Brain Research, 591 (1992) 1-7 © 1992 Elsevier Science Publishers B.V. All rights reserved 0006-8993/92/$05.00
BRES 18070
Research Reports
Polyclonal antibody localizes glia maturation factor fl-like immunoreactivity in neurons and glia B a i - R e n W a n g , A s g a r Z a h e e r a n d R a m o n Lim Department of Neurology, Division of Neurochemistry and Neurobiology, Unicersityof Iowa Collegeof Medicine and VeteransAffairs Medical Center, Iowa City, IA 52242 (USA) (Accepted 14 April 1992)
Key words: Maturation factor; Growth factor; Immunolocalization; Endogenous glia maturation factor-/3; Neuronal protein; Glial protein
A rabbit polyclonai antibody (91-01) was raised against recombinant human glia maturation factor /3 (r-hGMF-/3). The antibody did not cross-react with a number of other growth factors on ELISA test. When compared with the monoclonai antibody G2-09 previously obtained, 91-01 immunoblotted the same protein band in rat brain extract. However, unlike G2-09 which immunostained only astrocytes and Bergmann glia, 91-01 stained neurons as well. Many but not all neurons in the central and peripheral nervous system were positive for GMF-/3. The larger cell population stained by the polyclonal antibody was most likely due to its increased sensitivity, although other explanations are possible. The presence of GMF-/3-1ike immunoreactivity in both neurons and glia raises the possibility of a wider range of cell-cell interaction than was previously considered.
INTRODUCTION Olia maturation factor /3 (GMF-//) is a 17 kDa acidic protein isolated from the bovine brain ~. The pure protein exhibits a strong antimitogenic activity on cultured tumor cells s and appears to play a role in neural differentiation and regeneration. For example, OMF-/3 inhibits the proliferation of N18 neuroblastoma cells and stimulates their neurite outgrowth, along with an increase in neurofilamenP. When applied to the cerebral cortex following injury, GMF-/3 promotes the appearance of large, neurofilament-rich neurons 7. GMF-/3 induces glial fibrillary acidic protein (GFAP) in a medulloblastoma cell line (Berger, Keles, Silber and Lim, unpublished data). In the sciatic nerve, the expression of GMF-//in Schwann cells is under axonal regulation ~. We have sequenced the bovine GMF-//and showed that it has no significant homology to other proteins including the known growth factors and tumor suppressors t°. GMF-// is highly conserved, having identical amino acid sequence between human and cattle. The
human GMF-/3 has now been cloned and expressed in E. coil 2. As part of our effort to define the role of GMF-/3 in the nervous system, we previously localized the protein with a monoclonal antibody designated G2-09 which demonstrated the presence of GMF-/3-1ike immunoreactivity in spinal cord astrocytes and cerebellar Bergmann glia5. With the availability of recombinant human GMF-/~ (r.hGMF-/3), we have recently raised polyclonal antibodies against the recombinant protein and re-examined the localization of endogenous GMF/3. In this article we demonstrate the immunostaining of both neurons and glia by the rabbit polyclonal antibody and provide an overview of the distribution of endogenous GMF-/3 in the rat brain. MATERIALS AND METHODS
Sources of growth factors Recombinant human GMF-/3 (r-hGMF-/3)was isolated from E.
coli as previously described 2'11. The pure protein was used as immunogen and as antigen for the immunoassays. Acidic and basic fibroblast growth factors were gifts of M. Jaye and P.A. Maher, respectively. Platelet-derived growth factor was provided by D.J.
Correspondence: R. Lim, Department of Neurology, University of Iowa College of Medicine, Iowa City, IA 52242, USA. Fax: (1) (319) 356-4505.
Hicklin. All other growth factors were purchased from commercial sources.
development was terminated after 5 min by repeated washings with water.
Sources of antibodies
Enzyme-li, ked immunosorbent assay (ELISA)
The mouse monoclonal antibody G2-09, an lgG2b, was previously raised against a partially purified bovine GMF 4'5. The polyclonal antibody 91-01 was raised in a rabbit against r-hGMF-/3 as follows. The rabbit was injected subcutaneously in the back with 100 # g of the pure protein in 1 ml of a mixture consisting of 0.5 ml complete Freund's adjuvant and 0.5 ml of PBS (10 mM sodium phosphate, 0.15 NaCI, pH 7.4). Four weeks later, a booster was given intravenously with the same amount of protein in the same carrier where the complete Freund's adjuvant was replaced by the incomplete adjuvant. Tl~e animal was bled a week later by heart puncture. The blood was allowed to clot at room temperature for 2 h and then left overnight at 4°C. The antiserum was collected by centrifugation a~ 2,500 x g for 30 rain in the cold and subsequently stored at - 2 0 °.
Two-fold serially diluted antigen in 50 /~! of 50 mM sodium bicarbonate buffer, pH 9.6, was added to each well in a 96-well flat bottom micro-titration plate and incubated at 4°C overnight. The unoccupied binding sites in the wells were blocked by incubating with 3% (w/v) bovine serum albumin in PBS for 2 h at room temperature. Incubation with the primary antibody (91-01 or G2-09, at 1:500 dilution), was carried out at room temperature for 2 h. After three washings with 200 #l of PBS containing 0,05% Tween-20 (PBST), the plate was incubated for 2 additional hours with peroxidase-conjugated second antibody, goat anti-rabbit or anti-mouse igG. at ! :5,000 dilution. Alter washing three times with PBST and once with PBS, color reaction was developed with 2,2'-azinobis (3-ethylbenzothiazoline-6-sulfonic acid), diammonium salt (ABTS). The reaction was stopped after 5 min by adding 50/zi of 5% SDS solution in water and read at 415 nm using a Bio-Tek microplate reader.
