To obtain human neurons, AAVS1-TRE3G-NGN2 human embryonic stem cells (hESC-NGN), kindly provided by J.W. Harper, were used and differentiated as indicated by Ordureau et al., with minor modifications (Ordureau et al., 2018). Briefly, H9-NGN2 cells were seeded on day 0 in a Geltrex^®-coated six-well plate (2 × 10⁵ cells/well) in ND1 medium composed of DMEM/F12/NEAA (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA), human Brain-Derived Neurotrophic Factor (BDNF) (10 ng/ml; PeproTech, Thermo Fisher Scientific), human NT-3 (10 ng/ml, PeproTech, Thermo Fisher Scientific), human laminin (0.2 µg/ml, Invitrogen, Thermo Fisher Scientific), and doxycycline (2 µg/ml, Clontech, Takara Bio USA, San Jose, CA, USA). On day 1, ND1 medium was replaced with fresh ND1 medium; on day 2, ND1 medium was replaced with ND2 medium consisting of Neurobasal medium supplemented with B27/GlutaMAX (Invitrogen, Thermo Fisher Scientific), BDNF, NT3, and doxycycline (concentrations are the same as above). After day 2, 50% of the medium in each well was replaced every 2 days with fresh medium. On day 7, cells were detached by treatment with Accutase (Gibco, Thermo Fisher Scientific) and replated in Geltrex^®-coated six-well plates (4 × 10⁵ cells/well) in ND2 medium supplemented with Y27632 (10 µM, Thermo Fisher Scientific). On day 8, the medium was replaced with fresh ND2 Medium. After day 8, 50% of the medium in each well was replaced every 2 days until day 14, when cortical neurons reached the mature phase.

To obtain 3D human brain organoids (hBOs), the protocol developed by Lancaster and Knoblich was adopted with some modifications (Lancaster and Knoblich, 2014). Briefly, on day 0, human ES cells (H9, WiCell Institute, Madison, U.S.A.) were dissociated to a single-cell suspension and seeded in round-bottom, low-attachment 96-well plates (1 × 10⁵ cells/well) in mTeSR complete medium (StemCell Technologies, Cologne, Germany) supplemented with Y27632 (10 µM, Thermo Fisher Scientific). On day 3, the medium was replaced with fresh mTeSR. On day 5, the medium was replaced with Neural Induction Medium, composed of DMEM/F12/NEAA (Gibco, Thermo Fisher Scientific) with GlutaMAX/B27 (Thermo Fisher Scientific), heparin (1 mg/ml, Merck Millipore, Burlington, MA, USA), and antibiotic–antimycotic (Anti-Anti, 1% v/v, Gibco, Thermo Fisher Scientific). The embryoid bodies (EBs) were fed every other day. On day 11, the EBs were embedded on Matrigel (Corning™, Merck Millipore) droplets and grown in a low-attachment six-well plate (eight to nine EBs/well) in differentiation medium (DM), composed of Neurobasal/DMEM/F12 (Thermo Fisher Scientific) supplemented with GlutaMAX/NEAA/B27/N2 (Thermo Fisher Scientific), with human insulin (0.00025%, v/v, Merck-Milliore, Milan, Italy) and 2-mercaptoethanol (0.00035%, v/v, Gibco, Thermo Fisher Scientific) added. To prevent contamination, Anti-Anti (1%, v/v, Gibco, Thermo Fisher Scientific) was included. On day 13, DM was replaced with fresh medium. On day 15, the medium in each well was replaced with differentiation medium 2 (DM2), i.e., DM with B27 plus vitamin A (Thermo Fisher Scientific), and the plate was placed on an orbital shaker. From days 15 to 60, the hBOs were grown in an orbital shaker until the day of infection. The medium was changed twice per week.

