- Email: [email protected]
Isolation of the fifth component of the bovine complement system
Veterinary Immunology and Immunopathology, 24 (1990) 301-312 Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands
301
Isolation of the Fifth Component of the B o v i n e Complement S y s t e m W.P. ASTON, A. MHATRE and J. MACRAE
Department of Microbiology and Immunology, Queen's University, Kingston, Ont. K7L 3N6 (Canada) (Accepted 17 October 1989 )
ABSTRACT Aston, W.P., Mhatre, A. and Macrae, J., 1990. Isolation of the fifth component of the bovine complement system. Vet. Immunol. Immunopathol., 24: 301-312. Bovine C5 has been isolated from fresh bovine serum by a five-step procedure: polyethylene glycol precipitation, sequential ion-exchange chromatography on DEAE-Sephacel and CM-Sephadex, hydroxylapatite chromatography, and affinity chromatography. The purified C5 was a protein of apparent molecular weight 202 000 ___9000 composed of two chains: an alpha-chain of molecular weight 127 000 _+5000 and a beta-chain of molecular weight 74 000 _+2000. The alpha-chain was cleaved by Sepharose-CVF.Bb (a cobra venom factor (CVF)-induced C3/C5 alternative pathway convertase) in the absence of any C3 or C3b. The monocarboxylic acid form of K-76, a sesquiterpene compound isolated from the culture filtrates of Stachybotris complementi, inhibited the alternative pathway of bovine serum, and the inhibited hemolytic activity was restored, in a dose dependent manner, by bovine C5. This provided the basis for a C5 functional assay throughout the purification procedure. The purified C5 showed species specificity and was functionally distinct from bovine C3.
INTRODUCTION
The fifth component of complement, C5, is a plasma glycoprotein that is comprised of two disulphide linked polypeptide chains, an alpha- and a betachain (Tack et al., 1979). Activation of C5 involves the cleavage of the alpha chain by either the classical or alternative pathway C5-convertases, and results in the generation of two important biologically active fragments, C5a and C5b (Muller-Eberhard, 1985). The human C5a fragment has anaphylatoxin, chemotactic and immunoregulatory activities, and the C5b fragment provides the nucleus for the assembly of the complement membrane attack complex (MAC) which is responsible for the lytic action of the complement system. C5 has been isolated from a variety of mammals, including humans {Tack et al., 0165-2427/90/$03.50
© 1990 Elsevier Science Publishers B.V.
302
W.P. ASTON ET AL.
1979 and 1981), guinea pigs (Kinoshita et al., 1981), mice (Fukuoka et al., 1984) and rabbits (Giclas et al., 1981 ). Human C5 has an apparent molecular weight of 190 000 (Tack et al., 1979) and it has structural homology with C3 and C4 (Tack et al., 1981). This paper reports the isolation of C5 from bovine serum and presents a novel assay that was used to identify C5 during purification. The assay is based on the ability of C5 to reconstitute the alternative pathway hemolytic activity in sera treated with the monocarboxylic acid derivative of K- 76 (K- 76 COOH ), a sesquiterpene anticomplementary compound, isolated from the culture filtrate of Stachybotrys complementi nov. sp. K-76 (Hong et al., 1979). MATERIALSAND METHODS Bovine serum, bovine serum deficient in C3 (R3), and human erythrocytes were obtained as previously described (Pang and Aston, 1978; Mhatre and Aston, 1987b ). Rabbit and guinea pig erythrocytes were purchased from Woodlyn Laboratories Ltd., Guelph, Ont. Bovine factors B and D, and C3 were prepared by the methods of Pang and Aston (1978) and Menger and Aston (1984, 1985), and human C5 was prepared by Cordis Laboratories, Miami, FL. All column chromatography media were purchased from Pharmacia Fine Chemicals, Uppsala, Sweden, except the Bio-Gel H T hydroxylapatite (BioRad, Richmond, CA). The monocarboxylic acid form of K-76 (K-76 COOH) was a gift from Dr. Kozo Inoue, Osaka University Medical School, Japan. All other chemicals were obtained from Sigma, St. Louis, MO. PBS-EGTA, was phosphate buffered saline (PBS) at physiological ionic strength and pH 7.2, containing 5 m M MgC12 and 15 m M ethylene glycol bis-amino tetraacetate (EGTA). PBS-EDTA was PBS containing 20 m M ethylene diamine tetraacetate (EDTA). GBS-EGTA was barbital buffered saline at physiological ionic strength and pH 7.2 (Rapp and Borsos, 1970 ) containing 0.1% gelatin, 15 m M EGTA and 4 m M MgC12. GGBS-EGTA was GBS-EGTA adjusted to an ionic strength of 0.065 with aqueous glucose (Rapp and Borsos, 1970).
Hemolytic assays C5. Bovine and human R5 reagents (sera depleted of C5) were prepared by mixing the appropriate serum with an equal volume of K-76 COOH (5 m g / m l ) in GGBS-EGTA and gently agitating the mixture for 18 h at 4 ° C. The mixture was extensively dialyzed, at 4 ° C, against PBS-EDTA and then stored in small aliquots at - 70 ° C. C5 samples (10 #1) were assayed for their ability to reconstitute the alternative pathway hemolytic function of an R5 reagent (210/tl). The bovine R5 reagent was 10 ttl of bovine R5, 10 ttl 8% (v/v) guinea pig erythrocytes, and 190/tl of the diluent, GGBS-EGTA. The h u m a n R5 reagent was 20 #l of human R5, 10 ttl of 8% (v/v) guinea pig erythrocytes, and 180/tl of GGBS-EGTA. The
ISOLATION OF THE FIFTH COMPONENT OF THE BOVINE COMPLEMENT SYSTEM
303
mixtures (R5 r e a g e n t + C 5 sample) were incubated at 37°C for 30 min, and then the reaction was stopped by the addition of 2.0 ml ice-cold GGBS-EDTA. The extent of target cell lysis was determined by measuring the absorption of the fluid phase at 412 nm, and expressed as a percentage of the complete lysis of the erythrocytes in the reaction mixture. The substitution of diluent for the C5 sample served as a negative control. All assays were performed in triplicate. The tube hemolytic assays for C5 were readily adapted to hemolytic diffusion plate assays, similar to those described for bovine factor D and C3 (Menger and Aston, 1984; Mhatre and Aston, 1987b). The hemolytic plates were prepared by pouring 6 ml of a molten mixture of 1% agarose in GGBS-EGTA, containing 0.75% guinea pig erythrocytes and 5-10% R5, onto a 5 × 7.5 cm polyester film (GelBond, Pharmacia Fine Chemicals, Uppsala, Sweden) and then allowing the mixture to solidify. Samples of C5 (7.5/A) introduced into wells (4 m m diameter) in the plate produced rings of lysis after several hours at 37 ° C. Buffer controls produced no lytic rings.
