein concentration was determined using the Bradford assay. Aliquots containing 50 μg total protein were resolved on Mini-PROTEAN TGX^TM Precast Protein Gels, 4–15% (Biorad). The proteins were transferred to nitrocellulose membranes using a Transblot turbo apparatus (Biorad). Membranes were blocked at 25°C for 1 h with 2.5% BSA in TBS-T (0.05 M Tris, 0.15 M NaCl, 0.1% Tween 20, pH 7.5) and incubated overnight with anti-HLA-DR (1:1,000, Abcam, ab92511), anti-Tau phospho S396 [E178] (1:1,000, Abcam, ab32057), or anti-β-actin (1:10,000, Merck, MAB1501). Primary antibodies were diluted in TBS-T containing 2.5% BSA. After washing and incubation with HRP-conjugated secondary antibodies (GE Healthcare), membranes were developed using ECL-Prime (GE Healthcare). Quantitative analyses were made using Image J/Fiji.

Cell viability in the cultured slices was assessed using the MTT assay as in Mendes et al. (2018) and Fernandes et al. (2019). Briefly, slices were incubated with 0.5 mg/ml MTT for 3 h at 37°C, washed, and homogenized in isopropanol/0.01 N HCl. The homogenate was centrifuged at 5,000 rpm for 2 min and the supernatants were collected for quantification of O.D. at 540 nm. The values were normalized by the mass of each slice.

Statistical analysis was performed using Graph Pad Prism 6 (Graph Pad Software). Means were compared using Student’s t-test (for two groups) or One-Way ANOVA (for three groups), as indicated in the legends for the figures. The number of tissue donors included in each experiment has been referred to as “n” throughout the text.

Previous studies demonstrated neuroinfection by OROV in rodent models (Rodrigues et al., 2011; Santos et al., 2012, 2014; Proenca-modena et al., 2016). On the other hand, no conclusive information about OROV infection in human neural cells is yet available. To address this, we used a novel human brain slice culture model developed by our group (Mendes et al., 2018; Fernandes et al., 2019). The slices are prepared from tissue collected at the surgical room, from adult patients undergoing surgical resection of epileptic foci. In our previous works, we have observed no major constraints in cell viability and tissue morphology/cytoarchitecture in these human brain-derived slice cultures that could impose a significant limitation to the use of this model for studying virus neuroinfection (Mendes et al., 2018; Fernandes et al., 2019). The tissue fragment used to prepare the slice cultures correspond to the temporal cortex resected to provide access to the hippocampal epileptogenic area. A significant advantage of this model is to preserve the original cellular population and connections found in the adult human brain, as we have previously shown the presence of neurons, microglia and astrocytes by light microscopy in re-sectioned slices submitted to DAB-Nickel immunostaining (Fernandes et al., 2019).

To evaluate if human neural cells in a preserved tissue architecture are susceptible to OROV infection, we have exposed human brain slices in culture to OROV for either 24- or 48-h, and evaluated the presence of viral antigens in the cells using a polyclonal anti-OROV antibody (Rodrigues et al., 2011). We detected viral antigens in human neural cells at both time points tested, with a significant increase in the number of infected cells at 48 hours post infection (hpi) compared to 24 hpi (Figures 1A–C). Sequential images of cells positive for OROV in human brain slices showed virus antigens throughout the cell cytoplasm (Figure 1B). To investigate if human neural cells were permissive for the production of novel virus particles, we collected supernatants from OROV-infected slices and performed OROV titration in OROV-susceptible Vero cell monolayers by TCID₅₀. OROV progeny accumulated about 40-fold in supernatant of OROV-infected human brain slices at 48 hpi as compared to time zero, and had only a modest decrease at 72 hpi (Figures 1D,E). The supernatant collected at 24 hpi was not titrated because only a few cells were positive for OROV antigen by immunostaining at that time point (Figure 1). The presence of OROV virus particles in infected human brain slices was also observed by transmission electron microscopy at 48 hpi, which showed structures with typical morphology and diameter expected for family Peribunyaviridae (Talmon et al., 1987; Travassos Da Rosa et al., 2017; Figure 1F). These structures were not observed in mock-infected slices (data not shown). In conjunction, these data indicate that neural cells in adult human brain tissue are indeed susceptible to OROV.