DBTRG.05MG cells and human ESC-derived neurons were infected with SI-1812 or INMI TOSV strains at an MOI of 0.1. After 1 h of virus absorption, extensive washes were performed to remove the inoculum, and complete growth medium was added. Cell culture supernatants were collected at 1, 3, 5, 6, and 7 days postinfection (p.i.) and stored at − 80 °C for further analysis. hBOs were infected with SI-1812 and INMI TOSV strains at an MOI of 0.2. For infection, hBOs were transferred into a tube containing the required amount of virus. After inoculation, hBOs were centrifuged at 1,100 rpm for 30 min. Subsequently, hBOs were incubated at 37 °C with 5% CO₂ for 60 min to allow the adsorption of the virus. The medium was then removed, the organoids were washed with PBS twice, and fresh DM2 was added. The organoids were incubated at 37 °C, with 5% CO₂ in an orbital shaker. Supernatants of infected organoids were collected at different time points (1, 3, 6, 10, 16, 19, and 21 days postinfection) for further analysis. The organoids were also collected to perform immunofluorescence. The release of infectious viral particles by infected DBTRG.05MG cells, human ESC-derived neurons, and hBOs was determined by virus microtitration assay on Vero E6 cells cultured in 96-well plates, and the 50% Tissue Culture Infectious Dose (TCID₅₀/ml) end point titer was calculated using the Reed and Muench method (Reed and Muench, 1938). Results were reported as mean values ± standard deviation (SD) from at least three independent experiments.

Medium from DBTRG.05MG cells and human ESC-derived neurons, mock- or TOSV-infected, collected at 1, 3, 5, 6, and 7 days p.i as described above, and the medium from hBOs, mock- or TOSV-infected, collected at 1, 3, 6, 10, 16, 19, and 21 days p.i., were processed for IFN-β quantification by enzyme-linked immunosorbent assay (ELISA) using a commercial kit (VeriKine-HS Human IFN Beta TCM ELISA Kit; PBL Assay Science, Piscataway, NJ, USA), following manufacturer’s instructions. IFN-β level was expressed as mean cytokine concentrations (pg/ml) ± SD from at least three independent experiments.

Cell supernatant of mock-infected or TOSV-infected hBOs was centrifuged at 10,000 × g for 5 min, prior to performing the assay, and diluted 1:4 in Universal Assay Buffer. A multiplex biometric ELISA-based immunoassay, containing dyed microspheres conjugated with 10 different monoclonal antibodies specific for a target cytokine, was used according to the manufacturer’s instructions (Human Luminex Discovery Assay, Bio-Techne, Minneapolis, MN, USA). Magnetic bead concentrations were measured using a Luminex™ detection system (Luminex™ 200™ Instrument System, Bio-Techne, Minneapolis, MN, USA). The following cytokines were analyzed: IFN-β, Interferon gamma (IFN-γ), tumor necrosis factor (TNF-α), interleukin (IL)-1β, IL-2, IL-4, IL-6, IL-8, IL-10, and IL-12. The analyte concentration was calculated using a 7-point standard curve. Acquired data were analyzed using xPONENT v4.3.309.1, and cytokine concentrations were reported as picograms per milliliter.

