Inducible nitric oxide synthase deficiency promotes murine-β-coronavirus induced demyelination

Multiple sclerosis (MS) is the most widely studied neuroinflammatory and demyelinating disorder of the CNS and is common among young adults. It is characterized by the destruction of myelin and myelin-producing oligodendrocytes. The disease etiology is associated with autoimmunity, several genetic and environmental factors, and pathogens such as viruses. A recent groundbreaking study by Bjornevik et al. [1] suggested Epstein Barr virus (EBV) as one of the leading causes of MS confirming the involvement of virus etiology. Neuroinflammation in MS is mediated by the production of several inflammatory mediators, including cytokines, chemokines, and reactive oxygen and nitrogen species by activated glial cells, such as microglia and astrocytes, and CNS infiltrating myeloid and lymphoid cells [2, 3]. In response to an inflammatory stimulus, inducible nitric oxide synthase (iNOS or NOS2) catabolizes L-arginine to generate nitric oxide (NO) that has been reported to significantly contribute to the oligodendrocyte degeneration, demyelination, and neuronal damage in MS [4].

During inflammation, NOS2 is induced in all leukocytes, including macrophages, dendritic cells, NK cells, primary tumor cells, and to a certain extent, activated T cells [5]. NOS2 expression by microglia and macrophages is a prominent pro-inflammatory marker of neuroinflammation and is implicated in MS pathophysiology [4].

MS patients revealed NOS2 expression in the chronic active plaques in diverse cell types, including microglia/macrophages (MG/Mφ), ependymal cells, inflammatory cells, and to a certain extent astrocytes [6]. Studies using experimental animal models of MS have shown mixed results. Several experiments report improved disease symptoms upon inhibition of NOS2 or NOS2 deficiency [7‐12], while in others, it rendered the system vulnerable to pathogen attack or exacerbated the disease [13‐16]. Though studies implicating NOS2 in demyelination are limited, most indicate no effect or a detrimental effect of NOS2.

In this study, we investigated the effect of NOS2 on the pathological outcome of neuroinflammation and neuroinflammatory demyelination induced by RSA59, a spike gene recombinant strain of murine β-coronavirus (MHV-A59), in an effort to better understand the NOS2-mediated pro-inflammatory immune response [17]. RSA59 infection induces a biphasic disease in mice characterized by acute phase (day 5/6 post-infection [p.i.]) hepatitis, meningitis, encephalitis, and successive chronic phase (day 15–30 p.i.) demyelination and concomitant axonal loss [17‐22]. We previously showed an interplay between MG/Mφ and CD4 + T cells via CD40 receptor (CD40) and CD40 ligand (CD40L) interaction in the RSA59 induced model. The absence of either CD4 + T cells or CD40L impaired viral clearance from the CNS and aggravated the severity of demyelination in the chronic phase [23, 24]. Confirming evidence of the protective role of CD4 + T cells was further illustrated in a parallel study performed in Ifit2-/- mice. RSA59 infected Ifit2-/- mice presented with dampened activation of MG/Mφ resulting in low recruitment of CD4 + T cells in the CNS, which led to severe clinical disease and high mortality in the mice [25]. Having studied the adaptive arm and its crosstalk with the innate arm during the RSA59-induced immune response, we next examined the role of innate immune MG/Mφ to investigate the NOS2-mediated proinflammatory immune response. We have compared 4–5-week-old male C57BL/6 mice with NOS2-/- mice for pathological outcomes on RSA59 infection.

In the current study, we showed that NOS2 expression was upregulated at the acute phase of RSA59 infection and its absence led to more severe clinical disease score as compared to WT mice. Most strikingly NOS2-/- mice exhibited a drop in survival at the acute-adaptive transition phase (day 9/10 p.i.) associated with significant increase in the infiltration of macrophages and neutrophils in the brain. Resultantly, the transcript levels of TNFα, IFNγ and Ifit2 were higher in the NOS2-/- brains. However, the virus was successfully cleared from the CNS in both NOS2-/- and wildtype mice. In addition, demyelination pathology was observed as early as day 9/10 p.i. in infected NOS2-/- mice as compared to day 30 p.i. in the WT infected mice. The demyelinating plaques showed accumulation of amoeboid phagocytic MG/Mφ which was further confirmed by significant upregulation of phagocytic markers- CD206, TREM2 and Arginase1 in the CNS. Confocal imaging of spinal cords at day 9/10 p.i. showed engulfment of myelin basic protein (MBP) and proteolipid protein (PLP) by Iba1 + MG/Mφ. At the chronic phase as well NOS2-/- mice showed exacerbated demyelination. The results together indicated a protective role of NOS2 against murine β-coronavirus-induced demyelination.

The role of NOS2 in neurodegeneration has been studied in murine experimental models of MS utilising demyelination-inducing chemicals, peptides, or virus strains, the outcomes of which often depend on differential disease pathology and the role of infiltrating T cells. In the early 2000s Lane and Perlman laboratories showed that absence of NOS2 does not affect demyelination [11, 32]. These studies were carried out incorporating MHV-JHM, a highly neurovirulent strain of MHV as compared to RSA59 [33]. In the neurotoxicant cuprizone-induced demyelination model, on the other hand, absence of NOS2 exacerbated demyelination at early time points [15]. As compared to JHM where mice succumb to death at day 12 p.i. due to severe demyelination, RSA59 infection in WT mice causes prominent demyelination around day 30 p.i. allowing us to dissect the immune responses involved over time. The current study was thus initiated to revisit the role of NOS2 in demyelination using MHV-RSA59 infection as the model.

