NOX2 ablation attenuates JE progression

The role of NOX2, a major enzyme responsible for the production of ROS, has not been elucidated in viral encephalitis following neurotrophic viruses such as JEV and WNV. While various reports have indicated contrasting alterations in resistance to viral infections due to NOX2 deficiency or inhibition [46], there have been no attempts to investigate NOX2 role in the CNS inflammation disorder induced by actual JEV infection. In order to address this issue of NOX2 in JE, we first examined the susceptibility of NOX2-deficient (NOX2 KO) mice to viral encephalitis caused by JEV infection. After WT and NOX2 KO mice were infected with JEV, surviving mice were monitored until 14 dpi (Fig. 1A). Mice in both groups showed comparable clinical signs, starting with generalized piloerection, paresis, and rigidity, followed by progression into severe neurological signs such as postural imbalance, ataxia, and generalized tonic–clonic seizure from 6 to 10 dpi. However, somewhat interestingly, NOX2 KO mice showed enhanced resistance to JE with decreased mortality after showing neurological disorders (70% mortality for WT mice vs. 15% mortality for NOX2 KO mice). Likewise, NOX2 KO mice showed reduction in loss of body weight in the course of JE (Fig. 1B), and delayed and reduced encephalitis progression was observed in NOX2 KO mice, compared to WT mice (Fig. 1C). In support of this finding, NOX2-ablated mice showed delayed and reduced proportion of neurological disorders starting from 8 dpi, compared to WT mice showing neurological disorder started from 5 dpi (Fig. 1D). Thus, this result clearly indicates that NOX2 ablation results in enhanced resistance to JE with reduction of neurological disorder presentation. To clarify clinical signs of reduced JE in NOX2 KO mice, we monitored JEV-infected mice for clinical signs depending on seven signs in the course of JE. NOX2 KO mice developed obviously less severe clinical signs from 5 to 7 dpi, compared to WT mice (Fig. 1E). To further examine the enhanced resistance of NOX2 KO mice to JE progression, we assessed viral burden within lymphoid and the CNS tissues (Fig. 1F). NOX2 KO mice were found to exhibit five fold decreased viral load in the spleen, brain, and spinal cord, which indicates that NOX2 ablation induces the reduction of JEV replication in lymphoid and the CNS tissues. Similarly, we found that NOX2 KO mice exhibited lower levels of infectious JEV in sera compared to WT mice (Fig. 1G). Therefore, these findings paradoxically suggest that ROS produced by NOX2 could exacerbate JE progression with contributing to enhanced JEV replication.

Fig. 1

NOX2-ablated mice show enhanced resistance to JE. A–C Susceptibility to NOX2-ablated mice to JE. Wild-type (WT) and NOX2 KO mice (5 to 6 weeks old, n = 10–11) were inoculated i.p. with JEV (7.5 × 107 ffu) and examined over 14 days after their survival, body weight, and encephalitis. A Curve showing survival rate; B Changes in body eight; C Encephalitis score. D Ratio of mice showing neurological disorders during JE progression. Mice infected with JEV were examined every 6 h from 5 to 11 dpi. E Clinical signs. Clinical signs of JEV-infected WT and NOX2 KO mice were monitored and scored on day 5, 6, and 7 post-infection. F Viral burden in lymphoid and inflammatory tissues during JE progression. Viral burden in lymphoid (spleen) and inflammatory tissues (brain and spinal cord) of infected mice (n = 5–6) were assessed by real-time qRT-PCR at the indicated time points. Viral RNA load was expressed as viral RNA copy number targeted on JEV NS1 gene per microgram of total RNA. G Infectious JEV burden in sera. Following JEV infection, the levels of infectious JEV were determined by a focus-forming assay using sera collected at the indicated time points. Data show the mean ± SEM of the levels derived from at least 2 individual experiments (n = 5–6). *p < 0.05, **p < 0.01, ***p < 0.001 between the levels derived from WT and NOX2 KO mice

