106

Brain Research, 549 (1991) 106-111 © 1991 Elsevier Science Publishers B.V. 0006-8993/91/$03.50 ADONIS 000689939116603S

BRES 16603

Differential expression of protein kinase C isozymes in rat glial cell cultures Eliezer Masliah 1, Kazunari Yoshida 1, Shun Shimoharna 2, Fred H. Gage 1 and Tsunao Saitoh 1 1Department of Neurosciences, School of Medicine, and Centerfor Molecular Genetics, University of California, San Diego, La Jolla, CA 92093-0624 (U.S.A.) and 2Departmentof Neurology, Faculty of Medicine, Kyoto University, Sakyo-ku, Kyoto (Japan) (Accepted 11 December 1990)

Key words: Astroglia; Glia; Microglia; Protein kinase C

Protein kinase C (PKC) is a family of closely related enzymes implicated in molecular processes involved in growth and differentiation in a variety of cells. We studied the presence and distribution of 4 PKC isozymes in gliai cell cultures of the rat hippocampus employing antisera raised against synthetic peptides predicted from the cDNA sequences corresponding to the C-terminal portion of 4 PKC isoforms, a, ill, fill, and ~,. PKC(a) and -(fllI), but neither PKC(flI) nor -(y) isoforms were detected in glial cultures of the rat hippocampus. Anti-PKC(a) immunostained all glial cells, whereas anti-PKC(fllI) faintly stained about 20% of total glial cells resembling the type-2 astrocyte that were GFAP immunopositive, with few processes. Anti-PKC(fllI) did not stain about 80% of the glial fibriltary acidic protein (GFAP)-immunopositive cells with a few thick processes which resembled the type-1 astrocyte. A few cells that stained intensely with anti-PKC(flII) were GFAP immunopositive and possessed fine, but well-developed, multiple processes. Faint PKC(flII) immunoreactivity was also detected among anti-MBP-positive cells (possibly oligodendrocytes), RCA-l-positive cells (possibly microglia), and small, oval, anti-GFAP-positive cells. These results suggest the involvement of distinct PKC isoforms in different glial functions.

INTRODUCTION Protein kinase C (PKC), a family of enzymes associated with a variety of cell functions, including secretion, proliferation, and differentiation, is especially abundant in the central nervous system (CNS) 15. Until recently, brain P K C was thought to be a single entity. However, it now appears that the apparently homogeneous brain enzyme is a mixture of several types. Analyses of complimentary D N A clones of PKC by several laboratories suggest that more than 6 genes exist for the brain enzyme 3,936,17. The previous enzyme purification studies were conducted with brain preparations which consisted of many different cell types 7'8. Thus, it is not clear which PKC isozymes are derived from neurons and which are from glial cells. PKC in neurons has been extensively examined 7 whereas less attention has been paid to the role of PKC in glial cells, even though neurotransmitterdependent stimulation of phosphoinositide hydrolysis has been demonstrated in astrocytes ~8, and abundant PKC has been reported in these cultures 14. Moreover, immunocytochemical approaches suggest that PKC is localized in cells resembling oligodendrocytes in rat brain white matter 5'20 and in cells resembling glia in rat hippocampus

following hippocampal deafferentation 19. Furthermore, PKC activity has been measured in cultured glial cells 12' 14,22 while the differentiation and proliferation of glial cells are enhanced by phorbol esters 6'13, all indicating an important role for PKC in glial cells. To better understand the possible role of P K C in glial cells, we studied immunocytochemicaUy the presence and distribution of the PKC isozymes in glial cell cultures of the rat hippocampus. We found selective localization of PKC isozymes in glial cell cultures.

MATERIALS AND METHODS

Antibody preparation The preparation and characterization of polyclonai antisera to 4 PKC isozymes (a, ill, fllI, and ),) are described in detail elsewhere l°' zo. Briefly, antisera were raised in rabbits against synthetic peptides predicted from the human cDNA sequence corresponding to the C-terminal portion of 4 PKC isozymes, a, ill, fllI, and y. All antisera gave a titer between 1/2000 and 1/5000 on dot blots with the individual peptides. Antisera were further affinity-purified on immobilized antigen columns prepared by coupling free peptide to activated aldehyde-agarose (ActigelA, Sterogene, Arcadia, CA). These antisera stained an 80-kDa protein on Western blots of rat brain homogenate. The staining on blots was abolished by the inclusion of 20/~g/mi of the free corresponding peptide but not by other peptides, demonstrating the specificity of our antisera.

