Reactive Oxygen Species (ROS) Are Not a Key Determinant for Zika Virus-Induced Apoptosis in SH-SY5Y Neuroblastoma Cells

Introduction: ZIKV is a highly neurotropic virus that can cause the death of infected neuroprogenitor cells through mitochondrial damage and intrinsic apoptotic signaling. In this context, the role of reactive oxygen species (ROS) in neuronal cell death caused by ZIKV still remains elusive. Objective: We aimed at evaluating the role of these cellular components in the death of human undifferentiated neuroblastoma cell line infected with ZIKV. Results: ZIKV infection resulted in the extensive death of SH-SY5Y cells with the upregulation of several genes involved in survival and apoptotic responses as well as the colocalization of mitochondrial staining with ZIKV Envelope (E) protein. Notably, levels of intracellular reactive oxygen species (ROS) were not altered during ZIKV infection in undifferentiated SH-SY5Y cells, and consistent with these results, the treatment of infected cells with the widely studied ROS scavenger N-acetylcysteine (NAC) did not prevent cell death in these cells. Conclusion: Altogether, our results suggest that excessive ROS production is not the main trigger of SH-SY5Y cells death in ZIKV infection.


Introduction
The emergence of Zika Virus (ZIKV) in Latin America was followed by a major economic and social impact, mainly for pregnant women and for the families of children affected by Congenital Zika Syndrome [1]. As a highly neurotropic virus, ZIKV is able to infect the neural progenitor cells of unborn children and interferes with several key cellular processes, such as cell cycle [2], neuronal differentiation [3], and neuronal migration [4]. The loss of these important cellular functions induces premature neuronal death in the forming Central Nervous System (CNS) that underlies congenital neurological manifestations of ZIKV infection [5].
Apoptosis is a homeostatic cellular process that occurs during development and aging, but it also occurs as a defense mechanism following sustained proinflammatory responses or cell body damage [6,7]. In this way, programmed cell death plays a central role in controlling viral infection spread, and it can be triggered via two main pathways: (1) the extrinsic pathway, in which an external stress stimulus, e.g., proinflammatory cytokines, triggers the activation of death receptors pathways followed by caspase-8 activation; and (2) the intrinsic pathway, in which a stress signal, e.g., hypoxia-excessive ROS production, triggers mitochondrial membrane damage, releasing cytotoxic mitochondrial components (e.g., cytochrome c) followed by caspase-9 activation [8].

Cell Culture and Viral Infection
Human neuroblastoma cell line SH-SY5Y (ATCC ® CRL-2266™) was grown in Minimum Essential Medium (MEM) (Gibco/Invitrogen, CA, USA) supplemented with Ham's F12 Nutrient Mixture (Sigma-Aldrich, St. Louis, MO, USA), 10% (v/v) fetal bovine serum (FBS), 0.1 mM MEM non-essential amino acids (Gibco/Invitrogen, CA, USA), 1 mM sodium pyruvate (Sigma-Aldrich, St. Louis, MO, USA), 2 mM L-glutamine (Gibco/Invitrogen, CA, USA), 100 U/mL penicillin (Gibco/Invitrogen, CA, USA), and 0.1 g/mL streptomycin (Gibco/Invitrogen, CA, USA), at 37 • C with 5% CO 2 environment. We optimized SH-SY5Y cell culture protocol in order to minimize the number of passages in our in vitro cell culture system, and cell morphology was closely analyzed in each passage to verify the presence of its neuroblast-like characteristics morphological features [16]. For all experiments, cells were grown for 24-30 h in culture conditions before each experiment, and the maximum number of passages of cells employed in all experiments was 20.
The ZIKV PE243 strain was originally isolated in C6/36 cells from a patient with symptoms of febrile illness as previously reported [17] and was further propagated by conducting four passages in VERO cells. Viral stocks were clarified by centrifugation (2000× g at 4 • C for 10 min), filtered in Millipore Express ® PES Membrane Filters (0.22 µm) (Millipore, CO, USA), and stored at −80 • C. Virus titer was determined by plaque assay in VERO cells.
