The roles of herpes simplex virus in neuroscience
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Journal of NeuroVirology (1997) 3, 110 ± 125
ã 1997 Journal of NeuroVirology, Inc.
http://www.jneurovirol.com
Review
The roles of herpes simplex virus in neuroscience
Sharon L Turner1 and Frank J Jenkins2
2
Department of Pathology, School of Medicine; 2Department of Infectious Diseases and Microbiology, School of Public
Health; 1,2Division of Behavioral Medicine and Oncology, University of Pittsburgh Cancer Institute; 1,2Center for
Neuroscience, University of Pittsburgh, Pittsburgh, PA 15213, USA
Herpes simplex virus (HSV) has been a focus of research in many laboratories
during the last 30 ± 35 years, with the majority centered on the virus'
replication, molecular biology and pathogenesis. Recently, HSV has begun to
receive considerable attention in the field of neuroscience, where scientists
have begun to use the virus as a tool or model for several areas of investigation.
These areas include the construction and development of HSV-based vectors
for gene therapy and the use of HSV as a neuronal tracer, as a model for
demyelinating disease and to study interactions between the nervous, immune
and endocrine systems. The goal of this paper is to review these different roles
for HSV in the broad field of neuroscience.
Keywords: herpes simplex virus; neuroscience; viral vectors; latency;
demyelination; neuro-immune interactions
Introduction
Herpes simplex virus (HSV) refers to two closely HSV infections
related members of the human herpesvirus family; In a primary infection, HSV enters the body via
herpes simplex virus type 1 and herpes simplex mucosal membrane or abraded skin and establishes
virus type 2. HSV-1 and HSV-2 share a significant infection locally in epithelial cells. Active viral
homology at genetic and nucleic acid levels and can replication in these cells results in the amplification
produce similar diseases ranging from inapparent of virus, the formation of a `fever blister' and the
infections and self-limiting cutaneous lesions to stimulation of both cellular and humoral immune
fatal encephalitis (for a review, see Whitley and responses (as reviewed in Hill, 1985). HSV spreads
Gnann, 1993). In addition, the virus establishes and from infected epithelial cells to axons of sensory
maintains a latent state in sensory neurons from neurons innervating the site of local infection where
which recurrent HSV infections arise. it travels by retrograde axonal transport to the
The properties of HSV that make it a useful tool neuronal nuclei (Cook and Stevens, 1973). Studies
for studies in the field of neuroscience include its using animal models have indicated that infected
neurotropism, the ability to construct viral mutants neurons demonstrate either a limited viral replica-
that express foreign genes and its ability to establish tion or establish a latent infection (as discussed in
latent infections in neuronal cells. In this review we Cook and Stevens, 1973; Hill, 1985). While the exact
will discuss five different roles for HSV in mechanism(s) for the establishment of HSV latency
neuroscience which include (1) a model for have yet to be determined, recent studies have
demyelinating disease, (2) a tool for transneuronal suggested that the decision between active versus
tracing studies, (3) use as a viral vector for gene latent infection is made soon after neuronal
therapy, (4) a model for studying neuro-immune infection (Simmons et al, 1992; Speck and Sim-
interactions, and (5) a model for studying health mons, 1992).
effects of chronic stress. A complete discussion of HSV latency is characterized by the presence of
any of these roles would require its own separate viral genomes in the nuclei of infected neurons and
review and is beyond the scope of this review. the absence of viral replication or viral protein
Instead, we will briefly discuss each role and, in production (for review, see Hill, 1985). The infec-
instances where it is possible, point the reader to tion and establishment of latency within neurons
more detailed reviews on each subject. explains why HSV is termed a neurotrophic virus.
In a neuron in which the virus is latent, there are no
virus-specific proteins produced and thus the host's
Correspondence: FJ Jenkins immune system is unaware of the virus' presence.
Received 31 May 1996; revised 18 November 1996; accepted As a result, the immune system does not target the
5 December 1996 latently-infected neuron for destruction, and anti-HSV in neuroscience
SL Turner and FJ Jenkins
111
body levels to HSV antigens are not continuously its own transcription and the transcription of other
stimulated during viral latency. genes containing ICP4 repressor elements (DeLuca
A latent HSV infection is maintained for the life and Schaffer, 1988a; Muller, 1985; Watson and
of the host, but the virus can be reactivated Clements, 1980; Kristie and Roizman, 1986). ICP4
periodically to produce infectious virus and recur- deletion mutants must be maintained in cell culture
rent disease (as reviewed in Hill, 1985). During systems which provide the missing ICP4 gene in
reactivation, viral replication occurs within the trans to the virus. The use of the complementing
reactivated neuron, producing infectious viral cell lines to propagate ICP4 mutant viruses has
progeny. Mature virions are transported back down further delineated the functions of ICP4 as well as
the sensory axon where they may establish a assisted in the development of viral vectors.
recurrent infection in the epithelia of the skin. Similar to ICP4, another major regulatory pro-
Studies using both animal models and human tein, ICP0, exhibits more moderate transactivation
subjects have shown that viral reactivation can be of the b and g class genes (Cai and Schaffer, 1992).