Affinity purification and adsorptio, of amibodies Antibodies were affinity-purified from hybridoma conditioned medium (for G2-09) and from rabbit antiserum (for 91-01) as follows. The antibody sample was applied to a ! ml Protein A-sepharose (Pierce) gel, packed in a disposable polypropylene column (Econo-column. BioRad), which had been extensively washed and pre-equilibrated with PBS. The sample was recycled 3-4 times at room temperature. After washing the column with 10 column volumes of PBS, the bound material was eluted with 0.2 M glycine-HCI, pH 2.8. containing 0.5 M NaCI. Fractions (l ml) were collected and immediately neutralized with 50/zl of I M Tris. The pooled fractions were buffer-exchanged with PBS and concentrated to 0.25 mg protein/mi with Centricon-10 microconcentrator (Amicon), and used for all the immunological experiments described in this paper. When immunoadsorption was called for, the affinity purified anti-GMF-,8 antibody was incubated with 100-fold excess of pure r-hGMF-,8 at 4°C overnight, and subsequently centrifuge for 15 min at 12,000x ,¢.
t~,stem blot analy.ri,~' Brain extract from an adult rat was obtained by centrifuging a 25¢~ (w/v) homogenate in PBS at ll)l),(}tl0x g for I h (two times) and collecting the supernatant. Protein concentration was estimated by the bicinchoninic acid (BCA) method as recommended by Pierce Chemicals. Protein san)pies (40 ~g) were dissolved in a sample buffer containing 0.1 M dithiothreitol, 2% SDS, 15% glycerol, 2 mM phenylmethanesulfonyl fluoride, 2 mM EDTA, I mM N-ethylmaleimide, ! mM iodoacetic acid, 75 mM Tris.HCI (pH 6,8), and 0.l}01% Bromophenol blue, and boiled for 3 min. Polyacrylamide gel electrophoresis in SDS was carried out by the method of Laemmli -~, Protein samples were applied to a i-mm-thick discontinuous gel consisting of 4% stacking and 15% resolving gel made in the MiniProtean ll slab cell (BioRad Laboratories, Richmond, CA). Electrophoresis was carried out at a constant voltage of 150 V for stacking and 80 V for separation. The gels were fixed in a solution of 5()~[ methanol/12% acetic acid for 30 rain and then stained with 0.1¢q Coomassie blue R-250 in the same solution. I-or immunoblotting, the unfixed, unstained protein bands were transferred onto nitrocellulose membrane at a constant voltaEe of 100 V for ! h with cooling, using a Mini Trans-BIot Cell (BioRad,. The buffer used was 25 mM Tris, 92 mM glycine, pH 8.3, with 20% (v/v) methanol, After electroblotting, the nitrocellulose membrane was gently agitated in a blocking solution of TBS/BSA (20 mM Tris-HCI and 0.15 M NaCI, pH 7.4/3% bovine serum albumin) for ! h at room temperature. lmmunoreaction was carried out on the membrane by first reacting with the primary anti-GMF.~ antibody in TTBS (0,2% Tween-20 in TBS) overnight at 4°C. Dilution of primary antibodies was made at 1:250. The membrane was washed three times (10 rain each) with "FrBS and subsequently incubated for 2 h at room temperature with pcroxidase-conjugated second antibody (goat anti-rabbit or antimouse leG, as appropriate, diluted at !: i,000), After washing the membrane, three times with TTBS and once with TBS, it was ima:,:~¢d in the substrate solution (30 mg 2-chloronaphthol, 10 ml meti~anol, 50 ml TBS and 30 ~,1 of 30% hydrogen peroxide). Color
lmmtmocytochernistry Female Sprague-Dawley rats weighing about 200 g were anesthetized by intraperitoneal injection of sodium pentobarbital (60 mg/kg) and perfused through the heart with I00 ml of saline followed by 500 ml of fixative over 60 min. The fixative consisted of 2% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer, pH 7.4. Brains and spinal cords were removed and postfixed for 3 hr in the same fixative. Tissue blocks were then stored in 30% sucrose in 0.1 M phosphate buffer, pH 7.4, for 2-3 days until they sank. Sections, 50 p,m thick, were cut with a freezing microtome. Immunocytochemistry for GMF-/3 was performed using the ABC (avidin-biotin complex) method. Before reaction, the floating sections were treated for 1 h at room temperature with 0.1 M phosphate buffer, pH 7.4, containing 3% normal goat serum (NGS), 1% bovine serum albumin (BSA) and 0.3% Triton X-100. The sections were then incubated, with shaking, at 4°C for 2 days with the affinity-purlfied rabbit antiserum 91-01 at a d~iution of 1:1,000-2,000. The sections were rinsed and then treated at room temperature for 0.5 h with 0.l M phosphate buffer, pH 7.4, containing I% NGS, I% BSA and 0,3% Triton X.I{I[), The sections were incubated for 1,5 h at room temperature with biotinylated goat anti-rabbit l e g (Sigma Chemical) at 1:250 dilution, After another rinse, the sections were incubated for 1,5 h at room temperature with horseradish peroxidase (HRP)-streptavidin (Amersham International) at 1:250 dilution, The
A 1.o.
B
.
..~.g~t..-o ./I-=
-
,
, ,a~, m i l l
1 ,,1,
i /
o.."
/
I
GMF-~
F
J .FGF
t
to
too
500
~ z o s ~ (us/wen) Fig. I. Comparison of rabbit polyclonal antibody 91-01 with mouse monoclonal antibody G2-09 by ELISA, using recombinant human GMF-/3 (r-hGMF-B) as antigen, l , 91-01; o, G2-09. Acidic fibroblast growth factor (aFGF) was used as a negative control antigen. Inset, immunoblotting of r-hGMF-/3 using (A) 91-01 and (B) G2-09. Lane I, prestained molecular weight standards (top to bottom: 43 kDa, 29 kDa: 18.4 kDa, 14.3 kDa); lane 2, pure r-hGMF4/(50 ng per slot),
B
A
DE
,.,
m
~
~.~,.
~,~
i~~
12
1 2 12
12-12
Fig. 2. Comparison of rabbit polyclonal antibody 91-01 with mouse monoclonal antibody G2-09 by immunoblotting of rat brain extract. A: Coomassie blue-stained gel pattern of brain extract. B: Western blot of brain extract using 91-01. C: Western blot of brain extract using adsorbed 91-01. D: Western blot of brain extract using (32-09. E: Western blot of brain extract using adsorbed G2-09. Lane I, brain extract (40/zg protein per slot); lane 2, prestained molecular weight standards (top to bottom: 29 kDa, 18.4 kDa, 14.3 kDa, 6.2 kDa). •
~i~~i~i~¸:i ...........
