- Email: [email protected]
Human serum inhibitors of collagenase as revealed by preparative isoelectric focusing
Climco Chimicu Acta. Elsevier/North-Holland
CCA
I I7 (I 98 I) 2 19-225 Biomedical
219
Press
1956
Human serum inhibitors of collagenase as revealed by preparative isoelectric focusing W. Borth
a,*, E.J. Menzel
a The Instituie of Immunology,
a, M. Salzer
b and C. Steffen
a
University of Vienna, 1090 Viennu (Austria) und h Hospital of Orthopuedlcs, Gersthof- Viennu (Austria) (Received
March
5 th, I98 1)
summary Single step separation of pooled normal human serum by means of preparative isoelectric focusing in the range from pH 3.5-9.8 revealed more heterogeneous inhibition of collagenolytic activity than previously reported. Essentially three inhibition zones were resolved. According to their electrophoretic behaviour the respective serum fractions displaying inhibitory activity were designated CX-,p- and y-collagenase inhibitors. The main component responsible for collagenase inhibition in the o-zone was found to be a,-macroglobulin. In the p-zone inhibitory activity focused around pH 6.3. In the y-range a non-dialysable cationic component focusing at pH 9.2 was also able to decrease collagenolytic activity derived from rheumatoid arthritis synovial culture supematant. These findings were supported by single step separation of serum on DEAE-anion exchange chromatography.
Introduction In previous investigations three serum inhibitors of rheumatoid arthritis (RA) collagenase (EC 3.4.24.3) were described; (Y,-antitrypsin ((Y,-AT), (~z-macroglobulin ((~z -M) and j3, -anticollagenase ( p, -AC). As far as (Y,-AT is concerned, contradictory results about the interaction of this inhibitor with collagenase were obtained [l- 31. a,-M (it4, = 725000) with an isoelectric point (PI) around pH 5.0-5.5 [4] reacts with all subclasses of endopeptidases (EC 3.4.21-24). The affinity of cu,-M to collagenase, which cleaves the triple helix of native collagen at a point three-quarters of the way from the N-terminus [5], is lower than to other proteinases; therefore a retarded inhibition of collagenase by (~z-M in vivo is likely. The large molecular size inhibitor is restricted to the vicinity of the blood vessels [6]. Small molecular size inhibitors could be better distributed in remote tissue sections and might be of greater importance in determining physiological collagen breakdown. There is evidence that a small molecular mass inhibitor (M, = 40000) acts specifically on collagenase [2]. In the course of further characterization of the 0009-898
I /8 1/000&0000/$02.75
0 198 1 Elsevier/North-Holland
Biomedical
Press
220
inhibitor with p, -electrophoretic mobility using isoelectric focusing we observed that the inhibition of collagenolytic activity by serum components is very heterogeneous. Since high collagenase activities are correlated with forced joint destruction [7,8] elucidation of biologic collagenase inhibitors may be of clinical value. Materials Carrier ampholytes “Servalyt”, pl marker proteins and high purity reagents were from Serva, Heidelberg, FRG; the chromatography media as well as the instrumentation for electrophoresis were from Pharmacia, Uppsala, Sweden; a digital pH-meter from Seibold, Vienna, Austria in combination with a surface electrode from Ingold, Zurich, Switzerland were used for pH determination; antisera against human serum components were from Behring AG, Marburg, FRG. Methods Collagenase assay with reconstituted fibrils (microgel-assay) [ 91 200 pg of thermally reconstituted [ i4C]collagen fibrils (25000 dpm) in 200 ~1 microgel assay buffer (MAb; 0.05 mol/l Tris-HCl, 0.15 mol/l NaCl, 0.005 mol/l CaC12, 0.2 g/l NaN,) per test tube were incubated with 10 ~1 crude purified enzyme solution and 50 ~1 of the dialysed serum fractions for 16 h at 36°C. The reaction was stopped with EDTA within the linearity of the assay; the 14C-activity, rendered soluble by the action of the enzyme, was determined by counting an aliquot of the supernatant. Controls were run with each experiment as follows: (1) 10 ~1 trypsin solution (0.5 g trypsin/l) released less than 6% of the radioactivity of the buffer controls; (2) 10 ~1 bovine serum albumin (Sigma; fraction V powder; 50 g/l) so as to detect non-specific solubilisation of collagen