triggered by a variety of stressful or stress-related Transient expression assays have also suggested
stimuli including heat, UV light, fever, hormonal that ICP0 is capable of trans-activating ICP4 and is
changes, menses and physical trauma to the neuron critical for the replication of HSV following
(e.g. Segal et al, 1974; Spruance, 1985; Carlton and transfection of viral DNA (Cai et al, 1993a; Cai
Kilbourne, 1952; Pazin et al, 1978). While the virus and Schaffer, 1989). While the synthesis of ICP0 is
appears to be latent most of the time, HSV infection not essential for viral replication in cell culture
is probably best characterized by recurrent reactiva- (Sacks and Schaffer, 1987; Stow and Stow, 1986), it
tions and periods of latency. does appear to be involved in viral replication in
animal models and in viral reactivation triggered
HSV replication by ganglion explanation (Leib et al, 1989a,b). These
The HSV genome is a linear, double-stranded DNA properties are best demonstrated in a recent study
approximately 150 Kb in length (McGeoch et al, by Cai et al (1993b) which utilizes nonsense,
1988). There are at least 75 separate open reading insertion and deletion mutants of ICP0 to delineate
frames encoded within this DNA genome. Molecu- the effects of mutant ICP0 peptides on viral
lar studies have revealed that many of these genes replication and reactivation. Plasmid constructs
can be deleted without interfering with the virus' containing the mutated ICP0 genes (which were
ability to replicate in cell culture lines (as reviewed used to generate the ICP0 viral mutants) exhibited
in Ward and Roizman, 1994). In addition, it is unique levels of transactivating activity as mea-
possible to construct site-specific mutations, in- sured by the ability of the expressed ICP0 peptides
cluding the deletion of viral genes and the insertion to stimulate transcription of ICP0 promoter-respon-
of foreign genes, into the viral genome (for example, sive genes in vitro. In a murine model for HSV
Post and Roizman, 1981; Roizman and Jenkins, replication and latency, using the ICP0 mutants, it
1985; Jenkins and Roizman, 1986). was reported that the efficiency of viral replication
HSV genes are divided into three major temporal in the eyes and trigeminal ganglia of inoculated
classes (a, b, and g) which are regulated in a mice, and the frequency of viral reactivation
coordinated, cascade fashion (for review see Roiz- induced by ganglion explant, correlated with the
man and Sears, 1993). The a or immediate-early (IE) levels of ICP0 transactivating activity. This data
genes contain the major transcriptional regulatory suggests a functional role for ICP0 in viral
proteins and their production is required for the replication and reactivation in vivo. However,
transcription of the b and g gene classes. b proteins levels of viral DNA as determined by PCR analysis,
are not produced in the absence of a proteins and did not correlate with the efficiency of viral
their synthesis is required for viral DNA replication. replication or reactivation, suggesting that other,
The b proteins consist primarily of proteins as yet undefined factors, may influence the ability
involved in viral nucleic acid metabolism. The g of the virus to establish neuronal latency.
proteins reach peak rates of synthesis after the onset HSV latency is defined by the presence of viral
of DNA replication and consist primarily of virus genomes and the absence of viral protein produc-
structural proteins. tion within infected cells. While this definition
Two HSV-1 a proteins that are relevant for the suggests that viral gene transcription is absent, a
discussions in this review are called ICP0 and ICP4 small subset of viral transcripts, termed the latency
(ICP=infected cell protein). ICP4 is the major associated transcripts (LATs) are transcribed during
regulatory protein of HSV and deletion studies have active and latent infections (Wechsler et al, 1988;
demonstrated that it is absolutely required for viral Wagner et al, 1988; Stevens et al, 1988; Krause et al,
replication (DeLuca and Schaffer, 1988b; Dixon and 1990; Rock et al, 1987; Roizman and Sears, 1993).
Schaffer, 1980; Preston, 1979; DeLuca et al, 1984, The function of the LAT transcripts is unclear. Viral
1985). The ICP4 protein is involved in the deletion mutants indicate that the LATs are not
transactivation of both b and g genes (as reviewed required to establish latent infection (Ho and
in Roizman and Sears, 1993), but may also repress Mocarski, 1989; Steiner et al, 1989). In addition,HSV in neuroscience
SL Turner and FJ Jenkins
112
contradicting data indicate variable roles for the CNS demyelination. Susceptible strains exhibit
LATs in viral reactivation. Some studies report multifocal demyelination, while moderately resis-
decreased reactivation frequencies in LAT mutants tant strains exhibit unifocal demyelination and
whereas others document no difference in reactiva- resistant strains simply exhibit an increased pre-
tion frequencies in LAT mutants and wild type sence of inflammatory cells at the TREZ (Kastrukoff
virus (Natarajan et al, 1991; Block et al, 1990; Perng et al, 1987). This study was followed by a more
et al, 1996; Maggioncalda et al, 1994). Such recent report which described investigations on the
inconsistencies may be partially explained by the effects of immunosuppression on TREZ demyelina-
different mutations in the different LAT mutants tion in strains of immunocompetent mice, immune
which include deletions in LAT promoter se- deficient mice, and immunocompetent mice that
quences and LAT coding sequences. An additional had been immunosuppressed with either cyclospor-
problem arises from the fact that the LAT gene is in A or cyclophosphamide (Kastrukoff et al, 1993).
located antisense to the ICP0 gene (Stevens et al, Animals that were immune deficient or immuno-
1987). The overlap between the LAT and ICP0 genes suppressed either lacked evidence of demyelination
makes it difficult to ascertain possible functions of or exhibited reduced levels of demyelination when
LAT since many LAT mutations commonly disrupt compared to immunocompetent mice. Taken to-
ICP0 sequences. As previously discussed, ICP0 gether, these studies clearly demonstrate that the
plays an important role in viral replication and immune system plays an important role in the
reactivation in vivo and therefore LAT mutations development of demyelination during an acute
may not always delineate the functions of LAT from HSV infection and that this response may be
those of ICP0. influenced by genetic determinants. In addition, it
was demonstrated that infectious virus was required
A model for virus-induced demyelinating disease for the development of demyelinating lesions and
Animal studies using peripheral inoculations of that the presence of virus alone was not sufficient.
HSV on the cornea, snout, footpad or vagina have Townsend (1981b) also reported that following a
demonstrated that the virus spreads from peripheral peripheral HSV infection, the virus spreads from the
epithelial cells to both the peripheral nervous PNS to the CNS where it initially infects and lyses
system (PNS) and central nervous system (CNS; astrocytes. This infection occurs prior to onset of
Townsend and Baringer, 1978a,b; Kristensson et al, demyelination. Following the infection of astro-
1978, 1979). Cellular damage induced by the virus is cytes, immune cells appear and demyelination
markedly different in these two areas. Tissue damage begins. The results of this study were later confirmed
in the PNS is generally mild while the damage in the by Itoyama et al (1991). Thus HSV-induced demye-
CNS is often extensive within a local foci. A well lination of the CNS appears to be due to a
studied example of this concept is seen within the combination of viral-induced lysis of astrocytes
trigeminal root entry zone (TREZ) of the brainstem, a and oligodendrocytes and immuno-pathology
junction region between the PNS and CNS (Town- caused by immune cell infiltrates.
