Selective retention of herpes simplex virus-specific T cells in latently infected human trigeminal ganglia

Primary infection with herpes simplex virus 1 (HSV-1) and varicella zoster virus (VZV) results in lifelong latent infections of neurons in sensory ganglia such as the trigeminal ganglia (TG). It has been postulated that T cells retained in TG inhibit reactivation of latent virus. The acquisition of TG specimens of individuals within hours after death offered the unique opportunity to characterize the phenotype and specificity of TG-resident T cells in humans. High numbers of activated CD8(+) T cells expressing a late effector memory phenotype were found to reside in latently infected TG. The T cell infiltrate was oligoclonal, and T cells selectively clustered around HSV-1 but not VZV latently infected neurons. Neuronal damage was not observed despite granzyme B expression by the neuron-interacting CD8(+) T cells. The TG-resident T cells, mainly CD8(+) T cells, were directed against HSV-1 and not to VZV, despite neuronal expression of VZV proteins. The results implicate that herpesvirus latency in human TG is associated with a local, persistent T cell response, comprising activated late effector memory CD8(+) T cells that appear to control HSV-1 latency by noncytolytic pathways. In contrast, T cells do not seem to be directly involved in controlling VZV latency in human TG.     

Initiation and maintenance of latent herpes simplex virus infections: the paradox of perpetual immobility and continuous movement

During the acute phase of infection, herpes simplex virus (HSV) is taken up by nerve endings and travels, probably as a noninfectious nucleocapsid, toward the neurons of sensory ganglia. Infectious virus can be detected in ganglia for a limited period, after which the virus enters into a latent phase. It appears that synthesis of deoxyribonucleic acid is not required and that an early viral protein and at least one additional late virus gene product are involved in the establishment of latency. The distinction between a "static" and a "dynamic" form of latency depends on the ability to detect viral activities in neurons and on whether these observed activities are expressed continuously or intermittently. The development of recurrent lesions following virus reactivation is an occasional event and is controlled by inducing agents and the state of the organism. The maintenance of latency depends on the number of neurons that become latently infected after the primary episode, the number of neurons in which reactivation takes place, the fate of the neuron after virus reactivation, and the possibility of renewed neuronal infections after each recurrent episode. Exogenous reinfections may also contribute to the maintenance of latency since they can lead to latent infections in nearby or distantly located sensory ganglia. Multiple latent infections may result also from a single primary infection by dissemination of the virus to distantly located ganglia.     

The comparative effects of famciclovir and valacyclovir on herpes simplex virus type 1 infection, latency, and reactivation in mice.

Infections by herpes simplex virus (HSV) cannot yet be eliminated, but the severity of the disease can be reduced. Two newer drugs with established efficacy for such infections, famciclovir and valacyclovir, were tested in a mouse eye model of HSV infection. Both drugs significantly reduced mortality and titers of virus shed from the eyes of mice infected with an otherwise lethal dose of HSV type 1 (HSV-1). Similar titers of HSV-1 were found in the eyes, ganglia, and brains of treated animals. Although valacyclovir reduced the latent viral DNA load better in these studies than did famciclovir, rates of reactivation by explantation and UV exposure were the same. Thus, in this study, famciclovir and valacyclovir were equally effective in limiting the virulence and spread of HSV-1, despite their biochemical and pharmacologic differences.     

A model of lytic, latent, and reactivating varicella-zoster virus infections in isolated enteric neurons