SH-SY5Y monolayers were infected with ZIKV stocks at different Multiplicities of Infection (MOI), in a small volume of culture medium containing 2% (v/v) FBS, for 1-2 h at 37 • C. After virus adsorption, cells were replenished with fresh medium containing 2% (v/v) FBS and incubated at 37 • C with 5% CO 2 . At the indicated time points, ZIKV-and mockinfected cells (medium without virus) were harvested and processed for further analysis.

ZIKV Titration in SH-SY5Y Cells by Plaque Assay
Mock-and ZIKV-infected (MOI 0.5) SH-SY5Y cell supernatant samples were harvested at 2 dpi. Confluent monolayers of VERO cells were inoculated with 10-fold serial dilution of supernatant samples from SH-SY5Y supernatant samples (each dilution point was performed in duplicate) and incubated for 2 h at cell culture conditions. After incubation, inoculums were removed, and cells were replenished with Minimum Essential Medium (MEM) containing 2.5% carboxymethyl cellulose (CMC) and 2% FBS. After an incubation period of 6 days, cell monolayers were fixed using 10% (v/v) formalin diluted in phosphate-buffered saline (PBS), stained with crystal violet solution, and plaque numbers were counted. Viral titer was calculated in PFU mL −1 using the following formula: T = P/(D × V), where P is the average number of plaques in the duplicate, D is the dilution factor, and V is the volume of diluted virus added per well.

ZIKV Replication Kinetics by Real-Time PCR
Supernatants from mock-and ZIKV-infected samples were harvested at the indicated timepoints and stored at −80 • C. Viral RNA was extracted using a QIAamp Viral RNA Mini kit (QIAGEN, Hilden, Germany). Complementary DNA (cDNA) was synthesized with a High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific, Waltham, MA, USA). A ZIKV stock with a known viral titer was used to prepare a relative standard curve for estimation of the number of viral genome copies in each reaction. A standard curve was prepared by performing 10-fold serial dilutions of the viral stock RNA sample, which was followed by relating original viral stock titer in PFU with the qPCR cycle threshold (CT) amplification of each point of the dilution curve. Quantitative RT-PCR (RT-qPCR) reactions were performed with the QuantiNovaTM Probe RT-PCR Kit (QIAGEN, Hilden, Germany), ZIKV-specific primers and probes were employed, as previously described [18]. Reactions were performed with the ABI 7500 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA) under the following cycling conditions: 45 • C for 10 min; 95 • C for 5 min; 95 • C for 5 s; and 60 • C for 45 s (45 cycles).

ZIKV E Protein Labeling and FCM Analysis
Mock-and ZIKV-infected (MOI 5) SH-SY5Y cells were harvested at the indicated timepoints, cells were fixed using a BD Cytofix/Cytoperm™ Plus Fixation/Permeabilization Kit, and intracellular antibody labeling was performed following the manufacturer's instructions (BD Biosciences, Franklin Lakes, NJ, USA). Cells were labeled with the 4G2 mouse anti-flavivirus E monoclonal primary antibody and goat anti-mouse IgG FITCconjugated secondary antibody (Sigma-Aldrich #F0257, 1:100). Samples were incubated with antibody solutions for 30 min and then washed with PBS. FCM analysis was performed using the BD FACSAriaIII Cell Sorter (BD Biosciences, Franklin Lakes, NJ, USA). Results were analyzed using BD FACSDiva (BD Biosciences, Franklin Lakes, NJ, USA) and FlowJo (TreeStar, Woodburn, OR, USA) software.