send and Baringer, 1978a,b; Townsend, 1983). A puzzling aspect of HSV-induced demyelination
Within this region, following an acute HSV infec- is the development of lesions only within the CNS
tion, the peripheral myelin is untouched while the and not in the PNS. One explanation for this has
CNS side of the trigeminal root develops demyeli- been a reported inability of Schwann cells to be
nated lesions. Within the demyelinated lesions, productively infected with HSV. Electron micro-
there is an absence of myelin and the presence of scopic studies have reported that following a
both intact axons and a mononuclear cell (MNC) peripheral HSV infection (such as corneal inocula-
infiltrate. The exact mechanism of demyelination is tion) Schwann cells were found to contain viral
not known, but results from several studies suggest capsids, but not enveloped particles, suggesting an
that it is a combination of cellular infection and host abortive infection within these cells (Rabin et al,
immune response. Townsend and Baringer (1978b) 1968; Hill and Field, 1973). The inability of HSV to
reported that cyclophosphamide injections reduced productively infect and lyse Schwann cells would
the amount of both MNC infiltrate and myelin prevent the attraction of immune cells and therefore
destruction in the CNS following corneal HSV demyelination would not occur. This theory has
infections. Townsend (1981a) later reported that been supported in a study by Townsend and Collins
corneal infections in nude (athymic) mice resulted (1986) in which infection and lysis of Schwann cells
in an absence of demyelinating lesions. Kristensson and subsequent demyelination of the PNS was
et al (1982) demonstrated that the timing of achieved only by the direct inoculation of HSV into
administration of immunosuppressive agents such the sciatic nerve. A productive infection of the
as cyclophosphamide could either enhance or Schwann cells was established along with an
suppress the immune reaction producing either an immune cell infiltrate and subsequent demyelina-
increase or decrease in demyelination. Studies tion. These results suggest that during a peripheral
comparing genetically distinct strains of mice have HSV infection, there is some sort of block prevent-
identified strains which range in their resistance to ing the virus from productively infecting theHSV in neuroscience
SL Turner and FJ Jenkins
113
Schwann cells which is overcome by direct inocu- across synapses resulting in an absence of or
lation of the nerve. relatively weak labeling of linked neurons. In
Several investigators have also reported that addition, nonspecific labeling of adjacent neurons
following the development of demyelinating le- can occur at increased injection concentrations and
sions, some remyelination of the naked axons extended labeling times. Neurotropic viruses, spe-
occurs. Remyelination was first noted by Kristens- cifically herpesviruses, possess an advantage to
son et al (1979). Separate reports by Kristensson et these other contemporary methods in that they are
al (1982) and by Townsend (1983) reported that the able to replicate within neuron cell bodies provid-
remyelination appeared to be due to the migration ing signal amplification before infecting second-
of Schwann cells from the PNS to the CNS. and third-order neurons. The herpesviruses have
Schwann cell remyelination of CNS axons was later also been shown to specifically label neuronal
confirmed by Soffer and Martin (1988) using a connections in both the retrograde and anterograde
vaginal infection with HSV-2. In at least one study, direction (Ugolini et al, 1987; Norgren Jr, et al,
there was evidence of multiple episodes of demye- 1992).
lination followed by remyelination (Soffer and The most common viral transneuronal tracers are
Martin, 1988). In instances where HSV establishes herpes simplex virus 1 and 2 (HSV-1/HSV-2) and
latency and cycles through several rounds of pseudorabies virus (PRV). All three viruses belong
reactivation, it is reasonable to postulate that cyclic to the alpha herpesvirinae family and therefore are
episodes of demyelination followed by remyelina- neurotrophic (Roizman and Sears, 1993). The
tion also occur. ability of these DNA viruses to specifically infect
As all of the studies just described demonstrate, neurons contributes to their transneuronal trans-
HSV infection provides a good model for studies port. The most common method used to detect the
investigating the demyelination of CNS axons presence of these viruses in neuronal tissue is by
through a combination of cellular damage and immunohistochemical staining for viral antigen.
immunopathology (ie the cytotoxic effects of an Other methods that can be used include direct
immune cell infiltrate). Interestingly, HSV and HSV labeling of virus particles with S35 (by growth of
antibodies have been isolated from patients exhibit- viral stocks in the presence of S35-labeled amino
ing episodes of multiple sclerosis (MS), supporting acids) or by incorporation of the b-galactosidase
the relevance of this model to MS-related clinical gene in their DNA genome (Loewy et al, 1991). In
pathology (Bergstrom et al, 1989). Studies on the the former method, presence of virus is detected by
mechanisms of CNS demyelination and the long- photographic emulsion while in the latter method,
term effects of demyelination followed by remyeli- tissues are incubated with a chromogenic substrate
nation may serve as an excellent model for (such as X-gal) that detects the presence of the b-
trigeminal neuralgia (Burchiel, 1993). In addition, galactosidase enzyme.
the HSV model may be helpful in studies investigat- Viral tracings require certain precautions. In
ing other human demyelinating diseases such as order to correctly interpret viral tracing studies it
Bell's palsy and multiple sclerosis (Vahlne et al, is important to understand the mechanisms of viral
1985). uptake, transport, replication and release as well as
the generation of specific immune responses by host
A tool for transneuronal tracing studies immune cells and their associated neuropathology.
Neuroscientists have long desired to map chains of Careful analyses should identify transneuronal
neurons in order to identify communication path- connections by comparison of viral tracing results
ways from origin to termination. With the use of with conventional transneuronal tracers that are
antero- and retrograde tracers such as fast blue and injected concurrently to viral infection. Other
horseradish peroxidase it is possible to identify important considerations include the presence of
single neurons along with their axons and termina- alternate innervation pathways which could con-
tions. However, to identify synaptically linked tribute to nonspecific labeling; for example, non-
second and third order neurons, transneuronal specific labeling may result from dissemination of
tracers are required. virus from the site of inoculation to adjacent sites.
In order for a transneuronal tracer to be effective Experiments used to obtain transneuronal tracings
it must be specific for synaptically linked connec- are controlled by the strain of virus used, the host
tions, possess the ability to be transported antero- or animal, the site of injection, the amount of virus
retrogradely and be sufficiently tagged for efficient inoculated, the time of post-inoculation analysis,
and sensitive detection. Substances such as cholera and the extent of neuropathology associated with
toxin, tetanus toxin and wheat germ agglutinin are the infection. The importance of these parameters is
known to bind specifically at neuronal membranes apparent from many studies which have reported
and have been used as transneuronal tracers (Cabot false labeling under conditions of severe gliosis and
et al, 1984; Jansen et al, 1993; Hirakawa et al, 1993). at late survival times. (Norgren Jr and Lehman,
These methods have limitations however, since 1989; Strack and Loewy, 1990). However, careful
only small amounts of protein are transported analyses from several transneuronal tracing studiesHSV in neuroscience
SL Turner and FJ Jenkins
114
which contain proper controls have demonstrated of virus-specific transcription and translation, viral
the use of herpes viruses to identify transneuronal DNA synthesis, and the assembly of mature
connections (for example, Strack et al, 1989; nucleocapsids. Viral capsid envelopment and
Spencer et al, 1990; Jansen et al, 1992; Loewy et egress are complicated pathways which differ
al, 1991; Rotto-Percelay et al, 1992; Schramm et al, according to viral strain and the type of cell
1993; Hedner et al, 1993; Vizzard et al, 1995). infected. Regardless of the observed differences,
Reports from several studies have identified two processes seem most important for the produc-
properties of the herpesviruses that contribute to tion and passage of mature virions. The first
the specificity of transneuronal labeling. Vahlne involves the acquisition of a bilaminar membrane
and colleagues have studied the attachment and or transport vesicle which will permit fusion with
adsorption of HSV-1 and HSV-2 to glial cells, the plasma membrane and viral egress. Secondly,
neuronal perikarya and synaptosomal fractions of viral glycoproteins must be processed to their fully
brain homogenates (Vahlne et al, 1978, 1980). They functional form by association with the Golgi
reported that synaptosomal fractions of brain apparatus or golgi-derived vesicles which contain
homogenates exhibited the strongest viral adsorp- the enzymes required for these modifications.