Correspondence: T. Saitoh, Department of Neurosciences, School of Medicine (M-024), UCSD, La Jolla, CA 92093-0624, U.S.A.

107 Cell cultures The glial cell-predominant culture was obtained by the following method. From 1-day neonatal rats (Sprague-Dawley) the brains were aseptically removed. Hippocampus was dissected and collected in Dulbecco's modified Eagle's medium (DME). The brain fragments made by passing through stainless-steel mesh (pore size 250 #m) were suspended in DME containing 10% fetal calf serum (FCS), 4.5 g/l glucose, and 40 mg/l gentamicin, and seeded in a poly-L-lysine-coated 25-cm2 Coming flask. The cells were cultured for 9 days at 37 °C, under 5% CO2/95% air, and transferred to two 25-cm2 Coming flasks after trypsinization (0.25% trypsin, 0.02% EDTA in phosphate-buffered saline (PBS)) and cultured for another 7 days. The cells were then detached from the culture flasks by trypsinization, suspended in the culture medium, seeded on poly-L-lysine-coated coverslips (13 mm diameter) placed in a 24-well Costar culture plate (2 x 105 cells/well), and cultured for about 3 days.

Peroxidase immunocytochemistry Immunohistochemistry was performed according to a modification of the avidin-biotin peroxidase procedure. The coverslips were fixed in 4% paraformaldehyde in PBS and 80% ethanol-12 N HCI (100:1), rinsed 3 x 10 min in PBS followed by 3 x 5-min rinses in 0.1 M PBS, pH 7.4, with 0.25% Triton X-100 (PBST), incubated for 15 min with 0.3% H202 in PBST to inhibit the endogenous peroxidase, and for 30 min with 3% normal goat serum (NGS) in PBST to block the non-specific protein binding sites. The coverslips were rinsed in PBST and then incubated with one of 4 different antibodies against PKC(a), -(flI), -(flII), or -(Y) for 18 h at 4 °C on

a rotator in a humid environment with sodium azide added to prevent bacterial growth. The coverslips were rinsed 3 x 5 rain in PBST with 1% NGS and incubated with biotinylated anti-rabbit lgG (1:200, Vector Laboratories) in PBS with 1% NGS for 1.5 h. The coverslips were then rinsed 3 x 5 rain in PBST with 1% NGS and incubated in avidin D peroxidase (1:100, Vector Laboratories) in PBST with 1% NGS for 1 h and then rinsed 3 x 5 rain in Tris-buffered saline (TBS), pH 7.4. PKC-positive structures were visualized by incubating the tissue in 0.05% 3,3"-diaminobenzidine tetrachloride (DAB) with 0.01% H20 2 and 0.04% NiCI in 40 ml TBS for 5-15 min. Immunohistochemistry for glial fibrillary acidic protein (GFAP) and neurofilament protein was performed in the same manner, except using mouse monoelonal anti-human GFAP (Boehringer Mannheim, Indianapolis, IN) and mouse monoelonal anti-200-kDa neurofilament (Labsystem, Helsinki, Finland) as a primary antibody, respectively. Specificity of the immunocytochemical reaction with anti-PKC antibodies was confirmed by the absence of labeled profiles in tissue sections incubated with either NGS or with antibodies preadsorbed with 20/zg of the corresponding synthetic peptides.

Double-labeling and laser confocal imaging Cultured cells on coverslips were fixed for 15 min in 80% ethanol with 1:100 HCI. The coverslips were then rinsed twice with PBS pH 7.4, pretreated for 30 min with 0.3% H20 2 and incubated overnight at 4 °C in a mixture of the following: (a) anti-PKC(flII)/anti-GFAP, (b) anti-PKC(fllI)/biotinylated RCA-1 (microglial marker, Vector Labs, Burlingame, CA), and (c) anti-PKCfflII)/anti-myelin basic protein (MBP, Hybritech Labs, San Diego, CA). The coverslips

Fig. 1. PKC(a) (A) and -(flII) (C) immunoreactivity in glial cell cultures of the rat hippocampus. The majority of glial cells were PKC(a) immunoreactive, large, round, and had multiple processes, resembling type-I astrocytes. Strongly PKC(flII)-immunoreactive cells were very few, small, rectangular, and had a few well-developed but fine processes. Lightly stained PKC(flII)-positive cells are about 20% of total glial cells and with well-developed multiple processes resembling type-2 astrocytes. The corresponding synthetic peptide abolished the staining. (B) Adsorbed anti-PKC(a) and (D) adsorbed anti-PKC(flII). Bar = 100/~m.

