marker protein for astrocytes). However, there is no decrease in staining due to Mab J1-. 31 in sections incubated in antiserum to GFAP prior to incubation with ...
Bioscience Reports, Vol. 6, No. 1, 1986
A New "Marker" Protein for Astrocytes Rajkumari Singh, 1 Bhagirath Singh, and S. K. Malhotra 1 Received November 18, 1985 KEY WORDS:
A monoclonal antibody (Mab J1-31) has been produced by using human brain homogenate as immunogen in mouse. Double-label immunofluorescence microscopy on cryostat sections of human, rabbit and rat brain, reveals staining of cells that are also stained with antiserum to glial fibrillary acidic protein (GFAP, a commonly used marker protein for astrocytes). However, there is no decrease in staining due to Mab J131 in sections incubated in antiserum to GFAP prior to incubation with the J1-31 ascites fluid. Immunoprecipitation of aqueous and detergent extracts of brain homogenate gives a single band at 30K by SDS PAGE followed by autoradiography. Immunoelectron microscopy shows that the J1-31 antigen is associated with the cytoskeleton. Thus, the Mab J1-31 recognizes a new protein present in GFAP positive cells (astrocytes) in the brain. INTRODUCTION Currently, a few "marker" molecules are known for central nervous system (CNS); these include neurofilament (NF) protein for neurons (Steinert et al., 1984), galactocerebroside (Raft et al., 1979) and myelin basic protein (Sternberger et al., 1978) for oligodendrocytes, and glial fibrillary acidic protein (GFAP) for astrocytes (Bignami and Dahl, 1974; Bjorklund e~ al., 1984; Rafter al., 1983). Apart from their usefulness as specific markers, monoclonal antibodies (Mab) are invaluable in identifying novel neural and non-neural antigens (Hawkes et al., 1982; Lawson, 1983; Cole and Glaser, 1984) in differentiation and development. We have produced a Mab J1-31 (IgG2b isotype), against the crude homogenate of brain from the autopsy of patients with multiple sclerosis (MS, Malhotra et al., 1984). This monoclonal antibody recognizes a new "marker" protein of MW 30K in astrocytic cells in the CNS of humans and is designated as the J1-31 antigen. On the basis of immunofluorescence microscopy, a similar antigen was found to be present in the astrocytes of rat and rabbit brain. Besides 1 DepartmentsofZoologyand Immunology,Universityof Alberta,Edmonton,Alberta,CanadaT6G 2E9. 73 0144-8463/86/0100-0073505.00/0 9 1986Plenum Publishing Corporation
74
Singh, Singh and Malhotra
the CNS, no J1-31 antigen was detected in any of the several tissues of the rat tested by immunofluorescence microscopy.
MATERIALS AND METHODS
Characterization of J1-31 Antigen A crude human brain homogenate preparation previously used for immunization of mice to produce hybridomas (Malhotra et al., 1984) was utilized for these studies. A suspension of this homogenate in saline was sonicated for 2 rain at 4~ The resulting suspension was centrifuged at 700 g for 10 rain, and the pellet was discarded. The supernatant was centrifuged at 100,000g for 1 hr. The pellet and supernatant (supernatant I) were collected. The pellet was treated with 1% NP-40 detergent in 10 mM Tris-HCl-saline pH 7.2, containing 25 p M P M S F (phenyl methyl sulfonyl fluoride), 0.02% NAN3, and centrifuged at 100,000g for 1 hr. The supernatant (supernatant II) was collected. Both the supernatants I and II were found to contain J 1-31 antigen as determined by binding to the Mab J 1-31 in an ELISA (enzyme-linked-immunosorption assay). Therefore, they were pooled for subsequent experiments. The pooled supernatarhs were dialysed against phosphate buffered saline (PBS) and labelled with 1251 by the chloramine-T method (Greenwood et aL, 1963), This preparation was used for immunopreclpitation (Kessler, 1976): 2 x l0 s cpm of ~25I-labelled preparation was treated with 200#1 of protein ASepharose CL-4B for 30 rain at room temperature to remove non-specific binding. The supernatant obtained was mixed with 100 ~ti of Mab J I-31 ascites fluid for 1 hr at room temperature and then incubated with 100/~1 Protein A~Sepharose CL-4B. After 1 hr of further incubation, the gel was washed (four times) with i0 mM Tris-HCl-saline 1% NP40-0.1% BSA and centrifuged. The pellet thus obtained was used for SDS-PAGE (Laemmli, 1970) analysis under reducing conditions. For controls, normal mouse serum fl'om Balb/C mice, was used in place of J1-31 ascites. Rabbit anti-GFAP serum ( D a b ) and normal rabbit serum were also employed under similar conditions for comparison with the J1-31 antigen. G F A P was selected because it is widely used as a marker protein for astrocytes in CNS (Bignami and Dahl, 1974; Bjorklund et al., 1984; Raft et ai., 1983). 14C-iabelled molecular weight-marker proteins (BRL) were used as standards, and the gels were autoradiographed.