Neurons (10⁴ cells/cover slip) were grown on Geltrex^®-treated coverslips. Cells were fixed in paraformaldehyde (4%, v/v, Thermo Fisher Scientific) for 20 min and permeabilized with Triton X-100 (0.1%, v/v) in Dulbecco's modified Phosphate Buffered Solution (DPBS) (Gibco, Thermo Fisher Scientific) for 5 min at room temperature. Cells were then incubated with the following primary antibodies diluted 1:100 in BSA (1%, w/v, in PBS): serum anti-TOSV, beta-3 Tubulin (Thermo Fisher Scientific). Nuclei were stained with the DRAQ5™ intercalating fluorescent probe (Thermo Fisher Scientific), and cells were observed with a confocal microscope (Leica, Buffalo Grove, IL, USA) at × 63 magnification and processed using Fiji-ImageJ software. Human brain organoids were fixed in paraformaldehyde (4%, v/v, Thermo Fisher Scientific) for 40 min and washed three times in DPBS. hBOs were placed in 10% w/v sucrose solution for 30 min, in 20% w/v sucrose solution for 30 min, and finally in 30% w/v sucrose solution at + 4 °C, until embedding with OCT solution (VWR International PBI, Milan, Italy) in a plastic mold. To solidify the solution rapidly, molds were put in contact with liquid nitrogen vapors. These organoids were prepared for cryosectioning and immunostaining. The 20-µm-thick sections generated were washed three times with DPBS and incubated with 4% BSA/DPBS containing 0.2% Triton X-100 for 1 h at room temperature. After washing twice in DPBS, primary antibodies diluted in 4% BSA/PBS were added and incubated overnight at 4 °C. Cells were incubated with the following primary antibodies: serum anti-TOSV, anti-TUJI1 (Thermo Fisher Scientific), and anti-cleaved caspase-3 (Asp175; Cell Signaling Technology, Leiden, The Netherlands). Nuclei were stained with DRAQ5™ interlayer fluorescent probe (Thermo Fisher Scientific), and cells were observed confocal microscope (Leica, Buffalo Grove, IL, USA) and processed by Fiji-ImageJ software. The experiment was repeated three times, and a total of 12 slices of processed brain tissue were analyzed. Vero E6 cells were infected with medium collected from hBOs at 6 days p.i. and fixed with cold acetone:methanol solution for 10 min. Immunofluorescence staining was then performed using an anti-N TOSV antibody, followed by a FITC-conjugated secondary antibody. Images were acquired with a fluorescence-equipped Eclipse Ts2 microscope (Nikon, Milan, Italy) and processed using Fiji-ImageJ software. Corrected total cell fluorescence (CTCF) was calculated as: CTCF = Integrated Density − (Area of selected cell × Mean fluorescence of background readings). Data are presented as mean CTCF ± SD from at least three independent fields.

Endogenous Retinoic-acid Inducible Gene I (RIG-I) and TOSV SI-1812 or INMI NS expression was investigated in DBTRG.05MG cells infected as previously described and collected at 24, 48, and 72 h p.i. Whole-cell lysates were prepared in RIPA buffer (25 mM Tris-HCl [pH 7.5]; 150 mM NaCl; 1% Triton X-100) supplemented with COmplete™ Protease Inhibitor (Merck-Millipore), and protein concentration was determined by BCA assay (Pierce, Milan, Italy). Fifty micrograms of total proteins, supplemented with Laemmli sample buffer and denatured for 5 min at 95 °C, was separated by SDS-PAGE and then transferred to a NitroBind nitrocellulose membrane (Santa Cruz Biotechnology, Heidelberg, Germany). After blocking with 5% nonfat dry milk, the filters were incubated O/N at room temperature with anti-RIG-I CARDs (1:5,000), anti-NSs (1:200), or antiactin (1:2,500; Invitrogen, Thermo Fisher Scientific) as a loading control. After being washed three times with PBS containing 0.05% Tween-20 (PBS-T), membranes were incubated with a horseradish peroxidase (HRP)-conjugated anti-mouse IgG secondary antibody (Merck-Millipore). After several washes, the target bands were visualized using an enhanced chemiluminescence kit (Pierce), and images were acquired with a ChemiDoc instrument (BioRad, Milan, Italy).

Statistical significance was assessed with the two-tailed Chi-square test, paired or unpaired as appropriate. Results were considered statistically significant at p < 0.05. When appropriate, values are reported as mean ± SD. One-way ANOVA followed by a Bonferroni post hoc test was performed to compare the significance among different study groups. Data are presented as mean values ± SD of three independent experiments performed in triplicate, unless otherwise stated. All analyses were performed using GraphPad Prism software (v.8.0.1).

To compare the infection and replication efficiency of the two TOSV strains (INMI and SI-1812) in CNS cells, we used different human cell systems, i.e., the DBTRG.05 MG glioblastoma cell line, cortical neurons, and brain organoids. The DBTRG.05 MG cell line, isolated from a female patient with glioblastoma, although highly aneuploid, responds to IFN-β stimulus with expression of proinflammatory genes, thus suggesting that it originated from astrocytes. Both TOSV INMI and SI-1812 strains were able to infect and grow in DBTRG.05 MG glioblastoma cells, although the INMI strain reached higher titers at 5 days (2.1 × 10⁷ TCID₅₀/ml ± 1.05 × 10⁷ TCID₅₀/ml, p = 0.03) and 6 days p.i. (7.5 × 10⁵ TCID₅₀/ml ± 3.8 × 10⁵ TCID₅₀/ml, p = 0.04) than the SI-1812 strain (1.5 × 10⁶ ± 1.06 × 10⁶ and 1.4 × 10⁵ ± 5.5 × 10⁴ TCID₅₀/ml, respectively). At 7 days p.i., viral titer slightly decreased for both the strains ( Figure 1A ).

