Type-specific identification of herpes simplex and varicella-zoster virus antigen in autopsy tissues

Herpes simplex virus type 1 encephalitis in acquired immunodeficiency syndrome

Herpes simplex (HSV) infection of the central nervous system is uncommon in AIDS and usually has an atypical topography. This review is centred around the case of a 49-year-old homosexual patient with AIDS who died from diffuse encephalopathy. Neuropathological examination revealed necrotic and haemorrhagic changes involving both temporal lobes, insulae and cingulate gyri. Cowdry type A intranuclear inclusion bodies were abundant but inflammation was minimal. Electron microscopy confirmed characteristic herpes virus particles. Immunocyto-chemistry was positive for HSV type 1 and 2. In situ hybridization and PCR, however, were positive for HSV type 1 but excluded HSV type 2. There was associated cytomegalovirus ventriculitis but clearly separated from HSV encephalitis. There were no histological features of HIV encephalitis and HIV could not be demonstrated by immunocytochemistry or by PCR to demonstrate proviral DNA. Apoptotic neurons were numerous in areas with a severe macrophage reaction. Only two pathological cases with characteristic limbic distribution and necrotic haemorrhagic histologic have been reported previously. The rarity of these reports suggests that in advanced AIDS, the immune reaction causing a typical necrotizing encephalitis cannot be mounted. Distinction between HSV type 1 and 2 infection may be difficult by immunocytochemistry and usually requires in situ hybridization, tissue culture or PCR. In AIDS patients, HSV-1 has been identified as responsible for encephalitis whereas HSV-2 has been more responsible for myelitis. Associated productive HIV infection of the CNS was found in none of the cases. In contrast, cytomegalovirus encephalitis was found in nine of 11 cases of AIDS-associated HSV encephalitis.

Distribution of varicella zoster virus and herpes simplex virus in disseminated fatal infections

AIMS

To study the cutaneous and visceral distribution of herpes simplex virus (HSV) and varicella zoster virus (VZV) in fatal infections.

METHODS

Standard histology, immunohistochemistry (monoclonal antibodies VL8 and VL2 and polyclonal antibody IE63 directed against VZV; monoclonal antibodies IBD4 and HH2 and polyclonal antibodies directed against HSVI and HSVII) and in situ hybridisation (anti-HSV and anti-VZV probes) were applied to formalin fixed, paraffin wax sections.

RESULTS

On histological examination, Herpesviridae infection was evident in various organs including the lungs, liver and skin. In addition, immunohistochemistry and in situ hybridisation revealed the presence of HSV and VZV antigens and nucleic acids in several cell types and tissues showing no cytopathological alterations suggestive of Herpesviridae infection. The organs with histological evidence of infection also contained VZV or HSV antigens and their genes.

CONCLUSIONS

These findings suggest that organ failure in disseminated VZV and HSV infections is primarily caused by HSV or VZV induced cell damage and lysis. They also indicate that immunohistochemistry and in situ hybridisation can provide an accurate, type-specific diagnosis on formalin fixed, paraffin wax embedded tissue even when classic histological and cytological characteristics are lacking.

Distribution of varicella-zoster virus gpI and gpII and corresponding genome sequences in the skin

In the course of varicella-zoster virus (VZV) infection, some viral capsid antigens are found in the epidermis and dermis. The aim of this study was to investigate the localisation of two major VZV glycoproteins (gpI and gpII) and of their respective genes in the skin. The distribution of VZV gpI and II in 27 formalin fixed paraffin embedded skin biopsies from herpes zoster eruptions were compared by immunohistochemistry. Double immunostaining was carried our to identify infected cells. The presence of viral nucleic acids coding for gpI and gpII was examined by in situ hybridisation. The distribution of gpI and gpII and their corresponding genome sequences was similar in the epidermis. gpI and gpII were also detected in dermal FXIIIa positive dendrocytes, in Mac 387 and CD68 positive macrophages, and in perineural and endothelial cells. However, the corresponding viral nucleic acids were rarely and barely detected in these cells of the dermis. It is concluded that VZV infection of epithelial cells follows a different course than in dermal cells.

Autopsy findings in two cases of neonatal herpes simplex virus infection: detection of virus by immunohistochemistry, in situ hybridization and the polymerase chain reaction

We describe the pathological findings in two fatal cases of neonatal infection with herpes simplex virus. One had an encephalitis caused by herpes simplex virus type 2 (HSV-2); the other had a disseminated infection with herpes simplex virus type 1 (HSV-1). Confirmation of the diagnosis was obtained by use of the polymerase chain reaction to amplify viral DNA from paraffin sections of autopsy tissues. By using primers which amplify fragments of the HSV-1 thymidine kinase gene and HSV-2 glycoprotein gene respectively it was possible to discriminate between infection with HSV-1 and HSV-2. In contrast, immunohistochemistry and in situ hybridization using commercially available reagents did not distinguish between HSV-1 and HSV-2 infection. However, immunohistochemistry and in situ hybridization are probably more reliable than the polymerase chain reaction for assessment of the distribution of virus in different tissues.

