Mar 18, 1991 - peroxide concentration of reac- tion mixture, 55 pmol/L). The reference mean (and range) of serum catalase activity concentration was 50.5.
CLIN.CHEM.37/12, 2043-2047 (1991)
Serum Catalase: Reversibly Formed Charge Isoform of Erythrocyte Catalase L#{225}szl#{243} Gth The different electrophoretic mobilities of erythrocyte and serum catalase (EC 1.11.1.6) were confirmed and the causes responsible for their differences were examined. The presence of a catalase-binding protein in serum that could form a complex with erythrocyte catalase was excluded by incubating serum proteins with erythrocyte catalase. No new unequivocal catalase bands representing a catalase-binding protein were detected. The erythrocyte and serum catalase proved to be charge isoforms: their molecular masses, estimated by gel permeation chromatography or polyacrylamide gel electrophoresis in a nondenatunng system, were very similar, whereas their electrophoretic mobilities were different. Assay of serum catalase by gel permeation and hydrophobic chromatography yielded a product with the same electrophoretic mobility as that of erythrocyte catalase. Different dilution of erythrocyte catalase with human sera led to a gradual decrease of its mobility, 20-fold or greater dilution yielding the same results as for serum catalase. Similarly, when serum catalase was diluted 20-fold or more with 60 mmol/L phosphate buffer, it migrated similarly to erythrocyte catalase. I detected no effect of dialyzable serum ligands, NADPH, or protection of SH groups on the electrophoretic mobility of either catalase isoform. I conclude that formation of charge isoforms of catalase is caused by a reversible, conformational modification due to matrix effect of serum. The appearance of some intracellular proteins, especially enzymes, in serum can be used for diagnosis of several diseases. The estimation of organ-specific isoenzymes such as lactate dehydrogenase (EC 1.1.1.27) and a-amylase (EC 3.2.1.1) appearing in serum with regularly unchanged forms is a further diagnostic tool. On the other hand, modifications of intracellular proteins that get into the circulation could create species with different electrophoretic mobilities. The knowledge of the exact route of such modifications is important because the new variants may yield either further information or diagnostic error. The changes involved might be reversible or irreversible. Either type can yield differences in charge (charge isoforms) or molecular mass (molecular mass isoforms). Modification of the M subunit of creatine kinase (EC 2.7.3.2) by plasma carboxypeptidase N (1), the proteolytic cleavage of ceruloplasmin (2), and stable macmenzyme complexes (3) are examples of the irreversibly formed proteins formed during circulation. The stable complexes are high-molecular-mass complexes formed Department of Laboratory, Municipal Hospital, Sumeg, Hungary H-8330. Received March 18, 1991; accepted September 26, 1991.
by self-polymerization (mitochondrial creatine kinase) or by association with immunoglobulins or lipids in sera; another example is the complex of erythrocyte hemoglobin with its binding protein (haptoglobin) of serum.
In contrast, reversible changes that yield conformationally different forms of proteins also occur in serum. The main routes causing reversible alterations include the following: First, some ligands in serum produce allosteric modifications of enzymes (4). Second, the electrophoretic mobility of transport proteins is changed by their bound ligand, i.e., albumin by bilirubin, vitamin D-binding protein by vitamin D, and transferrin
by iron.
Third, the extracellular proteins may also differ from intracellular proteins by containing disulfide bonds that provide stability and slightly different electrophoretic mobility, as well as the folded structure of large plasma proteins during recirculation (5). The serum activity of heme-containing catalase (EC 1.11.1.6) is highly increased in hemolytic diseases (6) and in acute pancreatitis (7). Serum catalase has no isoenzymes (8), and its decisive source is the erythrocyte pool in hemolytic disease (9), in acute pancreatitis (10), and under normal circumstances (9). Serum and eryth-
rocyte catalases are two distinct forms, differing in their electrophoretic mobilities (8, 11), pH optima, and activation energy (12). Furthermore, the isoform of catalase specific for blood serum looks likely to be a product of the processing presumably takes fluid itself (8, 11). There are no data on any catalase-binding proteins in serum, nor on whether erythrocyte and serum catalase are molecular-mass or charge isoforms. Furthermore, from the literature it is not clear whether the formation of serum catalase from erythrocyte catalase is reversible erythroid isoform. This place in the extracellular
