CLINICAL
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
5,
258-263
Electra-Immunoassay A Comparison
with
(1976)
for Properdin
the Radioimmunoassay
M. A. SCHRAGER,
J. CHAPITIS,*
Division
Diseases, Department of Medicine, Universify of Medicine, Farmington. Connecficuf 56532
of Rheumatic School
Received Properdin immunoassay methods was reproducible immunoassay appear to be
N. F.
August
ROTHFIELD,
AND
1.
H.
LEPOW
oj’ C’onnecticur
15. 1975
concentrations in 22 human sera were determined by both electroand radioimmunoassay. A close correlation between the results of the two noted (r = 0.8945). Values obtained by electro-immunoassay were readily and did not change after 2 weeks of aging at 4°C. Since the electrois simpler to perform and requires less sophisticated equipment, it would the method of choice for serum determinations.
INTRODUCTION
Properdin is a normal serum protein first discovered by Pillemar et nl. in 1954 (1). Recent work in several laboratories has indicated the existence of a series of proteins, distinct from the early classical complement components (C 1, C4, C2), which is capable of activating the terminal components (C3, C5-C9) of the complement sequence (2). Properdin is one of the proteins in this sequence (the alternative pathway) and, in addition, may also participate in the feedback loop which recruits alternative pathway components following activation of C3 by the classical convertase, ClaZ (3). Serum properdin levels have been shown to be reduced in a number of disease entities including chronic membrano-proliferative glomerulonephritis, poststreptococcal glomerulonephritis, lupus nephritis (4-6) and active systemic lupus erythematosus (SLE) without clinical nephritis (Schrager and Rothfield, unpublished observations). Workers in most of these studies have employed single radial immunodiffusion for this measurement. The average reproducibility reported in one study was & 12.5% for a given serum (7). A sensitive radioimmunoassay (RIA) for properdin was recently reported by Minta ef al. (8). We have used a modified version as the standard for comparison. The electroimmunoassay, the “rocket” technique, is a means of determining protein concentrations and employs the principles of electrophoresis and immunoprecipitation in agarose. The method has been well summarized by Laurel1 (9). We have shown that electro-immunoassay is of comparable accuracy on clinical materials. while being simpler and requiring less sophisticated equipment and reagents than the RIA. ’ This work was supported by the United States Public Health Service. National Institutes of Health grants (No. AM 16576 and AI-08251). an Arthritis Foundation Clinical Research Center grant. and the University of Connecticut Research Foundation. * National Institutes of Health Postdoctoral Research Fellow, Grant No. I F27 AI 01 128-01. and Training Grant No. .5-TOl-AI-00438.
ELECTRO-IMMUNOASSAY
MATERIALS
FOR
PROPERDIN
259
AND METHODS
Serum samples. Blood was obtained from 48 normal volunteers and allowed to clot at room temperature. Aliquots of serum were stored at -70°C until 3-7 days prior to properdin determination and then aged at 4°C. One aliquot of each serum sample was used in the standard pool. Properdin concentrations in an additional 12 sera from patients with SLE as well as 10 sera from normal individuals were determined both by radioimmunoassay and electro-immunoassay. Additional studies were performed on sera which had been kept at 4°C for 2 weeks and with serial dilutions of individual sera. Purified properdin. Human properdin for use in the radioimmunoassay was isolated from pooled normal serum (commercial blood, Knickerbocker, New York, N.Y.) by the zymosan-adsorption and column chromatographic method described by Pensky et al. (10). The purified protein at a concentration of about 1 mg/ml in phosphate-buffered saline (PBS) was sterilely filtered (0.45~pm disposable filter unit, Millipore Corporation, Bedford, Mass.) and stored at 4°C. Monospeci$c anti-human properdin (anti-P). Anti-P used in both the electroimmunoassay and the radioimmunoassay was raised in rabbits and rendered monospecific by immunoadsorption with human IgG. Antisera were stored at -70°C. Electra-immunoassay.
