N e u r o c h e m i c a l R e s e a r c h (1) 93-111 (1976)
THE B R E A K D O W N OF M Y E L I N - B O U N D PROTEINS BY INTRA- AND EXTRACELLULAR PROTEASES N. MARKS, A. GRYNBAUM, and A. LAJTHA New York State Research Institute for Neurochemistry and Drug Addiction , Ward's Island, New York, New York, 10035
Accepted December 23, 1975
Changes in protein components of purified myelin were measured following incubation in vitro with purified intra- and extracellular enzymes. Incubation with calf brain cathepsin D did not result in a significant release of acid-soluble peptides as measured by ninhydrin analysis but was accompanied by a large loss of myelin proteins as determined on SDS-acrylamide gels. After 5 hr at 37~ there was a loss of about 25% for fast and slow basic proteins and the Agrawal proteolipid, but only a 5-10% loss for the Folch-Lees and Wolfgram components. Rat brain cathepsin D prepared by affinity chromatography gave a 3060% breakdown of basic proteins and proteolipids. In general, breakdown using lyophilized myelin was increased over two-fold as compared to experiments with fresh myelin. Breakdown induced by cathepsin D was completely inhibited by the pentapeptide pepstatin. Incubation of myelin at physiological pH resulted in an endogenous breakdown of about 12% for basic proteins in freshly prepared, and 50% for lyophilized material. Addition of a soluble neutral proteinase that splits hemoglobin did not induce additional breakdown except for a small change in the Folch-Lees component. The extracellular enzymes pepsin and TPCKtreated trypsin resulted in a larger breakdown of all components as compared to brain enzymes. Present results demonstrate that all protein components of myelin are accessible to hydrolases and vulnerable to breakdown to varying extents by brain enzymes. These facts are consistent with the known rates for myelin protein turnover and may have a bearing on changes associated with demyelinating diseases.
INTRODUCTION Although some findings suggest that a part of myelin is metabolically stable (1) a number of recent studies indicate that a significant pro93 ~) 1976PlenumPublishingCorporation,227 West 17thStreet, New York, N.Y. 10011. No part of this publicationmay be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying,microfilming,recording,or otherwise,withoutwrittenpermissionof the publisher.
94
N. MARKS ET AL.
portion of the proteins and lipids of myelin are in a dynamic state (210). Turnover includes synthesis and breakdown, which means, in the case of myelin proteins, that proteases have access to different portions of the myelin sheath. The nature and origin of these endogenous enzymes involved in myelin turnover, both synthesis and breakdown, still need to be better established. Enzymes involved in protein breakdown are of clinical interest, since there are now many reports that they are increased in CNS tissue as a result of experimental demyelination, canine distemper, and multiple sclerosis (11-19). Earlier studies found that myelin basic protein is split by brain cathepsin D, leading to the formation of only three peptide fragments (20-22). However, the breakdown by brain enzymes of basic protein complexed with lipids within the sheath has not been established, although there are reports that crude lysosomal preparations can disrupt whole myelin preparations (23, 24). New methods capable of separating and determining myelin protein components present an opportunity to study the effects of purified endogenous enzymes, especially those known to be increased during pathological conditions. Degradation of basic proteins, still attached to membranes would be consistent with the hypothesis that toxic polypeptides are formed that can interact at the sites involved in autoimmune disorders associated with demyelination (11, 13, 17). Hitherto, only the effects of trypsin have been studied in detail, using isolated myelin fragments (25-29). Breakdown of proteins in membrane preparations using enzymes as probes has provided useful information on membrane structure (30) and may help to reveal the topographical arrangement of proteins within the myelin complex. In the following, the effects of endogenous enzymes on myelin protein will be compared to the effects of trypsin and pepsin. A preliminary report of some of these results has appeared (31).
EXPERIMENTAL PROCEDURE Preparation of Myelin and its Components. Fresh myelin was prepared in our laboratory by the method of Norton (32). Briefly, the rat brain homogenate was prepared in 0.32 M sucrose, layered over 0.85 M sucrose, and centrifuged at 75,000g. Crude myelin was collected from the interface, subjected to two water shocks, and further centrifuged at 12,000g to remove microsomes. The final step involved another discontinuous gradient centrifugation of myelin in 0.32 M sucrose over 0.85 M sucrose to obtain the purified preparation. The lyophilized preparation of myelin was isolated by the method of Agrawal et al. (33), and was supplied by Dr. Harish Agrawal, Washington University, St. Louis, Missouri. Basic protein was prepared from calf spinal cord (34), proteolipid from bovine white matter (35) and was supplied by Dr. Marjorie Lees (McLean Hospital, Belmont, Massachusetts).