fibrils depending on the protein concentration in the reaction mixture; (3) 10 ~1 (Y*-M solution (1 g/l); inhibition control; inhibited collagenase up to 80%. Purified LY*-M was obtained by Zn2+ chelate affinity chromatography [lo]. Rheumatoid synovial collagenase was prepared as described [ 111. The crude RA synovial culture supernatants were lyophilised, dissolved in MAb and dialysed extensively against the same buffer. The final enzyme solution had a protein concentration of 15 g/l; 10 ~1 of this solution had a collagenolytic activity of 4pg collagen degradation/h as determined by the microgel assay. Partial purification of collagenase was obtained by the method of Bauer et al. [ 121. 10 ~1 of the purified enzyme solution (2 g protein/l) degraded 10 pg collagen/h. Preparation and (‘4C]acetic anhydride Iabelling of colIagen was performed as described [ 131. The specific activity was 125 000 dpm per mg pepsinised type I collagen. Preparative isoelectric focusing (pIEF) was carried out as described by Radola [14]; briefly: the thoroughly washed Sephadex G-200 gel was mixed with 10 ml Servalyt AG 2-l 1, 10 ml Servalyt TG 4-9, 3 ml Servalyt AG 3-5 and 13 ml pooled normal human serum. The final volume of the slurry was 280 ml. The electrodes were soaked in the above-mentioned ampholyte mixture, which was made for this
221
purpose 1.5 g/l. The experiment was run under N,-atmosphere at 7°C for 15 h. A total of 7 500 V/h were applied. The pH of the fractionated gel-23 fractions were obtained-was measured from anode to cathode. The proteins were eluted and subsequently dialysed against MAb. The serum dilution caused by these procedures corresponded to 1: 10. Analytical isoelectric focusing (aIEF) was performed as described elsewhere [ 151. The polyacrylamide gel was made T5% and C 3.3% [16]; ampholytes were used as mentioned above at 2.5 g/l; a total of 2000 V/h were applied; the pH gradient was estimated by means of isoelectric point marker proteins (Serva). The gel was fixed and stained according to the perchloric acid/Coomassie G-250 method [ 171. aIEF was used to prove and document the separation quality of pIEF (Fig. 3). Immobilized anti-a,-M antibodies: anti human a,-M immunoglobulins were isolated using Protein A Sepharose from rabbit antisera and coupled to CNBr-activated Sepharose CL 4B. Known inhibitqrs and other proteins were measured by single radial immunodiffusion using Partigen@ plates from Behring AG, Marburg, FRG. Results Three significant inhibition zones were resolved. The respective inhibitors were preliminarily called cy-, /3- and y-collagenase inhibitors (-CI) according to their electrophoretic mobility (Fig. 1). In the (Y-CI zone the presence of the following inhibitors were substantiated-fraction with maximum titer underlined-(Fig. 2): (Y,-antichymotrypsin (fraction 3-5; pH 4.29-4.72) inter-a-trypsin inhibitor (3-6;4; 4.29-4.93; additional immunoreactive IaI was found at higher pH values corresponding to aggregated or degraded IaI) [I81 (Y,-antitrypsin (5-6; 4.72-4.93) a,-macroglobuhn (6-10; 7; 4.93-5.70) antithrombin III (7-9; 5.25-5.50). To determine theinhibitory activity displayed in the a-zone in the absence of cu,-M, this inhibitor was removed from the fractions comprising anionic proteins (fr. l- 16) by immobilized anti-a,-M antibodies. The efficiency of the elimination was tested and the complete lack of cw,-M was verified. Thus inhibitory capacity was reduced from the original 75% to 15% in fraction 7, the serum fraction with the highest inhibition activity. The collagenolytic inhibition of the p-zone was not affected by this procedure. The p-C1 zone (fr. 11-14; pH 5.96-6.73) revealed up to 50% inhibition by fr. 12. This fraction was 10 times concentrated and immunoelectrophoresis carried out; mainly P-proteins could be detected such as lipoproteins, complement components and &-GP I. In the y-C1 zone (fr. 20-22; pH 8.70-9.40) a lower, though significant, inhibitory activity was found (up to 40%). These findings were essentially supported by single step separation of human serum on DEAE-anion exchange chromatography (Fig. 4). y-C1 coeluted with immunoglobulins and the complement component C,, and was not dialysable. p-C1 coeluted with P-proteins (e.g. lipoproteins, transferrin, complement components and
Prealb.
