An analysis of astrocytic cell lines with different abilities to promote axon growth

The adult mammalian central nervous system (CNS) lacks the capacity to support axonal regeneration. There is increasing evidence to suggest that astrocytes, the major glial population in the CNS, may possess both axon-growth promoting and axon-growth inhibitory properties and the latter may contribute to the poor regenerative capacity of the CNS. In order to examine the molecular differences between axon-growth permissive and axon-growth inhibitory astrocytes, a panel of astrocyte cell lines exhibiting a range of axon-growth promoting properties was generated and analysed. No clear correlation was found between the axon-growth promoting properties of these astrocyte cell lines with: (i) the expression of known neurite-outgrowth promoting molecules such as laminin, fibronectin andN-cadherin; (ii) the expression of known inhibitory molecules such tenascin and chondroitin sulphate proteoglycan; (iii) plasminogen activator and plasminogen activator inhibitor activity; and (iv) growth cone collapsing activity. EM studies on aggregates formed from astrocyte cell lines, however, revealed the presence of an abundance of extracellular matrix material associated with the more inhibitory astrocyte cell lines. When matrix deposited by astrocyte cell lines was assessed for axon-growth promoting activity, matrix from permissive lines was found to be a good substrate, whereas matrix from the inhibitory astrocyte lines was a poor substrate for neuritic growth. Our findings, taken together, suggest that the functional differences between the permissive and the inhibitory astrocyte cell lines reside largely with the ECM.     

Astroglial alterations in rat hippocampus during chronic lead exposure.

The present study was performed in order to follow the response of astroglial cells in the rat hippocampus to chronic low-level lead exposure. The experiments combined immunohistochemistry using anti-glial fibrillary acidic protein (GFAP) antibody and conventional transmission electron microscopy (EM). Chronic administration with drinking water [1 g% w/v (subclinical dose) of lead acetate dissolved in distilled water] was started through the mother's milk when pups were 7 days old. Following weaning, experimental offspring were treated for 3 months with the same concentration of adulterated water. The group of intoxicated animals and their controls were sacrificed by perfusion-fixation at 30, 60, and 90 days of exposure. After 60 days of lead treatment, staining of GFAP-positive cells demonstrated an astroglial transformation from the quiescent to the reactive state, characterized by an increase in GFAP. In control rats no changes in GFAP immunostaining were observed. The intensity of the astroglial response was enhanced after 90 days of lead intoxication, showing an increment of GFAP immunoreactivity. Quantification of these changes was made by computerized image analysis, confirming that the sectional areas of the astroglia in lead-exposed animals were larger than those in controls. These results are consistent with the ultrastructural alterations. Simultaneously with the increment in gliofilaments, intranuclear inclusions were seen in some astrocytes. The mechanisms by which lead affects astrocytes are unknown. Probably the astroglial changes induced by lead intoxication produce microenvironmental modifications that may disturb the neuronal function.     

Distribution of glutathione and glutathione-related enzyme systems in mitochondria and cytosol of cultured cerebellar astrocytes and granule cells

The cellular and regional distribution of glutathione (GSH) and GSH-related enzyme systems involved in cellular defense against reactive oxygen species and electrophilic xenobiotics in the nervous system has been extensively studied. However, little is known about the subcellular distribution of GSH systems in brain tissue and cultured neural cells. The present study investigates the distribution of mitochondrial and cytosolic GSH and GSH-related enzymes in cultured cerebellar astrocytes and granule cells, and compares them with levels in the adult rat cerebellum. Cytosolic GSH levels and cytosolic activities of glutathione reductase (GR), glutathione peroxidase (GPx) and glutathione-S-transferase (GST) in astrocytes were 57, 153, 245, and 92% higher than those found in granule cells, respectively. In contrast, granule cells contained significantly higher mitochondrial GSH levels than astrocytes. Granule cells also demonstrated comparable mitochondria/cytosolic concentrations of GSH and GR, GPX and GST activities to those observed in the cerebellar tissue, whereas ratios in astrocytes were markedly lower. Although in vitro treatments with 100 μM ethacrynic acid depleted both cytosolic and mitochondrial GSH in cultured astrocytes and granule cells in a time-dependent fashion, cellular GSH in granule cells was more resistant to the GSH-depleting agent than astrocytes. These results suggest that although GSH and GSH-related enzymes are abundant in cytosolic compartments of astrocytes, mitochondrial pools are relatively small. Since brain mitochondria are sites of significant hydrogen peroxide generation, the mitochondrial localization of GSH and its associated enzymes in neural cells provide important defenses against toxic oxygen species in the nervous system. Differences in subcellular distribution of GSH systems in individual neural cell types may provide a basis for selective cellular and/or subcellular expression of neurotoxicity.     