' ~i!!. . . . . . . .
sections were rinsed again and the color was developed with 0.02% diaminobenzidine (DAB) and 0.005% H aO z in 0.05 M Tris buffer, pH 7.6, for 10-20 rain. Each rinsing step was conducted with 0.02 ~4 phosphate buffered saline (0.15 M NaCl) for 3 changes of 10 rain each. The first antibody, second antibody, and HRP-strepta~idin were all prepared with a solution of 0.1 M phosphate buffer, pH 7.4, containing 1% NGS, 1% BSA and 0.3% Triton X-100. Finally, the sections were dehydrated and covered with Permount for light microscopy. RESULTS
Characterization of polyclonal antibody R a b b i t polyclonal a n t i b o d y 9 1 - 0 1 , d e v e l o p e d a g a i n s t p u r i f i e d r-hGMF-/3, was c o m p a r e d with the m o u s e m o n o c l o n a l a n t i b o d y G 2 - 0 9 previously p r o d u c e d in this l a b o r a t o r y . Fig. 1 shows that, with E L I S A study, the two a n t i b o d i e s e x h i b i t e d identical r e a c t i o n curve t o w a r d a s a m p l e of r-hGMF-/3. W h e n tested a g a i n s t b o t h a n t i b o d i e s , the following growth factors d e m o n -
,~/:i~~~i/~'~,~, !!
¢
~
~ ~*~llg~
,.
.....
!B ....
IC* Fig. 3. lmmunostaining for GMF-/3 in spinal cord. A: white matter showing positively stained astrocytes. B: gray matter (anterior horn) showing positively stained neurons. Bars = 50 ,~m,
Fig. 4. Immunostaining for GMF-/3 in cerebellum. A: cerebellar cortex showing intensely stained Purkinje cell bodies and Bergmann glial processes, the latter presenting as parallel lines extending from the level of the Purkinje cell layer traversing through the molecular layer toward the surface of the cortex. B: negative control for 'A' using adsorbed antibody. C: a cerebellar nucleus showing GMF-/3positive neurons. D: cerebellar white matter showing positively stained astrocytes. Bars -- 50 ~m.
strated less than 1% cross-reactivity with respect to GMF-/3: acidic and basic fibroblast growth factors (FGF), epidermal growth factor (EGF), nerve growth factor (NGF), transforming growth factor a and /3 (TGF-a; TGF-/3), interleukin lt~ (IL-la), interleukin 2 (IL-2), tumor necrosis factor (TNF), platelet-derived growth factor (PDGF-AA), insulin, and insulin-like growth factcr II (IGF-II). (For clarity of presentation, only the reaction curve of acidic FGF is included in Fig. 1.) On immunoblotting of rat brain extract (Fig. 2), both 91-01 and G2-09 recognized a single protein band having a molecular weight of authentic GMF-fl.
Fig. 6. Immunostaining for GMF-/3 in the lateral vestibular (Deiters) nucleus. Neurons of various sizes, including their processes, were stained. Bar = 50/~m,
Adsorption of both antibodies with r-hGMF-/3 completely abolished the immunoreactivity.
lmmunostaining of rat na~ral tissue with polyclonai antibody
Fig, 5. lmmunostaining for GMF-/3 in the midbrain. A: oculomotor nuclei showing positively stained neurons. B: red nucleus showing positively stained neurons. C: substantia nigra showing positively stained neurons in pars compacta (upper right half) and pars reticulata (lower left half), Bar --- 50 ~m.
Neural tissue from adult rat was processed and immunostained with rabbit polycional antibody 91-01. Transverse sections of the spinal cord showed strongly positive staining of astrocytos having star-shaped cell bodies with elongated processes which are in parallel formation and perpendicular to the surface of the cord (Fig. 3A). In addition, motor neurons in the anterior horn also stained positively for GMF-/3 (Fig. 3B). lmmunocytoehemistry of the cerebellar cortex (Fig. 4A) showed strong staining of the Purkinje cell perikarya and Bergmann gliai processes. In contrast, the dendritic arborizations of the Purkinje ceils were not stained. Most of the granule cells were negative. In the deeper structures of the cerebellum, neurons of the cerebellar nuclei (Fig. 4C) and astrocytes in the cerebellar white matter (Fig. 4D) were clearly positive for GMF-/3. The astrocytes exhibited typical stellate morphology. Many neurons in the midbrain were GMF-,8-positire, including those in the oculomotor nucleus, the red nucleus and substantia nigra (Fig. 5). In the medulla, numerous neurons in the lateral vestibular nucleus, including their long processes, showed positive-staining for GMF-/3 (Fig. 6).
