key: cord-0010055-gtersx2l
authors: Lascano, Eduardo F.; Berría, María I.; Avila, María M.; Weissenbacher, Mercedes C.
title: Astrocytic reaction predominance in chronic encephalitis of junin virus‐infected rats
date: 2005-12-09
journal: J Med Virol
DOI: 10.1002/jmv.1890290420
sha: f88dc2955b85dd3203612a6ecae7715257e22e86
doc_id: 10055
cord_uid: gtersx2l

Junin virus antigen distribution and astrocytic reaction to prolonged infection were characterized in rat brain by the PAP technique. During the acute stage of neurologic disease following intracerebral inoculation, Junin antigen was detected in 100% of animals, strongly in most neurons but also to a much lesser degree in scattered astrocytes, dropping to 20% of rats at 540 days postinfection. Initially labeled in all brain areas, viral antigen gradually disappeared from hippocampus but persisted irregularly in cerebral cortex, basal ganglia, Purkinje cells, pons, and medulla oblongata. Such a pattern suggests that specific neuronal subpopulations, in spite of apparently unaltered cell morphology, may persistently harbor the virus, leading on occasion to a delayed neurologic syndrome. During both the acute and chronic stages of disease, a mild inflammatory exudate was observed, characterized by the presence of T and B lymphocytes, as well as macrophages and unidentified round cells. GFAP immunostaining showed increased astrocytic reaction as infection lapsed into chronicity. Corpus callosum, hippocampus, and cerebellum exhibited the sharpest reactive astrocytosis, followed by basal ganglia, pons, and medulla oblongata, whereas in cerebral cortex it was considerably less. Astrocyte activation, which failed to correlate with viral antigen presence in neurons, seems to result from a generalized condition, possibly including diffusible brain factors triggered by viral infection. Such widespread astroglial reaction may thus contribute to the outcome of the late neurologic syndrome.

models. As previously reported [Weissenbacher et al., 19861 , following intracerebral inoculation of newborn rats with XJ strain three different stages could be discerned: an early phase of acute disease, up to 30 days postinfection (pi), with 5% mortality; a n intermediate one, extending to 280 days pi, without clinical signs but with evident viral persistence; and a final period of chronic illness, featuring a clinical neurologic syndrome.

However, a more expansive histopathologic study of samples corresponding to these sequential stages was lacking. Therefore, on the basis of brain tissue findings, the purpose of the present study was to characterize the topographical distribution of viral antigen and the degree of astrocytic reaction to prolonged infection. Since chronic neurologic disease in rats seems associated with virus persistence and long-lasting immunologic stimulation [Weissenbacher et al., 19861, factors involved in faulty clearance of brain virus could explain the mechanism through which the viral infection may ultimately damage the CNS.

Extrapolation of data from a useful model of Junin virus chronic neurologic disease, such as the rat has proved to be, may help to shed some light on the physiopathogenesis of the late neurologic syndrome observed a t times in Argentine hemorrhagic fever patients [Maiztegui et al., 1979; Melcon and Herkovits, 19811 .

The XJ prototype strain of Junin virus was used.

Junin virus, a member of the Arenaviridae family and etiologic agent of Argentine hemorrhagic fever, is known to induce neurologic disease in some animal 

Techniques employed to identify brain structures included hematoxylin-eosin (HE) for routine examination and special staining methods such as Luxol fast blue-Holmes silver nitrate for myelin and neurofibrils, the Nissl procedure for neurons, and a n expressly developed silver impregnation method for neurons and glial cells. Solutions for the silver impregnation method comprised 1) 2% aqueous potassium bichromate; 2) 1% aqueous silver nitrate plus 0.4% aqueous potassium carbonate, and the resulting precipitate just dissolved by dropwise addition of ammonium hydroxide; and 3) 0.25% ( w h ) gelatin dispersed in warm distilled water followed by addition of 10% paraformaldehyde. The steps for staining were as follows: 1) Samples were deparafflnized, hydrated, and sensitized with potassium bichromate solution for 2 min; 2) preparations were washed several times in distilled water; 3) one drop of pure pyridine was add-ed for each milliliter of filtered silver solution and then used to impregnate samples at 60°C for 15 min; and 4) Formalin-gelatin mixture was employed to reduce silver, the reaction left to rest for 10 min. If the staining was weak, as usually occurs, steps 2-4 were repeated as required until intensity and sharpness were achieved.