or not. Here I report an examination of the molecular differences between erythrocyte and serum catalase. For this purpose, the first step was to exclude the presence of catalase-binding protein in serum. I then demonstrated that erythrocyte and serum catalase are reversibly convertible charge isoforms with the same molecular mass and different electrophoretic mobilities. Methods and MaterIals Catalase electrophoresis. For catalase electrophoresis cellulose acetate membrane (Electrophoresis Strip; Sartorius, G#{244}ttingen,F.R.G.). I applied 2 pL of sample onto a 4 x 7 cm strip and electrophoresed this for 5 h (5 mA per strip in phosphate buffer, 60 mmol/L, pH 7.2). I then placed the strips on the surface of starch gel (50 gIL in the phosphate buffer) for 2 h. After blotting the samples, catalase in the starch gel was made visible
I used
CLINICAL CHEMISTRY, Vol. 37, No. 12, 1991
2043
by a negative-staining method (8). The white catalase bands were characterized by their distances from the point of sample application. Detection of “catalase-binding protein” in serum. Serum proteins were separated on cellulose acetate membrane and blotted into starch gel as described above. The proteins in the starch gel were incubated for 30 mm with fresh human erythrocyte hemolysate (containing catalase activity of 500 kUfL). After incubation the hemolysate was removed and the gel was rinsed with 50 g/L human albumin in phosphate-buffered saline (PBS; 60 mmolJL phosphate buffer, pH 7.4, and 0.14 mol/L NaC1) and stained as described above. In the blank test, erythrocyte hemolysate in the incubation was replaced by 50 g/L albumin in PBS. To veriir the noncatalase nature of the new spot, I used
serum
samples
with
highly
increased
a-amylase
activity
(1170 and 4086 UIL) as well as human salivary cr-amylase (50 400 UIL). Furthermore, I made a 10-mmlong vertical pin-scratch at the point of sample application, artificially damaging the surface of the starch gel, to see whether such a mark could be confused with the stained bands. Molecular-mass estimation with gel-permeation chromatography. The apparent molecular mass of human serum and erythrocyte catalase was assayed by gelpermeation chromatography of 2-mL samples on a 1.5 x 30 cm column of Sepharose 6B (Pharmacia, Bromma, Sweden). The eluent was 60 mmolJL phosphate buffer, pH 7.4, at a flow rate of 6 mL cm2 h-i; 2-mL fractions were collected. The catalase activity of the serum samples ranged from 130 to 440 kUIL. The molecular mass (kDa) standards were ferritin (450), aldolase (158), uricase (100), and albumin (68), obtained from Boshringer Mannheim (Mannheim, F.R.G.), and glucose oxidase (186) from Sigma Chemical Co. (St. Louis, MO). For comparison I assayed bovine liver catalase (240 kDa) from Boehringer Mannheim. In the eluted fractions the molecular mass markers were detected at 280 nm, the purified catalases at 280 and 405 nm; I also measured the activity of serum catalase (13). Molecular mass estimation with polyacrylamide gel electrophoresis in a nondenaturing system. This determination, which allows the retention of the characteristics of “native” serum catalase, was performed according to the manufacturer’s information (Technical Bulletin No. MKR-137, Nondenatured protein molecular weight marker kit; Sigma Chemical Co.). I used a set of electrophoresis gels of various (4.0%, 4.5%, 5.0%, 5.5%, and 6.0%) polyacrylamide gel concentrations. The molecular-mass markers were urease (272 and 545 kDa) from Sigma, bovine liver catalase (240 kDa) and ferritin (450 and 900 kDa) from Boehringer Mannheim. These markers and purified human erythrocyte catalase were stained with Coomassie Brilliant Blue, whereas unpurified catalases were detected by their enzymatic activity (14). The mean cat#{225}lase activity in the serum samples was 60.6 kUIL for normal sera and 210 kUIL for pathological 2044
CLINICAL CHEMISTRY, Vol. 37, No. 12, 1991
sera. Per gel, 5-30 L