To gain information on the pattern of OROV infection in neocortical areas, we evaluated virus antigens distribution along cortical layers in OROV-infected brain slices. Although some care must be taken when analyzing cortical tissue obtained from pharmaco-resistant epileptic patients, as it is known that in some cases neurons can present abnormal orientations (Al Sufiani and Ang, 2012), in our previous work we have been able to clearly identify all neuronal cortical layers in similar slice cultures (Fernandes et al., 2019). Here, we have seen that OROV infection is present in most of the cortical layers identified, being more readily detected in deeper layers (Figure 2). In order to determine which neural cell types are infected by OROV in human brain slices in culture, we performed double immunostainings to detect the viral antigens and one of the following neural cell markers: Iba1 for microglia; NeuN for neurons; or GFAP for astrocytes. We observed that at 48 hpi both microglial cells and neurons were positive for OROV antigen, while no infected astrocytes were detected (Figure 3A). Despite the fact that microglia represent 10% of the total cells present in slices (Figure 3B), approximately 35% of these cells were positive for OROV antigens (Figure 3C). On the other hand, neurons represent 33% of total cells (Figure 3B), but only approximately 2% of them were OROV-positive (Figure 3D). Astrocytes correspond to 11% of total cells (Figure 3B), but although being a significant cell population in these slices, a careful search (11 fields, including 0.5 μm z stacks), indicated a lack of co-localization between OROV proteins and GFAP positive cells in adult human brain slices. In conjunction, these quantifications indicate that microglial cells are preferentially infected by OROV in adult human brain cortex.

To further investigate the OROV potential to infect human neurons, we induced differentiation of the human neuroblastoma cell line SH-SY5Y into a mature neuronal phenotype and exposed these cultures to OROV. We observed intense cytopathic effect at 24-, 36-, and 48 hpi (Figure 4A), and infection was confirmed by the detection of strong OROV antigens by immunostaining (Figure 4B). In addition, infectious OROV particle production by infected differentiated SH-SY5Y cells was confirmed by TCID₅₀ assay in supernatants and remaining cell homogenates, which was time-dependent and reached a plateau at 48 h post infection (Figure 4C), roughly in agreement with the time when most cells presented cytopathic effect. These data reinforced the notion that OROV can infect human neurons, even though the degree of neuronal infection by OROV in the context of preserved human brain tissue seemed to be significantly less abundant than that of microglia.

The demonstration of OROV infection of both microglia and neurons in human brain slices raised the possibility of augmented release of inflammatory signals by these cells in response to OROV infection. We therefore determined the levels of cytokines in the conditioned medium of OROV-infected human brain slices by ELISA. Elevated levels of cytokines have been described as an important initial response to experimental virus infection, including against OROV (Rempel et al., 2005; Kawanokuchi et al., 2006; Amorim et al., 2020). Interestingly, we observed a significant increase in TNF-α release by brain slices infected with OROV, as compared to control slices (Figure 5A). On the other hand, no significant differences were found in IL-12 or IL-10 levels between mock and OROV-infected human brain slices (Figure 5A). We also detected an increase in TNF-α mRNA levels in OROV-infected brain slices in samples from a single donor (Supplementary Figure 1). This inflammatory response seems to be not associated to tissue processing prior culturing, since we have observed a clear change in microglial morphology, including increased cell body area and reduced amount and length of branches, known features of activated microglia (Tay et al., 2017; Koss et al., 2019; Tan et al., 2020), after a pro-inflammatory stimulus with LPS at DIV2 (Supplementary Figure 2), which indicates microglia was not highly activated at the time of experimental infection with OROV.

Since microglia has been described as the main source of TNF-α in the brain (Riazi et al., 2008; Wang et al., 2015), we further investigated microglial activation status in OROV infected-brain slices. Firstly, we have evaluated the expression of iNOS, a functional marker of M1 (pro-inflammatory) microglia phenotype (Lisi et al., 2017). In line with the notion of microglial activation upon OROV infection, we have observed an increase in iNOS mRNA levels in samples from a single donor (Supplementary Figure 1). Moreover, we evaluated the protein levels of a classical activation marker for microglial cells, HLA-DR, a MHC class II cell surface receptor known to play a role in virus clearance (Walker and Lue, 2015; Hendrickx et al., 2017; Klein et al., 2019; Chhatbar and Prinz, 2021). Remarkably, we found an augment in HLA-DR levels in all OROV-infected slices tested (Figures 5B,C). In addition to elevated iNOS and TNF-a mRNA levels, and HLA-DR and TNF-α proteins in OROV-infected human brain slices, we also observed microglia with apparent amoeboid morphology near infected cells, suggesting microglial activation (Supplementary Figure 3).

Finally, to investigated the impact of OROV infection on neuronal viability, we evaluated Tau phosphorylation (Serine396) levels in OROV-exposed adult human brain slices, as increased pTau (Ser396) has been shown to be a solid marker of neurodegeneration (Wang et al., 2013). We observed an increase in pTau (ser396) levels in 2 out of 4 slices tested, indicating neuronal damage in part of the infected slices (Figures 5D,E). Furthermore, OROV-infection led to a reduction in tissue viability, as determined by the MTT assay (Figure 5F). This reduction in viability was not associated with a reduction in mitochondrial content (Supplementary Figure 4), suggesting that it corresponds to cell death. Collectively, these data indicated that OROV infection triggers a microglia-mediated inflammatory response and neurotoxicity in adult human brain slices.