Agarose, 1% (Bio-Rad), was dissolved in boiling Veronal buffer, 0.05 M, pH 8.6, with 0.001 M calcium lactate. After cooling to 60°C rabbit anti-human properdin, 0.1% was added and the solution poured between two glass plates separated by a plastic spacer to form a gel layer 9.5 x 18 x 0.15 cm. After the gel had set, the plates were separated, and a series of 3-mm wells were punched parallel to the longer margins of the plate. Eight-microliter aliquots of a 2:3 dilution in buffer of the test sera and of the pooled normal serum were applied in the wells using a lo-p1 Hamilton syringe. The plate was then placed in a water-cooled electrophoresis apparatus (MRA Corp., Boston, Mass.) and wicks fashioned of triple-thickness Whatman No. 3 filter paper were used to connect the plate with the buffer reservoirs. After covering with a heavy glass cover, a constant current of 50 mA was applied across the gel by using a Buchler Model 3-1014A power supply (Buchler Instruments, Fort Lee, N. Y.). Electrophoresis was continued for 18 hr during which the buffer was circulated continuously between the two reservoirs by means of a peristaltic pump and passive U-tube. This was done to maintain the composition of the buffer and to prevent abundant salt deposition at the anode. During the electrophoretic run, the applied voltage across the plate usually decreased from approximately 235 to 180 V because of a decrease in the electrical resistance of the gel. Plates were then washed overnight in saline, rinsed in distilled water, blotted dry, stained with Coomassie blue (0.25%) and then destained. In all cases, a well-defined cathodally migrating peak was found (Fig. 1). The height of each peak was measured from the top of the well and expressed as a percentage of the height of the pooled normal serum. Each serum was run on duplicate plates and the results averaged. Discordant results were repeated. Radiolabeling with iodine-125 was performed Radioiodination of properdin. according to the iodine monochloride method of Helmkamp et al. (11). The
260
FIG. aliquots
SCHRAGER
ET AL
1. Typical electro-immunoassay plate showing properdin “rockets” of the normal pool (peaks 3. 9 and 14). The cathode is at the top.
of 13 sera
and three
radiolabeled protein was buffered in PBS containing 0.5% bovine serum albumin, sterilely filtered, and stored at refrigerator temperatures. Measurement of radioactivity. Radioactivity was measured by placing 12 x 75-mm Polystyrene assay tubes into 16 x 125-mm plastic carrier tubes (AmershamKearle Corporation, Arlington Heights, III.) and counting for gamma radiation at optimal instrument settings in a scintillation well counter (Model 4230. Nuclear-Chicago Corporation, Des Plaines, Ill.). Solid phase radioimmunoassay of human properdin. This was performed by modifications of the published method (8). Polystyrene tubes (12 x 75 mm, Falcon Plastics, Oxnard, Calif.) were coated at room temperature for 2 hr with 1 ml of monospecific rabbit antiserum to human properdin. diluted 1: 1000 in 0.06 M sodium barbital buffer, pH 9.6. Unbound antibody was aspirated and the tubes were then filled immediately with 1 ml of mixtures containing dilutions of unlabeled serum antigen (unknown serum) and a constant amount of 1251-labeled properdin antigen. PBS fortified with 1% bovine serum albumin (Fraction V powder, Sigma Chemical Company, St. Louis, MO.) and 0.02% sodium azide constituted the diluent. A control serum pool, assumed to contain a normal serum concentration of properdin (P) (25 wg of P/ml), was included in each assay to serve as a reference with which to compare patient sera. By using 16 x loo-mm disposable glass tubes and starting with a stock dilution of 1:27.8 (0.1 ml of serum + 2.68 ml of diluent), 10 serial 1: 1.5 dilutions were made by successive transfer of 2 ml of the preceding dilution into 1 ml of diluent. One milliliter of radiolabeled properdin at twice the standard concentration was added to the dilutions. One milliliter of the resultant solutions contained lz51-labeled P at the standard concentration (lo-20 pg of P/ml, - 800- 1800 cpm/ml) and decreasing amounts of serum
ELECTRO-IMMUNOASSAY
FOR
PROPERDIN
261
(11.99-0.31 ~1). Equilibrium conditions between antibody irreversibly bound to the tube wall and antigen were established after incubation overnight (16-24 hr) in a temperature-controlled (37 ? l”C), humidified incubator (double-chamber incubator, Model 17060-002, National Appliance Co., Portland, Oreg.). Additional tubes included in each assay that held only the standard concentration of lz51labeled P were placed directly into plastic carrier tubes and counted for 10 min for gamma emission; the average total count (T/ was expressed in counts per minute (cpm). For all remaining tubes, including several that contained only 125-i-labeled P, the free (unbound) antigen was separated from the bound antigen by aspirating and discarding the liquid contents (free antigen) prior to counting