Immunofluorescence Reaction and Confocal and Epifluorescence Microscopy Assays
Mock-and ZIKV-infected (MOI 5) SH-SY5Y cells (1 × 10 5 ) grown in 0.13 mm round glass slides were processed at 2 dpi for mitochondrial staining using a Mitotracker™ Deep Red FM probe (Molecular Probes, Invitrogen, OR, USA). Briefly, cells were incubated with 500 µM of mitotracker probe for 45 min and then fixed in PBS/3% formaldehyde for 15 min at room temperature (RT). For antibody labeling, cells were fixed in PBS/4% paraformaldehyde for 15 min at RT and then permeabilized with PBS/0.2% Triton X-100 (Sigma-Aldrich, St. Louis, MO, USA) for 10 min at RT. Slides containing the fixed cells were blocked with PBS/porcine gelatin solution (0.2%) (Sigma-Aldrich, St. Louis, MO, USA) for 30 min at 37 • C. After blocking step, cells were incubated with primary antibody for 1 h at 37 • C as follows: mouse anti-flavivirus E protein monoclonal antibody 4G2 (1:10, ATCC: HB-112). Then, cells were incubated with AlexaFluor 488-conjugated donkey antimouse IgG secondary antibody (1:1000, Molecular Probes, Invitrogen, OR, EUA) for 1 h at 37 • C. Nuclei were stained with Hoechst 33342. Samples were mounted using ProLong ® Diamond Antifade Mountant (Molecular Probes, Eugene, OR, USA) or Fluoromount G mounting medium (EM Sciences, PA, USA) and analyzed using Leica DMI4000 microscope or SPII2-AOBS confocal laser scanning microscope (Leica Microsystems, Wetzlar, Germany). Images were processed and analyzed using ImageJ (National Institutes of Health, Bethesda, MD, USA). infected cells were treated 2 h before or immediately after viral inoculation. After 2 h of viral inoculation, cell culture media only (mock-treated) or containing NAC (NAC-treated) was renewed in both mock-and ZIKV-infected samples, and samples were incubated for the indicated times post-infection. Treated and mock-treated cell culture media were renewed every two days, and cells were stained with Annexin V-FITC and 7-AAD after the indicated timepoints and analyzed by FCM.  Gene expression analysis was performed using TaqMan ™ Array Human Apoptosis via Death Receptors (Thermo Fisher Scientific, Waltham, MA, USA), which consists of 96-well plates pre-configured with 42 sets of probes and primers specific to genes involved in apoptosis and survival response pathways (Supplemental Table S1) and six endogenous control candidate genes (ACTA1, ACTB, 18S, GAPDH, GUSB, HPRT). The amount of cDNA samples added per reaction was 3.5 ng and qPCR reactions were performed using an ABI 7500 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA). Gene expression analysis was performed using 7500 v2.0.6 (Thermo Fisher Scientific, Waltham, MA, USA), Microsoft Excel, and GraphPad Prism (Graph Pad Software Inc., San Diego, CA, USA) software.

SDS-PAGE and Immunoblot
Among endogenous control candidates, GUSB showed the most stable expression in our experimental conditions (standard deviation = 0.26) and was thus selected as the endogenous control gene. Analysis of gene expression levels was performed as follows: (1) Amplification CT values ≤ 36 were considered as above detection limit, and CT values ≥ 36 were considered below the detection limit and set to 36 for the purpose of gene expression analysis; (2)

Statistical Analysis
Data are shown as mean ± standard deviation (SD). Shapiro-Wilk and Kolmogorov-Smirnov normality tests were performed before statistical analysis with parametric tests. Statistical analysis was performed using Student's t-test or variance analysis (one-way or two-way ANOVA) followed by Tukey's multiple comparisons test, as indicated in the figure legends. All statistical analyses were performed using GraphPad Prism 8 software (GraphPad Software, Inc., La Jolla, CA, USA), and the results were considered significant when the p-value was below 0.05 (significance level of 5%).