tion followed by glial cells and neuronal perikarya. These processes are necessary to ensure that virions
In addition, it was observed that HSV-2 bound less will be released from the cell with the ability to bind
efficiently than HSV-1 in each respective fraction of via glycoprotein receptors to infect adjacent cells.
rat, human or monkey brain preparations. However, Our current understanding of the mechanisms by
HSV-1 and HSV-2 demonstrated equal binding in which virions traverse the cell are the result of
mouse glial cell preparations. Pseudorabies virus limited analyses offered by electron microscopy. In
uptake and retrograde transport studies have HSV infected neurons, nucleocapsids have been
reported the presence of limited viral specific observed obtaining a primary envelope at the inner
receptors on axon terminals which mediate viral lamina of the nuclear membrane. Virions are
uptake (Marchand and Schwab, 1986). These subsequently released into the endoplasmic reticu-
findings suggest preferential infectivity at the lum as naked nucleocapsids by a de-envelopment
synapse, but also illustrate that host species type, fusion with the outer nuclear membrane (Marchand
viral strain and different cell types influence the and Schwab, 1986; Lycke et al, 1988). HSV
path of infection. nucleocapsids receive a secondary envelopment
Electron microscopic studies using HSV and PRV by the endoplasmic reticulum and follow antero-
have demonstrated that fusion of the viral envelope grade transport to the cellular membrane by a
with the cellular plasma membrane of neuronal transport vesicle. Virus observed in the cytoplasm
extensions is followed by retrograde axonal trans- of dorsal root ganglion (DRG) neurons appear to be
port of unenveloped nucleocapsids along axonal contained within these transport vesicles. Fusion of
microtubules (Ugoline et al, 1987). Although this is vesicles with the plasma membrane at axonal
the primary mode of viral transport to the neuronal terminals releases infectious, enveloped virions
nucleus, it is not exclusive. Other studies have capable of infecting adjacent cells. Similar replica-
shown anterograde transport of virus (Norgren Jr, et tive processes in neurons of the vagal motor nucleus
al, 1992; Card et al, 1990), and at least one report have been reported in a study by Card and
suggests that the direction of transneuronal trans- colleagues using pseudorabies virus (PRV; Card et
port may be strain dependent (Zemanick et al, al, 1993). In this study, the authors reported that
1991). By analyzing labeled neurons at progressive PRV nucleocapsids obtained a single envelope from
timepoints, it has been determined that retrograde the inner nuclear membrane and traversed the inner
transport occurs much faster than anterograde lamina to gain access to the endoplasmic reticulum
transport (Norgren Jr, et al, 1992). Consideration of (ER). From the ER, virus fused with the outer ER
the difference in transport rate is important in membrane, entering the cytoplasm as naked nu-
tracing analyses and can be useful in determining cleocapsids which subsequently become enveloped
connections between groups of neurons. For exam- in a bilaminar structure by the trans cisternae of the
ple, in groups of neurons which are highly Golgi. Fusion of the outer virion envelope with
connected by collaterals, one must consider the synaptic membranes occurred at adjacent term-
fact that individual neuron labeling could be due to inals, releasing single enveloped virions with sur-
either antero- or retrograde transport, and in such face glycoproteins in the correct orientation to
instances, there may be no way to distinguish transneuronally fuse.
between originating and target cells. It is particu- Although the release of herpes viruses occurs at
larly important to consider this limitation when neuronal terminals, sites of virion egress do not
interpreting the specificity of viral tracing studies. always occur directly into synaptic clefts. Herpes-
Once delivered to the neuronal cell body, herpes containing vesicles have been reported to fuse at
virus nucleocapsids are observed at the pores of presynaptic terminals, releasing enveloped virus
nuclear membranes where they release their DNA which fuses to postsynaptic membranes adjacent to
into the nucleus. The viral replicative cycle consists presynaptic terminals. This results in the entry ofHSV in neuroscience
SL Turner and FJ Jenkins
115
naked nucleocapsids into the neuron. Astrocytes expression of the target gene within the neuronal
are also susceptible to PRV and HSV infection, but cells. Herpesviruses, particularly herpes simplex
infected cells are only observed subsequent to an virus type 1, have unique characteristics of infec-
adjacent neuronal infection. Ultrastructural ana- tion, replication and pathogenesis which make
lyses of PRV-infected astrocytes have revealed a them potentially ideal candidates for the develop-
defect in the cytoplasmic envelopment of viral ment of viral vectors capable of altering endogenous
nucleocapsids. This defect renders nucleocapsids gene expression or delivery of foreign genes both in
incapable of plasma membrane fusion, resulting in vivo and in vitro. The reader is directed to several
an absence of viral egress and an accumulation of reviews on these subjects (Glorioso et al, 1994,
virion particles within the cellular cytoplasm (Card 1995; Neuwelt et al, 1995; Kwong and Frenkel,
et al, 1993). The resulting abortive infection 1995).
effectively prevents astrocytic PRV virions from Herpes viruses have several advantages which
contributing to nonspecific extracellular spread. At lend to their ability to act as neuronal vectors. HSV
present, no such mechanisms are known for HSV. In exists as a large approximately 150 Kb double
fact, HSV has been reported to be quite capable of stranded DNA virus which has been entirely
establishing a productive infection in astrocytes sequenced and extensively studied (McGeoch et
(Itoyama et al, 1991). The inability of PRV to al, 1988). As a result of many years of intense
establish a productive infection in astrocytes research, a general knowledge exists of which genes
provides a great advantage to PRV in ensuring and DNA sequences may be deleted and at which
specific transneuronal transport and is a major sites foreign DNA may be inserted into the DNA
reason why PRV is considered by many to be the genome (as reviewed in Ward and Roizman, 1994).
