<<

. 47
( 61 .)



>>



2.0 kb LAT
E
1.5 kb LAT


Fig. 1. HSV- I transcription in LAT region. (A) Graphic representation of HSV-1
genome which codes 75 proteins. L, long; S, short; TR, terminal repeat; IR, internal
repeat. (B) LAT genome is in 60th TRL and IRL region. Only the segment in the IR, is
shown. (C) Two HSV proteins (ICPO and ICP 34.5). (D) Transcription of 8.5-kb LAT
during acute infection. (E) The region 2.0-kb LAT and region 1.5-kb LAT.


ganglia, demonstrating that strains with very limited replication are capable of
establishing latent infection (54). Also, latent infection was found in ganglionic
neurons that showed no prior synthesis of viral proteins (55). Finally, neurons
with different developmental surface markers were found to differ in their capac-
ity to harbor latent viral DNA, providing evidence for the presence of neuronal
cell surface phenotypes that favor the establishment of latent infection (56).
Although the molecular mechanisms by which HSV establishes and maintains
latency are a focus of intense study, much remains unknown. LAT is the only HSV-
RNA that has been detected during neuronal latency. HSV- 1 LAT consists of a fam-
ily of overlapping transcripts encoded by diploid genes that map to the long repeat
region of the viral genome (Fig. 1B). Two colinear LAT, 2.0 and 1.5 kb, have been
detected by Northern blot analysis and/or ISH of RNA from latently infected ganglia
(Fig. 1E) (41,57,58). The LAT overlap the 3™-terminus of ICPO n-RNA (Fig. 1C) and
are transcribed from the complementary DNA strand (Fig. ID) (40,41,57,58). The
function of LAT in neuronal latency, however, has not been identified.
Hill, Wen, and Halford
306
The role that LAT mtght play m HSV-I replication and m the pathogenesrs of
latent mfectron m rabbits has been examined with mutant vrrus constructs
(17,18,.59-61) These constructs, which are deficient m the productron of LAT,
are capable of establishing and maintaining latent mfecttons in vtvo Addrtron-
ally, all of the mutants tested so far are replication-competent m cell culture and
m the rabbit eye. Thus, LAT 1s not required for HSV-1 rephcatron or for the
establishment or maintenance of latent infection (I 7,60,62)
LAT expressed during HSV-1 latency is encoded by an unusual gene Theo-
retrcally, either thts gene could act via a gene product or its function could be the
outcome of the action of a functional RNA. The 3™ end of the LAT region over-
laps with that of ICPO Therefore, LAT was suggested to act as an antisense
agent (58), but no supportmg evidence has been published. One LAT promoter IS
located more than 650 nucleottdes upstream of the 5™ end of the most abundant
LAT (2.0 kb). Analysts of LAT promoter viral mutants mdrcates that transcrtp-
non is mttlated close to and/or possibly interacts with components such as the
TATA box sp 1, LAT promoter bmdmg factor, CAMP response element,
long-term expression region, and CAAT box The transcription and sphcmg of
the 8.5 kb LAT are very complex and have not been elucidated fully (63-65)
Also, the relationshtp between the neuronal expression of this LAT promoter and
the repression of other possible HSV- 1 promoters during latent mfectron 1s not
understood (63-65). To date, no HSV-specific LAT translational product has
been Identified m vivo (63-6.5), and whether any HSV- 1 LAT protein is expressed
during latency remains to be determined.
2 Molecular mechanisms of reactrvatton Activity m neural tissue and shedding at
peripheral site: The molecular mechanisms of reacttvation, mcludmg the precise
event(s) that serve as triggers, remam unknown. A key unanswered question
regarding reactivation is how replication begins in the absence of Vmw65, a viral
protein that is assumed to stimulate the early stage of the HSV- 1 replication cycle
Perhaps a cructal step in the ability of HSV-1 to reactivate is the presence of
some viral or cellular factors that compensate for the initial lack of Vmw65, per-
mitting transactivation of immediate-early gene expression (63,65). Nerve growth
factor (NGF) could have a role m rendering a cell nonpermtsstve for viral reph-
cation. Deprivation of NGF m the rabbit ocular model resulted in reactivation of
latent HSV- 1 (66)
HSV-DNA encodes 75 protems. However, only 37 are required for growth m
culture (64) No virally encoded functions are required for establishment of the
latent state, but a specific gene(s) could be essential for efficient reactrvatton of
latent vtrus (64). The 8 5-kb LAT, or more likely some derivative of it, probably
functions to facilitate adrenergtcally induced (but not spontaneous) reactivation
m the rabbit eye model (17,18)
Although LAT is not required for establishment of latent infection in rabbit
trtgeminal ganglia, this transcript appears to be important for efficient reactiva-
tion of latent HSV-1 in vwo (I 7,18) Recent studies have shown that the rates of
ocular reactivation m VIVO are significantly lower for constructs lacking the LAT
HS V Latency 307