C3 and factor H. Bovine C3 activity was determined by the alternative pathway method of Mhatre and Aston (1987b), using guinea pig erythrocytes as the target activators. Bovine factor H activity was assessed by the capacity to inhibit lysis of human erythrocytes by a limited amount of bovine serum (Mhatre and Aston, 1987a). Isolation of bovine C5 The C5 protein was isolated from fresh bovine serum by polyethylene glycol (PEG) precipitation and a combination of chromatographic procedures on DEAE-Sephacel, CM-Sephadex C-50 and hydroxylapatite adapted from the procedures described by Tack and Prahl (1976), Tack et al. (1979) and Sim et al. (1981). A column of Sepharose-factor H was used as the final step, to remove any C3 contamination of the isolated C5 material. The Sepharosefactor H (2 mg of factor H per ml of Sepharose 4B ) was prepared as described by Scott and Fothergill ( 1982 ). Eight hundred ml of bovine serum at 4°C was mixed with 400 ml of 15% (w/v) PEG 4000 (polyethylene glycol) in 0.1 M phosphate buffer, pH 7.4 containing 0.15 M NaC1, 0.015 M EDTA and 0.03 M epsilon amino caproic acid (EACA). The resultant precipitate was removed by centrifugation. The supernatant, containing 5% PEG, was adjusted to 12% PEG by the slow addition of an appropriate volume of 33% PEG in the phosphate buffer. The resultant precipitate was collected by centrifugation, washed with 12% PEG and then dissolved in 5 m M phosphate buffer, pH 7.4 containing 20 m M NaC1, 5 m M EDTA and 10 m M EACA. The solution was applied to a DEAE-Sephacel colu m n (5 × 30 cm) equilibrated in the 5 m M phosphate buffer. The column was washed with the equilibrating buffer and then developed, at a flow rate of 60
304
W.P. ASTON ET AL,
ml/h, with a 3 1 salt gradient to 0.3 M NaC1. Column fractions (10 ml) containing C5 activity (fractions 80 to 105) were pooled, concentrated and dialysed against a 20 m M tris-acetate buffer, pH 6.8 containing 20 m M NaC1 and 5 m M EDTA. The material was applied to a 5 × 25 cm column of CM-Sephadex equilibrated with the 20 m M tris-acetate buffer. The column was washed with the equilibrating buffer and then developed at a flow rate of 40 m l / h with a 1.5 1 gradient to 0.3 M NaC1. Fractions (10 ml) containing C5 activity (fractions 145 to 175) were pooled and then dialyzed against 5 m M sodium phosphate buffer at pH 7.6, containing 5 m M EDTA and 100 m M NaC1. One hundred and fifty ml of the sample was applied to an hydroxylapatite column (2.5 × 30 cm) equilibrated in the 5 m M phosphate buffer. The column was washed with the equilibrating buffer and then developed (flow rate 30 m l / h ) sequentially: (a) with 700 ml of the 5 m M p h o s p h a t e buffer containing 2 M NaC1, (b) with 900 ml of the 5 m M phosphate buffer containing 0.075 M potassium phosphate, (c) with 700 ml of the 5 m M phosphate buffer containing 0.15 M potassium phosphate and finally, (d) with 1000 ml of the 5 m M phosphate buffer containing 0.3 M potassium phosphate. The C5 was eluted from the column with the 5 m M phosphate buffer containing 0.15 M potassium phosphate. Column fractions containing the peak of the C5 activity were pooled (100 ml) and dialyzed against PBS-EDTA at pH 8.0, and then treated with 300 m M methylamine (CH3NH2) for 18 h at 4°C to convert any contaminating C3 to C3 (CH~NH2) (Mhatre and Aston, 1987b). The sample was dialyzed against 25 m M tris-HC1 buffer at pH 7.6 and passed through a column ( 1 × 1 0 cm) containing the Sepharose-factor H to remove the C3 (CH3NH:). The purified C5 was concentrated by positive pressure ultrafiltration and stored at - 70 ° C. All column chromatography procedures were performed at 4°C, all buffers contained 0.04% sodium azide, and at each step C5 was identified by a hemolytic plate assay.
Polyacrylamide gel electrophoresis (PAGE) Slab gel electrophoresis (9% polyacrylamide) in the presence of sodium dodecyl sulphate (SDS) was performed as previously described (Menger and Aston, 1984).
Protein concentration determinations A bicinchonnic acid (BCA) micro protein assay system was performed according to the monograph "BCA Protein Assay Reagent" ( # 23225 from Pierce Chemical Co., Rockford, IL).
ISOLATION OF T H E FIFTH COMPONENT OF THE BOVINE COMPLEMENT SYSTEM
305
Cleavage of C3 and C5 by Sepharose-CVF.Bb Sepharose-CVF and Sepharose-CVF.Bb were prepared as previously described (Menger and Aston, 1985); the CVF (cobra venom factor), was isolated from the venom of Naja naja atra (Sigma, St. Louis, MO) and factor B was isolated from bovine serum. Aliquots (100 ttl) of a 50% (v/v) suspension of the immobilized CVF or CVF.Bb were washed with the PBS-EGTA. The packed Sepharose-CVF or Sepharose-CVF-Bb preparations were then resuspended in 30 #l of either bovine C3 (approx. 300 ttg/ml ) or bovine C5 (approx. 300 ttg/ml) in PBS-EGTA, and the reaction mixtures were maintained at 37 ° C for 45 min. The insoluble matrix was pelleted by centrifugation and 25 pl of the supernatant was analyzed by SDS-PAGE. RESULTS
The alternative pathway hemolytic activity of bovine serum was inhibited by K-76 COOH. The activity could be restored, in a dose-dependent manner by bovine serum depleted of factor B (50°C, 20 min) and C3 (methylamine treatment), and also by purified bovine C5; this provided the basis of an assay for bovine C5. The optimum treatment of serum to produce an effective R5 was the addition of 5 mg/ml of K-76 COOH in GGBS-EGTA to an equal volume of serum and an incubation period of 18 h at 4 ° C. Reconstitution of the R5 by C5 samples could be performed either in a tube, or a hemolytic diffusion plate assay. Hemolytic plate assays were employed during the purification of C5. Bovine C5 closely resembled the reported behaviour of human C5 during PEG-precipitation, and the sequential ion-exchange column chromatography on DEAE-Sephacel and CM-Sephadex. The C5 active material eluted from the CM-Sephadex column also contained C3. This material was fractionated further on a column of hydroxylapatite (Fig. 1 ). Most of the C5 applied to the column was eluted by a 0.15 Mphosphate buffer (pH 7.6). The C3, and a small amount of C5, eluted later with the 0.3 M phosphate buffer. SDS-PAGE analysis of the reduced and non-reduced C5 containing fractions eluted from the hydroxylapatite column with the 0.15 M phosphate, showed an apparently pure macromolecule composed of two polypeptide chains. However, the pooled and concentrated fractions contained a relatively small amount of hemolytically active C3, and gave a very weak immunodiffusion reaction with anti-bovine C3. To remove the C3 contamination, the C5 preparation was treated with 300 m M CH3NH2 to convert the C3 contaminant to C3 (CH3NH2), which was subsequently bound by a column of immobilized factor H. The column eluate was highly purified C5 that was free from hemolytically active and immunochemically detectable C3. A typical yield of C5 from 800 ml of serum was between 5 to 10 mg protein.
306
W.P. ASTON E T AL.
2"5-
C3
A28o C5
1"5
0.5A
B
400
500 FRACTION
I 600
NUMBER
Fig. 1. Purification of C5 on hydroxylapatite. The C3/C5 material from CM-Sephadex was dialyzed against 5 mM phosphate buffer, pH 7.6 containing 5 mM EDTA and 100 mM NaC1. It was applied to an hydroxylapatite column (2.5 X 30 cm) equilibrated in the same phosphate buffer. The column was first eluted with the equilibrating buffer (fractions 1-270), and then developed sequentially with the phosphate buffer containing: (i) 2 M NaCl (fractions 271-360 ); (ii) 75 mM potassium phosphate (fractions 361-410); (iii) 150 mM potassium phosphate (start indicated by A in elution profile, fractions 411-480 ); (iv) 300 mM potassium phosphate (start indicated by B, fractions 481-580 ). The flow rate was 30 ml/h. Fractions ( 10 ml ) were measured for hemolytic C3 and C5 activities and for absorbance at 280 nm (A2s0).