Immunization-induced inflammatory infiltration of the central nervous system in transgenic mice expressing a microbial antigen in astrocytes

Transgenic mice expressing a defined microbial antigen from central nervous system (CNS) cell type-specific promoters can be utilized to investigate the consequences of induction of peripheral immune responses to foreign antigens produced by different CNS cell types. Immunization of mice expressing β-galactosidase (β-gal) in astrocytes with this protein resulted in antigen-dependent infiltration of the CNS by mononuclear cells, principally CD4⁺ T lymphocytes and monocyte/macrophages. The perivascular and intraparenchymal infiltrates, which were located predominantly in the hippocampal formation and cerebellum, the areas of highest β-gal expression, were associated with astrocytosis, microgliosis, and a generalized increase in blood-brain barrier permeability. The resemblance of these pathological changes to aspects of human immune inflammatory CNS disorders e.g. multiple sclerosis, suggests that an initiating step in the process by which such complex diseases are produced could be the induction of peripheral immune responses to antigens expressed in astrocytes.     

Characterization of “plasma proteins” secreted by cultured rat macroglial cells

The brain is isolated behind a blood-tissue barrier that restricts the access of circulating proteins to neural cells. There is evidence that some of these proteins are synthesized within the central nervous system. The present study examines the synthesis and secretion of such proteins by cultured macroglial cells. Primary glial cultures were derived from cortical and subcortical regions of neonatal rat brains, and subsequent secondary cultures were enriched in type-1 astrocytes, type-2 astrocytes, or oligodendrocytes. Newly synthesized proteins were immunoprecipitated from the culture media using antisera directed against whole rat serum. All three types of glial cells secreted a range of plasma proteins. In general, type-1 astrocytes secreted more of these proteins than did type-2 astrocytes or oligodendrocytes, although the one-dimensional polyacrylamide gel electrophoresis (PAGE) profiles were specific for each cell type. Antisera directed against specific plasma proteins identified three of the most abundant proteins secreted by type-1 astrocytes as transferrin, α-2-macroglobulin, and ceruloplasmin. Northern blot analysis of cellular RNA confirmed that type-1 astrocytes contained transferrin mRNA, and that it was more abundant in cultures derived from subcortical regions than from cortical regions. In situ hybridization studies revealed that virtually all type-1 and type-2 astrocytes contained transferrin mRNA. Since the proteins identified in this study have been proposed to have a variety of neurotrophic roles in the central nervous system, these data further extend the range of possible functions that glial cells may serve in the CNS.     

纳络酮对脂多糖诱导原代培养星形胶质细胞释放谷氨酸的抑制效应

目的探讨纳络酮对脂多糖诱导原代培养星形胶质细胞释放谷氨酸的抑制效应。方法体外培养星形胶质细胞于融合状态,随机分为5组(1)对照组(L0 N0);(2)1μg·ml-1脂多糖组(L1 N0);(3)1μg·ml-1脂多糖 0.5μmol·L-1纳络酮组(L1 N0.5);(4)1μg·ml-1脂多糖 1.0μmol·L-1纳络酮组(L1 N1.0);(5)1μg·ml-1脂多糖 2.0μmol·L-1纳络酮组(L1 N2.0)。各组培养液均换成Neurobasal/B27无血清培养液后,在相应组中加入上述相应终浓度的脂多糖和纳络酮,继续培养2h。用反相高效液相色谱分析方法测定各组细胞外液中谷氨酸含量。结果L1 N0组中谷氨酸含量明显高于L0 N0组,其差异有极显著性意义(P<0.05);L1 N0.5、L1 N1.0、L1 N2.0组内谷氨酸的含量则随着纳络酮用量的增加而逐渐降低,其中L1 N2.0组与L1 N0组相比,谷氨酸含量差异有显著性(P<0.05)。结论纳络酮能抑制脂多糖刺激大脑皮质星形胶质细胞释放谷氨酸,具有一定的脑保护作用。     