B R E A K D O W N OF M Y E L I N P R O T E I N S
95
Enzymes. Cathepsin D was purified from bovine brain (36) or from rat brain by the method of affinity chromatography on Sepharose hemoglobin (37). It was homogeneous when tested on sodium dodecyl sulfate (S DS) acrylamide gel and was free of other known proteolytic enzymes. Cathepsin D was kept as a lyophilized powder and is stable when stored at 4~ Cathepsin A (lysosomal carboxypeptidase A) was prepared from frozen calf brain by the method of Grynbaum and Marks (38). Enzyme was stored as a frozen solution stable at -20~ For neutral proteinase the postmitochondrial supernatant of rat brain was used (36). Pepsin was obtained (as 3x crystalline powder) from Nutritional Biochemicals (Cleveland, Ohio), and L-l-tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin was obtained from Sigma (St. Louis, Missouri). Reagents'. Hemoglobin substrate powder was obtained from Worthington Biochemicals (Freehold, New Jersey), carbobenzoxy-Glu-Tyr (CBZ-Glu-Tyr) from Fox Chemical (Los Angeles, California) acrylamide, N,N'-methylene-bisacrylamide, and tetramethylethylene diamine (temed) from Eastman Kodak (Rochester, New York), ammonium persulfate from Fisher Scientific (Faiflawn, New Jersey), SDS and Fast Green F C F from Matheson, Coleman and Bell (East Rutherford, New Jersey). Incubation of Myelin. 2-rag portions of myelin from rat brain were suspended in 10-ml cortical centrifuge tubes, each containing 0.5 ml of the appropriate buffer, and incubated with or without enzyme at 37~ in a shaking water bath. Incubations were terminated by placing tubes in a freezing mixture (dry ice and acetone) and immediately lyophilized in the cooling chamber of a mechanical lyophilizer (Metrovac; Long Island, New York) equipped with a small heating coil. The lyophilized powder was then defatted in the same tubes to avoid losses. Enzyme Determination. Pepsin (EC 3.4.4.1), cathepsin D (EC 3.4.23.5), and neutral proteinase were measured using hemoglobin as the substrate at the appropriate pH (acid proteinase in 50 mM sodium citrate buffer pH 3.2; neutral proteinase in 40 mM Tris-HCl pH 7.6). Activity was determined by the ninhydrin procedure of Serra et al. (39) or by an automated ninhydrin procedure (see below). Cathepsin A was measured in 0.5 ml of 40 mM acetate buffer pH 5.5 containing 16/xg of enzyme protein (specific activity 5.5 /zmol/mg protein) and l tzmol of Z-Glu-Tyr. The mixture was incubated for 1.5-2 hr at 37~ and the reaction was terminated by the addition of 0.5 ml of 6% sulfosalicylic acid. After centrifugation at 40,000g, 0.5 ml of the supernatant was diluted with 2 ml of 400 mM acetate buffer p H 5.5 and placed in a sample cup of a Sampler I1 model (Technicon, Tarrytown, New York), with suitable controls and the ninhydrin-positive materials measured. This method is capable of detecting as little as 1-2 nmol of a-amino group (Tyr in the case of Cathepsin A assay) using the expanded scale of a linear-output two-channel colorimeter coupled to a suitable recorder (Technicon, Tarrytown, New York). This method was modified for microdetermination of ninhydrinpositive materials produced on incubation of myelin and its components with enzyme. Portions (250-500 p,g) of myelin basic protein, or proteolipid, were placed in 6 x 50-mm disposable culture tubes in a final volume of 110 /xl of citrate buffer. For myelin and proteolipid, Lubrol (WX-5070, ICI, Amer. Div., New York) was added to solubilize these components. Reaction was terminated after 2 hr incubation by adding 100 ~1 of 6.2% sulfosalicylic acid and spinning down the pellet in a microhematocrit centrifuge for 15 min at 3000g. 100 tzl of the supernatant was diluted with 1.0ml of 0.1 M acetate buffer pH 5.5, and the ninhydrin-positive material was measured in a Technicon Autoanalyzer at 570 nm using arginine as a standard. Trypsin was measured with benzoyl arginyl fl-naphthylamide (BANA) by determining the release of/3-naphthylamine in 1.0 ml of 50 mM Tris-HCl pH 7.6, containing 1.5 ~mol of B A N A and 3 /zg of enzyme in a cuvette with continuous fluorometric readings in a
96
N. MARKS ET AL.
Turner model III (peak activation 360 rim, and the secondary 2A filter with cutoff at 415 rim).
