- Email: [email protected]
II. Application of high performance liquid chromatography to the preparation of encephalitogenic myelin basic protein
Life Sciences, Vol. 30, pp. 989-993 Printed in the U.S.A.
Pergamon Press
MINISYMPOSIUM II. APPLICATION OF HIGH PERFORMANCE LIQUID CHROMATOGRAPHY TO THE PREPARATION OF ENCEPHALITOGENIC MYELIN BASIC PROTEIN I Channing L. Hinman, Helene C. Rauch and Robert F. Pfeifer Department of Immunology and Microbiology Wayne State University School of Medicine Detroit, MI 48201 Waters Associates, Inc. Milford, MA 01757 Summary Two preparations of myelin basic protein (MBP) were derived from an acid extraction of chloroform-methanol defatted bovine spinal cord. The first was purified by ion-exchange chromatography using guanidineHCI; the second, by high performance liquid chromatography (HPLC) using a triethylamine eluant. Both methods of preparation yield MBP which is identical on acid-urea polyacrylamide gel elec~rophoresis and which has identical encephalitogenic potency. Because of the greater timeefficiency of the HPLC system with no deleterious side effects due to buffer contamination, this latter method can be recommended for MBP purification. Myelin basic protein (MBP) is a low molecular weight (18,500) constituent of the myelin membrane which ensheaths axons in the central nervous system (CNS). Although the precise role of MBP in the immunopathology and etilogy of demyelinating diseases such as multiple sclerosis remains to be defined, MBP (or its fragments) in cerebrospinal fluid (CSF) is of major value for clinical assessment of the extent of myelin destruction which occurs in CNS pathologic processes (Cohen et al., 1978; Wicher et al., 1979; Frick and Stickl, 1980). Immunoassays using radioisotopic or fluorometric markers to detect the presence of protein(s) or peptide(s) that bind antibody directed against MBP (Tennenbaum et al., submitted for publication) correlate well with the occurrence of clinical exacerbations in MS and of strokes (Cohen, op. cit.). MBP, injected with adjuvant, is capable of inducing autoimmune encephalomyelitis (EAE) in a variety of species (Rauch and Einstein, 1974). EAE is a useful experimental model for the study of primary demyelinating diseases, as well as of autoimmunity in general." Specifically, the induction of EAE with MBP is used to study mechanisms of immunoregulation in autoimmune diseases. For example, MBP given without adjuvant (either pre- or post-challenge) can protect animals from the development of inflammatory lesions and/or clinical symptoms, which otherwise occur in the CNS following challenge with MBP in adjuvant (Alvord et al., 1965; Rauch and Einstein, 1971; Wilenborg et al., 1978; Alvord et al., 1979; Chou et al., 1980). Adoptive transfer studies in guinea pigs indicate that cells from protected animals are responsible for the unresponsive state in challenged recipients (Rauch, in press). Moreover, it has been reported that a peptide fragment of the rat MBP molecule can lead to the production of cells which cause EAE, while another portion of the same molecule stimulates a different population of cells which suppress disease (Swanborg et al., 1974; Welch et al., 1980). iSupported in part by grants NS 12754 and AI 07118 from the Department of Health and Human Services. 0024-3205/82/120989-05503.00/0 Copyright (e) 1982 Pergamon Press Ltd.
990
HPLC for MBP Puriflcation
Vol.