成年大鼠神经元和星形胶质细胞细胞周期蛋白依赖性激酶抑制因子表达的差异研究

目的比较研究成年大鼠细胞周期蛋白依赖性激酶抑制因子在神经元和星形胶质细胞的表达差异。方法应用免疫荧光和激光扫描共聚焦显微镜观察成年大鼠生理状态下大脑皮层或海马CA1、CA3、DG区神经元和星形胶质细胞细胞周期蛋白依赖性激酶抑制因子(CDKI)p15Ink4b、p21cipl的表达。结果成年大鼠海马区和大脑皮层的神经元有p15Ink4b和p21cipl的表达,细胞核和细胞浆均有表达,且以胞核为主;星形胶质细胞也有上述细胞周期调控蛋白的表达,便细胞数目较少,并且表达这些指标的星形胶质细胞多聚集在海马区。结论成年大鼠大脑皮层和海马区的神经元和星形胶质细胞均表达p15Ink4b和p21cipl,而其在神经元的表达较星形胶质细胞更为普遍。     

C-terminal tail of β-adrenergic receptors: immunocytochemical localization within astrocytes and their relation to catecholaminergic neurons in N. tractus solitarii and area postrema

β-Adrenergic receptors (βAR) in the medial nuclei of tractus solitarii (m-NTS) and area postrema (AP) may bind to catecholamines released from neurons, whereas only the AP has fenestrated capillaries allowing access to circulating catecholamines. Since varied autonomic responses are seen following βAR activation of the dorsal vagal complex, including the m-NTS and AP, we hypothesized that there might be a cellular basis for varied responses to βAR stimulation that depends pn the differential access to circulating catecholamines. Therefore, we comparatively examined the ultrastructural localization of the βAR in relation to catecholaminergic neurons in these regions. An antibody directed against the C-terminal tail (amino acids 404–418) of hamster β-adrenergic receptor (βAR404) was used in this study. The localization of βAR404 was achieved by the avidin-biotin peroxidase complex (ABC) technique in combination with a pre-embed immunogold labeling method to localize tyrosine hydroxylase (TH), the catecholamine-synthesizing enzyme. Within m-NTS and at subpostremal border, labeling for βAR404 was evident along the intracellular surface of plasma membranes of small, apparently distal, astrocytic processes. Astrocytic processes with βAR404-immunoreactivity formed multiple, thin lamellae around TH-labeled and non-TH neuronal cell bodies and dendrites. βAR404-immunoreactive astrocytes also extended end-feet around blood vessels and surrounded groups of axon terminals that were directly juxtaposed to each other. Some, but not all, of these axons demonstrated TH-immunoreactivity. Fewer βAR404-immunoreactive astrocytes were detected in AP, regardless of their proximity to catecholaminergic processes or blood vessels. The present astrocytic localization of βAR404, together with the earlier, neuronal localization of βAR's third intracellular loop, suggest that the βAR may be substantially different between neurons and astrocytes. The regional difference in the prevalence of βAR404-immunoreactive astrocytes suggests that these receptive sites may either: (i) be preferentially activated by catecholamines released from terminals rather than circulating catecholamines; or (ii) be down-regulated in AP due to blood-born substances, such as catecholamines. The extensive localization of βAR in the border between m-NTS and AP also suggests that catecholaminergic activation of these astrocytes may dictate the degree of diffusion of catecholamines which are of neuronal or vascular origin. The specific localization of βAR404-immunoreactivity to the more distal portions of astrocytes suggests the possibility that astrocytes have restrictive distributions of βAR and that the β-adrenergic activation lead to morphological or chemical changes that are also localized to the distal portions of astrocytes. Additionally, the detection of βAR404 in astrocytes contacting non-TH-immunoreactive neurons suggests the possibility for catecholaminergic modulation of non-catecholaminergic neurons via the activation of astrocytes.     