ZIKV-Induced Apoptosis Is Partially Dependent on Caspase Activation in SH-SY5Y Cells
As reported previously, ZIKV efficiently infects human and mouse neuronal progenitor cells leading to cell dysfunction, apoptosis, and inhibition of neuronal differentiation [19][20][21]. Here, we used a human undifferentiated neuroblastoma cell line denominated SH-SY5Y to model ZIKV infection. First, we analyzed the virus growth in SH-SY5Y cells by infecting these cells with a high viral load (MOI 5) and detected an increase in the number of envelope (E) protein-positive cells over time, mainly between 1 and 2 dpi ( Figure 1A,B). Next, we infected cells with low (MOI 0.1) to high (MOI 5) viral loads and detected a significant increase in viral RNA levels in cell supernatant between 1 and 2 dpi, reaching a growth plateau at 3 and 4 dpi ( Figure 1C). ZIKV infection at MOI 0.5 leads to the production of infectious particles, although it was not highly productive (less than 2 × 10 4 at 2 dpi) (Supplemental Figure S1A). Infected cells showed a significantly higher death rate as evidenced by Annexin-V/7-AAD staining at 3 and 4 dpi ( Figure 1D) and by morphological changes suggestive of apoptotic cells mainly at 3 dpi ( Figure 1E). The extensive cell death rate observed at these timepoints is associated with a decline in ZIKV growth and in the rate of ZIKV-infected cells, as observed in Figure 1B,C. Next, to better characterize the intracellular processes engaged in ZIKV-induced cell death, we assessed the downstream signaling pathways that could be activated during infection. We observed a significant increase in the fluorescence levels of cleaved nuclear protein poly-(ADP-ribose)-polymerase (PARP) and SYTOX, which is a high-affinity nucleic acid stain that penetrates cells with compromised membrane integrity, mainly between 2 and 3 dpi ( Figure 1F). This process was preceded by a significant increase of activated caspase-3/7 observed as early as 2 dpi ( Figure 1F). In addition, the blockage of caspases activation with a pan-caspase inhibitor (Q-VD-OPh) significantly increased the survival rate of infected cells by nearly 50% at 3 dpi compared to media-treated cells (M) or vehicle-treated cells (V) ( Figure 1G, and Supplemental Figure S1B), and this protective effect was not related to a diminished virus replication rate (Supplemental Figure S1C). Thus, these results indicate that caspase-3/7 activation plays an important, albeit not exclusive, role in triggering cell death, and that PARP cleavage occurs subsequently to caspase-3/7 activation as a downstream process of cell death.

ZIKV Infection Triggers Intrinsic Apoptotic Pathway
We next analyzed the expression profile of genes related to survival and death signaling pathways during ZIKV infection in SH-SY5Y cells. Considering that extensive cell death is already observed at 3 dpi ( Figure 1D,E), we sought to assess the gene expression profile before cell death onset to better encompass the primary signaling pathways triggered early in ZIKV infection. Hence, gene expression profiling was performed at 1 and 2 dpi in cells infected with MOI 5, since cell death and SYTOX Red staining are almost at baseline levels at these timepoints (Figure 1D,F,G). This analysis included a total of 42 genes related to the death receptors' signaling pathway and survival as well as anti-apoptotic and proapoptotic responses (Figure 2, Supplemental Figure S2 and Supplemental Table S1). Although the cell death process is not apparent at the analyzed timepoints ( Figure 1D-F), a total of 32 genes was already significantly modulated at either 1 or 2 dpi, among which 13 genes were modulated at 1 dpi (12 genes upregulated and one gene downregulated) and 29 genes upregulated at 2 dpi (Figure 2A,B,C,E, Supplemental Figure S2 and Supplemental Table S1). Among genes upregulated by ZIKV infection, only 11 genes showed significant upregulation equal or greater than two-fold at either 1 and/ or 2 dpi ( Figure 2B-D).
