- Email: [email protected]
Simplified assays of hemolytic activity of the classical and alternative complement pathways
Journal oflmmunologicalMethods, 72 (1984) 49-59
49
Elsevier JIM03133
Simplified Assays of Hemolytic Activity of the Classical and Alternative Complement Pathways U.R. Nilsson and B. Nilsson The Blood Center, University Hospital, Uppsala, Sweden
(Received 14 December 1983, accepted 21 March 1984)
Simplified hemolytic assays for the classical (CP) and alternative (AP) pathways of complement (C) were developed. The CP function was tested with sensitized sheep erythrocytes in a diluent containing Ca 2+ and Mg 2+ , while AP was tested with unsensitized rabbit erythrocytes in a diluent containing Mg2+-EGTA. In contrast to the co{nmonly used hemolytic titration (CH50) assays, the present techniques tested the activity in reaction mixtures containing C at final dilutions which would not affect its function. These ranges fell between 1/1 and approximately 1/20 for CP and between 1/1 and approximately 1/3 for AP. With the adopted assay techniques single aliquots of serum were tested at single final dilutions of 1/8 for CP and 1/2 for AP, in the presence of excess target cells. Hemolysis was allowed to take place at 37°C for 20 min. The number of cells lysed by CP and AP under these conditions was directly proportional to the dose of serum and unaffected by the presence of a large excess of target cells. Each pathway was tested independently of the other. Serum C levels, measured as described, correlated strongly with those determined by standard hemolytic titration (CH50) assays. The modified assays should offer less laborious alternatives for the functional assay of C than current routine procedures. Key words: complement - hemolytic assay - classical pathway - alternative pathway
Introduction
Since the 1930's, hemolytic complement (C) activity has usually been estimated in terms of the 50% hemolytic unit of C, designated CH50 (Mayer, 1961). This unit defines the amount of C which is required for 50% hemolysis under certain standardized conditions, for example with respect to reaction volume, concentration of sheep erythrocytes (E), the quantity of antibody (A) required for sensitization of E, the nature of A, the concentrations of Ca2+ and Mg 2÷, the pH, the reaction time and the temperature. The C function of an unknown sample cannot be determined by assay of a single dose of the test specimen, but a series of dilutions of the serum has to be tested in order to secure points of measurement within the range of partial lysis. The dose-response curve is sigmoidal and ranges from 0 to 100% hemolysis. The curve is described by the equation of Van Krogh and can be transformed into 0022-1759/84/$03.00 © 1984 Elsevier Science Publishers B.V.
50 its logarithmic form which furnishes a linear function that can be used for a graphic evaluation of CH50. The described hemolytic assay, which has become the standard functional test of C, actually only yields information on the state of the classical pathway (CP). The functional state of the alternative pathway (AP) has been determined by a less commonly used test, which was described more recently (Platts Mills and Ishizaka, 1974). The potential value of these tests lies in their applicability for the screening of sera for C abnormalities, particularly when used in combination, since all the C proteins are involved in the mediation of hemolysis when both activation pathways are tested. However, the difficulty in performing these tests in their current modifications has largely prevented their application for screening purposes on a large scale. In attempts to minimize these difficulties we have considered 2 well-known characteristics of the hemolytic C reaction (Mayer, 1961): (1) that the degree of hemolysis on an absolute basis is constant even in the presence of target cells in large excess, and (2) that the hemolytic activity of a constant dose of serum varies inversely with the volume in which the reaction takes place. We have developed a procedure which tests the hemolytic function at a relatively high concentration of serum, where the dilution does not affect the function of C, and where a sufficiently large excess of target cells is present to prevent any serum dose tested from being able to cause more than partial hemolysis. C function can therefore be determined by assay of a single dose of serum and without need for repeated analysis in the event of very high or very low hemolytic activity, which is often necessary when the hemolytic activity of C is measured by the methods of Mayer (1961) or Platts-Mills and Ishizaka (1974). Materials and Methods
Diluents and chelating agents Veronal-buffered saline (VB 2-) was prepared as described by Mayer (1961). When required, Ca 2÷ and Mg 2÷ from a stock solution containing 1 M MgC12 and 0.3 M CaCI 2 were added to VB 2- to obtain final Ca z÷ and Mg 2+ concentrations of 0.15 and 0.5 mM respectively (VB2+). Gelatin was dissolved in these diluents (0.1% w / v ) as a 'stabilizing' agent (GVB 2- and GVB 2+ respectively). Sodium ethylenediamine tetraacetate (EDTA) was prepared as a stock solution (concentration 0.2 M pH 7.5). MgCI 2 and sodium ethyleneglycol tetraacetate (EGTA) were added to a mixed stock solution at concentrations of 0.05 and 0.2 M respectively. From this stock solution and GVB 2-, diluents containing Mg and EGTA at 0.002 and 0.008 M (GVB-Mg EGTA) or 0.004 and 0.016 M (GVB-Mg EGTA ( × 2)) respectively were prepared. VB z- and GVB 2- containing 0.01 M EDTA are referred to as VB-EDTA and GVB-EDTA respectively.
Antibody to sheep erythrocytes (SE) Rabbit antisera against SE stromata were prepared according to the method of
51 Mayer (1961). The anti-SE antisera were separated by gel chromatography on a Sephacryl $300 (Pharmacia, Uppsala) column, 2.5 x 100 cm, equilibrated with 0.5 M NaCI and buffered with 0.046 M NaPO 4 at pH 7.0. Eluted fractions were added to mixtures of unsensitized SE and fresh guinea pig serum diluted 1 / 2 0 in GVB 2÷ and incubated at 37°C. Hemolytic activity was demonstrated in 2 well separated peaks, which corresponded to the distributions of IgM and IgG. Fractions from the IgM area of the eluates were pooled, frozen, and stored in small aliquots at - 7 0 ° C and used as a source of hemolysin.