In the basal ganglia (Fig. 7), a distinctive contrast was seen between the caudate nucleus and globus pallidus. While in the former only occasional neurons were stained, in the latter the majority of neurons were GMF-/3 positive. In the sensory/motor cortex (Fig. 8A) and entorhihal cortex (Fig. 8C), some GMF-/3-containing neurons were scattered on a background studded with GMF-/3positive astrocytes. There were relatively more immunoreactive neurons in the deep layers of the cortex compared to superficial layers. Likewise, in the hippocampal formation, GMF-//-positive astrocytes were numerous, while GMF-/3-positive neurons were few. In the hippocamptls proper, pyramidal cells were all unstained whereas the scattered smaller sized neurons were intensely stained. In the dentate gyrus, most of the granule cells were u_,stained except for a single row of neurons in the bottom of the granule cell layer which were highly immunoreactive (Fig. 8D). GMF-//-like immunoreactivity was also found in the dorsal root ganglia (Fig. 9). Here, both the perikarya of sensory neurons and the surrounding capsule cells were stained. The intensity of staining in the capsule cells was particularly strong. DISCUSSION The current work was undertaken to compare a recently obtained polyclonal antibody (rabbit serum
91-01) derived from recombinant GMF-//with a monoclonal antibody (G2-09) previously raised against natural GMF. The two antibodies exhibited similar immunochemicai specificity: they both recognized the same GMF-/3 protein, whether in the pure state or from a crude brain extract, and they both showed no cross-reactivity with other growth factor proteins. However, despite these similarities, the two antibodies gave different immunocytochemical reactions. Thus, while the monoclonal antibody stained only astrocytes in the spinal cord and Bergmann glia in the cerebellum 5, the polyclonal antibody stained also astrocytes in other areas of the nervous system and neurons as well. There are several explanations for the larger cell population stained by the polyclonal antibody. The simplest explanation is that the polyclonal antibody was more sensitive because it recognized multiple epitopes, as opposed to the monoclonal antibody which recognized a single ¢pitope. Thus, whereas the monoclonal antibody stained only cells that contained the highest level of GMF-/3, the polyclonal antibody also stained cells that contained lower levels of the protein. Another attractive explanation is that GMF-~ may belong to a family of related proteins, so that while the monoclonal antibody G2-09 recognized an epitope unique to GMF-~, the polyclonal antibody 91-01 also recognized epitopes that were in common with other members of the family. However, we have not been able to support this by observing extra protein bands,
Fig. 7. lmmunostainingfor GMF-// in basal ganglia.A: caudate nucleusshowingselectivestainingof a few neurons. B: globuspallidus showing positive stainingof numerous neurons. Bar = 50 ~m.
with either 1-D or 2-D gels, that immunoblotted the polyclonal antibody. A more definitive answer must await the isolation of monoclonal antibodies that can stain both astrocytes and neurons. It should be noted that not all neurons were positive for GMF-/], and not all the positive neurons were stained with equal intensity. Also, in both neurons and astrocytes, only the cytoplasm and not the nucleus was stained, indicating that GMF-/3 is a cytoplasmic rather than nuclear protein. In astrocytes, both cell processes and the perinuc[ear region were equally stained. This was not true for neurons, as staining of dendritic processes did not always accompany staining of the perikaryon (as in Purkinje cells). On the other hand, oligodendroglia and white matter were consistently negative for endogenous GMF-/3. No attempt was made
Fig. 9. immunostaining for GMF-# in spinal ganglia (L,O, showing positively stained neurons and capsule cells surrounding neuronal pcrikarya. Bar =50 #m. Inset: details with I.Sxfurther magnification.
•
u
t
Fig. 8. Immunostaining for GMF-# in cerebral cortex, A: sensory/motor cortex showing positive staining of many astrocytes and a few neurons. B: negative control for 'A" using adsorbed antibody. C: entorhinal cortex showing positive staining of numerous astrocytes and occasional neurons, D: dentate gyrus of hippocampal fi)rmation showing many positively stained astrocytes in the molecular layer (upper i~rtion), some positively stained neurons in the granular layer (middle zone), and a few positively stained neurons in the hilar region (lower portion), Bar = 50/~m,
in this study to immunostain organs outside the nervous system because, as we have demonstrated with mRNA hybridization, GMF-# is synthesized mainly in the nervous system (Zaheer, Fink and Lira, unpublished data). The demonstration of endogenous GMF-/3 in neurons as well as astrocytes suggests a wider range of cell-cell interaction than was previously thought. Thus, GMF-# could mediate communication among astrocytes, among neurons, and between neurons and astrocytes. The coexistence in the brain of a growth suppressor such as GMF-/3 and growth promoters such as the fibrobast growth factors, etc., implies the importance of checks and balances in cellular regulation that could be crucial to the development, degeneration and regeneration of the nervous system. Acktiowledgemet~t. We thank Brian D, Fink, Joseph C, Behr and Jeffrey Hwang for technical assistance, Dr. John E. Butler for help with rabbit immunization, and Dr. H,K. Kultas-llinsky for advice and discussions, This work was supported by the following grants to R.L.: Veterans Affairs Merit Review Award, National Science Foundation Grant BNS-8917665, and Diabetes-Endocrinology Center Grant DK25295,
REFERENCES 1 Bosch, E.P., Zhong, W. and Lim, R., Axonal signals regulate expression of glia maturation factor /3 in Schwann cells: an immunohistochemical study of injured sciatic nerves and cultured Schwann cells, J. Neurosci., 9 (1989) 3690-3698. 2 Kaplan, R., Zaheer, A., Jaye, M. and Lira, R., Molecular cloning and expression of biologically active human glia maturation factor fl, J. Neurochem., 57 (1991) 483-490. 3 Laemmli, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature, 227 (1970) 680-685. 4 Lira, R., Miller, J.F., Hicklin, D.J. and Andresen, A.A., Purification of bovine glia maturation factor and characterization with monoclonal antibody, Biochenffstry, 24 (1985) 8070-8074. 5 Lira, R., Hicklin, D.J., Miller, J.F., Williams, T.H. and Crabtree, J.B., Distribution of immunoreactive glia maturation factor-like molecule in organs and tissues, Dec. Brabz Res., 33 (1987) 93-100.
6 Lim, R., Miller, J.F. and Zaheer, A., Purification and characterization of glia maturation factor fl: a growth regulator for neurons and glia, Proc. Natl. Acad. Sci. USA, 86 (1989) 3901-3905. 7 Lira, R. and Huang, L., Gila maturation factor/3 promotes the appearance of large neurofilament-rich neurons in injured rat brains, Brain Res., 504 (1989) 154-158. 8 Lira, R., Zhong, W. and Zaheer, A., Antiproliferative functior~ of glia maturation factor/3, Cell Regul., 1 (1990) 741-746. 9 Lim, R., Liu, Y. and Zaheer, A., Glia maturation factor fl regulates the growth of N18 neuroblastoma cells, Dec. Biol., 137 (1990) 444-450. 10 Lim, R., Zaheer, A. and Lane, W.S., Complete amino acid sequence of bovine glia maturation factor fl, Proc. Natl. Acad. Sc£ USA, 87 (1990) 5233-5237. 11 Lim, R. and Zaheer, A., Preparation of glia maturation factor fl, Methods Neurosci., 6 (1991) 321-337.