In situ polymerase chain reaction: localization of HSV-2 DNA sequences in infections of the nervous system

To detect and localize a herpes simplex virus type 2 (HSV-2) thymidine kinase gene sequence in paraffin sections of brains and trigeminal ganglia of infected mice, an in situ polymerase chain reaction (ISPCR) protocol was developed. Using a single pair of primers, a 110 base pair DNA target sequence, and incorporation of a digoxigenin-labelled nucleotide during amplification, this procedure permitted rapid, specific, reproducible detection of infected cells. During acute brain infection, cells labelled by ISPCR were in the same infected foci that, in adjacent sections, contained viral antigen. This, together with controls, gave evidence of method specificity. In mice surviving acute infection, latently infected cells were labelled by ISPCR. In brains, focal areas contained labelled cell nuclei, and in trigeminal ganglia, neuronal nuclei were likewise labelled. Latent infection was confirmed by several methods, including identification of an HSV-specific sequence in DNA extracts of brains and ganglia, virus isolation from explanted ganglia, and HSV-2 latency-associated transcript (LAT) RNA localization in ganglionic neurons by in situ hybridization. Evidence in brains of ISPCR-labelled cells in regions where HSV-2 LAT-positive cells were not detected, and in ganglia of more ISPCR-labelled neurons than were LAT-positive, indicated that ISPCR is more sensitive in detecting latently infected cells than previous methods.

Detection of viral DNA in neonatal herpes encephalitis autopsy tissues by solution-phase PCR: comparison with pathology and immunohistochemistry

To detect DNA sequences of herpes simplex virus (HSV) in neural and non-neural tissue sections in disseminated human neonatal HSV infection, a solution polymerase chain reaction (PCR) protocol was developed which amplified HSV thymidine kinase and host genomic DNA sequences that were hybridized with sequence-specific probes in Southern blots. Serial sections of formalin-fixed, paraffin embedded autopsy tissues were tested by PCR and compared to histology and HSV antigen detection. The sensitivity, specificity and reproducibility of this PCR protocol were determined on uninfected and HSV-infected mouse tissues and on HSV DNA from infected tissue culture cells. Samples estimated to contain as few as 60 copies of preserved HSV DNA target sequence gave a positive PCR result. In nine neonates that died during acute HSV infection, all non-neural tissues and a minority of neural tissues with histological lesions had HSV antigen; when DNA could be amplified, HSV DNA sequences were detected by PCR. Together, these findings indicate a direct role for virus in the pathogenesis of these lesions. In the same cases, some or all brain samples were negative for HSV antigen, but nevertheless had HSV DNA sequences detected by PCR. The possible explanations for this finding are discussed. In one neonate dying seven weeks after birth, HSV sequences were found in brain lesions in the absence of HSV antigen; neither HSV DNA nor antigen were found in non-neural tissues, suggesting a latent HSV infection in brain. It is practical to apply PCR methods to detect minute quantities of viral DNA in formalin-fixed, paraffin embedded autopsy tissues.(ABSTRACT TRUNCATED AT 250 WORDS)

Comparative immunohistochemical study of herpes simplex and varicella-zoster infections

Herpes simplex (HSV) and varicella-zoster (VZV) skin infections share so many histological similarities that distinguishing between them may prove to be impossible. We developed and characterized a new monoclonal antibody, VL8, IgG kappa isotype, directed to the VZV envelope glycoprotein gpI. Immunohistochemistry with VL8 appeared highly sensitive and specific on formalin-fixed paraffin-embedded biopsies and a clear-cut distinction between HSV and VZV infections was possible. The pattern of VL8 immunolabelling in VZV infections was strikingly different from that found in HSV infections studied with polyclonal antibodies to HSV I and II. Double immunolabelling revealed the VL8 positivity of sebaceous cells, endothelial cells, Mac 387- and CD68-positive monocyte-macrophages, and factor XIIIa-positive perivascular, perineural and interstitial dendrocytes. Intracytoplasmic VL8 labelling of endothelial cells and perivascular dendrocytes was found at the site of leukocytoclastic vasculitis.

Fatal varicella-zoster virus meningoradiculitis without skin involvement

A 77-year-old man with T-cell lymphoma developed an acute fatal meningoradiculitis of cranial nerve roots and cauda equina, pathologically and virologically confirmed to be caused by varicella-zoster virus. This is the first report of fatal varicella-zoster virus-induced neurological disease in the absence of skin lesions. Varicella-zoster virus should be included in the differential diagnosis of acute radiculoneuropathy in the immunocompromised patient, particularly because antiviral treatment for varicella-zoster virus exists.