Research Article: Type I Interferon Programs Innate Myeloid Dynamics and Gene Expression in the Virally Infected Nervous System

Date Published: May 30, 2013

Publisher: Public Library of Science

Author(s): Debasis Nayak, Kory R. Johnson, Sara Heydari, Theodore L. Roth, Bernd H. Zinselmeyer, Dorian B. McGavern, Glenn F. Rall.


Viral infections of central nervous system (CNS) often trigger inflammatory responses that give rise to a wide range of pathological outcomes. The CNS is equipped with an elaborate network of innate immune sentinels (e.g. microglia, macrophages, dendritic cells) that routinely serve as first responders to these infections. The mechanisms that underlie the dynamic programming of these cells following CNS viral infection remain undefined. To gain insights into this programming, we utilized a combination of genomic and two-photon imaging approaches to study a pure innate immune response to a noncytopathic virus (lymphocytic choriomeningitis virus) as it established persistence in the brain. This enabled us to evaluate how global gene expression patterns were translated into myeloid cell dynamics following infection. Two-photon imaging studies revealed that innate myeloid cells mounted a vigorous early response to viral infection characterized by enhanced vascular patrolling and a complete morphological transformation. Interestingly, innate immune activity subsided over time and returned to a quasi-normal state as the virus established widespread persistence in the brain. At the genomic level, early myeloid cell dynamics were associated with massive changes in CNS gene expression, most of which declined over time and were linked to type I interferon signaling (IFN-I). Surprisingly, in the absence of IFN-I signaling, almost no differential gene expression was observed in the nervous system despite increased viral loads. In addition, two-photon imaging studies revealed that IFN-I receptor deficient myeloid cells were unresponsive to viral infection and remained in a naïve state. These data demonstrate that IFN-I engages non-redundant programming responsible for nearly all innate immune activity in the brain following a noncytopathic viral infection. This Achilles’ heel could explain why so many neurotropic viruses have acquired strategies to suppress IFN-I.

Partial Text

The central nervous system (CNS) is an immunologically specialized compartment consisting of the brain and spinal cord [1]. These structures are lined by what is referred to as the meninges. Most blood vessels within the CNS are non-fenestrated, meaning the endothelial cells that comprise these vessels are connected by tight junctions which limit the influx of vascular materials into the CNS [2]. Tight junctions are a key feature of the blood brain and blood cerebral spinal fluid barriers that help protect the CNS from peripheral challenges. Despite this elaborate barrier structure, many infectious agents have evolved clever strategies to access the CNS [3]. This tissue must therefore be equipped to mount an immune response to preserve its cellular inhabitants, some of which are non-replicative (e.g. neurons). Because most immune responses begin with pattern recognition or the sensing of “danger” [4], [5], tissues often possess elaborate networks of innate immune sentinels that typically serve as the first responders to infectious agents. Despite its immunoprivileged status [6], the CNS is no different from the periphery in this regard. The most abundant innate immune sentinels in the CNS are referred to as microglia [7]. These cells are ramified and distributed evenly throughout the CNS parenchyma. In addition, recent intravital imaging studies have demonstrated that microglia processes are highly dynamic and continually scan the CNS [8], [9]. The meninges, choroid plexus, and perivascular spaces in the CNS are also inhabited by specialized macrophages as well as dendritic cells (DCs) [10], [11], [12], [13]. Unlike microglia [14], these cells are hematopoietically-derived and turnover at regular intervals [12]. In fact, a recent study demonstrated that DCs residing in the meninges and choroid plexus are Flt3-ligand responsive and have a 5–7 day half-life [12]. Thus, the innate immune composition of the CNS lining in some ways resembles that observed in peripheral tissues.

Innate protection of the CNS is mediated by an elaborate network of innate immune sentinels that consist of microglia, specialized macrophages, and DCs [13]. These often serve as the first responders against invading infectious agents and must hold these microbes in check prior to the arrival of adaptive immune cells. In this study, we unexpectedly uncovered that the brain has an Achilles’ heel in its defense against a non-cytopathic arenavirus. Specifically, all gene expression and innate myeloid cell dynamics were completely abrogated in the absence of IFN-I signaling. That LCMV induced an IFN-I signature of gene expression in the brain was not surprising given that the virus is detected by RIG-I/MDA5 and is known to trigger IFN-I production [22]. It was not predicted, however, that all differentially regulated genes (both up and down) would be linked exclusively to this pathway. During the early stage of infection, when LCMV localized primarily to the meninges and superficial parenchyma, a robust innate response was observed at the genomic and cellular levels. At this time point, 5 logs of infectious virus were detected in the brain, 585 genes were differentially regulated, microglia had transformed morphologically, and vascular patrolling by innate myeloid cells was markedly elevated. These were all indicators of a successful innate response. During the later stages of persistence, no consistent pattern of viral mutation was observed, nor was there any gain in the efficiency of infectious virion generation. However, both IFN-β synthesis and the network of innate immune gene expression were largely silenced as the virus moved into the brain parenchyma. This coincided with the restoration of myeloid cell dynamics to a quasi-normal state by 140 days post-infection, suggesting equilibration between LCMV and its murine host. Importantly, when IFNR−/− mice were infected with LCMV, no anti-viral response was observed at the genomic or dynamic levels despite increased viral loads and infection of microglia. These data indicate that the brain has only one way to respond innately to LCMV. This Achilles’ heel has in turn been exploited by LCMV and most other arenaviruses, which were demonstrated in vitro to suppress IFN-I synthesis and signaling [22], [39], [40], [41]. This could also explain why many other neurotropic viruses have acquired strategies to dampen the IFN-I pathway [20].




0 0 vote
Article Rating
Notify of
Inline Feedbacks
View all comments