Myelin Basic Protein and Myelin Basic Protein ... - Springer Link

Neurochemical Research, Vol. 24, No. 2, 1999, pp. 255-260

Myelin Basic Protein and Myelin Basic Protein Peptides Induce the Proliferation of Schwann Cells Via Ganglioside GM1 and the FGF Receptor* Shun-Fen Tzeng,2 Gladys E. Deibler,3 and George H. DeVries1,4 (Accepted August 14, 1998)

Myelin basic protein (MBP) and two peptides derived from MBP (MBP1-44 and MBP152-167) stimulated Schwann cell (SC) proliferation in a cAMP-mediated process. The two mitogenic regions of MBP did not compete with one another for binding to SC suggesting a distinctive SC receptor for each mitogenic peptide. Neutralizing antibodies to the fibroblast growth factor receptor blocked the mitogenic effect of the myelin-related SC mitogen found in the supernatant of myelin-fed macrophages. The binding of 125I-MBP to Schwann cells was specifically inhibited by basic fibroblast growth factor (bFGF) and conversely the binding of 125I-bFGF was competitively inhibited by MBP. These data suggested that the mitogenic effect of one MBP peptide was mediated by a bFGF receptor. The binding of MBP to ganglioside GM1 and the ability of MBP peptides containing homology to the B subunit of cholera toxin (which binds ganglioside GM1) to compete for the binding of a mitogenic peptide (MBP1-44) to SC, identified ganglioside GM1 as a second SC receptor. Based on these results, we conclude that MBP1-44 and MBP152_167 associate with ganglioside GM1 and the bFGF receptor respectively to stimulate SC mitosis.

KEY WORDS: Myelin basic protein; mitogens; Schwann cell; ganglioside.

INTRODUCTION

produced during Wallerian degeneration may be the local source of the mitogenic signal (4-6). Our laboratory has reported that soluble mitogens related to MBP are produced by macrophages which ingest myelin (6-9). Three lines or evidence implicate MBP-related peptides as the mitogens produced by macrophage proteolytic processing of myelin (9). First, the mitogenic effect of the macrophage produced soluble mitogen can be corn-

After nerve injury, SC1 are stimulated to proliferate (1,2). Without this proliferative response SC cannot remyelinate the regrowing nerve (3). The origin of the mitogen operative during nerve degeneration is not certain, although evidence has been presented that debris Research Service, Hines/VA Hospital, Hines, Illinois 60141, Department of Cell Biology, Neurobiology and Anatomy, Stritch School of Medicine-Loyola University Chicago, Maywood, Illinois 60153. 2 Present Address: Department of Medical Research and Education, Taichung VA Hospital, Taichung, Taiwan, Republic of China. 3 Laboratory of Cerebral Metabolism, NIMH-NIH, Bethesda, MD. 20892 4 To whom reprint requests should be addressed at: Research 151, Hines VA Hospital, 5th Avenue and Roosevelt Rd, Hines, Illinois 60141. * Special issue dedicated to Dr. Kunihiko Suzuki 1

Abbreviations: ABTS-2,2'-amino-di-3-ethyIbenzol thiazoline sulfonic acid; aFGF-acidic fibroblast growth factor; bFGF-basic fibroblast growth factor; BSA-bovine serum albumin; FGF-flbroblast growth factor; GAl-asialo ganglioside GM1; GD1a-disialosyl ganglioside; GM1monosialosyl ganglioside; Li-labeling index; MBPXX-ZZ-Myelin basic protein peptide containing amino acids xx to zz (inclusive); MBPmyelin basic protein; MEF-myelin enriched fraction; DMEM-Dulbecco's modified Eagle medium; Li-labeling index; PBS-Phosphate buffered saline; SC-Schwann cells

255 0364-3190/99/0200-0255$16.00/0 C 1999 Plenum Publishing Corporation

256 pletely blocked by a specific antiserum to MBP. Secondly, myelin isolated from the Shiverer mutant mice, which is devoid of MBP, is not mitogenic for SC. Thirdly, liposomes in which MBP peptide has been incorporated are mitogenic for SC. Additionally we have previously reported that the mitogenic potency of the myelin was directly related to its MBP content (8). In this study, we utilized mitogenic assays and a competitive binding assay with I 125 labeled and unlabeled MBP and MBP peptides and related molecules to identify the peptides of MBP which stimulate mitosis and to identify two SC receptors which bind to MBP and MBP peptides.

