Degradation of myelin basic protein in myelin by protease in cerebrospinal fluid and effects of protease inhibitors

Neutral protease is shown to be present in cell-free human cerebrospinal fluid. Incubation of heated human myelin with CSF at 25°C resulted in a marked reduction of myelin basic protein (MBP) with time. Degradation products appeared at apparent mol wt 14 KDa and 12 KDa on polyacrylamide gel electrophoresis. Optimal pH of the protease was 7.0. This protease was activated by calcium ion. Degradation of MBP was inhibited by FOY305 (camostat mesilate), Trasylol^®, and Leupeptin, but not a specific calcium-activated neutral protease inhibitor, E-64-a. FOY305, which is a synthesized specific serine protease inhibitor, was the strongest inhibitor of all. The role of this protease in CSF has not been elucidated. In may be related to the physiological turnover of MBP, and may affect myelin maintenance in pathological conditions such as demyelination.     

An improved radioimmunoassay for myelin basic protein-like immunoreactive material in cerebrospinal fluid

A modified radioimmunoassay protocol is described which can measure elevated levels of myelin basic protein-like immunoreactive material in the cerebrospinal fluid of some patients with multiple sclerosis or head injury and in rats developing an acute demyelinating form of experimental allergic encephalomyelitis. The assay uses synthetic peptides that differ in their sequences from natural myelin basic protein for both standard and radioligand, but in apparent contrast to previously published assays serial dilutions of samples produce the expected dose estimates when interpolated from the standard curve. A second radioimmunoassay was produced with high sensitivity and specificity for myelin basic protein peptides with a carboxyl terminus at phenylalanine 89. This assay was used in attempts to detect the myelin basic protein-like immunoreactivity recently reported to occur in human urine. This radioimmunoassay failed to detect specific immunoreactivity in urine samples from control and multiple sclerosis donors. The specificities of both assays were studied using a wide range of synthetic peptides and the importance of this information to the design of future assays is addressed. Our work reinforces that of others in suggesting the complex situation involved in the design of assays for MBP fragments in body fluids. We are willing to distribute the reagents for use in our CSF assay to researchers who request them.     

Preparation and properties of monoclonal antibodies to myelin basic protein and its peptides

Techniques are described which have enabled the production and characterisation of monoclonal antibodies to myelin basic protein. These are shown by enzyme immunoassay to react with six different epitopes. Two of these are to peptide 82–91, a region claimed to be present in the spinal fluid of patients with demyelinating disease. One of these, an IgG_2a, is shown to react only with peptides in which the 91–92 phe-phe bond has broken. The other, an IgM, also reacts with whole myelin basic protein. The IgG_2a antibody is shown to have an affinity suitable for use in immunoassay of peptide 82–91. The enzyme immunoassay procedures described help to minimise the work load involved in the preparation and characterisation of monoclonal antibodies to this protein.     

D-aspartic acid in purified myelin and myelin basic protein

The presence of the biologically uncommon D-isomer of aspartic acid in the white matter of human brains has been reported previously from this laboratory (1). We now report that the level of D-aspartate in human brains is higher in purified myelin than in white matter and is even higher in the myelin basic protein fraction. There also appears to be a difference in the level of D-aspartate found in human brain as compared to bovine brain, possibly a species or age-related difference.     

Deimination of the myelin basic protein decelerates its proteasome-mediated metabolism

Deimination of myelin basic protein (MBP) by peptidylarginine deiminase (PAD) prevents its binding to the proteasome and decelerates its degradation by the proteasome in mammalian cells. Potential anticancer drug tetrazole analogue of chloramidine 2, at concentrations greater than 1 µM inhibits the enzymatic activity of PAD in vitro. The observed acceleration of proteasome hydrolysis of MBP to antigenic peptides in the presence of PAD inhibitor may increase the efficiency of lesion of the central nervous system by cytotoxic lymphocytes in multiple sclerosis. We therefore suggest that clinical trials and the introduction of PAD inhibitors in clinical practice for the treatment of malignant neoplasms should be performed only after a careful analysis of their potential effect on the induction of autoimmune neurodegeneration processes.     