lmmunoelectron Microscopy 50 ~tm-thick sections of the rat cerebellum (fixed in 4% paraformaldehyde) were cut on an Oxford vibratome, incubated with the Mab J1-31 ascites fluid or with normal mouse serum as control (1:1,000 dilution) and peroxidase-antiperoxidase (PAP) reagents, fixed in osmium tetroxide before dehydration and embedding in Araldite (Malhotra et al., 1984). This sections were examined without further staining with heavy metal salts. Human autopsy material used in this investigation was obtained in frozen condition and the quality of preservation was not suitable for electron microscopy. Therefore only rat brain was examined by immunoelectron microscopy.
Marker Protein for Astrocytes
75
Immunofluorescence Microscopy Human, rabbit, and rat brain specimens were fixed in 4% paraformaldehyde in phosphate buffer (0.02 M, pH 7.2) and 8 I0 #m-thick sections were cut on a cryostat. The sections were picked up on a cover-glass for staining. They were incubated in the following order: 30% goat or rabbit serum (as appropriate) in PBS, primary antibody 01-31 or anti-GFAP antiserum at 1 : 1,000 dilution), washed in PBS, second antibody (goat or rabbit anti-mouse or goat anti-rabbit IgG conjugated to FITC or rhodamine), washed in PBS, anti-GFAP antiserum or J 1-31 (as appropriate for double labelling at the same dilution as above), washed in PBS, second antibody conjugated to FITC or rhodamine, washed in PBS and double distilled water before mounting in glycerol with p-phenylenediamine. The sections were examined by epifluorescence. Normal mouse or rabbit serum was run as control at the same dilution as the primary antibody (1:1,000). RESULTS AND DISCUSSION A single protein band at approximately MW 30K was obtained after immunoprecipitation followed by SDS-PAGE and autoradiography with Mab J 1-31 (Fig. 1). Under similar conditions, with anti-GFAP antiserum there were two to three bands but distinct from the 30K protein. GFAP was selected for comparison with J 1-31 antigen because it is widely used as a marker protein for astrocytes (Bignami and Dahl, 1974; Bjorkiund et al., 1984; Raft et al., 1983); and our results with double label immunofluorescence microscopy on the same cryostat section indicated that the staining due to Mab J1-31 is preferentially localized in GFAP positive cells (see below). Iodination and immunoprecipitation of the supernatants (I and II) with Mab J 131 gives a single protein band at 30K. However, there is a rare possibility that some proteins are not iodinated and may be recognized by the Mab. Also the possibility has not been ruled out that the detergent insoluble pellet contains some proteins reactive to the Mab. The appearance of this single protein band is particularly noteworthy because both supernatants I and II were mixed before iodination and subsequent immunoprecipitation and gel electrophoresis. Supernatant I is an aqueous extract whereas supernatant II is a detergent (NP-40) extract of the brain homOgenate. These results suggest that the MW 30K antigen exists in situ in human brain such that some of the protein is readily extracted, whereas the rest is extracted with detergent (Wang et al., 1984; Malfroy et al., 1985). In respect of the solubility of the J1-31 antigen protein in detergent and in aqueous extracts, it is of interest that similar solubility properties have been reported for a high-molecular-weight protein (MW 100K) from the rat brain extracts (Malfroy et al., 1985). Therefore, it is conceivable that the 30K protein is associated with the cytoskeleton, some of it being readily solubilized (aqueous extract perhaps monomeric form) and some more tightly integrated into the cytoskeleton (and perhaps polymeric form), and requires detergent for extraction. Thus J1-31 antigen differs from the known cytoskeleton proteins because it is present in aqueous as well as detergent extracts, whereas cytoskeletal proteins, in particular intermediate filaments, are generally identified by their detergent insolubility.