108 were then incubated with a combination of secondary antibodies tagged with FITC and Texas Red. For detection of the biotinylated RCA-1 binding, the glial cells were reacted with avidin conjugated to FITC or Texas Red. The double-labeled coverslips were transferred to gelatin-coated slides and mounted with antifading media (4% n-propyl gallate, Sigma, St. Louis, MO). The imaging of the double-labeled glial cells was done with the BioRad MRC-600 laser confocal scanning microscope mounted on a Nikon Optiphot microscope. This system permits the simultaneous analysis of double-labeled cells in the same optical plane. Images of serial 1-/~m optical sections were recorded and the digitized video images were stored on an optical disk for subsequent processing and analysis. The image files were subsequently transfered and converted to the Macintosh format. Quantitative analysis (measurement of sizes of cells and processes) of these digitized images displaying the different types of glial cells were carried out with the aid of the software package 'image' running on a Mac Ilci. The image software is a public domain package written by Wayne Rasband of NIH. RESULTS In the present study, we used the cultures o b t a i n e d after two passages from the primary cultures. A l t h o u g h a significant n u m b e r of i m m a t u r e glia ( G F A P - n e g a t i v e , vimentin-positive cells) exist in the early stage of our

GFAP

PI

PKC 13I I

II

cultures, such cells could not be seen at the stage we used for this study. We seeded these cells at higher cell density, and these cultures reached confluency in 24 h. Thus, we decided to use the cells 3 days after passage (19 days in total) for the present study. The great majority of rat neonatal h i p p o c a m p a l glial cells grown for 19 days stained with the astrocyte-selective monoclonal antiG F A P protein. Neurons, which are positive for 200-kDa neurofilament protein, did not survive in this culture condition (data not shown). To examine closely the localization of each P K C isozyme within individual gliai cells, we studied the distribution of immunoreactivity. A t this stage of differentiation, a differential pattern of P K C immunoreactivity was present. W e did not find a remarkable difference in the staining patterns of PKCs between 3 days and 7 days after the last passages. First, there were P K C ( a ) - and P K C ( f i l I ) - i m m u n o r e a c tive glial cells in culture (Fig. 1A,C), but neither PKC(fiI) nor -(y) immunoreactivity was d e t e c t e d (not shown). The P K C ( a ) and -(fill) i m m u n o r e a c t i o n s were a d s o r b e d by incubation of antibodies with their corresponding pep-

PKC 1311

I

G FA P

GFAP

Fig. 2. Confocal laser imaging of double-immunolabeled glial cell cultures. In each case the image on the left side corresponds to the FITC signal and the image on the right side is Texas Red. A: small GFAP-positive glial cells with fine PKC(fllI)-positive processes (arrowheads). In addition, PKC(fllI) immunolabeled small oval cells (arrow). B: large GFAP-positive glial cells display PKC(flII) immunoreactivity in some areas of the cytoplasm and in some of their thick, cellular processes. C: RCA-1 bound to large cells with abundant cytoplasm and fine cellular processes (arrow). In contrast, PKC(flII) immunoreacted with astrocyte-like cells that are wrapped with thick, RCA-l-positive processes (arrowhead). D: RCA-l-positive cells are GFAP negative, and GFAP-positive astrocytes with fine cellular processes are RCA-1 negative. Bar = 25/zm. •

109

!