-
Albumin V-
V
fr.7
i
Fig. I. Screening electrophoresis in Tris-barbiturate buffer, (samples were concentrated IO times; i=a,- macroglobulin; v = normal human serum).
tit - antichymotrypstn inter. d. trypsinhibltor *t -antltr.ypsin d2 -macroglobul~n antithrombln
(main
12
fr.
ii
fr.20
iii
pH 8.6, of pIEF ii=/?,-transfertin:
iv
fractions 7, 12 and 20 iii=&-GP I: iv=‘?,,:
band)
Ill
75 60 45 % InhIbItIon 30 15 0
2
4
6
6
10
12
14
16
18
19
20
21
22
23
fractions
Fig, 2. 50 yl samples from pIEF were assayed for inhibition of crude (0 q !) or partially purified n) collagenax: arrows indicate the highest relative inhibitor content as measured by single (A----radial immunodiffusion: ( ) pH gradient (2 I “C).
223
1
2345670
9
12
13
TR
17
18
13
20
21
22
23
Fig. 3. Analytical PAGIF (T 5%. C 3.3%); fraction l-23 from pIEF; TR= transferrin 5.6); IPM = isoelectric point marker proteins. Samples were applied near the anode.
Fig. 4. DFAE-Sephacel” anion exchange ) A,,,, (- - - - - -) salt gradient;
(
chromatography; arrows indicating
column 2.5X 22 cm; a-, p- and +JI zones.
.
/PM
(main band
flow rate
pH
40 ml/h;
&-GP I). With (u-Cl emerged a,-M. Besides these three clear-cut inhibition zones, crude and also purified collagenase were inhibited by fraction 3 (pH 4.30), which had the highest relative cu,-antichymotrypsin content. In addition fraction 16, situated between /I- and y-Cl, inhibited slightly crude collagenase.
Discussion In this study, three zones with significant inhibitory activity on mammalian collagenase were resolved. (~z-macroglobulin is mainly responsible for collagenolytic inhibition of the cw-Cl zone as confirmed in the present investigation, comparing the of the inhibitory action in the presence of (Ye-M (up to 80%) and after elimination inhibitor. The remaining inhibitory activity of 15% could be due to another neutral protease inhibitor with broad specificity such as antithrombin Ill (M, = 65 000) the highest relative content of which was detected in fraction 7. It is of interest that antithrombin Ill and a,-AT have a common evolutionary origin, expressed by extensive homology of their primary structure [19]. The inhibitory capacity of the /?-Cl zone was not reduced after immunoabsorbtion of a,-M, indicating the presence of another potent though low molecular inhibitor of collagenase. In the y-Cl zone weak though significant inhibition around pH 9 was found.
224
While cationic collagenase inhibitors derived from different tissue explants have been described [20], to our knowledge no inhibitory serum component with yelectrophoretic mobility has been described before. Since it has been shown that mammalian collagenase is able to interact with the yZ-globulin C ,q via its collagen-like fragment, it cannot be excluded that the inhibition of the y-zone is caused by this complement component by way of substrate competitioir [21,22]. In our hands aIEF of C,, did not result in a clear-cut pattern, even in large pore agarose gels. However, on DEAE anion chromatography, C,, could be detected together with immunoglobulins in the y-zone. Concluding, it is evident from this study that the most important collagenase inhibitory activity is mainly due to (Y*-M present in the a-zone followed by inhibitory activity in the P-region; the relative importance of the y-inhibitory activity remains to be substantiated by further studies. Acknowledgement This work was supported Forschung, Nr. 4027.