The apparent molecular weights of bovine C5 and the alpha- and beta-chains were calculated after SD S-PAGE analysis of the non-reduced and reduced protein (Fig. 2). The non-reduced protein had an apparent molecular weight of 202 000 +_9000 (n = 5 ), and the molecular weights of the alpha- and beta-chains were 127 000 + 5000 (n = 5 ) and 74 000 + 2000 (n-- 5 ), respectively. Table 1 shows that bovine C5 and human C5 were able to reconstitute the homologous alternative complement pathways, activated by guinea pig erythrocytes. Neither species of C5 was able to give any significant reconstitution of the heterologous alternative pathway, under the same conditions. Bovine and human R3 reagents were able to reconstitute the hemolytic activity of heterologous R5 reagents, using guinea pig erythrocytes as target cells (results not shown). The results in Table 1 also establish that the isolated C5 protein
ISOLATIONOF THE FIFTH COMPONENTOF THE BOVINECOMPLEMENTSYSTEM
a
b
C5 iiiiiii~ :~:~i~i:
i~iiii~iiii!~!iiiiiiiiiiii,i
c <
origin
<
2ozo,oo
127
000
74000
>
>
>
307
d
....
-chain
~
~,,~
/3 - c h a i n
i iiiiiiii!i!ii i, ~~i!i!ii!iiiiiii!iii!!~!iii
ii~iii!~i~ i:iiiili~
Fig. 2. SDS-PAGE of purified bovine C5. Tracks a and b: non-reduced C5. Tracks c and d: reduced C5. The apparent molecular weights shown in the figure were calculated with reference to reduced marked proteins (not shown). TABLE 1 Comparison of activities present in preparations of bovine C3 and C5 and human C5 Activity
Bovine C51
Restoration of alternative pathway hemolytic activity 2 Bovine R5 > 95% Human R5 <5% Bovine R3 < 5% Inhibition by methylamine 3 0% Bovine factor H activity 4 0%
Bovine C31
< 5% <5% > 95% 100% 0%
Human C51
< 5% > 95% < 5% ND ND
ZThe concentration range of C3 and C5 was 12.5-125/lg, and the activity of the human C5 (obtained from Cordis Laboratories, Miami, FL) was 100 CHso units per ml. 2Expressed as a percentage of the target guinea pig erythrocytes lysed by the alternative complement pathway. 3Performed as described by Mhatre and Aston (1987b). 4Performed as described by Mhatre and Aston (1987a).
308
W . P . A S T O N E T AL.
was functionally distinct from bovine C3, because the C5 was unable to reconstitute bovine R3 and C3 was unable to reconstitute bovine R5. Furthermore, the hemolytic activity of the bovine C5 was not affected by a concentration of methylamine that completely inhibited the hemolytic activity of bovine C3. Also, in contrast to C3, the C5 activity was found to be stable to repeated freezing and thawing, although there was some breakdown of C5 during prolonged storage (e.g. > 6 months at - 7 0 ° C , see track a in Fig. 3). Factor H activity, a common contaminant in preparation of human C3 and C5, was apparently absent from the bovine C5 preparation (Table 1 ). Immobilized CVF.Bb, prepared from bovine factors B and D, and Naja naja atra cobra venom factor (CVF), cleaved the alpha-chain of bovine C3 (Menger and Aston, 1985 ) as shown in Fig. 3. Fig. 3 also shows that the CVF.Bb cleaved a
b <
1>
d
•
>
f ~ origin
>
J
~< 1 .je
3
origin
c
chain
>
~
~_~-~ _ ~ _ _ m * ~ .....
< o~-chain
> ~
-~ 2
>
-c 3
< ~-chain
-c 4
45
i ~/~i ~i!!!
Fig. 3. Sepharose-CVF.Bb cleavage of bovine C3 and C5 analysed by SDS-PAGE. Track a: C5; track b; C5 exposed to Sepharose-CVF; track c: C5 exposed to Sepharose-CVF.Bb; track d: C3 exposed to Sepharose-CVF.Bb; track e: C3 exposed to Sepharose-CVF; track f: C3. The positions of the marker proteins are indicated by arrows: 1,200 000; 2,116 250; 3, 92 500; 4, 66 200; 5, 45 000.
ISOLATION OF THE FIFTH COMPONENT OF T H E BOVINE COMPLEMENT SYSTEM
309
the alpha chain of C5 in the absence of C3b. The C5 alpha' chain produced had a molecular weight of about 10 000 less than the native alpha chain. The CVF.Bb appeared to cause more cleavage of C3 than of C5. Neither C5 nor C3 was cleaved by immobilized CVF alone. DISCUSSION K-76 monocarboxylic acid (K-76 COOH), an oxidized form of the sesquiterpene compound: 6,7-dihydroxy-2,5,5,8a-tetramethyl-l,2,3,4,4a,5,6,7,8,8adecahydronaphthalene- 1-spiro-2' - (7'-carboxyl-6'- formyl-4'-hydrozyl-2' ,3' dihydrobenzofuran), strongly inhibits the classical and alternative hemolytic pathways of human complement (Hong et al., 1979). The primary site of inhibitions is the generation of active C5b sites; K-76 COOH blocks, or destroys, the active structure of native C5, K-76 also binds with factor I and inhibits the degrading activity of factors I and H on C3b and C4b (Hong et al., 1981a; Pangburn and Muller-Eberhard, 1985). K-76 COOH was found to be an effective inhibitor of the bovine alternative complement pathway. It was also found that bovine sera deficient in components of the alternative pathway C3/C5 convertase (e.g. RB and R3 reagents) could reconstitute the hemolytic alternative complement pathway of the K-76 COOH inhibited bovine sera, thus indicating that the site of inhibition of the bovine serum was associated with the terminal complement pathway components, C5-C9. Subsequently, it was shown that purified C5 could reconstitute hemolytic activity to the inhibited serum samples and this reaction was used to provide an assay for the purification of C5 from bovine serum. Other bovine complement components affected by K-76 COOH were not identified. If bovine factor I resembles human factor I in it's sensitivity to K-76 COOH, this would presumably potentiate the assay for C5 by prolonging the half-life of the C3/C5 convertase enzyme, and also reduce the extent of the conglutination reaction (Linscott et al., 1978). Bovine C5 behaved in a manner similar to human C5 throughout the isolation procedure. The final purification step which removed traces of C3 from the C5 preparation was based upon structural differences between the two proteins; C5 does not have the internal thioester bond that is so important in the functional behaviour of C3 (Janatova et al., 1980; Law et al., 1980; Pangburn and Muller-Eberhard, 1985). The thioester bond in C3 can be cleaved in a variety of ways, including methylamine treatment and exposure to repeated freezing and thawing, yielding a modified protein, C3u, that is C3b-like in it's capacity to bind, with strong affinity, to factor H (Pangburn and Muller-Eberhard, 1985). Methylamine treatment of C5 samples containing C3 results in the convertion of C3 to C3u, in this particular case C3 (CH3NH2) ( Mhatre and Aston, 1987b). The C3u can then be removed from the sample by passing it through a column of Sepharose-factor H.
310
W.P. A S T O N E T AL.