NOX2 ablation attenuates neuroinflammation caused by JEV infection

CNS-infiltration of Ly-6C + monocytes and Ly-6G + neutrophils derived from the myeloid cells represents a hallmark neuroinflammation in the progression of JE [47]. To further characterize the CNS inflammation in NOX2 KO mice following JEV infection, we assessed the infiltration of CD11b + Ly-6C + monocytes and CD11b + Ly-6G + neutrophils. Our results revealed a significantly lower frequency of infiltrated Ly-6C + monocytes and Ly-6G + neutrophils in the brain of NOX2 KO mice at 3 and 5 dpi, compared to that observed in the brain of WT mice (Fig. 2A). Notably, NOX2 KO mice displayed greatly reduced infiltration of Ly-6C + monocytes at 5 dpi, compared to WT mice. Similarly, the absolute number of infiltrated Ly-6C + monocytes and Ly-6G + neutrophils in the brain of NOX2 KO mice decreased two- and five-fold at 5 dpi, respectively (Fig. 2B). In particular, the reduction in CNS-infiltration of Ly-6C + monocytes was more pronounced in NOX2 KO mice compared to Ly-6G + neutrophils. The involvement of microglia cells in the pathogenesis of encephalitis induced by certain neurotrophic viruses, such as WNV, has been demonstrated [48, 49]. Consequently, triple-color staining (CD11c/CD11b/CD45) was employed to discern between the resting and activated microglia. Based on the CNS myeloid cell classification of Ford et al. [50], we observed diminished frequency and absolute count of activated CD11b + CD45hi microglia in the brain of NOX2 KO mice following JEV infection. Conversely, resting CD11b + CD45int/lo microglia cells exhibited a modest decrease in NOX2 KO mice compared to WT mice (Fig. 2C and D). Additionally, NOX2 KO mice displayed a sustained reduction in the number of CD4 + , CD8 + T cells, and NK cells in the brain following JEV infection, in contrast to WT mice (Fig. 2E). These results indicate that NOX2 ablation leads to a reduction in the infiltration of inflammatory cells from the peripheral sites following JEV infection. To better understand less CNS inflammation in NOX2 KO mice, histopathological examinations were performed. Histopathological examinations revealed that decreased BBB permeability in JEV-infected NOX2 KO mice was associated with reduced observation of perivascular cuffing, compared to those of WT mice (Fig. 2F). Consistent with this observation, NOX2 KO mice exhibited a decreased inflammation score in the brain, as determined by the extent of inflammatory cell infiltration (Fig. 2G). Hence, these results suggest that the absence of NOX2 leads to a milder progression of CNS inflammation, marked by a reduced influx of inflammatory cells from peripheral sites following JEV infection.

Fig. 2

Ameliorated JE progression in NOX2-ablated mice. A, B The frequency and number of Ly-6C + monocytes and Ly-6G + neutrophils in the brain. Wild-type (WT) and NOX2 KO mice were inoculated with JEV (7.5 × 107 ffu), and the frequency (A) and total number (B) of Ly-6C + monocytes and Ly-6G + neutrophils in the brain were determined by flow cytometric analysis at 3 and 5 dpi using vigorous heart perfusion. Values in the dot-plots show the average percentage of each population after gating on CD11b + cells. C and D The frequency and number of activated and resting microglia in the brain. The frequency (C) and total number (D) of CD11b+CD45hi (activated microglia) and CD11b+CD45int/lo (resting microglia) cells were determined by flow cytometric analysis at 7th dpi. (E) Total number of lymphoid cells in the brain. The total number of lymphoid cells (CD4 + and CD8 + T cells, NK cells) were assessed on the 7th dpi. F Histopathological examinations of the brain tissues derived from infected WT and NOX2 KO mice. Brain sections of WT and NOX2 KO mice were prepared and stained with H&E 5 days after JEV infection via intraperitoneal route. Representative photomicrographs of the brain were obtained from blood vessel areas, meninges, and ventricles. G Inflammatory score of the brain tissues derived from WT and NOX2 KO mice. Inflammation was scored based on the degree of infiltration of inflammatory cells at 5th dpi. H The attenuated expression of inflammatory cytokines and chemokines in NOX2 KO mice. The expression of cytokines and chemokines in the spleen and brain was determined by real-time qRT-PCR at the indicated time points after WT and NOX2 KO mice were inoculated i.p. with JEV (7.5 × 107 ffu). The expression of cytokines and chemokines was normalized to β-actin expression and displayed as the average of at least four independent samples, according to the indicated color on a log2 scale. Data show the mean ± SEM of the levels derived from at least 2 individual experiments (n = 5–6). *p < 0.05, **p < 0.01, ***p < 0.001 between the levels derived from WT and NOX2 KO mice