30, No. 12, 1982
Because of the need in the above situation for efficiently prepared, hlgh yields of purified MBP, we have investigated alternatlve methods for its isolation and purification. In this communlcation, we compare MBP purified by HPLC with MBP derived from conventional ion-exchange chromatography (IE). We have found that the yields of encephalitogenic MBP are equivalent, that the HPLC buffers used for elution do not alter the immunologic potency of the MBP, and that they are not themselves encephalitogenic. The remarkable efficiency of the HPLC system recommends this method for the purification of MBP. Methods Preparation
of MBP
Myelin basic protein (MBP) was isolated from bovine spinal cord by an acid extraction procedure (pH=l.90) following chloroform-methanol separation from homogenized tissue (Nakao et al., 1966; Oshiro and Eylar, 1970). Purificatlon of MBP was inltiated for one sample of 200 mg dlssolved in 4 ml eluant by cation-exchange (IE) chromatography using a 1.6 x 30 cm column (Pharmacia) packed with 50 ml of Blorex-70, sodium form, minus 400 mesh. The eluting buffer was a linear 5-15% gradient of guanidlne-HCl in 0.i M NaH2PO4, adjusting to pH 6.5 with 1 N NaOH, and the flow rate was 0.25 ml/mln. A second, identical sample from the original batch of isolate was processed using a Model 204 HPLC system, and a 7.8 mm x 30 cm protein analysis column, 1-125 (Waters Assoc., Milford, Mass). The isocratlc solvent consisted of a 2.5% trlethylamine buffer adjusted to pH 4.5 with glacial acetic acid, flowing at 1 ml/min. Sample concentratlon was 4 mg per ml of eluant. Followlng chromatography eaCh set of pooled material from the peak of interest was dialyzed separate]y against doubly-distilled water at 4 ° C. After lyophllization of the supernatants, each o preparation was stored over dessicant at -70 C. until use. Yields of purlfied MBP were calculated as: mg MBP (from major peak)/total mg crude sample applied to the column. Comparison of the two samples was made using two different pH polyacrylamide gel electrophoresis systems (Diebler et al., 1972; Benuck, et al., 1975). Each sample was dissolved in 1.0 M acetlc acid at a concentratlon of 2 mg/ml; urea was added, to a final concentration of 8M; and 20 - 40 ul of each dissolved sample was loaded onto each 5% gel. For the alkaline pH electrophoresis, samples were dissolved at a concentration of 1 mg/ml in 5 m M Tris-glyclne at pH 8.7; sucrose was added, to a final concentratlon of 30% (w/v); and 40 ul was loaded onto each 10% ~el. In both systems, electrophoresis was done at 2.5 mA/ gel for 1-2 hours. Assay of MBP for biological actlvity To test for encephalitogenic activity, each of the two preparations was injected subcutaneously into separate groups of adult guinea plgs, and clinical corre lates of the MBP-induced EAE were observed. Briefly, each of 3 animals in each group received 20 ug MBP via a single 0.i ml injection into each hind footpad of an emulsion which consisted of 0.2 mg ~ P dissolved In 1.0 ml of sterile physiologicaly saline and subsequently emulsified wlth an equal volume of complete Freund's adjuvant (Difco). Animals were observed dally for clinical signs of disease, indicated by weight loss, incontlnence, ataxia, and hindlimb paralysis. Animals were sacrificed either wlthin 24 hours of the onset of paralysis or 26 days post-injection. Brains and spinal cords were fixed in formalin, embedded in paraffin, and stained by hematoxylln and eosin following saglttal sectionlng. Histology was scored as mild, moderate, or severe perivascular monomuclear cell infiltration of the CNS white matter. The relatively low doses of MBP (normal=40-50 ug/animal) were used in an effort to separate the
Vol. 30, No. 12, 1982
HPLC for MBP Purification
991
effects of the two differentially purified preparations of MBP samples. Results Comparisons of HPLC and conventional IE chromatography can be made using at least three independent criteria. One basis for comparison is biochemical purity; another is efficiency of product acquisition; a third and a sometimes neglected area is the biological activity of the material following chromatography. After dlalysis and lyophilization, both methods of chromatography gave comparable yields: 51% from HPLC and 61% from IE. Because of losses associated with dialysis, there is no significant difference between these yields. In addition, the MBP derived by these two methods was identical on both acidurea and Tris-glycine polyacrylamide gels at disparate pH's. Purity was confirmed by the appearance of only a single, narrow