Disruption of astroglial interlaminar processes in Alzheimer's disease.

A palisade of long, interlaminar astroglial processes in supragranular layers of the cerebral cortex is characteristic of adult individuals of anthropoid species. In the present study, this distinctive cytoarchitectonic feature was analyzed in tissue deriving from the neocortex of cases affected by Alzheimer's disease (n=14) and age-matched control cases (n=10). Samples of different cortical areas, and in particular prefrontal, temporal and striate fields, were analyzed. Astroglia was labeled by glial fibrillary acidic protein immunoreactivity, that allowed a clear distinction between the classical, stellate intralaminar astroglia and the interlaminar glial processes. The occurrence and relative density of neuritic plaques were ascertained in the same specimens with Bielchowsky staining. In most cortical regions of cases diagnosed as severe Alzheimer's disease by the donor institutions, interlaminar astroglia was found to be markedly altered or absent, and replaced by hypertrophic intralaminar astrocytes. Cases diagnosed as milder or uncertain Alzheimer's disease showed a less consistent involvement of the interlaminar glial palisade. Alterations of the interlaminar palisade in the cortex affected by Alzheimer's disease did not strictly correlate with the density of neuritic plaques in the examined specimens. The findings indicate that loss/severe disruption of the interlaminar palisade of astroglial processes is part of the array of neuropathological changes occurring in the cerebral cortex during Alzheimer's disease. In addition, our data indicate that different types of neocortical astrocytes (namely intralaminar and interlaminar astrocytes) respond differently to the pathobiology of Alzheimer's disease in the neocortex, inasmuch as interlaminar processes tend to disappear while intralaminar processes become reactive.

Tissue printing of astroglial interlaminar processes from human and non-human primate cerebral cortex.

Astroglial interlaminar processes are unique features of the cerebral cortex of adult primates, including man. The functional role of these processes in the primate cerebral cortex is largely unknown. The development and standardization of procedures that could maximize the utilization of primate brain samples is required for the experimental analysis of the individual and collective dynamic properties of interlaminar glial processes. With this aim and in order to assess the relative stability of these glial processes in ex vivo conditions, "tissue printing" procedures were applied. "Tissue printing" allows for the acute transfer of cellular elements from fresh tissue onto an artificial substrate. Human, monkey (Cebus apella), and rat brain samples were subjected to "tissue printing" procedures followed by cell culture and immunohistochemistry. For the purpose of comparing the efficiency of this procedure on the transfer of other long glial processes, "tissue prints" of radial glial processes from neonatal rats and of Bergmann glia from cerebellar samples of adult rats were included. Nitrocellulose (with and without added fibronectin or laminin) produced the best attachment results. Interlaminar processes were not modified following 24-h incubation in a cell culture medium, with the addition of agents known to modify astroglial morphotypes in vitro (cyclic adenosine monophosphate, 40 mM K(+), or fetal calf serum). It is concluded that glia with interlaminar processes can be detached from fresh tissue using "tissue printing" procedures, can be maintained for at least 24 h in standard culture conditions, and showed a stable morphological phenotype.

Dissociated functional recovery in parkinsonian monkeys following transplantation of astroglial cells.

Bilateral astroglial transplantation into the neostriatum of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-treated monkeys resulted in significant performance improvement in a spatial delayed response task, but failed to modify perseveration in an object retrieval detour task, or to improve motor clinical rating. Results suggest that brain circuits subserving various motor and cognitive performances can be functionally dissociated, and that remaining resources for the reorganization of neural circuits involved in spatial working memory performance in parkinsonian monkeys, appear to be responsive to striatal transplantation of subcultured, fetal striatal astroglial cells.

Differential effect of oestradiol and astroglia-conditioned media on the growth of hypothalamic neurons from male and female rat brains.