SDS-Disk Gel Electrophoresis. Digests and controls were separated by S DS-disk gel electrophoresis on 15% acrylamide using a slightly modified procedure of Greenfield et al. (40). Briefly, the lyophilized mixtures were extracted with ether-ethanol (3:2 v/v) followed by sonification in a protein solvent (50 mM Tris, pH 6.8, 1% SDS, 1% mercaptoethanol, 0.001% bromophenol blue, 10% glycerol). Aliquots corresponding to 100 /zg of protein (based on the amount present in the original myelin incubation) were subjected to electrophoresis on 10-cm gels at constant voltage of 45 V for 19 hr with equilibration of the buffer in the two chambers during the run by means of a small pump. The gels were then removed by breaking the glass tubes, kept for 24 hr in fixing solution of acetic acidmethanol-water (10:45:45, v/v/v), and then kept for 2 hr in fixing solution containing 1% solution of Fast Green FCF. Gels were destained by diffusion in several changes of the same solution, and the density of the different bands was measured at 580 nm using a Gilford 2400 spectrophotometer equipped with a linear transport mechanism.
RESULTS In a preliminary set of experiments, an attempt was made to monitor breakdown by measurements of the release of ninhydrin-soluble materials. Incubation of myelin membranes with purified cathepsin D from calf brain failed to yield any significant amounts of acid-soluble products (Table I). Even in the presence of 1% Lubrol, which resulted in a clear Opalescent solution, the release of acid-soluble products amounts to a rate of only 0.05%/hr. Lubrol at this concentration did not
TABLE I BREAKDOWN OF MYELIN PROTEINS BY CATHEPSIN D ESTIMATED BY THE APPEARANCE OF SOLUBLE N I N H Y D R I N - P o s I T I V E COMPOUNDS a
Substrate and incubation conditions Myelin Myelin + 1% Lubrol Proteolipid (Folch-Lees) Proteolipid + 1% Lubrol Basic protein
Ninhydrin-positive material (nmol)
Breakdown (%)
0 1.8 0 1.5 8.0
0 0.14 0 0.06 0.31
a The reaction mixture contained 0.5 mg myelin or proteolipid or 0.25 mg basic protein in 0.5 ml of sodium citrate buffer pH 3.2 and 3 /~g of purified cathepsin D enzyme protein (enzyme-to-substrate ratio 1:80160); it was incubated in the presence or absence of 1% Lubrol for 3 hr at 37~ Enzyme action was terminated by addition of 0.5 ml of 6% sulfosalicylie acid, and after centrifugation the supernatant was used for determination of ninhydrin-positive materials.
B R E A K D O W N OF M Y E L I N PROTEINS
97
interfere with cathepsin D activity. Similarly, a purified preparation of Folch-Lees proteolipid incubated in 1% Lubrol with the enzyme resulted in only trace levels of breakdown products. Bovine basic protein, which is very soluble, gave a breakdown equivalent to only 0.1%/hr. These results imply that any breakdown occurring may have resulted in the formation of acid-insoluble polypeptides. The effects of different proteolytic enzymes were evaluated, therefore, by measuring the changes in the level of proteins following their separation on acrylamide gels in the presence of SDS. Myelin of rat brain gave a reproducible characteristic pattern of proteins: two bands for basic proteins, a major band (Folch-Lees) and a minor band (Agrawal) for proteolipid, and several high-molecular-weight components, including the Wolfgram protein (Figure 1).
Effect of Brain Enzymes Incubation of fresh myelin membranes with purified cathepsin D from calf or rat brain led to a marked decrease in the intensity for several of the major protein bands (Table II, Figure 2). Treatment with calf enzyme for 5 hr (enzyme-to-substrate ratio 1:30) resulted in decrease of the basic proteins and the Agrawal proteolipids, but no significant change of the Folch-Lees proteolipids and the Wolfgram protein. Incubation with the rat enzyme with high specific activity prepared by affinity chromatography (enzyme-to-substrate ratio 1:150) resulted in a greater breakdown of basic proteins and of proteolipids but no detectable breakdown of the Wolfgram protein. Addition of 0.1 t~M pepstatin, a known inhibitor of brain cathepsin D (41), completely blocked the degradation of myelin proteins (Table II). Myelin prepared in sucrose gradients is reported to contain neutral endopeptidase capable of digesting protein components (2, 39, 42). Incubation of myelin at pH 7.6 in the absence of added enzyme resulted in a decrease in basic protein; addition of a rat brain cytosol known to contain enzymes degrading hemoglobin and hormones (36, 43, 44) was without further effect except for a small change in the Folch-Lees proteolipid. Interference on gels by high-molecular-weight components of the crude cytosol fractions prevented the measurement of Wolfgram proteins (Table II).