Differentiated cortical neurons were infected with TOSV INMI and SI-1812 at an MOI of 0.1 and assayed during the next 7 days. These cells were susceptible to TOSV infection and permissive to replication, as determined by titration of infectious viral particles and immunofluorescence. As shown in Figure 1B , in this neural cell system, TOSV INMI replication was slower than that of the SI-1812 strain in the first 3 days p.i., yielding significantly lower infectious viral titer in cell supernatant (day 3 p.i., SI-1812 titer 1.2 × 10⁶ TCID₅₀/ml ± 1.7 × 10⁵ TCID₅₀/ml vs. 6.6 × 10³ TCID₅₀/ml ± 2.5 × 10³ TCID₅₀/ml of the INMI strain, p = 0.028). However, the INMI strain appeared more capable of spreading in these cells than the SI-1812 strain. Indeed, although both strains reached a similar viral titer at 6 days p.i. (p = 0.074), immunofluorescence showed a higher number of positive cells in neurons infected with the INMI strain (p = 0.028) ( Figure 2A ). Furthermore, the fold-change fluorescence relative to the mock sample confirmed the enhanced capability of the INMI strain to spread in ESC-derived neurons ( Figure 2B ).

hBOs were infected at day 60 of differentiation, when they presented complex cell–cell and cell–environment interactions similar to the in vivo condition. Briefly, hBOs were infected with TOSV INMI and SI-1812 strains at a 0.2 MOI. Experiments were conducted in biological triplicate, with five hBOs per experiment. Following infection, morphological sizing of hBOs was monitored up to day 21 p.i. Viral infection kinetics and replication efficiency were assessed by infectious virus titration for each experiment. Additionally, immunofluorescence staining was performed on hBOs, with a total of six slides per organoid ( Figure 3 ). Also in this cell system, the TOSV INMI strain grew more slowly than the SI-1812 strain for the first 3 days p.i. (p = 0.045). At day 6 p.i., the viral titer (about 10⁵ TCID₅₀/ml) was similar for the two strains (p = 0.808). At 10 days p.i., however, the INMI strain showed a higher titer (p = 0.017) (8.1 × 10⁵ TCID₅₀/ml ± 3.1 × 10⁵ TCID₅₀/ml) than the SI-1812 strain (9.1 × 10⁴ TCID₅₀/ml ± 2.7 × 10⁴ TCID₅₀/ml) ( Figure 3A ). The difference further increased in the follow-up due to massive cell death and hBO disruption. Indeed, as shown in Figure 4 , the size of SI-1812-infected hBOs was smaller than that of hBOs infected with INMI at day 16 p.i. (p < 0.0009). Immunofluorescence performed on TOSV-infected hBOs at day 6 p.i. showed a similar distribution of infected cells for both strains ( Figure 3B ), but a higher level of apoptosis induction in INMI-infected hBOs compared to SI-1812-infected hBOs, as shown by coimmunostaining for cleaved caspase-3 (p < 0.0008, Figure 3C ).

The two virus strains isolated from brain organoids at 6 days p.i., showing similar titer, were then inoculated in Vero E6 cells. At 3 days p.i., SI-1812 TOSV showed a CPE that was slightly more pronounced than the other virus. In contrast, at 4 days p.i., the CPE was markedly more extended in cells infected with the INMI strain than those infected with SI-1812, suggesting a greater spreading capability for this strain ( Figures 5A–D ). This result was further confirmed by immunofluorescence, which demonstrated the presence of the virus in a larger number of cells infected with the INMI strain at 4 days p.i. ( Figures 5E–H ), which, in turn, exhibited a reduced cytopathic effect and a decreased neural cell death. Indeed, the mean CTCF for infected INMI cells was 7.45 × 10⁷ ± 1.87 × 10⁷, while for SI-1812 samples, it was 2.56 × 10⁷ ± 1.27 × 10⁷ (p = 0.002).