A comparison of HMGB1 concentrations between cerebrospinal fluid and blood in patients with neurological disease

Aims: To determine whether a correlation exists between paired cerebrospinal fluid (CSF) and serum levels of a novel inflammatory biomarker, high-mobility group box 1 (HMGB1), in different neurological conditions.

Methods: HMGB1 was measured in the serum and CSF of 46 neurological patients (18 idiopathic intracranial hypertension [IIH], 18 neurological infection/inflammation [NII] and 10 Rasmussen’s encephalitis [RE]).

Results: Mean serum (±?SD) HMGB1 levels were 1.43?±?0.54, 25.28?±?27.9 and 1.89?±?1.49?ng/ml for the patients with IIH, NII and RE, respectively. Corresponding mean (±?SD) CSF levels were 0.35?±?0.22, 4.48?±?6.56 and 2.24?±?2.35?ng/ml. Both CSF and serum HMGB1 was elevated in NII. Elevated CSF HMGB1 was demonstrated in RE. There was no direct correlation between CSF and serum levels of HMGB1.

Conclusion: Serum HMGB1 cannot be used as a surrogate measure for CSF levels. CSF HMGB1 was elevated in NII and RE, its role as a prognostic/stratification biomarker needs further study.     

Specificity of zinc binding to myelin basic protein

Z²⁺ appears to stabilize the myelin sheath but the mechanism of this effect is unknown. In a previous report we have shown that zinc binds to CNS myelin basic protein (MBP) in the presence of phosphate and this results in MBP aggregation. For this paper we used a solid phase zinc blotting assay to identify which myelin proteins bind zinc. MBP and a 58 kDa band were found to be the major targets of⁶⁵Zn binding. Moreover, using fluorescence, light scattering and electron microscopy we investigated the binding of zinc and other cations to purified MBP in solution. Among the cations tested for their ability to interfere with the binding of zinc, the most effective were cadmium, mercury and copper, but only cadmium and mercury increased the scattering intensity, whereas MBP aggregation was not inhibited by copper ions. Thus, the effect of zinc on the formation of MBP clusters seems to be specific.     

Covalent linkage of phosphoinositides to myelin basic protein: in vitro incorporation of [32P] phosphoinositides to myelin basic protein

We have previously reported that the covalent attachment of phosphoinositides to myelin basic protein (MBP) occurs both in vivo and in vitro [Smith, R. A. et al. (1986) Biochemistry 25:2677-2681; Biochemistry 25:2682-2686; and Biochem. Biophys. Res. Comm. 316:426-432]. Phosphoinositidation of MBP was also detected when [32P] phosphoinositides were incubated with myelin pretreated with Triton X-100 and EGTA. Less than 10% of this covalent linkage of phosphoinositides to MBP survived after acidic treatment (0.1 N HCl at 37 degrees C for 10 min). MBP is predicted to lack sufficient hydrophobicity to bind to membranes as shown by analysis of its amino acid sequence for hydrophobic regions and thus its phosphoinositidation may provide an anchor for this purpose.     