Extracellular Enzymes Incubation of myelin with pepsin (enzyme-to-substrate ratio 1:700) for 0.5-3 hr resulted in a decrease of all protein components except proteolipids. At higher concentration of pepsin (ratio 1:350) a decrease
98
N. MARKS ET AL.
W
Pfl Bs Bf
A
B
C
FI6. 1. The pattern of myelin protein components of rat brain myelin after separation on a 15% acrylamide gel containing 1% SDS; W, Wolfgram protein; Pf~ Folch-Lees proteolipid; Pa, Agrawal proteolipid; Bs, slow basic protein, and Bf, fast protein component. The three gels illustrate the solubility of proteins after incubation in a 50 mM sodium citrate buffer, pH 3.2, as described in the text. A, complete incubation mixture after 5 hr at 37~ B, pellet fraction remaining after centrifugation; C, supernatant fraction. For conditions of incubation see Table IV.
BREAKDOWN
OF MYELIN
99
PROTEINS
Z
=.~
~
~
0,,~
~.~ ~
~ ~
~,~
z
z m
m~ =z
9 Z < Z
N
~'= M~M
~,..
~d~
~'~ 9~ . ~ =
~ ~~
~ ~
-~-~-
,~.~ ~ ~,~ ~
~ ~
100
N.
MARKS
ET
AL.
1,5
~g 1,0 t~ v
b-g g ~
Z t.t.a
_..1
0,5
DO
MOB I L I TY (CM)
B R E A K D O W N OF MYELIN PROTEINS
101
in proteolipids also occurred. Incubation of fresh myelin with trypsin (TPCK treated) resulted in a decrease of all proteins at 0.5 hr and in an almost complete breakdown at 3 hr (Table III).
Solubility of Proteins Susceptibility to breakdown may be influenced by the amount of proteins in solution under the conditions of incubation. In order to test this, myelin preparations were incubated in the absence of enzymes, then centrifuged, and the protein pattern of the supernatant and pellet fractions was analyzed. The results showed that there was a partial solubilization of basic proteins and Agrawal proteolipid in sodium citrate buffer (50 raM, p H 3.2), which was greater for lyophilized than for freshly prepared myelin (Table IV). The total quantity of proteins in the soluble and pellet fractions accounted for that present in the untreated myelin, indicating an absence of endogenous breakdown at this pH.
Breakdown of Lyophilized Myelin To determine if lyophilized myelin, in view of its increased solubility, was more susceptible to breakdown, it was treated with calf cathepsin D and then analyzed. Results show that calf enzyme gave a 2-3-fold increase in breakdown of basic, Agrawal, and Wolfgram proteins and about 7-fold increase in breakdown of the Folch-Lees components in lyophilized as compared to fresh myelin (Tables II and V). Incubation in the absence of enzyme at p H 3.2 resulted in no changes and confirmed the absence of endogenous breakdown. At p H 7.6, however, there was a decrease of about 50% in basic proteins, showing that a larger endogenous breakdown occurred in lyophilized than in fresh myelin (Tables II and V). Upon addition of rat cytosol (containing a hemoglobin-splitting enzyme) no further breakdown was observed except in the case of the Folch-Lees proteolipid (Table V). It was again not possible to measure FIG. 2. Densitometric scans of myelin protein components following electrophoresis on SDS-acrylamide gels and staining with Fast Green as described in the methods section. The scans were made with a Gilford 2400 spectrophotometer equipped with a linear transport mechanism and read at 580 nm. Rates of breakdown were calculated from the change in area of control (C) as compared to the action of cathepsin D (B) and pepsin (A) incubated as described in the Experimental Procedure. The width of the curves for purposes of measurement are indicated by arrows. The abbreviations are identical to those of Fig. 1. The alignment of bands with an actual run is illustrated in the gel in the lower portion of the figure. The first band for gels A and B indicates breakdown products of about 10,000 daltons and which are largely absent in the control C.