Among genes involved in survival and antiapoptotic responses, BIRC3 showed strong upregulation at 1 and 2 dpi, whereas BIRC2, CFLAR, NFKB1, and NFKBIA showed significant upregulation of less than two-fold and NFKB2, NGF, and NGFR showed upregulation greater than two-fold at 2 dpi (Supplemental Figure S2A,B). Among proapoptotic genes, CASP3 was the only gene with two-fold upregulation starting at 1 dpi and increasing at 2 dpi ( Figure 2E), which also coincides with caspase-3/7 activation ( Figure 1G). APAF1, BAX, BID, DIABLO, CASP2, CASP3, CASP9, and PARP1 were also significantly upregulated during infection, and TP53 was the only proapoptotic gene slightly downregulated at 1 dpi with unremarkable modulation at 2 dpi ( Figure 2E). Although upregulation of the APAF1 gene was detected at 2 dpi, a significant increase in APAF1 protein levels was already detectable at 1 dpi ( Figure 3A,C), together with a significant increase in cleaved caspase-9 levels at 1 and 2 dpi followed by a significant decrease in procaspase-9 levels at 3 dpi ( Figure 3B,C). This is also followed by a notable intracellular redistribution of mitochondrial staining that was also observed, showing a clustered staining pattern juxtaposed to the nucleus of ZIKV-infected cells at 2 dpi (Supplemental Figure S3, lower panels) in contrast to the widespread filament-like staining pattern observed in uninfected cells (Supplemental Figure S3, upper panels). These results indicate that ZIKV infection in SH-SY5Y cells triggers caspase-9 activation and an intrinsic apoptotic pathway. Among the genes involved in the death receptors pathway with modulation of more than two-fold, CHUK, MAP3K5, and TNF were upregulated at both 1 and 2 dpi, and FAS, MAPK9, and TNFRSF10A were upregulated only at 2 dpi (Supplemental Figure  S2C-E). The expression of TNFRSF10D, TNFSF10 (TRAIL/APO2L encoding gene), and FASLG (FasL/CD95L encoding gene) were not detected in either mock-or ZIKV-infected cells (Supplemental Figure S2C,D). Seven other genes involved in the death receptors pathway were also significantly upregulated albeit less than two-fold (Supplemental Figure  S2C,E,F). Then, we analyzed intracellular levels of TNF-α protein and three other death receptors ligands, FasL/CD95L, TWEAK/ APO3L, and TRAIL/APO2L. TRAIL/APO2L levels were unaltered at 1 dpi and FasL/CD95L, TWEAK/ APO3L, and TNF-α levels remained unaltered throughout the first 48 h of ZIKV infection (Supplemental Figure  S4A-D). Moreover, despite the strong upregulation of TNF-α encoding gene at 1 and 2 dpi (Supplemental Figure S2D), soluble TNF-α was not detected in the supernatant of both mock-and ZIKV-infected cells at any of the analyzed timepoints post-infection (Supplemental Table S2). To further assess the possible role of TNF-α in ZIKV-induced SH-SY5Y cell death, we blocked TNF-α synthesis by treating mock-and ZIKV-infected cells with pentoxifylline (PTX), which is a widely used phosphodiesterase inhibitor that inhibits TNF-α synthesis possibly at the transcriptional level [22,23]. PTX treatment did not prevent death caused by ZIKV infection in SH-SY5Y cells, as quantified by LIVE/DEAD staining followed by FCM analysis (Supplemental Figure S4E,F), suggesting that TNF-α synthesis does not seem to be determinant for ZIKV-induced apoptosis in SH-SY5Y cells.  