band following the application of either 40 or 80 ug samples to each gel. An alternative method was also used to verify the extent of removal of contaminants by HPLC. Material collected from the elution peak was re-injected and chromatographed. Even with increased senslvity, only a slngle, symmetric peak occurred (8.95 min at a flow rate of 0.7 ml/min), and fluctuations in the baseline were less than + 3% of peak absorbance through the 20 min. run. Repeated sample injections at a flow rate of 0.7 ml/min, all yielded identical peak elution times (8.92 + 0.06 min.) with no evidence of protein subsequently trailing off the column. After as many as 15 consecutive 20-min. chromatography runs, elution profiles remained unchanged. Identical profiles were obtained on subsequent days. A major difference in the two methods is time-efficiency. At least a 2day process is involved in IE to repack, equilibrate, and run the column with an average load of 200 mg MBP and a 20 hr elution time. To process this same 200 mg of MBP with the analytic HPLC 1-125 column would require less than 5 hrs (the MBP peak eluted from the HPLC column at 7.35 min. following each injection, as seen in Figure i; the maximum injection volume for the equiment used was 500 ul; concentrations up to at least 20 mg/ml can be applied to the 1-125 column). Other factors which determine preparative efficiency include the flow rate needed for adequate peak separation and the availability of automated in]ectlon and collectlon apparatus. Even with manual injection and without using a preparative column, the efficiency of HPLC is seen to be much greater than that of conventional chromatography. With regard to encephalitogenic potency of the MBP purlfied by each method, the same incidence of encephalomyelitis was observed with both the IE- and HPLC-preparations of purified MBP. ~By day ii, two animals in each group were incontinent. By day 14, one of the two incontinent animals in either group was paralyzed, and by day 19 the third animal of each group had shown at least a 5% weight loss (approx. 20 g) maintained for at least two consecutive days, with subsequent recovery. Histological examination of the brain and spinal cord sections of animals from both groups revealed that comparable pathology had occurred. In each of the animals, there were areas of perivascular cuffing and infiltratlon of mononuclear cells, indicative of EAE. As expected, the animals which had been paralyzed showed the most extensive pathology. Although the triethylamine buffer used to elute MBP from the HPLC column should be readlly removed by dialysis, it was important to verify that triethylamine was not itself capable of facilitating EAE induction. Thus, concentrations of triethylamine up to 100x the concentration used for the column elution were injected into mice. Mice were chosen for these experiments because of the
992
HPLC for MBP Purification
Vol. 30, No. 12, 1982
MBP
0
2
4
6 8 10 12 14 TIME (min)
FIG. ] HPLC separation of myelin basic protein from bovine spinal cord. Injecttion volume, 200~I; Detection, 280nm at 0.20 AUFS; flow, l ml/min; Mobile Phase, TEA acetate. high numbers of animals required and because it has been demonstrated (Montgomery and Rauch, 1979) that bovine MBP can produce both clinical and pathological signs of EAE in mice. Three different strains of mice (SJL/J, BALB/c, and C57BI/6) were tested, with 5 mice in each strain compared against 5 age- and sex-matched controls. Control mice received footpad injections of physiological saline. Although clinical and histological signs of EAE are present in MBP-challenged mice between 12-21 days post-injection, the triethylamine-injected animals showed neither weight loss nor altered behavior over 40 days when compared with their controls. Histologic examination of brain and upper spinal cord sections confirmed the normal state of saline control and triethylamine challenged mice. Conclusion The above results indicate that the MBP purified by HPLC is identical to the MBP purified by ion-exchange chromatography, with respect to their ability to induce autoimmune EAE in guinea pigs. Significant differences in the ease and time efficier~cy o~ HPLC v e r ~ s IE chromatography strongly recommend the HPLC system for large scale purification of the basic protein of myelin.
Vol.
30, No. 12, 1982
HPLC for MBP Purification
993
Acknowledgement This work was supported in part by grants NS 12754 and AI 07118 from the Department of Health and Human Services. References i. 2. 3. 4. 5. 6. 7. 8. 9. i0. ii. 12. 13. 14. 15. 16. 17.