To determine whether soluble products from different CNS regions differ in their ability to support oestrogen-stimulated neurite growth, hypothalamic neurons from sexually segregated embryos were cultured with astroglia-conditioned medium (CM) derived from cortex, striatum and mesencephalon, with or without 17-beta-oestradiol 100 nM added to the medium. After 48 h in vitro, neurite outgrowth was quantified by morphometric analysis. Astroglia-CM from mesencephalon (a target for the axons of hypothalamic neurons) induced the greatest axogenic response in males and in this case only a neuritogenic effect could be demonstrated for oestradiol. On the other hand, astroglia-CM from regions that do not receive projections from ventromedial hypothalamus inhibited axon growth. A sexual difference in the response of hypothalamic neurons to astroglia-CM and oestradiol was found; growth of neurons from female foetuses was increased by astroglia-CM from mesencephalon, but no neuritogenic effect could be demonstrated for oestradiol in these cultures. Blot immunobinding demonstrated the presence of receptors for neurotrophic factors in cultures of hypothalamic neurons; Western blot analysis of these cultures demonstrated that oestradiol increased the concentration of trkB and IGF-I Rbeta, whereas trkA was not detected and the concentration of trkC was not modified. These results support the hypothesis that target regions produce some factor(s) that stimulate the growth of axons from projecting neurons and further indicate that in the case of males this effect is modulated by oestradiol, perhaps mediated through the upregulation of trkB and IGF-I receptors.

Heterochronous maturation of regional brain astroglia: neuronal modulation of striatal glial cells differentiation ex vivo.

Subcultured astroglial cells from striatum, cerebral cortex and ventral mesencephalon obtained from primary cultures of fetal (E14, E17 and E21) or postnatal (days 5-6) rats showed different regional, age-dependent morphological response (stellation) to cyclic AMP. While most of the cerebral cortex and ventral mesencephalic astroglial cell population was responsive at all ages tested, striatal cells at E14 and E17 were not. At age E21 striatal astroglia showed a significant shift toward a mature-like type of response to cyclic AMP. Postnatal striatal astroglia responded to cyclic AMP as the cortical and ventral mesencephalic astroglia did, with generalized stellation. Prenatal striatal astroglia was characterized immunocytochemically as A2B5+, fibronectin+, vimentin+, S-100+ and GFAP-. Failure of early prenatal (E14, E17) striatal astroglia to differentiate in response to cyclic AMP, was overcome by previous (5-7 days) co-culture with primary cell dissociates from postnatal-, but not from prenatal donors, from all brain regions tested including a non-target region for striatal cells, such as septum. This effect was duplicated when striatal astroglia was co-cultured with cell populations enriched in neurons through Percoll gradients. Only cell-to-cell contact co-cultures were able to induce a change in the studied response. Dead neuron-enriched populations obtained following various types of physical treatments were also able to change significantly striatal cell response toward cyclic AMP. Enriched astroglial populations from postnatal donors did not change striatal astroglial response toward cyclic AMP, except for ventral mesencephalic astroglia which induced a comparatively reduced but significant increase in striatal cell responsiveness. It is concluded that astroglial maturation and potential for phenotype expression during brain development proceeds with regional heterochrony. Also, that maturation of prenatal striatal astroglia responsiveness toward cyclic AMP is inducible by non-diffusible factors, probably of neuronal origin, expressed in live or dead primary cultures from various, homotopic and heterotopic, postnatal brain regions. It is further suggested that striatal afferents and/or mature local striatal neurons express membrane associated molecules that regulate responsiveness for phenotype expression of striatal glial cells, thus reinforcing the concept of a highly interactive, continuous neuron-glial developmental process that takes place during brain organization.

"Rodent-like" and "primate-like" types of astroglial architecture in the adult cerebral cortex of mammals: a comparative study.