The purified C5 was shown to be functionally and structurally distinct from bovine C3 in the following ways: (a) C3 activity was not inhibited by K-76 COOH when the protein was exposed to comparable amounts to the compound that inhibited the C5 activity. The mechanism of inhibition was not determined but it was found to be irreversible since dialysis of the treated protein did not reverse the inhibition. (b) C5 activity was not inhibited by methylamine treatments that inhibited the activity of C3. This is a consequence of the absence of a thioester bond in the alpha-chain of C5. (c) Another consequence of the absence of the thioester bond in C5 was the relative stability of the protein to repeated freezing and thawing. Prolonged storage of C5 did result in some degradation of the molecule. (d) Both C3 and C5 macromolecules are composed to two covalently-linked polypeptide chains. The molecular weights of these two chains differ slightly in both molecules; the molecular weights of the C5 alpha and beta chains are higher than those reported for bovine C3. Bovine C5, like human C5, shows species specificity. Purified bovine C5 was not effective in a human C5 deficient hemolytic system that involved the alternative pathway. In contrast, a C3 deficient bovine serum was able to interact with a human C5 deficient serum to induce the alternative pathways lysis of guinea pig erythrocytes, and similarly a bovine R5 could be reconstituted with a human R3. Since human C3 and bovine C3 are functionally interchangeable in the lysis of guinea pig red cells by either the human or bovine alternative pathways (Mhatre and Aston, 1987b), both human and bovine C3b molecules must be capable by interacting with either human or bovine C5. Therefore it is probable that the incompatibility between bovine C5 and the human R5 involves an incompatible reaction between bovine C5 and one or more of the human terminal components C6-C9. CVF isolated from the venom of the species Naja naja atra can generate a C3 convertase, CVF-Bb that will bind both human and guinea pig C3 and C5, and cause cleavage of their alpha-chains (Zabern et al., 1980). The cleavage of C5 occurs in the absence of C3 or C3b. Bovine C3 and C5 also act as substrates for the CVF.Bb, generated from CVF obtained from Naja naja atra, and yield C3b and C5b respectively. Such a cleavage of C5 could provide a means for studying the biological activities of bovine C5a and C5b. This investigation has extended the knowledge of the bovine complement system by providing methods for assaying and isolating bovine C5, the first terminal component of the bovine complement system to be obtained in a purified form. Functionally active components of the bovine terminal pathway have been isolated previously, but not in a form that permitted any structural characterization (Barta et al., 1976). Assays for bovine C5 have also been described (Linscott and Triglia, 1980), but these have involved the preparation of heterologous intermediates which are expensive and complicated to prepare. The assay presented in this communication is simple to perform and the detection reagent is simple to prepare.
ISOLATION OF THE FIFTH COMPONENT OF THE BOVINE COMPLEMENT SYSTEM
311
ACKNOWLEDGEMENT
The study was supported by a grant from the Natural Sciences and Engineering Research Council of Canada.
REFERENCES Barta, 0., Nelson, R.A., Jr. and Kuo, C.Y., 1976. Separation of six bovine complement components and one inactivator. Immunol. Commun., 5: 75-86. Fukuoka, Y., Seino, J., Okuda, T. and Tachibana, T., 1984. Purification of the fourth, second and fifth components of mouse complement. Immunology, 51: 493-501. Giclas, P.C., Keeling, P.J. and Henson, P., 1981. Isolation and characterization of the third and fifth components of rabbit complement. Mol. Immunol., 18: 113-123. Hong, K., Kinoshita, T., Miyazaki, W., Izawa, T. and Inoue, K., 1979. An anticomplementary agent, K-76 monocarboxylic acid: Its site and mechanism of inhibition of the complement activation cascade. J. Immunol., 122: 2418-2423. Hong, K., Kinoshita, T., Kitajima, H. and Inoue, K., 1981a. Inhibitory effect of K-76 monocarboxylic acid, an anticomplementary agent, on the C3b inactivator system. J. Immunol., 127: 104-108. Hong, K., Kinoshita, T. and Inoue, K., 1981b. Simple method for preparing EACl,4b,2a,3b and EAC4b,3b with human or guinea pig complement components using an anticomplementary agent K-76 monocarboxylic acid. J. Immunol., 127: 109-114. Janatova, J., Tack, B.F. and Prahl, J.W., 1980. Third component of human complement: Structural requirements for its function. Biochemistry, 19: 4479-4485. Kinoshita, T., Hong, K., Kondo, K. and Inoue, K., 1981. Fifth component of guinea pig complement: Purification and characterization. J. Immunol., 126: 2414-2418. Law, S.K., Lichtenberg, N.A., Holcombe, F.H. and Levine, R.P., 1980. Interaction between the labile binding sites of the fourth (C4) and fifth (C5) complement proteins and erythrocyte cell membranes. J. Immunol., 125: 634-639. Linscott, W.D. and Triglia, R.P., 1980. Methods for assaying nine bovine complement components and C3b-inactivator. Mol. Immunol., 17: 729-740. Linscott, W.D., Ranken, R. and Triglia, R.P., 1978. Evidence that bovine conglutinin reacts with an early product of C3b degradation, and an improved conglutinin assay. J. Immunol., 121: 658-664. Menger, M. and Aston, W.P., 1984. Factor D of the alternative pathway of bovine complement: Isolation and characterization. Vet. Immunol. Immunopathol., 7: 325-336. Menger, M. and Aston, W.P., 1985. Isolation and characterization of the third component of bovine complement. Vet. Immunol. Immunopathol., 10: 317-331. Mhatre, M. and Aston, W.P., 1987a. Isolation of bovine complement factor H. Vet. Immunol. Immunopathol., 14: 357-375. Mhatre, A. and Aston, W.P., 1987b. A simple hemolytic assay for bovine complement component C3. Vet. Immunol. Immunopathol., 15: 239-251. Muller-Eberhard, H.J., 1985. The membrane attack complex. In: H.J. Muller-Eberhard and P.A. Miescher (Editors), Complement. Springer-Verlag, New York, N.Y., pp. 227-275. Pang, A.S.D. and Aston, W.P., 1978. The alternative complement pathway in bovine serum: The isolation of a serum protein with factor B activity. Immunochemistry, 15: 529-534. Pangburn, M.K. and Muller-Eberhard, H.J., 1985. The alternative pathway of complement. In: H.J. Muller-Eberhard and P.A. Miescher (Editors), Complement. Springer-Verlag, New York, N.Y., pp. 185-214.
312
W.P.ASTONET AL.
Rapp, H.J. and Borsos, T., 1970. Reagents commonly used in complement research. In: Molecular basis of complement action. Appleton-Century-Crofts, New York, N.Y., pp. 75-109. Scott, J.D. and Fothergill, J.E., 1982. A general method for affinity purification of complement component C3b using factor H-Sepharose. Biochem. J., 205: 575-580. Sim, R.B., Twose, T.M., Paterson, D.S. and Sim, E., 1981. The covalent-binding reaction of complement C3. Biochem. J., 193: 115-127. Tack, B.F. and Prahl, J.W., 1976. Third component of human cc.mplement: Purification from plasma and physicochemical characterization. Biochemistry, 15: 4513-4521. Tack, B.F., Morris, S.G. and Prahl, J.W., 1979. Fifth component of human complement: Purification from plasma and polypeptide chain structure. Biochemistry, 18: 1490-1497. Tack, B.F., Janatova, J., Thomas, M.L. Harrison, R.A. and Hammer, C.H., 1981. The third, fourth, and fifth components of human complement: Isolation and biochemical properties. Methods Enzymol., 80: 64-101. Zabern, I.V., Hinsch, B., Przyklenk, H., Schmidt, G. and Vogt, W., 1980. Comparison ofNaja n. naja and Naja h. haje cobra-venom factors: Correlation between binding affinity for the fifth component of complement and mediation of its cleavage. Immunobiology, 157: 499-514.