E.C. ALVORD, C.-M. SHAW, S. HRUBY and M.W. KIES, Ann. N.Y. Acad. Sci. 122 333 (1965). E.C. ALVORD, C.-M. SHAW and S. HRUBY, Ann. Neurol. 6 469 (1979). M. BENUCK, N. MARKS and G.A. HASHIM, Eur. J. Biochem. 52 615 (1975). F.C. CHOU, C.-H. CHOU, R.B. FRITZ and R.F. KIBLER, Ann. Neurol. 7 336 (1980). S.R. COHEN, M.J. BURNE, R.M. HERNDON and G.M. MCKAHNN, Adv. Exp. Med. Biol. I00 513 (1978). G.E. DIEBLER, R.E. MARTENSON and M.W. KIES Prep. Biochem. 2 139 (1972). E. FRICK and H. STICKL, J. Neurol. Sci. 46 187 (1980). I.N. MONTGOMERY and H.C. RAUCH, Fed. Proc. 38 1289 (1979). A. NAKAO, W.J. DAVIS and E. EINSTEIN, Biochem. Biophys. Acta. 130 163 (1966). Y. OSHIOR AND E.H. EYLAR, Arc. Biochem. Biophys. 138 392 (1970). H.C. RAUCH and E.R. EINSTEIN, Fed. Proc. 30 306 (1971). H.C. RAUCH and E.R. EINSTEIN, Rev. Neurosc. 1 283 (1974). H.C. RAUCH, Cell, Immunol. 60 220. R.H. SWANBORG, J.E. SWIERKOSZ and R,G. SAIEG, J. Immunol. 112 594 (1974). A.M. WELCH, J.H. HOLDA and R.H. SWANBORG, J. Immunol. 125 186 (1980). V. WICHER, W. OLSZEWSKI and F. MILGROM, Int. Arch. Allergy Appl. Immunol. 59 49 (1979). D.O. WILENBORG, E.A. STATEN and G.F. WITTING, Exp. Neurol. 61 527 (1978).
Pergamon Press
MINISYMPOSIUM II. APPLICATION OF HIGH PERFORMANCE LIQUID CHROMATOGRAPHY TO THE PREPARATION OF ENCEPHALITOGENIC MYELIN BASIC PROTEIN I Channing L. Hinman, Helene C. Rauch and Robert F. Pfeifer Department of Immunology and Microbiology Wayne State University School of Medicine Detroit, MI 48201 Waters Associates, Inc. Milford, MA 01757 Summary Two preparations of myelin basic protein (MBP) were derived from an acid extraction of chloroform-methanol defatted bovine spinal cord. The first was purified by ion-exchange chromatography using guanidineHCI; the second, by high performance liquid chromatography (HPLC) using a triethylamine eluant. Both methods of preparation yield MBP which is identical on acid-urea polyacrylamide gel elec~rophoresis and which has identical encephalitogenic potency. Because of the greater timeefficiency of the HPLC system with no deleterious side effects due to buffer contamination, this latter method can be recommended for MBP purification. Myelin basic protein (MBP) is a low molecular weight (18,500) constituent of the myelin membrane which ensheaths axons in the central nervous system (CNS). Although the precise role of MBP in the immunopathology and etilogy of demyelinating diseases such as multiple sclerosis remains to be defined, MBP (or its fragments) in cerebrospinal fluid (CSF) is of major value for clinical assessment of the extent of myelin destruction which occurs in CNS pathologic processes (Cohen et al., 1978; Wicher et al., 1979; Frick and Stickl, 1980). Immunoassays using radioisotopic or fluorometric markers to detect the presence of protein(s) or peptide(s) that bind antibody directed against MBP (Tennenbaum et al., submitted for publication) correlate well with the occurrence of clinical exacerbations in MS and of strokes (Cohen, op. cit.). MBP, injected with adjuvant, is capable of inducing autoimmune encephalomyelitis (EAE) in a variety of species (Rauch and Einstein, 1974). EAE is a useful experimental model for the study of primary demyelinating diseases, as well as of autoimmunity in general." Specifically, the induction of EAE with MBP is used to study mechanisms of immunoregulation in autoimmune diseases. For example, MBP given without adjuvant (either pre- or post-challenge) can protect animals from the development of inflammatory lesions and/or clinical symptoms, which otherwise occur in the CNS following challenge with MBP in adjuvant (Alvord et al., 1965; Rauch and Einstein, 1971; Wilenborg et al., 