procaspase-9 levels at 3 dpi ( Figure 3B,C). This is also followed by a notable intracellular redistribution of mitochondrial staining that was also observed, showing a clustered staining pattern juxtaposed to the nucleus of ZIKV-infected cells at 2 dpi (Supplemental Figure  S3, lower panels) in contrast to the widespread filament-like staining pattern observed in uninfected cells (Supplemental Figure S3, upper panels). These results indicate that ZIKV infection in SH-SY5Y cells triggers caspase-9 activation and an intrinsic apoptotic pathway. Among the genes involved in the death receptors pathway with modulation of more than two-fold, CHUK, MAP3K5, and TNF were upregulated at both 1 and 2 dpi, and FAS, MAPK9, and TNFRSF10A were upregulated only at 2 dpi (Supplemental Figure S2C-E). The expression of TNFRSF10D, TNFSF10 (TRAIL/APO2L encoding gene), and FASLG (FasL/CD95L encoding gene) were not detected in either mock-or ZIKV-infected cells (Supplemental Figure S2C,D). Seven other genes involved in the death receptors pathway were also significantly upregulated albeit less than two-fold (Supplemental Figure  S2C,E,F). Then, we analyzed intracellular levels of TNF-α protein and three other death receptors ligands, FasL/CD95L, TWEAK/ APO3L, and TRAIL/APO2L. TRAIL/APO2L

NAC Treatment Is Not Able to Reverse Cell Death in SH-SY5Y Cells Infected with ZIKV
Considering that a common trigger of intrinsic apoptotic pathway is excessive free radical generation [24,25], we quantified ROS intracellular levels throughout the first 48 h of infection in ZIKV-infected SH-SY5Y cells. ROS levels were significantly decreased mainly at 48 hpi, and this decrease was notably observed in cells infected with higher MOI ( Figure 4A). To further assess whether ROS production would be associated with cell death in ZIKV infection, we treated mock-and ZIKV-infected cells with the antioxidant ROS scavenger NAC, and remarkably, NAC treatment did not affect cell death caused by ZIKV infection at any of the tested concentrations and analyzed timepoints post-infection, (Figure 4B,C, and Supplemental Figure S5A). Moreover, the cell death kinetics of ZIKV-infected NAC-treated samples is comparable to the cell death kinetics of ZIKV-infected mock-treated samples at any of the analyzed NAC concentrations (Supplemental Figure S5B). These results suggest that excessive ROS is probably not the main trigger of the intrinsic apoptotic pathway in SH-SY5Y cells infected with ZIKV. 21, 13, 2111 13 of 18

Discussion
Herein, we demonstrate that a contemporary South American strain of ZIKV is able to infect human undifferentiated SH-SY5Y cells, resulting in extensive cell death, mainly through the engagement of intrinsic apoptosis, further corroborating previous evidence [10]. Previous reports showed that ZIKV-induced neuronal cell death is largely dependent on caspases activation, mainly caspase-3/7 and -9, and it also involves PARP cleavage [26][27][28][29]. Our findings support these findings and also indicate that caspase-3/7 activation is an upstream step to PARP cleavage in ZIKV infection, considering that the blockage of caspase activation only partially rescued cells from death. These data suggest that ZIKV infection can also possibly result in a non-canonical (caspase independent) cell death mechanism.
Mitochondria seem to play a central role in ZIKV-induced cell death in SH-SY5Y cell line, as the redistribution of mitochondrial staining intracellular distribution is altered in ZIKV-infected cells. Previous reports showed that ZIKV and DENV nonstructural protein NS4B disturbs mitochondrial morphodynamics [30]. Changes in the mitochondrial membrane potential directly affect mitochondrial function, causing disturbances in the mitochondrial homeostasis and leading to mitochondrial cytochrome c release [31]. This is also a common feature of other flaviviruses infections. JEV, DENV2, and West Nile virus (WNV) infections induce caspase-9 activation and cytochrome c release in human medulloblastoma cells [32][33][34]. Thus, we suggest that the apoptosis of human neuroblastoma SH-SY5Y cells infected with ZIKV is mainly caused by the disruption of mitochondrial homeostasis and activation of the intrinsic apoptotic pathway.