by the Fonds
zur Forderung
der wissenschaftlichen
References I Werb Z, Burleigh C, Barret AJ, Starkey PM. The interaction of a,-macroglobulin with proteinases. Binding and inhibition of mammalian collagenases and other metal proteinases. Biochem J 1974: 139: 359-368. 2 Wooley DE, Roberts DR, Evanson JM. Small molecular weight p, serum protein which specifically inhibits human collagenases. Nature 1976; 261: 325-327. 3 Ohlsson K. Polymorphonuclear leucocyte collagenase. In: Wooley DE, Evanson JM, eds. Collagenase in normal and pathological connective tissues. Chichester: John Wiley & Sons, 1980: 209-222. 4 Frenoy JP, Bourillon R. Studies on the structure of human a,-M: analysis of the microheterogeneity by isoelectric focusing. Biochim Biophys Acta 1974; 371: I68- 176. 4 Seifter S, Harper E. Collagenases. In: Perlmann GE, Lorand L, eds. Methods in enzymology, Vol. 19. New York: Academic Press, 1970: 6 13-634. 6 Ballidin G, Laurel1 CB, Ohlsson K. Increased catabolism of a,-macroglobulins after intravenous infusion of trypsin a,-antitrypsin complexes in dogs. Hoppe-Seyler’s Z Physiol Chem 1978; 359: 699-708. 7 Harris ED, Dibona DR. Krane SM. Collagenases in human synovia. J Clin Invest 1969; 48: 2104. 8 Menzel J, Steffen C. Detection of collagenase activity in rheumatoid arthritis synovial fluids using “C-1abeled collagen. Z Rheumatol 1977; 36: 364-377. 9 Wooley DE, Glanville RW, Crossley MJ, Evanson JM. Purification of rheumatoid synovial colIagenase and its action on soluble and insoluble collagen. Eur J Biochem 1975; 54: 61 l-622. IO Porath J, Carlsson J, Olsson I, Belfrage G. Metal affinity chromatography, a new approach to protein fractionation. Nature 1975; 258: 598-599. I I Steffen C, Zeitelhofer J, Zielinski CH, Menzel J, Smolen J. Collagenase-induced experimental arthritis. Z Rheumatol 1979; 38: I-IO. 12 Bauer EA. Jeffrey JJ, Eisen AE. Preparation of three vertebrate collagenases in pure form. Biochim Biophys Res Commun I971 ; 44: 8 13. I3 Menzel J. Radioimmunoassay for anticollagen-antibodies using “C-labe1ed collagen. J Immunol Meth 1977; 15: 77-95. I4 Radola BJ. Isoelectric Biochim Biophys Acta
focusing in layers of granulated 1974; 386: 18 I- 195.
gels. II. Preparative
isoelectric
focusing.
225
15 Delincee H, Radola BJ. Determination of isoelectric points in thin-layer isoelectric focusing: the importance of attaining steady state and the role of CO, interference. Anal Biochem 1978; 90: 609-623. 16 Chrambach A, Nguyen NY. Electrofocusing in buffers: Formation of natural gradients, flexibility, gradient stability, relation to isotachophoresis and preparative potential. In: Radola BJ, Graesslin D, eds. Electrophocusing and isotachophoresis. Berlin: Walter de Gruyter, 1977: 58. footnote. 17 Reisner AH, Nemes P, Bucholth C. The use of Coomassie Brillant Blue G 250 perchloric acid solution for staining in electrophoresis and isoelectric focusing in polyacrylamide gels. Anal Biochem 1975; 64: 509-5 19. 18 Steinbuch M. The inter-a-trypsin inhibitor. In: Lorand L, ed. Methods in enzymology, Vol. 45: 76 I-772. 19 Magnusson S. Thrombin and prothrombin. In: Blomback B. Hanson LA, eds. Plasma proteins. Chichester: John Wiley & Sons, 1980: 254-276. 20 Murphy G, Sellers A. The extracellular regulation of collagenase activity. In: Wooley DE, Evanson JM, eds. Collagenase in normal and pathological connective tissues. Chichester: John Wiley & Sons, 1980: 65-81. 21 Menzel J. Wechselwirkungen zwischen Kollagen und C,,: Ihre Bedeutung fur die Rheumatologie. Z Rheumatol 1980; 39: 251-283. (Abstract in English.) 22 Nagai Y, Shinkai H, Ninomiya Y. The release of collagenase inhibitors from procollagen additional peptides by pepsin treatment. Proc Japan Acad 1978; 54, Ser. B: 140-144.
