Neurochemical Research (1) 261-273 (1976)

P R E P A R A T I O N OF G L I A L P L A S M A MEMBRANE FROM A CELL F R A C T I O N E N R I C H E D IN ASTROCYTES FRITZ A . H E N N AND ANDERS HAMBERGER University of Iowa College of Medicine Iowa City, Iowa 52242 Institute of Neurobiology, Gothenberg, Sweden

Accepted February 18, 1976

A technique for obtaining glial plasma m e m b r a n e has been developed, starting with a bulk-prepared glial cell-enriched fraction from rabbit cerebral cortex. The astrocytic-enriched fraction was hand-homogenized in isotonic sucrose media, and the crude membrane fraction sedimented at 3,000g. The isolation of a membrane-enriched fraction was accomplished with sucrose density gradient centrifugation. The plasma membrane fraction was collected at the interphase between 31.5% and 25.5% sucrose. Enzymatic and electron-microscopical analyses indicated a 4-7-fold enrichment in plasma membrane, and a 15-20% contamination with microsomal and mitochondrial material. Some multilaminar membrane structures were also seen in the fraction.

INTRODUCTION The astrocytic glial cell that surrounds the neuronal perikaryon, axon, and nerve ending has recently been a focus for renewed interest in questions on regulation of extraneuronal ion levels (1,2), amino acid transport (3,4), and transmitter inactivation (5,6). All these problems relate to glial membrane composition and transport properties. Functional studies show distinct electrophysiological characteristics of the glial membrane that differ, qualitatively, from those of neurons (7). The behavior and components of glial plasma membrane have to some extent been studied in tumor material and tissue culture. However, tumors

261 9 1976 Plenmn PublishingCorporation, 227 West 17th Street, New York, N.Y. 1001I. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical. photocopying, microfilming,recording, or otherwise, without written permissionof the publisher.

262

HENN AND HAMBERGER

contain large numbers of mesenchymal cells and blood vessels in addition to the tumor cells, and both these samples may reflect a pathological rather than a normal state of membrane composition and metabolism. More recently, the plasma membrane of an oligodendroglial fraction has been isolated and reported (8). Beginning with fractions of predominantly astrocytic cells, prepared by the method of Blomstrand and Hamberger (9,10), we have worked out a technique to obtain glial plasma membrane from rabbit cerebral cortex. The membrane fraction has been studied by biochemical and morphological criteria. The results suggest that the final fraction contains better than 80% glial plasma membrane.

EXPERIMENTAL PROCEDURE Preparation of a Glial Cell-Enriched Fraction. A fraction, enriched in glial cells, was prepared from whole brains of white rabbits, according to the method of Bloomstrand and Hamberger (9). For each preparation, 10-12 rabbit brains were used. Cortical tissue was chopped into pieces 0.4 mm thick, which were incubated at 37~ for 45 min in a shaking water bath. When the tissue had been softened by the incubation, mild mechanical disruption was carried out by passing the material through 1-mm pore size nylon mesh attached to the opening of a syringe. After passage of the suspension through nylon meshes of 243/.~ and 48 t~, a final f~tration through a double layer of 48-/z mesh was carded out. The filtrate was centrifuged at 150g. The sediment was resuspended in Ficoll, and glial cells were isolated by centrifugation for 120 rain at 50,000g in a discontinuous sucrose-Ficoll gradient. The glial fraction was collected between 12% and 15% Ficoll and pelleted by centrifugation at 10,000g for 20 min after dilution with 2-3 vol isotonic sucrose. The fraction was checked under the light microscope after staining with methylene blue. The fraction contained 80-85% glial cells, and the different types of glial cells were represented; however, the predominant species was astrocytic (11). The small contamination consisted of blood capillaries and shorn cell processes, some of which could be of neuronal origin. Preparation of Glial Plasma Membranes. The glial cell pellet was resuspended in 0.25 M sucrose containing 10 mM MgCI2 and 5 mM Tris-HCl, pH 7.4. The fractionation scheme is shown in Fig. 1. It involves gentle homogenization by hand in loose glass-Teflon homogenizers with 12-15 up-and-down strokes with little twisting. The total amount of the glial cell-enriched fraction from 10 rabbits was homogenized in 5 ml medium, then diluted to 15 ml and centrifuged for 10 min at 150g to sediment free nuclei and unbroken cells. The supernatant, which contained the plasma membrane fragments, was saved, and the pellet rehomogenized and diluted with the sucrose solution as described. This washing of the pellet was repeated 4-5 times, until at least 50% of the glial nuclei were released, as checked by light microscopy after methylene blue staining. Care was taken to avoid destruction of nuclei during homogenization, and in order to ensure gentle conditions, the homogenization was discontinued when between 50% and 60% of the glial nuclei appeared to be released. The combined supernatants were centrifuged for 20 min at 3,000g. The pellet was resuspended in the sucrose medium and homogenized lightly, and the suspension was mixed with 4 parts 60% sucrose to obtain a final sucrose concentration of 50%.