PKCP l i

RCA- I

RCA- I

PKC pll I

PKCPli

MBP

adsorbed

L

Fig. 3. Laser confocalimagingof double-labeledglial cells. A: large groups of big RCA-l-positive cells presented faint to negligiblePKC(fllI) immunoreactivity(arrow). B: the majority of PKC(fllI)-positivecells are not RCA-1 positive, one RCA-l-positive cell (arrow) presented light PKC(fllI) immunoreactivity.C: anti-MBP-positivecellswere relativelysmallcompared to the astrocytesand containedfew fine processes; some of these cellspresented light PKC(fllI)reactivity.D: the correspondingsyntheticpeptide completelyadsorbed anti-PKC(fllI)immunoreaetivity. Bar = 25/tm. tides (Fig. 1B,D) but not with irrelevant peptides. Second, all the glial cells in culture were PKC(a) positive but not all of the gliai cells were PKC(fllI) immunoreactive. Third, PKC(fllI)-immunoreactive cells had a distinct shape relative to the majority of PKC(a)-immunopositive cells. We further studied in detail the PKC(fllI)-immunoreactive cells using a double-labeling technique. Cells, double-immunolabeled with PKC(fllI)/GFAP, showed that anti-GFAP immunostained mainly two types of astrocytes: one small (approximately 150 pm 2) with fine processes (less than 1 pm in diameter) (Fig. 2A) and the other large (approximately 250 /tm 2) with thick (approximately 2.5/~m in diameter) interconnected processes (Fig. 2B). The latter were also co-immunolabeled with anti-PKC(fllI). Occasionally, anti-PKC(fllI) immunoreacted with oval cells (approximately 60 pm 2) with multiple processes which were lightly GFAP positive, similar to mature Type 2 astroglia (Fig. 2A). In contrast, RCA-1 reacted in a granular fashion in large cells (approximately 200 pm 2) with abundant cytoplasm and fine, short (approximately 2.5/~m in diameter) processes (Fig. 2C). The great majority of these cells were GFAP and PKC(fllI) negative (Fig. 2C,D). Occasionally,

GFAP-positive glial cells which had interconnections with thick RCA-l-positive processes displayed PKC(flII) immunoreactivity (Fig. 2C). Another group of RCA1-positive cells, which were GFAP and PKC(flII) negative, was characterized by the presence of thick cellular processes and moderate amount of cytoplasm (Figs. 2D and 3A,B). Anti-MBP-positive cells (approximately 100 ~tm 2) were less abundant and morphologically were characterized by a few thin cellular processes (less than 1/~m in diameter), scant cytoplasm, and peripheral nuclei (Fig. 3C). These cells were lightly labeled with anti-PKC(flII) (Fig. 3C) and were RCA-1 and GFAP negative. Immunolabeled profiles were absent on the sibling cultures incubated with the anti-PKC(fllI) antibody preadsorbed with the peptide (Fig. 3D). DISCUSSION This study indicates: (1) PKC(a) and -(fllI) isozymes are detectable in glial cultures of the rat hippocampus; PKC(flI) and -0') isozymes are not. (2) The majority of GFAP-immunopositive cells having

110 multiple processes are PKC(a) immunoreactive but not PKC(filI) immunoreactive. (3) PKC(flII)-immunoreactive cells were divided into 5 classes: (a) a few intensely stained cells which were GFAP immunonegative and possessed only a few processes, (b) faintly stained cells (about 20% of all glial cells) which were GFAP immunopositive and possessed multiple processes, (c) a few faintly stained small oval cells with multiple processes with GFAP immunoreactivity, (d) a few faintly stained anti-MBP-positive cells without RCA-1 or anti-GFAP reactivity (possibly oligodendrocytes), and (e) a few faintly stained RCA-1positive cells, possibly microglial cells (although the majority of RCA-l-positive cells were PKC(flII) negative). We have shown that all 4 PKC isozymes were expressed differentially during development in primary neuronal cultures of the rat cerebellum using the same antibodies, and that PKC might play an important role in neuronal growth and synaptic development (in preparation). In contrast to the neuronal cultures, only PKC(a) and -(flII) isozymes were detected in glial cultures, suggesting the differential role of PKC in neurons and glia. Although neurotransmitter-dependent stimulation of phosphoinositide hydrolysis has been demonstrated in primary astrocyte cultures TM, and abundant phorbol ester binding sites were reported in these cultures TM, the role for each PKC isozyme in glial cultures is yet to be studied. As the differentiation and proliferation of glial cells in culture are enhanced by phorbol esters 6'~3, a role for PKC in the astrocyte is suggested. Our present study showed the differential expression of each PKC isozyme in in vitro cultured glial cells. Among the 4 PKC isozymes, PKC(a) and -(fill) might play a role in in vitro glial cells. We have reported the presence of PKC immunoreactivity in glial cells in vivo 1°'2°. Although no PKC transcripts (/31, fill, and ~) were noted in cell types other than neurons in in situ hybridization studies 2'23, a small number of glia-like cells were labeled with anti-PKC(filI) antibody in the non-lesioned rat hippocampus and lesioninduced reactive gliaqike cells were labeled not only with anti-PKC(filI) antibody but also with anti-PKC(a) antibody tg. The glia-like cells were not labeled with anti-PKC(fiI) antibody, suggesting the differential cell expression between PKC(fiI) and -(fill). Immunohistochemistry with anti-PKC(a) antibody revealed reactive glial cells around senile plaques in brains from AlzheiREFERENCES 1 Benowitz, L.I. and Rounenberg, A., A membrane phosphoprotein associated with neural development, axonal regeneration, phospholipid metabolism and synaptic plasticity, Trends

