Virus Infections in the Nervous System

View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector Cell Host & Microbe Review

Virus Infections in the Nervous System

Orkide O. Koyuncu,1 Ian B. Hogue,1 and Lynn W. Enquist1,* 1Department of Molecular Biology, Princeton University, Princeton, NJ 08544 USA *Correspondence: [email protected] http://dx.doi.org/10.1016/j.chom.2013.03.010

Virus infections usually begin in peripheral tissues and can invade the mammalian nervous system (NS), spreading into the peripheral (PNS) and more rarely the central (CNS) nervous systems. The CNS is protected from most virus infections by effective immune responses and multilayer barriers. However, some viruses enter the NS with high efficiency via the bloodstream or by directly infecting nerves that innervate peripheral tissues, resulting in debilitating direct and immune-mediated pathology. Most viruses in the NS are opportu- nistic or accidental pathogens, but a few, most notably the alpha herpesviruses and rabies virus, have evolved to enter the NS efficiently and exploit neuronal cell biology. Remarkably, the alpha herpesviruses can establish quiescent infections in the PNS, with rare but often fatal CNS pathology. Here we review how viruses gain access to and spread in the well-protected CNS, with particular emphasis on alpha herpes- viruses, which establish and maintain persistent NS infections.

Introduction tions caused by West Nile virus (WNV) and Nipah virus also Most acute and persistent viral infections begin in the periphery, represent zoonotic infections. When these zoonotic RNA viruses often at epithelial or endothelial cell surfaces. Infection of cells at infect fully differentiated neurons (in the brain and spinal cord), these sites usually induces a tissue-specific antiviral response virus clearance by the immune system is a major challenge. that includes both a cell-autonomous response (intrinsic immu- Because CNS neurons in particular are irreplaceable, canonical nity) and paracrine signaling from the infected cell to surrounding T-cell-mediated cytolysis of these infected cells is not a favor- uninfected cells by secreted cytokines (innate immunity). This able strategy. Thus, as understood by animal models, recovery local inflammatory response usually contains the infection. After from viral infection and the clearance of viral RNA require a several days, the adaptive immune response may be activated slower process involving the actions of virus-specific antibodies and the infection cleared by the action of infection-specific anti- and interferon gamma (IFN-g) secretion from T cells (Griffin and bodies and T cells (acquired immunity). Viral infections that Metcalf, 2011). escape local control at the site of primary infection can spread In contrast to the dangerous zoonotic CNS infections to other tissues, where they can cause more serious problems described above, some human-adapted viruses gain access to due to robust virus replication or overreacting innate immune the CNS as a result of diminished host defenses that fail to limit response. This latter reaction is sometimes called a ‘‘cytokine peripheral infections (e.g., Epstein-Barr virus, human cytomega- storm’’ because both proinflammatory and anti-inflammatory lovirus, JC virus). Others are the result of the iatrogenic introduc- cytokines are elevated in the serum leading to vigorous systemic tion of the infection directly into the CNS, bypassing peripheral immune activity (Figure 1). Such a response in the brain is usually barriers (e.g., use of adenovirus or lentivirus vectors, or prions). devastating and can lead to meningitis, encephalitis, meningo- Remarkably, many alpha herpesviruses efficiently enter the pe- encephalitis, or death. ripheral nervous system (PNS) of their natural hosts and establish The capacity to invade the host nervous system (NS) appears a quiescent infection with little or no CNS pathogenesis. Their to be a rather poor evolutionary path because of the potentially initial peripheral infection stimulates a well-controlled intrinsic damaging and lethal nature of these infections. In most cases, and innate immune response, as well as a long-lived adaptive acute infection of the central nervous system (CNS) has no immune response. The alpha herpesvirus genomes remain apparent selective advantage for the host or the pathogen. quiescent in PNS neurons for the life of their hosts, only occa- These infections, when they occur, often represent a zoonotic sionally reactivating to produce virions that can reinfect periph- infection: an incursion of a virus from its natural host into a eral tissues and spread to other hosts. Perhaps the prosurvival, ‘‘dead-end host,’’ which has little chance to further disseminate cytolysis-resistant nature of mature neurons facilitates establish- the infection. Zoonotic virus infections are often minimally path- ment of such persistent and reactivatable infections. ogenic in their natural hosts, but they can be highly virulent and In this review, after providing a general overview on how neuroinvasive in nonnatural hosts. The virulence and lethality of viruses gain access to the NS, we will focus on intra- and inter- zoonotic infections may result from the cytokine storm evoked neuronal virus trafficking and spread tactics and discuss how after the primary infection. The 1918 influenza pandemic and alpha herpesviruses establish persistent infections in the NS. more recent H5N1 (bird flu) outbreak in humans are good exam- We will finish by discussing our views on the control of virus ples of this phenomenon (Teijaro et al., 2011). Patients infected invasion in the NS. with H5N1 virus die, not because of robust virus replication, but because of acute respiratory distress syndrome (ARDS) trig- Entering the Nervous System gered by the cytokine storm. Fatal infection by rabies virus While the PNS is relatively more accessible to peripheral infec- (RABV), and the less known but nonetheless devastating infec- tions because nerves are in direct contact with tissues of all

Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. 379 Cell Host & Microbe Review

Figure 1. Immune Control of Viral Infections Cellular pattern recognition receptors (PRRs) recognize viral molecules after attachment and entry. This initial recognition starts a cell-autono- mous intrinsic defense involving increased syn- thesis of many antiviral proteins, and several cytokines, including type I interferons (IFN-a/b). If intrinsic defenses fail to stop virus replication, cytokines and infected cell death activate sentinel cells (e.g., dendritic cells and macrophages), which produce copious cytokines and present antigens to trigger T-cell-mediated immunity. The innate immune response, which is activated quickly and robustly, also involves the actions of the complement system, natural killer (NK) cells, neutrophils, and other granulocytes. Overreaction of this response causes immunopathology termed a ‘‘cytokine storm.’’ The adaptive, acquired immune response is slow, systemic, and pathogen specific and leads to the induction of immuno- logical memory. The cell-mediated ‘‘Th1’’ response involves the action of CD4+ T helper cells and CD8+ cytotoxic T cells (CTLs). CTLs are important to kill infected cells and to produce type II interferon (IFN-g) and tumor necrosis factor (TNF-a). The ‘‘Th2’’ humoral response involves CD4+ T helper cells and antibody-producing B cells. Most of the time, the clinical symptoms of virus infection (e.g., fever, pain, and tissue damage) are caused by the inflammatory action of the innate and adaptive immune systems. Due to their mostly irreplacable nature, nervous system tissues rely predominantly on the intrinsic and innate immune responses and avoid the extensive inflammation and cytotoxic effects of the adaptive immune response. types, the CNS proper has several layers of protection. Spread of zoster virus (VZV). Well-studied animal alpha herpesviruses infection from the blood to the cerebral spinal fluid and CNS cells include pseudorabies virus (PRV) and bovine herpes virus is limited by the blood brain barrier (BBB). The BBB is mainly (BHV). The cellular adhesion molecule Nectin-1 is a major composed of endothelial cells, pericytes, astrocytes, and the neuronal receptor for these viruses (except for VZV) (Smith, basement membrane (Figure 2D). Pericytes project finger-like 2012). Alpha herpesvirus particles enter sensory nerve endings processes to ensheath the capillary wall and coordinate the neu- by membrane fusion and engage dynein motors for retrograde rovascular functions of the BBB. The star shaped astrocytes (i.e., transport to the neuronal cell body (i.e., soma) (Smith, 2012). astroglia) are the major glial cell type in the CNS. The astrocyte The capsids then dock at the nuclear pore, and the viral DNA endfeet projections ensheath the capillary, regulating BBB is instilled in the nucleus, where an acute or a quiescent, latent homeostasis and blood flow. Brain microvascular endothelial infection is established. These viruses are unique in maintaining cells (BMVECs) that line the CNS vasculature are connected by quiescent infections in the PNS neurons of their hosts. This tight junctions not found in capillaries in other tissues. These quiescent state can be reversed by certain stress stimuli, result- junctions restrict the egress of bacteria, virus particles, and large ing in the production of large numbers of progeny in a relatively protein molecules from the lumen of the blood vessel, while short period of time (1–2 days) (Camarena et al., 2010). Despite allowing the transport of metabolites, small hydrophobic pro- the direct synaptic connection of PNS neurons to the CNS, teins, and dissolved gasses in and out of the CNS. The basement spread of alpha herpesvirus infection to the CNS is rare, but membrane, a thick extracellular matrix, also surrounds these debilitating, in natural hosts (Tirabassi et al., 1998). Alpha capillaries, further limiting movement of pathogens. Perivascular herpesvirus trafficking, latency, and reactivation is further macrophages (i.e., microglia) residing between the endothelial reviewed in the following sections. and glial cells further provide immune surveillance in the CNS Invasion of Motor Neurons at Neuromuscular Junctions. tissue. Virus infections that leave the periphery and find their Neuromuscular junctions (NMJs) are specialized synapses way into the PNS or CNS do so either by direct infection of nerve between muscles and motor neurons that facilitate and control endings in the tissues or by infecting cells of the circulatory muscle movement. NMJs can be gateways for many viruses to system that ultimately carry the infection through the BBB into spread into the CNS (Figure 2B). Most motor neurons have their the CNS (Table 1 and Figure 2). cell bodies in the spinal cord, which, in turn, are in synaptic con- Direct Infection of Nerve Endings tact with motor centers in the brain. Poliovirus and RABV infec- Invasion of Sensory Nerve Endings. Some viruses can enter the tions spread into the CNS through NMJs. While RABV particles PNS by binding to receptors on axon termini of sensory and enter the NMJ directly after a bite from an infected animal, polio- autonomic neurons, which respectively convey sensory and virus particles reach the NMJ by a more circuitous route (Raca- visceral information. Most alpha herpesviruses use this route to niello, 2006). Poliovirus infection begins with the ingestion of enter the PNS and establish a life-long persistent infection (Tira- virus particles followed by replication in the intestinal mucosa bassi et al., 1998)(Figure 2A). Human alpha herpesviruses and secretion of new virus particles in feces, typically with no include herpes simplex type-1 (HSV-1), HSV-2, and varicella CNS infection. However, about 1%–2% of poliovirus infections

380 Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. Cell Host & Microbe Review

Figure 2. Virus Entry Routes into the CNS (A) Alpha herpesviruses (e.g., HSV-1, VZV, and PRV) infect pseudounipolar sensory neurons of PNS ganglia. CNS spread is rare and requires anterograde axonal transport of progeny virions toward the spinal cord. (B) RABV and poliovirus spread via neuromuscular junctions (NMJs) from muscles into somatic motor neurons in the spinal cord. (C) Several viruses may infect receptor neurons in the nasal olfactory epithelium. Spread to the CNS requires anterograde axonal transport along the olfactory nerve into the brain. (D) Infiltration through the BBB. The BBB is composed of brain microvascular endothelium cells (BMVECs) with specialized tight junctions, surrounding basement membrane, pericytes, astrocytes, and neurons. Infected leukocytes can traverse this barrier carrying virus into the brain parenchyma. (E) Alternatively, virus particles in the bloodstream can infect BMVECs, compromising the BBB. result in the infection of motor neurons that leads to the well- spread of progeny virions to motor centers in the brain results known motor dysfunction poliomyelitis (Racaniello, 2006). The in muscle paralysis (Kuss et al., 2008; Ohka et al., 2012). receptor for poliovirus is an Ig superfamily member called RABV, a member of the Rhabdoviridae family, infects many CD155, which is found on axonal membranes (Ren and Raca- animals, including bats, skunks, foxes, and dogs, and rhabdovi- niello, 1992). Work in transgenic mice expressing human ruses can even infect insects and plants. RABV in animal saliva CD155 has led to a model in which the virus replicates in motor spreads between hosts via bites or scratches. Animals infected neurons within the spinal cord and subsequent retrograde with RABV can survive for months and even years secreting

Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. 381 Cell Host & Microbe Review

Table 1. Classification, Properties, and CNS Entry Routes of Viruses that Can Cause Nervous System Infections Genome Virus Family N/E Specifics Viruses CNS Entry dsDNA Adenoviridae N MAV-1 BBB and BMVECs Herpesviridae E alpha herpesvirus HSV-1, HSV-2, VZV, PRV, and BHV sensory nerve endings and ORN E beta herpesvirus HCMV BBB and BMVECs E gamma herpesvirus EBV BBB and BMVECs Polyomaviridae N JCV BBB and BMVECs dsRNA Reoviridae N mouse model T3a BBB; peripheral nerve? (+)ssRNA Coronaviridae E mouse model MHVb peripheral nerve and ORN Flaviviridae E hepacivirus HCVc BBB and BMVECs E zoonotic WNVc and JEVc peripheral nerve, BBB, ORN, and BMVECs Picornaviridae N enterovirus poliovirusc and EV71c NMJs and BBB N mouse model TMEVd BBB Togaviridae E zoonotic CHIKVc ORN? E mouse model SINVe ORN (–)ssRNA Arenaviridae E zoonotic LCMV BBB Bornaviridae E zoonotic BDV ORN Bunyaviridae E zoonotic LACV ORN and BBB? Orthomyxoviridae E zoonotic influenza Ac peripheral nerve and ORN Paramyxoviridae E morbilivirus MV BBB E rubulavirus MuV BBB E henipavirus (zoonotic) HeVc and Nipahc BBB and ORN? Rhabdoviridae E zoonotic RabV and VSV NMJs and ORN ssRNA-RT Retroviridae E lentivirus HIVc and HTLV BBB dsRNA, double-stranded RNA; ssRNA, single-stranded RNA; N, naked; E, enveloped; ORN, olfactory receptor neuron; RT, reverse transcriptase; MHV, mouse hepatitis virus. aRichardson-Burns et al. (2002). bBergmann et al. (2006). cEmerging virus. dTsunoda and Fujinami (2010). eMetcalf and Griffin (2011).

infectious particles in their saliva. In contrast, human infection ronment. The olfactory nerve, which belongs to the PNS, inner- results in fatal acute myeloencephalitis unless treated with rabies vates the olfactory epithelium and terminates in the olfactory antiserum. RABV particles enter the axons of motor neurons at bulb in the CNS proper (Figure 2C). The olfactory epithelium is the NMJ by binding to nicotinic acetylcholine receptors (nAchRs) well protected from most common infections by mucus and and neural cell adhesion molecules (NCAMs) (Ugolini, 2011). the presence of several pathogen recognition receptor systems Transneuronal spread occurs exclusively between synaptically (Kalinke et al., 2011). However, in laboratory models, the olfac- connected neurons, and the infection moves unidirectionally tory portal can be used by some viruses. There is evidence from postsynaptic to presynaptic neurons (retrograde spread). that HSV-1, vesicular stomatitis virus (VSV), Borna disease virus A unique feature of RABV infection of humans is the relatively (BDV), RABV, influenza A virus, parainfluenza viruses, and prions long asymptomatic incubation period after the initial infection, can enter CNS through olfactory route (Detje et al., 2009; Mori which can last from weeks to 1 year. This feature provides et al., 2005). valuable time for intervention with antiserum, which blocks Some emerging zoonotic infections may also spread to the CNS infection. Once rabies infection reaches the CNS, marked CNS by the olfactory pathway. For example, the paramyxovi- behavioral and neurological symptoms begin, and death almost ruses Hendra virus (HeV) and Nipah virus are natural pathogens always ensues. If the host survives long enough, virus particles of bats. HeV infection of humans causes lethal pneumonia but can spread from the CNS to peripheral tissues, particularly to also causes encephalitis by unknown mechanisms. In mice, the salivary glands. HeV is able to infect the CNS after intranasal infection via the Invasion via the Olfactory Epithelium and Olfactory Neurons. olfactory system, which may also be the infection route in The olfactory system provides a unique and directly accessible humans (Dups et al., 2012). An alphavirus, Chikungunyavirus portal of entry to the CNS from the periphery. The bipolar olfac- (CHIKV), from the Togaviridae family, is a mosquito-borne path- tory receptor neurons have dendrites in the olfactory epithelium ogen causing high fever, joint pain, and skin rash. Although at the roof of the nasal-pharyngeal cavity, where odors are CHIKV is not considered a neurotropic virus, CHIKV may spread detected. The CNS literally is one synapse away from the envi- into the CNS especially in young or elderly patients (Arpino et al.,

382 Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. Cell Host & Microbe Review

2009). In a mouse model, CHIKV could enter the CNS using the (HCMV) (Fish et al., 1998), and mouse adenovirus 1 (MAV-1) olfactory path (Powers and Logue, 2007). Similarly, La Crosse (Gralinski et al., 2009) all can infect BMVECs (Figure 2E). Infec- virus (LACV), a mosquito-borne bunyavirus, causes rare but tion of these cells often leads to the disruption of BBB integrity, severe meningoencephalitis, particularly in children. How LACV which is accompanied by the uncontrolled migration of immune spreads into the CNS is controversial, but it may occur via cells into the brain parenchyma. Inflammation in the brain tissue viremia after the initial replication in muscle cells (Johnson, induced by the activity of these cells is usually the cause of the 1983), or the virus may travel from the olfactory epithelium to observed neurological disorders. the CNS via olfactory neurons (Bennett et al., 2008). WNV, an arthropod-borne flavivirus, causes encephalitis and Invasion by Infected Circulating Leukocytes meningitis in a small percentage of infected humans after the Some viruses gain access to the NS without infecting neurons. initial replication in keratinocytes and Langerhans cells in the Instead, these viruses infect leukocytes, which circulate in the skin (Lim et al., 2011). CNS infections often are associated with blood, and may infiltrate in the brain parenchyma. This mecha- inefficient immune clearance of infection in the peripheral tissues nism is known as ‘‘Trojan horse’’ entry since the pathogens are and the resulting robust virus replication and subsequent viremia hidden within these immune defense cells, which are naturally (Cho and Diamond, 2012). WNV can access the CNS either by able to traverse the BBB (McGavern and Kang, 2011) infecting sensory nerve endings or olfactory neurons or through (Figure 2D). Lentiviruses, including simian immunodeficiency blood circulation, but not via NMJs (Lim et al., 2011). A hallmark virus (SIV) and human immunodeficiency virus (HIV), and another of WNV neuropathogenesis is the disruption of the BBB, result- retrovirus, human T cell leukemia virus (HTLV), may access the ing in uncontrolled entry of immune cells into the brain. The infec- CNS by the Trojan horse mechanism after infection of mono- tion not only stimulates the loss of tight junction proteins in both cytes or macrophages that infiltrate through the BBB (Kaul epithelial and endothelial cells (Xu et al., 2012), but also induces et al., 2001). HIV primarily targets CD4+ T lymphocytes, but it production of matrix metalloproteinases that degrade the base- also infects macrophages/microglia, which function as special- ment membrane. As a result, leukocytes traverse the capillaries ized phagocytic and antigen-presenting cells in the CNS, and into surrounding tissue and release cytokines upon recognition astrocytes, which express the cell-surface receptor CD4 and of WNV double-stranded RNA (dsRNA) via toll-like receptor 3 appropriate chemokine coreceptors (Martin et al., 1997; (TLR3) and tumor necrosis factor alpha (TNF-a) signaling (Verma Schweighardt and Atwood, 2001). Viral envelope glycoproteins et al., 2010; Wang et al., 2004). These events are unique to WNV gp120 and gp41 mediate virion attachment to these receptors. and are not observed with other hemorrhagic flaviviruses such as HIV replication is extensively influenced by the immune response Dengue virus (Xu et al., 2012). HIV, HTLV-1, and MAV-1 infec- induced by the infection. For instance, cytokines produced dur- tions also may disrupt the BBB by affecting tight junction pro- ing infection can either stimulate virus replication or suppress teins (Afonso et al., 2008; Gralinski et al., 2009). MAV-1 infects viral entry and replication, depending on the nature of the both BMVECs and macrophages, leading to acute encephalo- released cytokine and the state of the cell. CNS invasion occurs myelitis in susceptible mouse strains, which might explain how when infected memory T cells or infected monocyte/macro- human adenovirus infections in immunodeficient patients leads phage precursors traverse the BBB, which then differentiate to encephalitis with mortality rate approaching 60% (Ashley into perivascular microglia. Virus replication in microglia and et al., 2009; Gralinski et al., 2009). the subsequent induction of an inflammatory cytokine response Another flavivirus, HCV, which primarily infects hepatocytes, damages the CNS, causing depression and cognitive disorders. has also been associated with CNS abnormalities such as cogni- Notably, 10%–20% of infected patients exhibit HIV-associated tive dysfunction, fatigue, and depression (reviewed in Fletcher dementia (HAD) (McArthur et al., 1997). HIV infection also has and McKeating, 2012). BMVECs are the only cell type in the brain an effect on the PNS. Between 30%–60% of AIDS patients suffer that express all four receptors required for HCV entry (scavenger from HIV-associated sensory neuropathy characterized by the receptor BI, CD81, claudin-1, and occludin). They also support degeneration of myelinated and unmyelinated nerve fibers HCV infection in vitro, suggesting that they may be the CNS (Kamerman et al., 2012). reservoir of HCV (Fletcher et al., 2012). CNS infection by HCV The picornavirus enterovirus 71 (EV71) (Solomon et al., 2010) is also associated with increased expression of proinflammatory and a ubiquitous human polyomavirus, JC virus (JCV) (Boothpur cytokines (interleukin-1 [IL-1], TNF-a, IL-12, and IL18), choline, and Brennan, 2010), can also infiltrate into the CNS by the Trojan creatine, and inositol, all of which trigger microglia and astrocyte horse mode of entry. In the case of JCV infection, infiltration of activation, leading to pronounced neurological disorders infected B cells into the CNS in immune suppressed patients (Fletcher and McKeating, 2012). can result in the infection of oligodendrocytes (the myelin- Prevalent human DNA viruses that establish life-long, well- producing cells) and astrocytes, leading to a fatal inflammatory controlled persistent infections in other tissues can also damage disease in the brain called progressive multifocal leukoencephal- the CNS by infecting BMVECs under immune-suppressive con- opathy (PML) (Boothpur and Brennan, 2010). ditions. The beta herpesvirus HCMV (Fish et al., 1998), the Infection of Brain Microvascular Endothelium gamma herpesvirus EBV (Casiraghi et al., 2011), and the polyo- In some instances, virus particles in the circulatory system can mavirus JCV (Chapagain et al., 2007) are among such patho- reach and infect BMVECs, a major constituent of the BBB. gens. HCMV establishes life-long latency, predominantly in the RNA viruses such as WNV (Verma et al., 2010), Hepatitis C virus cells of the myeloid lineage (Goodrum et al., 2004). If the primary (HCV) (Fletcher et al., 2012), HTLV-1 (Afonso et al., 2008), and infection happens during pregnancy, transmission to the not- DNA viruses such as JCV (Chapagain et al., 2007), Epstein- yet-immune-competent fetus may result in a variety of fatal Barr virus (EBV) (Casiraghi et al., 2011), human cytomegalovirus CNS abnormalities, such as mental retardation and hearing

Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. 383 Cell Host & Microbe Review

loss. Besides BMVECs, astrocytes, pericytes, neurons, micro- with influenza infection have also been reported after nonneuro- glial cells, and, importantly, neural stem cells all can be infected tropic H1N1 pandemics. Recent work shows that increased with HCMV (Cheeran et al., 2009; Fish et al., 1998). Astrocytes production of proinflammatory cytokines and activation of and microglial cells naturally produce copious quantities of in- microglia upon peripheral, nonneurotropic influenza (A/PR/8/ flammatory cytokines in response to HCMV infection, and this 34) infection contributes to altered hippocampal structures and response can promote fetal neurodevelopmental diseases. cognitive dysfunction in mice in the absence of neuroinvasion EBV establishes life-long latency in memory B cells (Speck (Jurgens et al., 2012). and Ganem, 2010). EBV infection in young adults causes infec- Paramyxoviruses, measles virus (MV), and mumps virus tious mononucleosis. In rare cases, EBV infection leads to B (MuV), which are the causative agents of highly communicable cell lymphomas (e.g., Burkitt’s lymphoma) and epithelial cell diseases, can also lead to serious CNS infections. Primary MV carcinomas (e.g., nasopharyngeal carcinoma). Horwitz and col- and MuV infections begin in the upper respiratory tract, and leagues showed that EBV can infect human BMVECs, where it infection of lymphoid tissue causes viremia and spread to other becomes latent (Casiraghi et al., 2011). Virus reactivation in tissues. MuV is highly neurotropic and can cause acute enceph- these cells increases inflammatory cytokine and chemokine alopathy in children with high incidence. Elevated levels of expression, affecting the integrity of the local BBB, which might multiple cytokines were detected in cerebrospinal fluids of chil- lead to progression of the inflammatory neurological disease dren diagnosed with MuV-associated acute encephalopathy multiple sclerosis (MS) (Casiraghi et al., 2011). (Watanabe et al., 2013). Unlike MuV, MV infection spreads to Virus Induced Immune-Mediated CNS Pathogenesis the CNS in approximately 0.1% of the cases, causing several The BBB may be disrupted by mechanisms other than endothe- types of devastating neurological diseases, such as fatal sub- lial cell infection. Local production of cytokines in the CNS due to acute sclerosing panencephalitis (SSPE), which manifests viral infections affects tight junctions of the BBB as well. For weeks to years after infection (reviewed in Buchanan and example, expression of the chemokine monocyte chemoattrac- Bonthius, 2012). Although the molecular mechanisms of CNS tant protein-1 (MCP-1, CCL2), which is induced by viral infec- infection are not well understood, transgenic rodent models tions, is sufficient to disrupt tight junction proteins in the BBB reveal that a complex interplay of innate and adaptive immune (Stamatovic et al., 2005) and predispose the brain tissue for system responses are involved including disrupted Th1/Th2 immune-mediated neuropathology. An interesting example of balance and defective T cell proliferation (Schneider-Schaulies virus-induced immune-mediated CNS pathogenesis is the rare, et al., 2003). but fatal, Borna disease virus (BDV) infection. BDV, a zoonotic These selected examples of neuropathogenesis caused by RNA virus infecting warm-blooded animals, establishes persis- diverse virus families demonstrate that, while virus replication tent, noncytolytic infection in the CNS, particularly in neurons itself may cause neuropathologies, the activated immune of the limbic system (cortex and hippocampus) (Planz et al., system also contributes to the neuronal damage in an effort to 2009). The transmission route into the CNS is probably via the eradicate the infection. Moreover, immune-mediated changes olfactory system, followed by axonal transport and transsynaptic in the CNS induced by one virus infection might stimulate the spread to other parts of the brain (Lipkin et al., 2011). Depending neuropathogenicity of other viruses, as is seen in coinfections on the immune state of the host, BDV replication interferes with a with HIV and HCV, JCV, or HCMV. To better comprehend multitude of signaling cascades, deranging the expression of the effects of viral invasion on the NS, we must understand cytokines, neurotrophins that control neuron survival and func- how viruses interact with, replicate in, and spread in between tion, and apoptosis-related genes, leading to neuropathogene- neurons. sis. Although controversial, human BDV infections might be associated with a variety of neurological disorders, including Cell Biology of Neuronal Infection and Spread unipolar depression, bipolar disorder, and schizophrenia (Lipkin Neurons are highly polarized, with distinct dendrite and axonal et al., 2011). processes. Single axons can extend meters, many thousands Lymphocytic choriomeningitis virus (LCMV) is a rodent-borne of times the diameter of a typical cell. The maintenance and human pathogen that causes CNS disease. Pathogenesis the communication of distal axons with their cell bodies requires occurs through the combinatorial effects of virus replication highly specialized signal transduction, intracellular sorting, traf- and the subsequent host immune response. Humans acquire ficking, and membrane systems. As obligate intracellular para- this infection by inhalation of aerosols or dust suspensions. After sites, viruses are dependent on these cellular functions that are the initial replication in lung cells, LCMV virions move into the often cell-type specific. Accordingly, neurotropic viruses can circulation and can enter the CNS. In animal models, LCMV be divided into two categories: those that accidentally or oppor- preferentially infects nonneuronal cells in the brain (i.e., glial tunistically spread to the specialized NS (e.g., WNV, HCV, and cells), which triggers a robust inflammatory response and inter- EBV) and those that have adapted to the NS (e.g., alpha herpes- feres with neuronal migration (Bonthius, 2012). viruses and RABV). From a human health perspective, these Influenza A (i.e., H5N1, H1N1), an orthomyxovirus that is both fundamental differences in virus-host interaction result in the zoonotic and endemic in human population, primarily replicates broad range of viral pathogenesis. In this section, we will sum- in the lungs after transmission through the respiratory tract. marize how viral infections engage neuronal cell biology to enter, Some of the Influenza A isolates (i.e., HK483) are neurovirulent traffic, and spread between neurons. and can spread to the CNS through olfactory, vagus, trigeminal, Entry and sympathetic nerves in mouse models (Park et al., 2002). In general, virions may enter cells by direct fusion with or pene- Interestingly, neurological and cognitive effects associated tration of the plasma membrane or by endocytosis. Alpha

384 Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. Cell Host & Microbe Review

Figure 3. Cell Biology of Virus Entry, Transport, and Spread in the NS (A) Virons enter axons either via direct fusion with the axonal membrane after receptor attach- ment (e.g., alpha herpesviruses) or by endo- cytosis (e.g., RABV; most neurotropic viruses). All particles entering the axon cytoplasm require minus-end-directed dynein motors for long- distance retrograde transport on microtubules toward the cell body. Microtubules in axons are uniformly oriented with plus ends facing the axon terminus and the minus ends in the cell body. In the cell body and dendrites, microtubules have mixed polarity. (B) Postreplication trafficking and spread of progeny virions. Progeny alpha herpesvirus virions in transport vesicles can be transported antero- grade in axons, dependent on the viral protein Us9 associating with the microtubule motor kinesin-3/ KIF1A. Egress from axons allows anterograde virus spread from presynaptic to postsynaptic neurons. Virions in transport vesicles can also traffic to and egress from the somatodendritic compartment, allowing retrograde spread from postsynaptic to presynaptic neurons. Accordingly, herpesviruses can spread bidirectionally in neu- ronal circuits. In contrast, enveloped RNA virions (e.g., RABV and MV) acquire their membranes by budding though the plasma membrane during egress. For these viruses, the location of envel- opment/egress and the directionality of spread may depend on the transport and intracellular localization of viral glycoproteins. These viruses tend to spread unidirectionally, only from post- synaptic to presynaptic neurons.

and it is mediated by many neuron- specific isoforms of the conserved endocytosis machinery. At nerve termi- nals, in particular, the most prominent and best studied endocytic activity is a herpesviruses enter neurons by direct fusion of their virion specialized mode of clathrin-mediated endocytosis for neuro- envelopes with the plasma membrane (Figure 3A) (Lycke transmitter reuptake and synaptic vesicle recycling (Saheki et al., 1988). Nectin-1, a member of the immunoglobulin (Ig) and De Camilli, 2012). These endocytosis-derived recycled superfamily, is a receptor for HSV and PRV virions and engages synaptic vesicles typically remain near the presynaptic active the conserved viral glycoprotein gD on the virion surface to zone for ongoing rounds of endo- and exocytosis, and initiate the entry process. After membrane fusion, the capsid therefore their role in mediating viral entry is unclear. Other with inner tegument proteins is deposited into the axonal vesicle populations, including those with canonical early and cytoplasm, ready for subsequent intracellular transport steps late endosomal markers like Rab5 and Rab7 GTPases, are (Smith, 2012). present in axons, but their genesis, fate, and role in virus entry Most other neurotropic viruses enter neurons by endocytosis, is also uncertain. using cellular receptors that are concentrated at nerve terminals Upon endocytosis, the lumen of endocytic vesicles is rapidly (Figure 3A). The following neuromuscular junction proteins acidified by vacuolar ATPases. In nonneuronal cell types, this serve as viral receptors: NCAMs and nAChRs for RABV (Ugolini, pH change is often a trigger for virions to fuse with or otherwise 2011), CD155 for poliovirus (Ohka et al., 2004), and the coxsack- disrupt the vesicle membrane to reach the cytoplasm (Mercer ievirus and adenovirus receptor (CAR) (Salinas et al., 2009). In et al., 2010). Interestingly, in neurons, most virions remain inside addition, many viruses that do not infect neurons during natural the endocytic vesicle for subsequent long-distance retrograde infections can enter nerve terminals by endocytosis when trafficking to the neuronal cell body (e.g., rabies virus [Klingen artificially introduced, for example, during experimental or iatro- et al., 2008], polio [Ohka et al., 2009], and adenoviruses genic infections with adenovirus (Salinas et al., 2009), adeno- [Salinas et al., 2009]). Studies of the endocytosis and trafficking associated virus (AAV) (Cearley and Wolfe, 2007), and lentivirus of canine adenovirus 2 (CAV-2) and tetanus neurotoxin (TeNT) vectors (Mazarakis et al., 2001). revealed a population of endocytic vesicles with a neutral The molecular mechanisms of endocytosis are well studied pH, suggesting that such nonacidified vesicles may generally in a variety of cell types (Mercer et al., 2010). Compared to promote the retention and subsequent trafficking of neurotropic nonneuronal cells, neuronal endocytosis is highly regulated, viruses (Salinas et al., 2010).

Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. 385 Cell Host & Microbe Review

Postentry Retrograde Transport virion-containing endocytic vesicle. For example, the poliovirus After axonal entry, virus particles must undergo retrograde trans- cellular receptor CD155, which is endocytosed along with the port to the cell body using motor proteins that move along micro- virion, binds the dynein light chain Tctex-1 to engage the retro- tubules (Figure 3A). Generally, in polarized cells and particularly grade transport machinery (Ohka et al., 2004). Similarly, the in axons, microtubules are oriented with the plus end at the cell envelope glycoprotein (G protein) of RABV and related rhabdovi- periphery or axon terminal and the minus end toward the cell ruses can confer endocytosis and efficient retrograde transport body. The kinesin family of motor proteins generally mediates functions to heterologous lentivirus vector particles (Desmaris plus-end-directed anterograde transport (from the soma to the et al., 2001; Mazarakis et al., 2001), suggesting that interaction axon terminus), and dynein motor complexes generally mediate between the G protein and a cellular receptor (possibly minus-end-directed retrograde transport (from axon terminus to p75NTR [Langevin et al., 2002]; see above) mediates retrograde soma) (Kapitein and Hoogenraad, 2011). transport. After membrane fusion and entry into axons, alpha herpes- While these examples provide some insight into the molecular virus capsid-inner tegument complexes undergo very efficient mechanisms of postentry retrograde transport, it is not clear retrograde transport to the cell body (Figure 3A). The viral pro- whether virions endocytosed at nerve terminals engage a com- tein VP1/2 (UL36), a major component of the inner tegument, mon or default retrograde transport pathway and to what extent recruits minus-end-directed dynein motors for this retrograde they alter or modulate these intracellular trafficking mechanisms. transport (Zaichick et al., 2013). However, retrograde axonal After entry and retrograde transport to the cell body, viral nu- transport is punctuated by brief anterograde movements, puri- cleic acid genomes replicate and express messenger RNA fied capsid/inner tegument complexes also bind kinesin motors, (mRNA) and proteins. Remarkably, the biosynthetic machinery and VP1/2 is a multifunctional protein involved in multiple intra- (i.e., ribosomes and secretory pathway organelles) of neurons cellular transport steps during virus replication (Smith, 2012). is distributed in axons and dendrites, as well as in the cell bodies. Thus, motor recruitment by VP1/2, and other viral and cellular It is not clear whether this distribution of cellular machinery factors, is likely a complex and regulated process requiring facilitates compartment-specific viral replication and gene further study. expression. Like alpha herpesvirus particles, RABV virions were thought to Postreplication Transport, Egress, and Spread immediately undergo membrane fusion, releasing the viral nucle- Many neurotropic virus infections (e.g., RABV, alpha herpesvi- ocapsid into the axon cytoplasm, as typically occurs in non- ruses, MV, and WNV) spread between chains of connected neu- neuronal cell types. The nucleocapsid-associated P protein rons. Virions do not typically spread by free diffusion, but rather was proposed to mediate retrograde axonal transport by binding between tightly connected neurons via neurochemical synapses dynein light chain LC8. However, other studies demonstrated or other close cell-cell contacts. Infected PNS neurons often are that P protein binding to LC8 is not necessary for RABV neuropa- in direct synaptic contact with CNS neurons, providing a direct thogenesis (Mebatsion, 2001), and RABV virions instead remain route of spread from the periphery. Consequently, CNS infection within endocytic vesicles during retrograde transport (Klingen is determined in part by the directionality of virus spread along et al., 2008)(Figure 3A). neuronal circuits, which is influenced by directional intracellular The fact that most neurotropic viruses, particularly those not virus transport and egress. Motor neurons receive input from adapted to the nervous system, are transported within endocytic CNS neurons via synapses on the somatodendritic plasma vesicles presents an inherent topological problem. Unlike alpha membrane (i.e., postsynaptic contacts) and relay these signals herpesvirus particles, which are exposed to the axon cytoplasm to neuromuscular junctions at axon termini. RABV, which immediately after entry, these endocytosed virions are trapped typically infects motor neurons, spreads exclusively in the retro- in the lumen of an endosome and therefore can neither recruit grade direction along neuronal circuits, exiting from the somato- motor complexes nor directly modulate the transport machinery, dendritic plasma membrane to spread toward the CNS but may remain passive cargo of the pre-existing axonal (Figure 2B). Most neurotropic viruses are capable of spreading transport pathways. Several viruses appear to use a common retrograde in neural circuits, suggesting that there may be com- retrograde transport pathway involved in neurotrophin signal mon or default somatodendritic transport pathways. In contrast transduction. In this signaling pathway, neurotrophins (e.g., to motor neurons, sensory neurons are typically pseudounipolar, nerve growth factor [NGF] and brain-derived neurotrophic factor) with two axon-like projections: one innervating peripheral and their receptors (e.g., p75NTR and TrkB) are incorporated tissues, and the other making presynaptic contact with CNS into Rab5-positive early-endosome-like vesicles, which mature neurons (Figure 2A). To be able to spread within this pseudouni- into Rab7-postive late-endosome-like vesicles. Rab7 mediates polar architecture, the alpha herpesviruses have evolved the recruitment of dynactin (p50/glued), part of the dynein complex, capacity to spread in the anterograde direction, exiting from and these vesicles undergo retrograde transport to the neuronal axon termini. The alpha herpesviruses are among the very few cell body (Deinhardt et al., 2006). This pathway is implicated in viruses that are capable of bidirectional spread (both antero- the retrograde transport of poliovirus (Ohka et al., 2009), rabies grade and retrograde direction) (Figure 3B). virus (Langevin et al., 2002), CAV-2, and TeNT (Salinas et al., Like postentry transport, postreplication transport of viral 2009, 2010). These endosomal transport vesicles appear to components also requires microtubule motors. The cellular remain at a neutral pH, which may stabilize pH-dependent mechanisms that determine whether cargo is transported into viruses within the endosome (Salinas et al., 2009, 2010). the axon or into dendrites are not entirely clear; however, several In some cases, the interaction between virions and their mechanisms are apparent (Kapitein and Hoogenraad, 2011). cellular receptors may direct the retrograde transport of the First, the axon initial segment contains a mesh of actin filaments

386 Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. Cell Host & Microbe Review

proposed to function as a molecular sieve, preventing macromo- native VSV G proteins conferred anterograde (presynaptic to lecular complexes like transport vesicles or virions from entering postsynaptic) virus spread (Beier et al., 2011). While it is difficult the axon nonspecifically. Some motors, for example kinesin-1/ to draw conclusions about the functions of viral proteins outside KIF5, efficiently penetrate this barrier, providing a level of spec- of their natural context, these studies contribute toward a mech- ificity so that only particular cargo/adaptor/motor combinations anistic understanding of viral spread in the NS. But important are transported into the axon. Second, stable microtubules questions remain: What are the molecular determinants of viral acquire posttranslational modifications, most notably C-terminal glycoprotein sorting in neurons? How is the transport of nucleo- detyrosination and K40 acetylation of alpha-tubulin. Both axons capsids and other viral factors coordinated to ensure efficient and dendrites contain acetylated microtubules, and axons are assembly and egress? And, does this glycoprotein-centric enriched in detyrosinated microtubules. Kinesin-1/KIF5 prefer- model hold true for other enveloped viruses? entially binds these modified microtubules, suggesting that post- The transport of nascent alpha herpesvirus particles is now translational modifications may also regulate cargo transport understood in more molecular detail. These viruses replicate into axons. Finally, microtubule polarity differs between axons their genome and assemble capsids in the nucleus, traverse and dendrites. While microtubules in axons have a uniform, par- the nuclear envelope, acquire tegument proteins, and obtain allel orientation, with plus ends at the axon terminus and minus their envelope by budding into trans-Golgi-like membranes ends in the soma, microtubules in dendrites have mixed, antipar- (Smith, 2012). There has been some debate as to whether alpha allel orientations. This difference in microtubule polarity would herpesviruses fully assemble in the neuronal cell body prior to allow only plus-end-directed kinesin motors to enter axons, axonal transport or whether subassemblies are transported to whereas the minus-end-directed dynein complex could enter distant sites of final assembly. The details of this debate are re- dendrites (Kapitein and Hoogenraad, 2011). Indeed, one study viewed elsewhere (Diefenbach et al., 2008; Smith, 2012), but it is using a drug-inducible protein dimerization system showed clear that alpha herpesviruses have evolved neuron-specific viral that experimental recruitment of kinesin-1/KIF5 or kinesin-2/ mechanisms to direct anterograde transport and spread along KIF17 induces cargo transport into axons, whereas recruitment circuits of synaptically connected neurons. For both HSV of dynein induces transport into dendrites (Kapitein et al., 2010). and PRV, three conserved viral membrane proteins, gE, gI, and While enveloped virions may exit cells by budding though US9, are necessary for efficient anterograde transport and cellular membranes, which preserves membrane topology, non- spread (Smith, 2012). While the role of US9 in HSV is debated, enveloped viruses (e.g., poliovirus and Ad) must breach mem- US9-null PRV mutants are completely defective in anterograde branes to enter or exit a cell. How nonenveloped viruses cross transport and spread in neurons, and these phenotypes appear cellular membranes, particularly neuronal membranes, without to be neuron specific (Lyman et al., 2007). US9 is dispensable in causing extensive cytopathology is not well understood (re- nonneuronal cells and for retrograde somatodendritic egress in viewed in Kirkegaard, 2009). However, a study on Theiler’s neurons. A recent study demonstrated that PRV US9 associates murine encephalomyelitis virus (TMEV), a nonenveloped picor- with the neuron-specific plus-end-directed microtubule motor navirus like poliovirus, revealed that TMEV can exit from neurons kinesin-3/KIF1A, providing a molecular basis for this neuron- into oligodendrocyte myelin sheaths without axon lysis or degen- specific phenotype (Kramer et al., 2012). According to this eration (Roussarie et al., 2007). It is not clear whether other model, PRV US9 mediates the recruitment of KIF1A, which me- nonenveloped viruses can spread between neurons without diates both sorting into axons and anterograde transport leading cell lysis or axon degeneration. to spread of infection from presynaptic to postsynaptic cells in a In the case of many enveloped viruses (e.g., RABV and MV), neuronal circuit (Figure 3B). Thus, in bipolar neuronal circuitry, final assembly and egress are coordinated processes (Fig- PRV serves as an illustrative example of how directional spread ure 3B). Viral components and subassemblies, such as nucleo- along neuronal circuits depends on the directionality of intracel- capsids and membrane proteins, are transported separately lular transport and egress. Ongoing research to identify the and meet at the plasma membrane. Virions then acquire their cellular and molecular mechanisms of viral transport and egress envelope by budding through the plasma membrane, simulta- in neurons will be necessary to fully understand how other neuro- neously completing assembly and egress from the cell. For tropic viruses spread through the NS. decades, researchers have recognized that viral membrane pro- However, directional transport in neurons is not the sole deter- teins (i.e., glycoproteins and matrix proteins) are sorted either to minant of spread in neuronal circuits. For example, PRV and other the apical or basolateral side of polarized epithelial cells and that alpha herpesviruses naturally infect pseudounipolar sensory release of enveloped viruses is often polarized accordingly (see neurons that have two axon-like projections. Thus, anterograde Jayakar et al., 2004, for relevant discussion). Analogously, if viral axonal transport could direct progeny virions down either branch, envelope proteins are preferentially sorted into the axons or den- into peripheral tissues or into the CNS (Figure 4). Yet, spread to drites, virion assembly, egress, and transsynaptic spread may the CNS is typically rare or inapparent. In this case, the direction- also be polarized. In support of this model, a recent study ality of spread in the nervous system may depend on factors other showed that the directionality of viral spread along synaptic con- than intracellular transport, such as intrinsic immune responses nections depends on the viral glycoprotein G (Beier et al., 2011). and latency, as described in the following section. In this study, the authors produced recombinant VSV express- ing G proteins from RABV (a related rhabdovirus), LCMV (an Alpha Herpesvirus Invasion of the Nervous System: unrelated arenavirus), or its native VSV G protein. The RABV G The Potential for a Lifetime Association protein conferred retrograde (postsynaptic to pre-synaptic) The life-long persistence of herpesvirus genomes in their natural spread across synaptic connections, whereas the LCMV or hosts coupled with the capacity to reactivate the quiescent

Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. 387 Cell Host & Microbe Review

Figure 4. Establishment and Control of Alpha Herpesvirus Latency in PNS Neurons HSV-1 enters the host organism through mucosal/epithelial surfaces. Initial replication in epithelial cells results both in cytokine secretion and viral spread. Progeny virions enter sensory nerve terminals and undergo long-distance retrograde transport to the neuronal cell body. Localized intracellular signaling and protein translation in distal axons induced by cytokines and viral proteins during entry supports retrograde particle transport and also may influence the establishment of latency in neuronal cell bodies. Once in the nucleus, viral latency genes are activated (e.g., LAT and miRs), and the HSV-1 genome is repressed. CD8+ CTLs secreting IFN-g and granzyme-B and continuous retrograde NGF-PI3K-mTOR signaling from distal axons contribute to maintaining this latent state of infection. If this homeostasis is disrupted by environmental or nutritional stress, the HSV-1 genome reactivates lytic gene expression, leading to virus replication. Newly replicated virions traffic back by anterograde axonal transport mechanisms, to reestablish infection at epithelial tissues. It is still not understood why CNS invasion is rare after reactivation from latency, even though both axonal projections of the pseudounipolar sensory neuron are accessible to HSV-1 virion transport. If there is no difference in transport, it may be that HSV-1 does invade the CNS but is strongly suppressed or silenced by an effective intrinsic immune response in CNS neurons of healthy individuals. infection to produce infectious particles that transmit to other neurons is a multistep process and depends on the efficiency of hosts is remarkable in many ways (reviewed in Wagner and several parameters: (1) virus replication in nonneuronal cells un- Bloom, 1997). Three subfamilies of the herpesviridae establish der the pressure of local intrinsic, innate, and humoral immune a latent persistent infection in different cell types: alphaherpes- responses, (2) virion entry into axon terminals and retrograde virinae (PNS neurons), betaherpesvirinae (monocytes and mac- transport of nucleocapsids to the neuronal cell body, and (3) rophage precursors), and gammaherpesvirinae (B and T cells) intrinsic and innate immune responses in the infected neuron (Speck and Ganem, 2010). In most cases, the quiescent herpes- and ganglia. After the latent infection is established, mainte- virus genome is not integrated into the host genome and either nance of this silent state requires continuous action of surround- exists as a nonreplicating, quiescent histone-covered molecule, ing glial and virus-specific CD8+ T cells that suppress lytic gene or as a replicating extra ‘‘chromosome’’ (reviewed in Speck and expression through secretion of IFN-g and granzyme-B (St Leger Ganem, 2010). Recently, this paradigm was broken by human and Hendricks, 2011). herpesvirus 6, a beta herpesvirus that can integrate into the Recent studies with HSV-1 suggest that infection of neurons host genome (Arbuckle et al., 2010). after axonal entry is inherently biased toward a latent infection Alpha herpesvirus genomes can remain in a latent, but reacti- because when the HSV-1 virions enter axons, some critical tegu- vatable, state in neurons of essentially all PNS ganglia. HSV-1 ment proteins are transported toward the cell body separately and HSV-2 are two human alpha herpesviruses causing cold from the capsid (Roizman and Sears, 1987). When neuronal sores and genital herpes. A third human alpha herpesvirus, cell bodies are exposed directly to purified virions, the infection VZV, causes chickenpox during primary infection of children is instead productive. This bias toward latency presumably and zoster (shingles) when the virus reactivates in the ganglia occurs because productive infection may require that both tegu- that transmit sensory signals toward the brain (e.g., dorsal root ment proteins and capsids containing the viral genome arrive at ganglion and trigeminal ganglion). Infection initiates in epithelial the nucleus at the same time (Hafezi et al., 2012; Kristie and Roiz- cells at a mucosal surface. The primary infection is usually well man, 1988). Besides this long-distance trafficking, other factors contained by host defenses, and overt pathogenic effects are such as cytokine and IFN production also influence the outcome. rare. Some viral progeny invariably enter axon terminals that The cytokines produced after primary infection of epithelial cells innervate the infected tissue and move on microtubules to PNS or the virion attachment to axonal receptors might induce an cell bodies, where a productive or latent infection may be estab- antiviral state in neuronal soma before the genome reaches the lished (Figure 4). Establishment of the quiescent infection in PNS nucleus. Indeed, De Regge et al., using an in vitro two-chamber

388 Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. Cell Host & Microbe Review

neuron culture system, showed that treatment of trigeminal (Orvedahl et al., 2007). This interaction has little effect or advan- ganglion neurons with IFN-a before HSV-1 or PRV infection is tage in nonneuronal cells, but it promotes HSV-1 persistence in sufficient to suppress lytic gene expression and induce a quies- neurons and also pathogenesis if the infection spreads to the cent infection (De Regge et al., 2010). CNS (Yordy et al., 2012). A hallmark of alpha herpesvirus latency is the expression of a After reactivation, it is not clear whether progeny herpesvirus latency-associated transcript (LAT) and several microRNAs virions are transported toward the CNS and, if so, what prevents (miRNAs) (Farrell et al., 1991; Umbach et al., 2008). Although devastating CNS infection and pathogenesis in most immune- the exact role of the LAT transcripts is still not well understood, competent individuals. A recent elegant study revealed the there is evidence that these noncoding RNAs are involved in importance of neuronal intrinsic immunity in the control of blocking apoptosis of the infected neuron and in the control of HSV-1 spread in the NS. Children with genetic defects in the reactivation from latency (reviewed in Proenc¸a et al., 2011). antiviral TLR3-interferon pathway have a higher risk of herpes Besides LAT, HSV-1 miRNAs target the key regulators of the viral simplex encephalitis (HSE) during primary infection with HSV- lytic cycle (i.e., ICP0) to control and stabilize latency. Remark- 1. Using pluripotent stem cells from susceptible children, the ably, for VZV, six different transcripts and proteins are produced authors showed that a TLR3-dependent intrinsic immune during latency. Interestingly, one of the most abundant proteins, response in neurons is crucial to restrict HSV-1 spread in the ORF63, is predominantly cytoplasmic during latency but local- NS (Lafaille et al., 2012). It is likely that the lack of this defense izes mostly in the nucleus during lytic infection (Cohen, 2007). system in neurons prevents the establishment of latency in the Nevertheless, almost 90% of the viral genome remains transcrip- peripheral ganglia, promoting further viral spread into the brain. tionally silent during latent HSV-1, HSV-2, or PRV infection, and, Once in the brain, where HSV-1 replication is normally tightly importantly, the infected neuron is not recognized or eliminated restricted, TLR-3 deficient neurons may continue to support by the immune system. virus replication leading to HSE. Such long-term maintenance of a silenced viral genome in a The herpesvirus latency-reactivation cycle is understood in state that can be reactivated requires precise control and a principle, but important details are lacking. Small-animal models high threshold for reactivation. Stress responses, including nutri- (i.e., mouse peripheral infection and rabbit eye models to study tional, hormonal, and environmental, can reactivate a latent HSV-1 and guinea pig model for HSV-2 latency) have been quite genome. Strikingly, although reactivation can recur throughout useful, but they do not recapitulate all the nuances of latency in the life of the host, it seems to have little adverse effect on humans (Wagner and Bloom, 1997; Wilson and Mohr, 2012). PNS function. This is, in part, because few latently infected neu- However, latency and reactivation have been addressed with rons undergo reactivation: in vivo animal experiments show that PNS neurons cultured in vitro, and these studies have provided only one to three neurons, among thousands of latently infected some mechanistic understanding of these phenomena (Camar- cells in the trigeminal ganglion, display signs of reactivation in a ena et al., 2010; Wilcox et al., 1990). HSV-1 latency is proposed single recurrence (Sawtell and Thompson, 1992). to be maintained by the continuous binding of NGF to the TrKA Furthermore, neurons have evolved to inherently avoid cell receptor tyrosine kinase (RTK) and activation of phosphatidylino- death, and viruses that persist in neurons have also adapted to sitol 3-kinase (PI3K)-Akt signaling pathway (Camarena et al., prevent the death of their host cells. These neuronal and viral 2010; Wilcox et al., 1990). Recently, Kobayashi et al. further prosurvival strategies clearly allow neurons to survive during a showed that the key target of PI3K-Akt signaling is mTORC1 in latent herpesvirus infection, but it may be that these same stra- axons (Kobayashi et al., 2012). The mTOR system regulates tegies promote neuronal survival during reactivation as well. It the translational repressor eIF4E-binding protein (4E-BP), which appears that a fine balance exists between the immune system influences the state of latent viral genomes in the neuronal of a naturally immunized host and the virus replication machin- nucleus. Although the cellular effectors of this pathway that ery, limiting virus replication and subsequent axon-to-cell spread suppress the viral lytic gene expression are still unknown, any of the progeny even during reactivation (Kobiler et al., 2010; Tay- interference with this signaling cascade (e.g., NGF depletion, in- lor et al., 2012). Indeed, it appears that neurons can tolerate not hibition of mTOR activity, or hypoxia) breaks latency control and only high viral genome inputs but also abundant viral gene promotes reactivation. It appears that a variety of physiological expression, which possibly is controlled by the suppressing ac- stimuli originating in the periphery are integrated in the local tion of nearby CD8+ T cells (St Leger and Hendricks, 2011). mTOR system in axons and relayed to the cell body. This local Furthermore, autophagy is a neuronal prosurvival strategy to monitoring in axons and long distance signaling to the cell block several neurotropic virus infections (i.e., Sindbis virus body could alarm and protect neuronal cell bodies but at the [SINV], HCMV, HSV-1) (Chaumorcel et al., 2008; Orvedahl same time might be repurposed for the replication of opportu- et al., 2007, 2010). The autophagy machinery normally is used nistic viruses that invade the nervous system (Figure 4). by cells to engulf cytosolic material in a double-membrane vesicle to be degraded by the lysosomes. HSV-1 encodes pro- What Controls Invasion of the Nervous System: teins that modulate autophagy in a neuron-specific manner. A Herpesvirus Perspective For example, autophagy is activated in response to HSV-1 infec- Alpha herpesvirus infection of immune-competent hosts tion through interferon-induced protein kinase R (PKR), which presents an apparent two-part paradox: First, during primary induces an antiviral signaling cascade in response to viral dsRNA infection, the infection passes with high efficiency from epithelial (Tallo´ czy et al., 2002). The HSV-1 protein ICP34.5 interferes with surfaces to the nerve terminals and on to the soma of PNS this pathway by binding to and inactivating Beclin-1, a host pro- ganglia neurons. Indeed, it is difficult to block primary PNS inva- tein responsible for the recruitment of other autophagy proteins sion once epithelial cells are infected. Second, once in the PNS,

Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. 389 Cell Host & Microbe Review

infection rarely spreads further to the brain. We learned from cles to the neuronal nuclei. Interestingly, retrograde transport of animal models and mutant virus infections that a complex inter- poliovirus to the CNS in mice is inefficient unless there is muscle play exists between the host immune system and the NS adapt- damage at the site of infection (Lancaster and Pfeiffer, 2010). For ed viral replication strategy. For instance, virulent PRV infection PRV, we and others have seen almost the opposite: tissue dam- of the skin induces a violent, inflamed peripheral pruritus with age blocks neural invasion (Duale et al., 2009; Koyuncu et al., almost no further spread to the brain in mice (Brittle et al., 2013). Perhaps these local events in the PNS ganglia or near 2004). By contrast, the attenuated mutant strain PRV Bartha, the infected axons contribute to determination of lytic versus which replicates as well as wild-type virus, produces almost no latent infection states and control of further neuroinvasion. tissue damage and peripheral inflammation yet is massively neu- It is still a mystery why spread of alpha herpesviruses to the roinvasive (Brittle et al., 2004). This property makes PRV Bartha CNS after the initial peripheral infection or after reactivation an excellent neural tracer (Brittle et al., 2004). from latency is rare. There is no evidence that the block is at What regulates controlled alpha herpesvirus spread in the NS, the cellular and molecular level of axonal transport to the CNS and how does the infection proceed differently in neurons, non- neurons. Perhaps the limitation is in the replication of virus in neuronal cells, PNS neurons, and CNS neurons? The answer CNS neurons. Like PNS neurons, these neurons may be less may lie within the specialized signaling and gene expression pat- permissive to lytic herpesvirus infection. CNS and PNS neurons terns that maintain highly polarized morphology and function of may both have specialized intrinsic immune responses that pro- neurons. Such a differentiated system calls for a long-distance mote abortive infections and long-term maintenance of latency. communication between the axon terminals that are in contact The suggestion that herpesvirus particles or genomes are pre- with epithelial or muscle cells and the cell bodies in the ganglia. sent in the brain of latently infected individuals draws attention This axon-to-cell-body communication must be finely tuned, so to possible correlation between herpesvirus infection and neuro- that cell bodies are well informed about distant events near axon pathologies like Alzheimer’s disease (Hill et al., 2005). Such cor- terminals but at the same time should not overreact. It is note- relations remain difficult to study due to the lack of appropriate worthy that PNS neurons produce little type I IFN after HSV-1 experimental models. infection (Leib et al., 1999; Szpara et al., 2010). However, when herpes virions invade epithelial cells, prior to infecting nerve Future Perspectives terminals, these infected cells produce a variety of inflammatory The interaction of viruses with their hosts is noteworthy in many and antiviral cytokines, including type I IFN. Such potent ways. Studying the complex interplay between these tiny intra- signaling molecules are likely to have an effect on local axon cellular entities and the infected cell in a culture dish has taught function and subsequent viral invasion of PNS neurons (Cun- us many valuable lessons in cell biology. However this interac- ningham and Mikloska, 2001). Indeed, previous work on human tion between virus and host in vivo is even more complicated dorsal root ganglion neurons in a chambered culture system by the hierarchical organization of cells, tissues, and systems. showed that IFN-a and IFN-g, added to the epidermal cell This complexity, while providing layers of protection against compartment, after axonal transmission of HSV-1 infection, microbial invasions, also requires strictly regulated communica- inhibit infection and spread (Mikloska and Cunningham, 2001). tion among cells and tissues to produce the appropriate protec- Certainly, information about events at nerve termini (i.e., dam- tive response. If this combined reaction to viral attack is not age, inflammation, and infection) must be transmitted to the sufficient, balanced, and controlled, severe pathology is likely. distant neuronal cell bodies to produce an appropriate response It is striking that there is no prototypical virus that spreads to at the site of axonal insult. Because some axons can be centime- the CNS. Viruses with RNA or DNA genomes and enveloped or ters to meters long, efficient and rapid transmission presents a naked virions, members of diverse virus families, all can spread potential problem for homeostasis of the neuron. How this long into the CNS under the right conditions. In general, even viruses distance communication is accomplished remains a funda- capable of infecting a variety of cell types tend to spread from mental problem in neuroscience and in virology. New findings one tissue to another until they reach a barrier. If the infection suggest that new protein synthesis in axons upon environmental is not contained locally (due to inefficient intrinsic and innate stimuli and subsequent transport of local signaling molecules to immune reaction), it can spread to vital organs, causing severe the nuclei might serve this function. Indeed a subset of mRNAs pathologies. Damage to cells can result from viral replication or and the complete protein synthesis machinery are localized to by the action of the activated immune system. Virus spread to axons in uninfected neurons, far away from the cell body (Vuppa- CNS can be deadly, not only because infected neurons may lanchi et al., 2009). Local translation in axons, not in the distant die, but also because of the immune-mediated inflammation in cell body, regulates growth cone navigation and integrity during the brain. We now recognize that we are constantly infected development and regeneration, and retrograde communication and colonized with numerous microorganisms, some of which with the cell body in adult neurons and plays a key role in axonal can affect spread and pathogenesis of invading microbes (the damage signaling (Jung et al., 2012). microbiome). This complex interaction of host and microbes We have found that local events involving axonal protein syn- was not recognized and was routinely neglected in in vitro and thesis in or near the initially infected axon termini facilitate neural in vivo experimental models. However, recent pioneering work invasion and spread (Koyuncu et al., 2013). Our results also sug- (Kuss et al., 2011; Matullo et al., 2011) suggests methods and gest that virus infection stimulates an axonal response similar to models of how to approach these complex and multifactorial injury signaling and that these damage or danger signals induced biological problems. by initial axonal infection and subsequent inflammatory events at Historically, research on PNS-adapted alpha herpesviruses, the site of primary infection affect axonal transport of viral parti- zoonotic RABV, and enteric poliovirus taught us much of what

390 Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. Cell Host & Microbe Review

we know about viral spread in the NS. Developing techniques Buchanan, R., and Bonthius, D.J. (2012). Measles virus and associated central such as live-cell and intravital imaging together with neuron nervous system sequelae. Semin. Pediatr. Neurol. 19, 107–114. culturing methods (in chambers or microfluidic devices, and co- Camarena, V., Kobayashi, M., Kim, J.Y., Roehm, P., Perez, R., Gardner, J., cultures with other cell types) accompanied by the ability to Wilson, A.C., Mohr, I., and Chao, M.V. (2010). Nature and duration of growth factor signaling through receptor tyrosine kinases regulates HSV-1 latency in construct recombinant viruses now allow researchers to study neurons. Cell Host Microbe 8, 320–330. some of the fundamental characteristics of virus replication Casiraghi, C., Dorovini-Zis, K., and Horwitz, M.S. (2011). Epstein-Barr virus and spread within and between neurons. These approaches infection of human brain microvessel endothelial cells: a novel role in multiple will enable new mechanistic understanding of how neurotropic sclerosis. J. Neuroimmunol. 230, 173–177. viruses gain access to and spread in the NS. In vitro culture of hu- Cearley, C.N., and Wolfe, J.H. (2007). A single injection of an adeno-associ- man neuronal/pluripotent stem cells, and closer-to-natural-host ated virus vector into nuclei with divergent connections results in widespread models such as ‘‘humanized mice’’ carrying human-specific vector distribution in the brain and global correction of a neurogenetic disease. genes, cells, or tissues will reveal host-specific aspects of lytic J. Neurosci. 27, 9928–9940. or latent infections caused by human viruses. Novel techniques, Chapagain, M.L., Verma, S., Mercier, F., Yanagihara, R., and Nerurkar, V.R. such as deep sequencing of viral nucleic acid from clinical (2007). Polyomavirus JC infects human brain microvascular endothelial cells independent of serotonin receptor 2A. Virology 364, 55–63. samples or single-molecule sequencing will facilitate identifica- tion of more neurovirulent and/or neuroinvasive virus mutants Chaumorcel, M., Souque` re, S., Pierron, G., Codogno, P., and Esclatine, A. (2008). Human cytomegalovirus controls a new autophagy-dependent cellular and will provide a genetic view of host barriers and viral bypass antiviral defense mechanism. Autophagy 4, 46–53. mechanisms. Screening of diseased tissues by these techniques Cheeran, M.C., Lokensgard, J.R., and Schleiss, M.R. (2009). Neuropathogen- will enable the detection of minute quantities of viral genomic esis of congenital cytomegalovirus infection: disease mechanisms and material, possibly revealing more roles for viruses in a variety prospects for intervention. Clin. Microbiol. Rev. 22, 99–126. of neuropathologies even if we cannot propagate them in the Cho, H., and Diamond, M.S. (2012). Immune responses to West Nile virus laboratory. infection in the central nervous system. Viruses 4, 3812–3830.

ACKNOWLEDGMENTS Cohen, J.I. (2007). VZV: molecular basis of persistence (latency and reactiva- tion). In Human Herpesviruses: Biology, Therapy, and Immunoprophylaxis, A. Arvin, G. Campadelli-Fiume, E. Mocarski, P.S. Moore, B. Roizman, R. Whitley, We thank the members of the Enquist lab, especially M.P. Taylor, R. Song, E. and K. Yamanishi, eds. (Cambridge: Cambridge University Press). Engel, K. Lancaster, and A. Ambrosini, for critical reading of the manuscript. L.W.E. acknowledges support from US National Institutes of Health grants Cunningham, A.L., and Mikloska, Z. (2001). The Holy Grail: immune control NS033506 and NS060699. I.B.H. is supported by fellowship PF-13-050-01- of human herpes simplex virus infection and disease. Herpes 8(Suppl 1 ), MPC from the American Cancer Society. 6A–10A.

De Regge, N., Van Opdenbosch, N., Nauwynck, H.J., Efstathiou, S., and REFERENCES Favoreel, H.W. (2010). Interferon alpha induces establishment of alphaherpes- virus latency in sensory neurons in vitro. PLoS ONE 5, e13076. Afonso, P.V., Ozden, S., Cumont, M.C., Seilhean, D., Cartier, L., Rezaie, P., Mason, S., Lambert, S., Huerre, M., Gessain, A., et al. (2008). Alteration of Deinhardt, K., Salinas, S., Verastegui, C., Watson, R., Worth, D., Hanrahan, S., blood-brain barrier integrity by retroviral infection. PLoS Pathog. 4, e1000205. Bucci, C., and Schiavo, G. (2006). Rab5 and Rab7 control endocytic sorting along the axonal retrograde transport pathway. Neuron 52, 293–305. Arbuckle, J.H., Medveczky, M.M., Luka, J., Hadley, S.H., Luegmayr, A., Abla- shi, D., Lund, T.C., Tolar, J., De Meirleir, K., Montoya, J.G., et al. (2010). The Desmaris, N., Bosch, A., Salau¨ n, C., Petit, C., Pre´ vost, M.C., Tordo, N., Perrin, latent human herpesvirus-6A genome specifically integrates in telomeres of P., Schwartz, O., de Rocquigny, H., and Heard, J.M. (2001). Production and human chromosomes in vivo and in vitro. Proc. Natl. Acad. Sci. USA 107, neurotropism of lentivirus vectors pseudotyped with lyssavirus envelope 5563–5568. glycoproteins. Mol. Ther. 4, 149–156.

