SEARCH

SEARCH BY CITATION

Keywords:

  • cell adhesion molecule;
  • CNS;
  • cysteine proteases;
  • glial cells;
  • myelin;
  • proteolysis

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The myelin-associated glycoprotein (MAG) is a transmembrane cell adhesion molecule participating in myelin formation and maintenance. Calcium-activated/-dependent proteolysis of myelin-associated glycoprotein by calpain and cathepsin L-like activities has already been detected in purified myelin fractions, producing a soluble fragment, called degraded (d)MAG, characterized by the loss of the transmembrane and cytoplasmic domains. Here, we demonstrate and analyze dMAG formation from pure human brain myelin-associated glycoprotein. The activity never exhibited the high rate previously reported in human myelin fractions. Degradation is time-, temperature-, buffer- and structure-dependent, is inhibited at 4°C and by denaturation of the sample, and is mediated by a trans-acting factor. There is no strict pH dependency of the proteolysis. Degradation was inhibited by excess aprotinin, but not by 1–10 µg/mL aprotinin and was not eliminated by the use of an aprotinin-sepharose matrix during the purification. dMAG formation was not enhanced by calcium, nor inhibited by a wide variety of protease inhibitors, including specific calpain and cathepsin L inhibitors. Therefore, while cysteine proteases may be present in human myelin membrane fractions, they are not involved in dMAG formation from highly purified human brain myelin-associated glycoprotein preparations.

Abbreviations used
AEBSF

4-(2-aminoethyl)benzenesulfonyl fluoride

BSA

bovine serum albumin

CAM

cell adhesion molecule

dMAG

degraded MAG (extracellular domain fragment)

L-MAG

large isoform of MAG

MAG

myelin-associated glycoprotein

MAGCT

cytoplasmic domain of MAG

S-MAG

small isoform of MAG

PBS

phosphate-buffered saline

PMSF

phenylmethylsulfonyl fluoride

SDS–PAGE

sodium dodecyl sulfate–polyacrylamide gel electrophoresis

The 100-kDa, transmembrane, myelin-associated glycoprotein (MAG) (Quarles et al. 1973) is thought to contribute to myelination and long-term myelin maintenance in both the central (CNS) and peripheral (PNS) nervous systems (Schachner and Bartsch 2000). MAG exists as two isoforms, known as small MAG (S-MAG) and large MAG (L-MAG), that are differentially expressed during development through alternative RNA splicing (Frail and Braun 1984), and differ only by the peptide sequences (Arquint et al. 1987; Lai et al. 1987a; Salzer et al. 1987; Fujita et al. 1989; Sato et al. 1989; Spagnol et al. 1989) of their respective C-terminal cytoplasmic domains. The heavily glycosylated extracellular domain of MAG includes five immunoglobulin (Ig)-like regions presenting a strong homology with members of the Ig superfamily of cell adhesion molecules (CAM) (Lai et al. 1987b; Cunningham et al. 1987), of which MAG is a member (Lai et al. 1987b; Salzer et al. 1987). The extracellular domain of MAG was shown to bind to neuronal membranes (Poltorak et al. 1987; Johnson et al. 1989; Sadoul et al. 1990), possibly via neuronal microtubule-associated protein 1B (Franzen et al. 2001). Isoform-specific ligands linking MAG to the glial cell cytoskeleton and, possibly, signal transduction pathways, have been identified for both L-MAG and S-MAG (Kursula et al. 1999, 2001), which are known to be differentially phosphorylated by both serine/threonine kinases and tyrosine kinases (Kirchhoff et al. 1993; Umemori et al. 1994; Kursula et al. 1998, 2000). These observations point to roles for MAG in intercellular and/or intermembrane adhesion and, particularly in the case of L-MAG, in signal transduction. It is probable that these functions rely on the structural integrity of this transmembrane protein, and are therefore likely to be compromised upon the partial, or total loss of either the extracellular or the cytoplasmic domain. Furthermore, the purification of the individual MAG isoforms for the study of isoform-specific properties and functions is strictly dependent on the presence of their respective cytoplasmic domains.

Native human brain MAG is highly susceptible to degradation in purified myelin fractions (Möller 1996), forming a 90-kDa, soluble derivative, known as dMAG, comprising only the extracellular domain of the native protein (Sato et al. 1982). Although the presence of dMAG in the cerebrospinal fluid is thought to be the result of a normal physiological turnover of MAG (Yanagisawa et al. 1985), the activity producing dMAG was shown to be higher in white matter from patients with multiple sclerosis than in healthy brain (Sato et al. 1984a; Möller et al. 1987; Quarles 1989). It was recently shown that dMAG released from damaged brain white matter is able to inhibit axonal regeneration (Tang et al. 2001). It has been suggested that the formation of dMAG is mediated by an endogenous neutral cysteine protease (calpain) found in myelin (Sato et al. 1984b; Yanagisawa et al. 1988), or by a cathepsin L-like activity, also found in myelin (Stebbins et al. 1997, 1998).

In the present study, we report and characterize dMAG formation in highly purified native human brain MAG preparations, and demonstrate that this activity is not mediated by a cysteine protease.

Materials

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The human brain samples were obtained at autopsy from the Oulu Central Hospital (Finland). Permission to use human brain tissue for research purposes was obtained from the Finnish Medico-Legal council (permit number 102/32/200/99). The Protein A Sepharose CL-4B was from Pharmacia (Helsinki, Finland). The aprotinin–agarose matrix, aprotinin, leupeptin, calpain inhibitor peptide, E64d, phenylmethylsulfonyl fluoride (PMSF) and the commercial inhibitor cocktail, the mouse monoclonal anti-HNK-1 IgGs, the horseradish peroxidase-conjugated goat anti-mouse and goat anti-rabbit IgGs were from Sigma (Helsinki, Finland). The rabbit polyclonal anti-MAG cytoplasmic domain (MAGCT) IgGs have been described previously (Heape et al. 1999). The Biomax 30-kDa cut-off centrifugal filter devices were from Millipore Corporation (Bedford, MA, USA).

Purification of MAG

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The brain samples, stored at − 20°C in 1 mm EDTA/H2O, were taken at autopsy from 45- to 88-year-old patients with no known neurological diseases. The time between death and autopsy varied from 10 h to 7 days. Total MAG, a mixture of both isoforms, was purified from human brain using a protocol modified from that described previously for the rat protein (Heape et al. 1999). For each MAG purification, a 2-cm3 block of human brain white matter was homogenized, on ice, in 10 mL (total volume) of solubilization buffer [final concentrations: 10 mm Tris-HCl (pH 8), 140 mm NaCl, 300 mm KCl, 4 mm NaN3, 0.5% Triton X-100, 12 mm sodium deoxycholate, 1 mm EDTA]. The homogenate was mixed on a nutator for 2 h at 4°C, and then diluted to 30 mL with column loading buffer [10 mm Tris-HCl (pH 8), 140 mm NaCl, 4 mm NaN3, 0.5% Triton X-100, 12 mm sodium deoxycholate, 1 mm EDTA], sonicated 4 × 20 s and mixed for an additional 30 min at 4°C. The solubilisate was centrifuged 2 × 15 min at 10 000 g and the supernatant (cleared lysate) was used for further purification. The cleared lysate was diluted to 50 mL with column loading buffer giving final concentrations of 10 mm Tris-HCl (pH 8), 140 mm NaCl, 60 mm KCl, 4 mm NaN3, 0.5% Triton X-100, 12 mm sodium deoxycholate and 1 mm EDTA, and was mixed with 1 mL bed-volume of anti-MAGCT–sepharose matrix overnight at 4°C. The anti-MAGCT–sepharose matrix was poured into a column and was washed with 1 × 5 mL, followed by 4 × 2 mL column loading buffer, 1 × 5 mL and 4 × 2 mL TTS (50 mm Tris-HCl, 500 mm NaCl, 0.1% Triton X-100, 1 mm EDTA) pH 8, 1 × 5 mL and 4 × 2 mL TTS pH 9 and finally 2 × 5 mL TTS pH 8. Bound MAG was eluted from the column with 4050 µL of elution buffer (50 mm triethanolamine, 0.1% Triton X-100, 150 mm NaCl, pH 11.5) and the eluate was immediately neutralized with 450 µL of 1 m Tris-HCl, pH 6.8. The eluted MAG was mixed with 100 µL bed volume of Protein A Sepharose CL-4B in phosphate-buffered saline (PBS) for 2 h at 4°C to remove IgGs leaking from the column. The non-bound fraction was washed 3 times with 12 mL 50 mm Hepes, 0.01% Tween-20, pH 7, by ultrafiltration using a 30-kDa cut-off centrifugal filter device. The sample was concentrated to 200 µL, mixed with 100 µL glycerol (final storage buffer composition: 22 mm Hepes, 0.006% Tween-20, 33% glycerol, pH 7), and stored at − 20°C until use.