1978; Alvord et al., 1979; Chou et al., 1980). Adoptive transfer studies in guinea pigs indicate that cells from protected animals are responsible for the unresponsive state in challenged recipients (Rauch, in press). Moreover, it has been reported that a peptide fragment of the rat MBP molecule can lead to the production of cells which cause EAE, while another portion of the same molecule stimulates a different population of cells which suppress disease (Swanborg et al., 1974; Welch et al., 1980). iSupported in part by grants NS 12754 and AI 07118 from the Department of Health and Human Services. 0024-3205/82/120989-05503.00/0 Copyright (e) 1982 Pergamon Press Ltd.
990
HPLC for MBP Puriflcation
Vol.
30, No. 12, 1982
Because of the need in the above situation for efficiently prepared, hlgh yields of purified MBP, we have investigated alternatlve methods for its isolation and purification. In this communlcation, we compare MBP purified by HPLC with MBP derived from conventional ion-exchange chromatography (IE). We have found that the yields of encephalitogenic MBP are equivalent, that the HPLC buffers used for elution do not alter the immunologic potency of the MBP, and that they are not themselves encephalitogenic. The remarkable efficiency of the HPLC system recommends this method for the purification of MBP. Methods Preparation
of MBP
Myelin basic protein (MBP) was isolated from bovine spinal cord by an acid extraction procedure (pH=l.90) following chloroform-methanol separation from homogenized tissue (Nakao et al., 1966; Oshiro and Eylar, 1970). Purificatlon of MBP was inltiated for one sample of 200 mg dlssolved in 4 ml eluant by cation-exchange (IE) chromatography using a 1.6 x 30 cm column (Pharmacia) packed with 50 ml of Blorex-70, sodium form, minus 400 mesh. The eluting buffer was a linear 5-15% gradient of guanidlne-HCl in 0.i M NaH2PO4, adjusting to pH 6.5 with 1 N NaOH, and the flow rate was 0.25 ml/mln. A second, identical sample from the original batch of isolate was processed using a Model 204 HPLC system, and a 7.8 mm x 30 cm protein analysis column, 1-125 (Waters Assoc., Milford, Mass). The isocratlc solvent consisted of a 2.5% trlethylamine buffer adjusted to pH 4.5 with glacial acetic acid, flowing at 1 ml/min. Sample concentratlon was 4 mg per ml of eluant. Followlng chromatography eaCh set of pooled material from the peak of interest was dialyzed separate]y against doubly-distilled water at 4 ° C. After lyophllization of the supernatants, each o preparation was stored over dessicant at -70 C. until use. Yields of purlfied MBP were calculated as: mg MBP (from major peak)/total mg crude sample applied to the column. Comparison of the two samples was made using two different pH polyacrylamide gel electrophoresis systems (Diebler et al., 1972; Benuck, et al., 1975). Each sample was dissolved in 1.0 M acetlc acid at a concentratlon of 2 mg/ml; urea was added, to a final concentration of 8M; and 20 - 40 ul of each dissolved sample was loaded onto each 5% gel. For the alkaline pH electrophoresis, samples were dissolved at a concentration of 1 mg/ml in 5 m M Tris-glyclne at pH 8.7; sucrose was added, to a final concentratlon of 30% (w/v); and 40 ul was loaded onto each 10% ~el. In both systems, electrophoresis was done at 2.5 mA/ gel for 1-2 hours. Assay of MBP for biological actlvity To test for encephalitogenic activity, each of the two preparations was injected subcutaneously into separate groups of adult guinea plgs, and clinical corre lates of the MBP-induced EAE were observed. Briefly, each of 3 animals in each group received 20 ug MBP via a single 0.i ml injection into each hind footpad of an emulsion which