BRES 18070
Research Reports
Polyclonal antibody localizes glia maturation factor fl-like immunoreactivity in neurons and glia B a i - R e n W a n g , A s g a r Z a h e e r a n d R a m o n Lim Department of Neurology, Division of Neurochemistry and Neurobiology, Unicersityof Iowa Collegeof Medicine and VeteransAffairs Medical Center, Iowa City, IA 52242 (USA) (Accepted 14 April 1992)
Key words: Maturation factor; Growth factor; Immunolocalization; Endogenous glia maturation factor-/3; Neuronal protein; Glial protein
A rabbit polyclonai antibody (91-01) was raised against recombinant human glia maturation factor /3 (r-hGMF-/3). The antibody did not cross-react with a number of other growth factors on ELISA test. When compared with the monoclonai antibody G2-09 previously obtained, 91-01 immunoblotted the same protein band in rat brain extract. However, unlike G2-09 which immunostained only astrocytes and Bergmann glia, 91-01 stained neurons as well. Many but not all neurons in the central and peripheral nervous system were positive for GMF-/3. The larger cell population stained by the polyclonal antibody was most likely due to its increased sensitivity, although other explanations are possible. The presence of GMF-/3-1ike immunoreactivity in both neurons and glia raises the possibility of a wider range of cell-cell interaction than was previously considered.
INTRODUCTION Olia maturation factor /3 (GMF-//) is a 17 kDa acidic protein isolated from the bovine brain ~. The pure protein exhibits a strong antimitogenic activity on cultured tumor cells s and appears to play a role in neural differentiation and regeneration. For example, OMF-/3 inhibits the proliferation of N18 neuroblastoma cells and stimulates their neurite outgrowth, along with an increase in neurofilamenP. When applied to the cerebral cortex following injury, GMF-/3 promotes the appearance of large, neurofilament-rich neurons 7. GMF-/3 induces glial fibrillary acidic protein (GFAP) in a medulloblastoma cell line (Berger, Keles, Silber and Lim, unpublished data). In the sciatic nerve, the expression of GMF-//in Schwann cells is under axonal regulation ~. We have sequenced the bovine GMF-//and showed that it has no significant homology to other proteins including the known growth factors and tumor suppressors t°. GMF-// is highly conserved, having identical amino acid sequence between human and cattle. The
human GMF-/3 has now been cloned and expressed in E. coil 2. As part of our effort to define the role of GMF-/3 in the nervous system, we previously localized the protein with a monoclonal antibody designated G2-09 which demonstrated the presence of GMF-/3-1ike immunoreactivity in spinal cord astrocytes and cerebellar Bergmann glia5. With the availability of recombinant human GMF-/~ (r.hGMF-/3), we have recently raised polyclonal antibodies against the recombinant protein and re-examined the localization of endogenous GMF/3. In this article we demonstrate the immunostaining of both neurons and glia by the rabbit polyclonal antibody and provide an overview of the distribution of endogenous GMF-/3 in the rat brain. MATERIALS AND METHODS
Sources of growth factors Recombinant human GMF-/3 (r-hGMF-/3)was isolated from E.
coli as previously described 2'11. The pure protein was used as immunogen and as antigen for the immunoassays. Acidic and basic fibroblast growth factors were gifts of M. Jaye and P.A. Maher, respectively. Platelet-derived growth factor was provided by D.J.
Correspondence: R. Lim, Department of Neurology, University of Iowa College of Medicine, Iowa City, IA 52242, USA. Fax: (1) (319) 356-4505.
Hicklin. All other growth factors were purchased from commercial sources.
development was terminated after 5 min by repeated washings with water.
Sources of antibodies
Enzyme-li, ked immunosorbent assay (ELISA)
The mouse monoclonal antibody G2-09, an lgG2b, was previously raised against a partially purified bovine GMF 4'5. The polyclonal antibody 91-01 was raised in a rabbit against r-hGMF-/3 as follows. The rabbit was injected subcutaneously in the back with 100 # g of the pure protein in 1 ml of a mixture consisting of 0.5 ml complete Freund's adjuvant and 0.5 ml of PBS (10 mM sodium phosphate, 0.15 NaCI, pH 7.4). Four weeks later, a booster was given intravenously with the same amount of protein in the same carrier where the complete Freund's adjuvant was replaced by the incomplete adjuvant. Tl~e animal was bled a week later by heart puncture. The blood was allowed to clot at room temperature for 2 h and then left overnight at 4°C. The antiserum was collected by centrifugation a~ 2,500 x g for 30 rain in the cold and subsequently stored at - 2 0 °.
Two-fold serially diluted antigen in 50 /~! of 50 mM sodium bicarbonate buffer, pH 9.6, was added to each well in a 96-well flat bottom micro-titration plate and incubated at 4°C overnight. The unoccupied binding sites in the wells were blocked by incubating with 3% (w/v) bovine serum albumin in PBS for 2 h at room temperature. Incubation with the primary antibody (91-01 or G2-09, at 1:500 dilution), was carried out at room temperature for 2 h. After three washings with 200 #l of PBS containing 0,05% Tween-20 (PBST), the plate was incubated for 2 additional hours with peroxidase-conjugated second antibody, goat anti-rabbit or anti-mouse igG. at ! :5,000 dilution. Alter washing three times with PBST and once with PBS, color reaction was developed with 2,2'-azinobis (3-ethylbenzothiazoline-6-sulfonic acid), diammonium salt (ABTS). The reaction was stopped after 5 min by adding 50/zi of 5% SDS solution in water and read at 415 nm using a Bio-Tek microplate reader.