mer's disease patients w, This in vivo immunohistological study is consistent with that of the present in vitro study, suggesting PKC(a) and -(filI) might play an important role in glial cells both in vitro and in vivo. It is not a priori obvious that we can get 'reactive astrocytes' in our cultures. However, our data showed the expression of PKC(a) in cultured astrocytes, which could be detected in reactive astrocytes in vivo. Furthermore, the astrocytes obtained in our culture condition usually express vimentin, which is transiently expressed by reactive astrocytes in vivo in addition to GFAP. These findings suggest that cultured astrocytes used in the present study possess characteristics similar to 'reactive astrocytes' in vivo. It should be noted that, in contrast to anti-PKC(ct) immunoreactivity, only a portion of GFAP-immunopositive cells in vitro as well as in vivo were anti-PKC(fiII) immunoreactive. Astrocytes generally defined as GFAPimmunopositive cells have been subclassified into types 1 and 2 depending on morphology and cell surface antigens H. They might be further subclassified according to intracellular signal transduction systems, such as PKC, as well. In the current study, the shape of about 80% of the astrocytes resembled type 1 and the others resembled type 2, although we did not attempt to further identify the type 1 and type 2 astrocytes. The conclusive distinction between type 1 and type 2 astrocytes is based on molecular and immunological markers. Heterogeneity in the shape of cultured astrocytes might result from different growth stages of individual cells, which is not controlled in the current study. An investigation of the PKC substrates in glial components will be needed to clarify the physiological significance of PKC in the glial system. A glial protein, vimentin, is a substrate for PKC tS. Furthermore, GAP-43 seems to be expressed in type-2 astrocytes and oligodendrocytes4. GAP-43 has been thought to be a PKC substrate specific to neurons ~. But recent evidence suggested that this protein is more widely localized and involved in the regulation of G protein-mediated signal transduction pathways 21. Thus, it is probable that PKC in glial cells plays an important role in the response to neurodegeneration and regeneration.

Acknowledgements. This work was supported by grants from the McKnight Endowment Fund for Neuroscience, the Pew Charitable Trust, Institute for Research on Aging, UCSD, and the National Institutes of Health (AG-05131, AG-08201, AG-08205, and NS28121). Neurosci., 10 (1987) 527-532. 2 Brandt, S.J., Niedel, J.E., Bell, R.M. and Young, III, W.S., Distinct patterns of expression of different protein kinase C mRNAs in rat tissues, Cell, 49 (1987) 57-63. 3 Coussens, L., Parker, P.I., Rhee, L., Yang-Feng, T.L., Chen,