virus of choice for transneuronal tracing studies. Research has also defined the minimal require-
Additional host mechanisms restricting virus ments for viral replication and packaging (as
spread to non-neuronal cells are provided by the discussed in Kwong and Frenkel, 1995). HSV-based
host's immune response to both PRV and HSV vector strategies rely on the ability of HSV to infect
infections. Resident microglia, monocytes and neuronal cells and establish a latent infection.
macrophages are activated in the nervous system The two main strategies for HSV-based vectors
during viral infection and may effectively phagocy- are genetically-engineered viruses and plasmid
tose virus and degenerating cellular debris (Rina- derived `amplicon' vectors. The first strategy
man et al, 1993). The importance of these involves the construction of recombinant viruses
mechanisms is apparent considering the large viral containing deletions in one or more viral genes
load which may be released from necrotic cells to whose expression is essential for viral replication
the extracellular space. T-lymphocytes may also (for reviews, see Glorioso et al, 1994, 1995). The
play a role in the delineation of viral spread (Zhao et deletion of genes whose expression is essential for
al, 1995). Factors regulating these mechanisms have virus replication (for example, ICP4), results in the
yet to be elucidated, but most likely involve production of a virus that cannot replicate in
immune-mediated cytokine production and the normal cells. Replication of these defective viruses
induction of major histocompatibility antigen ex- requires a cell line that provides the missing gene
pression within the nervous system. (or genes) in trans (termed a complementing cell
line). Foreign genes can be inserted into these
A viral vector for gene therapy mutated viral genomes with the goal of producing
The advances of modern molecular biology and in a virus vector that will infect the target cell (ie.,
vivo gene therapy have challenged neuroscientists neurons), and express the foreign gene without
with the potential prospect of gene manipulation in killing the cell. The second strategy involves the use
postmitotic neurons. The ability to alter gene of plasmid derived vectors containing HSV-1
expression in these cells would open the door origins of DNA replication and DNA packaging
towards potential therapies for several disorders signals which enable multiple copies of the vector
such as Parkinson's disease, Huntington's disease genomes to be packaged into helper virus virions
and amyotrophic lateral sclerosis. Gene therapy (for reviews see Neuwelt et al, 1995; Kwong and
using viral-based vectors has received considerable Frenkel, 1995). Helper viruses can be either
attention and represents a major focus of ongoing recombinant viruses containing a deletion within
research in many laboratories. Viral vectors using an essential viral gene or viruses containing
several different human viruses such as adeno- temperature-sensitive mutations that prevent repli-
viruses, retroviruses and herpes viruses are cur- cation at 378C (normal body temperature). In the
rently being developed. Gene therapy directed case of the former, the replication of the helper virus
towards neuronal cells however, presents unique and packaging of the amplicon vector DNA must
problems. These problems include the genetic occur in a cell line capable of complementing the
manipulation of post-mitotic (ie., non-dividing) mutations in the helper virus. Plasmid-derived
cells, the ability to specifically infect neurons, vectors (amplicons) are advantageous because the
long-term maintenance of the vector DNA and DNA constructs can be easily manipulated to testHSV in neuroscience
SL Turner and FJ Jenkins
116
endogenous, foreign, antisense or promoter gene Originally, strong promoter systems such as the
expression in the target cell. Although the effi- human cytomegalovirus IE promoter and the Rous
ciency of delivery of these multiple copy vectors is sarcoma virus long terminal repeat (RSV LTR) along
high, the primary disadvantage of this system is the with several HSV promoters were used to drive gene
fluctuating helper virus to amplicon ratio with expression. Although such promoter systems were
passage. These fluctuations must be monitored to capable of expression they were only active
ensure high amplicon delivery and experimental transiently in the CNS and did not result in long-
reproducibility in the absence of wild-type recom- term gene expression (as reviewed in Glorioso et al,
binants (Kwong and Frenkel, 1995). 1995). Neuronal specific promoters (such as the
Regardless of the vector system used, two neurofilament and neuronal-specific enolase pro-
primary goals must be achieved to enable long-term moters) which are believed to be constitutively
gene expression in neuronal cells. The first goal active in neurons, also produced only transient
involves the construct of mutant vectors which expression in the CNS (Andersen et al, 1993).
themselves are non-cytotoxic to cells. Several A second approach to achieving long term
studies have noted active expression of foreign expression in the CNS has been to use the LAT
genes by HSV vector constructs which subsequently promoters. As mentioned earlier, during HSV
become inactivated (for examples see Kwong and latency the only viral transcripts consistently
Frenkel, 1995; Johnson et al, 1992, 1994). Reasons detected are the latency associated transcripts
for the inactivation of the HSV construct are not (LATs). The possibility that the LATs are constitu-
apparent, but evidence suggests that it is a result of tively expressed in latently infected neurons has
cytotoxic effects induced by the vector. This made them strong candidates for long-term gene
problem has been shown in both plasmid derived expression in neuronal systems. Two latency active
and recombinant virus vectors alike. Several lines promoter (LAP) regions upstream of the LAT
of evidence suggest that viral immediate early (IE) transcripts have been identified, termed LAP1 and
gene expression may induce CPE (Johnson et al, LAP2 (Dobson et al, 1989; Goins et al, 1994).
1992). Two approaches to circumvent this include Transient expression assays and construction of
HSV vectors with multiple IE gene deletions and specific deletion mutants have suggested that the
vectors with deletions in both IE genes and the HSV LAP1 promoter is the major latency promoter.
VP16 gene. VP16 is a structural HSV protein Deletion of this promoter results in the production
responsible for transactivating the HSV IE genes of a virus that can establish latency but does not
(as reviewed in Roizman and Sears, 1993). It has express LAT transcripts (Dobson et al, 1989; Nicosia
been reported that a HSV mutant containing et al, 1993). The role of LAP2 is less clear, but
deletions in two of the five HSV IE genes, namely appears to be required for full expression of the LAT
ICP4 and ICP27, fails to express early and late viral transcripts (Glorioso et al, 1995). The use of LAT
proteins and is less cytotoxic than the parent virus promoters for gene expression has met with mixed
(Glorioso et al, 1994, 1995). Also, it has been success. LAP1 constructs have failed to produce
reported that a virus lacking VP16 and ICP4 exhibits long term expression in the CNS (Margolis et al,
reduced CPE when compared to an ICP4 deletion 1993; Lokensgard et al, 1994), but have been
mutant (Glorioso et al, 1995). HSV expresses a reported to demonstrate reasonable expression in
virion host shut-off (VHS) protein which is respon- the PNS (Dobson et al, 1989; Wolfe et al, 1992).
sible for degrading host cell RNA during an active Glorioso and coworkers have reported that a HSV
viral replication cycle (Read and Frenkel, 1983). construct containing a lac Z gene under the control
This protein may also be involved in cytopathic of the LAP2 promoter is expressed in the brain for at
effects although a recent study reported no reduc- least 41 h following inoculation (Glorioso et al,
tion in CPE using vectors containing mutations in 1995). Thus, while short term expression in either
the VHS gene (Johnson et al, 1994). Other packaging the CNS or PNS appears achievable, long term
alternatives may also be explored as ways to deliver expression in the CNS is still not a definite reality
HSV vectors that are incapable of establishing a and more research is required to produce an
lytic infection. One such method involves the use of acceptable HSV vector.