In some experiments, as specified, 1 µg/mL aprotinin was present at all steps of the purification. When specified, two treatments of the sample were performed with 2.5 mL bed volume of aprotinin–agarose matrix: the first treatment (1 h, 4°C) after sonication of the brain homogenate, and the second (1 h, 4°C) after clearing the solubilisate. The matrix from the first treatment was eliminated during the centrifugation aiming to clear the solubilisate, and that from the second treatment was removed by passing the whole suspension through an empty, disposable chromatography column, collecting the protease-depleted cleared lysate as the flow-through. All other steps were performed as described above. In these experiments the sonicated homogenate was divided into two equal parts, and one of the latter was used as a non-treated control, undergoing all steps except for the aprotinin-agarose treatment.

MAG degradation experiments

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The stability of purified human MAG samples was studied under different conditions. Duplicate samples were made for all assays; one of the samples was placed at 4°C (control) and the other at 37°C, all other conditions being identical for both samples. Unless stated otherwise, all samples in a given series of experiments were from the same brain sample and MAG purification batch. The degradation of MAG to dMAG was routinely monitored by analyzing the whole assay volume (see below) by non-reducing sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and silver staining and, in some cases, by western blotting with anti-HNK-1 and anti-MAGCT antibodies (see below).

Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

In the mixed stability MAG assay, approximately equal amounts of stable and unstable purified MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, were combined and incubated overnight at 37°C. For a set of similar assays studying the effect of the presence of the unstable MAG on the stable MAG, the unstable MAG was allowed to degrade at 37°C for 48 h prior to the assays. Aliquots (either 2.5, 5, or 7.5 µL) of the predegraded unstable MAG were then mixed with 4.5-µL aliquots of the stable MAG and incubated at 37°C for 6, 12, and 24 h. All of the samples were prepared at the same time and then kept at 4°C until the start of the incubation. The incubations were stagger-started to finish at the same time. In these assays, controls comprised the non-mixed MAG samples (same total amount of protein as used in the mixed assay). The incubations were stopped by adding 5 µL of SDS–PAGE sample buffer (Laemmli 1970), without β-mercaptoethanol, and boiling the samples for 5 min. The relative amounts of MAG and dMAG were determined by analyzing the corresponding pixel densities of the scanned digital images of the silver-stained SDS–PAGE gels, using imagej (version 1.12) software (Wayne Rasband, NIH, USA).

Buffer and pH dependencies

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The buffer and pH dependencies of MAG stability were studied with 50 mm Hepes, 0.01% Tween-20 (Hepes) buffer and 70 mm Bis-Tris, 50 mm Tris, 0.01% Tween-20 (Bis-Tris/Tris) buffer, at pH 6, 7, 8 and 9. The buffer changes were carried out by rinsing the sample three times with a 10-fold excess of the new buffer using an ultrafiltration 30-kDa cut-off centrifugal filter device. Aliquots (10 µL) of the purified human MAG were incubated overnight at 4°C and at 37°C in the specified buffers. Incubations were terminated and the samples were analyzed by SDS–PAGE and silver staining, as above.

Effects of protease inhibitors

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The protease inhibitors employed during this analysis included: 1–50 µg/mL aprotinin, an inhibitor of serine proteases (including trypsin, trypsinogen, chymotrypsin, plasmin, kallikrein and urokinase), 10 µg/mL leupeptin, a reversible inhibitor of trypsin-like serine proteases and most cysteine proteases (including trypsin, papain, cathepsin B and calpain), 2–10 µm calpain inhibitor peptide, a specific inhibitor of calpain activity, 2–10 µm E64d, a synthetic membrane-permeable analogue of E-64, that irreversibly inhibits cysteine proteases (including papain, calpain, cathepsins B and L) and 1 mm PMSF (an inhibitor of cysteine proteases and all serine proteases). Inhibitor mixtures comprising 50 µg/mL aprotinin and 50 µg/mL leupeptin, and 50 µg/mL aprotinin, 10 µg/mL leupeptin and 5 mm PMSF were also tested, as was a commercial inhibitor cocktail containing (final concentrations) 4 µm aprotinin, 105 µm leupeptin, 70 µm E64, 5.2 mm AEBSF (an inhibitor of serine proteases), 180 µm bestatin (an inhibitor of aminopeptidases) and 75 µm pepstatin A (an inhibitor of acid proteases). The protease inhibitors were added from concentrated stock solutions to 10-µL aliquots of purified human MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, and the samples were incubated as above.

Western blotting

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

Following SDS–PAGE as above, the proteins were electrotransferred onto nitrocellulose membranes (Towbin et al. 1979) and, after blocking the membrane with 1% bovine serum albumin (BSA) in PBS, were probed using mouse monoclonal anti-HNK-1 IgGs, which recognize a sulfated carbohydrate epitope on the extracellular domain of human MAG (1 : 4000 dilution), and rabbit polyclonal anti-MAG cytoplasmic domain (MAGCT) IgGs, which recognize the native and denatured cytoplasmic domains of both L-MAG and S-MAG (1 : 2000 dilution). Secondary antibodies were horseradish peroxidase-conjugated goat anti-mouse IgG (1 : 16 000 dilution) and goat anti-rabbit IgG (1 : 8000 dilution). All antibody dilutions were made with 1% BSA in PBS. Immunoreactive proteins were visualized with 0.5 mg/mL 4-chloro-1-naphthol and 0.5 µg/mL H2O2 in PBS.

