Microtubule-associated Proteins and Astrocyte Differentiation In Vitro

Primary cultures of mouse brain astrocytes have been used to identify the microtubule-associated proteins (MAPs) present in this cell type at different stages of in vitro differentiation. The MAPs of the astrocyte have been identified by polyacrylamide gel electrophoresis and immunological detection. Two antisera were raised against two brain MAPs, tau and MAP-2. These antisera were also used to label the microtubular network in the intact astrocytes at different stages of the culture. The mature astrocyte contains a variety of MAP-like proteins. Anti-MAP-2 serum detected several proteins of high molecular weight (380,000, 260,000, 205,000 and 165,000 mol wt) and one microheterogeneous peak of 83,000 mol wt. Anti-tau also detected high molecular weight components (380,000 to ~200,000 tool wt) but not the 165,000-mol-wt peak; in addition two microheterogeneous peaks of 83,000 and 62,000 tool wt were detected by the anti-tau serum. The 62,000-mol-wt peak was therefore detected only by the anti-tau serum whereas the 83,000-mol-wt component cross-reacted with both antisera. At early stages of the culture the immature cell contained about two times less immunoreactive material than at mature stages. Qualitative changes of the high molecular weight components were also observed. In the intact cell both antisera revealed a dense fibrous network. At early stages of the culture the astroblasts were stained by the antisera but the reaction was very diffuse in the cytoplasm; few fibrous cells were intensively stained. Morphological differentiation, which began after serum deprivation and which was accelerated by forskolin (a drug that induces cyclic AMP accumulation), led to high labeling of both the cell body and the cellular processes. In the presence of colchicine the staining regressed, the processes shortened, and the cell returned to a less-apparently differentiated state. Although less elaborated than that of the neuron, the cell form of the astrocyte changes markedly during differentiation. Astrocytes elaborate cell processes that resemble those of the neurons but they are, on the average, much shorter. Although the astrocytes develop different types of expansions and are interconnected among themselves, with the neurons, with the blood vessels, and with the basal lamina of the pial surface, their function(s) remain(s) largely unknown (for review see reference 21). In recent years it has been increasingly clear that the shape of cells is determined to a large extent by their cytoskeleton. Neurite outgrowth, for instance, requires microtubule assembly (19, 44, 53); of particular interest have been the so-called TH[ JOURNAL OF CELL BIOLOGY VOLUME 101 DECEMBER 1985 2 0 9 5 2 1 0 3 © The Rockefel ler Univers i ty Press • 0021-9525 /85 /12 /2095 /09 $1.00 microtubule-associated proteins (MAPs),1 which seem to play an important role both in neurite outgrowth and in the specification of what will be an axon or a dendrite (for review see reference 22). In contrast, very little is known about the microtubule organization of the astrocytes, their role during the differentiation process, and the growth of the cellular processes. Some data on the microtubules or the MAPs present in glioma cell lines (8, 17, 18, 39) have been published, t Abbreviations used in this paper: A-MAP, astrocyte microtubuleassociated-type protein; B-MAP-1 and B-MAP-2, brain MAP-1 and -2, respectively; B-tau, brain microtubule-associated tau protein; MAP, microtubule-associated protein. 2095 on N ovem er 6, 2017 jcb.rress.org D ow nladed fom but nothing is known of the MAPs of normal astrocytes. MAPs co-assemble with tubulin and decorate the surface of microtubules not only in brain but also in various other tissues and cell types (see reference 22). Brain tissue contains high molecular weight MAPs (3 l, 39, 48), or B-MAP-1 and B-MAP-2; and low molecular weight MAPs, or B-tau proteins (15, 16), which are themselves heterogeneous. B-MAP-2 (~300,000 mol wt), for example, has been resolved into two discrete peaks, both of which are thermostable proteins (40); at early stages of brain development B-MAP-2 is present but as a single peak (4). B-MAP-1 (~350,000 mol wt) splits into three bands (8) which are all thermolabile (27, 28) and are therefore eliminated during heat treatment of the microtubules (24). Recent developments (4 l) have shown that another high molecular weight MAP (MAP4) is composed of a triplet of proteins (215,000-240,000 tool wt). B-tau, the low molecular weight MAP, is composed at an adult stage of brain development of a series of closely spaced bands of 58,000-65,000 mol wt (15, 16) which can be resolved by two-dimensional gel electrophoresis to at least 20-30 entities (40). At an early stage of brain development