Partial rescue of herpes simplex virus neurovirulence with a 3.2 kb cloned DNA fragment

Herpes simplex virus type 1 mutants containing the KOS strain ICP34.5 gene in place of the McKrae ICP34.5 gene have McKrae-like spontaneous reactivation but non-McKrae-like virulence

Herpes simplex virus type 1 (HSV-1) strain McKrae is neurovirulent in rabbits infected by the ocular route, causing fatal encephalitis in approximately 50% of the animals, and has a high-level spontaneous reactivation phenotype, with 10% of rabbit eyes containing reactivated virus at any given time. In contrast, HSV-1 strain KOS is completely avirulent (no rabbits die) and has a completely negative spontaneous reactivation phenotype. Mutations of the ICP34.5 gene can reduce the neurovirulence of HSV-1 strains McKrae and 17syn(+) by up to 100000-fold. ICP34.5 mutants also have reduced spontaneous reactivation phenotypes. To determine whether differences in the ICP34.5 gene might be involved in the reduced neurovirulence and spontaneous reactivation phenotypes of KOS compared with McKrae, we constructed chimeric viruses containing the KOS ICP34.5 gene in place of the McKrae ICP34.5 gene. Rabbits ocularly infected with the chimeric viruses had a high spontaneous reactivation phenotype indistinguishable from McKrae. In contrast, neurovirulence of the chimeric viruses was decreased compared with McKrae. Thus, one or more 'defects' in the KOS ICP34.5 gene appeared to be at least partially responsible for the reduced neurovirulence of KOS compared with McKrae. However, there appeared to be no 'defect' in the KOS ICP34.5 function required for efficient spontaneous reactivation.

Histopathological responses in the CNS following inoculation with a non-neurovirulent mutant (1716) of herpes simplex virus type 1 (HSV 1): relevance for gene and cancer therapy

The RL1 gene of herpes simplex virus (HSV) encodes a polypeptide, ICP34.5 which is a specific virulence determinant. RL1 null mutants fail to replicate in both the PNS and CNS and are incapable of causing encephalitis. Additionally, RL1 null mutants have the capacity to replicate in actively dividing cells but fail to replicate in growth arrested or terminally differentiated cells. This selective replication phenotype has highlighted their use as both tumour killing agents and gene delivery vehicles particularly to the nervous system. Before their full potential can be assessed, however, it is necessary to determine the pathological and immune responses induced following direct intracerebral inoculation. Fourteen mice were injected in the left cerebral hemisphere with a high dose of the HSV-1, RL1 null mutant 1716. At regular time intervals up to 28 days, the mice were killed and the distribution of virus antigen, histopathological changes and immune responses in the CNS determined by H & E staining and immunohistochemistry. Control mice were injected with either wild type HSV-1 or buffer. At early times post-inoculation with 1716, there is a low grade meningoencephalitis with a limited inflammatory response. This is accompanied by virus antigen expression confined to the site of inoculation. By 28 days the CNS is histopathologically normal; virus antigen and immune responses are no longer detectable. These findings demonstrate that infection of the CNS by RL1 null mutants of HSV results in a finite, self-limiting response and highlights their potential for therapeutic use.

Herpes simplex virus genome replication and transcription during induced reactivation in the rabbit eye