Characterization of human brain MAG and its degradation products

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

Purified human brain MAG and the products of its degradation were characterized by SDS–PAGE, silver staining and western blotting using both anti-HNK-1 and anti-MAGCT antibodies (Fig. 1a). As expected (Heape et al. 1999), the silver-stained pure native MAG exhibited an apparent molecular weight of approximately 100 kDa following SDS–PAGE under reducing conditions (lane 1), and an increased mobility (∼90 kDa) under non-reducing conditions (lane 2). This mobility shift can be attributed to the reduction of the intrachain disulfide bridges of the extracellular immunoglobulin-like domains in the presence of β-mercaptoethanol, and does not correlate with the appearance of supplementary protein bands, illustrating the purity of the MAG preparation. The protein was identified as intact MAG by western blot analysis with a monoclonal anti-HNK-1 antibody (lane 3), which recognizes a carbohydrate epitope abundantly expressed on the extracellular domain of human MAG, and a polyclonal anti-MAGCT antibody (lane 4) specific for the MAG cytoplasmic domain, and recognizing the entire juxtamembrane cytoplasmic region common to both isoforms, as well as the L-MAG-specific C-terminal region. Figure 1(a) (lanes 5–7) also shows a MAG preparation that has undergone partial, postpurification proteolysis (note that the affinity purification procedure employed in this study relies on the presence of the MAGCT). The intact 90-kDa, HNK-1-positive, MAGCT-positive protein (MAG) has undergone partial degradation, producing an ∼80-kDa, HNK-1-positive, MAGCT-negative band (dMAG), and several smaller bands of 15–20 kDa that are HNK-1-negative and MAGCT-positive and which undoubtedly correspond to proteolytic fragments of the cytoplasmic domains of both isoforms: in the absence of specific antibodies, it was not ascertained as to whether the transmembrane domain was present in any of these fragments, which are not analyzed further in this study. When formed, dMAG is relatively stable compared with the intact protein (Fig. 1b), exhibiting either no further significant degradation, even after extended incubations at 37°C (upper panels), or slow degradation (lower panel).

image

Figure 1. Characterization of purified human MAG and its degradation products. (a) Freshly purified human brain MAG samples in 50 mm Hepes, 0.01% Tween-20, pH 7, were submitted to SDS–PAGE under reducing (lane 1) and non-reducing (lanes 2–7) conditions and were analyzed by either silver-staining (Ag: lanes 1, 2 and 5), or western blotting with monoclonal anti-HNK-1 IgGs (lanes 3 and 6), or polyclonal anti-MAGCT IgGs (lanes 4 and 7). The positions of the molecular weight markers (in kDa) are shown on the left and the relative positions of MAG, dMAG and the MAGCT-positive fragments are indicated on the right. Note the quantitative electrophoretic mobility shift of the 90-kDa band to 100 kDa upon treatment with β-mercaptoethanol (ME). Under non-reducing conditions, stable intact MAG (lanes 2–4) is characterized by the presence of a single 90-kDa band that is both HNK-1- and MAGCT-positive, while unstable MAG samples (lanes 5–7) present variable proportions of intact MAG and an additional 80-kDa band (dMAG) that is HNK-1-positive and MAGCT-negative, as well as several 15–20-kDa fragments that are HNK-1-negative and MAGCT-positive. Note, in lane 5, that the minor bands visible above the intact MAG and between dMAG and the small MAGCT-positive fragments are recognized by neither antibody, and probably correspond to carry-over contaminants from the purification protocol in this sample. The occasional presence of these bands does not correlate with the degree of stability of the MAG sample. (b) Aliquots of predegraded samples in 50 mm Hepes, 0.01% Tween-20, pH 7, containing only minor amounts of intact MAG, were incubated at 37°C for 0, 4, 8, or 24 h and analyzed by silver staining (left panels) and western blotting with anti-HNK-1 IgGs (top right panel). The relative positions of MAG and dMAG are indicated on the left. Note the stability of the dMAG in the sample presented in the upper panels, and the slow, but significant degradation of that in the lower panel, even when intact MAG is no longer detectable.

Download figure to PowerPoint

Inter-sample variability of MAG stability

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

As illustrated by the representative samples in Fig. 2, MAG purified from brain samples taken from different individuals were either stable (a), moderately unstable (b), or highly unstable (c). On the other hand, different MAG preparations (A and B) purified from the same brain exhibited no significant differences in their stability, indicating that the variations of MAG stability from one brain sample to another are not determined by a low reproducibility of the efficiency of the purification protocol. The variation in the stability of MAG samples purified from different brains does not seem to correlate with the age of the individual (not shown), although this would need to be confirmed employing a significantly larger number of brain samples than that employed in this study. Nevertheless, both highly stable and unstable MAG preparations have been purified from subjects over 70 years of age and from middle-aged persons. Similarly, the time between death and autopsy (between 1 and 7 days), after which the brain samples are stored frozen, does not appear to correlate with MAG instability (not shown), as both high and low stabilities have been observed after 5 days.

image

Figure 2. Inter-sample variability of dMAG formation. Independently purified MAG preparations in 50 mm Hepes, 0.01% Tween-20, pH 7 (A and B) from different brain samples (a, b and c) were incubated overnight at 4°C (lanes 1, 3, 5, 7, 9 and 11), or 37°C (lanes 2, 4, 6, 8, 10 and 12) under identical conditions, but in separate assays, and analyzed by SDS–PAGE and silver staining. The positions of intact MAG and dMAG are indicated on the right. The stability of the MAG samples purified from different brains can be high, with no visible degradation (a), or moderate, with partial degradation (b), or low, characterized by a quasi-total degradation (c) to dMAG during the overnight incubation at 37°C. In all cases, different MAG preparations, independently purified from the same brain sample, exhibited very similar stabilities (compare A and B panels for each brain).

Download figure to PowerPoint

A trans-acting factor is implicated in the degradation of human MAG

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The implication of the trans-acting factor in the degradation of MAG to dMAG is indicated by the results of an assay where stable and unstable MAG preparations were incubated together (Fig. 3a). While the stable MAG alone was still intact after an overnight incubation at 37°C, it was completely degraded when incubated under identical conditions in the presence of an unstable MAG preparation. These results were confirmed in a similar mixed degradation assay, where a stable MAG preparation was incubated in the presence of increasing amounts of a predegraded partially unstable MAG preparation (Fig. 3b). No significant change in the MAG/dMAG ratio was observed during the 24 h incubation when the predegraded and fresh stable MAG preparations were incubated alone. However, MAG degradation was clearly observed in the mixed assays. A small dose effect of the unstable MAG on the degradation in the mixed assays is also apparent. The low amplitude of this effect may be attributed to the ongoing further degradation of the dMAG during the assay (not shown).

image

Figure 3. Trans-acting factor in MAG degradation. (a) Equal amounts (total protein) of unstable (U) and stable (S) MAG preparations were incubated alone, or in a 1 : 1 mixture (M), overnight at 37°C. The control (ctr) was an identical mixture incubated overnight at 4°C. All samples were analyzed by SDS–PAGE and silver staining. Note the total degradation of the MAG in the mixed sample (M). (b) A partially unstable MAG was predegraded for 48 h at 37°C prior to the assays. Aliquots (2.5, 5 or 7.5 µL) of the predegraded unstable MAG (P) were then mixed with 4.5-µL aliquots of fresh stable MAG (F) in 50 mm Hepes, 0.01% Tween-20, pH 7 (mix 1, mix 2 and mix 3, respectively), and incubated at 37°C for 6 h, 12 h, and 24 h. All of the samples, including a 4°C 24-h control, were prepared at the same time and then kept at 4°C until the start of the incubations. The incubations were stagger-started to finish at the same time. The controls (triplicate assays) comprised the fresh stable MAG and the predegraded unstable MAG on their own. The relative amounts of MAG and dMAG were determined by analyzing the corresponding pixel densities of the scanned digital images of the silver-stained SDS–PAGE gels, using ImageJ software. No significant change in the relative amount of MAG is observed when the predegraded and fresh stable MAG preparations were incubated alone, but MAG degradation was clearly observed in the mixed assays. The results are expressed as the ratio MAG/(MAG + dMAG)%. The values for the predegraded and fresh MAG samples are given as means ± standard deviation of triplicate assays.