Preparation of target erythrocytes Sheep blood in Alsever's solution was obtained alternatingly from 3 sheep, which were used throughout this study. Sheep erythrocytes (SE) were washed, standardized and optimally sensitized as described by Mayer (1961). Purified rabbit anti-SE IgM (see above) served as the source of sensitizing antibody. Complexes between SE and rabbit anti-SE antibodies will be referred to as SEA. Suspensions of SEA at 2.5% ( v / v ) in GVB 2÷ were stored for a maximum of 1 week. Rabbit blood which was obtained alternatingly from 5 rabbits, which were used throughout this study, was let by puncturing the central artery of the ear with an injection needle (diameter 0.8 mm). The blood was allowed to flow freely to the 10 ml mark of a centrifuge tube, prefilled with 0.5 ml of 0.2 M EDTA. The cells were washed and standardized by the same procedures as for SE. Suspensions of rabbit erythrocytes (RE) at 2.5% (v/v) in GVB 2÷ were stored for a maximum of 1 week. Complexes between SEA and human C4 (SEAC4) were prepared by previously described procedures (Nilsson et al., 1974).
Serum, serum reagents and purified C proteins Human serum was obtained from apparently healthy laboratory personnel and blood donors at the Blood Center of the University Hospital, Uppsala. Serum specimens with hemolytic activity ranging from low to high levels were selected among sera submitted for routine analysis at the Section of Clinical Immunology. Most of these sera derived from patients with diseases predominantly leading to a response of the C system as an acute-phase reactant (e.g., rheumatoid arthritis, certain forms of chronic nephritis, systemic sclerosis) or from patients with diseases leading to a consumptive depletion of C (e.g., systemic lupus erythematosus, rheumatoid arthritis with vasculitis, post-infectious nephritis, Gram-negative sepsis). Guinea pig serum was obtained from normal animals or from animals with a congenital deficiency of C4 by cardiac puncture under ether anesthesia; the blood specimens were allowed to clot for approximately 1 h at room temperature and then centrifuged at 6000 x g at 4°C. The serum was removed, dispensed into several aliquots per specimen and immediately frozen and stored at - 7 0 ° C . Human serum was depleted of functionally active C3, C4 and C5 by exposure to 1 M KSCN as described previously (Nilsson et al., 1974). Human serum was depleted of factor D by repeated cycles of gel filtration on Sephadex G-75. EDTA was added to a volume of serum to a final concentration of 0.01 M and the p H was adjusted to 9.0. The serum aliquot was then passed through a Sephadex G-75. (Pharmacia Uppsala) column, 2.6 × 80 cm, equilibrated against 0.5 M NaCI, containin~ 0.01 M EDTA
52 and 0.1 M Tris buffer with a pH of 9.0. The filtration procedure was repeated 3 times, in between which the portion of the eluate corresponding to the main protein peak of V0 was concentrated by ultra filtration in an Amicon device (Amicon Mod. 3, 3 ml), using a YM 5 filter. The final serum preparation, the factor D-depleted serum, which was reconcentrated to its original protein concentration by the same procedure, was fully hemolytically active by CP but lacked such activity by AP. The latter function was completely restored, however, after recombination with fractions that followed the main protein peak, which was eluted at V0 in the first cycle of the separation procedure. Partially purified factor D was obtained from fractions of this eluate by selecting those fractions which were active in combination with the factor D-depleted serum, but which on their own showed no activity. C3 (Nilsson et al., 1975) and C5 (Nilsson et al., 1972) were isolated as described in previous publications.
Hemolytic assay procedures Titration of the hemolytic activity of C by CP was performed as described by Mayer (1961). For titration of the hemolytic function by AP, a modification of the procedure of Platts-Mills and Ishizaka (1974) was used. The latter technique allowed interaction between RE (0.1 ml of 1% suspension) and diluted test serum (0.2 ml) in proportionate amounts and otherwise under comparable conditions to those of the assay for CP function. GVB-Mg EGTA was used as diluent. The data obtained ( X = dose of serum; Y = fraction of hemolyzed cells), by both procedures, plotted as log X vs. log ( l ~ x y ) , yielded straight lines. The dose of serum which caused 50% lysis of the cells was extrapolated from the graph. Our new hemolytic assay procedures were arrived at by various studies, the conditions of which are described in Results. The standardized conditions which we now use routinely, however, are as follows: Hemolytic activity by CP. GVB 2÷ is used as diluent. SEA are adjusted to a 40% v / v suspension. 0.1 ml of test serum diluted 1 / 5 is carefully mixed with 0.1 ml of 40% suspension and vigorously shaken at 37°C. A blank tube containing SEA and 0.1 ml of GVB 2÷ and, if necessary, a serum color blank, 0.1 ml of serum dilution without cells, are incubated simultaneously. After 20 min the reaction is stopped by the addition of 3 ml of ice-chilled VB-EDTA. The optical density (OD) is measured spectrophotometrically at 541 nm and the appropriate blanks are subtracted. The fraction of lysed cells is calculated from the O1)541 of the test and from that of an aliquot of cells which have been completely lysed in a total volume of 3.2 ml of 0.1% Na2CO 3. The activity of the test serum is expressed in percent of that of a normal pool obtained by mixing equal aliquots from 40 sera obtained from normal blood donors and laboratory personnel. Hemolytic activity by AP. RE, preserved in GVB 2+ as a 5% (v/v) suspension, are washed once and resuspended in GVB-Mg EGTA ( × 2). 0.05 ml of undiluted test serum, carefully mixed with 0.1 ml of 50% RE suspension, are vigorously agitated at 37°C. Blanks and 100% lysis samples are
53
prepared analogously to those of the procedure for CP activity. After 20 min the reaction is stopped by the addition of 3 ml of VB-EDTA and the hemolysis is evaluated in the same way as described above. The activity of the test serum is compared with that of a reference serum. From this relationship the activity is calculated in percent of the mean as determined by individual measurements on 40 individual reference sera.