the tubes; the counts per minute of tubes that had contained mixtures of labeled and unlabeled antigen were designated bound counts (B), and the average counts of the tubes that had contained only lz51-labeled P at the standard concentration were referred to as maximum bound counts (B,). One criterion for an acceptable assay was met if the ratio (BJT) x 100 fell between 68 and 100%. The B values corresponding to the 10 dilutions of one serum was first converted to Y values by the formula Y = B/B,,, and then Y values were converted to the logit values Y/( 1- Y). The 10 pairs of points (microliters of serum/dilution, Y/( 1 -Y) when plotted on two-cycle log-log graph paper yielded a set of points through which a straight line could be interpolated. The value microliters of serum/dilution (ordinate) of the point corresponding to Y/( 1-Y) = 1 (determined from the straight line) and referred to as 50% BIB, was converted to the concentration of properdin in the serum sample (pg of P/ml) by the calculation: [(50% BIB, of the control serum)/(50% B/B,, of the serum sample)] x (25 pg of P/ml). RESULTS
In the electro-immunoassay, a single, well-defined cathodally migrating peak (“rocket”) was observed with all of the test sera and also with the pooled sera. This peak was sharply tapered, indicating complete electrophoretic development of the plate. Peak heights varied from 10 to 40 mm and were more readily measured if projected with an overhead projector. Serial dilutions of the test sera produced a straight line when rocket height was plotted against concentration, also indicating complete development (9). Electrophoretic runs of 6-8 hr under otherwise identical conditions, on the other hand, frequently produced peaks with blunt apices and nonlinear dilution curves in sera with high concentrations of properdin. Even though properdin has been shown to exist in two electrophoretic forms (12), a second peak was never observed. Most sera did produce background staining in both anodal and cathodal directions, but this did not have the welldefined borders or characteristic shape of the specific precipitin peak. Duplicate runs on different days did not vary by more than + 5%. Aliquots of a single serum were frozen immediately at -70°C. Individual aliquots were thawed and kept at 4°C for 0, 4, 5, 6, 8, 11, 12 and 14 days prior to being assayed. The results of duplicate determinations of each aliquot are shown in Fig. 2. There was no significant change in peak height with aging. The mean (*SD) properdin concentration of this serum was 112 ? 2.7%. The properdin concentrations of the 48 sera comprising the normal pool ranged
262
SCHRAGER
ET AL
Fit.. 2. Electra-immunoassay plate showing duplicate samples of aliquots cated number of days in addition to three aliquots of the standard pool (3.
kept
at 4°C for the indi-
from 66 to 146% of the pool with a mean (2 1 SD) of 103 + 18%. Hence, the normal range (mean + 2 SD) was 66138%. The values for the normal pool conformed to a Gaussian distribution. Values obtained for each serum by RIA and electro-immunoassay. compared by linear regression analysis, are shown in Fig. 3. Pearson’s r for this comparison was 0.8945, indicating a close correspondence between the two methods. A still unexplained phenomenon in the electro-immunoassay was the apparent increase in height of the pooled serum peak when the pool was kept at 4°C for several days. Comparison of freshly thawed aliquots of the pool with those aged at 4°C for several days showed a net increase in height of nearly 25% during a 5-day period after which the height stabilized. Therefore. only such aged aliquots were used. This phenomenon did not occur with individual test sera. CONCLUSION
Electra-immunoassay is a simple and accurate technique which is applicable to the measurement of serum properdin levels. The method is reproducible and agrees well with the radioimmunoassay. The latter technique is obviously much more sensitive, but since the normal concentration of properdin in serum is estimated to be approximately 25 &ml, this sensitivity is not necessary for clinical purposes. In our hands, the results are more reproducible than those reported for radial immunodiffusion . The long period necessary for the completion of the electrophoretic run and the cathodal orientation of the peak are due to the fact that at pH 8.6 properdin exists with a slight positive charge in contrast to other plasma proteins which bear a net negative charge. It is possible that native properdin may be converted to the activated form under the assay conditions described. Because of the close agree-
ELECTRO-IMMUNOASSAY
FOR
263
PROPERDIN
601 IO
I I5 PROPERDIN
1 20 CONCENTRATION
I 25 BY RIA
,
(w~/ml)
3. Scattergram of properdin concentrations in 22 individual sera determined independently by electro-immunoassay (ordinate) and RIA (abscissa). There are 12 SLE sera (0) and 10 normal control sera (a). Regression coefficient (r) is 0.8945. FIG.