Association of myelin basic protein with detergent micelles

Equilibrium measurements of the binding of central nervous system myelin basic protein to sodium dodecyl sulphate, sodium deoxycholate and lysophosphatidylcholine have been obtained by gel permeation chromatography and dialysis. This protein associates with large amounts of each of these surfactants: the apparent saturation weight ratios (surfactant/protein) being $\text{3.58 ± 0.12}\text{and}\text{2.30 ± 0.15}$ for dodecyl sulphate at ionic strengths 0.30 and 0.10, respectively, $\text{1.34 ± 0.10}$ for deoxycholate (at 0.12 ionic strength) and $\text{4.0 ± 0.5}$ for lysophosphatidylcholine. Binding to the ionic surfactants increases markedly close to their critical micelle concentrations. Sedimentation analysis shows that at 0.30 ionic strength in excess dodecyl sulphate the protein is monomeric. It becomes dimeric when the binding ratio falls below 1 at a free detergent concentration of approximately 0.25 mM: below this concentration much of the protein and detergent forms an insoluble complex. The amount of dodecyl sulphate bound at high concentrations and at both above-mentioned ionic strengths corresponds closely to that expected for interaction of a single polypeptide with two micelles. Variability of deoxycholate micelle size on interaction with other molecules precludes a similar analysis for this surfactant. Association was observed only with single micelles of lysophosphatidylcholine. The results provide strong evidence for dual lipid-binding sites on basic protein and indicate that lipid bilayer cross-linking by this protein may be effected by single molecules.     

Deimination of myelin basic protein. 1. Effect of deimination of arginyl residues of myelin basic protein on its structure and susceptibility to digestion by cathepsin D

The effect of deimination of arginyl residues in bovine myelin basic protein (MBP) on its susceptibility to digestion by cathepsin D has been studied. Using bovine component 1 (C-1) of MBP, the most unmodified of the components with all 18 arginyl residues intact, we have generated a number of citrullinated forms by treatment of the protein with purified peptidylarginine deiminase (PAD) in vitro. We obtained species containing 0-9.9 mol of citrulline/mol of MBP. These various species were digested with cathepsin D, a metalloproteinase which cleaves proteins at Phe-Phe linkages. The rate of digestion compared to component 1 was only slightly affected when 2.7 or 3.8 mol of citrulline/mol of MBP was present. With 7.0 mol of citrulline/mol of MBP, a large increase in the rate of digestion occurred. No further increase was observed with 9.9 mol of citrulline/mol of MBP. The immunodominant peptide 43-88 (bovine sequence) was released slowly when 2.7 and 3.8 mol of citrulline/mol of MBP was present, but it was released rapidly when 7.0 mol of citrulline/mol of MBP was present. The dramatic change in digestion with 7.0 mol of citrulline/mol of MBP or more could be explained by a change in three-dimensional structure. Molecular dynamics simulation showed that increasing the number of citrullinyl residues above 7 mol/mol of MBP generated a more open structure, consistent with experimental observations in the literature. We conclude that PAD, which deiminates arginyl residues in proteins, decreases both the charge and compact structure of MBP. This structural change allows better access of the Phe-Phe linkages to cathepsin D. This scheme represents an effective way of generating the immunodominant peptide which sensitizes T-cells for the autoimmune response in demyelinating disease.     

Modifications of myelin basic protein which occur during its isolation.

Abstract— Chromatography of myelin basic protein (BP) on carboxymethylcellulose gives a pattern of multiple components, of which three are major. Component 1 is considered the unmodified species of BP while component 2 has been found to be modified primarily by deamidation and component 3 by phosphorylation (Chou et al., 1976). ³ 3 The numbering system for the components is that used by DEIBLER & MARTENSON (1973) for guinea pig BP and is preferred over the reverse system of numbering used by CHOU et al. (1976); i.e. components 1, 2 and 3 of DEIBLER & MARTENSON (1973) are the same as components 6, 5 and 4 of CHOU et al. (1976), respectively.
In contrast to BP prepared from tissue delipidated in the standard fashion in chloroform–methanol (CM powder), BP prepared from tissue delipidated first in acetone and then in chloroform–methanol (ACM powder) gave an elution pattern on carboxymethylcellulose characterized by a decrease in component 1 and an increase in the earlier eluting, less basic components. Studies with radiolabelled component 1 showed that this difference in elution patterns was due to the partial conversion of component 1 to less basic components during the extraction of ACM powder at neutral pH. The components derived from component 1 (D2, D3 and D4) were then isolated and subjected to tryptic peptide map analyses and determination of their carboxy-terminal arginine content and content of phosphorus. None of the derived components contained phosphorus but tryptic peptide map analyses did show the presence of two minor peptides, T14M₂ and T20M, previously found in component 2 from CM powder and considered to be the deamidation products of their parent peptides T14 and T20 (Chou et al., 1976). In addition components D3 and D4 were shown to have lost appreciable arginine from their carboxy-termini. Since none of the efforts to reduce enzyme activity in vitro had any appreciable effect on components 2 and 3 it was concluded that phosphorylation probably occurs exclusively in vivo, that deamidation occurs both in vivo and in vitro and that loss of carboxy-terminal arginine occurs exclusively in vitro.     