the low molecular weight MAPs present in the tau region clearly differ in number, molecular weight, peptide mapping, assembly promoting activity of tubulin (25, 36) and encoding mRNAs (20, 26). These "young B-MAPs" have been identified at an immature stage of development, i.e., when the brain contains few differentiated neurons and few glial cells. Little is known of the functional significance of the heterogeneity of B-MAPs and of the developmental changes that occur during brain differentiation. Some data suggest that the different MAPs are not evenly distributed within the brain and in different parts of the neurons (3, 7, 14, 29, 30, 37, 51). Some MAPs, such as MAP-2 and tau, also interact with microfilaments and neurofilaments (2, 6, 27, 34, 43, 46). To understand better the functional significance of the heterogeneity of B-MAPs we decided to use in vitro culture systems of differentiating brain cells as homogeneous as possible. We report in this work the results obtained with a primary culture that has been shown to be a highly enriched population of astrocytes (1, 2, 5, 49). Such cultures begin spontaneous differentiation when deprived of serum; serum deprivation is performed when contact inhibition of growth occurs. The differentiation can be accelerated by dibutyryl cyclic AMP (5) or by forskolin (this work), a drug that increases the intracellular level of cyclic AMP. Two polyclonal antibodies directed against adult B-tau and B-MAP-2, respectively, have been used (a) to identify the thermostable MAPs eventually present in the immature and differentiating astrocytes (A-MAPs); and (b) to stain the microtubule network in the intact cell at different stages of the culture. MATERIALS AND METHODS Products: GTP and 2-(N-morpholino)ethane sulfonic acid were from Boehringer Mannheim Diagnostics, Mannheim, FRG. EDTA was from Prolabo, Paris). Forskolin was from Calbiochem-Behring Corp., La Jolla, CA. EGTA, colchicine, leupeptin, phenylmethylsulfonyl fluoride, trypsin inhibitor, benzamidine, Tween-20, and bovine albumin (fraction V) were purchased from Sigma Chemical Co., St. Louis, MO. Acrylamide, bisacrylamide, SDS, and nitrocellulose sheets were from Bio-Rad Laboratories, Richmond, CA. Antirabbit (Ig ~2~I)-labeled whole antibody from donkey (specific activity, 5-20 uCi/ ug) was from Amersham International, Amersham, U.K. Cell Culture Conditions: Astroglial primary cultures were obtained as previously described (1, 9) from cerebral hemispheres of neonatal mice. These cultures contain an astroglial population constituted mostly by proto2096 THE JOURNAL OF CELL BIOLOGY • VOLUME 101, 1985 plasmic-like cells and by a few fibrous cells. These cells grow rapidly for 2 wk in the presence of 20% fetal calf serum and reach a spontaneous state of differentiation at around the third week of culture. Immunohistochemical staining with anti-glial fibrillary acidic protein and anti-glutamine synthetase antisera, two astroglial markers, labeled -80% of the cell population (49). The negative labeling with neuronal markers such as glutamate decarboxylase or ~,enolase confirms the absence of any detectable neuronal contamination. Preparation of Yhermostable 100,000 g Brain Supernatants: Thermostable supernatants were prepared from adult rat brain. Brains were homogenized in l ml/g tissue of buffer A: 0.1 M 2-(N-morpholino)etbane sulfonic acid (pH 6.4) containing 0.5 mM MgCI2, 1 mM EGTA, 0.1 mM EDTA, 1 mM GTP, l mM 2-mereaptoethanol, l0 mM benzamidine, 5 #M leupeptin, 25 mg/l trypsin inhibitor, and 1.5 ml/1 phenylmethylsulfonyl fluoride (50 mg/ml dimethylsulfoxide). The homogenate was incubated at 4°C for 30 min and then centrifuged at 100,000 g for 60 min. The supematant was incubated at 4"C for 60 min in the presence of 0.75 M NaCI and 2 mM dithiothreitol. The preparation was immersed in boiling water for 5 min. After centrifugation (20,000 g, 30 rain), the supernatant was dialyzed overnight at 40(3 against buffer A. Protein concentration was determined as described in reference 10 with bovine serum albumin as standard. Preparation of Astrocyte Extracts: Since most of the B-MAPs are thermostable, heat-stable extracts were prepared from the astrocytes to achieve a partial purification. After culture medium was removed, each dish was rinsed three times with I ml phosphate-buffered saline (PBS). The cells were then scraped with 0.5 ml buffer A, and each lot of four dishes was rinsed with 0.5 ml of the same buffer. The suspension was homogenized at 4*C with a Potter Dounce homogenizer (model 06001, 20/55 u, Poly-Labo Block, Strasbourg, France) and treated as described above for the thermostable 100,000 g supernatant. Slab gel electrophoresis of these extracts followed by staining with Coomassie Blue showed that most of the heat-stable proteins had a molecular weight lower than ~50,000