Results
The effect of dilution on the hemolytic activity of serum was tested by interacting fixed amounts of serum and target cells in the presence of varying amounts of diluent (Fig. 1). The CP activity of undiluted serum was unaffected by dilutions to approximately 1/20. The corresponding dilution limit for AP function was approximately 1/3. The hemolytic activity of fixed doses of human serum maintained at a constant dilution was tested in the presence of increasing doses of target cells (Fig. 2A and B, top). In the lower dose range all added cells were lysed, whereafter the number of lysed target cells remained constant irrespective of the number of cells added. The hemolytic activity of increasing doses of human serum was tested in the presence of constant, large amounts of target cells (10 9 SEA (Fig. 2A, bottom) or 109 RE (Fig. 2B, bottom)). In order to keep the serum dilution at a constant level, the volumes of the reaction mixtures were increased with each increment of serum. There was a linear dose response up to levels of 60-70% and 30-40% hemolysis of the available cells in the CP and AP assays respectively. The time course study of the CP-mediated and AP-mediated hemolytic reaction (Fig. 3A and B) yielded steep initial reaction curves which attained distinctly more
2O
6o
% ¢~
40
U)
2O
I
I
1/116
i
i
i
i
1/64 1/32 118 114 112 111 Dilution of serum
o.
Fig. 1. Hemolysis in the presence of a constant dose of serum at varying dilutions, zx : 2.5 × 109 S E A + 0 . 2 ml of human serum, dilution 1 / 5 + 0 , 0.2, 0.6--,12.6 ml GVB 2+. O : 1.2×109 RE + 0.05 ml of human serum undiluted, containing Mg and EGTA at 0.002 and 0.008 M concentration respectively+0, 0.05, 0.15 -vl.55 ml GVB-Mg EGTA. 37°C, 1 h.
B/y
54 3.2
3.2
O~ I 0 x
w
~/~. LYS,S
1.6
¢o
0.8
z
~.1.6
o.8
0.4
LYSIS
0.4
I.'6
0.40.8
312
0.4 0.8
4:0
1.6
= 3.2
No. of RE x l O - 9
NO. of SEA x l O - 9 I I0
1.0
0.5
0.5
.,0.3 ne
x x
LU ¢./)
°0.1
6
z
r
I
I
6
I
Z
0.250 0.500 0.750 ml of serum dilution 1/10
S
0.025 0.100 0.200 ml of serum dilution 111
Fig. 2. Hemolysis as a function of an increasing number of SEA in the presence of 0.5 ml of h u m a n serum, diluted 1 / 1 0 in GVB 2+ (top A), or of RE in the presence of 0.05 ml of h u m a n serum, undiluted, with Mg E G T A (top B). Hemolysis as a function of increasing volumes of serum at constant concentrations in the presence of 109 SEA (no chelating agent) (bottom A) or in the presence of 109 RE, and Mg (0.002 M) and EGTA (0.008 M) (bottom B).
A
~75
•~ 50 c o
o25 a.
0 I.U
L
i
3O
~20 ¢:
~ 10 Q.
o
1~
2'o
3'o
Time
(min)
4'0
Fig. 3. Kinetic hemolysis curve for CP (A): 2.5×109 S E A + 0 . 2 ml of h u m a n serum, diluted 1/20, in GVB2+; or for AP (B): 1.33 × 109 R E + 0 . 1 ml of h u m a n serum, diluted 1/2, in GVB-Mg EGTA.
55 horizontal courses after 10-20 and 5-10 min respectively. An incubation period of 20 min was therefore chosen for the routine performance of both tests. The concentration of chelating agent required to prevent activation of CP in the assay of AP function was determined by performing hemolytic tests on SEA and RE in the presence of varying concentrations of Mg and EGTA. Unchelated serum (0.05 ml) was mixed with 0.1 ml of 50% (v/v) cells suspended in GVB-Mg EGTA. With suspensions in 0.004 M Mg and 0.016 M EGTA full lysis of RE but no lysis of SEA was reproducibly observed. These concentrations, which yield final Mg and EGTA molarities in the fluid phase of the reaction mixture of 0.002 M and 0.008 M respectively, were chosen for routine purposes. To rule out the possibility that AP activation might have disturbed the test of CP function, the experiment illustrated in Fig. 4 was performed. Factor D-depleted serum was added to SEA and to RE in 2 separate series of test tubes. From 0 to 1/~1 of partially purified factor D was added per 25 #1 undiluted serum to each of the series, respectively. In the absence of factor D significant lysis of SEA occurred, while RE were almost completely unaffected. Upon addition of increasing amounts of factor D, full hemolysis of RE was observed, whereas the parallel addition of factor D to the CP test system did not affect the degree of hemolysis of SEA. Furthermore, C4-deficient guinea pig serum (0.2 ml, dilution 1/10 and 2.5 x 109 SEA, 37°C, 30 min) was completely devoid of hemolytic activity in the CP assay. Replacement of 5% of the C4-deficient guinea pig serum by normal human serum fully restored the activity. Two groups of patient sera were tested hemolytically by the procedures of Mayer (1961) and Platts-Mills and Ishizaka (1974) and the methods described in this paper. The sera of group A (n = 29) were selected on the basis of a wide CP activity range, as determined according to the method of Mayer, while the sera of group B (n = 30) were chosen correspondingly in regard to AP function, employing the procedure of
20
0
t5
0
0
~to n
0.5
1.0
Partially purilied factor D ( p l ) Fig. 4. Hemolytic purified factor D. GVB 2+ + 0 ~ 1 ~l D-depleted serum,
function of factor D-depleted human s e r u m + reconstituting amounts of partially O : 1 . 5 x 1 0 9 S E A + 0 . 1 ml of factor D-depleted serum, diluted 1/10, in of factor D per 25/~l of undiluted serum, zx : 6.1 × 10 g R E + 0.05 ml of factor d i l u t e d 1 / 2 , in GVB-Mg EGTA + 0 - , 1 ~l of factor D per 25 /~l of undiluted serum.
56
o J~
5O
O/
E c ¢0 ¢D =. J~
25
2
g
S
I
I
100 200 CP function according to Mayer (CH 50 units)
Fig. 5. Determinations of CP-mediated hemolytic activity by 2 methods: Mayer (1961) and the present method (1.3 × 109 SEA + 0.4 ml of human serum, diluted 1/10, in GVB 2+ ).