Interferon-β expression was evaluated in human DBTRG.05MG glioblastoma cells, ESC-derived human cortical neurons, and hBOs infected with the two strains of TOSV at different time points. Infection with TOSV INMI led to a significantly lower induction of IFN-β than the SI-1812 strain in the two cell types and hBOs ( Figure 6 ). In infected DBTRG.05MG cells, the highest level of IFN-β was recorded at day 6 p.i. (478.46 pg/ml ± 59.51 pg/ml of the INMI strain vs. 1,427.57 pg/ml ± 80.19 pg/ml of the SI-1812 strain; p < 0.0001). It then began to drop only in cells infected with SI-1812, conceivably as a consequence of cell death due to cytopathic effect ( Figure 6A ). Similarly, along the 7-day time course in ESC-derived neurons, IFN-β was produced at lower levels in cells infected with the INMI strain (mean concentration: 3.6 pg/ml ± 0.93 pg/ml) than in those infected with the SI-1812 strain (mean concentration: 7.8 pg/ml ± 3.7 pg/ml), which showed a significant and progressive increase in IFN-β secretion, doubling its level after the third day p.i. (6.1 pg/ml ± 0.92 pg/ml) until the seventh day p.i. (10.2 pg/ml ± 1.41 pg/ml) (p < 0.0001) ( Figure 6B ).

In hBOs, the difference of IFN-β induction by the two strains was more pronounced, with IFN-β levels about 10-fold lower in cells infected with the TOSV INMI strain (161 pg/ml ± 35.4 pg/ml) than in cells infected with the SI-1812 strain (1,148 pg/ml ± 161.3 pg/ml) at 3 days p.i. (p = 0.0005) ( Figure 6C ). In cells infected with the INMI strain, IFN-β increased at 6 days p.i. (398 pg/ml ± 55.4 pg/ml) and showed no significant differences in the following days until day 21 p.i. (552 pg/ml ± 74.23 pg/ml) ( Figure 6C ). Multiplex cytokine analyses carried out on the growth medium of infected hBOs confirmed the higher production of IFN-β (day 10 p.i.: 1.9 × 10⁴ pg/ml ± 1.5 × 10³ pg/ml, p = 0.0001), as well as the proinflammatory cytokine IL-6 (day 10 p.i.: 3.4 × 10³ pg/ml ± 2.1 × 10² pg/ml, p < 0.0001) and IL-8 (day 10 p.i.: 922 pg/ml ± 57.2 pg/ml, p = 0.0035) in cells infected with SI-1812, compared to those infected with INMI ( Figure 7 ). The other assayed interleukins were not detected. The high expression of IL-6 and IL-8 in the supernatant of SI-1812-infected cells demonstrated that this strain unveiled the neuropathogenesis of the brain by inducing a strong inflammation and destroying cells, in contrast to the INMI strain, which persisted longer in hBOs.

Previous data showed that the inhibitory effect of TOSV NSs on RIG-I was involved in the signaling cascade for type I IFN production, leading to degradation of RIG-I upon binding. In this study, we examined whether the interaction between RIG-I and NSs could have a functional consequence when DBTRG.05MG cells were infected with TOSV INMI or SI-1812 strain. For this purpose, we assessed the presence of RIG-I and NSs in these infected cells at 24, 48, and 72 h p.i. As shown in Figure 8 , the RIG-I protein became detectable at 2 days p.i., while NSs were more prominent at 3 days p.i. in cells infected with the SI-1812 strain. In contrast, RIG-I was observed only at 3 days p.i., whereas NSs were strongly expressed as early as 1 day p.i. in cells infected with the INMI strain, suggesting that early and efficient expression of NSs in INMI-infected cells led to RIG-I degradation and hence impaired IFN-β response. This feature also confirmed that the different amount of IFN-β expressed by these cells depends on the TOSV strain used for infection.