Here, we found that TNF transcript levels were upregulated after ZIKV infection, although we did not detect significant intracellular and soluble levels of TNF-α protein, and treatment with TNF-α synthesis inhibitor PTX did not affect ZIKV apoptosis in SH-SY5Y cells. TNF-α is synthesized as an integral membrane protein, which is further cleaved in its soluble form by TNF-α converting enzyme (TACE), which is a plasma membrane member of the ADAM family of metalloproteases [35]. Considering this, our theory is that post-transcriptional regulation and silencing mechanisms may be playing a role in regulating levels of TNF-α [36][37][38][39], and further studies are needed in order to investigate this possibility. Furthermore, our results suggesting that TNF-α is not a determinant for cell death triggered in ZIKV infection are in consonance with two previous studies reporting that the inhibition of caspase-8 activation did not prevent cell death during ZIKV infection, further suggesting that the extrinsic (death receptors) apoptotic pathway is not triggered in these experimental conditions [2,10]. Moreover, a previous study reported a lack of FasL and TNF-α production during ZIKV infection in human neural progenitor cells [40], and another previous report showed that a model of primary human astrocytes also failed to produce detectable levels of TNF-α following ZIKV infection [41], further corroborating our data. Conversely, a previous report showed that ZIKV infection in murine neuronal cells leads to neurotoxicity and death and that TNF-α release plays an important role in this process [42], suggesting that inflammatory mechanisms triggered by ZIKV infection may differ across different experimental conditions. Although the death receptors pathway may not be the main trigger of cell death in our experimental conditions, we do not rule out that ZIKV infection triggers the production of proinflammatory cytokines in SH-SY5Y cells and that these may play a role in the process. Accordingly, previous data from our group showed that ZIKV infection in SH-SY5Y cells induces the production of several proinflammatory molecules including the proapoptotic chemokine CXCL10 [43], suggesting that indeed, ZIKV infection triggers an intricate inflammation process that needs to be further explored.
To the best of our knowledge, this is the first study that evaluates the autocrine role of ROS in ZIKV-induced death in human neuronal-like cells in vitro. Our data excluded excessive ROS production as a key process in ZIKV-induced death of SH-SY5Y cells. Interestingly, a previous report in human glioblastoma and human liver carcinoma cell lines showed that ZIKV induces oxidative stress with the downregulation of antioxidant enzymes [44], and another previous study about in vitro ZIKV infection in human astrocytes showed that it leads to oxidative stress that can be reversed by ascorbic acid (antioxidant) treatment [15]. Altogether, these results suggest that the oxidative stress response pathways triggered by ZIKV may be different across different cell types and needs further assessment in order to identify the particularities of each experimental setting. In fact, this corroborates the notion that each cell displays a unique redox environment and is highly dependent on the metabolism of cells and compartmentation of redox regulation components [45,46]. In this sense, one possibility is that ZIKV induces antioxidant compensatory responses that inhibit ROS detrimental effects. A previous study on WNV infection showed that ROS induction is followed by an antioxidant compensatory response that is sufficient to also overcome mitochondrial damage induced by both viral infection and Arsenite treatment (a commonly used cellular stressor) [47]. Moreover, another previous report showed that Usutu virus (USUV), a flavivirus closely related to ZIKV and DENV, showed antioxidant capacity in both murine neuroblastoma (Neuro2a) and Vero cell lines [48]. Whether ZIKV triggers a similar antioxidant mechanism to avoid oxidative stress in SH-SY5Y cells still remains to be determined.