The PAP method was employed to label Junin virus antigen and glial fibrillary acidic protein (GFAP) as described elsewhere [Lascano and Berria, 19831 and to stain T lymphocytes, B lymphocytes, and macrophages as previously reported [Lascano et al., 19861 .

Rat brains harvested up to 30 days pi were considered representative of the acute stage of neurologic disease, while the remainder were regarded as depicting chronic infection. Table I summarizes findings from virus antigen-PAP immunostaining, gradually dropping to 2/10 at phages, together with a few B lymphocytes and a variable number of unidentified round cells. This immunoperoxidase profile persisted until the very end of the observation period. Table I1 outlines the results of GFAP immunostaining and also staining by the PAP technique. Degrees of astrocytic reaction degree, which increased wth the successive stages of the disease, are depicted. Specifically, the cell activation, representing faster morphologic and immunocytochemical maturation versus control animals, featured hyperplasia and hypertrophy of cell bodies and processes. As from the onset, cerebellum exhibited sharp GFAP staining of astrocytes and of Bergmann fibers (Fig. 4a) . Although the glial reaction was less evident in basal ganglia, isolated cell areas of strong labeling were observed (Fig. 4b) . On occasion, foci of bizarre astrocytes were seen throughout the brain, including the thalamus (Fig. 4c) . Toward the end of the observation period, Bergmann fibers gradually stained even more heavily for GFAP (Fig. 4d) . Hyperplasia and hypertrophy were often depicted by a n increase in number and size of vascular end-feet, making up a sharp continuous border. Furthermore, a few degenerative foci appeared as from 30 days pi, represented by amorphous GFAP-positive masses or granules scattered within brain tissues. Table I1 also shows that the corpus callosum, hippocampus, and cerebellum had the sharpest astrocytic activation, followed by basal ganglia, pons, and medulla oblongata, whereas in cerebral cortex glial reaction was considerably weaker. slight staining of axons surrounding neurons and glial cells; 540 days pi. X 500. c: Extending along corpus callosum fibers, rows of oligodendrocytes; staining allows a few processes to be visualized; 540 days pi. x300. Fig. 3 . Luxol fast blue-Holmes method. Single row of Purkinje cells between molecular and granular layers; a t bottom left, axons with spared myelin sheaths; 540 days pi. x 400.

As shown by general and specialized staining methods, the X J strain of Junin virus induced mild histological changes in rat brain since the inflammatory reaction was moderate and, except astrocytic activation, no apparent neuronal, oligodendroglial, or myelin alterations were observed. The fact that during the chronic period infectious virus could only be rescued by coculture [Weissenbacher et al., 19861 indicates that Junin virus may persist in brain tissues, either causing a low-grade productive infection or remaining in an abortive state during which synthesis or assembly of viral components is incomplete.

In common with other CNS viral infections I Bernard et al., 1983; de la Monte et al., 1983; Fekadu et al., 1982; Kristensson et al., 1983; Sethi and Lipton, 19851 , Junin virus antigen was identified throughout the brain tissues during 3-4 weeks pi. The greatest number of Junin-labeled neurons was observed in cerebral cortex whereas within the cerebellum, selective expression of viral antigen was exhibited by Purkinje cells. At later times pi, immunostaining became restricted to certain areas such as cerebral cortex, basal ganglia, cerebellum, pons, and medulla oblongata, but disappeared from hippocampus as from 82 days pi. The finding that persistent infection was limited to specific neuronal subpopulations strongly suggests that certain cell types may harbor the virus during the prolonged asymptomatic stage, thereby providing a reservoir that at times triggers a delayed neurologic syndrome.