of serum samples and 20 pg of purified proteins were used. In another experiment a serum with highly increased (1156 kUIL) catalase was diluted 10-fold, either with 60 mmolfL phosphate buffer (pH 7.4) or with pooled normal serum (catalase activity 52.4 kUIL). The slope of a plot of 100’ log (Rf x 100) vs the percentage of the gel concentration was considered the retardation coefficient. The logarithm of negative slope was then plotted against the logarithm of the mass of the molecular mass standards, yielding the calibration curve. Each sample was run in duplicate, and results of four series were used for the final conclusion. Chromatography on Phenyl-Sepharose CL-4B. Human serum catalase was partially purified by hydrophobic interaction chromatography on 2 mL of Phenyl-Sepharose CL-4B gel (Pharmacia) packed into 2.5-mL minicolumns (Bio-Rad Labs., Richmond, CA) and equilibrated with 5 mL of 60 mmol/L phosphate buffer (pH 7.4) containing 1.5 mol of ammonium sulfate per liter. Each column was then loaded with 4-7 mL of serum (catalase activity 363-5770 kU/L), washed (protein-free) with the phosphate/ammonium sulfate buffer. The bound catalase was then eluted with 60 mmol/L phosphate buffer, pH 7.4. One-milliliter fractions were collected and analyzed for protein contents (15) and catalase activities (13). Effect of dilution on electrophoretic mobility of catalase. In this experiment, human sera with pathological catalase activity, erythrocyte hemolysate, and partially purified serum catalase obtained by hydrophobic interaction chromatography of sera containing high amounts of serum catalase were diluted with (a) 60 mmol/L phosphate buffer, pH 7.4, (b) pooled human sera with normal catalase activity, or (c) stabilizing reagent (EDTA, 2.7 mmol/L, pH 7.0, and 0.7 mmol/L 2-mercaptoethanol) to protect the enzymes’ SH groups (16). Effect of dialysis on electrophoretic mobility of catalase. I dialyzed 1 mL of sample against 500 mL of 60 mmol/L phosphate buffer (pH 7.4) for 24 h at 4#{176}C with two changes of buffer. The electrophoretic mobility of the dialyzed catalase was compared with that of untreated catalase stored at the same temperature. Effect of NADPH on electrophoretic mobility of catalase. I dissolved 1 mg of NADPH in 200 pL of catalasecontaining sample and incubated it for 48 h at 4#{176}C. The final NADPH concentration of this mixture was 6 mmolJL. The electrophoretic mobility of NADPHtreated catalase was compared with that of the untreated enzyme stored at the same temperature. Determination of catalase activity. Catalase activity was determined by spectrophotometric assay with hydrogen peroxide substrate, as described elsewhere (13). One unit (1 U) of catalase equals the enzyme activity that decomposes 1 Mmol of hydrogen peroxide in 608 at 37 #{176}C (initial hydrogen peroxide concentration of reaction mixture, 55 pmol/L). The reference mean (and range) of serum catalase activity concentration was 50.5 (14.3-104.8) kUIL. Pooled normal serum catalase was collected from sera
of hospitalized these patients
patients. The serum catalase activity of was within the reference range. Pathological sera with above-normal catalase activity were obtained from hospitalized patients suffering from hemolytic diseases or acute pancreatitis. Erythrocyte hemolysate. The erythrocyte hemolysate was prepared from freshly drawn blood from hospitalized patients with EDTA as anticoagulant. The erythrocytes were washed three times with PBS, then lysed with five volumes of distilled water, and centrifuged at 2500 x g for 20 mm. This hemolysate, containing catalase activity of about 10500 kUIL, was used for the dilution experiment. For detection of catalase-binding protein the hemolysate was diluted with PBS to yield a catalase activity of 500 kU/L. To test the possible role of SH groups, I diluted the washed erythrocytes ninefold with stabilizing solution and rapidly froze the resulting solution (16). After thawing and centrifuging, the supernate contained about 12400 kU of catalase activity per liter and was used for the dilution experiment. Human erythrocyte catalase. Human erythrocyte catalase was purified as described (11). Its absorptivity ratio at two wavelengths, e/e, was 1.21; its specific activity was 20.6 (SD 4.5) x iO U/g. Results
Catalase-binding protein. The results of separating the serum proteins and incubating them with erythrocyte catalase for detection of catalase-binding protein, if any, are shown in Figure 1. The white band for serum catalase is clearly visible on the sample and blank. Furthermore, a weak extra band became visible between the point of sample application and the y-globulin range (Figure 1, lane a). The surface of this new spot increased with the a-amylase activity of serum (lanes b and c), being largest when salivery a-amylase (50400 U/L) was applied (lane d). After blotting, it was noticeable that the surface of starch gel had become rough at