Affinity purification and adsorptio, of amibodies Antibodies were affinity-purified from hybridoma conditioned medium (for G2-09) and from rabbit antiserum (for 91-01) as follows. The antibody sample was applied to a ! ml Protein A-sepharose (Pierce) gel, packed in a disposable polypropylene column (Econo-column. BioRad), which had been extensively washed and pre-equilibrated with PBS. The sample was recycled 3-4 times at room temperature. After washing the column with 10 column volumes of PBS, the bound material was eluted with 0.2 M glycine-HCI, pH 2.8. containing 0.5 M NaCI. Fractions (l ml) were collected and immediately neutralized with 50/zl of I M Tris. The pooled fractions were buffer-exchanged with PBS and concentrated to 0.25 mg protein/mi with Centricon-10 microconcentrator (Amicon), and used for all the immunological experiments described in this paper. When immunoadsorption was called for, the affinity purified anti-GMF-,8 antibody was incubated with 100-fold excess of pure r-hGMF-,8 at 4°C overnight, and subsequently centrifuge for 15 min at 12,000x ,¢.
t~,stem blot analy.ri,~' Brain extract from an adult rat was obtained by centrifuging a 25¢~ (w/v) homogenate in PBS at ll)l),(}tl0x g for I h (two times) and collecting the supernatant. Protein concentration was estimated by the bicinchoninic acid (BCA) method as recommended by Pierce Chemicals. Protein san)pies (40 ~g) were dissolved in a sample buffer containing 0.1 M dithiothreitol, 2% SDS, 15% glycerol, 2 mM phenylmethanesulfonyl fluoride, 2 mM EDTA, I mM N-ethylmaleimide, ! mM iodoacetic acid, 75 mM Tris.HCI (pH 6,8), and 0.l}01% Bromophenol blue, and boiled for 3 min. Polyacrylamide gel electrophoresis in SDS was carried out by the method of Laemmli -~, Protein samples were applied to a i-mm-thick discontinuous gel consisting of 4% stacking and 15% resolving gel made in the MiniProtean ll slab cell (BioRad Laboratories, Richmond, CA). Electrophoresis was carried out at a constant voltage of 150 V for stacking and 80 V for separation. The gels were fixed in a solution of 5()~[ methanol/12% acetic acid for 30 rain and then stained with 0.1¢q Coomassie blue R-250 in the same solution. I-or immunoblotting, the unfixed, unstained protein bands were transferred onto nitrocellulose membrane at a constant voltaEe of 100 V for ! h with cooling, using a Mini Trans-BIot Cell (BioRad,. The buffer used was 25 mM Tris, 92 mM glycine, pH 8.3, with 20% (v/v) methanol, After electroblotting, the nitrocellulose membrane was gently agitated in a blocking solution of TBS/BSA (20 mM Tris-HCI and 0.15 M NaCI, pH 7.4/3% bovine serum albumin) for ! h at room temperature. lmmunoreaction was carried out on the membrane by first reacting with the primary anti-GMF.~ antibody in TTBS (0,2% Tween-20 in TBS) overnight at 4°C. Dilution of primary antibodies was made at 1:250. The membrane was washed three times (10 rain each) with "FrBS and subsequently incubated for 2 h at room temperature with pcroxidase-conjugated second antibody (goat anti-rabbit or antimouse leG, as appropriate, diluted at !: i,000), After washing the membrane, three times with TTBS and once with TBS, it was ima:,:~¢d in the substrate solution (30 mg 2-chloronaphthol, 10 ml meti~anol, 50 ml TBS and 30 ~,1 of 30% hydrogen peroxide). Color
lmmtmocytochernistry Female Sprague-Dawley rats weighing about 200 g were anesthetized by intraperitoneal injection of sodium pentobarbital (60 mg/kg) and perfused through the heart with I00 ml of saline followed by 500 ml of fixative over 60 min. The fixative consisted of 2% paraformaldehyde and 0.2% picric acid in 0.1 M phosphate buffer, pH 7.4. Brains and spinal cords were removed and postfixed for 3 hr in the same fixative. Tissue blocks were then stored in 30% sucrose in 0.1 M phosphate buffer, pH 7.4, for 2-3 days until they sank. Sections, 50 p,m thick, were cut with a freezing microtome. Immunocytochemistry for GMF-/3 was performed using the ABC (avidin-biotin complex) method. Before reaction, the floating sections were treated for 1 h at room temperature with 0.1 M phosphate buffer, pH 7.4, containing 3% normal goat serum (NGS), 1% bovine serum albumin (BSA) and 0.3% Triton X-100. The sections were then incubated, with shaking, at 4°C for 2 days with the affinity-purlfied rabbit antiserum 91-01 at a d~iution of 1:1,000-2,000. The sections were rinsed and then treated at room temperature for 0.5 h with 0.l M phosphate buffer, pH 7.4, containing I% NGS, I% BSA and 0,3% Triton X.I{I[), The sections were incubated for 1,5 h at room temperature with biotinylated goat anti-rabbit l e g (Sigma Chemical) at 1:250 dilution, After another rinse, the sections were incubated for 1,5 h at room temperature with horseradish peroxidase (HRP)-streptavidin (Amersham International) at 1:250 dilution, The
A 1.o.
B
.
..~.g~t..-o ./I-=
-
,
, ,a~, m i l l
1 ,,1,
i /
o.."
/
I
GMF-~
F
J .FGF
t
to
too
500
~ z o s ~ (us/wen) Fig. I. Comparison of rabbit polyclonal antibody 91-01 with mouse monoclonal antibody G2-09 by ELISA, using recombinant human GMF-/3 (r-hGMF-B) as antigen, l , 91-01; o, G2-09. Acidic fibroblast growth factor (aFGF) was used as a negative control antigen. Inset, immunoblotting of r-hGMF-/3 using (A) 91-01 and (B) G2-09. Lane I, prestained molecular weight standards (top to bottom: 43 kDa, 29 kDa: 18.4 kDa, 14.3 kDa); lane 2, pure r-hGMF4/(50 ng per slot),
B
A
DE
,.,
m
~
~.~,.
~,~
i~~
12
1 2 12
12-12
Fig. 2. Comparison of rabbit polyclonal antibody 91-01 with mouse monoclonal antibody G2-09 by immunoblotting of rat brain extract. A: Coomassie blue-stained gel pattern of brain extract. B: Western blot of brain extract using 91-01. C: Western blot of brain extract using adsorbed 91-01. D: Western blot of brain extract using (32-09. E: Western blot of brain extract using adsorbed G2-09. Lane I, brain extract (40/zg protein per slot); lane 2, prestained molecular weight standards (top to bottom: 29 kDa, 18.4 kDa, 14.3 kDa, 6.2 kDa). •
~i~~i~i~¸:i ...........