Viral encephalitis caused by neurotrophic viruses is indirectly derived from CNS degeneration, stemming from robust immunological responses characterized by uncontrolled secretin of cytokines and chemokines, thereby leading to the activation of microglia and astrocytes [11, 12]. In contrast, the proper expression of cytokines and chemokines in peripheral tissues is crucial for viral clearance, thereby reducing viral burden to invade the CNS tissues [8, 10, 51]. Thus, measuring the expression of cytokines and chemokines in peripheral lymphoid and CNS tissues can offer valuable insights into the progression of JE in NOX2 KO mice. To gain a deeper understanding of JE progression in NOX2 KO mice, we examined the expression of cytokines and chemokines in the primary target lymphoid organ, the spleen, and the brain, a key inflammatory CNS tissues, following JEV infection. We found that JEV infection in NOX2 KO mice led to significantly reduced expression of cytokines and chemokines in the peripheral lymphoid tissues (spleen) during the early stage (1–3 dpi) after JEV infection (Fig. 2H). Consequently, NOX2 KO mice exhibited lower expression of cytokine and chemokines in CNS tissues than WT mice during the later stage when neurological symptoms manifested. This outcome was likely influenced by the viral burden in spleen and brain. Furthermore, the reduced expression of chemokines, such as CCL2, CCL5, and CXCL2, in the brain of NOX2 KO mice was thought to contribute to the reduction of inflammatory cell infiltration in the brain, ultimately leading to less neuroinflammation.

NOX2-ablated mice show enhanced JEV-specific T cell responses with reduced CD4 + Foxp3 + Treg number

Antiviral adaptive immune responses, mediated by effector Ag-specific CD4 + and CD8 + T cells, are required for the regulation of JE progression by controlling and clearing JEV in extraneural tissues and the CNS [13, 14, 52]. When we examined the frequency and number of CD4 + Foxp3 + Treg cells in both WT and NOX2 KO mice, WT mice showed slightly increased frequency levels of CD4 + Foxp3 + Treg cells at 3 and 5 dpi, as previously described [52], but the frequency of CD4 + Foxp3 + Treg cells was not changed in NOX2 KO mice (Fig. 3A). Conversely, NOX2 KO mice exhibited a reduction in the frequency of CD4 + Foxp3 + Treg cells in response to JEV infection. In support, NOX2 KO mice contained significantly decreased number of CD4 + Foxp3 + Treg cells in the spleen with the levels peaked at 3 dpi, compared to WT mice. Although some of both WT and NOX2 KO mice infected with JEV exhibited neurological disorders at 5–6 dpi, a time before fully induced functional adaptive immune responses, we examined the generation of JEV-specific CD4 + and CD8 + T cell responses in surviving WT and NOX2 KO mice at 7 dpi. Our results revealed that NOX2 KO mice exhibited significantly increased frequencies of JEV-specific CD4 + T cells producing IFN-γ and TNF-α upon stimulation with the epitope peptides NS1132-145 and NS3563-574, compared to those in WT mice (Fig. 3B). Also, IFN-γ + CD4 + and TNF-α + CD4 + T cell number were higher in the spleen of NOX2 KO mice upon stimulation with JEV CD4 + T cell epitope peptides at 7 dpi, compare to WT mice (Fig. 3C). Similarly, NOX2 KO mice exhibited significantly increased JEV-specific CD8 + T cell responses, as shown by higher frequency and number of IFN-γ and TNF-α-producing CD8 + T cells upon stimulation with CD8 + T cell epitope peptide NS4B115–223 (Fig. 3D and E). Therefore, the reduced population of CD4  Foxp3 + Treg cells is presumed to contribute to the heightened Th1 CD4 + and CD8 + T cell responses producing IFN-γ and TNF-α upon JEV infection. These responses may play a role in controlling JE progression in NOX2 KO mice by reducing viral burden at the periphery during later stage. In contrast with enhanced JEV-specific Th1 CD4 + and CD8 + T cell responses in NOX2 KO mice, JEV E protein-specific IgM and IgG in sera were found to be detected at similar levels in both NOX2 KO and WT mice (Fig. 3F), which indicate that NOX2 ablation has no discernible impact on the induction of JEV-specific humoral immune responses.