76
Singh, Singh and Malhotra
Fig. 1. Autoradiograph of the immunoprecipitation of JI-3l antigen. ~zSI-labelled human brain homogenate was immunoprecipitated with various antibodies i.e.,J1-31, normal mouse serum (NMS), antiGFAP, normal rabbit serum (NRS). The precipitates were separated on 10% SDS-PAGE under reducing conditions and developed after two weeks of exposure. 14C-labelled protein standards were used for molecular weight determinations.
The a b o v e suggestion that the J1-31 antigen (30K protein) is associated with cytoskeleton is s u p p o r t e d by the results from i m m u n o e l e c t r o n microscopy. An e x a m i n a t i o n of thin sections of the rat cerebellum (without further staining with heavy metal salts) in the electron m i c r o s c o p e showed that the i m m u n o s t a i n i n g was localized in association with cytoskeletal elements in certain cells (Fig. 2). These cells are
Marker Protein for Astrocytes
77
Fig. 2. Electron micrographs of the thin sections of rat cerebellum showing staining of the cytoskeleton (arrows) due to Mab J 1-31 followedby PAP technique of Sternberger (see Malhotra et al., 1984for details). astrocytes, based upon results of immunofluorescence microscopy discussed below. No staining was seen in thin sections of the cerebellum incubated with normal mouse serum as controls. Double label immunofluorescence microscopy on cryostat sections of human brain (autopsy samples from normal and MS patients) revealed that immunostaining due to Mab J 1-31 overlaps with that due to anti-GFA P rabbit antiserum. This staining for G F A P positive cells with Mab J1-31 was confirmed in double-labelled sections of the rat cerebellum (Fig. 3), in which structures corresponding to fibres of Bergmann (processes of astrocytic cells or Golgi epithelial cells (Palay and Chan-Palay, 1974)) are
78
Singh, Singh and Malhotra
Fig. 3. Double-label immunofluorescence micrographs ofcryostat sections of the rat cerebellum showing staining due to Mab J 1-31 (a and d) and ant{-GFAP antiserum (b and c) in the molecular layer (a and b) and granular layer (c and d) of the vermis. P, Purkinje cell layer. Arrows indicate comparable structures revealed by staining of the same section (a and b; c and d) with the two antibodies. In a and b, staining with Mab J 1-31 was performed first followed by anti-GFAP serum; c and d, staining with anti-GFAP antiserum (1:1000 dilution) was done followed by staining with the Mab Jl-31.