Table 3
Strain Specificity of HSV-1 Ocular Reactivation
Adrenergic induction
Spontaneous reactivation
during Pi d 20-39, after Pl d 42,
PE/TE (%)
HSV-I strain PE/TE (%)”
l/6 (17)
Rodanus 818 (100)
I 10/120 (90)
McKrae 93/120 (78)
17 Syn+ 9112 (75) 8/12 (67)
RE 214 (50) o/4 (0)
E-43 7/14 (50) 8114 (57)
316 (50) O/6 (0)
KOS
F 4/8 (50) O/8 (0)
3/14 (21) 3/12 (25)
SC-16
Macmtyre 2/20 (10) o/20 (0)
CGA-3 O/10 (0) o/20 (0)
OPE/TE = HSV-1 sheddmg positive eyes/total eyes (68)


region, compared with the rates for strains with the LAT region intact (I 7,18,59-
61) Perng et al (59) reported that rabbits latently infected with an HSV-1
McKrae strain mutant (dLAT2903) having a LAT deletion showed a 33%
decrease m the spontaneous reactlvatlon rate, compared with the parental strain.
These authors proposed the following possibility. In the rabbit ocular model of
HSV- 1 latency, two-thirds of spontaneous reactivatlons are LAT-dependent and
could, therefore, represent m vivo induced reactivation for which the inducing
factors have not yet been identified. The other one-third of spontaneous reactiva-
tlons are Independent of LAT and represent a subgroup of spontaneous reactlvatlon.
3. Spontaneous shedding, With HSV-1 latent m the trlgemmal ganglia, shedding of
virus on the ocular surface is the most commonly used sign of viral reactivation
in experimental models, including the rabbit. However, reactivation rates and
rates of spontaneous and induced shedding vary among HSV- 1 strains (Table 3)
In general, the McKrae and 17 Syn+ strains have the highest rates of both sponta-
neous (67,68) and induced (68) shedding, makmg them the most useful in rabbit
model studies of latency and reactlvatlon. The Rodanus strain, which has a high
rate of spontaneous reactivation, cannot be induced, which limits its value m the
laboratory The Macmtyre and CGA-3 strains show nexther spontaneous nor
induced shedding, suggesting that for these strains, the mechanisms that allow
the establishment and maintenance of neuronal latency are separate from the pro-
cesses that trigger reactivation and ocular shedding m vivo (68).
Shedding frequencies for mdlvidual HSV-1 strains, as determined by the ocu-
lar swab procedure, vary with time after maculation. For the McKrae stram, spon-
taneous shedding was found in 80% of 20 eyes during postinoculation d 40-80,
but only 35% over d 8 l-180, and 25% over d 181-220 (69). In another study,
Hill, Wen, and Ha/ford
308
shedding was observed m 78% of 120 eyes over d 20-39 pi, however, approxi-
mately three-fourths of these episodes occurred m the first half of this period and
only one-fourth m the second half (23).
The frequencies of shedding after adrenergic mductron also vary significantly
with the strain, but not necessarily in correlation with spontaneous shedding
(Table 3). In a study of rabbits latently infected with 10 strains of HSV-1 (68),
four groups of viruses could be distmgmshed m terms of reactivation frequencies.
a. Very little or no spontaneous reactivation and no induced reactivatron
(Macmtyre, CGA-3),
b Little spontaneous or Induced reactivation (SC 16),
c Spontaneous reactivation, but no induced reactivation (Rodanus, RE, F, and
KOS), and
d Spontaneous and induced reactivation (McKrae, 17 Syn+, and E-43).
Despite the differences in spontaneous and induced reactivation frequencies, all of
the wild-type HSV-1 strains produce LAT However, as noted, studies with LAT
deletion mutant have demonstrated reduced spontaneous reactivation rates, sug-
gesting that LAT gene also has an effect on the frequency of spontaneous shedding
(59). Taken together, these results suggest that one or more components outside of
the LAT gene are involved in HSV spontaneous and induced reactivation (59,68)
4. Induction of HSV-1 specific cornea1 epithehal lesions Both adrenergic ionto-
phoresrs (46) and mrmunosuppression (49) can induce specific herpetic epithe-
ha1 lesions In the first report of adrenergrc induction of cornea1 lesions, Hill et
al (46) described a positive association of lesions and positive eye swabs after
ocular iontophoresis of 6hydroxydopamme followed by topical apphcation of
Propme (dipivefrin hydrochloride, a prodrug of epmephrme) m HSV- 1 latently
infected rabbits. Deep punctate, dendritic, and geographic cornea1 lesions were
observed Of the 36 eyes with positive swabs, 24 (67%) also had cornea1 lesions.
Of the 13 eyes with dendritic lesions, 10 (77%) were associated with positive