ment with the radioimmunoassay, however, such a phenomenon would not appear to affect the quantitative estimation of properdin protein concentration. ACKNOWLEDGMENTS We would like to thank Marilyn L. Cowles for typing the manuscript. We are grateful for the expert technical assistance of Bonnie J. Bordwell in performing the electro-immunoassays.
REFERENCES 1. Pillemer, L., Blum, L., Lepow, I. H., Ross, 0. A., Todd, E. W., and Wardlaw, A. C., Science 120, 279, 1954. 2. Gotze, O., and Mtiller-Eberhard. H. J., J. Exp. Med. 139, 44, 1974. 3. Fearon, D., and Austin, K. F., Fed. Proc. 34, 981, 1975. 4. Michael, A. F., Westberg, N. G., Fish, A. J., and Vernier, R. L.,J. Exp. Med. 134, 208s, 1971. 5. Penin, L. H., Lambert, P. H., and Miescher, P. A., Clin. Exp. Immunol. 16, 575, 1974. 6. McLean, R. H., and Michael, A. F., J. Clin. Invest. 52, 634, 1973. 7. Westberg, N. G., McLean, R. H., and Michael, A. F. Int. Arch. Allergy 44, 155, 1973. 8. Minta, J. O., Goodkofsky, I., and Lepow, I. H., Immunochemistry 10, 341, 1973. 9. Laurel], C. B., Stand. J. C/in. Lab. Invest. 29, Suppl. 124, 21, 1972. IO. Pensky, J., Hinz, C. F., Jr., Todd, E. W., Wedgewood, R. J., Boyer, J. T., and Lepow, I. H.,J. lmmunol. 100, 142, 1968. II. Helmkamp, R. W., Goodland, R. L., Bale, W. F., Spar, I. L., and Mutschler, L. E., Cancer Res. 20, 1495, 1960. 12. McLean, R. H., and Michael, A. F., Proc. Sot. Exp. Biol. Med. 141, 403, 1972.
IMMUNOLOGY
AND
IMMUNOPATHOLOGY
5,
258-263
Electra-Immunoassay A Comparison
with
(1976)
for Properdin
the Radioimmunoassay
M. A. SCHRAGER,
J. CHAPITIS,*
Division
Diseases, Department of Medicine, Universify of Medicine, Farmington. Connecficuf 56532
of Rheumatic School
Received Properdin immunoassay methods was reproducible immunoassay appear to be
N. F.
August
ROTHFIELD,
AND
1.
H.
LEPOW
oj’ C’onnecticur
15. 1975
concentrations in 22 human sera were determined by both electroand radioimmunoassay. A close correlation between the results of the two noted (r = 0.8945). Values obtained by electro-immunoassay were readily and did not change after 2 weeks of aging at 4°C. Since the electrois simpler to perform and requires less sophisticated equipment, it would the method of choice for serum determinations.