Previous observations disclosed that astroglia with interlaminar processes were present in the cerebral cortex of adult New and Old World monkeys, but not in the rat, and scarcely in the prosimian Microcebus murinus. The present report is a more systematic and comprehensive comparative analysis of the occurrence of such processes in the cerebral cortex of several mammalian species. Brain samples were obtained from adult individuals from the following orders: Carnivora (canine), Rodentia (rat and mouse), Marsupialia (Macropus eugenii), Artiodactyl (bovine and ovine), Scandentia (Tupaia glis), Chiroptera (Cynopteris horsfieldii and C. brachyotis), and Primate: Prosimian (Eulemur fulvus), non-human primate species (Cebus apella, Saimiri boliviensis, Callithrix, Macaca mulatta, Papio hamadryas, Macaca fascicularis, Cercopithecus campbelli and C. ascanius) and from a human autopsy. Tissues were processed for immunocytochemistry using several antibodies directed against glial fibrillary acidic protein (GFAP), with or without additional procedures aimed at the retrieval of antigens and enhancement of their immunocytochemical expression. The cerebral cortex of non-primate species had an almost exclusive layout of stellate astrocytes, with only the occasional presence of long GFAP-IR processes in the dog that barely crossed the extent of lamina I, which in this species had comparatively increased thickness. Species of Insectivora and Chiroptera showed presence of astrocytes with long processes limited to the ventral basal cortex. Interlaminar GFAP-IR processes were absent in Eulemur fulvus, at variance with their limited presence and large within- and inter-individual variability as reported previously in Microcebus murinus. In New World monkeys such processes were absent in Callithrix samples, at variance with Cebus apella and Saimirí boliviensis. Overall, the expression of GFAP-IR interlaminar processes followed a progressive pattern: bulk of non-primate species (lack of interlaminar processes)--Chiroptera and Insectivora (processes restricted to allocortex) < strepsirhini < haplorhini (platirrhini < catarrhini). This trend is suggestive of the emergence of new evolutionary traits in the organization of the cerebral cortex, namely, the emergence of GFAP-IR long, interlaminar processes in the primate brain. Interlaminar processes may participate in a spatially restricted astroglial role, as compared to the one provided by the astroglial syncytium. It is proposed that the widely accepted concept of an exclusively astroglial syncytium is probably linked with a specific laboratory animal species ("rodent-type" or, rather, "general mammalian-type" model) that misrepresents the astroglial architecture present in the cerebral cortex of most anthropoid adult primates ("primate-type" model), including man.

Patterned distribution of immunoreactive astroglial processes in the striate (V1) cortex of New World monkeys.

The possibility of a modular organization of non-neuronal elements was analyzed in the opercular region of the striate neocortex in adult New World monkeys. For this purpose, and in order to follow possible correlations in the general organization of neuronal and astroglial elements, immunocytochemical procedures for Glial Fibrillary Acid Protein (GFAP) and Microtubule Associated Protein 2 (MAP-2), in addition to cytochrome oxidase (COX) histochemistry, were applied to tangential and coronal sections and analyzed by using computer-assisted procedures. Astroglial interlaminar processes stemming from superficial laminae did not traverse lamina IVA, and thus did not appear in deeper layers. Clearly definable interlaminar processes were predominantly concentrated in laminae II-III. A honeycomb- like lattice was observed in tangential sections, with a "cell" size distribution similar to the MAP-2-IR lattice, suggesting an intimate association with the pyramidal columns. Additionally, analysis of similar sections disclosed the periodic appearance of large patches with high density of interlaminar processes, indicating a nonhomogeneous distribution of GFAP-IR processes in the striate cortex. COX "blobs" appeared frequently to coincide with areas expressing high density of GFAP-IR elements. These findings add a new perspective to current concepts of astroglial organization in the striate cortex of primates and reveal the existence of a non-neuronal modular organization in the primate striate cerebral cortex, and suggest that possible correlations between relative distributions of neuronal and astroglial elements should be further analyzed in cortical areas with a clear modular organization such as the striate cortex.

Disruption of patterns of immunoreactive glial fibrillary acidic protein processes in the Cebus Apella striate cortex following loss of visual input.

Long, interlaminar, astroglial processes and its patterned organization in the striate cortex of adult primates was previously described. Loss of visual input following bilateral retinal detachment and degeneration in an adult Cebus apella monkey resulted three months later in reduction of interlaminar processes immunoreactive to Glial Fibrillary Acid Protein antibody, loss of the honeycomb-like pattern normally present in tangential sections, and loss of high density patches of terminal segments of those processes in the opercular striate. These results further indicate the highly interactive nature of neuron-glial cerebral cortex architecture, and the dynamic regulation of astroglial interlaminar processes.

Cell-cell coupling in cultures of striatal and cortical astrocytes of the monkey Cebus apella.