Immunoreactive LHRH in cerebrospinal fluid: the clinical significance in intracranial diseases

Cerebrospinal fluid (CSF) and plasma levels of luteinizing hormone-releasing hormone (LHRH) were measured by RIA in 46 patients with acute intracranial diseases, ie, cerebral bleeding (group A), cerebral thrombosis (B), head injury (C) and meningitis (D), and the results were compared to those obtained in 21 patients with non-intracranial diseases (group E; control). Immunoreactive LHRH concentrations in CSF (CSF IR-LHRH) of 8 postmenopausal women in group E ranged 1.3 to 6.1 (mean +/- SE: 3.1 +/- 0.6) pg/ml, and those of 5 other women and 8 men with group E ranged 1.0 to 5.6 (3.6 +/- 0.4)pg/ml. In 7 out of 15 patients in group A(7/15), CSF IR-LHRH were above the levels seen in group E. In group B, C and D, CSF IR-LHRH were above the control levels in 9/15, 1/9, 3/7, respectively. The changes in plasma LHRH were not clear in postmenopausal patients in groups A and B. Plasma IR-LHRH in other women and men in group A were above the control levels in 2 out of 9 patients (2/9). Those in groups B, C and D were above the control levels in 3/8, 1/9, 2/7, respectively. Moreover, both plasma and CSF IR-LHRH of 13 patients in group A or B in chronic stage were within the control ranges. In cases observed following the time course, the occasionally increased IR-LHRH in plasma and CSF tended to decrease following the abatement of the diseases.(ABSTRACT TRUNCATED AT 250 WORDS)     

Interaction of myelin basic protein with gangliosides and ganglioside-phospholipid mixtures

The interaction of myelin basic protein with monosialoganglioside GM1 was investigated. It was found that the emission maximum of the tryptophan of the protein is blue-shifted due to the interaction. In mixtures of the monosialoganglioside with phosphatidylcholine, the myelin basic protein induces phase separation of the lipids as inferred from differential scanning calorimetry experiments.     

Interaction of myelin basic protein with micelles of dodecylphosphocholine

Interactions of myelin basic protein (MBP) and peptides derived from it with micelles of dodecylphosphocholine (DPC) and perdeuterated DPC have been studied by proton nuclear magnetic resonance (NMR) at 400 MHz and by circular dichroism (CD). When MBP binds to DPC micelles, it acquires about 18% alpha-helicity. The CD spectra of various peptides derived by cleavage of MBP indicate that a major alpha-helical region occurs in residues 85-99 just before the sequence of three prolyl residues 100-102. From line broadenings by fatty acid spin-labels in the micelles and from changes in chemical shifts, the NMR data identify specific residues in MBP that participate in lipid binding. One such sequence is an alpha-helical region from residues 85 to 95, and others occur around methionine-21 and between residues 117 and 135. The different effects of C5, C12, and C16 spin-labels suggest that some segments of the protein may penetrate beyond the dipolar interfacial region of the micelles into the hydrophobic interior, but no part of the protein is protected by the micelles against rapid exchange of its amide groups with the aqueous environment. Even at a lipid to protein molar ratio of 200/1, most NMR resonances from side chains of amino acid residues are not appreciably broadened, suggesting that much of the polypeptide remains highly mobile.