102
N. M A R K S E T AL.
%
,.d k)
THE B R E A K D O W N OF M Y E L I N - B O U N D PROTEINS BY INTRA- AND EXTRACELLULAR PROTEASES N. MARKS, A. GRYNBAUM, and A. LAJTHA New York State Research Institute for Neurochemistry and Drug Addiction , Ward's Island, New York, New York, 10035
Accepted December 23, 1975
Changes in protein components of purified myelin were measured following incubation in vitro with purified intra- and extracellular enzymes. Incubation with calf brain cathepsin D did not result in a significant release of acid-soluble peptides as measured by ninhydrin analysis but was accompanied by a large loss of myelin proteins as determined on SDS-acrylamide gels. After 5 hr at 37~ there was a loss of about 25% for fast and slow basic proteins and the Agrawal proteolipid, but only a 5-10% loss for the Folch-Lees and Wolfgram components. Rat brain cathepsin D prepared by affinity chromatography gave a 3060% breakdown of basic proteins and proteolipids. In general, breakdown using lyophilized myelin was increased over two-fold as compared to experiments with fresh myelin. Breakdown induced by cathepsin D was completely inhibited by the pentapeptide pepstatin. Incubation of myelin at physiological pH resulted in an endogenous breakdown of about 12% for basic proteins in freshly prepared, and 50% for lyophilized material. Addition of a soluble neutral proteinase that splits hemoglobin did not induce additional breakdown except for a small change in the Folch-Lees component. The extracellular enzymes pepsin and TPCKtreated trypsin resulted in a larger breakdown of all components as compared to brain enzymes. Present results demonstrate that all protein components of myelin are accessible to hydrolases and vulnerable to breakdown to varying extents by brain enzymes. These facts are consistent with the known rates for myelin protein turnover and may have a bearing on changes associated with demyelinating diseases.
INTRODUCTION Although some findings suggest that a part of myelin is metabolically stable (1) a number of recent studies indicate that a significant pro93 ~) 1976PlenumPublishingCorporation,227 West 17thStreet, New York, N.Y. 10011. No part of this publicationmay be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying,microfilming,recording,or otherwise,withoutwrittenpermissionof the publisher.
94
N. MARKS ET AL.
portion of the proteins and lipids of myelin are in a dynamic state (210). Turnover includes synthesis and breakdown, which means, in the case of myelin proteins, that proteases have access to different portions of the myelin sheath. The nature and origin of these endogenous enzymes involved in myelin turnover, both synthesis and breakdown, still need to be better established. Enzymes involved in protein breakdown are of clinical interest, since there are now many reports that they are increased in CNS tissue as a result of experimental demyelination, canine distemper, and multiple sclerosis (11-19). Earlier studies found that myelin basic protein is split by brain cathepsin D, leading to the formation of only three peptide fragments (20-22). However, the breakdown by brain enzymes of basic protein complexed with lipids within the sheath has not been established, although there are reports that crude lysosomal preparations can disrupt whole myelin preparations (23, 24). New methods capable of separating and determining myelin protein components present an opportunity to study the effects of purified endogenous enzymes, especially those known to be increased during pathological conditions. Degradation of basic proteins, still attached to membranes would be consistent with the hypothesis that toxic polypeptides are formed that can interact at the sites involved in autoimmune disorders associated with demyelination (11, 13, 17). Hitherto, only the effects of trypsin have been studied in detail, using isolated myelin fragments (25-29). Breakdown of proteins in membrane preparations using enzymes as probes has provided useful information on membrane structure (30) and may help to reveal the topographical arrangement of proteins within the myelin complex. In the following, the effects of endogenous enzymes on myelin protein will be compared to the effects of trypsin and pepsin. A preliminary report of some of these results has appeared (31).
EXPERIMENTAL PROCEDURE Preparation of Myelin and its Components. Fresh myelin was prepared in our laboratory by the method of Norton (32). Briefly, the rat brain homogenate was prepared in 0.32 M sucrose, layered over 0.85 M sucrose, and centrifuged at 75,000g. Crude myelin was collected from the interface, subjected to two water shocks, and further centrifuged at 12,000g to remove microsomes. The final step involved another discontinuous gradient centrifugation of myelin in 0.32 M sucrose over 0.85 M sucrose to obtain the purified preparation. The lyophilized preparation of myelin was isolated by the method of Agrawal et al. (33), and was supplied by Dr. Harish Agrawal, Washington University, St. Louis, Missouri. Basic protein was prepared from calf spinal cord (34), proteolipid from bovine white matter (35) and was supplied by Dr. Marjorie Lees (McLean Hospital, Belmont, Massachusetts).