263

GLIAL CELL M E M B R A N E S Isolated glial cells

I

Gentle homogenization x 3 - 5 (see text) ISO g

I

IlO min I

IPl

Supernatant 3000 g 20 min

l

_ - _-

l

crude nuclei

t

Suspended2MSucrose 54,000 g 180 min

Supernatant 10,O00 g 15 min

i

f Supernatant

crude mitochondria

lO0,000 g 60 min

P3 Nuclei

Supernatant cake crude capillaries

Density gradient 54,O00 g 180 min

1

~qicrosomes

I

i

1P4

I

19; w/w Sucrose

~

Any retained myelin

25.5 w/w Sucrose

U

Plasmamembrame

31.5

50

w/w

Sucrose

Crude mitochondria

Sucrose + Sample

w/w

FIG. 1. Flow chart showing the preparative scheme used in the isolation of a glial plasma membrane fraction.

This suspension, which had a final volume of 15 ml, was transferred to a tube for the SW 27 rotor (Beckman-Spinco), followed by successive additions of 10 ml each 31.5%, 25.5%, and 19% sucrose (wt/wt) to fill the tube. The gradient was centrifuged immediately for 180 rain at 54,000g. A small band appeared at the 19/25.5% interphase. It had no appreciable enzyme activity and was not used. The 25.5/31.5% interphase contained the plasma membrane fraction, and consisted of two bands close to each other. Both bands were collected together, since preliminary experiments had shown that they did not differ appreciably with respect to the specific activity of the marker enzymes. They were diluted with 0.32 M sucrose and spun at 10,000g to collect the plasma membrane pellet. The 31.5/ 50% interphase contained mitochondria and the other membranous material. No appreciable pellet was obtained. The mitochondrial fraction was obtained from the combined material from the 31.5/50% interphase and the pellet after centrifugation of the 3,000g supernatant at 10,000g. The

264

HENN AND HAMBERGER

resuspended pellet was layered on top of a discontinuous sucrose gradient (12), and the purified mitochondria were collected as the pellet after centrifugation for 120 min at 54,000g. The microsomal fraction was collected by centrifugation of the supernatant from the 10,000g centrifugation for 180 rain at 100,000g. Electron Microscopy. The plasma membrane pellets were fixed for 2 h in 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, and postfixed for 2 h in 2% OSO4. Fixation was followed by washing in 0.1 M cacodylate buffer, dehydration in ethanol, and embedding in Epon 812. Thin sections were cut with an LKB Ultratom III, and were examined in a Siemens Elmskop 1A after staining with uranyl acetate and lead citrate. Enzyme Assays. Glucose-6-phosphatase was assayed according to Hfibscher and West (13). Lysosomal contamination was examined by assaying for acid phosphatase at pH 5 with /3-glycerophosphate as a substrate (14). Cytochrome oxidase activity was measured spectrophotometrically according to the method of Appelmans et al. (15), and monoamine oxidase was measured via the procedure of Schnaitman et al. (16). The N A D P H - C y t . C reductase was measured according to the procedure of Dallner et al. (17). The N A D H Cyt. C oxidoreductase was assayed according to Mahler and Cotman (18). The N a - K +and Mg~+-stimulated ATPase were assayed in a media containing 50 mM Tris-HC1, pH 7.4, 5 mM MgC12, and for Na+-K + stimulation, 100 mM NaC1 and 10 mM KC1 were added. The increment of activity on adding Na++K + was taken as the Na+-K + ATPase activity, and was consistently inhibited 80-100% by ouabain. The procedure of Touster et al. (19) was used to assay 5' nucleotidase. Analytical Procedures. Protein concentration was determined by the procedure of Lowry et al. (20), and D N A was determined via the modification method reported by Waldman and Alto (21). Phosphate was determined according to Fiske and SubbaRow (22) in all cases where inorganic phosphate was liberated.