RSA59 infection in C57BL/6 WT mice led to upregulation of NOS2 mRNA and protein at the acute phase, i.e., day 5/6 p.i. confirming its role in RSA59-induced acute neuroinflammation. Further NOS2 deficiency led to increased mortality and uncommon demyelination pathology at the acute-adaptive transition phase. However, it did not affect viral clearance, which is in line with the above reports. We have previously reported the crucial role of CD4 + T cells in mounting anti-viral immune response against RSA59. In the absence of either CD4 + T cells or their activation marker CD40L which mediates the MG/Mφ-CD4 + T cell crosstalk, the virus persisted in the CNS. This in turn lead to continual activation of MG/Mφ concurrent with high expression of phagocytic markers to eliminate both the debris and virus-infected cells resulting in severe chronic demyelination as a bystander effect [23, 24]. In addition, ligation of CD40 on microglial cells has been reported to induce NOS2 expression [34]. In the absence of NOS2 the crosstalk between MG/Mφ and CD4 + T cell was not affected as evident from successful viral clearance and comparable CD4 + T cell numbers in infected NOS2-/- mice, however there was an increased infiltration of macrophages and neutrophils in infected NO2-/- mice brains at the acute-adaptive transition phase. The infected NOS2-/- brains also showed higher levels of inflammatory markers TNFα, IFNγ and Ifit2 which could be attributed to the increased infiltration of macrophages and neutrophils though the inflammatory status needs to be reconfirmed at protein expression level as well. Many studies have reported direct role of NO in inhibiting leukocyte recruitment by reducing the leukocyte-endothelial interactions [35]. Thus, lower levels of NO in NOS2-/- mice could be one of the reasons for high macrophage and neutrophil infiltration. A slightly similar alteration in myeloid cell infiltration was also reported in a relatively recent study in the experimental autoimmune encephalomyelitis (EAE) model of MS where it was showed that in the absence of NOS2 there was high recruitment of pathogenic CD11b + F4/80 − Gr-1 + cells and low infiltration of regulatory CD11b + F4/80 − cells in the brain during the priming phase and in the spinal cord during the effector phase [36]. This indicated that effect of NOS2 was phase-dependent which is also apparent in the current study. NOS2 production is highest at the acute phase (day 5/6 p.i.), its deficiency however did not affect CNS pathology during acute neuroinflammation (day 5/6 p.i.) but resulted in early demyelination at the acute-adaptive transition phase (day 9/10 p.i.) implying that the system adopted auxiliary mechanisms which could not be sustained over time. Demyelination pathology at the acute-adaptive transition phase in the spinal cords of NOS2-/- mice was accompanied by amoeboid Iba1 + MG/Mφ. Early studies by Wu and Perlman in MHV-JHM model have shown a correlation of increase in the number of F4/80 + macrophages with demyelination and showed the presence of distinct activated ameboid F4/80 + macrophages in the demyelinating plaques [33]. The early demyelination observed in the RSA59 infected NOS2-/- mice was accompanied by significant upregulation of MG/Mφ phagocytic markers CD206, TREM2 and Arg1 in the CNS confirming the anti- inflammatory phagocytic phenotype of MG/Mφ [37‐41]. The phagocytic markers were not further verified at protein expression level however, confocal imaging of the demyelinating plaques confirmed considerably more loss of MBP due to engulfment by Iba1 + MG/Mφ in infected NOS2-/- spinal cord sections. A lucid study in the EAE induced model of MS targeted the expression of NOS2 and Arg1 by mononuclear phagocytes showing that the cells specify and adapt to their phenotype locally guided by the CNS-derived signals and can change their individual phenotypes over time accordingly [42]. Another aspect of regulation of polarization in macrophages was shown by Van den Bossche et al. [43] where inhibition of NOS2 led to the rescue of NO suppressed mitochondrial respiration and enhanced inflammatory to anti-inflammatory phenotypic repolarization of macrophages. This proved that the local redox and immune milieu is important in deciding the fate of the cells and thereby the pathological outcome. It can thus be concluded that changes in brain inflammatory milieu due to more influx of macrophages and neutrophils led to early progression of MG/Mφ into an amoeboid phagocytic phenotype responsible for early demyelination in the absence of NOS2. Though anti-inflammatory phagocytic phenotype has been shown to be beneficial in various inflammatory diseases [44], persistent activation could damage the healthy tissue as observed by RSA59-induced early demyelination in this study which worsened by day 30 p.i. at which point majority of the myelin MBP was stripped off from the white matter axons. Current study thus suggests that the absence of NOS2 led to the loss of regulation in the recruitment of macrophages and neutrophils leading to enhanced inflammation which potentially caused the increased disease severity and mortality in the mice.

Contrary to the previously reported pathogenic role of NOS2 in MS and related autoimmune animal models, we show that β-coronavirus-RSA59 infected NOS2 deficient mice exhibit exacerbated disease with accelerated demyelination. This was accompanied by increased accumulation of neutrophils and macrophages and a high prevalence of phagocytic MG/Mφ. We thus show a protective role of NOS2 in RSA59-induced demyelination and the study adds to the limited knowledge of the role of NOS2 in modulating the demyelination process in experimental animal models of MS.