consisted of 0.2 mg ~ P dissolved In 1.0 ml of sterile physiologicaly saline and subsequently emulsified wlth an equal volume of complete Freund's adjuvant (Difco). Animals were observed dally for clinical signs of disease, indicated by weight loss, incontlnence, ataxia, and hindlimb paralysis. Animals were sacrificed either wlthin 24 hours of the onset of paralysis or 26 days post-injection. Brains and spinal cords were fixed in formalin, embedded in paraffin, and stained by hematoxylln and eosin following saglttal sectionlng. Histology was scored as mild, moderate, or severe perivascular monomuclear cell infiltration of the CNS white matter. The relatively low doses of MBP (normal=40-50 ug/animal) were used in an effort to separate the
Vol. 30, No. 12, 1982
HPLC for MBP Purification
991
effects of the two differentially purified preparations of MBP samples. Results Comparisons of HPLC and conventional IE chromatography can be made using at least three independent criteria. One basis for comparison is biochemical purity; another is efficiency of product acquisition; a third and a sometimes neglected area is the biological activity of the material following chromatography. After dlalysis and lyophilization, both methods of chromatography gave comparable yields: 51% from HPLC and 61% from IE. Because of losses associated with dialysis, there is no significant difference between these yields. In addition, the MBP derived by these two methods was identical on both acidurea and Tris-glycine polyacrylamide gels at disparate pH's. Purity was confirmed by the appearance of only a single, narrow band following the application of either 40 or 80 ug samples to each gel. An alternative method was also used to verify the extent of removal of contaminants by HPLC. Material collected from the elution peak was re-injected and chromatographed. Even with increased senslvity, only a slngle, symmetric peak occurred (8.95 min at a flow rate of 0.7 ml/min), and fluctuations in the baseline were less than + 3% of peak absorbance through the 20 min. run. Repeated sample injections at a flow rate of 0.7 ml/min, all yielded identical peak elution times (8.92 + 0.06 min.) with no evidence of protein subsequently trailing off the column. After as many as 15 consecutive 20-min. chromatography runs, elution profiles remained unchanged. Identical profiles were obtained on subsequent days. A major difference in the two methods is time-efficiency. At least a 2day process is involved in IE to repack, equilibrate, and run the column with an average load of 200 mg MBP and a 20 hr elution time. To process this same 200 mg of MBP with the analytic HPLC 1-125 column would require less than 5 hrs (the MBP peak eluted from the HPLC column at 7.35 min. following each injection, as seen in Figure i; the maximum injection volume for the equiment used was 500 ul; concentrations up to at least 20 mg/ml can be applied to the 1-125 column). Other factors which determine preparative efficiency include the flow rate needed for adequate peak separation and the availability of automated in]ectlon and collectlon apparatus. Even with manual injection and without using a preparative column, the efficiency of HPLC is seen to be much greater than that of conventional chromatography. With regard to encephalitogenic potency of the MBP purlfied by each method, the same incidence of encephalomyelitis was observed with both the IE- and HPLC-preparations of purified MBP. ~By day ii, two animals in each group were incontinent. By day 14, one of the two incontinent animals in either group was paralyzed, and by day 19 the third animal of each group had shown at least a 5% weight loss (approx. 