HSV- 1 eye swabs. Conversely, of 12 1 eyes with negative swabs, 105 (87%) were
also negative for HSV-1 cornea1 lesions In the first report of nnmunosuppres-
sion-induced ocular herpetic lesions (49), deep punctate, dendrmc, and geo-
graphic cornea1 epithehal lesions were observed after mtravenous administration
of cyclophosphamrde and dexamethasone (49) Also, a positive correlation was
noted between the induced lessons and recovery of HSV- 1 from ocular swabs
5. HSV- 1 cornea1 latency One of the most controversral aspects of HSV- 1 latency
1swhether the herpesvirus is capable of existing in the latent state m the cornea,
and if so, can cornea1 latency lead to vnus reactivation, shedding, and disease
The definition of HSV-1 latency proposed by Gordon et al. (70) states that the
virus must be capable of exrstmg m a host cell for an extended period of time in
a virion-free state and, upon appropriate stimulation, must reactivate to produce
infectious progeny virions. To verify the existence of cornea1 latency, four neces-
sary and sufficient experimental conditions must be fulfilled:
a. A functional HSV-1 genome capable of reactrvation to produce mfectious
HSV- 1 progeny must be detected,
309
HSV Latency
b. No intact vnion must be vistble by electron microscopy;
c. No HSV-1 transcrtption can occur, with the exception of LAT; and
d. No HSV- 1 protein expression can occur.
Three alternatives that must be ruled out are.
a. Residual defective HSV-1 DNA from a previous cornea1 infection;
b. Viral persistence in the cornea with a very slow turnover of a very small num-
ber of infectious vlrtons; and
c. Spontaneous shedding of virus from neuronal cells via anterograde axoplas-
mtc flow with coincidental detection in the cornea (with or without replication).
One problem, however, is that these strictures are based on the assumption that
cornea1 latency mimics gangliomc latency; if this is not the case, some or all of
these criteria may not apply.
ISH studies (72) have found that HSV-specific probes hybridized to DNA
extracted from cornea1 eptthelial, stromal, and endothelial cells obtained from
HSV-1 latently infected rabbits More recently, PCR demonstrated LAT-RNA m
9% of corneas and m 100% of trigeminal ganglia from latently infected rabbits
during postinoculation d 41-147 (72); HSV-DNA was detected in 60% of the
corneas In one study (73); HSV was recovered from 8-l 1% of long-term cell
cultures from rabbit corneas with no signs of active chmcal disease at postmocu-
lation d 118 (73); in another study (74), HSV-specific DNA (TK gene region)
was recovered in 55% of cornea1 cell cultures from HSV-infected rabbits at post-
inoculatton d 118, even though no vn-us was detected tn the cell cultures over a
44-d culture period. Taken together, these findings indicate that at least two popu-
lations of HSV-containing cells exist in the cornea after herpettc keratms One
population (approx 10%) contains vtrus that can be reactivated using conventional
culture methods and a second population (approx 50%) contains viral DNA, but in a
form that does not reactivate with routine cell culture techniques.
Although the results of many studies provide data that are consistent with the
idea of HSV-1 cornea1 latency m rabbits (15,72-74), a review of the literature
indicates that no single study has investigated the issue of cornea1 latency using
all of the three essential approaches: cocultivatton, electron microscopy, and bio-
chemical analyses (70). Thus, definitive answers remam elusive.
In the rabbit model, a variety of evidence points to the largely neuronal origin
of reactivated HSV- 1. Twenty-five years ago, Brown and Kaufman (75) reported
the recovery of HSV from the ocular secretions of eight of 22 enucleated rabbits
latently infected with HSV- 1. This study, mdicatmg that virus liberation occurs
in the absence of the globe, further confirms that the globe, I.e., the cornea itself,
cannot be the sole source of HSV- 1 reacttvatton.
Transplantation studies (76) have also indicated that, at least for induced reac-
tivation in the rabbtt model, the source of the virus is not the cornea. Unmfected
rabbits that received donor cornea1 tissue from latently infected rabbits did not
shed virus after either adrenergic iontophoresis within the limits of the donor
cornea or systemic immunosuppresston with cyclophosphamlde and dexa-
methasone. However, latently infected rabbits receiving corneas from uninfected
310 Hill, Wen, and Halford