L particles, which are defective HSV virions lacking
viral DNA, yet are capable of entering cells without A model for studying interactions among the
causing host cell damage (Glorioso et al, 1994; nervous, immune and endocrine systems
Rixon et al, 1992). In immunocompetent hosts, active herpes simplex
The second goal in the development of HSV virus infections rapidly stimulate immune re-
vectors capable of long-term gene expression sponses which function to restrict viral replication
involves designing stable, active promoters capable and the spread of virus. In addition, the host's
of expressing appropriate levels of the foreign immune response is most likely involved in the
protein. The specific promoter involved in indivi- establishment of latent infections. In fact, the
dual therapies may change according to the type, establishment of latency could be viewed as a
status, and activity of the neuronal cell of interest. double-edged sword. It provides the host with aHSV in neuroscience
SL Turner and FJ Jenkins
117
mechanism to limit viral spread and cellular by a cellular protein termed Oct-1 (Gerster and
damage which, at the same time, ensures the Roeder, 1988; Kristie and Roizman, 1987; O'Hare
persistence and survival of the virus. In HSV- and Goding, 1988). The Oct-1 DNA complex is then
infected neonates and immunocompromised hosts, bound by a viral protein called the alpha trans-
virus replicates to high titers, is often widely inducing factor (aTIF) or VP16 (Gerster and Roeder,
disseminated and the resulting viral pathology 1988; McKnight et al, 1987; O'Hare et al, 1988;
generally extensive. While the mechanisms in- Preston et al, 1988). The aTIF protein is contained
volved in controlling HSV replication and establish- within infectious HSV viral particles and is there-
ment of latency are not completely understood, it fore introduced into the cell upon viral infection.
has become increasingly evident that they involve The binding of aTIF to the Oct-1/DNA complex
interactions among nervous, immune and endo- results in a transactivation of the HSV alpha gene
crine systems. The strongest evidence for these promoters and is responsible for the rapid induction
interactions is seen in the establishment of HSV of the HSV alpha genes immediately following
latency. infection. Since the complex between aTIF and
As mentioned earlier, several studies have sug- Oct-1 is required for the transcription of the a genes,
gested that within neurons, the decision to establish the Oct-1 protein represents a positive cellular
either latency or active viral replication occurs very transcriptional factor used in the regulation of
soon after infection (Simmons et al, 1992; Speck and HSV genes. The cellular Oct-1 protein belongs to a
Simmons, 1992). This is indicated by the identifica- family of proteins that are capable of binding to
tion of at least three distinct populations of neurons in DNA sequences containing octomer motifs such as
infected ganglia. These populations include neurons TAATGARAT.
that express HSV antigens but not LAT, neurons that Several laboratories have reported that HSV
express LAT but not HSV antigens, and a small replication is restricted in the mouse neuroblastoma
proportion of neurons that coexpress LAT and HSV cell line C1300 (Vahlne and Lycke, 1978; Kemp and
antigens (Simmons et al, 1992; Speck and Simmons, Latchman, 1989) and in neuronal cell lines derived
1992). It is reasoned that neurons which express HSV from dorsal root ganglia (Wheatley et al, 1990). The
antigens and no LAT represent active infections restriction of viral replication within these cells
while those that express LAT and no antigen occurs at the level of HSV a gene expression and
represent latent infections. Early divergence (to appears to be due to the presence of a cellular
active replication or latency) is further supported by repressor that binds to the TAATGARAT motif
the realization that lytic replication commits the cell (Kemp and Latchman, 1989; Werstuck et al, 1990).
to viral control and cell death as discussed below. Lillycrop et al (1991) defined the cellular repressor
The infection of cells by HSV rapidly results in as a neuronal homolog to the Oct-2 transcriptional
cell death. Consequences of HSV infection include factor. Oct-2 is related to Oct-1 and has been shown
the disruption of polyribosomes, inhibition and to be capable of binding to the TAATGARAT
shut-off of host macromolecular synthesis and sequence (Gerster and Roeder, 1988). Presumably
margination of nuclear chromatin (for review see Oct-2 repression occurs as a result of binding
Roizman and Furlong, 1974). The establishment of a (without the aTIF protein) to the TAATGARAT
latent HSV infection requires that the infected sequence in a promoters. They also demonstrated
neuron must survive infection. In order to prevent that this Oct-2 protein was expressed in dorsal root
cell death by HSV, viral replication must be stopped ganglion neurons. Increasing the levels of Oct-2 in
relatively early in its replication cycle. Once the the neuron cultures by transfecting a plasmid
infected cell is committed to viral DNA replication, containing the Oct-2 gene decreased the amount of
it is targeted for death. Therefore, to establish a HSV a gene transcription in these cells. It is
latent infection, the HSV replicative cycle must be important to note that Oct-2 was originally defined
stopped prior to b protein synthesis. Inhibition at as a transcriptional activator of genes in B cells
the level of a gene transcription would achieve this (Singh et al, 1986; Staudt et al, 1986) and had not
goal. The establishment and maintenance of HSV been attributed with any repressor activity. More
latency therefore, may be due to either the absence recent results have demonstrated that differential
of specific cellular transcriptional regulatory pro- splicing results in the formation of different iso-
teins that are involved in the positive induction of forms of the Oct-2 protein (Wirth et al, 1991), and
HSV a genes or to the presence of cellular proteins the predominant form in lymphocytes is distinct
that repress the specific transcription of the a genes. from the predominant form found in neurons
At the molecular level, the HSV a genes are (Lillycrop and Latchman, 1992). In transient ex-
unique with regard to the other HSV genes in that pression assays the lymphocyte form was reported
their promoter DNA sequences contain one or more to act as a positive transactivator of HSV IE
copies of the octomer DNA sequence motif TAAT- promoters while the neuronal form was found to
GARAT (R=Adenine or Guanine; Mackem and be a repressor of IE transcription (Lillycrop and
Roizman (1982); Kristie and Roizman (1984). In a Latchman, 1992). In addition, Wood and co-workers
lytic viral replicative cycle, this DNA motif is bound (Wood et al, 1992) reported that NGF couldHSV in neuroscience
SL Turner and FJ Jenkins
118
upregulate (three- to fourfold) the level of Oct-2 in ment of HSV latency which is dependent on
neuronal cultures from dorsal root ganglia. NGF had interactions between immune cells, the endocrine
no effect on the levels of Oct-1 or a neuronal- system and cells of the PNS (Jenkins and Baum,
specific transcriptional factor Brn-3. These results 1995, Figure 1). During an active HSV infection
suggested that NGF could mediate the transcrip- virus replicates peripherally and is transported to
tional effects of Oct-2 in neurons. the nuclei of the sensory ganglia innervating the site
Studies using nerve transection or nerve organ of infection. Viral replication occurs both at the
cultures have found that the NGF levels in non- peripheral site and in some neurons, resulting in an
neuronal cells increase following nerve injury immune response characterized by the attraction of
(Heumann, 1987; Heumann et al, 1987a,b). This activated macrophages, at both the peripheral site
increase in NGF was due, at least in part, to the and within the sensory ganglia. The activated
presence of activated macrophages that were macrophages produce a local increase in IL-1 and
responding to the site of injury, and this macro- TNF which in turn results in an increase of NGF.