Download figure to PowerPoint

Structure and temperature dependency and rate of MAG degradation

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The structure dependency of dMAG formation was demonstrated in overnight degradation assays employing native and denatured human MAG (Fig. 4a). No dMAG was observed in the denatured MAG sample (boiled in SDS sample buffer), while minimal degradation is observed in the native sample incubated at 4°C. dMAG formation was significantly increased when the native protein was incubated at 37°C. The degradation was also temperature dependent (Fig. 4a,b), being slower at room temperature than at 37°C. At 4°C, in 50 mm Hepes buffers, purified human MAG was stable (displaying minimal degradation) for at least a month, while an identical sample incubated at room temperature displayed a higher level of dMAG formation. As shown in Fig. 2, the rate of dMAG formation varies from one brain to another. However, even for the highly unstable MAG samples, the half-reaction time was found to be several hours at 37°C (Fig. 4c), contrasting sharply with the values reported earlier (∼5 min) in purified human myelin preparations (Möller 1996).

image

Figure 4. Structure and temperature dependency and rate of dMAG formation. (a) Equal aliquots of native (N) and denatured (D: boiled in the presence of non-reducing SDS sample buffer) intact human MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, were incubated overnight at 4°C (N), or 37°C (N and D), and analyzed by SDS–PAGE and silver staining. The positions of intact MAG and dMAG are indicated on the right. No dMAG formation is observed in the denatured MAG sample. (b) Equal aliquots of a relatively stable sample of native MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, were incubated overnight (left panel) or for 1 month (right panel), at 4°C, room temperature (RT), or 37°C, and analyzed by SDS–PAGE and silver staining. The control (C) was an identical aliquot boiled in the presence of SDS sample buffer prior to being incubated overnight at 4°C. The positions of intact MAG and dMAG are indicated on the right. In the overnight samples, trace degradation is visible at room temperature, but is clear at 37°C. After 1 month, dMAG formation is detectable at 4°C and is significantly increased at room temperature. (c) Equal aliquots of a highly unstable human brain MAG in 50 mm Hepes, 0.01% Tween-20, pH 7 (sample (c) in Fig. 2), were incubated for 1, 2, 4, or 20 h at 37°C and analyzed by SDS–PAGE and silver staining. The control (C) was an identical aliquot incubated for 20 h at 4°C. The positions of intact MAG and dMAG are indicated on the right. The time required for half of the intact MAG to be degraded to dMAG can be estimated at approximately 3 h.

Download figure to PowerPoint

The effect of buffer and pH on MAG stability

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

Using a partially stable MAG preparation, we investigated the effect of the pH (6, 7 and 8) on dMAG formation during an overnight incubation at 4°C and 37°C in 50 mm Hepes buffers (Fig. 5a, left panels). As expected, no degradation was observed at 4°C, while partial degradation to dMAG was observed at 37°C. The pH had no significant effect on the degradation of the intact MAG, nor on the amount of dMAG formed. It should be noted that the useful buffering range of Hepes is pH 6.8–8.2, so the buffering efficiency in the sample at pH 6 can be expected to be quite low. As others have shown that dMAG formation in purified myelin exhibits pH-dependent variations, we repeated the experiments above using the same MAG preparation and a wide-range Bis-Tris/Tris buffer system (useful buffering range: pH 6–9). Again, no dMAG formation was observed in the samples incubated at 4°C (Fig. 5a, right panels). However, at 37°C, the degree of degradation of the intact MAG was significantly greater than that observed in Hepes buffers at the same pHs, as almost no intact MAG was detected: trace amounts were visible only in the sample at pH 8. Note that, at both 4°C and 37°C, there appears to be less protein in the samples buffered at pH 6. In order to further test the lower stability in the Bis-Tris/Tris buffers, we repeated the overnight degradation assays above using a MAG preparation that was shown to be almost completely stable in neutral Hepes buffer (Fig. 5b). Almost total degradation was observed in the Bis-Tris/Tris buffers at all pHs at 37°C. In addition, there is clearly less protein in the pH 6 samples at both 4°C and 37°C (see also Fig. 5a), suggesting that the buffer may be inducing a pH-dependent precipitation of both MAG and its degradation products. We tested this by incubating another relatively stable MAG sample overnight, at 4°C and 37°C, in Bis-Tris/Tris buffers ranging from pH 6–9 (Fig. 5c). For the samples incubated at both 4°C and 37°C, there was a clear pH-dependent decrease in the amount of MAG detected in the silver-stained gel, with no concomitant increase of low molecular weight bands. Some degradation to dMAG can be seen in the samples incubated at 37°C, but there does not seem to be any pH-dependency of the amount formed. The results indicate that there is a pH-dependent precipitation of MAG in Bis-Tris/Tris buffers, but not in Hepes buffer, and that intact MAG may be more prone to precipitation than dMAG.

image

Figure 5. Effect of buffer type and pH on MAG degradation. Equal aliquots of pure human brain MAG that were shown to be partially stable (a), or highly stable (b) in Hepes buffer at pH 7, were incubated overnight at 4°C, or 37°C in 50 mm Hepes, 0.01% Tween-20, or 70 mm Bis-Tris, 50 mm Tris, 0.01% Tween-20 at pH 6, 7, or 8, as indicated. The relative positions of MAG and dMAG are indicated by arrows. Note the absence of a clear effect of pH on dMAG formation in either buffer type, the increased degradation in Bis-Tris/Tris buffer compared with that in Hepes buffer, and the disappearance of both MAG and dMAG in Bis-Tris/Tris buffer at pH 6, at both 4°C and 37°C. (c) Equal aliquots of stable MAG (as assessed in Hepes buffer at pH 7) were incubated overnight at 4°C, or 37°C in 70 mm Bis-Tris, 50 mm Tris, 0.01% Tween-20 at pH 6, 7, 8, or 9, as indicated. Note the temperature-dependent dMAG formation, and the pH-dependent disappearance of the total protein at 4°C and 37°C.

Download figure to PowerPoint

The effect of calcium and EDTA on MAG degradation

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

Calpains have been proposed as being responsible for dMAG formation in human brain myelin. In neutral Hepes buffer, dMAG formation from a relatively stable MAG sample was not enhanced by the presence of 1 mm CaCl2(Fig. 6a), suggesting that, if calpain is present in the purified MAG preparation, it has already been totally activated prior to the purification (EDTA is present during all steps preceding the anti-MAGCT affinity column pre-elution washes, and Ca2+ is not added). Likewise, dMAG formation from a more unstable MAG sample, which had already undergone partial degradation prior to the assay, was not enhanced in neutral Tris buffer containing 2 mm CaCl2, nor was it inhibited by the addition of 2 mm EDTA (Fig. 6b). These results suggest that, if present, the copurifying calpain activity is calcium-activated, rather than calcium-dependent, and has been totally activated prior to the purification.

image

Figure 6. Effect of calcium and EDTA on MAG degradation. (a) Equal aliquots of purified human MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, were incubated at 4°C, or 37°C for 2, 4, or 20 h with (right panel), or without (left panel) 1 mm CaCl2, and analyzed by SDS–PAGE and silver staining. The positions of intact MAG and dMAG are indicated on the right. 1 mm CaCl2 had no effect on the degradation rate of purified human MAG. (b) Equal aliquots of partially degraded human MAG in 10 mm Tris-buffered saline, 0.01% Tween-20, pH 8, were incubated overnight at 37°C with, or without 2 mm EDTA, and/or 2 mm CaCl2, as indicated, and analyzed by SDS–PAGE and silver staining. A control sample free of EDTA and CaCl2 was incubated overnight at 4°C. The positions of intact MAG and dMAG are indicated on the right. Neither 2 mm EDTA, nor 2 mm CaCl2 had any effect on the degradation.

Download figure to PowerPoint

The effect of protease inhibitors on MAG degradation

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The effects of several broad-range and specific protease inhibitors on MAG degradation were tested in an attempt to characterize the proteolytic activity responsible for dMAG formation.