Astrocytes were cultured from striatum and neocortex of fetal (embryonic day 90) monkeys (Cebus apella). The cultures grew well, and the cells retained viability after freeze-storage and thawing. The cells displayed depolarized membrane potentials (-19 and -33 mV, for striatal and cortical cells, respectively) and the vast majority of cells were dye-coupled to a mean of 7 (1-18) neighbouring cells. Cell coupling was blocked by octanol (0.25 and 0.5 mM) but was independent of high K+ (10 and 50 mM) and glutamate (500 microm). Thus, cultures of fetal primate astrocytic cells are established as a model system for studies on astroglial cell-cell coupling.

Astroglial interlaminar processes in the cerebral cortex of prosimians and Old World monkeys.

Previous observations have shown that astrocytes with interlaminar processes are present in the cerebral cortex of humans and New and Old World monkeys, but not in the rodent. The present report furthers the analysis of possible evolutionary aspects regarding the expression of such astroglial features. A comparison between young and adult Microcebus murinus, a prosimian, and Old World monkeys (Macaca mulatta and Papio hamadryas) is presented. Brain samples were processed for glial fibrillary acidic protein (GFAP), vimentin, MAP2 and SMI 311 immunocytochemistry, using different procedures. The cerebral cortex of adult Microcebus showed the presence of long astroglial processes, albeit reduced in number and length with respect to those observed in Old World monkeys. Macaca and Papio showed dense packing of such processes extending in most cortical regions to a depth of approximately 700 micrometers. Based on double immunolabelling for GFAP and MAP 2 antigens, the location and extent of these processes was shown to overlap with areas traversed by bundles and individual apical dendrites. Aged Old World specimens depicted an increased thickness of terminal portions of interlaminar processes, with increased morphological alterations. Comparisons made between the average thickness of the "brush" composed of interlaminar processes and the thickness of lamina I among the species analyzed disclosed an absence of relationship between them. This suggests that interlaminar processes do not represent cellular adaptations to the increase in thickness in superficial cortical laminae, but rather to some other evolutionary pressure. Since astroglial interlaminar processes are already present in a prosimian, although in a comparatively reduced manner, it is suggested that such processes underwent an early expression within the primate order, with increasing presence in more recent primate species.

Cerebrospinal fluid from L-dopa-treated Parkinson's disease patients is dystrophic for various neural cell types ex vivo: effects of astroglia.

Cerebrospinal fluid from L-dopa-treated Parkinson's disease patients and subjects without neurodegenerative diseases (controls) was explored in its trophic properties as culture medium on a variety of cells from neural origin. Primary cultures of regional brain dissociates from rat and Cebus apella monkey fetuses, immature rat adrenal chromaffin cells, phaeochromocytoma (PC12), and neuroblastoma (NB69) cell lines as well as subcultured fetal rat astroglia were used as target cells for 24- to 48-h culture periods. Most cerebrospinal fluid samples from L-dopa-treated patients had a general dystrophic effect. This phenomenon was more apparent on striatum and ventral mesencephalon than on cerebral cortex cell dissociates. The deleterious effect of these samples was abolished by previous exposure to fetal astroglial cells. Neuroblastoma cells showed no differential response when exposed to samples from control and L-dopa-treated patients. Phaeochromocytoma cells did not grow processes under any of the samples assayed in the time interval explored, but neither showed evidence of dystrophy. The relevance of these findings to the transplantation of different cell types as one of the possible therapies for Parkinson's disease is discussed. The suggestion is made that CSF testing prior to transplantation may aid in anticipating its possible outcome. Cotransplantation of neuronal cells with subcultured astroglia may foster survival and growth of the former cells.

Immunocytochemical and electron microscope observations on astroglial interlaminar processes in the primate neocortex.

At variance with the rat, previous observations disclosed the presence of long interlaminar astroglial processes in the cerebral cortex of adult nonhuman primates. To examine its presence in human cerebral cortex, samples of frontal and temporal cerebral cortices were obtained during programmed brain surgery from a young patient with an intraventricular astrocytoma, and from one young and two adult patients with frontal and temporal lobe focal epilepsy, respectively. Samples of the visual cortex were also obtained at an autopsy of an 84-year-old woman without any known neurological disease. Brain tissues were processed for GFAP-IR immunocytochemistry. Long, interlaminar, GFAP-IR astroglial processes of usually 300-500 microm, but occasionally reaching almost 1,000 microm, were observed. These processes resembled those previously described in the cerebral cortex of adult New World monkeys. Available data suggest that they may represent a predominant characteristic in postnatal primate cerebral cortex. EM analysis of club-like endings disclosed a multilamellar organization of GFAP-IR intermediate filaments, and the presence of mitochondria and amorphous, electron dense material. Their possible function is yet to be determined.