RESULTS The initial problem in preparing a glial membrane fraction is to assure that one is dealing primarily with material of glial origin. We chose to approach this problem by working with a fraction of isolated cells containing a minimal amount of material of neuronal origin and heavily enriched in glial cells. This approach was feasible following the initial work of Rose (23), who had shown it was possible to prepare bulk fractions of isolated cells enriched in neurons and glia, respectively. The preparation developed by Blomstrand and Hamberger (9,10) used in these studies has been refined to the point where, on morphological grounds, there is virtually no cross contamination of intact neuronal and glial cells. The fractions appear to contain approximately 80% cellular material; however, a real problem in this study was to assure that synaptosomal membrane was not carried along and purified during the subsequent membrane purification. Hamberger et al. (24) have recently summarized the data regarding the purity of these cell fractions, and an

GLIAL CELL MEMBRANES

265

additional study has been carried out aimed at quantitating the degree of synaptosomal contamination in the isolated glial cell fraction. Rabbit brains were radioactively labeled in vivo, and a synaptosomal fraction was prepared and purified on a CsCI2 gradient (25). This labeled fraction was added to material from unlabeled rabbit brains, and the normal cell separations were carried out. The resultant glial fraction showed less than 10% synaptosomal contamination. The preparative scheme developed for obtaining plasma membrane utilized the formation of relatively large membrane fragments of glial cells via controlled gentle homogenization. This scheme allowed relatively easy separation from the bulk of contaminating microsomal material by collecting a pellet containing plasma membrane at 3,000g. At this low centrifugal force, a majority of microsomal material failed to sediment. The principal problem in the resulting fractionation was the design of the discontinuous sucrose gradient to minimize mitochondrial contamination. Glial mitochondria have a peak of apparent density of approximately 42% sucrose when run in a continuous gradient (26). Still, a large part of the plasma membrane fragments had to be sacrificed with the mitochondria, as the separation step had to be made as low as 31.5% sucrose to keep mitochondrial contamination of the plasma membrane fraction at a desired minimum.

FIG. 2. A low-power (15,000x) scan of the glial plasma membrane preparation, showing vesicular membrane patterns and occasional mitochondria.

266

HENN AND HAMBERGER

The morphology of the plasma membrane fraction is shown in Fig. 2. Most of the membranes appeared as closed vesicles, having the usual unit membrane structure. Multilaminar structures, resembling myelin, were also seen, often surrounding other structures. Occasional mitochondria were found, but rarely synaptic junctions, while ribosomes were virtually absent. The yield of glial plasma membrane protein was low. The starting material, 70-80 mg total glial cell protein from 10-12 rabbit brains, resulted in 0.1-0.2 mg purified plasma membrane proteins. The problem of myelin contamination was difficult to assess, since no marker enzyme appeared specific for the myelin fraction. In particular, 2'-3' cyclic nucleotide phosphorylase has been reported to be widely associated with membrane fragments assumed to be of glial origin (27). We attempted to minimize myelin contamination by using grey cortical matter as the starting material for the isolation of glial cells. The isolated glial fraction containing predominantly astrocytes and occasional microglia and oligodendroglia was examined for cerebroside and sulfatide content and compared to a myelin fraction. It was found that myelin contained approximately 15 times more cerebroside and 20 times more sulfatides than the isolated glial cell fraction on a basis of micromoles of lipid per milligram of protein (28). This finding indicates that a gross contamination of our glial membrane with myelin fragments is unlikely, especially since the oligodendrocyte serves as the precursor for myelin

FIG. 3. A higher magnification(34,000• of the glial membrane fraction, illustrating an area of multilamin.armembraneprofiles.

GLIAL CELL MEMBRANES

267

FIG. 4. High magnification(100,000x) of multilaminarmembranes, showingan absence of intrapefiod lines found in myelinpreparations. and would contain all the lipid constituents found in myelin (8). A lowpower view of this membrane preparation is given in Fig. 2. This view demonstrates a somewhat heterogeneous membrane fraction with an occasional poorly preserved mitochondrion. There are a variety of vesicular membrane fragments. In Fig. 3, a higher-power view demonstrates the variety of membrane profiles, including an area of multilaminar whorls reminiscent of myelin. In Fig. 4, these whorls are examined under high magnification, and it appears that they are composed of discrete trilaminar membranes without evidence of intraperiod lines, suggesting that some of these fragments may result from membranes other than myelin. As biochemical criteria for the purity of the plasma membrane fraction, we used the specific activities of enzymes known as markers for plasma membranes, mitochondria, and microsomes. Assays were done on the plasma membrane fraction, the purified mitochondrial fraction, and the microsomal fraction. Table I presents the specific activities of the enzymes, and the contaminations or enrichment calculated from these data. As shown by the specific activity of N A D P H cytochrome C oxidoreductase, the plasma membrane fraction contains a microsomal contamination of approximately 3%. The assays for the microsomal markers were all carried out immediately after isolation of the appropriate subcellular fractions in order to minimize the inactivation of these labile enzymes; in addition, the time

268

HENN

AND

HAMBERGER

,-:r4 +1

+1

+1

+,+,~ ~t~ 0

;5 c4~

Preparation of glial plasma membrane from a cell fraction enriched in astrocytes.

A technique for obtaining glial plasma membrane has been developed, starting with a bulk-prepared glial cell-enriched fraction from rabbit cerebral co...
1MB Sizes 0 Downloads 0 Views