When aprotinin, an inhibitor of serine proteases, was employed at normal effective concentrations (1–10 µg/mL), it did not prevent dMAG formation (Fig. 7a), although the rate of degradation appeared to be very slightly reduced under these conditions. However, when used at high concentrations (50 µg/mL; Fig. 7b, left panel), aprotinin did inhibit dMAG formation. In order to test whether aprotinin-sensitive proteases may reduce the yield of MAG affinity-purified from human brain with the anti-MAGCT-agarose matrix, we compared the stabilities of the MAG obtained from parallel purifications, using the same brain sample, performed in the absence of aprotinin and when 1 µg/mL aprotinin was present in all solutions employed prior to the elution from the anti-MAGCT-agarose column: no significant differences were observed (not shown). In a similar experiment, we included two aprotinin-agarose affinity steps prior to the anti-MAGCT affinity step in an attempt to remove copurifying serine proteases from the lysates: there was no significant effect on the yield of the final product, and the presence of similar amounts of both MAGCT fragments and dMAG, visible on the immunoblots, indicates that degradation had occurred in both samples during, or after, the anti-MAGCT affinity step of the purification protocol (Fig. 7b, right panel). The use of 50 µg/mL leupeptin, a cysteine and trypsin-like serine protease inhibitor, alone or in the presence of 50 µg/mL aprotinin, had no effect on dMAG formation (Fig. 7b, left panel); the inhibition observed in lane 5 could be attributed to the presence of the excess aprotinin.

image

Figure 7. Effect of inhibitor mixtures and serine protease inhibitors on the degradation of human MAG. (a) Equal aliquots of purified human MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, were incubated at 4°C, or 37°C for 1, 2, 4, or 20 h, as indicated, with (right panel) or without (left panel) 10 µg/mL (upper panel), or 1 µg/mL (lower panel) aprotinin, and analyzed by SDS–PAGE and silver staining. The control sample (C) was denatured and incubated at 4°C for 20 h. The positions of intact MAG and dMAG are indicated on the right. Neither concentration prevented dMAG formation. (b; left panel) Equal aliquots of purified human MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, were incubated overnight at 37°C with, or without 50 µg/mL aprotinin, or 50 µg/mL leupeptin, or both, and analyzed by SDS–PAGE and silver staining. The control sample was incubated overnight at 4°C in the absence of inhibitor. The positions of intact MAG and dMAG are indicated on the left. dMAG formation was only inhibited in the presence of an excess of aprotinin, while leupeptin had no effect. (b; right panel) Parallel purifications of human MAG were performed from the same brain solubilisate, with (+) or without (–) the inclusion of two rounds of aprotinin-agarose affinity chromatography, as described in Materials and methods. Equal aliquots of the final products in 50 mm Hepes, 0.01% Tween-20, pH 7, of both purifications were analyzed by SDS–PAGE and silver staining (left), and by western blotting with anti-HNK-1 (centre) or anti-MAGCT (right) antibodies. The positions of intact MAG, dMAG and MAGCT are indicated on the right. The use of aprotinin-agarose matrix in the purification protocol had no effect on the yield, nor on the stability of the final product. (c; left panel) Equal aliquots of unstable purified human MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, were incubated at 4 or 37°C for 4 or 20 h with (+) or without (–) 1 mm PMSF, as indicated, and analyzed by SDS–PAGE and silver staining. The control sample (0) was denatured and incubated for 20 h at 4°C. The positions of intact MAG and dMAG are indicated on the left. PMSF alone had no significant effect on the degradation. (c; right panels) Partially stable human MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, was incubated at 4°C, or 37°C for 20 h, in the presence of an inhibitor mix containing 50 µg/mL aprotinin, 10 µg/mL leupeptin and 5 mm PMSF, and analyzed by SDS–PAGE and western blot using anti-HNK-1 and anti-MAGCT antibodies, as indicated. The position of MAG and dMAG are indicated on the right. The inhibition by the protease inhibitor mix can be attributed to the presence of the excess aprotinin (compare with the silver-stained gels in b). (d) Equal aliquots of partially stable human MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, were incubated 20 h at 4°C, or 37°C in the presence (+), or absence (–) of a commercial protease inhibitor cocktail containing 5.2 mm AEBSF, 4 µm aprotinin, 105 µm leupeptin, 180 µm bestatin, 75 µm pepstatin A and 70 µm E64, as indicated, and analyzed by SDS–PAGE and silver staining. The positions of intact MAG and dMAG are indicated on the left. dMAG formation was not significantly affected.

Download figure to PowerPoint

PMSF (1 mm), another serine and cysteine protease inhibitor, had no effect on MAG degradation when used alone in an unstable sample (Fig. 7c, left panel). When an inhibitor mixture containing 50 µg/mL aprotinin, 10 µg/mL leupeptin, and 5 mm PMSF was employed with a more stable MAG preparation (Fig. 7c, right panel), dMAG formation was inhibited, probably due to the presence of the excess aprotinin, as also seen in Fig. 7(b). A more complex inhibitor cocktail, containing 4 µm aprotinin, 105 µm leupeptin, 5.2 mm AEBSF, 180 µm bestatin, 75 µm pepstatin A and 70 µm E64, had no effect on dMAG formation in a relatively stable MAG preparation (Fig. 7d).

As calpains and a cathepsin L-like activity have been proposed as being responsible for dMAG formation in purified brain myelin fractions, we tested the abilities of E64d, an irreversible cysteine protease inhibitor active towards calpains and cathepsin L, and calpain inhibitor peptide, a specific inhibitor of calpain activity, to inhibit the formation of dMAG in highly purified stable and unstable human brain MAG preparations. Both inhibitors were employed at 2 µm (unstable MAG) and 10 µm (stable MAG), representing the normal effective concentration ranges for these substances. Neither E64d, nor calpain inhibitor peptide had any noticeable effect on dMAG formation in either MAG sample (Fig. 8).

image

Figure 8. Effect of specific cysteine protease inhibitors, E64d and calpain inhibitor peptide (CIP), on the degradation of human MAG. Equal aliquots of unstable (upper panels), or partially stable (lower panels) MAG in 50 mm Hepes, 0.01% Tween-20, pH 7, were incubated at 4°C, or 37°C, for 1, 2, 4, or 20 h with 2 µm (upper panels), or 10 µm (lower panels) E64d (middle panels), or CIP (right panels), or without any protease inhibitor (left panels), as indicated, and analyzed by SDS–PAGE and silver staining. The positions of MAG, dMAG and MAGCT are indicated on the right. Neither E64d, nor CIP had any significant effect on the degradation of either of the MAG samples.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

Using purified myelin fractions, it has been shown repeatedly that the 100-kDa transmembrane MAG is prone to proteolysis, producing a large, relatively stable, 90-kDa fragment, comprising the entire extracellular domain, referred to as dMAG (Sato et al. 1982). It should be noted that these molecular weights refer to estimations made by SDS–PAGE under reducing conditions. In non-reducing SDS–PAGE, as employed in this study, these sizes are ∼90 and ∼80 kDa for intact MAG and dMAG, respectively (see Fig. 1). While the production of dMAG has received considerable attention, the fate of the transmembrane and cytoplasmic domains in the course of the degradation process has been totally ignored, thus depriving us of information that is potentially important for identifying the mechanism by which this degradation occurs. Nevertheless, the degradation of MAG to dMAG in human brain myelin was reported to be optimal at neutral pH, accelerated by millimolar concentrations of CaCl2, inhibited in buffers containing more than 100 µm EGTA, and partially inhibited by E-64, a specific cysteine protease inhibitor, suggesting that the activity was a calcium-activated neutral protease, calpain (Sato et al. 1984b), found in myelin membranes (Yanagisawa et al. 1988). Not unreasonably, it has been assumed that the primary cleavage site targeted by this proteolytic activity is the one producing the largest detectable fragment, dMAG. Based on the amino acid sequence of the C-terminal portion of this fragment, and the extracellular localization of cathepsin L, it was concluded that the activity responsible for dMAG formation is not that of a calpain, but rather a cathepsin L-like protease also found in myelin fractions, cleaving MAG at the extracellular domain/transmembrane domain junction, between residues Ala512 and Lys513 (Stebbins et al. 1997).