Platts-Mills and Ishizaka. Strong correlations were found between the results obtained by our methods and those obtained for CP function as described by Mayer (Fig. 5; Y = 0 . 2 1 X - 4 for 0 < X < 250; r = 0.91; P < 0.001) and for AP function as described by Platts-Mills and Ishizaka (Fig. 6; Y = 0.76X+ 23 for 0 _< X < 175; r = 0.72; P < 0.001). The relationship between the CP and AP functions of serum from the same individual has not yet been studied for pathological sera, but was
250 "0 0 J~
E
oo/
200
Q. >,0 ..Q C
o
O
/
°
~D
'- ==100 om
O0
c
~'~
e
50
0
0
0
0
0
O0
l
I
I
I
I
50
100
150
200
250
A P f u n c t i o n by P l a t t s - M i l l s ' modified method ( % of r e f e r e n c e s e r u m )
& Ishizaka's
Fig. 6. Determinations of AP-mediated hemolytic activity by 2 methods: modification of Platts-Mills and lshizaka (1974) and the present method, as described in Methods.
57 found to vary independently in 21 normal individuals (r = 0.03). CP- and AP-mediated activities were tested in sera from 40 normal individuals (blood donors and laboratory personnel) as follows: equal aliquots of serum were pooled. The pooled serum and each individual serum were tested for CP and AP function and the individual values were expressed as percent of the activity found in the serum pool. For CP, the mean value + SD was 101 + 26 and for AP 106 + 24. The finding of a mean value in excess of 100 for AP was consistent with the results for 2 other pools of serum and indicated a slight reduction of AP activity in the pooled material as compared with the mean of the individual sera making up the pool.
Discussion
C-mediated hemolysis is a function not only of the serum dose, but also of the dilution to which the serum is subjected by the test procedure. Methods were therefore developed for analyzing C function in reaction volumes which do not cause dilution of the test serum aliquots beyond the point which lowers the hemolytic activity (approx. 1 / 2 0 and 1 / 3 for CP and AP, respectively, Fig. 1). The actual, final dilutions of serum in the fluid phase of the reaction mixtures that we use routinely were 1 / 8 and 1 / 2 for CP and AP respectively. When increasing volumes of serum were added, at constant dilutions, within the ranges mentioned, the lytic response was linear from 0 to about 60 and 30% lysis for CP and AP respectively (Fig. 2, bottom). Therefore, in order to secure linearity of the dose response for the entire range of activity that may be anticipated during tests of unknown sera, the assay was routinely performed in the presence of a relatively large excess of cells. The possibility that this large number of cells could interfere with the assay in the event of low hemolytic activity of the test aliquot was ruled out by the finding, that the number of lysed cells remained unchanged even in a large excess of target cells (Fig. 2, top). Routinely, test mixtures for both CP and AP analysis were incubated for 20 min at 37°C. As illustrated in Figs. 3 and 4, the rate of hemolysis in both assays is relatively slow at this time and allows reproducible sampling, provided the reaction is promptly stopped by the addition of cold buffer, followed by centrifugation and decantation of the supernatants. For tests of AP function, unchelated test serum was added directly to the target cell suspension containing the chelating agent. In contrast to the experience of Platts-Mills and Ishizaka (1974), that chelation of the serum was necessary prior to mixing, we found that the described procedure efficiently inhibited CP-mediated lysis. The possibility that the methods described in this paper will come into routine use for simultaneous assessment of CP and AP as a functional test of C, raises the question of the accordance between the results of functional measurements obtained by our procedures and by those of Mayer (1961) and Platt-Mills and Ishizaka (1974). A comparison of results arrived at by these techniques and by ours showed a close
58
correlation between the data obtained by the different methods (Figs. 5 and 6). Rather than expressing C activity in terms of the dose of serum required for lysis of a standardized number of cells, analogously to the CH50 unit, we have chosen to express the hemolytic function in percent of a serum reference. The choice of this reference was based on studies of the hemolytic activity of individual specimens of serum and of pooled serum. When the CP function was expressed in percent of the activity of the serum pool made up from equal aliquots of the constituent sera, a mean value of 101 + 26% (1 SD, n = 40) was obtained, while the corresponding value for AP was 106 + 24%. The finding of a mean value in excess of 100% for the AP function was unanticipated, but was reproduced in similar studies of 2 additional pools and their corresponding individual sera. It would appear that the pooling of sera, for unknown reasons, leads to relative loss of AP function, while the CP function of the pool corresponds closely to the mean value of the individual sera. In view of these findings it was considered suitable to express the CP function of individual serum specimens in percent of the activity of a normal serum pool, determined on the same occasion. The AP function, on the other hand, was expressed in percent of the mean activity of the individual sera, set at 100%. Simultaneous hemolytic tests were performed on the test serum and on a reference serum with a known functional relationship to the mean relative function of reference serum relative function of the mean
=R.
The activity of the test serum in percent of the mean was arrived at by using the formula activitytest (%)
-
ODtest ODref ...... X
100 x R.
Currently, hemolytic titrations as described by Mayer (1961) and Platts-Mills and Ishizaka (1974) and radial hemolysis assays in agarose plates (Lint, 1982) are commonly employed techniques for assessment of C function in the diagnostic service laboratory. The advantage of the plate assays lies in their applicability in the screening of large numbers of sera, while their disadvantage is their insufficient quantitative accuracy. Abnormal results obtained by measuring C in the plate assays therefore need to be checked by a follow-up test by the more laborious titration procedures in order to assess more accurately the hemolytic levels quantitatively (Lint, 1982). We believe that the test techniques described in this paper, by being applicable both for screening and for quantitative analysis, combine the advantages of the plate and titration assays and should therefore be well suited for routine C measurements.
Acknowledgements The authors wish to acknowledge the excellent assistance of Mrs. Margita Nilsson, Mrs. Mariette Sjunneskog and Mr. Karl-Erik Svensson.
59 T h i s w o r k was s u p p o r t e d b y G r a n t s 5406 a n d 5647 f r o m the S w e d i s h M e d i c a l Research Council.
References Lint, T.F., 1982, Am. J. Med. Tech. 48, 743. Mayer, M.M., 1961, in: Experimental Immunochemistry, eds. E.A. Kabat and M.M. Mayer (Thomas, Springfield, IL) Ch. 4. Nilsson, U.R., R.H. Tomar and F.B. Taylor, 1972, Immunochemistry 9, 709. Nilsson, U.R., M.E. Miller and S. Wyman, 1974, J. Immunol. 112, 1164. Nilsson, U.R., R. Mandle and J. Mapes, 1975, J. Immunol. 114, 815. Platts-Mills, T.A.E. and K. lshizaka, 1974, J. Immunol. 113, 348.