PCR analysis of herpes simplex virus (HSV) genome replication and productive-cycle transcription was used to examine the role of the cornea in the latency-associated transcript (LAT)-mediated reactivation of HSV type 1 (HSV-1) in the rabbit eye model. The reduced relative reactivation frequency of 17 delta Pst (a LAT- virus) compared to those of wild-type and LAT+ rescuants correlated with reduced levels of viral DNA and transcription in the cornea following epinephrine induction. The timing of virus appearance in the cornea was most consistent with tissue peripheral to the cornea itself mediating a LAT-sensitive step in the reactivation process. Specific results include the following. (i) While viral DNA was found in the corneas of rabbits latently infected with either the LAT+ or LAT- virus prior to and during the first 16 to 24 h following induction, more was found in animals infected with the LAT+ virus. (ii) A significant increase in levels of viral DNA occurred 20 to 168 h following induction. (iii) The average relative amount of viral DNA was lower at all time points following reactivation of animals infected with the LAT- virus. (iv) Expression of productive-cycle transcripts could be detected in corneas of some rabbits latently infected with either the LAT+ or LAT- virus, and the amount recovered and the timing of appearance differed during the reactivation of rabbits latently infected with the LAT+ or LAT- virus. (v) Despite the reduced recoveries of LAT- virus DNA and productive-cycle transcripts in reactivating corneas in vivo compared to those of their LAT+ counterparts, such differences were not detected in cultured keratinocytes or in experiments in which relatively high titers of virus were superinfected into the eyes of latently infected rabbits. (vi) A number of LAT(+)-virus-infected rabbits expressed LAT in corneas isolated from uninduced rabbits. When seen, its amount was significantly higher than that of a productive-cycle (VP5) transcript.

Evidence that the herpes simplex virus type 1 uracil DNA glycosylase is required for efficient viral replication and latency in the murine nervous system

Herpes simplex virus (HSV) encodes a uracil DNA glycosylase (UNG; UL2), which has been shown to be dispensable for normal replication of HSV-1 in cultured cells (J. Mullaney, H.W. Moss, and D.J. McGeoch, J. Gen. Virol. 70:449-454, 1989). In adult neurons, UNG activity is undetectable (F. Focher, P. Mazzarello, A. Verri, U. Hubscher, and S. Spadari, Mutat. Res. 237:65-73, 1990), suggesting that the HSV-1 UNG may play an important role in viral replication in neurons acutely and/or following reactivation. To examine the contribution of the HSV-1 UNG in vivo, two independent strain 17 Syn+ Ung- mutants, designated uB1 and uB2, were examined in a mouse model of herpetic disease. Following direct intracranial inoculation, both mutants exhibited a 10-fold reduction in neurovirulence compared with the parental strain 17 Syn+. Inoculations by a peripheral route demonstrated that the Ung- mutants were at least 100,000-fold less neuroinvasive than 17 Syn+. Replication kinetics in vivo demonstrated that uB1 and uB2 replicated less well in both the mouse peripheral and central nervous systems. Latency was established by both of the mutants in 100% of the animals examined. Following transient hyperthermia, however, the frequency of reactivation of the mutants in vivo was dramatically reduced. Restoration of the UNG locus resulted in full neurovirulence, neuroinvasiveness, and the ability to reactivate in vivo. These findings suggest that the HSV-1 UNG plays an important role during acute viral replication in vivo and possibly in the reactivation process.

ICP34.5 mutants of herpes simplex virus type 1 strain 17syn+ are attenuated for neurovirulence in mice and for replication in confluent primary mouse embryo cell cultures

In a recent report, the neurovirulence of herpes simplex virus type 1 (HSV-1) was mapped to the ICP34.5 gene (J. Chou, E. R. Kern, R. J. Whitley, and B. Roizman, Science 250:1262-1266, 1990). In this report, specific mutations within ICP34.5 were constructed in HSV-1 strain 17syn+ to determine the effects of these mutations in a fully neurovirulent isolate. It was found that termination of the ICP34.5 gene after the N-terminal 30 amino acids resulted in a mutant, 17termA, which was 25- to 90-fold reduced in neurovirulence. This reduction of neurovirulence was associated with restricted replication of the mutant virus in mouse brain. The reduced replication phenotype was also evident in the trigeminal and dorsal root ganglia following inoculation at the periphery. 17termA was capable of replicating with wild-type kinetics in mouse footpads, and therefore the restriction seen in neural tissues was not due to a generalized replication defect in mouse cells. Significantly, replication of the mutant was also restricted in the mouse cornea in vivo and in confluent primary mouse embryo cells and mouse 10T1/2 cells in vitro. However, 17termA replicated with much greater efficiency in subconfluent mouse embryo cells, suggesting that the physiological state of the cell may be an important factor for productive replication of this mutant. Restoration of the ICP34.5 gene to the mutant resulted in a virus which displayed wild-type neurovirulence and replication kinetics in all cells and tissues tested.

