Herpes Simplex Virus Encephalitis: Toll-Free Access to the Brain

Abstract

Herpes simplex encephalitis (HSE) is a devastating infection of the central nervous system (CNS). Lafaille et al. (2012) show that susceptibility to HSE is due to mutations in Toll-like receptor response pathways that directly reduce the intrinsic resistance of neurons and other cells in the CNS to HSV infection.

Herpes simplex virus type 1 (HSV-1) is a highly evolved pathogen capable of a benign coexistence with its human host. Life-threatening complications are rare, especially given its near ubiquity in adult human populations. The high prevalence of HSV infection is fueled in part by the ability of the virus to establish life-long latent infections, which are punctuated by intermittent reactivation events (Nicoll et al., 2012). These reactivations are frequently asymptomatic and allow the virus to be shed and infect other susceptible individuals. The most common exceptions to this host-pathogen stalemate are sight-threatening corneal infections (stromal keratitis) and life-threatening central nervous system (CNS) infections (encephalitis). Herpes simplex encephalitis (HSE) predominantly affects children and the elderly, is one of the most common forms of viral encephalitis, and has remarkably poor outcomes despite the availability of good antiviral therapy (Sabah et al., 2012). Importantly, HSE can result from both primary infection and reactivation of latency, and until recently we had few clues as to why some children seem specifically susceptible to HSE. Certain inherited deficiencies of the innate immune system have now been shown to correlate strongly with childhood HSE (Casrouge et al., 2006; Herman et al., 2012; Pérez de Diego et al., 2010; Zhang et al., 2007). In a recent study, Lafaille et al. (2012) suggest that susceptibility to HSE is caused, at least in part, by defects in intrinsic innate responses of the nervous system itself.

Toll-like receptor (TLR) signaling is critical for the recognition of, and response to, a variety of pathogens. A number of TLRs are important for resistance to HSV in mice, but a critical aspect of the research of Casanova and others for some years has been the identification of TLR-driven pathways critical for HSV resistance in humans. Their initial observations were that an autosomal recessive mutation in UNC-93B, an endoplasmic reticulum (ER) protein with 12 membrane-spanning domains, renders children susceptible to HSE through impaired interferon (IFN) responses (Casrouge et al., 2006). UNC-93B is essential for TLR3, TLR7, and TLR9 signaling because it is required to deliver these TLRs from the ER to the endosome, where they recognize pathogens and initiate signaling cascades that activate immune responses (Kim et al., 2008; Tabeta et al., 2006). While an UNC-93B deficiency was apparently specific for resistance to HSV in humans, UNC-93B^−/− mice were susceptible to a variety of microbes. This important difference between the susceptibilities of UNC-93B^−/− mice and humans serves to highlight the importance of performing research using human patient material. Further strengthening the idea that UNC-93B and its associated innate response pathway are critical for HSE resistance is that humans with deficiencies in downstream factors, such as TLR3 and the signaling molecules TRAF3 and TBK, phenocopy almost exactly those with mutations in UNC-93B. While the penetrance of HSE in this population is incomplete, this previous work discovered a very strong association between the prevalence of HSE and defects in the TLR3 pathway. Some key issues, however, remained unresolved. For example, the idea that UNC-93B-deficient cells were highly susceptible to HSV infection arose from studies using SV40-transformed UNC-93B^−/− fibroblasts obtained from patients. While such data were consistent with clinical observations, it remained unknown whether the cells of the central nervous system (CNS) targeted by HSV during HSE showed a similar dependence on TLR3-mediated IFN-driven immunity for viral control.

Lafaille et al. (2012) in the Casanova, Studer, and Notarangelo laboratories therefore set about a collaborative tour de force study in which they obtained dermal fibroblast-derived pluripotent stem cells and induced their differentiation into neurons, oligodendrocytes, and astrocytes (Figure 1). These three cell types represent the primary nonhematopoetic cells that reside in the CNS. Extensive analysis of these cells through study of the cell-surface markers, gene-expression profiling, and electrophysiology was required to identify and characterize these newly derived cells. The authors then sought to examine TLR3 responses of these control and UNC-93B^−/− cells following stimulation with poly(I:C), a commonly used agonist of TLR3 frequently used for its mimicry of the effects of virus-derived dsRNA. As expected, the UNC-93C-deficient CNS cells behaved similarly to UNC-93C-deficient fibroblasts, with significantly reduced TLR3 pathway responses. Importantly, these responses were restored through provision of intact UNC-93B in trans.

Figure 1. The Derivation of CNS Cells from Human Pluripotent Stem Cells and Their Relative Susceptibilities to HSV-1 Infection.

Induced human pluripotent stem cells (black box, center) were generated from dermal fibroblasts from control, UNC-93B-deficient, and TLR3-deficient (not shown) subjects. These cells were then induced to differentiate to neural precursor cells, which in turn were differentiated with growth factors to produce neurons, oligodendrocyes, astrocytes, and neural stem cells. Infection of the differentiated CNS cells with a GFP-expressing HSV-1 strain showed a significant increase in fluorescence in the UNC93B^−/− neurons and oligodendrocytes (red boxes, below) relative to wild-type controls (above). Similar results were obtained with TLR3^−/− neurons and oligodendrocytes (not shown). Susceptibility of astrocytes and neural stem cells to HSV-1 infection remained high, unchanged by their UNC-93B genotype. These data suggest that the increased susceptibilities to HSV-1 encephalitis of humans with mutations in the TLR3 pathway result from intrinsic changes in the susceptibilities of their resident CNS cells to HSV-1 infection. Cell and virus illustrations created by Jonathan Leib.

The key experiments in this study, however, revealed that UNC-93C^−/− neurons and oligodendrocytes were significantly more susceptible to HSV-1 infection, as judged by increases in green fluorescence when infected with a HSV-1 recombinant strain with GFP fused to a structural protein (Figure 1). Interestingly, the UNC-93C^−/− neural stem cells (NSCs) and astrocytes did not show increased susceptibility over control cells, although these two cell types were generally more susceptible to HSV-1 infection than neurons and oligodendrocytes regardless of genotype. One caveat to this finding is that murine astrocytes require TLR3 for resistance to HSV-2 (Reinert et al., 2012), implying either that there is a difference between humans and mice in this regard or between astrocytes in vitro and in vivo. While viral titers and yields from these cells were not shown, the GFP fluorescence measurements strongly support the conclusion that neurons and oligodendrocytes require an intact TLR3 pathway for resistance to HSV. TLR3-driven innate resistance, intrinsic to the CNS, is critical for protecting the CNS, likely in addition to a requirement for such protective functions in the immune system in general. When combined with the clinical data, this pathway appears especially important for the protection from HSE during childhood. This study, and others currently emerging, are setting the stage for a veritable explosion of data pertaining to the study of host-pathogen interaction in humans, allowing confirmation or refutation of predictions based on small animal model systems.