49
Elsevier JIM03133
Simplified Assays of Hemolytic Activity of the Classical and Alternative Complement Pathways U.R. Nilsson and B. Nilsson The Blood Center, University Hospital, Uppsala, Sweden
(Received 14 December 1983, accepted 21 March 1984)
Simplified hemolytic assays for the classical (CP) and alternative (AP) pathways of complement (C) were developed. The CP function was tested with sensitized sheep erythrocytes in a diluent containing Ca 2+ and Mg 2+ , while AP was tested with unsensitized rabbit erythrocytes in a diluent containing Mg2+-EGTA. In contrast to the co{nmonly used hemolytic titration (CH50) assays, the present techniques tested the activity in reaction mixtures containing C at final dilutions which would not affect its function. These ranges fell between 1/1 and approximately 1/20 for CP and between 1/1 and approximately 1/3 for AP. With the adopted assay techniques single aliquots of serum were tested at single final dilutions of 1/8 for CP and 1/2 for AP, in the presence of excess target cells. Hemolysis was allowed to take place at 37°C for 20 min. The number of cells lysed by CP and AP under these conditions was directly proportional to the dose of serum and unaffected by the presence of a large excess of target cells. Each pathway was tested independently of the other. Serum C levels, measured as described, correlated strongly with those determined by standard hemolytic titration (CH50) assays. The modified assays should offer less laborious alternatives for the functional assay of C than current routine procedures. Key words: complement - hemolytic assay - classical pathway - alternative pathway
Introduction
Since the 1930's, hemolytic complement (C) activity has usually been estimated in terms of the 50% hemolytic unit of C, designated CH50 (Mayer, 1961). This unit defines the amount of C which is required for 50% hemolysis under certain standardized conditions, for example with respect to reaction volume, concentration of sheep erythrocytes (E), the quantity of antibody (A) required for sensitization of E, the nature of A, the concentrations of Ca2+ and Mg 2÷, the pH, the reaction time and the temperature. The C function of an unknown sample cannot be determined by assay of a single dose of the test specimen, but a series of dilutions of the serum has to be tested in order to secure points of measurement within the range of partial lysis. The dose-response curve is sigmoidal and ranges from 0 to 100% hemolysis. The curve is described by the equation of Van Krogh and can be transformed into 0022-1759/84/$03.00 © 1984 Elsevier Science Publishers B.V.
50 its logarithmic form which furnishes a linear function that can be used for a graphic evaluation of CH50. The described hemolytic assay, which has become the standard functional test of C, actually only yields information on the state of the classical pathway (CP). The functional state of the alternative pathway (AP) has been determined by a less commonly used test, which was described more recently (Platts Mills and Ishizaka, 1974). The potential value of these tests lies in their applicability for the screening of sera for C abnormalities, particularly when used in combination, since all the C proteins are involved in the mediation of hemolysis when both activation pathways are tested. However, the difficulty in performing these tests in their current modifications has largely prevented their application for screening purposes on a large scale. In attempts to minimize these difficulties we have considered 2 well-known characteristics of the hemolytic C reaction (Mayer, 1961): (1) that the degree of hemolysis on an absolute basis is constant even in the presence of target cells in large excess, and (2) that the hemolytic activity of a constant dose of serum varies inversely with the volume in which the reaction takes place. We have developed a procedure which tests the hemolytic function at a relatively high concentration of serum, where the dilution does not affect the function of C, and where a sufficiently large excess of target cells is present to prevent any serum dose tested from being able to cause more than partial hemolysis. C function can therefore be determined by assay of a single dose of serum and without need for repeated analysis in the event of very high or very low hemolytic activity, which is often necessary when the hemolytic activity of C is measured by the methods of Mayer (1961) or Platts-Mills and Ishizaka (1974). Materials and Methods
Diluents and chelating agents Veronal-buffered saline (VB 2-) was prepared as described by Mayer (1961). When required, Ca 2÷ and Mg 2÷ from a stock solution containing 1 M MgC12 and 0.3 M CaCI 2 were added to VB 2- to obtain final Ca z÷ and Mg 2+ concentrations of 0.15 and 0.5 mM respectively (VB2+). Gelatin was dissolved in these diluents (0.1% w / v ) as a 'stabilizing' agent (GVB 2- and GVB 2+ respectively). Sodium ethylenediamine tetraacetate (EDTA) was prepared as a stock solution (concentration 0.2 M pH 7.5). MgCI 2 and sodium ethyleneglycol tetraacetate (EGTA) were added to a mixed stock solution at concentrations of 0.05 and 0.2 M respectively. From this stock solution and GVB 2-, diluents containing Mg and EGTA at 0.002 and 0.008 M (GVB-Mg EGTA) or 0.004 and 0.016 M (GVB-Mg EGTA ( × 2)) respectively were prepared. VB z- and GVB 2- containing 0.01 M EDTA are referred to as VB-EDTA and GVB-EDTA respectively.
Antibody to sheep erythrocytes (SE) Rabbit antisera against SE stromata were prepared according to the method of
51 Mayer (1961). The anti-SE antisera were separated by gel chromatography on a Sephacryl $300 (Pharmacia, Uppsala) column, 2.5 x 100 cm, equilibrated with 0.5 M NaCI and buffered with 0.046 M NaPO 4 at pH 7.0. Eluted fractions were added to mixtures of unsensitized SE and fresh guinea pig serum diluted 1 / 2 0 in GVB 2÷ and incubated at 37°C. Hemolytic activity was demonstrated in 2 well separated peaks, which corresponded to the distributions of IgM and IgG. Fractions from the IgM area of the eluates were pooled, frozen, and stored in small aliquots at - 7 0 ° C and used as a source of hemolysin.