301
Isolation of the Fifth Component of the B o v i n e Complement S y s t e m W.P. ASTON, A. MHATRE and J. MACRAE
Department of Microbiology and Immunology, Queen's University, Kingston, Ont. K7L 3N6 (Canada) (Accepted 17 October 1989 )
ABSTRACT Aston, W.P., Mhatre, A. and Macrae, J., 1990. Isolation of the fifth component of the bovine complement system. Vet. Immunol. Immunopathol., 24: 301-312. Bovine C5 has been isolated from fresh bovine serum by a five-step procedure: polyethylene glycol precipitation, sequential ion-exchange chromatography on DEAE-Sephacel and CM-Sephadex, hydroxylapatite chromatography, and affinity chromatography. The purified C5 was a protein of apparent molecular weight 202 000 ___9000 composed of two chains: an alpha-chain of molecular weight 127 000 _+5000 and a beta-chain of molecular weight 74 000 _+2000. The alpha-chain was cleaved by Sepharose-CVF.Bb (a cobra venom factor (CVF)-induced C3/C5 alternative pathway convertase) in the absence of any C3 or C3b. The monocarboxylic acid form of K-76, a sesquiterpene compound isolated from the culture filtrates of Stachybotris complementi, inhibited the alternative pathway of bovine serum, and the inhibited hemolytic activity was restored, in a dose dependent manner, by bovine C5. This provided the basis for a C5 functional assay throughout the purification procedure. The purified C5 showed species specificity and was functionally distinct from bovine C3.
INTRODUCTION
The fifth component of complement, C5, is a plasma glycoprotein that is comprised of two disulphide linked polypeptide chains, an alpha- and a betachain (Tack et al., 1979). Activation of C5 involves the cleavage of the alpha chain by either the classical or alternative pathway C5-convertases, and results in the generation of two important biologically active fragments, C5a and C5b (Muller-Eberhard, 1985). The human C5a fragment has anaphylatoxin, chemotactic and immunoregulatory activities, and the C5b fragment provides the nucleus for the assembly of the complement membrane attack complex (MAC) which is responsible for the lytic action of the complement system. C5 has been isolated from a variety of mammals, including humans {Tack et al., 0165-2427/90/$03.50
© 1990 Elsevier Science Publishers B.V.
302
W.P. ASTON ET AL.
1979 and 1981), guinea pigs (Kinoshita et al., 1981), mice (Fukuoka et al., 1984) and rabbits (Giclas et al., 1981 ). Human C5 has an apparent molecular weight of 190 000 (Tack et al., 1979) and it has structural homology with C3 and C4 (Tack et al., 1981). This paper reports the isolation of C5 from bovine serum and presents a novel assay that was used to identify C5 during purification. The assay is based on the ability of C5 to reconstitute the alternative pathway hemolytic activity in sera treated with the monocarboxylic acid derivative of K- 76 (K- 76 COOH ), a sesquiterpene anticomplementary compound, isolated from the culture filtrate of Stachybotrys complementi nov. sp. K-76 (Hong et al., 1979). MATERIALSAND METHODS Bovine serum, bovine serum deficient in C3 (R3), and human erythrocytes were obtained as previously described (Pang and Aston, 1978; Mhatre and Aston, 1987b ). Rabbit and guinea pig erythrocytes were purchased from Woodlyn Laboratories Ltd., Guelph, Ont. Bovine factors B and D, and C3 were prepared by the methods of Pang and Aston (1978) and Menger and Aston (1984, 1985), and human C5 was prepared by Cordis Laboratories, Miami, FL. All column chromatography media were purchased from Pharmacia Fine Chemicals, Uppsala, Sweden, except the Bio-Gel H T hydroxylapatite (BioRad, Richmond, CA). The monocarboxylic acid form of K-76 (K-76 COOH) was a gift from Dr. Kozo Inoue, Osaka University Medical School, Japan. All other chemicals were obtained from Sigma, St. Louis, MO. PBS-EGTA, was phosphate buffered saline (PBS) at physiological ionic strength and pH 7.2, containing 5 m M MgC12 and 15 m M ethylene glycol bis-amino tetraacetate (EGTA). PBS-EDTA was PBS containing 20 m M ethylene diamine tetraacetate (EDTA). GBS-EGTA was barbital buffered saline at physiological ionic strength and pH 7.2 (Rapp and Borsos, 1970 ) containing 0.1% gelatin, 15 m M EGTA and 4 m M MgC12. GGBS-EGTA was GBS-EGTA adjusted to an ionic strength of 0.065 with aqueous glucose (Rapp and Borsos, 1970).
Hemolytic assays C5. Bovine and human R5 reagents (sera depleted of C5) were prepared by mixing the appropriate serum with an equal volume of K-76 COOH (5 m g / m l ) in GGBS-EGTA and gently agitating the mixture for 18 h at 4 ° C. The mixture was extensively dialyzed, at 4 ° C, against PBS-EDTA and then stored in small aliquots at - 70 ° C. C5 samples (10 #1) were assayed for their ability to reconstitute the alternative pathway hemolytic function of an R5 reagent (210/tl). The bovine R5 reagent was 10 ttl of bovine R5, 10 ttl 8% (v/v) guinea pig erythrocytes, and 190/tl of the diluent, GGBS-EGTA. The h u m a n R5 reagent was 20 #l of human R5, 10 ttl of 8% (v/v) guinea pig erythrocytes, and 180/tl of GGBS-EGTA. The
ISOLATION OF THE FIFTH COMPONENT OF THE BOVINE COMPLEMENT SYSTEM
303
mixtures (R5 r e a g e n t + C 5 sample) were incubated at 37°C for 30 min, and then the reaction was stopped by the addition of 2.0 ml ice-cold GGBS-EDTA. The extent of target cell lysis was determined by measuring the absorption of the fluid phase at 412 nm, and expressed as a percentage of the complete lysis of the erythrocytes in the reaction mixture. The substitution of diluent for the C5 sample served as a negative control. All assays were performed in triplicate. The tube hemolytic assays for C5 were readily adapted to hemolytic diffusion plate assays, similar to those described for bovine factor D and C3 (Menger and Aston, 1984; Mhatre and Aston, 1987b). The hemolytic plates were prepared by pouring 6 ml of a molten mixture of 1% agarose in GGBS-EGTA, containing 0.75% guinea pig erythrocytes and 5-10% R5, onto a 5 × 7.5 cm polyester film (GelBond, Pharmacia Fine Chemicals, Uppsala, Sweden) and then allowing the mixture to solidify. Samples of C5 (7.5/A) introduced into wells (4 m m diameter) in the plate produced rings of lysis after several hours at 37 ° C. Buffer controls produced no lytic rings.