In the present study, we too demonstrate dMAG formation, but this time from highly pure, non-denatured, native, human brain MAG, in the absence of added enzyme. However, our results, many of which differ significantly from those reported in the earlier studies, clearly demonstrate that the dMAG formation observed in our experiments with pure human MAG is mediated by neither calpain nor cathepsin L activities.

Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The immunoaffinity purification of the native MAG employed in the present study relies entirely upon the presence of the cytoplasmic domain (MAGCT). Thus, any dMAG formed prior to the elution of the protein from the anti-MAGCT affinity column will not be retained by the column. Furthermore, as MAG has been reported not to bind to itself (Fahrig et al. 1987), it is unlikely that dMAG will copurify with the intact protein under the stringent wash conditions employed here. Consequently, all dMAG detected during the subsequent analyses is the result of postpurification proteolysis. This is obviously not the case for fragments including parts of the MAGCT, which may be carried over from all stages of the purification. However, if smaller than 30 kDa, these fragments should be largely eliminated during the final ultrafiltration steps of the purification protocol.

Of the nine possible glycosylation sites (Burger et al. 1993), all but one are at least partially glycosylated and apparently possess the L2/HNK-1 epitope. In view of the relatively regular distribution of these sites along the whole extracellular domain, all proteolytic fragments of more than 15–20 kDa and comprising a part of the extracellular domain would be likely to possess at least one HNK-1 epitope that can be detected by western analysis. The first 36 juxtamembrane amino acids of the two MAG isoform cytoplasmic domains are identical, and this common region, as well as the L-MAG-specific C-terminal, are recognized by the polyclonal anti-MAGCT antibodies employed here (Heape et al. 1999). Thus, for the purposes of this study, intact MAG is defined as the ∼90-kDa (non-reducing SDS–PAGE) protein band, which is both HNK-1 and MAGCT-positive as judged by western analysis, while dMAG is defined as the ∼80-kDa, HNK-1-positive, MAGCT-negative band, and the cytoplasmic domain-containing degradation products include all fragments with electrophoretic mobilities corresponding to sizes of ∼20 kDa or less, and that are HNK-1-negative and MAGCT-positive (see Fig. 1). In view of the results obtained earlier by other groups, and particularly by Stebbins et al. (1997), it is assumed that the transmembrane domain of the proteolysed protein is associated with the MAGCT-positive fragments. These definitions, as well as the properties exhibited by the detected molecular species, are compatible with the properties of dMAG reported in the earlier studies.

While there are only two detectable HNK-1-positive species (MAG and dMAG), and only one high molecular weight MAGCT-positive species (MAG), there are always four small, strongly MAGCT-positive, HNK-1-negative fragments. Two of these probably represent the transmembrane/cytoplasmic domains of the two MAG isoforms, L-MAG and S-MAG. The two other MAGCT-positive fragments could either represent subpopulations of L-MAG and S-MAG transmembrane/cytoplasmic domain fragments with differences in post-translational modifications, such as phosphorylation, or products of further degradation of one, or both, of the primary MAGCT-positive proteolytic products. As the dMAG is relatively stable, with little or no further degradation occurring during extended incubations at 37°C (see Fig. 1b), the latter possibility suggests the presence of a second proteolytic activity copurifying with the intact MAG and acting specifically on the MAGCT fragments.

Characterization of the formation of dMAG from native human MAG

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References

The observation that the stability of the MAG, as defined by its susceptibility to be degraded to dMAG, varies from brain to brain (Fig. 2), raises the possibility of the existence of destabilizing polymorphisms similar to that observed for the rat S-MAG, where an Arg-to-Pro substitution results in an increased rate of degradation of the S-MAG protein (Kursula et al. 1998). However, a recent study of nucleotide dimorphisms in the human MAG gene of several hundred Italian subjects revealed only one individual with an amino acid change in the MAG protein (D'Alfonso et al. 2002). We can therefore rule out the possibility that amino acid dimorphisms are responsible for the intersample variability of human MAG stability. This is further confirmed by the sample mixing assays (Fig. 3), where stable MAG preparations were degraded after mixing them with unstable MAG preparations, demonstating the direct involvement of a trans-acting factor in dMAG formation. Furthermore, as the proteolysis is also time-, temperature- and structure-dependent (Fig. 4, and Sato et al. 1982), and variable from one human brain to another, we cannot exclude the possibility that trace amounts of proteases may copurify with the MAG. If this is the case, the amount of active protease that copurifies with the MAG is not dependent on the efficiency, or reproducibility of the purification protocol itself, as separate purifications from the same brain sample yield a MAG with remarkably similar stabilities. This implies that different brain samples contain different levels of protease and/or that there is a specific association of the proteolytic activity, whatever its form, with MAG. We noticed no obvious correlation between the time elapsed between death and autopsy (1–7 days), when the brain samples are taken and frozen, or the age of the subject (45–80 years), and the stability of our purified MAG samples (not shown). However, an influence of these parameters cannot be definitively ruled out, because the relatively small number of brain samples available for this study (∼ 15) did not permit us to perform a significant statistical analysis. On the other hand, dMAG formation in purified rat brain myelin was shown not to be affected by a postmortem time up to 15 h (Sato et al. 1982).

The rate of the MAG to dMAG conversion has been studied in myelin fractions isolated from several species, revealing that the time required for a 50% conversion of MAG to dMAG was 18–24 h in rodents and bovine species, 10 min (chimpanzees) to 2 h in non-human primates, and only 5–20 min in humans (Sato et al. 1982; Möller 1996). Our results with pure human MAG clearly differ from the estimates made using myelin fractions, as, even in the most unstable pure MAG samples, the half-reaction time was about 3 h (see Figs 4, 7 and 8), which is comparable to the myelin fraction values obtained for rhesus monkeys and gorillas (Möller 1996). This could of course be interpreted to be the result of a partial elimination of the protease during our MAG purification procedure.