those spots that remained white during the staining procedure. Furthermore, if the surface of starch gel was scratched (lane e), a similar new spot was detectable. Contrary to these findings, the blanks (lanes A-D) stained regularly, without additional bands. Molecular mass estimation by gel-permeation chromatography. The molecular masses (±SD) for serum catalase (273.2 ± 7.1 kDa, n = 4) and for human erythrocyte catalase (271.3 ± 10.2 kDa, n = 5) were similar, and differed from that for bovine liver catalase (238.8 ± 7.7 kDa, n = 4). During the chromatography on Sepharose 6B, the serum catalase was diluted 60-fold, and had an electrophoretic mobility of 27.3 ± 0.9 mm. This value is higher than the 24.3 ± 0.9 mm measured for serum catalase beforeloadingitontothecolumnandisverysimilartothe 27.5 ± 1.0 mm for erythrocyte catalase (Figure 2). Molecular mass estimation with polyaciylamide gel electrophoresis in a nondenaturing system. This method of estimation also yielded very similar molecular masses for both serum catalases and human erythrocyte catalase (Table 1). Furthermore, 10-fold dilution of catalase from pathological serum, with either 60 minol/L phosphate buffer or normal sera, yielded different Hf values (0.176 ± 0.012 and 0.195 ± 0.08, respectively) but the same molecular mass (287.5 ± 10.9 and 283.9 ± 5.3 kDa). Partial purification of serum catalase with PhenylSepharose CL-4B. Most of the serum catalase (>90%) bound to Phenyl-Sepharose CL-4B, in contrast to the other plasma proteins. The subsequent elution eluted 67.3 (SD 11.1)% of the loosely-bound serum catalase with a 40.6 (SD 10.6)-fold increase in specific catalase activity (to 1342 ± 439 kUIg). Before the chromatography the electrophoretic mobility of serum catalase was 24.3 ± 1.2 mm. The mobility of the eluted catalase was 27.7 ± 0.9 mm, the same as that (27.5 ± 1.4 mm) of erythrocyte catalase (lanes b, d, e, Figure 2). Effect of dilution on electrophoretic mobility. The electrophoretic mobilities of diluted and untreated catalase samples are given in Figure 3 and Table 2. The electrophoretic mobility of catalase from the erythrocyte he-
abc
d
A
CD
FIg. 1. DetectIonof “catalase-binding protein” in serum with electrophoretic separation of serum proteinson cellulose acetate membrane and blotting onto starch gel a-e, samples Incubated with erythrocyte hemolysate (catalase activity 500 kU/L); A-D, blank samples incubatedwIth50 g/l. humanalbumin inPBS.After Incubationall samples were stained for catalase.a, b, cand A, B, Care serum samples with a-amytaseactivIty01350,1170, and4086 U/I, respectively;d, D, human sallvaiy a-am1ase, 50400 U/I; e, pin-scratchon the surfaceof starch gel before incubation with erythrocyte catalase. Affow, point of sample application; C, catalase band; CBP, likelycatalasespotdependingon a-amylase activity of sample
a
b
c
ci
4.
Fig. 2. Electrophoretlcmobility of serum catalase before and after chromatography , f, pathological serum; b, normal serum; c, serum catalase after gelpermeationchromatographyon Sepharose68 gel; d, erythrocyte catalase; e, serum catalase after hydrophobic chromatography on Phenyl-Sepharose CL-48 gel CLINICAL CHEMISTRY, Vol. 37, No. 12, 1991
2045
Table 1. Electrophoretlc Mobility In 4% Polyacrylamide Gel and Molecular Mass of Various Human Catalaae Samplesa EI.ctropho,.tlc mobIlity, R,
catsias. sours. Normal serum Pathological serum Diluted 10-fold with serum
Molscular mass, kDa
0.177 ± 0.008
± 10.7 290.4 ± 12.5
0.176 ± 0.012
287.5 ± 10.9
285.9
0.173 ± 0.013
Diluted10-fold
0.195 ± 0.08 283.9 ± 5.3 with buffer Erythrocyte 0.198 ± 0.08 282.4 ± 4.7 1Electrophoresls In 4% polyacr)lamlde gel, nondenaturing(n =4 each).
0) ‘I)
a a .4.a 1
I
0,0 5
10 ditution
15
2OfoLd
Fig. 3. Effect of dilution on electrophoreticmigration of catalase Isoforms Serum catalase was diluted with 60 mmoVL phosphate buffer, pH 7.4 (C); erythrocyte hemolysate catalase was diluted with pooled sera (0). Ba,s: ±1 SE. 0.0 representsthe migrationofuntreatedserum catalase;1.0, migrationof untreatederythrocyte catalase (27.6 ± 0.9 mm) did not change (27.3 ± 1.3 mm) after being diluted with the phosphate buffer. Dilution of the hemolysate with pooled human sera also yielded only one catalase band, but with a slower (24.2 ± 1.4 mm) migration. The dilution of pathological sera (24.5 ± 1.5 mm before dilution) with normal sera did not elicit changes (24.5 ± 1.4 mm after dilution) of electrophoretic mobility of catalase. Contrary to this, when these sera were diluted with the phosphate buffer, the electrophoretic mobility of the serum catalase increased (27.4 ± 1.3 mm), to about the value of the erythrocyte catalase (27.6 ± 0.9 mm). The partial purification of serum catalase increased its mobility from 24.5 ± 1.5 mm to 27.7 ± 1.3 mm. molysate
the dilution test this material behaved similarly to erythrocyte catalase. Its mobility (27.5 ± 1.0 mm) did not change (27.6 ± 0.9 mm) during its dilution with the phosphate buffer. Otherwise, its dilution with normal sera yielded one catalase band (24.6 ± 1.2 mm), the migration of which corresponded to that of serum catslase (24.5 ± 1.5 mm). Dilution of the catalase samples with either 60 mmol/L phosphate buffer or the SH-group stabilizing solution caused a