' ~i!!. . . . . . . .
sections were rinsed again and the color was developed with 0.02% diaminobenzidine (DAB) and 0.005% H aO z in 0.05 M Tris buffer, pH 7.6, for 10-20 rain. Each rinsing step was conducted with 0.02 ~4 phosphate buffered saline (0.15 M NaCl) for 3 changes of 10 rain each. The first antibody, second antibody, and HRP-strepta~idin were all prepared with a solution of 0.1 M phosphate buffer, pH 7.4, containing 1% NGS, 1% BSA and 0.3% Triton X-100. Finally, the sections were dehydrated and covered with Permount for light microscopy. RESULTS
Characterization of polyclonal antibody R a b b i t polyclonal a n t i b o d y 9 1 - 0 1 , d e v e l o p e d a g a i n s t p u r i f i e d r-hGMF-/3, was c o m p a r e d with the m o u s e m o n o c l o n a l a n t i b o d y G 2 - 0 9 previously p r o d u c e d in this l a b o r a t o r y . Fig. 1 shows that, with E L I S A study, the two a n t i b o d i e s e x h i b i t e d identical r e a c t i o n curve t o w a r d a s a m p l e of r-hGMF-/3. W h e n tested a g a i n s t b o t h a n t i b o d i e s , the following growth factors d e m o n -
,~/:i~~~i/~'~,~, !!
¢
~
~ ~*~llg~
,.
.....
!B ....
IC* Fig. 3. lmmunostaining for GMF-/3 in spinal cord. A: white matter showing positively stained astrocytes. B: gray matter (anterior horn) showing positively stained neurons. Bars = 50 ,~m,
Fig. 4. Immunostaining for GMF-/3 in cerebellum. A: cerebellar cortex showing intensely stained Purkinje cell bodies and Bergmann glial processes, the latter presenting as parallel lines extending from the level of the Purkinje cell layer traversing through the molecular layer toward the surface of the cortex. B: negative control for 'A' using adsorbed antibody. C: a cerebellar nucleus showing GMF-/3positive neurons. D: cerebellar white matter showing positively stained astrocytes. Bars -- 50 ~m.
strated less than 1% cross-reactivity with respect to GMF-/3: acidic and basic fibroblast growth factors (FGF), epidermal growth factor (EGF), nerve growth factor (NGF), transforming growth factor a and /3 (TGF-a; TGF-/3), interleukin lt~ (IL-la), interleukin 2 (IL-2), tumor necrosis factor (TNF), platelet-derived growth factor (PDGF-AA), insulin, and insulin-like growth factcr II (IGF-II). (For clarity of presentation, only the reaction curve of acidic FGF is included in Fig. 1.) On immunoblotting of rat brain extract (Fig. 2), both 91-01 and G2-09 recognized a single protein band having a molecular weight of authentic GMF-fl.
Fig. 6. Immunostaining for GMF-/3 in the lateral vestibular (Deiters) nucleus. Neurons of various sizes, including their processes, were stained. Bar = 50/~m,
Adsorption of both antibodies with r-hGMF-/3 completely abolished the immunoreactivity.
lmmunostaining of rat na~ral tissue with polyclonai antibody
Fig, 5. lmmunostaining for GMF-/3 in the midbrain. A: oculomotor nuclei showing positively stained neurons. B: red nucleus showing positively stained neurons. C: substantia nigra showing positively stained neurons in pars compacta (upper right half) and pars reticulata (lower left half), Bar --- 50 ~m.
Neural tissue from adult rat was processed and immunostained with rabbit polycional antibody 91-01. Transverse sections of the spinal cord showed strongly positive staining of astrocytos having star-shaped cell bodies with elongated processes which are in parallel formation and perpendicular to the surface of the cord (Fig. 3A). In addition, motor neurons in the anterior horn also stained positively for GMF-/3 (Fig. 3B). lmmunocytoehemistry of the cerebellar cortex (Fig. 4A) showed strong staining of the Purkinje cell perikarya and Bergmann gliai processes. In contrast, the dendritic arborizations of the Purkinje ceils were not stained. Most of the granule cells were negative. In the deeper structures of the cerebellum, neurons of the cerebellar nuclei (Fig. 4C) and astrocytes in the cerebellar white matter (Fig. 4D) were clearly positive for GMF-/3. The astrocytes exhibited typical stellate morphology. Many neurons in the midbrain were GMF-,8-positire, including those in the oculomotor nucleus, the red nucleus and substantia nigra (Fig. 5). In the medulla, numerous neurons in the lateral vestibular nucleus, including their long processes, showed positive-staining for GMF-/3 (Fig. 6).
In the basal ganglia (Fig. 7), a distinctive contrast was seen between the caudate nucleus and globus pallidus. While in the former only occasional neurons were stained, in the latter the majority of neurons were GMF-/3 positive. In the sensory/motor cortex (Fig. 8A) and entorhihal cortex (Fig. 8C), some GMF-/3-containing neurons were scattered on a background studded with GMF-/3positive astrocytes. There were relatively more immunoreactive neurons in the deep layers of the cortex compared to superficial layers. Likewise, in the hippocampal formation, GMF-//-positive astrocytes were numerous, while GMF-/3-positive neurons were few. In the hippocamptls proper, pyramidal cells were all unstained whereas the scattered smaller sized neurons were intensely stained. In the dentate gyrus, most of the granule cells were u_,stained except for a single row of neurons in the bottom of the granule cell layer which were highly immunoreactive (Fig. 8D). GMF-//-like immunoreactivity was also found in the dorsal root ganglia (Fig. 9). Here, both the perikarya of sensory neurons and the surrounding capsule cells were stained. The intensity of staining in the capsule cells was particularly strong. DISCUSSION The current work was undertaken to compare a recently obtained polyclonal antibody (rabbit serum
91-01) derived from recombinant GMF-//with a monoclonal antibody (G2-09) previously raised against natural GMF. The two antibodies exhibited similar immunochemicai specificity: they both recognized the same GMF-/3 protein, whether in the pure state or from a crude brain extract, and they both showed no cross-reactivity with other growth factor proteins. However, despite these similarities, the two antibodies gave different immunocytochemical reactions. Thus, while the monoclonal antibody stained only astrocytes in the spinal cord and Bergmann glia in the cerebellum 5, the polyclonal antibody stained also astrocytes in other areas of the nervous system and neurons as well. There are several explanations for the larger cell population stained by the polyclonal antibody. The simplest explanation is that the polyclonal antibody was more sensitive because it recognized multiple epitopes, as opposed to the monoclonal antibody which recognized a single ¢pitope. Thus, whereas the monoclonal antibody stained only cells that contained the highest level of GMF-/3, the polyclonal antibody also stained cells that contained lower levels of the protein. Another attractive explanation is that GMF-~ may belong to a family of related proteins, so that while the monoclonal antibody G2-09 recognized an epitope unique to GMF-~, the polyclonal antibody 91-01 also recognized epitopes that were in common with other members of the family. However, we have not been able to support this by observing extra protein bands,
Fig. 7. lmmunostainingfor GMF-// in basal ganglia.A: caudate nucleusshowingselectivestainingof a few neurons. B: globuspallidus showing positive stainingof numerous neurons. Bar = 50 ~m.