An important factor to be taken into account when considering the participation of calpain and cathepsin activities in the degradation of MAG to dMAG is the pH-dependency of the reaction. While the former display optimal activity at neutral pH, the latter are acid proteases that are, in most cases, including mammalian cathepsin L, rapidly and irreversibly inactivated at pHs above neutral (Bromme et al. 1993; Nomura et al. 1996). While the secreted form, procathepsin L, is relatively stable for several hours at pH 8.5 and 37°C (Nomura and Fujisawa 1997), the protease activity after activation, measured at acid pH, is reduced by 75% after 6 h under these conditions, and is abolished after 24 h. These data alone diminish, but do not exclude, the likelihood that a cathepsin L-like activity is responsible for the dMAG formation observed here, as the large majority of the purification protocol, lasting 24 h at 4°C, is performed at pHs equal to, or greater than 8, and which never descend under pH 7: procathepsin L is not processed to the active form under these conditions (Nomura and Fujisawa 1997). The non-involvement of cathepsin L-like activity is also indicated by the observations that dMAG formation occurred even at pH 8 in Hepes and Tris-based buffers, and that the major effect of decreasing the pH, at least in Tris buffers, was to induce what appeared to be the precipitation of MAG. Although degradation was observed, MAG was significantly more stable in Hepes buffers, in which pH had no noticeable effect on dMAG formation and which did not induce a pH-dependent precipitation of the protein, even outside of its useful buffering range (see Fig. 5). This suggests that dMAG formation from pure human MAG is more buffer type-dependent than pH-dependent, which may partly explain the faster rates observed in the myelin fractions, where 0.2–0.4 m ammonium bicarbonate buffers were employed (Sato et al. 1982; Möller 1996). In the latter buffers, dMAG formation exhibited a broad pH optimum in the neutral range, leading the authors to suggest the involvement of a neutral protease, and the presence of salts, including KCl, NaCl, Tris-HCl and CaCl2, was reported to accelerate the degradation (Sato et al. 1982). With the exception of the Bis-Tris/Tris-HCl buffers mentioned above, we observed no effect of any of these salts on dMAG formation, as shown in Fig. 6 for CaCl2 in both 50 mm Hepes and 10 mm Tris-buffered saline. In a subsequent study, Sato et al. (1984b) reported the stimulation of dMAG formation by CaCl2 in a purified human myelin fraction (in 10 mm Tris-HCl), its inhibition by EGTA, a divalent cation chelator, and its partial inhibition by 10 µm E64a, an analogue of the specific, irreversible inhibitor of cysteine proteases, E64. It was therefore suggested that the protease responsible for dMAG formation was a calcium-activated neutral cysteine protease, or calpain. Here again, our results using pure human MAG differ. First, CaCl2 had no stimulatory effect, in either Tris-HCl or in Hepes buffers, and EDTA, which also chelates Ca2+, did not inhibit it. This implies that, if present, the calpains are already fully activated. Second, neither 70 µm E64, included in a commercial inhibitor cocktail (Fig. 7), nor 2–10 µm E64d alone (Fig. 8), a membrane-permeable E64 analogue, had any effect whatsoever on dMAG formation in neutral Hepes buffer. Furthermore, calpain inhibitor peptide, a strong specific inhibitor of calpain I and II, also failed to inhibit dMAG formation.

Taken together, these observations lead us to conclude that, while both calpains and cathepsin L (also strongly inhibited by E64) are present in myelin and, at least in the case of calpains, may be capable of degrading MAG to dMAG in vitro, significant levels of dMAG formation can, and do also occur via a mechanism that is independent of cysteine protease.

In an attempt to identify the nature of the putative protease that may be copurifying with the human brain MAG, we also tested the effects of several protease inhibitors with a broad specificity, either alone, or as components of inhibitor cocktails. With the sole exception of the serine protease inhibitor aprotinin, we observed no significant effect of any inhibitor on dMAG formation (see Fig. 7), correlating well with the earlier results reported by Sato et al. (1982). Indeed, in our study, an inhibitory effect was only observed when the degradation assay was performed in a large excess of aprotinin. However, as the yield of intact MAG was not improved and degradation could not be eliminated, or diminished, by the presence of aprotinin during the entire purification protocol, nor by the removal of aprotinin-sensitive proteases with two rounds of aprotinin-agarose affinity chromatography, we conclude that the inhibitory effect of the high aprotinin concentration was non-specific. In support of this, it has been shown that, while MAG is partially degraded by trypsin and plasmin, this degradation does not lead to the formation of dMAG (Inuzuka et al. 1984).