Marker Protein for Astrocytes
79
readily recognizable in the molecular layer of the vermis (Fig. 3a-b), The staining due to Mab J1-31 was not abolished when the sections were first stained for G F A P and then with the Mag J1-31 (Fig. 3c~11. This suggests that the antigenic determinants recognized by the polyclonal antiserum to G F A P are distinct from those recognized by the J 1-31 Mab. This conclusion is further supported by the data on ELISA in which J 131 binding to the brain homogenate is not inhibited by antiserum to G F A P ldata not shown~. Similar experiments using doubleqabel immunofluorescence microscopy on cryostat sections of the rat cerebellum showed thal J 1-31 and S 100 ~Van Eldik er al.. 1984"1 antigens are distinct. Rabbit antiserum to S100 (Dako) did not inhibit i mmunostaining due to J1-31 Mab. $100 polypeptide is a heterogeneous fraction of lowmolecular-weight calcium binding proteins which have been reported from CNS. peripheral nervous system and several non-neuronal tissues (Van Eldik e~ al.. 1984~. The cytoskeleton of astrocytes is characterized by G F A P IMW 50KJ which is a component of t he intermediate filaments (Wang er al.. 19841. Recently, vimentin IMW 58K) previously thought to be a constituent of fibroblastic filaments has been also demonstrated in the GFAP-containing intermediate filaments in human glioma cells (Wang er al.. 19841. Therefore, the intermediate filaments in astrocytes may represent polymers of more than single protein. The J 1-31 antigen, i.e. the 30K protein, appears to be associated with the cytoskeleton of astrocytes. It is of interest that a protein termed epinemin (44.5K) has been reported in association with vlmentin filaments in nonneural cells (Lawson. 19831. Based upon results of immunofluorescence staining. J 1-31 antigen is specific to the CNS. We have detected immunostaining in cerebellum, cerebrum, optic nerve of human Enormal and MS), rabbit and rat. No immunostaining was detected in cryostat sections of rat sciatic nerve, skeletal muscle, heart, liver, kidney and pancreas and in sections of D r o s o p h i l a . Taken roger her. data from each of the above techniques suggests that Mab J 1-31 is specific for CNS and this antigen is associated with the cytoskeleton in astrocytes. The antigemc determinants recognized by Mab J 1-31 therefore represent a new marker for astrocytes that may be useful in ontogenic and diagnosnc studies on these cells in the nervous system.
ACKNOWLEDGMENTS We thank K. G. Warren, MD, for providing human autopsy brain tissue; F. Wong, P. Cumming, S. D. Ross, M. Wong, D. Luethe for invaluable technical assistance; M. Boyer, R. Hunziker, and Dr. T. Wegmann (Dept. of Immunology) for help with some of the experiments; and T. K. Shnitka, MD, and Dr. V. Manickavel for providing some of the antisera and PAP conjugate. This work was supported by the MSI Foundation, N S E R C of Canada, M R C of Canada, and A H F M R . REFERENCES Bitnami, A., and Dahl, D. (1974). Nature 252:55-56. Bjorktund, H., Dahl, D., Olson, L., and Seiger, A. (1984). J. Neurosci. 4:978-988. Cole, G. J., and Glaser, L. (1984). Proc. Nail, Acad, Sci. USA 81:2260-2264.
80
Singh, Singh and Malhotra
Greenwood, F. C., Hunter, W. M., and Glover, J. S. (1963). Biochem. J. 89:114-123. Hawkes, R., Niday, E., and Matus, A. (1982). Proc. Natl. Acad. Sci. USA 79:2410-2414. Kessler, S. W. (1976). J. Immunol. 89:1482-1490. Laemmli, U. K. (1970). Nature 227:680~585. Lawson, D. (1983). J. Cell Biol. 97:1891-1905. Malfroy, B., Bakhit, C., Bloom, F. E., Sutcliffe, J. G., and Milner, R. J. (1985). Proc. Natl. Acad. Sci. USA 82:2009-2013. Malhotra, S. K., Wong, F., Cumming, P., Ross, S. D., Shnitka, T. K., Manickavel, V., Warren, K. G. and Jeffrey, V. (1984). Microbios Letters 26:151-157. Palay, S. L., and Chan-Palay, V. (1974). Cerebellar Cortex, Springer-Verlag, New York Raft, M. C., Fields, K. L., Hakomori, S.-L, Mirsky, R., Pruss, R. M., and Winter, J. (t979). Brain Res. 174:283306. Raft, M. C., Miller, R. H. and Nobel, M. (1983). Cold Spring Harbor Syrup. Vol. XLVIII, 569-572. Steinert, P. M., Jones, J. C. R., and Goldman, R. D. (1984). J. Cell Biol. 99:22 27. Sternberger, N. H., Itoyama, Y., Kies, M. W., and Webster, H. de F. (1978). Proc. Natl. Acad. Sci. USA 75:2521-2524. Van Eldik, L. J., Ehrenfried, B., and Jensen, R. A. (1984). Proc. Natl. Acad. Sci. USA 81:6034~5038. Wang, E., Cairncross, J. G., and Liem, R. K. H. (1984). Proc. Natl. Acad. Sci. USA 81:2102-2106.