20 g) maintained for at least two consecutive days, with subsequent recovery. Histological examination of the brain and spinal cord sections of animals from both groups revealed that comparable pathology had occurred. In each of the animals, there were areas of perivascular cuffing and infiltratlon of mononuclear cells, indicative of EAE. As expected, the animals which had been paralyzed showed the most extensive pathology. Although the triethylamine buffer used to elute MBP from the HPLC column should be readlly removed by dialysis, it was important to verify that triethylamine was not itself capable of facilitating EAE induction. Thus, concentrations of triethylamine up to 100x the concentration used for the column elution were injected into mice. Mice were chosen for these experiments because of the
992
HPLC for MBP Purification
Vol. 30, No. 12, 1982
MBP
0
2
4
6 8 10 12 14 TIME (min)
FIG. ] HPLC separation of myelin basic protein from bovine spinal cord. Injecttion volume, 200~I; Detection, 280nm at 0.20 AUFS; flow, l ml/min; Mobile Phase, TEA acetate. high numbers of animals required and because it has been demonstrated (Montgomery and Rauch, 1979) that bovine MBP can produce both clinical and pathological signs of EAE in mice. Three different strains of mice (SJL/J, BALB/c, and C57BI/6) were tested, with 5 mice in each strain compared against 5 age- and sex-matched controls. Control mice received footpad injections of physiological saline. Although clinical and histological signs of EAE are present in MBP-challenged mice between 12-21 days post-injection, the triethylamine-injected animals showed neither weight loss nor altered behavior over 40 days when compared with their controls. Histologic examination of brain and upper spinal cord sections confirmed the normal state of saline control and triethylamine challenged mice. Conclusion The above results indicate that the MBP purified by HPLC is identical to the MBP purified by ion-exchange chromatography, with respect to their ability to induce autoimmune EAE in guinea pigs. Significant differences in the ease and time efficier~cy o~ HPLC v e r ~ s IE chromatography strongly recommend the HPLC system for large scale purification of the basic protein of myelin.
Vol.
30, No. 12, 1982
HPLC for MBP Purification
993
Acknowledgement This work was supported in part by grants NS 12754 and AI 07118 from the Department of Health and Human Services. References i. 2. 3. 4. 5. 6. 7. 8. 9. i0. ii. 12. 13. 14. 15. 16. 17.
E.C. ALVORD, C.-M. SHAW, S. HRUBY and M.W. KIES, Ann. N.Y. Acad. Sci. 122 333 (1965). E.C. ALVORD, C.-M. SHAW and S. HRUBY, Ann. Neurol. 6 469 (1979). M. BENUCK, N. MARKS and G.A. HASHIM, Eur. J. Biochem. 52 615 (1975). F.C. CHOU, C.-H. CHOU, R.B. FRITZ and R.F. KIBLER, Ann. Neurol. 7 336 (1980). S.R. COHEN, M.J. BURNE, R.M. HERNDON and G.M. MCKAHNN, Adv. Exp. Med. Biol. I00 513 (1978). G.E. DIEBLER, R.E. MARTENSON and M.W. KIES Prep. Biochem. 2 139 (1972). E. FRICK and H. STICKL, J. Neurol. Sci. 46 187 (1980). I.N. MONTGOMERY and H.C. RAUCH, Fed. Proc. 38 1289 (1979). A. NAKAO, W.J. DAVIS and E. EINSTEIN, Biochem. Biophys. Acta. 130 163 (1966). Y. OSHIOR AND E.H. EYLAR, Arc. Biochem. Biophys. 138 392 (1970). H.C. RAUCH and E.R. EINSTEIN, Fed. Proc. 30 306 (1971). H.C. RAUCH and E.R. EINSTEIN, Rev. Neurosc. 1 283 (1974). H.C. RAUCH, Cell, Immunol. 60 220. R.H. SWANBORG, J.E. SWIERKOSZ and R,G. SAIEG, J. Immunol. 112 594 (1974). A.M. WELCH, J.H. HOLDA and R.H. SWANBORG, J. Immunol. 125 186 (1980). V. WICHER, W. OLSZEWSKI and F. MILGROM, Int. Arch. Allergy Appl. Immunol. 59 49 (1979). D.O. WILENBORG, E.A. STATEN and G.F. WITTING, Exp. Neurol. 61 527 (1978).