In summary, we have demonstrated that highly purified human MAG is specifically degraded to fragments identified as comprising the extracellular domain (dMAG) and cytoplasmic domains (fragments < 20 kDa) of the protein. Due to its structure- and temperature-dependent nature, and its transmissibility from one sample to another, this degradation may indeed be due to a protease activity. However, our results demonstrate that dMAG formation in this experimental paradigm is not mediated by a calpain, or cathepsin L-like activity as suggested earlier (Sato et al. 1984b; Stebbins et al. 1997). Furthermore, in view of the spectrum of protease inhibitors tested, it does not seem likely that the degradation is mediated by any kind of serine or cysteine protease.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Materials
  5. Purification of MAG
  6. MAG degradation experiments
  7. Intersample variability
  8. Mixed degradation assays with stable and unstable MAG, and preparation of predegraded MAG
  9. Degradation rate
  10. Buffer and pH dependencies
  11. Effect of calcium ions
  12. Effects of protease inhibitors
  13. SDS–PAGE
  14. Western blotting
  15. Results
  16. Characterization of human brain MAG and its degradation products
  17. Inter-sample variability of MAG stability
  18. A trans-acting factor is implicated in the degradation of human MAG
  19. Structure and temperature dependency and rate of MAG degradation
  20. The effect of buffer and pH on MAG stability
  21. The effect of calcium and EDTA on MAG degradation
  22. The effect of protease inhibitors on MAG degradation
  23. Discussion
  24. Characterization of the purified human MAG and its proteolytic products, dMAG and MAGCT
  25. Characterization of the formation of dMAG from native human MAG
  26. Acknowledgements
  27. References
  • Arquint M., Roder J., Chia L. S., Down J., Wilkinson D., Bayley H., Braun P. and Dunn R. (1987) Molecular cloning and primary structure of myelin-associated glycoprotein. Proc. Natl Acad. Sci. USA 84, 600604.
  • Bromme D., Bonneau P. R., Lachance P., Wiederanders B., Kirschke H., Peters C., Thomas D. Y., Storer A. C. and Vernet T. (1993) Functional expression of human cathepsin S in Saccharomyces cerevisiae. Purification and characterization of the recombinant enzyme. J. Biol. Chem. 268, 48324838.
  • Burger D., Pidoux L. and Steck A. J. (1993) Identification of the glycosylated sequons of human myelin-associated glycoprotein. Biochem. Biophys. Res. Commun. 197, 457464.
  • Cunningham B. A., Hemperly J. J., Murray B. A., Prediger E. A., Brackenbury R. and Edelman G. M. (1987) Neural cell adhesion molecule: structure, immunoglobulin-like domains, cell surface modulation, and alternative RNA splicing. Science 236, 799806.
  • D'Alfonso S., Mellai M., Giordano M., Pastore A., Malferrari G., Naldi P., Repice A., Liguori M., Cannoni S., Milanese C., Caputo D., Savettieri G. and Momigliano-Richiardi P. (2002) Identification of single nucleotide variations in the coding and regulatory regions of the myelin-associated glycoprotein gene and study of their association with multiple sclerosis. J. Neuroimmunol. 126, 196204.
  • Damerval C., Guilloux M., Blaisonneau J. and Vienne D. (1987) A simplification of Heukeshoven and Dernick's silver staining of proteins. Electrophoresis 8, 158159.
  • Fahrig T., Landa C., Pesheva P., Kuhn K. and Schachner M. (1987) Characterization of binding properties of the myelin-associated glycoprotein to extracellular matrix constituents. EMBO J. 6, 28752883.
  • Frail D. E. and Braun P. E. (1984) Two developmentally regulated messenger RNAs differing in their coding region may exist for the myelin-associated glycoprotein. J. Biol. Chem. 259, 1485714862.
  • Franzen R., Tanner S. L., Dashiell S. M., Rottkamp C. A., Hammer J. A. and Quarles R. H. (2001) Microtubule-associated protein 1B: a neuronal binding partner for myelin-associated glycoprotein. J. Cell Biol. 155, 893898.
  • Fujita N., Sato S., Kurihara T., Kuwano R., Sakimura K., Inuzuka T., Takahashi Y. and Miyatake T. (1989) cDNA cloning of mouse myelin-associated glycoprotein: a novel alternative splicing pattern. Biochem. Biophys. Res. Commun. 165, 11621169.
  • Heape A. M., Lehto V. P. and Kursula P. (1999) The expression of recombinant large myelin-associated glycoprotein cytoplasmic domain and the purification of native myelin-associated glycoprotein from rat brain and peripheral nerve. Protein Exp. Purif. 15, 349361.
  • Inuzuka T., Sato S., McIntyre L. J. and Quarles R. H. (1984) Effects of trypsin and plasmin treatment of myelin on the myelin-associated glycoprotein and basic protein. J. Neurochem. 43, 582585.
  • Johnson P. W., Abramow-Newerly W., Seilheimer B., Sadoul R., Tropak M. B., Arquint M., Dunn R. J., Schachner M. and Roder J. C. (1989) Recombinant myelin-associated glycoprotein confers neural adhesion and neurite outgrowth function. Neuron 3, 377385.
  • Kirchhoff F., Hofer H. W. and Schachner M. (1993) Myelin-associated glycoprotein is phosphorylated by protein kinase C. J. Neurosci. Res. 36, 368381.
  • Kursula P., Lehto V. and Heape A. M. (2000) S100beta inhibits the phosphorylation of the L-MAG cytoplasmic domain by PKA. Mol. Brain Res. 76, 407410.
  • Kursula P., Lehto V. and Heape A. M. (2001) The small myelin-associated glycoprotein binds to tubulin and microtubules. Mol. Brain Res. 87, 2230.
  • Kursula P., Lehto V. P., Garbay B., Cassagne C. and Heape A. M. (1998) Expression of the amino acid dimorphism in the small myelin-associated glycoprotein cytoplasmic domain in rat peripheral nerves during postnatal development[published erratum appears in Mol. Brain Res. 1998, 62, 110]. Mol. Brain Res. 54, 252261.
  • Kursula P., Tikkanen G., Lehto V. P., Nishikimi M. and Heape A. M. (1999) Calcium-dependent interaction between the large myelin-associated glycoprotein and S100beta. J. Neurochem. 73, 17241732.
  • Laemmli V. K. (1970) Cleavage of structural proteins during the assembly of head bacteriophage T4. Nature 227, 680685.
  • Lai C., Brow M. A., Nave K. A., Noronha A. B., Quarles R. H., Bloom F. E., Milner R. J. and Sutcliffe J. G. (1987a) Two forms of 1B236/myelin-associated glycoprotein, a cell adhesion molecule for postnatal neural development, are produced by alternative splicing. Proc. Natl Acad. Sci. USA 84, 43374341.
  • Lai C., Watson J. B., Bloom F. E., Sutcliffe J. G. and Milner R. J. (1987b) Neural protein 1B236/myelin-associated glycoprotein (MAG) defines a subgroup of the immunoglobulin superfamily. Immunol. Rev. 100, 129151.
  • Möller J. R. (1996) Rapid conversion of myelin-associated glycoprotein to a soluble derivative in primates. Brain Res. 741, 2731.
  • Möller J. R., Yanagisawa K., Brady R. O., Tourtellotte W. W. and Quarles R. H. (1987) Myelin-associated glycoprotein in multiple sclerosis lesions: a quantitative and qualitative analysis. Ann. Neurol. 22, 469474.
  • Nomura T. and Fujisawa Y. (1997) Processing properties of recombinant human procathepsin L. Biochem. Biophys. Res. Commun. 230, 143146.
  • Nomura T., Fujishima A. and Fujisawa Y. (1996) Characterization and crystallization of recombinant human cathepsin L. Biochem. Biophys. Res. Commun. 228, 792796.
  • Poltorak M., Sadoul R., Keilhauer G., Landa C., Fahrig T. and Schachner M. (1987) Myelin-associated glycoprotein, a member of the L2/HNK-1 family of neural cell adhesion molecules, is involved in neuron-oligodendrocyte and oligodendrocyte–oligodendrocyte interaction. J. Cell Biol. 105, 18931899.
  • Quarles R. H. (1989) Myelin-associated glycoprotein in demyelinating disorders. Crit. Rev. Neurobiol. 5, 128.
  • Quarles R. H., Everly J. L. and Brady R. O. (1973) Evidence for the close association of a glycoprotein with myelin in rat brain. J. Neurochem. 21, 11771191.
  • Sadoul R., Fahrig T., Bartsch U. and Schachner M. (1990) Binding properties of liposomes containing the myelin-associated glycoprotein MAG to neural cell cultures. J. Neurosci. Res. 25, 113.
  • Salzer J. L., Holmes W. P. and Colman D. R. (1987) The amino acid sequences of the myelin-associated glycoproteins: homology to the immunoglobulin gene superfamily. J. Cell Biol. 104, 957965.
  • Sato S., Quarles R. H. and Brady R. O. (1982) Susceptibility of the myelin-associated glycoprotein and basic protein to a neutral protease in highly purified myelin from human and rat brain. J. Neurochem. 39, 97105.
  • Sato S., Fujita N., Kurihara T., Kuwano R., Sakimura K., Takahashi Y. and Miyatake T. (1989) cDNA cloning and amino acid sequence for human myelin-associated glycoprotein. Biochem. Biophys. Res. Commun. 163, 14731480.
  • Sato S., Quarles R. H., Brady R. O. and Tourtellotte W. W. (1984a) Elevated neutral protease activity in myelin from brains of patients with multiple sclerosis. Ann. Neurol. 15, 264267.
  • Sato S., Yanagisawa K. and Miyatake T. (1984b) Conversion of myelin-associated glycoprotein (MAG) to a smaller derivative by calcium activated neutral protease (CANP)-like enzyme in myelin and inhibition by E-64 analogue. Neurochem. Res. 9, 629635.
  • Schachner M. and Bartsch U. (2000) Multiple functions of the myelin-associated glycoprotein MAG (siglec-4a) in formation and maintenance of myelin. Glia 29, 154165.
  • Spagnol G., Williams M., Srinivasan J., Golier J., Bauer D., Lebo R. V. and Latov N. (1989) Molecular cloning of human myelin-associated glycoprotein. J. Neurosci. Res. 24, 137142.
  • Stebbins J. W., Jaffe H. and Möller J. R. (1998) Characterization of myelin-associated glycoprotein (MAG) proteolysis in the human central nervous system. Neurochem. Res. 23, 10051010.
  • Stebbins J. W., Jaffe H., Fales H. M. and Möller J. R. (1997) Determination of a native proteolytic site in myelin-associated glycoprotein. Biochemistry 36, 22212226.
  • Tang S., Qiu J., Nikulina E. and Filbin M. T. (2001) Soluble myelin-associated glycoprotein released from damaged white matter inhibits axonal regeneration. Mol. Cell Neurosci. 18, 259269.
  • Towbin H., Staehelin T. and Gordon J. (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Natl Acad. Sci. USA 76, 43504354.
  • Umemori H., Sato S., Yagi T., Aizawa S. and Yamamoto T. (1994) Initial events of myelination involve Fyn tyrosine kinase signalling. Nature 367, 572576.
  • Yanagisawa K., Quarles R. H., Johnson D., Brady R. O. and Whitaker J. N. (1985) A derivative of myelin-associated glycoprotein in cerebrospinal fluid of normal subjects and patients with neurological disease. Ann. Neurol. 18, 464469.
  • Yanagisawa K., Sato S., O'Shannessy D. J., Quarles R. H., Suzuki K. and Miyatake T. (1988) Myelin-associated calpain II. J. Neurochem. 51, 803807.