Although no morphological changes could be seen in infected neurons, it is quite possible that signs of disease may develop because of dysfunction in such highly specialized cells. The selective vulnerability of neuronal subpopulations t o viral infections has been reported for many RNA viruses [Kristensson and Norrby, 19861 , and, by comparing immunocytochemical findings to date, Purkinje cells stand out as a peculiar reservoir for persistent infection by virus such as JHM [Sorensen and Dales, 19851, human coronavirus [Pearson and Mims, 19831 , and reovirus type 3 [Spriggs et al., 19831 . It should be emphasized that the longlasting detection of Junin virus antigen observed in cerebral cortex neurons and in Purkinje cells has also been described for lymphocytic choriomeningitis virus [Rodriguez et al., 19831 , prototype of the Arenaviridae family to which Junin virus belongs. In our rat model, viral antigen was located mainly in neurons and, to a lesser extent, in astrocytes, closely paralleling a feature previously described for the mouse [Lascano and Berria, 19831 . Actually, neurons and astrocytes are precisely the two CNS cell types in which budding of and Berria, 19741 . Another pattern common to both animal models is that Junin antigen, though distributed all over the cytoplasm, was mostly concentrated at the cell periphery, where viral particle budding invariably takes place. Inflammatory infiltrate was mild throughout, in sharp contrast with the plentiful exudate observed in rat for the XJ-C1, strain [Lascano et al., 19861 . As regards its immunocytochemical profile, it agreed with the unique CNS features delineated by Griffin et al. [ 19871 in other experimental viral encephalitis diseases such as those induced by Sindbis and lymphocytic choriomeningitis viruses.

It is known that the astrocytic response to CNS injury is stereotyped, unidirectional, and characterized by proliferation [Latov et al., 19791, hypertrophy, and appearance of large quantitites of GFAP [Amaducci et al., 1981; Bignami and Dahl, 19761 . As a general rule, the intensity of the glial reaction is directly related to the proximity of the injury site [Eng and DeArmond, 1982J . However, in the present case, a comparison of the data given in Tables I and I1 indicates that there is no topographical relationship between viral antigen location in neurons and astrocytic activation detected by GFAP labeling. In fact, the hippocampus was Junin antigen free from 82 days pi, but this was the very area where astrocytic hyperplasia and hypertrophy reached a considerable degree. Conversely, during persistent infection, viral antigen was frequently detected in cerebral cortex neurons, where astrocytic activation was manifested to a lesser extent. Furthermore, at 280 and 450 days pi, there was a single case out of 10 animals that exhibited viral antigen in brain, although all rats presented a similar degree of astrocytic reaction. Resorting to in vitro systems [Berrla and Lascano, 19851, a direct Junin virus effect has been demonstrated on rat astroglial cell cultures, consisting of the induction of morphological and immunocytochemical differentiation. However, in the present model, the number of Junin antigen-labeled astrocytes in the whole animal was considerably lower that the number of activated astrocytes. Moreover, the gradual viral antigen restriction in neurons during the chronic stage was concomitant with widespread astrocytic reaction. Thus, the presence of viral antigen in both astrocytes and neurons seems unrelated to reactive astrocytosis.

Since control rats failed to show such changes, the recognized influence of age on GFAP labeling [Bignami and Dahl, 1974; Ludwin et al., 19761 had to be ruled out. Although complete information is still lacking about the triggering of glial reaction [Reier, 19861, several substances have been identified that may induce astrocyte proliferation or differentiation [Lindsay, 19861 , such as glia maturation factors present in brain tissues [Kato et al., 1981; Nieto-Sampedro et al., 1985; Pettmann et al., 1980; Sobue and Pleasure, 19841 and possibly released from damaged neurons, as well as from demyelinated areas [Kristensson et al., 19861 . Glial reaction has also been attributed to lymphokines produced by viral antigen-stimulated perivascular lymphocytes [Oldstone et al., 19861 or to direct effects caused by the infective agent itself [Manuelidis et al., 19871. Recent evidence has been presented that viral infection can impair neuronal activity even in the absence of histologic injury ILipkin et al., 19881. Although consistent morphologic changes were lacking in Junin antigen-labeled neurons, it is probable that cell infection can induce the release of mitogenic and morphogenic factors. Understandably, since the astrocyte is the most abundant neuroectodermal cell of the brain, its reactivity seems the most obvious histologic expression of Junin virus persistent infection. It may therefore be concluded that the indiscriminate and detrimental astrocytic activation is a contributory factor to the pathogenesis of the delayed neurologic disease observed in the adult rat. The present findings encourage further studies that include the identification of the astrocyte subtypes [Miller and Raff, 1984; Takamiya et al., 19881 that are involved in such glial reaction. It seems reasonable to assume the existence of regional differences in the astrocytic response to persistent viral infection of the CNS.