Interlaminar astroglial processes in the cerebral cortex of non human primates: response to injury.

At variance with the rodent, presence of long glial fibrillary acid protein-immunoreactive (GFAP-IR) astroglial processes traversing several cortical laminae (interlaminar processes) appears to be characteristic of the primate cerebral cortex. Their permanence or changes may constitute a significant factor in the functional alterations that develop after brain injury. The immediate and long term response of such astroglial processes to local application of KCl or lesioning, respectively, was analyzed in the striate cortex of adult Cebus apella monkeys. Intraparenchymal injection of 5 mM (within physiological range) or 50 mM (injury levels) KCl into the striate cortex of Cebus apella monkeys resulted three hours later in increased GFAP immunoreactivity in astroglial cells and processes, in the development of numerous foldings and thickenings of GFAP-IR filaments, as compared to mechanical lesioning alone. Such changes were not observed in a cortical region that approximately included laminae IV and V. These results suggest that the immediate GFAP-IR response to KCl cannot be solely explained on the basis of an exposure of new GFAP epitopes due to conformational changes following K-induced cell oedema, or to disaggregation of such filaments. Three months after mechanical lesioning of the frontal and striate cortex, interlaminar processes were absent up to 1.0 mm from the lesion site, and a predominant astrocytosis was present. Long term effects of mechanical lesioning on interlaminar processes resulted in a persistent reduction of these long processes in the vicinity of the lesion, suggesting a significant and prolonged alteration of the astroglial architecture in the adult primate cerebral cortex. It is speculated that these astroglial changes may bear a relationship with the functional alterations observed during the recovery process after brain injury.

Postnatal development of interlaminar astroglial processes in the cerebral cortex of primates.

UNLABELLED: Long astroglial processes traversing several cortical laminae appear to be characteristic of primate brains. Whether interlaminar processes develop as a modification of radial glia or are truly postnatal elements stemming from stellate astroglia, could be assessed by analyzing their early developmental stages. A survey of glial fibrillar acidic protein immunoreactive (GFAP-IR) astroglial interlaminar processes in the cerebral cortex of Ceboidea monkeys at various postnatal developmental ages, and in human cortical samples of a ten day and a seven year old child disclosed that such processes develop postnatally. At one month of age GFAP-IR interlaminar processes in monkeys were scarce and short in most frontal, parietal or occipital (striate) cortical areas, except for sulcal (principal and orbital sulci) and temporal cortical areas. Some processes were weakly positive for vimentin, and these were most abundant in ventral temporal cortical areas. At two months of age processes were present in all these areas, albeit in restricted patches and significantly shorter than in adults. The expression of this pattern was increased at seven months of age. At three years of age almost every area showed abundant processes and with lengths comparable to the adult Ceboidea individuals. In humans, at 10 days of age long interlaminar processes were readily apparent in a frontal cortex sample, becoming most apparent at the age of seven years although not reaching yet the adult characteristics as described previously. CONCLUSIONS: (1) GFAP-IR interlaminar processes develop postnatally, thus typifying a subtype of the classical stellate forms; (2) they bear no obvious direct relationship with radial glia; (3) their development is not contemporary among the various cortical regions. These long cellular processes represent an addition to those already described for other astroglial cell types in the adult mammalian brain (Golgi-Bergmann glia, tanicytes, Muller cells).

Ex vivo astroglial-induced radial glia express in vivo markers.

Ex vivo induction of radial-like glia has been previously reported to occur following exposure of cerebral cortex subcultures from fetal origin to cerebral cortex astroglial-conditioned medium. The present report further confirms similarities between in vivo and ex vivo radial glia, using additional criteria: adhesion of primary cell dissociates to glial processes, with presumptive cell migration along them, punctuate labelling for laminin, and immunolabelling with Rat-401 antisera.