Fig. 3

NOX2-ablated mice show enhanced JEV-specific T cell responses. A The frequency and total number of CD4 + Foxp3 + Tregs in the spleen of NOX2 KO mice. CD4 + Foxp3 + Tregs in the spleen of Wild-type (WT) and NOX2 KO mice were detected with intracellular Foxp3 and surface CD4 staining at 3 and 5 dpi. B and C JEV-specific CD4 + T cell responses. D and E JEV-specific CD8 + T cell responses. The splenocytes were prepared from surviving WT and NOX2 mice 7 days following infection with JEV and used for stimulation with JEV epitope peptide of CD4 + T cells (NS1132-145, NS3563-574) or CD8 + T cells (NS4B215-223) for 12 or 8 h, respectively. The frequency and absolute number of JEV-specific CD4 + and CD8 + T cells were determined by intracellular cytokine (IFN-γ and TNF-α) staining combined with surface CD4 and CD8 staining. F Serum levels of JEV E protein-specific IgM and IgG. Levels of JEV E protein-specific IgM and IgG in sera were determined by conventional ELISA using sera collected from surviving mice at 7 dpi. Values in representative dot-plots denote average percentage of indicated cell population. Bar charts show the mean ± SEM of the levels derived from at least 2 individual experiments (n = 5–6). *p < 0.05, **p < 0.01, ***p < 0.001 between the levels derived from WT and NOX2 KO mice. #p < 0.05, ##p < 0.01, ###p < 0.001 between the levels derived from mock and JEV-infected mice

Attenuated JE progression in NOX2-ablated mice is linked to reduced ROS production with the accumulation of M1-like macrophages