Glial fibrillary acidic protein in cryogenic lesions of the rat brain

Canine distemper infection in mice: Characterization of a neuro-adaoted virus strain and its long-term evolution in the mouse

Astrocyte differentiation induced by Junin virus in rat brain cell cultures

Astrocyte-protein and neuroglial differentiation. An immunofluorescent study with antibodies to the glial fibrillary acidic protein

The astroglial response to stabbing. Immunofluorescence studies with antibodies to astrocyte-specific protein (GFA) in mammalian and submammalian vertebrates

Immunoperoxidase detection of yellow fever after natural and experimental infections

Immunocytochemical studies of astrocytes in normal development and disease

Immunofluorescence, histologic and ultrastructural studies of the central nervous system

Immune responses in the central nervous system

Glia maturation factor in bovine brain: Partial purification and physicochemical characterization. Brain Research

Increased levels of myelin basic protein transcript gene in virus-induced demyelination

Persistence of RNA viruses in the central nervous system

Sandai virus infection in the brains of mice: Distribution of viral antigens studied with monoclonal antibodies

Ultrastructure of Junin virus in mouse whole brain and in mouse brain tissue cultures

Immunoperoxidase study of astrocytic reaction to Junin virus encephalomyelitis of mice

Brain inflammatory exudate in Junin virus-infected rats: Its characterization by the immunoperoxidase (PAP) technique

Fibrillary astrocytes proliferate in response to brain injury. A study combining immunoperoxidase technique for glial fibrillary acidic protein and radioautography of tritiated thymidine

Reactive gliosis

Viral infection of neurons can depress neurotransmitter mRNA levels without histologic injury

The topographical distribution of S-100 and GFA proteins in the adult rat brain. An immunohistochemical study using horseradish peroxidase-labelled antibodies

Efficacy of immune plasma in treatment of Argentine hemorrhagic fever and association between treatment and a late neurologic syndrome

Astrocyte gene expression in CreutzfeldtJakob disease

Complicaciones neurologicas tardias de la Fiebre Hemorragica Argentina

Fibrous and protoplasmic astrocytes are biochemically and developmentally distinct

The 333 control of glial populations in brain: Changes in astrocytic mitogenic and morphogenic factors in response to injury

Cytoimmunotherapy for persistent virus infection reveals a unique clearance pattern from the central nervous system

Selective vulnerability of neural cells and age-related susceptibility to Oc43 virus in mice

Isolation of a glial maturation factor from beef brain

Gliosis following CNS injury: The anatomy of astrocytic scars and their influences on axonal elongation

Ultrastructural localization of viral antigens in the CNS of mice persistently infected with lymphocytic choriomeningitis virus (LCM)

Location and distribution of virus antigen in the central nervous system of mice persistently infected with Theiler's virus

Astroglial proliferation and phenotype are modulated by neuronal plasma membrane

In vivo and in vitro models of demyelinating disease: JHM virus in the rat central nervous system localized by in situ cDNA hibridization and immunofluorescent microscopy

Hemagglutinin variants of reovirus type 3 have altered central nervous system tropism

Immunohistochemical studies on the proliferation of reactive astrocytes and the expression of cytoskeletal proteins following brain injury in rats

Chronic neurologic disease in Junin virus-infected rats

We thank Dr. Lawrence F. Eng, Pathology Research, Veterans Administration Medical Center, Palo Alto, CA, for the gift of anti-GFAP antiserum. This work was partially supported by grants from CONICET and SECYT, Argentina.