B R E A K D O W N OF M Y E L I N P R O T E I N S
95
Enzymes. Cathepsin D was purified from bovine brain (36) or from rat brain by the method of affinity chromatography on Sepharose hemoglobin (37). It was homogeneous when tested on sodium dodecyl sulfate (S DS) acrylamide gel and was free of other known proteolytic enzymes. Cathepsin D was kept as a lyophilized powder and is stable when stored at 4~ Cathepsin A (lysosomal carboxypeptidase A) was prepared from frozen calf brain by the method of Grynbaum and Marks (38). Enzyme was stored as a frozen solution stable at -20~ For neutral proteinase the postmitochondrial supernatant of rat brain was used (36). Pepsin was obtained (as 3x crystalline powder) from Nutritional Biochemicals (Cleveland, Ohio), and L-l-tosylamide-2-phenylethyl chloromethyl ketone (TPCK)-treated trypsin was obtained from Sigma (St. Louis, Missouri). Reagents'. Hemoglobin substrate powder was obtained from Worthington Biochemicals (Freehold, New Jersey), carbobenzoxy-Glu-Tyr (CBZ-Glu-Tyr) from Fox Chemical (Los Angeles, California) acrylamide, N,N'-methylene-bisacrylamide, and tetramethylethylene diamine (temed) from Eastman Kodak (Rochester, New York), ammonium persulfate from Fisher Scientific (Faiflawn, New Jersey), SDS and Fast Green F C F from Matheson, Coleman and Bell (East Rutherford, New Jersey). Incubation of Myelin. 2-rag portions of myelin from rat brain were suspended in 10-ml cortical centrifuge tubes, each containing 0.5 ml of the appropriate buffer, and incubated with or without enzyme at 37~ in a shaking water bath. Incubations were terminated by placing tubes in a freezing mixture (dry ice and acetone) and immediately lyophilized in the cooling chamber of a mechanical lyophilizer (Metrovac; Long Island, New York) equipped with a small heating coil. The lyophilized powder was then defatted in the same tubes to avoid losses. Enzyme Determination. Pepsin (EC 3.4.4.1), cathepsin D (EC 3.4.23.5), and neutral proteinase were measured using hemoglobin as the substrate at the appropriate pH (acid proteinase in 50 mM sodium citrate buffer pH 3.2; neutral proteinase in 40 mM Tris-HCl pH 7.6). Activity was determined by the ninhydrin procedure of Serra et al. (39) or by an automated ninhydrin procedure (see below). Cathepsin A was measured in 0.5 ml of 40 mM acetate buffer pH 5.5 containing 16/xg of enzyme protein (specific activity 5.5 /zmol/mg protein) and l tzmol of Z-Glu-Tyr. The mixture was incubated for 1.5-2 hr at 37~ and the reaction was terminated by the addition of 0.5 ml of 6% sulfosalicylic acid. After centrifugation at 40,000g, 0.5 ml of the supernatant was diluted with 2 ml of 400 mM acetate buffer p H 5.5 and placed in a sample cup of a Sampler I1 model (Technicon, Tarrytown, New York), with suitable controls and the ninhydrin-positive materials measured. This method is capable of detecting as little as 1-2 nmol of a-amino group (Tyr in the case of Cathepsin A assay) using the expanded scale of a linear-output two-channel colorimeter coupled to a suitable recorder (Technicon, Tarrytown, New York). This method was modified for microdetermination of ninhydrinpositive materials produced on incubation of myelin and its components with enzyme. Portions (250-500 p,g) of myelin basic protein, or proteolipid, were placed in 6 x 50-mm disposable culture tubes in a final volume of 110 /xl of citrate buffer. For myelin and proteolipid, Lubrol (WX-5070, ICI, Amer. Div., New York) was added to solubilize these components. Reaction was terminated after 2 hr incubation by adding 100 ~1 of 6.2% sulfosalicylic acid and spinning down the pellet in a microhematocrit centrifuge for 15 min at 3000g. 100 tzl of the supernatant was diluted with 1.0ml of 0.1 M acetate buffer pH 5.5, and the ninhydrin-positive material was measured in a Technicon Autoanalyzer at 570 nm using arginine as a standard. Trypsin was measured with benzoyl arginyl fl-naphthylamide (BANA) by determining the release of/3-naphthylamine in 1.0 ml of 50 mM Tris-HCl pH 7.6, containing 1.5 ~mol of B A N A and 3 /zg of enzyme in a cuvette with continuous fluorometric readings in a
96
N. MARKS ET AL.
Turner model III (peak activation 360 rim, and the secondary 2A filter with cutoff at 415 rim).