Macrophages are heterogenous immune cells differentiated from Ly-6C + and Ly-6C- monocytes and play an important role in neuroinflammation caused by neurotrophic viruses [18,19,20]. After viral infection, Ly-6C + monocytes can differentiate into M1 or M2 macrophages at inflammation sites depending on surrounding microenvironment, and thereby impacting the progression of inflammation caused by viral infection [27, 28]. Recent findings have indicated that ROS can influence M1 or M2 macrophage polarization [38]. Therefore, we examined whether ROS produced by NOX2 ultimately contribute to macrophage polarization, and consequently played a role in the resistance of NOX2 KO mice to JE progression. To this end, we examined ex vivo polarization of macrophages derived from peritoneal cavity and spleen of JEV-infected WT and NOX2 KO mice in response to brief LPS stimulation. Our results revealed that NOX2 KO mice exhibited more enhanced accumulation of M1-polarized macrophages in the peritoneal cavity and spleen, compared to WT mice (Fig. 4A). Specifically, IL-12p40 and iNOS-positive M1-polarized CD11b + F4/80 + macrophages were detected in the peritoneal cavity and spleen of NOX2 KO mice with higher levels than WT in response to brief LPS stimulation. In support, NOX2 KO mice contained more accumulated number of IL-12p40 and iNOS-producing M1-polarized macrophages in the peritoneal cavity and spleen, compared to WT mice (Fig. 4B). However, the accumulated number of IL-10-producing M2 macrophages was comparable in both WT and NOX2 KO mice. Hence, this finding indicates that NOX2 deficiency leads to an increased accumulation of IL-12p40 and iNOS-producing M1-polarized macrophages in primary inflammation and lymphoid tissues during JE progression following i.p. infection with JEV. Moreover, our result revealed that the reduced ROS production in sera of NOX2 KO mice was closely linked to an increased accumulation of M1 macrophages (Fig. 4C). Also, CD11b + F4/80 + macrophages derived from the peritoneal cavity and spleen of NOX2 KO mice displayed lower production of total ROS compared to those derived from WT mice (Fig. 4D), suggesting a strong correlation between reduced ROS production in macrophages and enhanced M1 polarization. To further confirm the impact of NOX2 on macrophage polarization, we prepared BMDM derived from BM cells of WT and NOX2 KO mice using GM-CSF and induced to undergo M1 polarization through stimulation with LPS and IFN-γ [53, 54]. Consequently, BMDM derived from BM cells of NOX2 KO mice exhibited more enhanced M1 polarization, as shown by increased IL-12p40 and iNOS-producing CD11b + F4/80 + macrophages in response to stimulation with LPS and IFN-γ (Fig. 4E). In support, BMDM derived from NOX2 KO mice displayed higher expression of M1 effector molecules (IL-12p40, IL-6, TNF-α, iNOS, CXCL9), compared to those of WT mice (Fig. 4F). However, the expression of M2 effector molecules (IL-4Rα, CD206, Fizz-1, Ym-1) showed no changes in BMDM derived from NOX2 KO mice, which indicates that NOX2 ablation may not affect M2 macrophage polarization. Also, IRF5, which plays a significant transcription role in the de novo differentiation of M1 macrophage by GM-CSF [52], demonstrated sustained upregulation in BMDM derived from NOX2 KO mice. It was also curious that IRF4, involved in the de novo differentiation of M2 macrophages by M-CSF [52, 53], exhibited notably heightened expression in NOX2-deficient BMDM. In addition, BMDM derived from NOX2 KO mice was found to produce TNF-α, IL-6, and IL-12p70 with higher levels in culture media, compared to those of WT mice (Fig. 4G). Collectively, these results suggest that the reduced ROS production due to NOX2 ablation leads to enhanced M1 polarization of macrophages in the peripheral inflammatory sites (peritoneal cavity) and lymphoid tissues (spleen), thereby contributing to the control of JE progression in NOX2 KO mice.