EXPERIMENTAL PROCEDURE Materials. Polyclonal antiserum to MBP, Rabbit MBP and MBP peptides were obtained as previously described (Tzeng et al., 1995). DMEM and fetal calf serum were purchased from Gibco (Gaithersburg, MD) and Hyclone (Logan, UT) respectively. Antibody to FGF receptor was purchased from UBI (Lakeland, NY). Dr. Robert K. Yu (Medical College of Virginia, Richmond, VA), supplied gangliosides. All other chemicals were purchased from Sigma Chemical Co (St. Louis, MO). 125I-bFGF was purchased from New England Nuclear (Boston, MA). Preparation of Schwann Cells. SC were prepared [by modification of the method of Brockes et al (11)] as previously described (10). The purified SC were replated in 48 well plates at a density of 40,000 cells per well in DMEM containing 10% fetal calf serum. Peptides or forskolin were added at the specified concentrations to each well for the indicated times and the medium was not changed for the duration of the experiment. BrdU Incorporation Assay. To measure proliferation, BrdU was added to the cultures at a final concentration of 10 \iM during the last 24 hours of treatment. Labeling indices (LI) were then determined after immunostaining for incorporated BrdU as previously described (10). All labeling indices were determined as the percentage of BrdU labeled cells in a total of 1,000 cells. lodination of MBP and MBP Peptide. MBP and MBP peptides were iodinated utilizing 125I-PIB reagent (New England Nuclear, Boston, MA.), and the previously described protocol (10). The specific activity of the labeled 125I-MBP utilized in these experiments was 50 uCi/nmol. The specific activity of 125I-MBP1-44 was 20 nCi/nmol. All radiolabeled MBP and MBP peptides were determined to retain their full mitogenicity after radiolabelling (data not shown). Competitive Binding Assays. The competitive binding assays were carried out as previously described (10). Briefly, SC were replated in a 96 well plate at a density of 20,000 cells per well. SC cells were incubated for 3 hours at 4°C in PBS containing 1% BSA and the indicated concentration of radiolabelled peptides and competitive unlabelled peptides. The cells were then washed three times with cold PBS and extracted with PBS containing 2% SDS and 10 mM NaOH. The amount of radioactivity in the cell extract was determined by liquid scintillation counting. Elisa Assay for MBP Binding to Gangliosides. A 96 well microtiter plate was coated with 1 |ig of ganglioside GM1, GD1a or GA1 as previously described (10) followed by incubation at room temperature for 2 hours with increasing amounts of MBP (10 -3 , 10-2, 0.1, 1,

Tzeng, Deibler, and DeVries 5, 10 \ig). The plate was incubated with rabbit anti-MBP antiserum at a dilution of 1:6000 for 1 hour. One hour after the addition of peroxidase conjugated anti-rabbit IgG (1:10,000), chromogenic substrate (ABTS) was applied to the plate, and after a 10 minute incubation at room temperature, the absorbance was read at 405 nm in a Bio-Tek microplate reader (10) (Winooski, VT). Stalisical Analysis. A T-test for independence was applied to data in order to evaluate the statistical significance. The T-values were calculated with a certainty of 95% to accept or reject the null hypothesis namely that two sets of data analyzed are significantly different.

RESULTS As shown in Fig. 1a, MEF is mitogenic for primary cultured Schwann cells as previously reported (8). In contrast, in the absence of elevated intracellular cAMP, MBP at 3 different doses was not mitogenic for SC. Many SC mitogens are known to require elevated intracellular cAMP to become mitogenic (12). Therefore, we evaluated the mitogenicity of MBP and a control peptide (cytochrome c, which has a molecular weight and isoelectric point similar to that of MBP) in the presence of 1 |o,M forskolin. Under these conditions there was a MBP dose-dependent increase in mitogenicity and the mitogenicity of MEF was further activated (Fig. la). Statistical analysis of the data (Student's West) revealed that each dose of either MEF or MBP was significantly more mitogenic than forskolin alone. In addition each increasing dose of MEF was increasingly and significantly more mitogenic. However in the case of increasing does of MBP in the presence of forskolin, only the 1 |^M and 2 uM doses were significantly different from each other while the mitogenicity of the 2 uM and 3 |j,M MBP (in the presence of forskolin) were not significantly different from each other. We then investigated which regions of MBP were responsible for the stimulation of SC proliferation using four peptides, which spanned the entire MBP molecule. As shown in Fig. Ib, in the presence of 1 (iM forskolin, SC proliferation was stimulated by either MBP1-44 or MBP152_167, but not by MBP45_87 or MBP88-151. Statistical analysis confirmed that the labeling index stimulated by either MBP1-44 or MBP152-167 were significantly different than the labeling index with no peptide present while the labeling index in the presence either MBP 45 _ 87 or MBP88_151 was not significantly different than the condition of no peptide present. We then further investigated the type of SC receptor(s) with which the MBP or MBP-related peptides interacted to stimulate mitosis. The mitogenic effect derived from MEF was completely inhibited by the addition of an anti-FGF receptor antibody: LI of control SC = 1.1% ± 0.4%; LI of SC treated with 40|ig/ml MEF = 14.5%