SDS-Disk Gel Electrophoresis. Digests and controls were separated by S DS-disk gel electrophoresis on 15% acrylamide using a slightly modified procedure of Greenfield et al. (40). Briefly, the lyophilized mixtures were extracted with ether-ethanol (3:2 v/v) followed by sonification in a protein solvent (50 mM Tris, pH 6.8, 1% SDS, 1% mercaptoethanol, 0.001% bromophenol blue, 10% glycerol). Aliquots corresponding to 100 /zg of protein (based on the amount present in the original myelin incubation) were subjected to electrophoresis on 10-cm gels at constant voltage of 45 V for 19 hr with equilibration of the buffer in the two chambers during the run by means of a small pump. The gels were then removed by breaking the glass tubes, kept for 24 hr in fixing solution of acetic acidmethanol-water (10:45:45, v/v/v), and then kept for 2 hr in fixing solution containing 1% solution of Fast Green FCF. Gels were destained by diffusion in several changes of the same solution, and the density of the different bands was measured at 580 nm using a Gilford 2400 spectrophotometer equipped with a linear transport mechanism.
RESULTS In a preliminary set of experiments, an attempt was made to monitor breakdown by measurements of the release of ninhydrin-soluble materials. Incubation of myelin membranes with purified cathepsin D from calf brain failed to yield any significant amounts of acid-soluble products (Table I). Even in the presence of 1% Lubrol, which resulted in a clear Opalescent solution, the release of acid-soluble products amounts to a rate of only 0.05%/hr. Lubrol at this concentration did not
TABLE I BREAKDOWN OF MYELIN PROTEINS BY CATHEPSIN D ESTIMATED BY THE APPEARANCE OF SOLUBLE N I N H Y D R I N - P o s I T I V E COMPOUNDS a
Substrate and incubation conditions Myelin Myelin + 1% Lubrol Proteolipid (Folch-Lees) Proteolipid + 1% Lubrol Basic protein
Ninhydrin-positive material (nmol)
Breakdown (%)
0 1.8 0 1.5 8.0
0 0.14 0 0.06 0.31
a The reaction mixture contained 0.5 mg myelin or proteolipid or 0.25 mg basic protein in 0.5 ml of sodium citrate buffer pH 3.2 and 3 /~g of purified cathepsin D enzyme protein (enzyme-to-substrate ratio 1:80160); it was incubated in the presence or absence of 1% Lubrol for 3 hr at 37~ Enzyme action was terminated by addition of 0.5 ml of 6% sulfosalicylie acid, and after centrifugation the supernatant was used for determination of ninhydrin-positive materials.
B R E A K D O W N OF M Y E L I N PROTEINS
97
interfere with cathepsin D activity. Similarly, a purified preparation of Folch-Lees proteolipid incubated in 1% Lubrol with the enzyme resulted in only trace levels of breakdown products. Bovine basic protein, which is very soluble, gave a breakdown equivalent to only 0.1%/hr. These results imply that any breakdown occurring may have resulted in the formation of acid-insoluble polypeptides. The effects of different proteolytic enzymes were evaluated, therefore, by measuring the changes in the level of proteins following their separation on acrylamide gels in the presence of SDS. Myelin of rat brain gave a reproducible characteristic pattern of proteins: two bands for basic proteins, a major band (Folch-Lees) and a minor band (Agrawal) for proteolipid, and several high-molecular-weight components, including the Wolfgram protein (Figure 1).
Effect of Brain Enzymes Incubation of fresh myelin membranes with purified cathepsin D from calf or rat brain led to a marked decrease in the intensity for several of the major protein bands (Table II, Figure 2). Treatment with calf enzyme for 5 hr (enzyme-to-substrate ratio 1:30) resulted in decrease of the basic proteins and the Agrawal proteolipids, but no significant change of the Folch-Lees proteolipids and the Wolfgram protein. Incubation with the rat enzyme with high specific activity prepared by affinity chromatography (enzyme-to-substrate ratio 1:150) resulted in a greater breakdown of basic proteins and of proteolipids but no detectable breakdown of the Wolfgram protein. Addition of 0.1 t~M pepstatin, a known inhibitor of brain cathepsin D (41), completely blocked the degradation of myelin proteins (Table II). Myelin prepared in sucrose gradients is reported to contain neutral endopeptidase capable of digesting protein components (2, 39, 42). Incubation of myelin at pH 7.6 in the absence of added enzyme resulted in a decrease in basic protein; addition of a rat brain cytosol known to contain enzymes degrading hemoglobin and hormones (36, 43, 44) was without further effect except for a small change in the Folch-Lees proteolipid. Interference on gels by high-molecular-weight components of the crude cytosol fractions prevented the measurement of Wolfgram proteins (Table II).