Fig. 4

Attenuated JE progression is associated with increased accumulation of M1 macrophages in NOX2-ablated mice. A Increased accumulation of M1 macrophages in peritoneal cavity and spleen of NOX2 KO mice. B Accumulated number of M1 macrophages in peritoneal cavity and spleen. Leukocytes were prepared from the peritoneal cavity and spleen of Wild-type (WT) and NOX2 KO mice 1 and 3 dpi, and briefly stimulated with LPS (200 ng/ml) for 6 h. M1 macrophages were then detected by intracellular IL-12p40 and iNOS staining combined with surface CD11b and F4/80 staining. C Reduced ROS levels in sera of JEV-infected NOX2 KO mice. ROS levels were determined by ROS ELISA kit using sera derived from JEV-infected WT and NOX2 KO mice at 2nd dpi. D Intracellular ROS levels of CD11b + F4/80 + macrophages derived from WT and NOX2 KO mice. Intracellular ROS levels in macrophages derived from peritoneal cavity and spleen were assessed using total ROS assay kit 2 days after JEV infection. Values in histograms show the average MFI ± SEM of total ROS in macrophages population after gating on CD11b + F4/80 + cells. E Enhanced M1 polarization of NOX2-deficient macrophage. BMDM derived from BM cells of WT and NOX2 KO mice were stimulated with LPS and IFN-γ for 6 and 12 h to drive M1 polarization. The M1 polarization of BMDM was evaluated by intracellular IL-12p40 and iNOS staining combined with surface CD11b and F4/80 staining. F Increased expression of M1 effector molecules in NOX2-deficient macrophages. Total RNA extracted from M1-polarized BMDM driven from WT and NOX2 KO mice was used for determining the expression of M1 and M2 effector molecules with real-time qRT-PCR. The expression of M1 and M2 effector molecules was normalized to β-actin expression and displayed as the average of at least four independent samples, according to the indicated color on a log2 scale. G Higher production of M1 effector cytokines from NOX2-deficient macrophages. BMDM derived from WT and NOX2 KO mice was stimulated with LPS and IFN-γ form the indicated times and the production of M1 effector cytokines (TNF-α, IL-6, IL-12p70) was determined by cytokine ELISA using culture media. Data show the mean ± SEM of the levels derived from at least 2 individual experiments (n = 5–6). *p < 0.05, **p < 0.01, ***p < 0.001 between the levels derived from WT and NOX2 KO mice

NOX2 ablation facilitates de novo differentiation of M1 macrophages to activate Th1 CD4 + T cells

Monocyte-derived M1 and M2 macrophages exhibit differential expression of surface markers. Consequently, M1 macrophages predominantly express high levels of CD80, CD86, MHC II, and TLR2, whereas CD206 and CXCR3 are highly expressed in M2 macrophages [55, 56]. Therefore, in order to further investigate the facilitated induction of M1 macrophage polarization due to NOX2 deficiency, the expression of surface molecules in M1 and M2 macrophages was analyzed. As anticipated, BMDM derived from NOX2 KO mice exhibited a significant increase in the expression of M1 macrophage surface markers, including CD80, CD86, MHC II, and TLR2, when compared to BMDM from WT mice (Fig. 5A). Conversely, the expression of M2 macrophage surface markers, CD206 and CXCR3, showed a slight decrease in BMDM derived from NOX2 KO mice. Furthermore, we examined the Ag-presenting capacity of BMDM derived from NOX2 KO mice, because the expression of molecules associated with Ag presentation (CD80, CD86, MHC II) was upregulated in M1-polarized macrophages [55, 56]. We conducted coculture experiments using CD4 + T cells isolated from OT-II Tg mice and BMDM derived from NOX2 KO and WT mice in the presence of OT-II CD4 + T cell epitope peptide (OVA323–339). Our results revealed that BMDM derived from NOX2 KO mice induced a stronger activation of OT-II CD4 + T cells. Notably, BMDM derived from NOX2 KO mice exhibited a significant increase of CD4 + T cells producing the autocrine growth factor IL-2 in the presence of OVA323-339 peptide when compared to those of WT mice (Fig. 5B). Similarly, BMDM derived from NOX2 KO mice induced increased production of IFN-γ from OT-II CD4 + T cells upon stimulation with OVA323-339 peptide, which indicates that NOX2-deficient BMDM facilitates Th1 differentiation of CD4 + T cells (Fig. 5C). When examining the expression of surface molecule indicatives of CD4 + T cell activation upon Ag presentation, BMDM derived from NOX2 KO mice were found to induce increased expression of CD44, CD154, as well as higher expression of CD25 and CD69, along with reduced expression of CD62L (Fig. 5D). This demonstrates that BMDM derived from NOX2 KO mice exhibit a stronger induction of OT-II CD4 + T cell activation compared to BMDM from WT mice. The primary ROS source initiates with the formation of a superoxide (O2−) and subsequently leads to the production of hydrogen peroxide (H2O2). This process is closely associated with NOX family, encompassing NOX1, NOX2, NOX3, NOX4, NOX5, DUOX1 and DUOX2 enzyme [57]. Thus, to ascertain the influence of H2O2 produced by NOX2 on de novo M1 and M2 macrophage polarization, we examined the effect of H2O2 addition on the de novo polarization of macrophages into M1 and M2 phenotypes. H2O2 addition in culture of BMDM derived from NOX2 KO mice markedly suppressed the production of M1 macrophage-related cytokines, IL-12p70, TNF-α, and IL-6 (Fig. 5E), which indicates that H2O2 produced by NOX2 could inhibit de novo polarization of macrophages into M1 phenotype. Furthermore, we investigated whether JEV infection could enhance M1 polarization of BMDM derived from NOX2 KO mice in the presence of IFN-γ. As a result, JEV infection was found to increase IL-12p40 and iNOS-producing macrophages in BMDM derived from NOX2 KO mice (Fig. 5F). Additionally, analysis of M1 macrophage-related cytokines, including IL-12p70, TNF-α, and IL-6, revealed that BMDM derived from NOX2 KO mice produced higher quantities of these cytokines in response to JEV infection (Fig. 5G). M1-polarized macrophages play a crucial role in inhibiting the replication of various viruses by producing antiviral cytokines and enhanced phagocytosis [33,34,35,36]. Therefore, we examined JEV replication in M1-polarized BMDM derived from NOX2 KO mice. As anticipated, it was observed that JEV replication in NOX2-deficient BMDM was significantly more inhibited than in BMDM derived from WT mice (Fig. 5H). Hence, these results suggest that macrophages derived from NOX2 KO mice undergo enhanced M1 macrophage polarization in response to JEV infection, thereby contributing to viral clearance at the early stage.