Preparation of target erythrocytes Sheep blood in Alsever's solution was obtained alternatingly from 3 sheep, which were used throughout this study. Sheep erythrocytes (SE) were washed, standardized and optimally sensitized as described by Mayer (1961). Purified rabbit anti-SE IgM (see above) served as the source of sensitizing antibody. Complexes between SE and rabbit anti-SE antibodies will be referred to as SEA. Suspensions of SEA at 2.5% ( v / v ) in GVB 2÷ were stored for a maximum of 1 week. Rabbit blood which was obtained alternatingly from 5 rabbits, which were used throughout this study, was let by puncturing the central artery of the ear with an injection needle (diameter 0.8 mm). The blood was allowed to flow freely to the 10 ml mark of a centrifuge tube, prefilled with 0.5 ml of 0.2 M EDTA. The cells were washed and standardized by the same procedures as for SE. Suspensions of rabbit erythrocytes (RE) at 2.5% (v/v) in GVB 2÷ were stored for a maximum of 1 week. Complexes between SEA and human C4 (SEAC4) were prepared by previously described procedures (Nilsson et al., 1974).
Serum, serum reagents and purified C proteins Human serum was obtained from apparently healthy laboratory personnel and blood donors at the Blood Center of the University Hospital, Uppsala. Serum specimens with hemolytic activity ranging from low to high levels were selected among sera submitted for routine analysis at the Section of Clinical Immunology. Most of these sera derived from patients with diseases predominantly leading to a response of the C system as an acute-phase reactant (e.g., rheumatoid arthritis, certain forms of chronic nephritis, systemic sclerosis) or from patients with diseases leading to a consumptive depletion of C (e.g., systemic lupus erythematosus, rheumatoid arthritis with vasculitis, post-infectious nephritis, Gram-negative sepsis). Guinea pig serum was obtained from normal animals or from animals with a congenital deficiency of C4 by cardiac puncture under ether anesthesia; the blood specimens were allowed to clot for approximately 1 h at room temperature and then centrifuged at 6000 x g at 4°C. The serum was removed, dispensed into several aliquots per specimen and immediately frozen and stored at - 7 0 ° C . Human serum was depleted of functionally active C3, C4 and C5 by exposure to 1 M KSCN as described previously (Nilsson et al., 1974). Human serum was depleted of factor D by repeated cycles of gel filtration on Sephadex G-75. EDTA was added to a volume of serum to a final concentration of 0.01 M and the p H was adjusted to 9.0. The serum aliquot was then passed through a Sephadex G-75. (Pharmacia Uppsala) column, 2.6 × 80 cm, equilibrated against 0.5 M NaCI, containin~ 0.01 M EDTA
52 and 0.1 M Tris buffer with a pH of 9.0. The filtration procedure was repeated 3 times, in between which the portion of the eluate corresponding to the main protein peak of V0 was concentrated by ultra filtration in an Amicon device (Amicon Mod. 3, 3 ml), using a YM 5 filter. The final serum preparation, the factor D-depleted serum, which was reconcentrated to its original protein concentration by the same procedure, was fully hemolytically active by CP but lacked such activity by AP. The latter function was completely restored, however, after recombination with fractions that followed the main protein peak, which was eluted at V0 in the first cycle of the separation procedure. Partially purified factor D was obtained from fractions of this eluate by selecting those fractions which were active in combination with the factor D-depleted serum, but which on their own showed no activity. C3 (Nilsson et al., 1975) and C5 (Nilsson et al., 1972) were isolated as described in previous publications.
Hemolytic assay procedures Titration of the hemolytic activity of C by CP was performed as described by Mayer (1961). For titration of the hemolytic function by AP, a modification of the procedure of Platts-Mills and Ishizaka (1974) was used. The latter technique allowed interaction between RE (0.1 ml of 1% suspension) and diluted test serum (0.2 ml) in proportionate amounts and otherwise under comparable conditions to those of the assay for CP function. GVB-Mg EGTA was used as diluent. The data obtained ( X = dose of serum; Y = fraction of hemolyzed cells), by both procedures, plotted as log X vs. log ( l ~ x y ) , yielded straight lines. The dose of serum which caused 50% lysis of the cells was extrapolated from the graph. Our new hemolytic assay procedures were arrived at by various studies, the conditions of which are described in Results. The standardized conditions which we now use routinely, however, are as follows: Hemolytic activity by CP. GVB 2÷ is used as diluent. SEA are adjusted to a 40% v / v suspension. 0.1 ml of test serum diluted 1 / 5 is carefully mixed with 0.1 ml of 40% suspension and vigorously shaken at 37°C. A blank tube containing SEA and 0.1 ml of GVB 2÷ and, if necessary, a serum color blank, 0.1 ml of serum dilution without cells, are incubated simultaneously. After 20 min the reaction is stopped by the addition of 3 ml of ice-chilled VB-EDTA. The optical density (OD) is measured spectrophotometrically at 541 nm and the appropriate blanks are subtracted. The fraction of lysed cells is calculated from the O1)541 of the test and from that of an aliquot of cells which have been completely lysed in a total volume of 3.2 ml of 0.1% Na2CO 3. The activity of the test serum is expressed in percent of that of a normal pool obtained by mixing equal aliquots from 40 sera obtained from normal blood donors and laboratory personnel. Hemolytic activity by AP. RE, preserved in GVB 2+ as a 5% (v/v) suspension, are washed once and resuspended in GVB-Mg EGTA ( × 2). 0.05 ml of undiluted test serum, carefully mixed with 0.1 ml of 50% RE suspension, are vigorously agitated at 37°C. Blanks and 100% lysis samples are
53
prepared analogously to those of the procedure for CP activity. After 20 min the reaction is stopped by the addition of 3 ml of VB-EDTA and the hemolysis is evaluated in the same way as described above. The activity of the test serum is compared with that of a reference serum. From this relationship the activity is calculated in percent of the mean as determined by individual measurements on 40 individual reference sera.
Results
The effect of dilution on the hemolytic activity of serum was tested by interacting fixed amounts of serum and target cells in the presence of varying amounts of diluent (Fig. 1). The CP activity of undiluted serum was unaffected by dilutions to approximately 1/20. The corresponding dilution limit for AP function was approximately 1/3. The hemolytic activity of fixed doses of human serum maintained at a constant dilution was tested in the presence of increasing doses of target cells (Fig. 2A and B, top). In the lower dose range all added cells were lysed, whereafter the number of lysed target cells remained constant irrespective of the number of cells added. The hemolytic activity of increasing doses of human serum was tested in the presence of constant, large amounts of target cells (10 9 SEA (Fig. 2A, bottom) or 109 RE (Fig. 2B, bottom)). In order to keep the serum dilution at a constant level, the volumes of the reaction mixtures were increased with each increment of serum. There was a linear dose response up to levels of 60-70% and 30-40% hemolysis of the available cells in the CP and AP assays respectively. The time course study of the CP-mediated and AP-mediated hemolytic reaction (Fig. 3A and B) yielded steep initial reaction curves which attained distinctly more
2O
6o
% ¢~
40
U)
2O
I
I
1/116
i
i
i
i
1/64 1/32 118 114 112 111 Dilution of serum
o.