Because human primary afferent neurons are not readily obtained, we sought to develop a model in which the lytic, latent, and reactivating phases of varicella-zoster virus (VZV) infection were recapitulated in neurons from an animal source. Enteric neurons were obtained from the small intestine of adult guinea pigs and from the bowel of fetal mice. Latency was established when these neurons were infected by cell-free VZV in the absence of fibroblasts or other cells of mesodermal origin. In contrast, lytic infection ensued when fibroblasts were present or when the enteric neurons were infected by cell-associated VZV. Latency was associated with the expression of a limited subset of viral genes, the products of which were restricted to the cytoplasm. Lysis was associated with the expression of viral glycoproteins, nuclear translocation of latency-associated gene products, and rapid cell death. Reactivation was accomplished by expressing VZV open reading frame (ORF) 61p or herpes simplex virus ICP0 in latently infected neurons. Isolated enteric neurons from guinea pigs and mice recapitulate latent gene expression in human cranial nerve and dorsal root ganglia. Expression of latency-associated VZV gene products was detected in 88% of samples of adult human intestine, suggesting that VZV not only infects enteric neurons but also is latent in the human enteric nervous system. This in vitro model should facilitate further understanding of latency and reactivation of VZV.     

Herpes simplex type 1 DNA in human brain tissue.

Herpes simplex virus type 1 (HSV-1) is known to reside latently in the trigeminal ganglia of man. Reactivation of this virus causes skin lesions and may occasionally infect other tissues, including the brain. To determine whether the brain tissue of humans free of clinical signs of HSV-1 infection contains any trace of HSV-1, we examined the DNA from brain tissue by endonuclease digestion, separation of the fragments by gel electrophoresis, and hybridization with labeled HSV-1 DNA probes. Hybrid bands were detected autoradiographically in experiments using cloned and virion-purified fragments of the HSV-1 genome. HSV-1 DNA sequences were found in 6 of 11 human brain DNA samples tested. In some cases, these bands corresponded to the bands expected for the complete viral genome, whereas others contained bands representing only a part of the genome. In some cases, the terminal fragments could be found, suggesting that the DNA was in a linear, nonintegrated form.     

Reactivated and latent varicella-zoster virus in human dorsal root ganglia.

Ganglia obtained at autopsy were examined by in situ hybridization from one patient with zoster (also called herpes zoster or shingles), two varicella-zoster virus (VZV)-seropositive patients with clinical evidence of zoster, one VZV-seronegative child, and one fetus. Ganglia positive for VZV had a hybridization signal in both neuronal and nonneuronal satellite cells. Ganglia obtained from the fetus and from the seronegative infant were consistently negative for VZV. Two striking observations were evident regarding the presence of VZV DNA in ganglia obtained from the individual with zoster at the time of death. First, ganglia innervating the sites of reactivation and ganglia innervating adjacent sites yielded strongly positive signals in neurons and satellite cells, whereas ganglia from distant sites were rarely positive. Second, VZV DNA was found in both the nuclei and the cytoplasm of neurons innervating areas of zoster. However, in neurons innervating zoster-free areas, VZV DNA was found only in the nucleus of neurons and their supporting satellite cells. Immunohistochemistry with a fluorescent monoclonal antibody to the VZV glycoprotein gpI, a late virus protein, revealed a positive signal in the cytoplasm of ganglia with clinical evidence of reactivation. These results illustrate that both neuronal and satellite cells become latently infected following primary VZV infection. The presence of VZV DNA and gpI in the cytoplasm of neurons demonstrates productive infection following reactivation at the site of latency.     

miR-H28 and miR-H29 expressed late in productive infection are exported and restrict HSV-1 replication and spread in recipient cells