Extracellular Enzymes Incubation of myelin with pepsin (enzyme-to-substrate ratio 1:700) for 0.5-3 hr resulted in a decrease of all protein components except proteolipids. At higher concentration of pepsin (ratio 1:350) a decrease
98
N. MARKS ET AL.
W
Pfl Bs Bf
A
B
C
FI6. 1. The pattern of myelin protein components of rat brain myelin after separation on a 15% acrylamide gel containing 1% SDS; W, Wolfgram protein; Pf~ Folch-Lees proteolipid; Pa, Agrawal proteolipid; Bs, slow basic protein, and Bf, fast protein component. The three gels illustrate the solubility of proteins after incubation in a 50 mM sodium citrate buffer, pH 3.2, as described in the text. A, complete incubation mixture after 5 hr at 37~ B, pellet fraction remaining after centrifugation; C, supernatant fraction. For conditions of incubation see Table IV.
BREAKDOWN
OF MYELIN
99
PROTEINS
Z
=.~
~
~
0,,~
~.~ ~
~ ~
~,~
z
z m
m~ =z
9 Z < Z
N
~'= M~M
~,..
~d~
~'~ 9~ . ~ =
~ ~~
~ ~
-~-~-
,~.~ ~ ~,~ ~
~ ~
100
N.
MARKS
ET
AL.
1,5
~g 1,0 t~ v
b-g g ~
Z t.t.a
_..1
0,5
DO
MOB I L I TY (CM)
B R E A K D O W N OF MYELIN PROTEINS
101
in proteolipids also occurred. Incubation of fresh myelin with trypsin (TPCK treated) resulted in a decrease of all proteins at 0.5 hr and in an almost complete breakdown at 3 hr (Table III).
Solubility of Proteins Susceptibility to breakdown may be influenced by the amount of proteins in solution under the conditions of incubation. In order to test this, myelin preparations were incubated in the absence of enzymes, then centrifuged, and the protein pattern of the supernatant and pellet fractions was analyzed. The results showed that there was a partial solubilization of basic proteins and Agrawal proteolipid in sodium citrate buffer (50 raM, p H 3.2), which was greater for lyophilized than for freshly prepared myelin (Table IV). The total quantity of proteins in the soluble and pellet fractions accounted for that present in the untreated myelin, indicating an absence of endogenous breakdown at this pH.
Breakdown of Lyophilized Myelin To determine if lyophilized myelin, in view of its increased solubility, was more susceptible to breakdown, it was treated with calf cathepsin D and then analyzed. Results show that calf enzyme gave a 2-3-fold increase in breakdown of basic, Agrawal, and Wolfgram proteins and about 7-fold increase in breakdown of the Folch-Lees components in lyophilized as compared to fresh myelin (Tables II and V). Incubation in the absence of enzyme at p H 3.2 resulted in no changes and confirmed the absence of endogenous breakdown. At p H 7.6, however, there was a decrease of about 50% in basic proteins, showing that a larger endogenous breakdown occurred in lyophilized than in fresh myelin (Tables II and V). Upon addition of rat cytosol (containing a hemoglobin-splitting enzyme) no further breakdown was observed except in the case of the Folch-Lees proteolipid (Table V). It was again not possible to measure FIG. 2. Densitometric scans of myelin protein components following electrophoresis on SDS-acrylamide gels and staining with Fast Green as described in the methods section. The scans were made with a Gilford 2400 spectrophotometer equipped with a linear transport mechanism and read at 580 nm. Rates of breakdown were calculated from the change in area of control (C) as compared to the action of cathepsin D (B) and pepsin (A) incubated as described in the Experimental Procedure. The width of the curves for purposes of measurement are indicated by arrows. The abbreviations are identical to those of Fig. 1. The alignment of bands with an actual run is illustrated in the gel in the lower portion of the figure. The first band for gels A and B indicates breakdown products of about 10,000 daltons and which are largely absent in the control C.
102
N. M A R K S E T AL.
%
,.d k)