111 E., Waterfield, M.D., Francke, U. and Ullrich, A., Mutiple, distinct forms of bovine and human protein kinase C suggest diversity in cellular signalling pathways, Science, 233 (1986) 859-866. 4 Da Cunha, A. and Vitkovic, L., Regulation of immunoreactive GAP-43 expression in rat cortical macroglia is cell type specific, J. Cell Biol., 111 (1986) 209-215. 5 Girard, P.R., Mazzei, G.J., Wood, I.G. and Kuo, J.E, Polyclonal antibodies to phospholipid/Ca÷ dependent protein kinase and immunocytochemical localization of the enzyme in rat brain, Proc. Natl. Acad. Sci. U.S.A., 82 (1985) 3030-3034. 6 Honegger, P., Protein kinase C-activating tumour promotors enhance the differentiation of astrocytes in aggregating fetal brain cell cultures, J. Neurochem., 46 (1986) 1561-1566. 7 Kikkawa, U., Kishimoto, A. and Nishizuka, Y., The protein kinase C family: heterogeneity and its implications, Annu. Rev. Biochem., 58 (1989) 31-44. 8 Kikkawa, U., Takai, Y., Minakuchi, R., Inohara, S. and Nishizuka, Y., Calcium-activated, phospholipid-dependent protein kinase from rat brain: subcellular distribution, purification, and properties, J. Biol. Chem., 257 (1982) 13341-13348. 9 Knopf, I.L., Lee, M.-H., Sultzman, L.A., Kriz, R.W., Loomis, C.R., Hewick, R.M. and Bell, R.M., Cloning and expression of multiple protein kinase C cDNAs, Cell, 46 (1986) 491-501. 10 Masliah, E., Cole, G., Shimohama, S., Hansen, L., DeTeresa, R., Terry, R.D. and Saitoh, T., Differential involvement of protein kinase C isozymes in Alzheimer disease, J. Neurosci., 10 (1990) 2113-2124. 11 Miller, R.H., French-Constant, C. and Raft, M.C., The macroglial cells of the rat optic nerve, Annu. Rev. Neurosci., 12 (1989) 517-534. 12 Murphy, J.A., Chapman, I.A., Suckling, A.I. and Rumsby M.G., Protein kinase C activity in soluble fractions from glial cells in primary culture and subcultures, Neurosci. Lett., 85 (1988) 255-260. 13 Murphy, S., McCabe, N., Morrow, C. and Pearce, B., Phorbol

ester stimulates proliferation of astrocytes in primary culture, Dev. Brain Res., 31 (1987) 133-135. 14 Neary, I.T., Norenberg, L.O.B. and Norenberg, M.D., Calcium-activated, phospholipid-dependent protein kinase and protein substrates in primary cultures of astrocytes, Brain Research, 385 (1986) 420-424. 15 Nishizuka, Y., Studies and prospectives of the protein kinase C family for cellular regulation, Cancer, 63 (1989) 1892-1903. 16 Ohno, S., Kawasaki, H., Imajoh, S., Suzuki, K., Inagaki, M., Yokokura, H., Sakoh, T. and Hidaka, H., Tissue-specific expression of three distinct types of rabbit protein kinase C, Nature, 325 (1987) 161-166. 17 Ono, Y., Fujii, T., Ogita, K., Kikkawa, U., Igarashi, K. and Nishizuka, Y., The structure, expression, and properties of additional members of the protein kinase C family, J. Biol. Chem., 263 (1988) 6927-6932. 18 Pearce, B., Cambray-Deakin, M., Morrow, C., Grimble, I. and Murphy, S., Activation of muscarinic and of al-adrenergic receptors on astrocytes results in the accumulation of inositol phosphates, J. Neurochem., 45 (1985) 1534-1540. 19 Shimohama, S., Saitoh, T. and Gage, EH., Protein kinase C in hippocampus and septum following fimbria-fornix transection, Soc. Neurosci. Abstr., 14 (1988) 19. 20 Shimohama, S., Saitoh, T. and Gage, EH., Differential expression of protein kinase C isozymes in rat cerebellum, J. Chem. Neuronat., 3 (1990) 367-375. 21 Strittmatter, S.M., Valenzuela, D., Kennedy, T.E., Neer, E.J. and Fishman, M.C., Go is a major growth cone protein subject to regulation by GAP-43, Nature, 344 (1990) 836-841. 22 Walker, A.G., Chapman, I.A., Bruce, C.B. and Rumsby, M.G., Immunocytochemical characterization of cell cultures grown from l-2-day post-natal rat cerebral tissue, J. Neuroimmunol., 17 (1985) 1-20. 23 Young, III, W.S., Expression of three (and a putative four) protein kinase C genes in brains of rat and rabbit, J. Chem. Neuronat., 1 (1988) 177-194.