C3 and factor H. Bovine C3 activity was determined by the alternative pathway method of Mhatre and Aston (1987b), using guinea pig erythrocytes as the target activators. Bovine factor H activity was assessed by the capacity to inhibit lysis of human erythrocytes by a limited amount of bovine serum (Mhatre and Aston, 1987a). Isolation of bovine C5 The C5 protein was isolated from fresh bovine serum by polyethylene glycol (PEG) precipitation and a combination of chromatographic procedures on DEAE-Sephacel, CM-Sephadex C-50 and hydroxylapatite adapted from the procedures described by Tack and Prahl (1976), Tack et al. (1979) and Sim et al. (1981). A column of Sepharose-factor H was used as the final step, to remove any C3 contamination of the isolated C5 material. The Sepharosefactor H (2 mg of factor H per ml of Sepharose 4B ) was prepared as described by Scott and Fothergill ( 1982 ). Eight hundred ml of bovine serum at 4°C was mixed with 400 ml of 15% (w/v) PEG 4000 (polyethylene glycol) in 0.1 M phosphate buffer, pH 7.4 containing 0.15 M NaC1, 0.015 M EDTA and 0.03 M epsilon amino caproic acid (EACA). The resultant precipitate was removed by centrifugation. The supernatant, containing 5% PEG, was adjusted to 12% PEG by the slow addition of an appropriate volume of 33% PEG in the phosphate buffer. The resultant precipitate was collected by centrifugation, washed with 12% PEG and then dissolved in 5 m M phosphate buffer, pH 7.4 containing 20 m M NaC1, 5 m M EDTA and 10 m M EACA. The solution was applied to a DEAE-Sephacel colu m n (5 × 30 cm) equilibrated in the 5 m M phosphate buffer. The column was washed with the equilibrating buffer and then developed, at a flow rate of 60
304
W.P. ASTON ET AL,
ml/h, with a 3 1 salt gradient to 0.3 M NaC1. Column fractions (10 ml) containing C5 activity (fractions 80 to 105) were pooled, concentrated and dialysed against a 20 m M tris-acetate buffer, pH 6.8 containing 20 m M NaC1 and 5 m M EDTA. The material was applied to a 5 × 25 cm column of CM-Sephadex equilibrated with the 20 m M tris-acetate buffer. The column was washed with the equilibrating buffer and then developed at a flow rate of 40 m l / h with a 1.5 1 gradient to 0.3 M NaC1. Fractions (10 ml) containing C5 activity (fractions 145 to 175) were pooled and then dialyzed against 5 m M sodium phosphate buffer at pH 7.6, containing 5 m M EDTA and 100 m M NaC1. One hundred and fifty ml of the sample was applied to an hydroxylapatite column (2.5 × 30 cm) equilibrated in the 5 m M phosphate buffer. The column was washed with the equilibrating buffer and then developed (flow rate 30 m l / h ) sequentially: (a) with 700 ml of the 5 m M p h o s p h a t e buffer containing 2 M NaC1, (b) with 900 ml of the 5 m M phosphate buffer containing 0.075 M potassium phosphate, (c) with 700 ml of the 5 m M phosphate buffer containing 0.15 M potassium phosphate and finally, (d) with 1000 ml of the 5 m M phosphate buffer containing 0.3 M potassium phosphate. The C5 was eluted from the column with the 5 m M phosphate buffer containing 0.15 M potassium phosphate. Column fractions containing the peak of the C5 activity were pooled (100 ml) and dialyzed against PBS-EDTA at pH 8.0, and then treated with 300 m M methylamine (CH3NH2) for 18 h at 4°C to convert any contaminating C3 to C3 (CH~NH2) (Mhatre and Aston, 1987b). The sample was dialyzed against 25 m M tris-HC1 buffer at pH 7.6 and passed through a column ( 1 × 1 0 cm) containing the Sepharose-factor H to remove the C3 (CH3NH:). The purified C5 was concentrated by positive pressure ultrafiltration and stored at - 70 ° C. All column chromatography procedures were performed at 4°C, all buffers contained 0.04% sodium azide, and at each step C5 was identified by a hemolytic plate assay.
Polyacrylamide gel electrophoresis (PAGE) Slab gel electrophoresis (9% polyacrylamide) in the presence of sodium dodecyl sulphate (SDS) was performed as previously described (Menger and Aston, 1984).
Protein concentration determinations A bicinchonnic acid (BCA) micro protein assay system was performed according to the monograph "BCA Protein Assay Reagent" ( # 23225 from Pierce Chemical Co., Rockford, IL).
ISOLATION OF T H E FIFTH COMPONENT OF THE BOVINE COMPLEMENT SYSTEM
305
Cleavage of C3 and C5 by Sepharose-CVF.Bb Sepharose-CVF and Sepharose-CVF.Bb were prepared as previously described (Menger and Aston, 1985); the CVF (cobra venom factor), was isolated from the venom of Naja naja atra (Sigma, St. Louis, MO) and factor B was isolated from bovine serum. Aliquots (100 ttl) of a 50% (v/v) suspension of the immobilized CVF or CVF.Bb were washed with the PBS-EGTA. The packed Sepharose-CVF or Sepharose-CVF-Bb preparations were then resuspended in 30 #l of either bovine C3 (approx. 300 ttg/ml ) or bovine C5 (approx. 300 ttg/ml) in PBS-EGTA, and the reaction mixtures were maintained at 37 ° C for 45 min. The insoluble matrix was pelleted by centrifugation and 25 pl of the supernatant was analyzed by SDS-PAGE. RESULTS
The alternative pathway hemolytic activity of bovine serum was inhibited by K-76 COOH. The activity could be restored, in a dose-dependent manner by bovine serum depleted of factor B (50°C, 20 min) and C3 (methylamine treatment), and also by purified bovine C5; this provided the basis of an assay for bovine C5. The optimum treatment of serum to produce an effective R5 was the addition of 5 mg/ml of K-76 COOH in GGBS-EGTA to an equal volume of serum and an incubation period of 18 h at 4 ° C. Reconstitution of the R5 by C5 samples could be performed either in a tube, or a hemolytic diffusion plate assay. Hemolytic plate assays were employed during the purification of C5. Bovine C5 closely resembled the reported behaviour of human C5 during PEG-precipitation, and the sequential ion-exchange column chromatography on DEAE-Sephacel and CM-Sephadex. The C5 active material eluted from the CM-Sephadex column also contained C3. This material was fractionated further on a column of hydroxylapatite (Fig. 1 ). Most of the C5 applied to the column was eluted by a 0.15 Mphosphate buffer (pH 7.6). The C3, and a small amount of C5, eluted later with the 0.3 M phosphate buffer. SDS-PAGE analysis of the reduced and non-reduced C5 containing fractions eluted from the hydroxylapatite column with the 0.15 M phosphate, showed an apparently pure macromolecule composed of two polypeptide chains. However, the pooled and concentrated fractions contained a relatively small amount of hemolytically active C3, and gave a very weak immunodiffusion reaction with anti-bovine C3. To remove the C3 contamination, the C5 preparation was treated with 300 m M CH3NH2 to convert the C3 contaminant to C3 (CH3NH2), which was subsequently bound by a column of immobilized factor H. The column eluate was highly purified C5 that was free from hemolytically active and immunochemically detectable C3. A typical yield of C5 from 800 ml of serum was between 5 to 10 mg protein.
306
W.P. ASTON E T AL.
2"5-
C3
A28o C5
1"5
0.5A
B
400
500 FRACTION
I 600
NUMBER
Fig. 1. Purification of C5 on hydroxylapatite. The C3/C5 material from CM-Sephadex was dialyzed against 5 mM phosphate buffer, pH 7.6 containing 5 mM EDTA and 100 mM NaC1. It was applied to an hydroxylapatite column (2.5 X 30 cm) equilibrated in the same phosphate buffer. The column was first eluted with the equilibrating buffer (fractions 1-270), and then developed sequentially with the phosphate buffer containing: (i) 2 M NaCl (fractions 271-360 ); (ii) 75 mM potassium phosphate (fractions 361-410); (iii) 150 mM potassium phosphate (start indicated by A in elution profile, fractions 411-480 ); (iv) 300 mM potassium phosphate (start indicated by B, fractions 481-580 ). The flow rate was 30 ml/h. Fractions ( 10 ml ) were measured for hemolytic C3 and C5 activities and for absorbance at 280 nm (A2s0).