We report on the properties and function of two herpes simplex virus-1 (HSV-1) microRNAs (miRNAs) designated “miR-H28” and “miR-H29.” Both miRNAs accumulate late in productive infection at a time when, for the most part, viral DNA and proteins have been made. Ectopic expression of miRNA mimics in human cells before infection reduced the accumulation of viral mRNAs and proteins, reduced plaque sizes, and at vey low multiplicities of infection reduced viral yields. The specificity of the miRNA mimics was tested in two ways. First, ectopic expression of mimics carrying mutations in the seed sequence was ineffective. Second, in similar tests two viral miRNAs made early in productive infection also had no effect. Both miR-H28 and miR-H29 are exported from infected cells in exosomes. A noteworthy finding is that both miR-H28 and miR-H29 were absent from murine ganglia harboring latent virus but accumulated in ganglia in which the virus was induced to reactivate. The significance of these findings rests on the principle that the transmission of HSV from person to person is by physical contact between the infected tissues of the donor and those of uninfected recipient. Diminished size of primary or recurrent lesions could be predicted to enhance person-to-person transmission. Reduction in the amount of reactivating latent virus would reduce the risk of retrograde transport to the CNS but would not interfere with anterograde transport to a site at or near the site of initial infection.Herpes simplex viruses (HSVs) infect and multiply in cells at the portal of entry, i.e., mouth or genitals (1). They then are exported by retrograde transport to sensory or autonomic neurons where they establish latent (silent) infection. In response to neuronal stress, the virus reactivates and is transported anterograde to a site at or near the site of initial infection. At that site the virus is available for transmission by physical contact between the infected tissues of the donor and those of the uninfected recipient (1, 2). In principle it would be expected that extensive lesions on initial infection and on reactivation would seed more neurons with latent virus and increase the transmission of virus from person to person. This report challenges the hypothesis that viral gene products uniformly enhance viral replication and spread.This report centers on the function of two newly detected viral microRNAs (miRNAs). Specifically HSV-1 and HSV-2 have been reported to express at least 27 miRNAs (3–13). Analyses of 17 of these miRNAs have shown that they differ in their requirements for synthesis in productively infected cells. Thus, some are made in higher amounts in the absence of viral protein synthesis, but others require viral protein synthesis for their synthesis (14). A confounding problem associated with the function of viral miRNAs is that miRNAs made in the course of latent infections in neurons are also made in productively infected cells, whereas some, but not all, miRNAs accumulating in neurons in which latent virus is in the process of reactivation are not made in appreciable amounts in productively infected cells (14). Here we report the properties of two newly identified miRNAs designated “miR-H28” and “miR-H29.” The key properties of these miRNAs are (i) they are made late in infection, i.e., after viral DNA and structural proteins have been made; (ii) they are exported in exosomes; and (iii) they accumulate in neurons in which the virus is in the process of reactivation from the latent state but are absent from neurons harboring latent virus. Additional studies showed that ectopic expression of mimics of these miRNAs in cells before infection reduces the rate of accumulation of viral proteins, decreases plaque size, and at the same time reduces viral yields in cells infected at a very low multiplicity of infection.One hypothesis that could explain the evolution of these miRNAs is the need to block excessive replication of reactivating virus for two reasons. First, repulsive lesions on the mouth or genitals would reduce physical contact between the infected tissues of the transmitter and the uninfected tissue of the recipient. A second reason may be to reduce the amount of virus made on reactivation to ensure that the reactivated virus is transmitted anterograde to the mouth or genitals rather than retrograde to the CNS. Retrograde transport would likely kill the host and block further viral spread. In essence these studies support the conclusion that HSV-1 down-regulates its replication to enhance its spread and that this process is executed by viral gene products.     

Genital herpes in guinea pigs: pathogenesis of the primary infection and description of recurrent disease

Guinea pigs inoculated intravaginally with herpes simplex virus type 2 (HSV-2) developed a self-limiting infection characterized by vesiculo-ulcerative lesions on the external genital skin, urinary retention, and hindlimb paralysis. Infection rarely resulted in death. Virologic, histologic, and immunoperoxidase data suggested the following scheme for viral pathogenesis: initial replication in the introitus, vagina, and bladder; spread via sensory nerves to the lumbosacral dorsal root ganglia and spinal cord, and transmission via peripheral nerves to the external genital skin to produce the characteristic lesions. After recovery from primary infection, animals developed recurrent vesicular lesions, shed virus from genital sites in the absence of lesions, and harbored latent HSV-2 in dorsal root ganglia. Genital infection in the guinea pig shares many features with genital herpes in humans and provides a model to explore mechanisms of latency and reactivation and to evaluate several methods for control of recurrent disease.