with either 1-D or 2-D gels, that immunoblotted the polyclonal antibody. A more definitive answer must await the isolation of monoclonal antibodies that can stain both astrocytes and neurons. It should be noted that not all neurons were positive for GMF-/], and not all the positive neurons were stained with equal intensity. Also, in both neurons and astrocytes, only the cytoplasm and not the nucleus was stained, indicating that GMF-/3 is a cytoplasmic rather than nuclear protein. In astrocytes, both cell processes and the perinuc[ear region were equally stained. This was not true for neurons, as staining of dendritic processes did not always accompany staining of the perikaryon (as in Purkinje cells). On the other hand, oligodendroglia and white matter were consistently negative for endogenous GMF-/3. No attempt was made
Fig. 9. immunostaining for GMF-# in spinal ganglia (L,O, showing positively stained neurons and capsule cells surrounding neuronal pcrikarya. Bar =50 #m. Inset: details with I.Sxfurther magnification.
•
u
t
Fig. 8. Immunostaining for GMF-# in cerebral cortex, A: sensory/motor cortex showing positive staining of many astrocytes and a few neurons. B: negative control for 'A" using adsorbed antibody. C: entorhinal cortex showing positive staining of numerous astrocytes and occasional neurons, D: dentate gyrus of hippocampal fi)rmation showing many positively stained astrocytes in the molecular layer (upper i~rtion), some positively stained neurons in the granular layer (middle zone), and a few positively stained neurons in the hilar region (lower portion), Bar = 50/~m,
in this study to immunostain organs outside the nervous system because, as we have demonstrated with mRNA hybridization, GMF-# is synthesized mainly in the nervous system (Zaheer, Fink and Lira, unpublished data). The demonstration of endogenous GMF-/3 in neurons as well as astrocytes suggests a wider range of cell-cell interaction than was previously thought. Thus, GMF-# could mediate communication among astrocytes, among neurons, and between neurons and astrocytes. The coexistence in the brain of a growth suppressor such as GMF-/3 and growth promoters such as the fibrobast growth factors, etc., implies the importance of checks and balances in cellular regulation that could be crucial to the development, degeneration and regeneration of the nervous system. Acktiowledgemet~t. We thank Brian D, Fink, Joseph C, Behr and Jeffrey Hwang for technical assistance, Dr. John E. Butler for help with rabbit immunization, and Dr. H,K. Kultas-llinsky for advice and discussions, This work was supported by the following grants to R.L.: Veterans Affairs Merit Review Award, National Science Foundation Grant BNS-8917665, and Diabetes-Endocrinology Center Grant DK25295,
REFERENCES 1 Bosch, E.P., Zhong, W. and Lim, R., Axonal signals regulate expression of glia maturation factor /3 in Schwann cells: an immunohistochemical study of injured sciatic nerves and cultured Schwann cells, J. Neurosci., 9 (1989) 3690-3698. 2 Kaplan, R., Zaheer, A., Jaye, M. and Lira, R., Molecular cloning and expression of biologically active human glia maturation factor fl, J. Neurochem., 57 (1991) 483-490. 3 Laemmli, U.K., Cleavage of structural proteins during the assembly of the head of bacteriophage T4, Nature, 227 (1970) 680-685. 4 Lira, R., Miller, J.F., Hicklin, D.J. and Andresen, A.A., Purification of bovine glia maturation factor and characterization with monoclonal antibody, Biochenffstry, 24 (1985) 8070-8074. 5 Lira, R., Hicklin, D.J., Miller, J.F., Williams, T.H. and Crabtree, J.B., Distribution of immunoreactive glia maturation factor-like molecule in organs and tissues, Dec. Brabz Res., 33 (1987) 93-100.
6 Lim, R., Miller, J.F. and Zaheer, A., Purification and characterization of glia maturation factor fl: a growth regulator for neurons and glia, Proc. Natl. Acad. Sci. USA, 86 (1989) 3901-3905. 7 Lira, R. and Huang, L., Gila maturation factor/3 promotes the appearance of large neurofilament-rich neurons in injured rat brains, Brain Res., 504 (1989) 154-158. 8 Lira, R., Zhong, W. and Zaheer, A., Antiproliferative functior~ of glia maturation factor/3, Cell Regul., 1 (1990) 741-746. 9 Lim, R., Liu, Y. and Zaheer, A., Glia maturation factor fl regulates the growth of N18 neuroblastoma cells, Dec. Biol., 137 (1990) 444-450. 10 Lim, R., Zaheer, A. and Lane, W.S., Complete amino acid sequence of bovine glia maturation factor fl, Proc. Natl. Acad. Sc£ USA, 87 (1990) 5233-5237. 11 Lim, R. and Zaheer, A., Preparation of glia maturation factor fl, Methods Neurosci., 6 (1991) 321-337.