phage-induced increase in NGF was subsequently The additional NGF is taken up by the neurons
found to be due to the secretion of the cytokine IL- creating an increase in the Oct-2 transcriptional
1 (Lindholm et al, 1987). Using cultured explants factor. This increase acts to suppress HSV a gene
of sciatic nerves, the addition of cultured medium transcription, leading to the establishment of
from activated macrophages or of IL-1 resulted in latency. Levels of Oct-2 would increase in neighbor-
a 14-fold increase in NGF mRNA (Lindholm et al, ing neurons as well, including both infected and un-
1987). This increase was transient and in order to infected cells. Neurons in which viral replication
maintain the higher levels of NGF mRNA, the IL-1 had proceeded beyond viral DNA replication, and
had to be replenished. This increase was not which express viral proteins, would most likely be
totally IL-1 specific since the addition of tumor targeted for destruction either through viral-in-
necrosis factor (TNF) or platelet-derived growth duced lysis or immune-mediated cytolysis. Under
factor resulted in a doubling of the NGF mRNA these conditions, latency would occur in those
levels. However, the addition of prostaglandin E2 neurons which contained viral DNA but had yet to
or fibroblast growth factor did not result in any initiate viral replication (ie expressing LAT but not
mRNA induction. In a subsequent study, Lind- HSV antigens). Thus the IL-1/NGF/Oct-2 mediated
holm and co-workers reported that the mechanism resistance to HSV transcription is somewhat analo-
of the IL-1-induced increase in NGF was an gous to viral resistance mediated by the interferons.
increase in the stability of the NGF mRNA That is to say that peripheral replication of HSV
(Lindholm et al, 1988). may induce a non-permissive environment for viral
Other evidence further supports a role for NGF in replication in neurons innervating the site of
the maintenance of neuronal latency. Wilcox et al replication. Animal models using corneal infec-
(1990); Wilcox and Johnson (1988); Wilcox and tions, have demonstrated that infectious HSV is not
Johnson Jr (1987), describe an in vitro model system detected within trigeminal ganglia until 48 ± 72 h
in which latency is established in cultured sympa- post-infection (Coen et al, 1989; Jenkins and Martin,
thetic and dorsal root ganglion neurons. This 1990; Tenser and Edris, 1987). This delay would
system relies upon the addition of NGF in addition allow for the establishment of a IL-1/NGF/Oct-2-
to antiviral drugs to establish latency. In this mediated non-permissive environment. This model
system, removal of NGF disrupts latency, lytic gene is based on studies performed in vitro using either
transcription resumes and viral reactivation occurs. established cell culture lines or primary cell lines. It
This model suggests that NGF is required for the will be important to determine if this model is
maintenance of neuronal latency since its removal correct for the establishment of HSV latency in vivo
results in viral reactivation. Later studies have using well-established animal models.
shown that the mechanism of NGF dependent
latency involves binding of NGF to the NGF A model for studying health consequences of
receptor (TrkA) and the stimulation of second psychological stress
messenger pathways (Smith et al, 1992). It has also In addition to the establishment and maintence of
been found that interruption of RNA transcription a latent infection, herpesvirus infections are also
or protein synthesis in these latently-infected cells characterized by recurrent infections that are
results in viral reactivation with similar kinetics to caused by stress or stress-related reactivations
NGF withdrawal. These data suggest the presence of from the latent state. Scientists in the fields of
a neuronal factor, regulated by NGF receptor neuroscience and behavioral medicine have long
binding, which contributes to the maintenance of desired to understand the mechanisms by which
the latent state. The similarities between this in psychological processes affect biological function.
vitro system and the in vivo condition have yet to be The reactivation of herpesviruses provides an
determined. excellent model for studying the direct relation-
Based on the studies described in this section, we ship between biobehavioral processes and dis-
have postulated a molecular model for the establish- ease.HSV in neuroscience
SL Turner and FJ Jenkins
119
Figure 1 Model for the establishment of HSV latency in sensory neurons. Sensory neurons are depicted as large triangular cells with
a circular nucleus. Surrounding the neurons are smaller satellite cells shown as small circles. A primary HSV infection occurs at a
peripheral site producing a `fever blister' (A). The virus is transported to the nuclei of sensory ganglia by retrograde transport along
axonal fibers (B). (1) and (2) represent un-infected and infected neurons respectively, surrounded by satellite cells. Virus replication
occurs in some of the neurons (3) resulting in the insertion of viral proteins on the surface of the neuron (4). Macrophages are attracted
to these neurons (5) where they become activated and begin producing IL-1 (6). The IL-1 causes an increase in the levels of NGF
produced by the satellite cells which is taken up by the neurons (7). Increased levels of NGF in the neuron result in an increase in the
levels of Oct-2 (8) which results in an inhibition of viral transcription resulting in the establishment of latency (9). Reprinted with
permission from Jenkins and Baum, 1995.