Arpino, C., Curatolo, P., and Rezza, G. (2009). Chikungunya and the nervous Detje, C.N., Meyer, T., Schmidt, H., Kreuz, D., Rose, J.K., Bechmann, I., Prinz, system: what we do and do not know. Rev. Med. Virol. 19, 121–129. M., and Kalinke, U. (2009). Local type I IFN receptor signaling protects against virus spread within the central nervous system. J. Immunol. 182, 2297–2304. Ashley, S.L., Welton, A.R., Harwood, K.M., Van Rooijen, N., and Spindler, K.R. (2009). Mouse adenovirus type 1 infection of macrophages. Virology 390, Diefenbach, R.J., Miranda-Saksena, M., Douglas, M.W., and Cunningham, 307–314. A.L. (2008). Transport and egress of herpes simplex virus in neurons. Rev. Med. Virol. 18, 35–51. Beier, K.T., Saunders, A., Oldenburg, I.A., Miyamichi, K., Akhtar, N., Luo, L., Whelan, S.P., Sabatini, B., and Cepko, C.L. (2011). Anterograde or retrograde Duale, H., Hou, S., Derbenev, A.V., Smith, B.N., and Rabchevsky, A.G. (2009). transsynaptic labeling of CNS neurons with vesicular stomatitis virus vectors. Spinal cord injury reduces the efficacy of pseudorabies virus labeling of sym- Proc. Natl. Acad. Sci. USA 108, 15414–15419. pathetic preganglionic neurons. J. Neuropathol. Exp. Neurol. 68, 168–178.

Bennett, R.S., Cress, C.M., Ward, J.M., Firestone, C.Y., Murphy, B.R., and Dups, J., Middleton, D., Yamada, M., Monaghan, P., Long, F., Robinson, R., Whitehead, S.S. (2008). La Crosse virus infectivity, pathogenesis, and immu- Marsh, G.A., and Wang, L.F. (2012). A new model for Hendra virus encephalitis nogenicity in mice and monkeys. Virol. J. 5, 25. in the mouse. PLoS ONE 7, e40308.

Bergmann, C.C., Lane, T.E., and Stohlman, S.A. (2006). Coronavirus infection Farrell, M.J., Dobson, A.T., and Feldman, L.T. (1991). Herpes simplex virus of the central nervous system: host-virus stand-off. Nat. Rev. Microbiol. 4, latency-associated transcript is a stable intron. Proc. Natl. Acad. Sci. USA 121–132. 88, 790–794.

Bonthius, D.J. (2012). Lymphocytic choriomeningitis virus: an underrecog- Fish, K.N., Soderberg-Naucler, C., Mills, L.K., Stenglein, S., and Nelson, J.A. nized cause of neurologic disease in the fetus, child, and adult. Semin. Pediatr. (1998). Human cytomegalovirus persistently infects aortic endothelial cells. Neurol. 19, 89–95. J. Virol. 72, 5661–5668.

Boothpur, R., and Brennan, D.C. (2010). Human polyoma viruses and disease Fletcher, N.F., and McKeating, J.A. (2012). Hepatitis C virus and the brain. with emphasis on clinical BK and JC. J. Clin. Virol. 47, 306–312. J. Viral Hepat. 19, 301–306.

Brittle, E.E., Reynolds, A.E., and Enquist, L.W. (2004). Two modes of pseu- Fletcher, N.F., Wilson, G.K., Murray, J., Hu, K., Lewis, A., Reynolds, G.M., dorabies virus neuroinvasion and lethality in mice. J. Virol. 78, 12951–12963. Stamataki, Z., Meredith, L.W., Rowe, I.A., Luo, G., et al. (2012). Hepatitis C

Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. 391 Cell Host & Microbe Review

virus infects the endothelial cells of the blood-brain barrier. Gastroenterology Kuss, S.K., Etheredge, C.A., and Pfeiffer, J.K. (2008). Multiple host barriers 142, 634–643, e6. restrict poliovirus trafficking in mice. PLoS Pathog. 4, e1000082.

Goodrum, F., Jordan, C.T., Terhune, S.S., High, K., and Shenk, T. (2004). Kuss, S.K., Best, G.T., Etheredge, C.A., Pruijssers, A.J., Frierson, J.M., Hoop- Differential outcomes of human cytomegalovirus infection in primitive hemato- er, L.V., Dermody, T.S., and Pfeiffer, J.K. (2011). Intestinal microbiota promote poietic cell subpopulations. Blood 104, 687–695. enteric virus replication and systemic pathogenesis. Science 334, 249–252.

Gralinski, L.E., Ashley, S.L., Dixon, S.D., and Spindler, K.R. (2009). Mouse Lafaille, F.G., Pessach, I.M., Zhang, S.Y., Ciancanelli, M.J., Herman, M., adenovirus type 1-induced breakdown of the blood-brain barrier. J. Virol. Abhyankar, A., Ying, S.W., Keros, S., Goldstein, P.A., Mostoslavsky, G., 83, 9398–9410. et al. (2012). Impaired intrinsic immunity to HSV-1 in human iPSC-derived TLR3-deficient CNS cells. Nature 491, 769–773. Griffin, D.E., and Metcalf, T. (2011). Clearance of virus infection from the CNS. Curr Opin Virol 1, 216–221. Lancaster, K.Z., and Pfeiffer, J.K. (2010). Limited trafficking of a neurotropic virus through inefficient retrograde axonal transport and the type I interferon Hafezi, W., Lorentzen, E.U., Eing, B.R., Mu¨ ller, M., King, N.J., Klupp, B., Met- response. PLoS Pathog. 6, e1000791. tenleiter, T.C., and Ku¨ hn, J.E. (2012). Entry of herpes simplex virus type 1 (HSV-1) into the distal axons of trigeminal neurons favors the onset of nonpro- Langevin, C., Jaaro, H., Bressanelli, S., Fainzilber, M., and Tuffereau, C. (2002). ductive, silent infection. PLoS Pathog. 8, e1002679. Rabies virus glycoprotein (RVG) is a trimeric ligand for the N-terminal cysteine- rich domain of the mammalian p75 neurotrophin receptor. J. Biol. Chem. 277, Hill, J.M., Gebhardt, B.M., Azcuy, A.M., Matthews, K.E., Lukiw, W.J., Steiner, 37655–37662. I., Thompson, H.W., and Ball, M.J. (2005). Can a herpes simplex virus type 1 Leib, D.A., Harrison, T.E., Laslo, K.M., Machalek, M.A., Moorman, N.J., and neuroinvasive score be correlated to other risk factors in Alzheimer’s disease? Virgin, H.W. (1999). Interferons regulate the phenotype of wild-type and mutant Med. Hypotheses 64, 320–327. herpes simplex viruses in vivo. J. Exp. Med. 189, 663–672.

Jayakar, H.R., Jeetendra, E., and Whitt, M.A. (2004). Rhabdovirus assembly Lim, S.M., Koraka, P., Osterhaus, A.D., and Martina, B.E. (2011). West Nile and budding. Virus Res. 106, 117–132. virus: immunity and pathogenesis. Viruses 3, 811–828.

Johnson, R.T. (1983). Pathogenesis of La Crosse virus in mice. Prog. Clin. Biol. Lipkin, W.I., Briese, T., and Hornig, M. (2011). Borna disease virus - fact and Res. 123, 139–144. fantasy. Virus Res. 162, 162–172.

Jung, H., Yoon, B.C., and Holt, C.E. (2012). Axonal mRNA localization and local Lycke, E., Hamark, B., Johansson, M., Krotochwil, A., Lycke, J., and Svenner- protein synthesis in nervous system assembly, maintenance and repair. Nat. holm, B. (1988). Herpes simplex virus infection of the human sensory neuron. Rev. Neurosci. 13, 308–324. An electron microscopy study. Arch. Virol. 101, 87–104.

Jurgens, H.A., Amancherla, K., and Johnson, R.W. (2012). Influenza infection Lyman, M.G., Feierbach, B., Curanovic, D., Bisher, M., and Enquist, L.W. induces neuroinflammation, alters hippocampal neuron morphology, and (2007). Pseudorabies virus Us9 directs axonal sorting of viral capsids. impairs cognition in adult mice. J. Neurosci. 32, 3958–3968. J. Virol. 81, 11363–11371.

Kalinke, U., Bechmann, I., and Detje, C.N. (2011). Host strategies against virus Martin, K.A., Wyatt, R., Farzan, M., Choe, H., Marcon, L., Desjardins, E., Rob- entry via the olfactory system. Virulence 2, 367–370. inson, J., Sodroski, J., Gerard, C., and Gerard, N.P. (1997). CD4-independent binding of SIV gp120 to rhesus CCR5. Science 278, 1470–1473. Kamerman, P.R., Moss, P.J., Weber, J., Wallace, V.C., Rice, A.S., and Huang, W. (2012). Pathogenesis of HIV-associated sensory neuropathy: evidence Matullo, C.M., O’Regan, K.J., Curtis, M., and Rall, G.F. (2011). CNS recruit- from in vivo and in vitro experimental models. J. Peripher. Nerv. Syst. 17, ment of CD8+ T lymphocytes specific for a peripheral virus infection triggers 19–31. neuropathogenesis during polymicrobial challenge. PLoS Pathog. 7, e1002462. Kapitein, L.C., and Hoogenraad, C.C. (2011). Which way to go? Cytoskeletal organization and polarized transport in neurons. Mol. Cell. Neurosci. 46, 9–20. Mazarakis, N.D., Azzouz, M., Rohll, J.B., Ellard, F.M., Wilkes, F.J., Olsen, A.L., Carter, E.E., Barber, R.D., Baban, D.F., Kingsman, S.M., et al. (2001). Rabies Kapitein, L.C., Schlager, M.A., Kuijpers, M., Wulf, P.S., van Spronsen, M., virus glycoprotein pseudotyping of lentiviral vectors enables retrograde axonal MacKintosh, F.C., and Hoogenraad, C.C. (2010). Mixed microtubules steer transport and access to the nervous system after peripheral delivery. Hum. dynein-driven cargo transport into dendrites. Curr. Biol. 20, 290–299. Mol. Genet. 10, 2109–2121.

Kaul, M., Garden, G.A., and Lipton, S.A. (2001). Pathways to neuronal injury McArthur, J.C., McClernon, D.R., Cronin, M.F., Nance-Sproson, T.E., Saah, and apoptosis in HIV-associated dementia. Nature 410, 988–994. A.J., St Clair, M., and Lanier, E.R. (1997). Relationship between human immu- nodeficiency virus-associated dementia and viral load in cerebrospinal fluid Kirkegaard, K. (2009). Subversion of the cellular autophagy pathway by and brain. Ann. Neurol. 42, 689–698. viruses. Curr. Top. Microbiol. Immunol. 335, 323–333. McGavern, D.B., and Kang, S.S. (2011). Illuminating viral infections in the ner- Klingen, Y., Conzelmann, K.K., and Finke, S. (2008). Double-labeled rabies vous system. Nat. Rev. Immunol. 11, 318–329. virus: live tracking of enveloped virus transport. J. Virol. 82, 237–245. Mebatsion, T. (2001). Extensive attenuation of rabies virus by simultaneously modifying the dynein light chain binding site in the P protein and replacing Kobayashi, M., Wilson, A.C., Chao, M.V., and Mohr, I. (2012). Control of viral Arg333 in the G protein. J. Virol. 75, 11496–11502. latency in neurons by axonal mTOR signaling and the 4E-BP translation repressor. Genes Dev. 26, 1527–1532. Mercer, J., Schelhaas, M., and Helenius, A. (2010). Virus entry by endocytosis. Annu. Rev. Biochem. 79, 803–833. Kobiler, O., Lipman, Y., Therkelsen, K., Daubechies, I., and Enquist, L.W. (2010). Herpesviruses carrying a Brainbow cassette reveal replication and Metcalf, T.U., and Griffin, D.E. (2011). Alphavirus-induced encephalomyelitis: expression of limited numbers of incoming genomes. Nat Commun 1, 146. antibody-secreting cells and viral clearance from the nervous system. J. Virol. 85, 11490–11501. Koyuncu, O.O., Perlman, D.H., and Enquist, L.W. (2013). Efficient retrograde transport of pseudorabies virus within neurons requires local protein synthesis Mikloska, Z., and Cunningham, A.L. (2001). Alpha and gamma interferons in axons. Cell Host Microbe 13, 54–66. inhibit herpes simplex virus type 1 infection and spread in epidermal cells after axonal transmission. J. Virol. 75, 11821–11826. Kramer, T., Greco, T.M., Taylor, M.P., Ambrosini, A.E., Cristea, I.M., and En- quist, L.W. (2012). Kinesin-3 mediates axonal sorting and directional transport Mori, I., Nishiyama, Y., Yokochi, T., and Kimura, Y. (2005). Olfactory transmis- of alphaherpesvirus particles in neurons. Cell Host Microbe 12, 806–814. sion of neurotropic viruses. J. Neurovirol. 11, 129–137.