Myelin Basic Protein and Schwann Cell Proliferation

257

Fig. 1. Mitogenio effect of MBP and MBP peptides on cultured SC in the presence or absence of 1 (O.M forskolin (a) SC were treated with MEF, MBP or cytochrome c at the indicated concentrations in the presence or absence of 1 (iM forskolin for 72 hours, followed by determination of the labeling index as described in Methods, (b) SC were treated with 3 nM of each MBP peptide in the presence of 1 ^M forskolin followed by determination of the labeling index as described in Methods.

± 1.5; LI of SC in the presence of 1:25 anti-FGF receptor = 5.2% ± 0.8; LI of SC treated with 40(ig/ml MEF plus 1:25 anti-FGF receptor = 7.4% ± 0.8% (n = 3). These data suggested that MBP-related factor derived from MEF might interact with the FGF receptor stimulate SC proliferation. Therefore we carried out competitive binding assays using 125I-MBP and unlabeled bFGF. As shown in Fig. 2a, increasing amounts of bFGF, but not acidic FGF(aFGF), attenuated the binding of 125I-MBP to cultured SC. Conversely, MBP at a concentration of 10p,M, could inhibit 80% of the specific binding of 125I-bFGF to SC (Fig. 2b). The ability of unlabeled MBP to compete for the radiolabelled MBP was not evident until the unlabeled MBP was present at concentrations from 1 nM to 1 mM. Although the unlabeled

MBP was not able to significantly displace the radiolabeled MBP until it reached these concentrations, the bFGF competed in a similar manner. This binding data is compatible with competition of MBP and bFGF for a common binding site. We have previously reported that neither TGF-0 nor PDGF-BB inhibited the binding of 125I-MBP when used in the same concentration range as bFGF (10). The specificity of the binding competition between MBP and bFGF to SC was supported by the fact that the control protein (cytochrome-c) did not influence the binding of 125I-bFGF to SC (Fig. 2b). Many reports have shown that aFGF and bFGF can bind to a common receptor family but with different affinities (13). We observed that aFGF at a concentration of 100 nM only partially inhibited the binding of bFGF to SC

258

Tzeng, Deibler, and DeVries

Fig. 2. Competitive binding of 125I-MBP (a) or 125I-bFGF (b) and unlabeled competing proteins, (a) SC were incubated with 186 mM 125I-MBP and the indicated concentrations of unlabeled aFGF, bFGF or MBP followed by determination of bound radioactivity as described in Methods, (b) SC were incubated with 100 pM 125I-bFGF and the indicated concentrations of unlabelled cytochrome c or MBP followed by determination of bound radioactivity as described in Methods. Inset: SC were incubated with 100 pM 125I-bFGF and the indicated concentrations of unlabelled aFGF or bFGF followed by determination of bound radioactivity as described in Methods.

(Fig. 2b, inset), and did not compete with the binding of 125I-MBP to SC (Fig. 2a). The inability of aFGF to effectively compete with bFGF or MBP binding to SC was consistent with the view that MBP specifically bound to a type of FGF receptor that contained a high affinity site for bFGF. These results demonstrated that part of the mitogenic effect of MBP on SC could be mediated by the FGF receptor. Since two mitogenic regions of MBP were identified (Fig. 1b), we investigated whether or not each of the mitogenic MBP peptides associated with a distinctive receptor. Using radiolabeled MBP1-44 we investigated whether or not MBP152_167 could compete for the same SC receptor. As shown in Fig. 3, at concentrations of

up to 1 mM, MBP152-167 clearly did not block the binding of MBP1-44 to SC. This result was consistent with the view that MBP152-167 and MPB1-44 interacted with distinctive and separate receptors. In contrast, the non-mitogenic peptide MBP88_151, was a more potent inhibitor for the binding of 125I-MBP1-44 than the unlabelled MBP1-44 (Fig. 3). These data suggest that MBP1-44 and MBP88_151 contain a related binding motif distinctive from that of MBP152_167. We have previously reported that both MBP1-44 and MBP88_151 could bind to ganglioside GM1 (10). In addition, both of these MBP peptides contain an amino acid sequence homologous to the binding subunit of cholera toxin (14). As shown in Fig. 4, using an enzyme-linked immunosorbent assay, we dem-