CCA
I I7 (I 98 I) 2 19-225 Biomedical
219
Press
1956
Human serum inhibitors of collagenase as revealed by preparative isoelectric focusing W. Borth
a,*, E.J. Menzel
a The Instituie of Immunology,
a, M. Salzer
b and C. Steffen
a
University of Vienna, 1090 Viennu (Austria) und h Hospital of Orthopuedlcs, Gersthof- Viennu (Austria) (Received
March
5 th, I98 1)
summary Single step separation of pooled normal human serum by means of preparative isoelectric focusing in the range from pH 3.5-9.8 revealed more heterogeneous inhibition of collagenolytic activity than previously reported. Essentially three inhibition zones were resolved. According to their electrophoretic behaviour the respective serum fractions displaying inhibitory activity were designated CX-,p- and y-collagenase inhibitors. The main component responsible for collagenase inhibition in the o-zone was found to be a,-macroglobulin. In the p-zone inhibitory activity focused around pH 6.3. In the y-range a non-dialysable cationic component focusing at pH 9.2 was also able to decrease collagenolytic activity derived from rheumatoid arthritis synovial culture supematant. These findings were supported by single step separation of serum on DEAE-anion exchange chromatography.
Introduction In previous investigations three serum inhibitors of rheumatoid arthritis (RA) collagenase (EC 3.4.24.3) were described; (Y,-antitrypsin ((Y,-AT), (~z-macroglobulin ((~z -M) and j3, -anticollagenase ( p, -AC). As far as (Y,-AT is concerned, contradictory results about the interaction of this inhibitor with collagenase were obtained [l- 31. a,-M (it4, = 725000) with an isoelectric point (PI) around pH 5.0-5.5 [4] reacts with all subclasses of endopeptidases (EC 3.4.21-24). The affinity of cu,-M to collagenase, which cleaves the triple helix of native collagen at a point three-quarters of the way from the N-terminus [5], is lower than to other proteinases; therefore a retarded inhibition of collagenase by (~z-M in vivo is likely. The large molecular size inhibitor is restricted to the vicinity of the blood vessels [6]. Small molecular size inhibitors could be better distributed in remote tissue sections and might be of greater importance in determining physiological collagen breakdown. There is evidence that a small molecular mass inhibitor (M, = 40000) acts specifically on collagenase [2]. In the course of further characterization of the 0009-898
I /8 1/000&0000/$02.75
0 198 1 Elsevier/North-Holland
Biomedical
Press
220
inhibitor with p, -electrophoretic mobility using isoelectric focusing we observed that the inhibition of collagenolytic activity by serum components is very heterogeneous. Since high collagenase activities are correlated with forced joint destruction [7,8] elucidation of biologic collagenase inhibitors may be of clinical value. Materials Carrier ampholytes “Servalyt”, pl marker proteins and high purity reagents were from Serva, Heidelberg, FRG; the chromatography media as well as the instrumentation for electrophoresis were from Pharmacia, Uppsala, Sweden; a digital pH-meter from Seibold, Vienna, Austria in combination with a surface electrode from Ingold, Zurich, Switzerland were used for pH determination; antisera against human serum components were from Behring AG, Marburg, FRG. Methods Collagenase assay with reconstituted fibrils (microgel-assay) [ 91 200 pg of thermally reconstituted [ i4C]collagen fibrils (25000 dpm) in 200 ~1 microgel assay buffer (MAb; 0.05 mol/l Tris-HCl, 0.15 mol/l NaCl, 0.005 mol/l CaC12, 0.2 g/l NaN,) per test tube were incubated with 10 ~1 crude purified enzyme solution and 50 ~1 of the dialysed serum fractions for 16 h at 36°C. The reaction was stopped with EDTA within the linearity of the assay; the 14C-activity, rendered soluble by the action of the enzyme, was determined by counting an aliquot of the supernatant. Controls were run with each experiment as follows: (1) 10 ~1 trypsin solution (0.5 g trypsin/l) released less than 6% of the radioactivity of the buffer controls; (2) 10 ~1 bovine serum albumin (Sigma; fraction V powder; 50 g/l) so as to detect non-specific solubilisation of collagen fibrils depending on the protein concentration in the reaction mixture; (3) 10 ~1 (Y*-M solution (1 g/l); inhibition control; inhibited collagenase up to 80%. Purified LY*-M was obtained by Zn2+ chelate affinity chromatography [lo]. Rheumatoid synovial collagenase was prepared as described [ 111. The crude RA synovial culture supernatants were lyophilised, dissolved in MAb and dialysed extensively against the same buffer. The final enzyme solution had a protein concentration of 15 g/l; 10 ~1 of this solution had a collagenolytic activity of 4pg collagen degradation/h as determined by the microgel assay. Partial purification of collagenase was obtained by the method of Bauer et al. [ 121. 