Fig. 5

NOX2 is negative regulator of de novo differentiation of M1 macrophages. A The expression of M1 and M2 surface markers in NOX2-deficient BMDM. BMDM derived from BM cells of WT and NOX2 KO mice were stimulated with LPS and IFN-γ for 12 and 24 h to drive M1 polarization, and used for evaluating the expression of surface markers by flow cytometric analysis. Values in the histograms show the average MFI ± SEM of each surface molecule expression in macrophage population after gating on CD11b + F4/80 + cells. B and C Enhanced Ag-presentation of NOX2-deficient BMDM. BMDM derived from the BM cells of WT and NOX2 KO mice was either stimulated with (M1) or without (M0) LPS and IFN-γ for 12 h, and co-cultured with purified OT-II CD4 + T cells at the varying ratios in the presence of OVA323-339 peptide for 24 h and 72 h to assess the generation of IL-2 and IFN-γ-producing cells, respectively. The Ag-presentation of BMDM to OT-II CD4 + T cells was evaluated by intracellular IL-2 and IFN-γ staining combined with surface CD4 staining. Values in dot-plots show the mean ± SEM percentage of IL-2 and IFN-γ-producing CD4 + T cells after gating on CD4 + cells. D Enhanced activation of OT-II CD4 + T cells stimulated with NOX2-deficient BMDM. The activation markers of OT-II CD4 + T cells stimulated with BMDM derived from WT and NOX2 KO mice were determined by flow cytometric analysis after surface staining with CD4 and activation marker antibodies. Data was expressed by the average MFI ± SEM levels of each activation marker. E Suppression of M1 macrophage polarization by NOX2-generated H2O2. BMDM derived from WT and NOX2 KO mice was stimulated with LPS and IFN-γ in the absence or presence of H2O2 (300 µM) for 18 h. The production of M1 effector cytokines (IL-12p40, TNF-α, IL-6) was determined by cytokine ELISA using culture media. F Facilitated de novo differentiation of M1 macrophage by JEV infection. BMDM derived from BM cells of WT and NOX2 KO mice were infected with JEV (5 MOI) in the presence of IFN-γ (10 ng/ml) for 48 h. The M1 polarization of infected BMDM was evaluated by intracellular IL-12p40 and iNOS staining combined with surface CD11b and F4/80 staining. G Higher production of M1 effector cytokines from NOX2-deificent macrophage by JEV infection. BMDM derived from WT and NOX2 KO mice was infected with JEV in the presence of IFN-γ, and the production of M1 effector cytokines (TNF-α, IL-6, IL-12p70) was determined by cytokine ELISA using culture media 48 h later. H Suppressed replication of JEV in NOX2-deficient macrophages. BMDM derived from WT and NOX2 KO mice were infected with JEV (1.0 MOI). Total RNA was extracted from infected BMDM and subjected to determine JEV replication with real-time qRT-PCR at the indicated time points post-infection. JEV RNA replication was expressed as viral RNA copy number targeted on JEV NS1 gene per microgram of total RNA. Data show the mean ± SEM of the levels derived from at least 2 individual experiments (n = 5–6). *p < 0.05, **p < 0.01, ***p < 0.001 between the levels derived from WT and NOX2 KO mice