Studies in behavioral medicine have demon- strated an increase in HSV antibody levels in
strated the ability of psychological stress to induce chronically-stressed individuals. McKinnon et al
viral reactivation. Several studies from the Glaser's (1989) reported that individuals living within a
laboratories at Ohio State University have demon- three mile radius of the Three Mile Island nuclear
strated that antibody titers to latent herpesviruses, plant in Middletown, Pennsylvania demonstrated
particularly Epstein-Barr virus (EBV) are elevated in higher antibody titers to HSV than did control
subjects who have been chronically stressed as subjects living in Fredrick, Maryland. This increase
compared to non-stressed control groups (Kiecolt- is specific for HSV and not due to a nonspecific
Glaser and Glaser, 1986; Kiecolt-Glaser et al, 1991; polyclonal B cell activation since antibody titers to
Glaser et al, 1985; Kiecolt-Glaser et al, 1987). In one Rubella virus did not differ between groups. In
example, stress associated with exam-taking in first addition, increases in HSV antibody titers were not
year medical students resulted in significant in- correlated to a concurrent increase in the presence
creases in EBV antibody titers (Glaser et al, 1994). In of HSV lesions (A Baum, personal communication)
addition, the modulation of EBV antibody levels has suggesting that viral reactivation resulted in either
been used to define effective coping strategies in asymptomatic shedding or an abortive reactivation.
psychotherapy patients (Esterling et al, 1994). It is unlikely that these increases occur in indivi-
These types of studies suggest that the observed duals who regularly reactivate HSV, since their
increases in viral antibody titers is due to reactiva- antibody levels would be extremely high at base-
tion of latent virus which acts to trigger the immune line. The notion of an abortive reactivation that
response. A similar type of study has also demon- results in increases in HSV antibody titers withoutHSV in neuroscience
SL Turner and FJ Jenkins
120
viral shedding or recurrent disease is particularly The most common endpoint for detecting viral
intriguing. In these instances, antibody levels could reactivation has been the isolation of infectious
be used as a marker for stress-induced reactivation virus from either ganglia or tissues. Although this is
in individuals who do not routinely present with a fairly sensitive method, it fails to detect abortive
recurrent HSV infections. Since a majority of replication and therefore may not detect a large
individuals in the general population are felt to percentage of reactivation events. In order to further
have latent HSV in their sensory ganglia while only define distinct reactivation triggers, analyses utiliz-
a small percentage actually demonstrate recurrent ing methods such as immunohistochemistry, in situ
disease (as reviewed in Whitley and Gnann, 1993), hybridization and RNA PCR may be necessary.
titers to HSV antibodies may serve as a useful Another reactivation approach involves modula-
marker for stress-induced abortive reactivations in tion of the host immune response. When mice are
most people. preimmunized with HSV specific antibodies latency
The investigation of the role(s) of psychological is established in the absence of observable pathology
or physical stress in reactivation of latent herpes- and the hosts immune system remains naive to viral
viruses needs an appropriate animal model. While antigen (Birmanns et al, 1993). Subsequent exposure
animal models for the establishment of herpes to reactivation stimuli leads to reactivation as
latency exist, useful models for reactivation are identified by virus recovery and increased antibody
lacking. Existing in vivo animal models usually titers. This preimmunization protocol results in
reactivate at low frequencies and require stimuli increased reactivation percentages as compared to
that are often difficult to associate with human non-immunized mice (Birmanns et al, 1993).
phenomena. Animal models for HSV latency Although direct mechanisms for this phenomena
include both mouse and rabbit models in which have yet to be elucidated, increased levels of viral
peripheral inoculation with wild-type HSV (eg. on DNA were observed in ganglia of preimmunized
scarified corneas or in footpads) results in the mice. It should be considered that this facilitation of
establishment of latency in neurons of the sensory reactivation may be a function of increased detect-
ganglia inervating the site of inoculation (such as ability due to an unprimed immune response, and
the trigeminal or dorsal root ganglia). The most not altered triggering stimuli or a prevention of
common method for reactivation of latent HSV reactivation by an active component of the immune
(from sensory ganglia) is the explanation of response. With this in mind, the importance of
latently-infected tissue into cell culture medium effective and sensitive measures of reactivation
(Devi-Rao et al, 1994; Tanaka et al, 1994; Cai et al, becomes apparent.
1993b). Following explant, infectious virus can be Stress and its effects on the body and on different
detected in these tissues 1 ± 5 days later. Some systems in the body have recently become more
studies have utilized this method to evaluate viral thoroughly defined in biological respects. Key
mutants and different strains for reactivation factors, including the hypothalamic-pituitary-adre-
competency in efforts designed to identify viral nal axis and glucocorticoid and catecholamine
genes that are involved in reactivation (for exam- regulation, have been shown to affect the nervous,
ple, see Cai et al, 1993b). This procedure, although endocrine and immune systems (for reviews, see
highly efficient at reactivation (generally 80 ± 95% Besedovsky and Del Rey, 1991; Dunn, 1989; Black,
effective), represents an in vitro method which 1994; Blalock, 1989). A combined goal of behavioral
should not be interpreted as an in vivo reactivation scientists, neuroscientists and herpes virologists
protocol, even though it results in the reactivation should be to define these responses with respect to
of virus from latently-infected host tissue. viral reactivation. In addition, systemic responses
Reactivation procedures representing in vivo to psychological stressors need to be defined locally
protocols include epinephrine-iontophoresis in rab- at the site of viral latency and reactivation. These
bits, intravenous administration of cyclophospha- goals will only be met using multi-disciplinary
mide and dexamethasone, corneal scarification, approaches and collaborations.
ultraviolet light exposure, surgical trauma, adminis-
tration of cadmium sulfate, and hyperthermia Conclusions
(Sawtell and Thompson, 1992; Fawl and Roizman, Herpes simplex virus, as a result of its rather unique
1993; Harbour et al, 1983; Blyth et al, 1976; Bloom et life cycle in humans, is a useful tool for many areas
al, 1994). Most of these suffer from the criticism that of research. Its use in the field of neuroscience
they do not have an analogous use in humans. represents the newest and certainly one of the most
Ideally, what is needed is a reactivation protocol interesting and complex directions of research.
that is based on a psychological stressor and results Studies utilizing HSV in gene therapy, neuronal
in significant levels of viral reactivation. However, at tracings, demyelinating diseases, neuro-immune
the present time, such a protocol does not exist. The interactions and health related effects of psycholo-
major stumbling block may be the simple fact that gical stress will continue for many years and should
HSV is a human pathogen which doesn't behave provide important insights in areas that up to now
identically in non-human animal models. have been most difficult to study.You can also read