10 ~1 of the purified enzyme solution (2 g protein/l) degraded 10 pg collagen/h. Preparation and (‘4C]acetic anhydride Iabelling of colIagen was performed as described [ 131. The specific activity was 125 000 dpm per mg pepsinised type I collagen. Preparative isoelectric focusing (pIEF) was carried out as described by Radola [14]; briefly: the thoroughly washed Sephadex G-200 gel was mixed with 10 ml Servalyt AG 2-l 1, 10 ml Servalyt TG 4-9, 3 ml Servalyt AG 3-5 and 13 ml pooled normal human serum. The final volume of the slurry was 280 ml. The electrodes were soaked in the above-mentioned ampholyte mixture, which was made for this
221
purpose 1.5 g/l. The experiment was run under N,-atmosphere at 7°C for 15 h. A total of 7 500 V/h were applied. The pH of the fractionated gel-23 fractions were obtained-was measured from anode to cathode. The proteins were eluted and subsequently dialysed against MAb. The serum dilution caused by these procedures corresponded to 1: 10. Analytical isoelectric focusing (aIEF) was performed as described elsewhere [ 151. The polyacrylamide gel was made T5% and C 3.3% [16]; ampholytes were used as mentioned above at 2.5 g/l; a total of 2000 V/h were applied; the pH gradient was estimated by means of isoelectric point marker proteins (Serva). The gel was fixed and stained according to the perchloric acid/Coomassie G-250 method [ 171. aIEF was used to prove and document the separation quality of pIEF (Fig. 3). Immobilized anti-a,-M antibodies: anti human a,-M immunoglobulins were isolated using Protein A Sepharose from rabbit antisera and coupled to CNBr-activated Sepharose CL 4B. Known inhibitqrs and other proteins were measured by single radial immunodiffusion using Partigen@ plates from Behring AG, Marburg, FRG. Results Three significant inhibition zones were resolved. The respective inhibitors were preliminarily called cy-, /3- and y-collagenase inhibitors (-CI) according to their electrophoretic mobility (Fig. 1). In the (Y-CI zone the presence of the following inhibitors were substantiated-fraction with maximum titer underlined-(Fig. 2): (Y,-antichymotrypsin (fraction 3-5; pH 4.29-4.72) inter-a-trypsin inhibitor (3-6;4; 4.29-4.93; additional immunoreactive IaI was found at higher pH values corresponding to aggregated or degraded IaI) [I81 (Y,-antitrypsin (5-6; 4.72-4.93) a,-macroglobuhn (6-10; 7; 4.93-5.70) antithrombin III (7-9; 5.25-5.50). To determine theinhibitory activity displayed in the a-zone in the absence of cu,-M, this inhibitor was removed from the fractions comprising anionic proteins (fr. l- 16) by immobilized anti-a,-M antibodies. The efficiency of the elimination was tested and the complete lack of cw,-M was verified. Thus inhibitory capacity was reduced from the original 75% to 15% in fraction 7, the serum fraction with the highest inhibition activity. The collagenolytic inhibition of the p-zone was not affected by this procedure. The p-C1 zone (fr. 11-14; pH 5.96-6.73) revealed up to 50% inhibition by fr. 12. This fraction was 10 times concentrated and immunoelectrophoresis carried out; mainly P-proteins could be detected such as lipoproteins, complement components and &-GP I. In the y-C1 zone (fr. 20-22; pH 8.70-9.40) a lower, though significant, inhibitory activity was found (up to 40%). These findings were essentially supported by single step separation of human serum on DEAE-anion exchange chromatography (Fig. 4). y-C1 coeluted with immunoglobulins and the complement component C,, and was not dialysable. p-C1 coeluted with P-proteins (e.g. lipoproteins, transferrin, complement components and
Prealb.
