Several members of the herpesvirus family including human herpesvirus 6, HSV-1, and HSV-2 can elicit damaging CNS inflammation [2, 32]. Acute HSV -1 infection or the reactivation of latent virus in the trigeminal ganglion can lead to the development of severe encephalitis that is associated with a high degree of morbidity and mortality. While acyclovir is currently employed in the treatment of HSV encephalitis, drug-resistant strains of HSV-1 are beginning to emerge [33, 34]. Furthermore, despite improvements in diagnosis such infections are associated with a 30% mortality rate and 62% of survivors recover with severe neurological deficits [33, 35, 36]. The treatment of HSV-1 associated encephalitis is especially challenging due to the rapid onset of disease and development of irreversible neurological damage in otherwise healthy individuals. These characteristics suggest that the innate immune responses of resident CNS cells play a pivotal role in disease progression, a notion that is supported by the ability of human microglia to produce key inflammatory mediators in response to in vitro HSV challenge or following in vivo infection [37–40]. However, much of our current understanding of HSV-1 encephalitis pathogenesis comes from rodent models in which intranasal HSV-1 administration results in an acute necrotizing encephalitis that closely resembles human disease [41–45]. In these models, HSV-1 infects both neurons and glial cells and elicits inflammatory mediator production that precedes leukocyte infiltration into the CNS [41, 46]. Such findings are consistent with the recognized ability of other viral pathogens to induce inflammatory cytokine production by microglia  and the susceptibility and responsiveness of astrocytes to productive HSV-1 infection [8, 48]. However, the mechanisms by which resident CNS cells recognize DNA viral pathogens such as HSV-1 have not been fully defined.
We and others have demonstrated that microglia and astrocytes express an array of cell surface and endosomal innate pattern recognition receptors, including TLR2 [9, 21], TLR3 [49, 50], TLR7 , and TLR9 , that are capable of recognizing viral motifs . Importantly, each of these cell surface sensors has been implicated in the perception of HSV by a variety of cell types [13, 14, 51, 52]. However, it is becoming increasingly apparent that cells, including glia, possess intracellular sensors that can detect compromise of the cytosolic compartment. We have recently demonstrated that murine and human glial cells functionally express retinoic acid-inducible gene I (RIG-I) and melanoma differentiation-associated antigen 5 (MDA5) [23, 24], two members of the RIG-I-like family of helicases that have been shown to function as intracellular pattern recognition receptors for replicative viral RNA motifs [51, 53]. It is possible that such receptors may also indirectly serve as sensors for viral and/or bacterial DNA via the actions of RNA polymerase III  although a role for this pathway in HSV recognition by immune cells remains controversial . Interestingly, a number of cytosolic proteins including DAI have been described that can directly mediate cellular responses to dsDNA [17, 18, 53]. It has therefore been suggested that such sensors could play a critical role in the perception of viral DNA and this notion has been supported by the report that DAI mediates immune molecule production by HSV-1-infected murine fibroblasts .
In the present study, we provide the first evidence that glial cells express DAI. Resting cultures of primary microglia expressed low levels of this intracellular viral sensor, a finding that is consistent with the very low DAI expression observed in ex vivo microglia. In contrast, cultured astrocytes constitutively expressed robust levels of this molecule, although it should be noted that this might be attributable, in part, to our in vitro culture conditions, as astrocytes acutely isolated from uninfected mice expressed somewhat lower DAI levels. Importantly, DAI expression was significantly elevated in microglia and astrocytes following either in vitro or in vivo HSV-1 challenge. Such upregulation was not specific to this neurotropic alphaherpesvirus as the leukotropic gammaherpesvirus, MHV-68, was also capable of elevating DAI expression by both cell types. As such, it is possible that glial perception of DNA viruses via this sensor could promote further DAI expression in a feed-forward manner. The constitutive expression of this viral sensor by glial cells and its upregulation following viral challenge provide circumstantial evidence of a role for DAI in glial responses to DNA viral pathogens. This notion is further supported by the robust constitutive expression by microglia and astrocytes of IPS-1 , STING, and RIP3, reportedly critical downstream effector molecules for DAI [29–31]. Importantly, we have confirmed that DAI is functional in primary murine glia by showing that cytoplasmic administration of the DAI ligand, B-DNA, is a potent stimulus for the production of key inflammatory mediators by both microglia and astrocytes. Finally, we have demonstrated a major role for this intracellular viral sensor in the immune responses of primary murine glia to a clinically relevant neurotropic DNA virus by demonstrating that DAI knockdown significantly and specifically inhibits HSV-1-induced inflammatory cytokine production by these cells.
Host responses to viral CNS infections are increasingly recognized to play a major role in disease pathology and the neuroinflammation elicited by HSV-1 has been suggested to underlie the neurological damage associated with infection [3–7]. However, it not clear whether this inflammatory damage is mediated primarily by infiltrating leukocytes or by the responses of the resident cells of the CNS themselves. In support of the latter possibility, activated glial cells are known to be capable of producing toxic mediators that can cause widespread CNS damage . Furthermore, the rapidity with which HSV-1 travels from the initial site of infection to the brain all but assures escape from recognition by adaptive immune cells. Based on these observations, it appears likely that the production of cytotoxic substances by HSV-1 challenged glial cells could play a significant role in neuronal cell dysfunction and/or loss and contribute to the neuropathology associated with HSV-1 encephalitis. In the present study, we have demonstrated that soluble factor(s) released by HSV-infected microglia and astrocytes elicit neuronal cell damage/death and that the production of this factor(s) is dependent, at least in part, on the expression of DAI. While the identification of this/these neurotoxic factor(s) is ongoing in our laboratory, we have shown that TNF-a is released by HSV-1 infected glia (Figure 4) and this cytokine can elicit neuronal cell death (Figure 6). In contrast, we have found that another likely candidate, nitric oxide, is not released by either microglia or astrocytes following HSV-1 infection as assessed by culture medium nitrite content (data not shown) and is therefore unlikely to be a contributing factor to neuronal cell death. Irrespective of the mediator(s) involved, our findings directly implicate DAI in the initiation of inflammatory immune responses by glial cells and suggest a novel mechanism underlying the neuropathology associated with acute DNA viral infections of the CNS.