The apparent molecular weights of bovine C5 and the alpha- and beta-chains were calculated after SD S-PAGE analysis of the non-reduced and reduced protein (Fig. 2). The non-reduced protein had an apparent molecular weight of 202 000 +_9000 (n = 5 ), and the molecular weights of the alpha- and beta-chains were 127 000 + 5000 (n = 5 ) and 74 000 + 2000 (n-- 5 ), respectively. Table 1 shows that bovine C5 and human C5 were able to reconstitute the homologous alternative complement pathways, activated by guinea pig erythrocytes. Neither species of C5 was able to give any significant reconstitution of the heterologous alternative pathway, under the same conditions. Bovine and human R3 reagents were able to reconstitute the hemolytic activity of heterologous R5 reagents, using guinea pig erythrocytes as target cells (results not shown). The results in Table 1 also establish that the isolated C5 protein
ISOLATIONOF THE FIFTH COMPONENTOF THE BOVINECOMPLEMENTSYSTEM
a
b
C5 iiiiiii~ :~:~i~i:
i~iiii~iiii!~!iiiiiiiiiiii,i
c <
origin
<
2ozo,oo
127
000
74000
>
>
>
307
d
....
-chain
~
~,,~
/3 - c h a i n
i iiiiiiii!i!ii i, ~~i!i!ii!iiiiiii!iii!!~!iii
ii~iii!~i~ i:iiiili~
Fig. 2. SDS-PAGE of purified bovine C5. Tracks a and b: non-reduced C5. Tracks c and d: reduced C5. The apparent molecular weights shown in the figure were calculated with reference to reduced marked proteins (not shown). TABLE 1 Comparison of activities present in preparations of bovine C3 and C5 and human C5 Activity
Bovine C51
Restoration of alternative pathway hemolytic activity 2 Bovine R5 > 95% Human R5 <5% Bovine R3 < 5% Inhibition by methylamine 3 0% Bovine factor H activity 4 0%
Bovine C31
< 5% <5% > 95% 100% 0%
Human C51
< 5% > 95% < 5% ND ND
ZThe concentration range of C3 and C5 was 12.5-125/lg, and the activity of the human C5 (obtained from Cordis Laboratories, Miami, FL) was 100 CHso units per ml. 2Expressed as a percentage of the target guinea pig erythrocytes lysed by the alternative complement pathway. 3Performed as described by Mhatre and Aston (1987b). 4Performed as described by Mhatre and Aston (1987a).
308
W . P . A S T O N E T AL.
was functionally distinct from bovine C3, because the C5 was unable to reconstitute bovine R3 and C3 was unable to reconstitute bovine R5. Furthermore, the hemolytic activity of the bovine C5 was not affected by a concentration of methylamine that completely inhibited the hemolytic activity of bovine C3. Also, in contrast to C3, the C5 activity was found to be stable to repeated freezing and thawing, although there was some breakdown of C5 during prolonged storage (e.g. > 6 months at - 7 0 ° C , see track a in Fig. 3). Factor H activity, a common contaminant in preparation of human C3 and C5, was apparently absent from the bovine C5 preparation (Table 1 ). Immobilized CVF.Bb, prepared from bovine factors B and D, and Naja naja atra cobra venom factor (CVF), cleaved the alpha-chain of bovine C3 (Menger and Aston, 1985 ) as shown in Fig. 3. Fig. 3 also shows that the CVF.Bb cleaved a
b <
1>
d
•
>
f ~ origin
>
J
~< 1 .je
3
origin
c
chain
>
~
~_~-~ _ ~ _ _ m * ~ .....
< o~-chain
> ~
-~ 2
>
-c 3
< ~-chain
-c 4
45
i ~/~i ~i!!!
Fig. 3. Sepharose-CVF.Bb cleavage of bovine C3 and C5 analysed by SDS-PAGE. Track a: C5; track b; C5 exposed to Sepharose-CVF; track c: C5 exposed to Sepharose-CVF.Bb; track d: C3 exposed to Sepharose-CVF.Bb; track e: C3 exposed to Sepharose-CVF; track f: C3. The positions of the marker proteins are indicated by arrows: 1,200 000; 2,116 250; 3, 92 500; 4, 66 200; 5, 45 000.
ISOLATION OF THE FIFTH COMPONENT OF T H E BOVINE COMPLEMENT SYSTEM
309
the alpha chain of C5 in the absence of C3b. The C5 alpha' chain produced had a molecular weight of about 10 000 less than the native alpha chain. The CVF.Bb appeared to cause more cleavage of C3 than of C5. Neither C5 nor C3 was cleaved by immobilized CVF alone. DISCUSSION K-76 monocarboxylic acid (K-76 COOH), an oxidized form of the sesquiterpene compound: 6,7-dihydroxy-2,5,5,8a-tetramethyl-l,2,3,4,4a,5,6,7,8,8adecahydronaphthalene- 1-spiro-2' - (7'-carboxyl-6'- formyl-4'-hydrozyl-2' ,3' dihydrobenzofuran), strongly inhibits the classical and alternative hemolytic pathways of human complement (Hong et al., 1979). The primary site of inhibitions is the generation of active C5b sites; K-76 COOH blocks, or destroys, the active structure of native C5, K-76 also binds with factor I and inhibits the degrading activity of factors I and H on C3b and C4b (Hong et al., 1981a; Pangburn and Muller-Eberhard, 1985). K-76 COOH was found to be an effective inhibitor of the bovine alternative complement pathway. It was also found that bovine sera deficient in components of the alternative pathway C3/C5 convertase (e.g. RB and R3 reagents) could reconstitute the hemolytic alternative complement pathway of the K-76 COOH inhibited bovine sera, thus indicating that the site of inhibition of the bovine serum was associated with the terminal complement pathway components, C5-C9. Subsequently, it was shown that purified C5 could reconstitute hemolytic activity to the inhibited serum samples and this reaction was used to provide an assay for the purification of C5 from bovine serum. Other bovine complement components affected by K-76 COOH were not identified. If bovine factor I resembles human factor I in it's sensitivity to K-76 COOH, this would presumably potentiate the assay for C5 by prolonging the half-life of the C3/C5 convertase enzyme, and also reduce the extent of the conglutination reaction (Linscott et al., 1978). Bovine C5 behaved in a manner similar to human C5 throughout the isolation procedure. The final purification step which removed traces of C3 from the C5 preparation was based upon structural differences between the two proteins; C5 does not have the internal thioester bond that is so important in the functional behaviour of C3 (Janatova et al., 1980; Law et al., 1980; Pangburn and Muller-Eberhard, 1985). The thioester bond in C3 can be cleaved in a variety of ways, including methylamine treatment and exposure to repeated freezing and thawing, yielding a modified protein, C3u, that is C3b-like in it's capacity to bind, with strong affinity, to factor H (Pangburn and Muller-Eberhard, 1985). Methylamine treatment of C5 samples containing C3 results in the convertion of C3 to C3u, in this particular case C3 (CH3NH2) ( Mhatre and Aston, 1987b). The C3u can then be removed from the sample by passing it through a column of Sepharose-factor H.
310
W.P. A S T O N E T AL.