Kristie, T.M., and Roizman, B. (1988). Differentiation and DNA contact points of Ohka, S., Matsuda, N., Tohyama, K., Oda, T., Morikawa, M., Kuge, S., and host proteins binding at the cis site for virion-mediated induction of alpha Nomoto, A. (2004). Receptor (CD155)-dependent endocytosis of poliovirus genes of herpes simplex virus 1. J. Virol. 62, 1145–1157. and retrograde axonal transport of the endosome. J. Virol. 78, 7186–7198.

392 Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. Cell Host & Microbe Review

Ohka, S., Sakai, M., Bohnert, S., Igarashi, H., Deinhardt, K., Schiavo, G., and St Leger, A.J., and Hendricks, R.L. (2011). CD8+ T cells patrol HSV-1-infected Nomoto, A. (2009). Receptor-dependent and -independent axonal retrograde trigeminal ganglia and prevent viral reactivation. J. Neurovirol. 17, 528–534. transport of poliovirus in motor neurons. J. Virol. 83, 4995–5004. Stamatovic, S.M., Shakui, P., Keep, R.F., Moore, B.B., Kunkel, S.L., Van Rooi- Ohka, S., Nihei, C., Yamazaki, M., and Nomoto, A. (2012). Poliovirus trafficking jen, N., and Andjelkovic, A.V. (2005). Monocyte chemoattractant protein-1 toward central nervous system via human poliovirus receptor-dependent and regulation of blood-brain barrier permeability. J. Cereb. Blood Flow Metab. -independent pathway. Front Microbiol 3, 147. 25, 593–606.

Orvedahl, A., Alexander, D., Tallo´ czy, Z., Sun, Q., Wei, Y., Zhang, W., Burns, Szpara, M.L., Kobiler, O., and Enquist, L.W. (2010). A common neuronal D., Leib, D.A., and Levine, B. (2007). HSV-1 ICP34.5 confers neurovirulence response to alphaherpesvirus infection. J. Neuroimmune Pharmacol. 5, by targeting the Beclin 1 autophagy protein. Cell Host Microbe 1, 23–35. 418–427.

Orvedahl, A., MacPherson, S., Sumpter, R., Jr., Tallo´ czy, Z., Zou, Z., and Tallo´ czy, Z., Jiang, W., Virgin, H.W., 4th, Leib, D.A., Scheuner, D., Kaufman, Levine, B. (2010). Autophagy protects against Sindbis virus infection of the R.J., Eskelinen, E.L., and Levine, B. (2002). Regulation of starvation- and central nervous system. Cell Host Microbe 7, 115–127. virus-induced autophagy by the eIF2alpha kinase signaling pathway. Proc. Natl. Acad. Sci. USA 99, 190–195. Park, C.H., Ishinaka, M., Takada, A., Kida, H., Kimura, T., Ochiai, K., and Umemura, T. (2002). The invasion routes of neurovirulent A/Hong Kong/483/ Taylor, M.P., Kobiler, O., and Enquist, L.W. (2012). Alphaherpesvirus axon-to- 97 (H5N1) influenza virus into the central nervous system after respiratory cell spread involves limited virion transmission. Proc. Natl. Acad. Sci. USA 109, infection in mice. Arch. Virol. 147, 1425–1436. 17046–17051.

Planz, O., Pleschka, S., and Wolff, T. (2009). Borna disease virus: a unique Teijaro, J.R., Walsh, K.B., Cahalan, S., Fremgen, D.M., Roberts, E., Scott, F., pathogen and its interaction with intracellular signalling pathways. Cell. Micro- Martinborough, E., Peach, R., Oldstone, M.B., and Rosen, H. (2011). Endothe- biol. 11, 872–879. lial cells are central orchestrators of cytokine amplification during influenza virus infection. Cell 146, 980–991. Powers, A.M., and Logue, C.H. (2007). Changing patterns of chikungunya virus: re-emergence of a zoonotic arbovirus. J. Gen. Virol. 88, 2363–2377. Tirabassi, R.S., Townley, R.A., Eldridge, M.G., and Enquist, L.W. (1998). Molecular mechanisms of neurotropic herpesvirus invasion and spread in Proenc¸a, J.T., Coleman, H.M., Nicoll, M.P., Connor, V., Preston, C.M., Arthur, the CNS. Neurosci. Biobehav. Rev. 22, 709–720. J., and Efstathiou, S. (2011). An investigation of herpes simplex virus promoter activity compatible with latency establishment reveals VP16-independent Tsunoda, I., and Fujinami, R.S. (2010). Neuropathogenesis of Theiler’s murine activation of immediate-early promoters in sensory neurones. J. Gen. Virol. encephalomyelitis virus infection, an animal model for multiple sclerosis. 92, 2575–2585. J. Neuroimmune Pharmacol. 5, 355–369.

Racaniello, V.R. (2006). One hundred years of poliovirus pathogenesis. Ugolini, G. (2011). Rabies virus as a transneuronal tracer of neuronal connec- Virology 344, 9–16. tions. Adv. Virus Res. 79, 165–202.

Ren, R., and Racaniello, V.R. (1992). Human poliovirus receptor gene expres- Umbach, J.L., Kramer, M.F., Jurak, I., Karnowski, H.W., Coen, D.M., and sion and poliovirus tissue tropism in transgenic mice. J. Virol. 66, 296–304. Cullen, B.R. (2008). MicroRNAs expressed by herpes simplex virus 1 during latent infection regulate viral mRNAs. Nature 454, 780–783. Richardson-Burns, S.M., Kominsky, D.J., and Tyler, K.L. (2002). Reovirus- induced neuronal apoptosis is mediated by caspase 3 and is associated Verma, S., Kumar, M., Gurjav, U., Lum, S., and Nerurkar, V.R. (2010). Reversal with the activation of death receptors. J. Neurovirol. 8, 365–380. of West Nile virus-induced blood-brain barrier disruption and tight junction proteins degradation by matrix metalloproteinases inhibitor. Virology 397, Roizman, B., and Sears, A.E. (1987). An inquiry into the mechanisms of herpes 130–138. simplex virus latency. Annu. Rev. Microbiol. 41, 543–571. Vuppalanchi, D., Willis, D.E., and Twiss, J.L. (2009). Regulation of mRNA trans- Roussarie, J.P., Ruffie´ , C., Edgar, J.M., Griffiths, I., and Brahic, M. (2007). Axon port and translation in axons. Results Probl. Cell Differ. 48, 193–224. myelin transfer of a non-enveloped virus. PLoS ONE 2, e1331. Wagner, E.K., and Bloom, D.C. (1997). Experimental investigation of herpes Saheki, Y., and De Camilli, P. (2012). Synaptic vesicle endocytosis. Cold simplex virus latency. Clin. Microbiol. Rev. 10, 419–443. Spring Harb. Perspect. Biol. 4, a005645. Wang, T., Town, T., Alexopoulou, L., Anderson, J.F., Fikrig, E., and Flavell, R.A. Salinas, S., Bilsland, L.G., Henaff, D., Weston, A.E., Keriel, A., Schiavo, G., and (2004). Toll-like receptor 3 mediates West Nile virus entry into the brain causing Kremer, E.J. (2009). CAR-associated vesicular transport of an adenovirus in lethal encephalitis. Nat. Med. 10, 1366–1373. motor neuron axons. PLoS Pathog. 5, e1000442. Watanabe, M., Suyama, K., Hashimoto, K., Sato, M., Ohara, S., Abe, Y., Salinas, S., Schiavo, G., and Kremer, E.J. (2010). A hitchhiker’s guide to the Kawasaki, Y., Yamaguchi, S., Saijo, M., and Hosoya, M. (2013). Mumps nervous system: the complex journey of viruses and toxins. Nat. Rev. Micro- virus-associated acute encephalopathy: case report and review of the litera- biol. 8, 645–655. ture. J. Child Neurol. 28, 243–245.

Sawtell, N.M., and Thompson, R.L. (1992). Rapid in vivo reactivation of herpes Wilcox, C.L., Smith, R.L., Freed, C.R., and Johnson, E.M., Jr. (1990). Nerve simplex virus in latently infected murine ganglionic neurons after transient growth factor-dependence of herpes simplex virus latency in peripheral hyperthermia. J. Virol. 66, 2150–2156. sympathetic and sensory neurons in vitro. J. Neurosci. 10, 1268–1275.

Schneider-Schaulies, J., Meulen, Vt., and Schneider-Schaulies, S. (2003). Wilson, A.C., and Mohr, I. (2012). A cultured affair: HSV latency and reactiva- Measles infection of the central nervous system. J. Neurovirol. 9, 247–252. tion in neurons. Trends Microbiol. 20, 604–611.

Schweighardt, B., and Atwood, W.J. (2001). Virus receptors in the human Xu, Z., Waeckerlin, R., Urbanowski, M.D., van Marle, G., and Hobman, T.C. central nervous system. J. Neurovirol. 7, 187–195. (2012). West Nile virus infection causes endocytosis of a specific subset of tight junction membrane proteins. PLoS ONE 7, e37886. Smith, G. (2012). Herpesvirus transport to the nervous system and back again. Annu. Rev. Microbiol. 66, 153–176. Yordy, B., Iijima, N., Huttner, A., Leib, D., and Iwasaki, A. (2012). A neuron- specific role for autophagy in antiviral defense against herpes simplex virus. Solomon, T., Lewthwaite, P., Perera, D., Cardosa, M.J., McMinn, P., and Ooi, Cell Host Microbe 12, 334–345. M.H. (2010). Virology, epidemiology, pathogenesis, and control of enterovirus 71. Lancet Infect. Dis. 10, 778–790. Zaichick, S.V., Bohannon, K.P., Hughes, A., Sollars, P.J., Pickard, G.E., and Smith, G.A. (2013). The Herpesvirus VP1/2 Protein Is an Effector of Dynein- Speck, S.H., and Ganem, D. (2010). Viral latency and its regulation: lessons Mediated Capsid Transport and Neuroinvasion. Cell Host Microbe 13, from the gamma-herpesviruses. Cell Host Microbe 8, 100–115. 193–203.

Cell Host & Microbe 13, April 17, 2013 ª2013 Elsevier Inc. 393