In conclusion, our study highlights some essential aspects of the apoptotic pathway triggered by ZIKV infection in SH-SY5Y cells, and our results indicate that excessive ROS production is not the main trigger of SH-SY5Y cells death in ZIKV infection, providing a better comprehension of the apoptotic mechanism triggered by ZIKV in neuronal-like cells.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/v13112111/s1. Figure S1: ZIKV-induced apoptosis is partially dependent on caspase activation (related to Figure 2). (a) Mock-and ZIKV-infected (MOI 0.5) SH-SY5Y cell supernatant were harvested at 2 dpi, and viral titer was analyzed by plaque assay using 10-fold serial dilution of mock and infected samples. (b) Mock-and ZIKV-infected (MOI 5) SH-SY5Y cells were treated with Q-VD-OPh (10 µM), and cell viability was assessed by MTS reduction assay at 5 dpi. Data represented as mean ± SD of one experiment performed in triplicate. Statistical analysis performed by one-way ANOVA followed by Tukey's post-test, ** p <0.01, n.s.-not-significant, according to Tukey's post-test. (c) qRT-PCR analysis of viral supernatant titers of mock-and ZIKV-infected (MOI 5) cells untreated or treated with 25 µM of Q-VD-Oph (Q-VD-Oph 25), or with vehicle (DMSO), and harvested at 3 and 4 dpi. Representative data of one experiment performed in triplicate, Figure S2: Expression profile of genes involved in apoptosis and survival response pathways in ZIKV-infected SH-SY5Y cells (related to Figure 3). Mock-and ZIKV-infected (MOI 5) SH-SY5Y cells were harvested and processed at 1 and 2 dpi for gene expression analysis. Expression levels of genes related to (a) antiapoptotic and (b) survival responses, (c) death receptors, (d) death receptors ligands, (e) death receptors pathway downstream kinases, and (f) death receptors adaptor proteins. Data represented as mean ± SD of one experiment performed in duplicate. N.D.-expression not detected. Statistical analysis performed by one-way ANOVA followed by Tukey's post-test. ns-non-significant, * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001, according to Tukey's post-test, Figure S3: ZIKV infection triggers intracellular redistribution of mitochondrial marker Mitotracker DeepRed FM. Uninfected (upper panels) and ZIKV-infected (MOI 5, lower panels) SH-SY5Y cells were harvested 2 dpi and processed for fluorescence microscopy. Cells were stained with Mitotracker DeepRed FM mitochondrial probe (red) and labeled with anti-flavivirus E (4G2 primary antibody) and AlexaFluor-488 anti-mouse IgG (green). Scale bar: 20 µM. Insets magnification:2X, Figure S4: ZIKV-induced apoptosis does not require TNF-α synthesis. (a-d) Mock-and ZIKV-infected (MOI 5) SH-SY5Y cells were treated with Brefeldin A for 1 h prior to antibody labeling and then processed for intracellular cytokine labeling and FCM analysis. Representative data of two independent experiments performed in duplicate. (a) Histograms showing mean fluorescence intensity (MFI) of unstained cells (gray), mock-(blue), and ZIKV-infected cells (red) of target ligands as indicated. Intracellular time-course detection of (b) TNF-α, (c) TWEAK, and (d) FasL performed at the indicated times post-infection. Data represented as mean ± SD of one experiment performed in duplicate. ns-not significant compared to mock-infected cells. Statistical analysis was performed by one-way ANOVA followed by Tukey's post-test. (e,f) Mock-and ZIKV-infected (MOI 5) SH-SY5Y cells were treated with pentoxifylline (PTX) at 1 mM and 2 mM, harvested at 4 dpi, stained with LIVE/DEAD Viability/Cytotoxicity Kit, and analyzed by FCM. (e) Cell death percentage within the population of mock-and ZIKVinfected cells. Values represent the frequency of live cells (calcein AM high/EtHD1-) and dead cells (calcein AM low/EtHD1+). (f) Frequency (% parent) of live cells in mock-and ZIKV-infected samples untreated or treated with PTX. Data represented as mean ± SD of one experiment performed in quadruplicate. ns-not-significant. Statistical analysis was performed by one-Way ANOVA followed by Tukey's post-test, Figure S5: Mock-and ZIKV-infected (MOI 5) SH-SY5Y cells treated with NAC (0.5 mM, 0.75 mM, and 1 mM) were harvested at the indicated timepoints post-infection, stained with Annexin-VI/7-AAD, and analyzed by FCM. (a) Frequency (percentage parent) of double-positive cells (Annexin-V+/7-AAD+). Data are represented as mean ± SD of one experiment performed in triplicate. (b) Cell death kinetics of mock-or NAC-treated SH-SY5Y cells during ZIKV infection. Pre-treatment before viral inoculation; post-treatment after viral inoculation, Table S1: Expression levels of 42 genes related to survival and death responses, Table S2: Experimental details relative to supernatant samples of MOCK-and ZIKV-infected cells analyzed for TNF-α detection by ELISA immunoassay.