In vitro induction of radial-like cells by leptomeningeal and cortical astroglial conditioned media. Effect of protease inhibitors.

A population of subcultured astroglia from rat fetal cortex was transformed into radial-like cells after exposure to cerebral cortex astroglial conditioned medium in vitro. Such changes were also induced by basal medium modified by fetal leptomeningeal subcultures, but not by postnatal leptomeninges nor by fetal skin fibroblasts. The radializing effects of astroglial conditioned medium were inhibited by previous heat treatment. The addition of protease inhibitors to the basal medium did not cause spontaneous radialization of subcultured cortical astroglia, but increased the length of cell processes and incidence of radial-like forms when added to cortical astroglial conditioned medium. It is concluded that cortical astroglia and leptomeningeal cells share the capability of synthesizing and releasing diffusible molecules into the culture medium which act as morphogenetic inducers in vitro. Based on the present results, it is suggested that such effects would depend on the presence of instructive factor(s) in the conditioned medium which are able to induce rearrangement of the cytoskeleton, rather than on secreted molecules able to modify cell adhesion to the substrate.

Regional differences in in vitro growth of neural cell processes during development.

Primary cell cultures from cerebral cortex, striatum and ventral mesencephalon obtained from rat fetal (embryonic day 17, E17) or postnatal (day 2, PN2) donors were grown either in media conditioned by subcultured astroglia from the same regions, an artificial trophic medium, normal human amniotic fluid, or in normal human cerebrospinal fluid. To estimate the presence of neuronal-like and non-neuronal cells, cell morphology and immunocytochemistry against microtubule-associated proteins and beta-tubulin were taken into consideration. The percentage of emitting neural cells and length of cell processes were determined after 24 hr in culture. Growth of cell processes in neuronal and non-neuronal cells from prenatal striatum was minimal compared with that in cerebral cortex and ventral mesencephalon, regardless of the culture condition. Nerve growth factor, basic fibroblast growth factor or epidermal growth factor did not significantly modify cell growth in E17 cultures, except for epidermal growth factor, which reduced the number of emitting cells in striatal cultures and increased it in cerebral cortex ones. Cultures derived from postnatal striatum showed a significant increase in neurite length when grown in an astroglial conditioned medium as compared to cultures derived from prenatal (E17) striatum. Results suggest significant regional differences in the brain regarding growth of cell processes at age E17, and reversal of striatal ability to grow cell processes by postnatal day 2. Reduced growth of cell processes showed by E17 striatum cultures was rather independent of the culture media. This fact could suggest that such early regional differences would depend on characteristics of sublineages present at this developmental stage, which would modulate the organization of regional neuropils. The restricted growth of cell processes in cultures from E17 striatum, no longer present in postnatal striatum, suggests that inputs to the striatum may modify expression of cell lineages at later stages of development.

Interlaminar astroglial processes in the cerebral cortex of adult monkeys but not of adult rats.

Three adult Cebus apella and two Saimiri sciureus monkeys, as well as four adult rat brains were processed for glial fibrillary acid protein (GFAP) immunocytochemistry. Irregular patches with relatively abundant, long astrocytic processes were found in monkey but not in rat brains, along the cingulate, as well as orbital, medial and dorsolateral cortices of the frontal lobe, and in temporal lobe cortical areas. The observed long astrocytic processes alternated in an apparent irregular fashion with others with a predominantly clear background and scattered GFAP-immunoreactivity (IR) astrocytes or with a predominant GFAP-IR-dense fibrillary meshwork. They usually stemmed from lamina I or II and coursed through deep laminae. Entorhinal cortex in rat brains showed occasional astroglial processes somewhat longer than those commonly observed in other cortical regions, albeit significantly shorter than those observed in Ceboidea monkeys and not truly interlaminar. These observations indicate a clear morphological difference in astroglial organization between nonhuman primates and a rodent brain. They further show that 'interlaminar astrocytes' are distributed nonhomogeneously in the frontal cortex of New World monkeys. No interlaminar processes were observed in cortical areas of rat brains. Even though the physiological role of interlaminar astrocytes in a primate brain remains unknown, their significant departure from the morphology of classical, stellate astrocytes suggest that they may subserve different cortical requirements.