Fig. 1. Hemolysis in the presence of a constant dose of serum at varying dilutions, zx : 2.5 × 109 S E A + 0 . 2 ml of human serum, dilution 1 / 5 + 0 , 0.2, 0.6--,12.6 ml GVB 2+. O : 1.2×109 RE + 0.05 ml of human serum undiluted, containing Mg and EGTA at 0.002 and 0.008 M concentration respectively+0, 0.05, 0.15 -vl.55 ml GVB-Mg EGTA. 37°C, 1 h.
B/y
54 3.2
3.2
O~ I 0 x
w
~/~. LYS,S
1.6
¢o
0.8
z
~.1.6
o.8
0.4
LYSIS
0.4
I.'6
0.40.8
312
0.4 0.8
4:0
1.6
= 3.2
No. of RE x l O - 9
NO. of SEA x l O - 9 I I0
1.0
0.5
0.5
.,0.3 ne
x x
LU ¢./)
°0.1
6
z
r
I
I
6
I
Z
0.250 0.500 0.750 ml of serum dilution 1/10
S
0.025 0.100 0.200 ml of serum dilution 111
Fig. 2. Hemolysis as a function of an increasing number of SEA in the presence of 0.5 ml of h u m a n serum, diluted 1 / 1 0 in GVB 2+ (top A), or of RE in the presence of 0.05 ml of h u m a n serum, undiluted, with Mg E G T A (top B). Hemolysis as a function of increasing volumes of serum at constant concentrations in the presence of 109 SEA (no chelating agent) (bottom A) or in the presence of 109 RE, and Mg (0.002 M) and EGTA (0.008 M) (bottom B).
A
~75
•~ 50 c o
o25 a.
0 I.U
L
i
3O
~20 ¢:
~ 10 Q.
o
1~
2'o
3'o
Time
(min)
4'0
Fig. 3. Kinetic hemolysis curve for CP (A): 2.5×109 S E A + 0 . 2 ml of h u m a n serum, diluted 1/20, in GVB2+; or for AP (B): 1.33 × 109 R E + 0 . 1 ml of h u m a n serum, diluted 1/2, in GVB-Mg EGTA.
55 horizontal courses after 10-20 and 5-10 min respectively. An incubation period of 20 min was therefore chosen for the routine performance of both tests. The concentration of chelating agent required to prevent activation of CP in the assay of AP function was determined by performing hemolytic tests on SEA and RE in the presence of varying concentrations of Mg and EGTA. Unchelated serum (0.05 ml) was mixed with 0.1 ml of 50% (v/v) cells suspended in GVB-Mg EGTA. With suspensions in 0.004 M Mg and 0.016 M EGTA full lysis of RE but no lysis of SEA was reproducibly observed. These concentrations, which yield final Mg and EGTA molarities in the fluid phase of the reaction mixture of 0.002 M and 0.008 M respectively, were chosen for routine purposes. To rule out the possibility that AP activation might have disturbed the test of CP function, the experiment illustrated in Fig. 4 was performed. Factor D-depleted serum was added to SEA and to RE in 2 separate series of test tubes. From 0 to 1/~1 of partially purified factor D was added per 25 #1 undiluted serum to each of the series, respectively. In the absence of factor D significant lysis of SEA occurred, while RE were almost completely unaffected. Upon addition of increasing amounts of factor D, full hemolysis of RE was observed, whereas the parallel addition of factor D to the CP test system did not affect the degree of hemolysis of SEA. Furthermore, C4-deficient guinea pig serum (0.2 ml, dilution 1/10 and 2.5 x 109 SEA, 37°C, 30 min) was completely devoid of hemolytic activity in the CP assay. Replacement of 5% of the C4-deficient guinea pig serum by normal human serum fully restored the activity. Two groups of patient sera were tested hemolytically by the procedures of Mayer (1961) and Platts-Mills and Ishizaka (1974) and the methods described in this paper. The sera of group A (n = 29) were selected on the basis of a wide CP activity range, as determined according to the method of Mayer, while the sera of group B (n = 30) were chosen correspondingly in regard to AP function, employing the procedure of
20
0
t5
0
0
~to n
0.5
1.0
Partially purilied factor D ( p l ) Fig. 4. Hemolytic purified factor D. GVB 2+ + 0 ~ 1 ~l D-depleted serum,
function of factor D-depleted human s e r u m + reconstituting amounts of partially O : 1 . 5 x 1 0 9 S E A + 0 . 1 ml of factor D-depleted serum, diluted 1/10, in of factor D per 25/~l of undiluted serum, zx : 6.1 × 10 g R E + 0.05 ml of factor d i l u t e d 1 / 2 , in GVB-Mg EGTA + 0 - , 1 ~l of factor D per 25 /~l of undiluted serum.
56
o J~
5O
O/
E c ¢0 ¢D =. J~
25
2
g
S
I
I
100 200 CP function according to Mayer (CH 50 units)
Fig. 5. Determinations of CP-mediated hemolytic activity by 2 methods: Mayer (1961) and the present method (1.3 × 109 SEA + 0.4 ml of human serum, diluted 1/10, in GVB 2+ ).