Myelin Basic Protein and Schwann Cell Proliferation

259

Fig. 3. Competitive binding of 125I-MBP1-44 and unlabeled competing MBP peptides. SC were incubated with 580 pM 125I-MBP1-44 and the indicated concentrations or unlabeled peptides MBP1-44, MBP88-151 or (inset) MBP45_87 or MBP152_167 followed by determination of bound radioactivity as described in Methods.

Fig. 4. Biding of MBP to ganglioside GM1, GD1 a or GA1. Ganglioside coated ELISA plate wells were incubated with MBP at the indicated concentrations, washed, and the amount of bound MBP was determined as described in Methods.

onstrated that MBP bound to ganglioside GM1 in a dose dependent manner, whereas no association of MBP with ganglioside GD1a or GA1 is noted. These data suggested a specific association of MBP with GM1. DISCUSSION The data presented in this investigation are consistent with the view that two distinct regions of MBP interact with two cellular receptors in SC to stimulate mitosis. The specific association of MBP to ganglioside GM1 (Fig. 4) suggests that SC mitosis is a consequence of the binding of MBP and MBP peptides to this cell

surface molecule. This mitotic response is not unexpected since cholera toxin also binds to ganglioside GM1 and cholera toxin is a potent mitogen, which stimulates SC proliferation (15). The homology of MBP or MBP peptides to either the A or B subunits of cholera toxin may be related to the response of SC to a given MBP-related molecule. After the binding of the B subunit of cholera toxin to ganglioside GM1, there is a transfer of the cholera toxin A subunit through the membrane to mediate ADP-ribosylation of G-type proteins (15). MBP has been shown to be ADP-ribosylated and to bind GTP, indicating a possible G-protein like role in signal transduction (16, 17). Therefore, MBP peptides

Tzeng, Deibler, and DeVries

260 with homology to both the A and B subunits of cholera toxin (e.g., MBP1-44) should be mitotically active whereas peptides with homology to only the B subunit of cholera toxin will bind to SC but will not be mitotically active (e.g. MBP88_151). Our data support this view, since MBP1_44 is mitotically active, while MBP88_151 is not a mitogen (fig. 1b) but can inhibit the binding of MBP1-44 to SC (Fig. 3). MBP152_167 did not compete with the binding of MBP1-44 (Fig. 3, inset), raising a question as to the identity of the receptor for MBP152_167. In addition, we have previously shown that MBP can be crosslinked to SC receptors, which have molecular weights consistent with the bFGF receptor (10). MBP competed with 125I-bFGF for SC binding sites and conversely bFGF competed with 125I-MBP for binding sites on SC. These data coupled with the fact that MBP152_167 did not block the binding of MBP1-44 suggests that MBP152_167 may interact with FGF receptor to stimulate SC proliferation. We conclude that there are two mitogenic regions in MBP that stimulate SC proliferation and that each region uses distinct cellular receptors. One mitogenic region, located on the first forty-four residues of the amino-termmus of MBP molecule may associate with ganglioside GM1. The other mitogenic region, located on the last fifteen residues of the carboxyl-terminus of MBP may interact with the bFGF receptor. Both MBPrelated mitogens require elevated intracellular cAMP to become mitogenic for SC. Our results are consistent with the following scenario. After peripheral nerve injury macrophages ingest myelin debris and via proteolysis produce mitotically active MBP-peptides. In addition, activated macrophages produce factor(s) which increase intracellular cAMP levels (10). The MBP-peptides then interact with SC bFGF receptors and ganglioside GM1 to stimulate SC mitosis which occurs after nerve degeneration (1,2). Current experiments are underway to determine whether or not this data can be confirmed and extended by in vivo studies. Experimental approaches include determining which MBP peptides are produced during Wallerian degeneration and the identity of mitogenic MBP peptides in vivo. The importance of this SC mitotic response in Wallerian degeneration is underscored by the fact that reinnervating axons are not mitogenic for SC (18) and that SC mitosis is a prerequisite for remyelination (3). ACKNOWLEDGMENTS This research was sponsored by a USPHS grant from the National Institute of Neurological Disease and Stroke (NS15408) and by a grant