Administration of ROS scavenger enhances resistance to JE

Ultimately, our results suggest that ROS produced by NOX2 hinder M1 macrophage polarization, thereby preventing the proper induction of viral clearance in the peripheral tissues and potentially promoting JE progression. Butylated hydroxyanisole (BHA), known as a ROS scavenger, has been utilized in various diseases and health supplements with underscoring its potential utility in this context [58, 59]. Therefore, we became interested in whether BHA possesses inhibitory capabilities against JE progression caused by JEV infection. To assess the potential impact of BHA on the JE progression, we infected WT mice and orally administered BHA at a daily dose of 300 mpk. Subsequently, we observed changes in susceptibility to JE progression caused by JEV infection. Our results showed that BHA administration reduced susceptibility to JE progression caused by JEV infection. Specifically, WT mice administered with BHA showed approximately 30% mortality in response to JEV infection, while WT mice administered with a vehicle (corn oil) showed around 70% mortality (Fig. 6A). Furthermore, the body weight loss induced by JEV infection was less pronounced in WT mice treated with BHA (Fig. 6B), and symptoms of encephalitis were also less severe in BHA-treated mice (Fig. 6C). In support of this finding, BHA-treated WT mice showed reduced proportion of neurological disorders till 11 dpi, compared to vehicle-treated WT mice (Fig. 6D). Thus, this result indicate that BHA administration increases resistance to JE progression induced by JEV infection with reducing the manifestation of neurological symptoms. To clarify the effect of BHA on JE progression, we monitored BHA-treated WT mice for clinical signs depending on seven signs in the course of JE. Our results showed that BHA administered WT mice exhibited significantly reduced clinical signs from 5 to 7 dpi compared to vehicle-treated WT mice (Fig. 6E). To further examine the effect of BHA administration on JE progression, we assessed viral burden within lymphoid and the CNS tissues. As expected, BHA-administered WT mice contained less JEV burden in peripheral lymphoid tissue (spleen) and the CNS tissues (brain and spinal cord) (Fig. 6F). In particular, WT mice administered with BHA showed lower JEV detection in the peripheral lymphoid tissue during the early stage of infection and lower JEV detection in the CNS tissues later on. Therefore, these results paradoxically suggest that BHA administration as ROS scavenger inhibits JEV replication in the peripheral tissues, ultimately reducing the viral load invading the CNS, and lowering manifestation of acute encephalitis.

Fig. 6

Administration of ROS scavenger attenuates JE progression. A–C Increased resistance of BHA-administered mice to JE. Wild-type (WT) C57BL/6 mice