Localization and characterization of epidermal growth-factor receptors in the developing rat medial septal area in culture

The presence and binding properties of epidermal growth-factor receptors (EGF-Rs) in different cell types purified from the rat medial septal area in culture were investigated. We report that astrocytes, oligodendrocytes and neurons from this area possess EGF-Rs while microglia do not. EGF-binding sites are detectable on astrocytes derived from the medial septum of both embryonic and neonatal rats. Scatchard analysis of the data for astrocytes from the fetal rats show that EGF specifically binds to both high- (K_d = 7.21 × 10⁻¹⁰ M, B_max = 3602 receptors/cell) and low-affinity (K_d = 3.99 × 10⁻⁸ 10⁻⁸ M, B_max = 6,265 receptors/cell) receptors on these cells. On the other hand, astrocytes purified from neonatal tissue possess a greater number of high-affinity receptors (B_max = 10,938 receptors/cell) when compared with the embryonic astroglia. With time in culture, the number of both types of receptors on neonatal astrocytes decreases. Oligodendrocytes also possess high- and low-affinity EGF-Rs with dissociation constants of 3.25 × 10⁻¹⁰ M and 3.85 × 10⁻⁸ M, respectively. The number of receptors on oligodendrocytes is significantly lower than those on neonatal astrocytes (B_max = 1185 and 25,081 receptors/cell for high- and low-affinity binding sites, respectively). Finally, neurons from this area also exhibit two different EGF-R types with dissociation constants similar to those described for astrocytes. As the number of receptors/neuron (B_max = 136 and 1159 receptors/cell for high- and low-affinity binding sites, respectively) appears to be extremely low, it is possible that EGF specifically binds only to a subpopulation of neurons from this area. These studies demonstrate which cell types in the developing medial sepal area posses EGF-Rs and provide a detailed characterization of these binding sites. These EGF-R-bearing cells may be potential targets for this growth factor or for transforming growth factor α in this brain area.     

Growth factor pre-treatment differentially regulates phosphoinositide turnover downstream of lysophospholipid receptor and metabotropic glutamate receptors in cultured rat cerebrocortical astrocytes

Reactive gliosis is an aspect of neural plasticity and growth factor (GF) stimulation of astrocytes in vitro is widely regarded as a model system to study astrocyte plasticity. Astrocytes express receptors for several ligands including lysophosphatidic acid (LPA) and sphingosine-1-phosphate (S1P), agonists for the G-protein-coupled lysophospholipid receptors (lpRs). Activation of lpRs by LPA or S1P leads to multiple pharmacological effects including the influx of calcium, phosphoinositide (PI) hydrolysis, phosphorylation of extracellular receptor regulated kinase (ERK), release of arachidonic acid, and induces mitogenesis. Treatment of astrocytes in vitro with a growth factor cocktail (containing epidermal growth factor [EGF], basic fibroblast growth factor [bFGF] and insulin) led to a marked attenuation of lpR-induced PI hydrolysis. In contrast, under identical conditions, GF treatment led to marked potentiation of PI hydrolysis downstream of activation of another abundantly expressed G-protein coupled receptor, mGluR5. Quantitative gene expression analysis of GF-treated or control astrocytes by TaqMan RT-PCR indicated that GF treatment did not change gene expression of lpa1 and s1p1, but increased gene expression of s1p5 which is expressed at very low levels in basal conditions. These results suggest that GF differentially affected PLC activation downstream of mGluR5 versus lpR activation and that the changes in mRNA levels of lpRs do not account for marked attenuation of agonist-induced phosphoinositide turnover.