-
Albumin V-
V
fr.7
i
Fig. I. Screening electrophoresis in Tris-barbiturate buffer, (samples were concentrated IO times; i=a,- macroglobulin; v = normal human serum).
tit - antichymotrypstn inter. d. trypsinhibltor *t -antltr.ypsin d2 -macroglobul~n antithrombln
(main
12
fr.
ii
fr.20
iii
pH 8.6, of pIEF ii=/?,-transfertin:
iv
fractions 7, 12 and 20 iii=&-GP I: iv=‘?,,:
band)
Ill
75 60 45 % InhIbItIon 30 15 0
2
4
6
6
10
12
14
16
18
19
20
21
22
23
fractions
Fig, 2. 50 yl samples from pIEF were assayed for inhibition of crude (0 q !) or partially purified n) collagenax: arrows indicate the highest relative inhibitor content as measured by single (A----radial immunodiffusion: ( ) pH gradient (2 I “C).
223
1
2345670
9
12
13
TR
17
18
13
20
21
22
23
Fig. 3. Analytical PAGIF (T 5%. C 3.3%); fraction l-23 from pIEF; TR= transferrin 5.6); IPM = isoelectric point marker proteins. Samples were applied near the anode.
Fig. 4. DFAE-Sephacel” anion exchange ) A,,,, (- - - - - -) salt gradient;
(
chromatography; arrows indicating
column 2.5X 22 cm; a-, p- and +JI zones.
.
/PM
(main band
flow rate
pH
40 ml/h;
&-GP I). With (u-Cl emerged a,-M. Besides these three clear-cut inhibition zones, crude and also purified collagenase were inhibited by fraction 3 (pH 4.30), which had the highest relative cu,-antichymotrypsin content. In addition fraction 16, situated between /I- and y-Cl, inhibited slightly crude collagenase.
Discussion In this study, three zones with significant inhibitory activity on mammalian collagenase were resolved. (~z-macroglobulin is mainly responsible for collagenolytic inhibition of the cw-Cl zone as confirmed in the present investigation, comparing the of the inhibitory action in the presence of (Ye-M (up to 80%) and after elimination inhibitor. The remaining inhibitory activity of 15% could be due to another neutral protease inhibitor with broad specificity such as antithrombin Ill (M, = 65 000) the highest relative content of which was detected in fraction 7. It is of interest that antithrombin Ill and a,-AT have a common evolutionary origin, expressed by extensive homology of their primary structure [19]. The inhibitory capacity of the /?-Cl zone was not reduced after immunoabsorbtion of a,-M, indicating the presence of another potent though low molecular inhibitor of collagenase. In the y-Cl zone weak though significant inhibition around pH 9 was found.
224
While cationic collagenase inhibitors derived from different tissue explants have been described [20], to our knowledge no inhibitory serum component with yelectrophoretic mobility has been described before. Since it has been shown that mammalian collagenase is able to interact with the yZ-globulin C ,q via its collagen-like fragment, it cannot be excluded that the inhibition of the y-zone is caused by this complement component by way of substrate competitioir [21,22]. In our hands aIEF of C,, did not result in a clear-cut pattern, even in large pore agarose gels. However, on DEAE anion chromatography, C,, could be detected together with immunoglobulins in the y-zone. Concluding, it is evident from this study that the most important collagenase inhibitory activity is mainly due to (Y*-M present in the a-zone followed by inhibitory activity in the P-region; the relative importance of the y-inhibitory activity remains to be substantiated by further studies. Acknowledgement This work was supported Forschung, Nr. 4027.