Platts-Mills and Ishizaka. Strong correlations were found between the results obtained by our methods and those obtained for CP function as described by Mayer (Fig. 5; Y = 0 . 2 1 X - 4 for 0 < X < 250; r = 0.91; P < 0.001) and for AP function as described by Platts-Mills and Ishizaka (Fig. 6; Y = 0.76X+ 23 for 0 _< X < 175; r = 0.72; P < 0.001). The relationship between the CP and AP functions of serum from the same individual has not yet been studied for pathological sera, but was
250 "0 0 J~
E
oo/
200
Q. >,0 ..Q C
o
O
/
°
~D
'- ==100 om
O0
c
~'~
e
50
0
0
0
0
0
O0
l
I
I
I
I
50
100
150
200
250
A P f u n c t i o n by P l a t t s - M i l l s ' modified method ( % of r e f e r e n c e s e r u m )
& Ishizaka's
Fig. 6. Determinations of AP-mediated hemolytic activity by 2 methods: modification of Platts-Mills and lshizaka (1974) and the present method, as described in Methods.
57 found to vary independently in 21 normal individuals (r = 0.03). CP- and AP-mediated activities were tested in sera from 40 normal individuals (blood donors and laboratory personnel) as follows: equal aliquots of serum were pooled. The pooled serum and each individual serum were tested for CP and AP function and the individual values were expressed as percent of the activity found in the serum pool. For CP, the mean value + SD was 101 + 26 and for AP 106 + 24. The finding of a mean value in excess of 100 for AP was consistent with the results for 2 other pools of serum and indicated a slight reduction of AP activity in the pooled material as compared with the mean of the individual sera making up the pool.
Discussion
C-mediated hemolysis is a function not only of the serum dose, but also of the dilution to which the serum is subjected by the test procedure. Methods were therefore developed for analyzing C function in reaction volumes which do not cause dilution of the test serum aliquots beyond the point which lowers the hemolytic activity (approx. 1 / 2 0 and 1 / 3 for CP and AP, respectively, Fig. 1). The actual, final dilutions of serum in the fluid phase of the reaction mixtures that we use routinely were 1 / 8 and 1 / 2 for CP and AP respectively. When increasing volumes of serum were added, at constant dilutions, within the ranges mentioned, the lytic response was linear from 0 to about 60 and 30% lysis for CP and AP respectively (Fig. 2, bottom). Therefore, in order to secure linearity of the dose response for the entire range of activity that may be anticipated during tests of unknown sera, the assay was routinely performed in the presence of a relatively large excess of cells. The possibility that this large number of cells could interfere with the assay in the event of low hemolytic activity of the test aliquot was ruled out by the finding, that the number of lysed cells remained unchanged even in a large excess of target cells (Fig. 2, top). Routinely, test mixtures for both CP and AP analysis were incubated for 20 min at 37°C. As illustrated in Figs. 3 and 4, the rate of hemolysis in both assays is relatively slow at this time and allows reproducible sampling, provided the reaction is promptly stopped by the addition of cold buffer, followed by centrifugation and decantation of the supernatants. For tests of AP function, unchelated test serum was added directly to the target cell suspension containing the chelating agent. In contrast to the experience of Platts-Mills and Ishizaka (1974), that chelation of the serum was necessary prior to mixing, we found that the described procedure efficiently inhibited CP-mediated lysis. The possibility that the methods described in this paper will come into routine use for simultaneous assessment of CP and AP as a functional test of C, raises the question of the accordance between the results of functional measurements obtained by our procedures and by those of Mayer (1961) and Platt-Mills and Ishizaka (1974). A comparison of results arrived at by these techniques and by ours showed a close
58
correlation between the data obtained by the different methods (Figs. 5 and 6). Rather than expressing C activity in terms of the dose of serum required for lysis of a standardized number of cells, analogously to the CH50 unit, we have chosen to express the hemolytic function in percent of a serum reference. The choice of this reference was based on studies of the hemolytic activity of individual specimens of serum and of pooled serum. When the CP function was expressed in percent of the activity of the serum pool made up from equal aliquots of the constituent sera, a mean value of 101 + 26% (1 SD, n = 40) was obtained, while the corresponding value for AP was 106 + 24%. The finding of a mean value in excess of 100% for the AP function was unanticipated, but was reproduced in similar studies of 2 additional pools and their corresponding individual sera. It would appear that the pooling of sera, for unknown reasons, leads to relative loss of AP function, while the CP function of the pool corresponds closely to the mean value of the individual sera. In view of these findings it was considered suitable to express the CP function of individual serum specimens in percent of the activity of a normal serum pool, determined on the same occasion. The AP function, on the other hand, was expressed in percent of the mean activity of the individual sera, set at 100%. Simultaneous hemolytic tests were performed on the test serum and on a reference serum with a known functional relationship to the mean relative function of reference serum relative function of the mean
=R.
The activity of the test serum in percent of the mean was arrived at by using the formula activitytest (%)
-
ODtest ODref ...... X
100 x R.
Currently, hemolytic titrations as described by Mayer (1961) and Platts-Mills and Ishizaka (1974) and radial hemolysis assays in agarose plates (Lint, 1982) are commonly employed techniques for assessment of C function in the diagnostic service laboratory. The advantage of the plate assays lies in their applicability in the screening of large numbers of sera, while their disadvantage is their insufficient quantitative accuracy. Abnormal results obtained by measuring C in the plate assays therefore need to be checked by a follow-up test by the more laborious titration procedures in order to assess more accurately the hemolytic levels quantitatively (Lint, 1982). We believe that the test techniques described in this paper, by being applicable both for screening and for quantitative analysis, combine the advantages of the plate and titration assays and should therefore be well suited for routine C measurements.
Acknowledgements The authors wish to acknowledge the excellent assistance of Mrs. Margita Nilsson, Mrs. Mariette Sjunneskog and Mr. Karl-Erik Svensson.
59 T h i s w o r k was s u p p o r t e d b y G r a n t s 5406 a n d 5647 f r o m the S w e d i s h M e d i c a l Research Council.
References Lint, T.F., 1982, Am. J. Med. Tech. 48, 743. Mayer, M.M., 1961, in: Experimental Immunochemistry, eds. E.A. Kabat and M.M. Mayer (Thomas, Springfield, IL) Ch. 4. Nilsson, U.R., R.H. Tomar and F.B. Taylor, 1972, Immunochemistry 9, 709. Nilsson, U.R., M.E. Miller and S. Wyman, 1974, J. Immunol. 112, 1164. Nilsson, U.R., R. Mandle and J. Mapes, 1975, J. Immunol. 114, 815. Platts-Mills, T.A.E. and K. lshizaka, 1974, J. Immunol. 113, 348.