The purified C5 was shown to be functionally and structurally distinct from bovine C3 in the following ways: (a) C3 activity was not inhibited by K-76 COOH when the protein was exposed to comparable amounts to the compound that inhibited the C5 activity. The mechanism of inhibition was not determined but it was found to be irreversible since dialysis of the treated protein did not reverse the inhibition. (b) C5 activity was not inhibited by methylamine treatments that inhibited the activity of C3. This is a consequence of the absence of a thioester bond in the alpha-chain of C5. (c) Another consequence of the absence of the thioester bond in C5 was the relative stability of the protein to repeated freezing and thawing. Prolonged storage of C5 did result in some degradation of the molecule. (d) Both C3 and C5 macromolecules are composed to two covalently-linked polypeptide chains. The molecular weights of these two chains differ slightly in both molecules; the molecular weights of the C5 alpha and beta chains are higher than those reported for bovine C3. Bovine C5, like human C5, shows species specificity. Purified bovine C5 was not effective in a human C5 deficient hemolytic system that involved the alternative pathway. In contrast, a C3 deficient bovine serum was able to interact with a human C5 deficient serum to induce the alternative pathways lysis of guinea pig erythrocytes, and similarly a bovine R5 could be reconstituted with a human R3. Since human C3 and bovine C3 are functionally interchangeable in the lysis of guinea pig red cells by either the human or bovine alternative pathways (Mhatre and Aston, 1987b), both human and bovine C3b molecules must be capable by interacting with either human or bovine C5. Therefore it is probable that the incompatibility between bovine C5 and the human R5 involves an incompatible reaction between bovine C5 and one or more of the human terminal components C6-C9. CVF isolated from the venom of the species Naja naja atra can generate a C3 convertase, CVF-Bb that will bind both human and guinea pig C3 and C5, and cause cleavage of their alpha-chains (Zabern et al., 1980). The cleavage of C5 occurs in the absence of C3 or C3b. Bovine C3 and C5 also act as substrates for the CVF.Bb, generated from CVF obtained from Naja naja atra, and yield C3b and C5b respectively. Such a cleavage of C5 could provide a means for studying the biological activities of bovine C5a and C5b. This investigation has extended the knowledge of the bovine complement system by providing methods for assaying and isolating bovine C5, the first terminal component of the bovine complement system to be obtained in a purified form. Functionally active components of the bovine terminal pathway have been isolated previously, but not in a form that permitted any structural characterization (Barta et al., 1976). Assays for bovine C5 have also been described (Linscott and Triglia, 1980), but these have involved the preparation of heterologous intermediates which are expensive and complicated to prepare. The assay presented in this communication is simple to perform and the detection reagent is simple to prepare.
ISOLATION OF THE FIFTH COMPONENT OF THE BOVINE COMPLEMENT SYSTEM
311
ACKNOWLEDGEMENT
The study was supported by a grant from the Natural Sciences and Engineering Research Council of Canada.
REFERENCES Barta, 0., Nelson, R.A., Jr. and Kuo, C.Y., 1976. Separation of six bovine complement components and one inactivator. Immunol. Commun., 5: 75-86. Fukuoka, Y., Seino, J., Okuda, T. and Tachibana, T., 1984. Purification of the fourth, second and fifth components of mouse complement. Immunology, 51: 493-501. Giclas, P.C., Keeling, P.J. and Henson, P., 1981. Isolation and characterization of the third and fifth components of rabbit complement. Mol. Immunol., 18: 113-123. Hong, K., Kinoshita, T., Miyazaki, W., Izawa, T. and Inoue, K., 1979. An anticomplementary agent, K-76 monocarboxylic acid: Its site and mechanism of inhibition of the complement activation cascade. J. Immunol., 122: 2418-2423. Hong, K., Kinoshita, T., Kitajima, H. and Inoue, K., 1981a. Inhibitory effect of K-76 monocarboxylic acid, an anticomplementary agent, on the C3b inactivator system. J. Immunol., 127: 104-108. Hong, K., Kinoshita, T. and Inoue, K., 1981b. Simple method for preparing EACl,4b,2a,3b and EAC4b,3b with human or guinea pig complement components using an anticomplementary agent K-76 monocarboxylic acid. J. Immunol., 127: 109-114. Janatova, J., Tack, B.F. and Prahl, J.W., 1980. Third component of human complement: Structural requirements for its function. Biochemistry, 19: 4479-4485. Kinoshita, T., Hong, K., Kondo, K. and Inoue, K., 1981. Fifth component of guinea pig complement: Purification and characterization. J. Immunol., 126: 2414-2418. Law, S.K., Lichtenberg, N.A., Holcombe, F.H. and Levine, R.P., 1980. Interaction between the labile binding sites of the fourth (C4) and fifth (C5) complement proteins and erythrocyte cell membranes. J. Immunol., 125: 634-639. Linscott, W.D. and Triglia, R.P., 1980. Methods for assaying nine bovine complement components and C3b-inactivator. Mol. Immunol., 17: 729-740. Linscott, W.D., Ranken, R. and Triglia, R.P., 1978. Evidence that bovine conglutinin reacts with an early product of C3b degradation, and an improved conglutinin assay. J. Immunol., 121: 658-664. Menger, M. and Aston, W.P., 1984. Factor D of the alternative pathway of bovine complement: Isolation and characterization. Vet. Immunol. Immunopathol., 7: 325-336. Menger, M. and Aston, W.P., 1985. Isolation and characterization of the third component of bovine complement. Vet. Immunol. Immunopathol., 10: 317-331. Mhatre, M. and Aston, W.P., 1987a. Isolation of bovine complement factor H. Vet. Immunol. Immunopathol., 14: 357-375. Mhatre, A. and Aston, W.P., 1987b. A simple hemolytic assay for bovine complement component C3. Vet. Immunol. Immunopathol., 15: 239-251. Muller-Eberhard, H.J., 1985. The membrane attack complex. In: H.J. Muller-Eberhard and P.A. Miescher (Editors), Complement. Springer-Verlag, New York, N.Y., pp. 227-275. Pang, A.S.D. and Aston, W.P., 1978. The alternative complement pathway in bovine serum: The isolation of a serum protein with factor B activity. Immunochemistry, 15: 529-534. Pangburn, M.K. and Muller-Eberhard, H.J., 1985. The alternative pathway of complement. In: H.J. Muller-Eberhard and P.A. Miescher (Editors), Complement. Springer-Verlag, New York, N.Y., pp. 185-214.
312
W.P.ASTONET AL.
Rapp, H.J. and Borsos, T., 1970. Reagents commonly used in complement research. In: Molecular basis of complement action. Appleton-Century-Crofts, New York, N.Y., pp. 75-109. Scott, J.D. and Fothergill, J.E., 1982. A general method for affinity purification of complement component C3b using factor H-Sepharose. Biochem. J., 205: 575-580. Sim, R.B., Twose, T.M., Paterson, D.S. and Sim, E., 1981. The covalent-binding reaction of complement C3. Biochem. J., 193: 115-127. Tack, B.F. and Prahl, J.W., 1976. Third component of human cc.mplement: Purification from plasma and physicochemical characterization. Biochemistry, 15: 4513-4521. Tack, B.F., Morris, S.G. and Prahl, J.W., 1979. Fifth component of human complement: Purification from plasma and polypeptide chain structure. Biochemistry, 18: 1490-1497. Tack, B.F., Janatova, J., Thomas, M.L. Harrison, R.A. and Hammer, C.H., 1981. The third, fourth, and fifth components of human complement: Isolation and biochemical properties. Methods Enzymol., 80: 64-101. Zabern, I.V., Hinsch, B., Przyklenk, H., Schmidt, G. and Vogt, W., 1980. Comparison ofNaja n. naja and Naja h. haje cobra-venom factors: Correlation between binding affinity for the fifth component of complement and mediation of its cleavage. Immunobiology, 157: 499-514.