by the Fonds
zur Forderung
der wissenschaftlichen
References I Werb Z, Burleigh C, Barret AJ, Starkey PM. The interaction of a,-macroglobulin with proteinases. Binding and inhibition of mammalian collagenases and other metal proteinases. Biochem J 1974: 139: 359-368. 2 Wooley DE, Roberts DR, Evanson JM. Small molecular weight p, serum protein which specifically inhibits human collagenases. Nature 1976; 261: 325-327. 3 Ohlsson K. Polymorphonuclear leucocyte collagenase. In: Wooley DE, Evanson JM, eds. Collagenase in normal and pathological connective tissues. Chichester: John Wiley & Sons, 1980: 209-222. 4 Frenoy JP, Bourillon R. Studies on the structure of human a,-M: analysis of the microheterogeneity by isoelectric focusing. Biochim Biophys Acta 1974; 371: I68- 176. 4 Seifter S, Harper E. Collagenases. In: Perlmann GE, Lorand L, eds. Methods in enzymology, Vol. 19. New York: Academic Press, 1970: 6 13-634. 6 Ballidin G, Laurel1 CB, Ohlsson K. Increased catabolism of a,-macroglobulins after intravenous infusion of trypsin a,-antitrypsin complexes in dogs. Hoppe-Seyler’s Z Physiol Chem 1978; 359: 699-708. 7 Harris ED, Dibona DR. Krane SM. Collagenases in human synovia. J Clin Invest 1969; 48: 2104. 8 Menzel J, Steffen C. Detection of collagenase activity in rheumatoid arthritis synovial fluids using “C-1abeled collagen. Z Rheumatol 1977; 36: 364-377. 9 Wooley DE, Glanville RW, Crossley MJ, Evanson JM. Purification of rheumatoid synovial colIagenase and its action on soluble and insoluble collagen. Eur J Biochem 1975; 54: 61 l-622. IO Porath J, Carlsson J, Olsson I, Belfrage G. Metal affinity chromatography, a new approach to protein fractionation. Nature 1975; 258: 598-599. I I Steffen C, Zeitelhofer J, Zielinski CH, Menzel J, Smolen J. Collagenase-induced experimental arthritis. Z Rheumatol 1979; 38: I-IO. 12 Bauer EA. Jeffrey JJ, Eisen AE. Preparation of three vertebrate collagenases in pure form. Biochim Biophys Res Commun I971 ; 44: 8 13. I3 Menzel J. Radioimmunoassay for anticollagen-antibodies using “C-labe1ed collagen. J Immunol Meth 1977; 15: 77-95. I4 Radola BJ. Isoelectric Biochim Biophys Acta
focusing in layers of granulated 1974; 386: 18 I- 195.
gels. II. Preparative
isoelectric
focusing.
225
15 Delincee H, Radola BJ. Determination of isoelectric points in thin-layer isoelectric focusing: the importance of attaining steady state and the role of CO, interference. Anal Biochem 1978; 90: 609-623. 16 Chrambach A, Nguyen NY. Electrofocusing in buffers: Formation of natural gradients, flexibility, gradient stability, relation to isotachophoresis and preparative potential. In: Radola BJ, Graesslin D, eds. Electrophocusing and isotachophoresis. Berlin: Walter de Gruyter, 1977: 58. footnote. 17 Reisner AH, Nemes P, Bucholth C. The use of Coomassie Brillant Blue G 250 perchloric acid solution for staining in electrophoresis and isoelectric focusing in polyacrylamide gels. Anal Biochem 1975; 64: 509-5 19. 18 Steinbuch M. The inter-a-trypsin inhibitor. In: Lorand L, ed. Methods in enzymology, Vol. 45: 76 I-772. 19 Magnusson S. Thrombin and prothrombin. In: Blomback B. Hanson LA, eds. Plasma proteins. Chichester: John Wiley & Sons, 1980: 254-276. 20 Murphy G, Sellers A. The extracellular regulation of collagenase activity. In: Wooley DE, Evanson JM, eds. Collagenase in normal and pathological connective tissues. Chichester: John Wiley & Sons, 1980: 65-81. 21 Menzel J. Wechselwirkungen zwischen Kollagen und C,,: Ihre Bedeutung fur die Rheumatologie. Z Rheumatol 1980; 39: 251-283. (Abstract in English.) 22 Nagai Y, Shinkai H, Ninomiya Y. The release of collagenase inhibitors from procollagen additional peptides by pepsin treatment. Proc Japan Acad 1978; 54, Ser. B: 140-144.