Herpes simplex virus type 1 dUTPase mutants are attenuated for neurovirulence, neuroinvasiveness, and reactivation from latency

Herpes simplex virus type 1 (HSV-1) encodes a dUTPase which has been shown to be dispensable for normal viral replication in cultured cells (S. J. Caradonna and Y. Cheng, J. Biol. Chem. 256:9834-9837, 1981; F. B. Fisher and V. G. Preston, Virology 148:190-197, 1986). However, the importance of this enzyme in vivo has not been determined. In this report, HSV-1 strain 17 syn+ and two isogenic engineered dUTPase-negative mutants were characterized in the mouse model. Both mutants replicated with wild-type kinetics and achieved wild-type titers in cultured cells. The mutants were 10-fold less neurovirulent than 17 syn+ following intracranial inoculation and more than 1,000-fold less virulent following footpad inoculation. The dUTPase- mutants replicated with wild-type kinetics in the footpad and entered and replicated efficiently in the peripheral nervous system of the mouse. However, their replication in the central nervous system was significantly reduced. The dUTPase- strains established latent infections but displayed a greatly reduced reactivation frequency in vivo. Neurovirulence, neuroinvasiveness, and reactivation frequency were all restored by recombination with wild-type dUTPase sequences. These results have important implications with regard to anti-herpesvirus therapeutic strategies.

Mutations in the UL53 gene of HSV-1 abolish virus neurovirulence to mice by the intracerebral route of infection

The cell fusion protein, the product of the UL53 gene, is responsible for intracerebral (IC) pathogenicity of HSV-1. Recombinant HSV-1 R15 is apathogenic to mice by the IC route of inoculation, while intratypic recombinants, in which the UL53 gene in R15 was replaced by an analogous sequence from the pathogenic strain R19, regained IC pathogenicity. The nucleotide sequence of the UL53 gene of HSV-1 strains R15 (apathogenic) and R19 (pathogenic) was determined and compared to that of other pathogenic strains. Four mutations were found which are thought to be responsible for the apathogenic phenotype of HSV-1 strain R15. Northern blot hybridization of RNA extracted from BSC-1 cells infected with several HSV-1 strains indicated that all of the virus strains tested expressed equal amounts of UL53 mRNA in infected cell cultures. Demonstration of the expression of UL53 mRNA in brains of mice infected with HSV-1 strains was made possible by the combined use of a rapid method for mRNA extraction (Oligo dT-linked magnetic beads) and a highly sensitive technique for detection of the existence of the UL53-specific mRNA (cDNA synthesis followed by PCR). It was shown that both pathogenic (KOS and P42) and apathogenic (R15) HSV-1 strains expressed the UL53 gene in brains of IC infected mice.

Herpes simplex virus neurovirulence and productive infection of neural cells is associated with a function which maps between 0.82 and 0.832 map units on the HSV genome

A herpes simplex virus (HSV) intertypic recombinant (RE6) has been shown to be completely and specifically non-neurovirulent in mice. Direct intracranial inoculation of 10(8) PFU of RE6 does not result in a lethal encephalitis. Neurovirulent recombinant viruses were generated by cotransfection of RE6 DNA with DNA fragments cloned from the pathogenic HSV-1 strain 17 syn+. It was found that a 1.6-kb fragment mapping between 0.82-0.832 m.u. could restore the neurovirulent phenotype. Recombinants which incorporated at least part of this fragment were at least 100,000-fold more neurovirulent than RE6. The recombinants displayed a greatly enhanced capacity to replicate in mouse brain in vivo, but did not display enhanced replication over that of RE6 in cultured mouse cells at 38.5 degrees. Immunohistochemical analysis of infected mouse brain tissue revealed that the permissive host cell range of the recombinants was dramatically altered from that of RE6. While antigen positive cells were extremely rare in mouse brain tissue infected with RE6, the neurovirulent recombinants produced antigens in many cell types including neurons. Thus, wild-type HSV-1 sequences mapping between 0.82-0.832 m.u. can donate a highly neurovirulent phenotype to RE6.