donors did shed virus with systemic tmmunosuppresslon, but not wtth adrenergrc
tontophoresis only wrthm the limrts of the donor cornea This suggests that neu-
ronal latency m the host IS necessary for reactivation m the rabbit model, and that
intact cornea1 nerves are necessary for adrenerglc lontophoretlc mductlon of ocu-
lar shedding of latent HSV- 1

Acknowledgments
This work was supported in part by US Publtc Health Service grant EY063 11
and EY02377 and a subcontract to EY09171 from the National Eye Instttute, as
well as a subcontract to A106246, National Institutes of Health, Bethesda, MD.

References
1 Cushmg, H. (1905) Surgtcal aspects of maJor neuralgia of trtgemrnal nerve.
reports of 20 cases of operation on the Gasserlan ganghon with anatomtc and
physrologtc notes on the consequences of removal JAMA 44, 1002-1008.
2 Gruter, W (1920) Experrmentelle und klmische Untersuchungen uber den Sog
Herpes corneae Kiln Monatsbl Augenhedk LXV, 398
3 Lowenstem, A. (19 19) Aettologtsche Untersuchungen uber den fieberhaften Her-
pes Munch Med Woch LXVI, 769
4 Goodpasture, E W and Teague, 0. (1923) Transmrsslon of the virus herpes
febrths along nerves m expertmentally infected rabbrts J Med Res 44, 12 l-l 84
5 Goodpasture, E. W. (1929) Herpetic mfecttons wtth special reference to mvolve-
ment of the nervous system Medtcme 8,223-243.
6. Stevens, J G. and Cook, M. L. (1971) Latent herpes stmplex V˜I.IS m spinal gan-
glia of mice. Scrence 173, 843-845
7 Stevens, J. G , Nesburn, A B., and Cook, M. L (1972) Latent herpes simplex
virus from trtgemmal gangha of rabbtts with recurrent eye infection Nature New
Blol 235,216217
8 Cook, M. L. and Stevens, J G. (1973) Pathogenests of herpetic neuritis and gan-
ghomtis m mice. evrdence of mtra-axonal transport of mfectron Infect Immunol
7,272-288.
9. Wagner, E. K , Guzowskt, J F., and Smgh, J. (1995) Transcription of the herpes
simplex vuus genome during productrve and latent infectton, in Progress zn
Nucleic Acid Research and Molecular Biology, vol. 51, (Cohn, W H and
Moldave, K., eds.), Academic, pp 123-165
10 Roizman, B. (1990) Whither Herpesvirus? Immunoblology and Prophylaxis of
Human Herpesvzrus Infections (Lopez, C , Mori, R , Rorzman, B., and Whltley,
R. J , eds ), Plenum, NY, pp 285-291
11 Liesegang, T J (1992) Biology and molecular aspects of herpes stmplex and
Varrcella-zoster virus mfectrons. Ophthalmology 99, 781-799
12. Kaufman, H. E and Rayfield, M A. (1988) The Cornea vnal conJunctivitts and
keratrtrs (Kaufman, H. E , Barron, B. A , McDonald, M B., and Waltman, S. R ,
eds.), Church111 Lrvmgstone, NY, pp 299-33 1
HSV Latency 311

13 Nesburn, A B., Burke, R L , Ghiasi, H , Slanina, S., and Wechsler, S. L (1994)
Vaccine therapy for ocular herpes simplex virus (HSV) infection* Periocular vac-
cination reduces spontaneous ocular HSV type 1 sheddmg m latently infected
rabbits. .J. Vwol 68, 5084-5092
14 Centtfanto-Fitzgerald, Y. M., Varnell, E. D., and Kaufman, H. E. (1982) Initial

<<

. 47
( 61 .)



>>