from the Medical Research Service Department of Veterans Affairs. The authors thank Dr. Timothy J. Neuberger for helpful discussion, Dr. AH Badache for a critical reading of the manuscript, Charity Preussler for help with the preparation of the manuscript, and Dr. Weichun Xu (Cooperative Studies Program, Hines VA Hospital) for help with the statistical analysis.

REFERENCES 1. Abercrombie M., and Johnson M.L. 1946. Quantitative histology of Wallerian degeneration. I. Nuclear population in rabbit sciatic nerve. J. Anat. 80:37-50. 2. Griffin J.W., Stockes E.A., Fahnestock K., Praagh A.V., Trapp B.D. 1990 Schwann cell proliferation following lysolecithin-induced demyelination. J. Neurocytol. 19:367-384. 3. Hall S.M., Gregson N. 1974 The effects of mitomycin C on remyelination in the peripheral nervous system. Nature 252:303305. 4. DeVries G.H. 1993. Schwann cell proliferation, Peripheral Neuropathy in Pages 290-298, (Dyck, P., Thomas, P., 3rd (ed.), W.B. Saunders, Philadelphia. 5. Sobue G., Krieder B., Asbury A.K., and Pleasure D. 1983. specific and potent mitogenic effect of axolemmal fraction on Schwann cells from sciatic nerves of rat in serum-containing and defined media. Brain Res. 280:263-275. 6. Baichwal R.R., and DeVries G.H. 1992. A mitogenic factor derived from the myelin membrane. In Myelin: biology and chemistry (Martenson RE), pp633-652, CRC Press, Boca Raton. 7. Yoshino J.E., Dinneen M.P., Lewis B.L., Meador-Woodruff J.H., and DeVries G.H. 1984. Differential responses of cultured Schwann cells to axolemma and myelin enriched fractions. I. Biochemical studies. J. Cell Biol. 99:2309-2313. 8. Baichwal R.R., Bigbee J.W., and DeVries G.H. 1988. Macrophage-mediated myelin-related mitogenic factor for cultured Schwann cells. Proc. Natl. Acad. Sci. U.S.A. 88:1701-1705. 9. Baichwal R.R., and DeVries G.H. 1989. A mitogen for Schwann cells is derived from myelin basic protein. Biochem. Biophys. Res. Commun. 164:883-888. 10. Tzeng S.-F., Deibler G.E., Neuberger T.J., and DeVries G.H., 1995. Two mitogenic regions of MBP interact with different receptors to induce Schwann cell proliferation in a eAMP mediated process. J Neurosci. Res. 42:758:767. 11. Brockes J.P., Fields K.L., and Raff M.C. 1979. Studies on cultured rat Schwann cells. I. Establishment of purified populations from cultured of peripheral nerve. Brain Res. 165:105-118. 12. Davis J.B., and Stroobant P. 1990. Platelet-derived growth factors and fibroblast growth factors are mitogens for rat Schwann cells. J. Cell Biol. 110:1353-1360. 13. Johnson D.E., Lee P.L., Lu J., and Williams L.T. 1990. Diverse forms of a receptor for acidic and basic fibroblast growth factors. Mol. Cell. Biol. 10:4728-4736. 14. Caamano C.A., and Zand R. 1989. Homologous sequences in cholera toxin A and B subunits to peptide domains in myelin basic protein. FEES Lett. 252:88-90. 15. Moss J., and Vaughan M. 1979. Activation of adenylate cyclase by choleragen. Ann. Res. Biochem. 48:581-600. 16. Chan K.C., Ranwani J., and Moscarello M.A. 1988. Myelin basic protein binds GTP at a single site in the N-terminus. Biochem. Biophys. Res. Commun. 143:908-912. 17. Boulias C., and Moscarello M.A. 1989. ADP-ribosylation of myelin basic protein charge isoforms. Trans. Am. Soc. Neurochem. 20:253. 18. Siironen J., Collany Y. and Roytta M. 1994. Axonal reinnervation does not influence Schwann cell proliferation after rat sciatic nerve transection. Brain Res. 654:303-311.