INTRODUCTION
Properdin is a normal serum protein first discovered by Pillemar et nl. in 1954 (1). Recent work in several laboratories has indicated the existence of a series of proteins, distinct from the early classical complement components (C 1, C4, C2), which is capable of activating the terminal components (C3, C5-C9) of the complement sequence (2). Properdin is one of the proteins in this sequence (the alternative pathway) and, in addition, may also participate in the feedback loop which recruits alternative pathway components following activation of C3 by the classical convertase, ClaZ (3). Serum properdin levels have been shown to be reduced in a number of disease entities including chronic membrano-proliferative glomerulonephritis, poststreptococcal glomerulonephritis, lupus nephritis (4-6) and active systemic lupus erythematosus (SLE) without clinical nephritis (Schrager and Rothfield, unpublished observations). Workers in most of these studies have employed single radial immunodiffusion for this measurement. The average reproducibility reported in one study was & 12.5% for a given serum (7). A sensitive radioimmunoassay (RIA) for properdin was recently reported by Minta ef al. (8). We have used a modified version as the standard for comparison. The electroimmunoassay, the “rocket” technique, is a means of determining protein concentrations and employs the principles of electrophoresis and immunoprecipitation in agarose. The method has been well summarized by Laurel1 (9). We have shown that electro-immunoassay is of comparable accuracy on clinical materials. while being simpler and requiring less sophisticated equipment and reagents than the RIA. ’ This work was supported by the United States Public Health Service. National Institutes of Health grants (No. AM 16576 and AI-08251). an Arthritis Foundation Clinical Research Center grant. and the University of Connecticut Research Foundation. * National Institutes of Health Postdoctoral Research Fellow, Grant No. I F27 AI 01 128-01. and Training Grant No. .5-TOl-AI-00438.
ELECTRO-IMMUNOASSAY
MATERIALS
FOR
PROPERDIN
259
AND METHODS
Serum samples. Blood was obtained from 48 normal volunteers and allowed to clot at room temperature. Aliquots of serum were stored at -70°C until 3-7 days prior to properdin determination and then aged at 4°C. One aliquot of each serum sample was used in the standard pool. Properdin concentrations in an additional 12 sera from patients with SLE as well as 10 sera from normal individuals were determined both by radioimmunoassay and electro-immunoassay. Additional studies were performed on sera which had been kept at 4°C for 2 weeks and with serial dilutions of individual sera. Purified properdin. Human properdin for use in the radioimmunoassay was isolated from pooled normal serum (commercial blood, Knickerbocker, New York, N.Y.) by the zymosan-adsorption and column chromatographic method described by Pensky et al. (10). The purified protein at a concentration of about 1 mg/ml in phosphate-buffered saline (PBS) was sterilely filtered (0.45~pm disposable filter unit, Millipore Corporation, Bedford, Mass.) and stored at 4°C. Monospeci$c anti-human properdin (anti-P). Anti-P used in both the electroimmunoassay and the radioimmunoassay was raised in rabbits and rendered monospecific by immunoadsorption with human IgG. Antisera were stored at -70°C. Electra-immunoassay.
Agarose, 1% (Bio-Rad), was dissolved in boiling Veronal buffer, 0.05 M, pH 8.6, with 0.001 M calcium lactate. After cooling to 60°C rabbit anti-human properdin, 0.1% was added and the solution poured between two glass plates separated by a plastic spacer to form a gel layer 9.5 x 18 x 0.15 cm. After the gel had set, the plates were separated, and a series of 3-mm wells were punched parallel to the longer margins of the plate. Eight-microliter aliquots of a 2:3 dilution in buffer of the test sera and of the pooled normal serum were applied in the wells using a lo-p1 Hamilton syringe. The plate was then placed in a water-cooled electrophoresis apparatus (MRA Corp., Boston, Mass.) and wicks fashioned of triple-thickness Whatman No. 3 filter paper were used to connect the plate with the buffer reservoirs. After covering with a heavy glass cover, a constant current of 50 mA was applied across the gel by using a Buchler Model 3-1014A power supply (Buchler Instruments, Fort Lee, N. Y.). Electrophoresis was continued for 18 hr during which the buffer was circulated continuously between the two reservoirs by means of a peristaltic pump and passive U-tube. This was done to maintain the composition of the buffer and to prevent abundant salt deposition at the anode. During the electrophoretic run, the applied voltage across the plate usually decreased from approximately 235 to 180 V because of a decrease in the electrical resistance of the gel. Plates were then washed overnight in saline, rinsed in distilled water, blotted dry, stained with Coomassie blue (0.25%) and then destained. In all cases, a well-defined cathodally migrating peak was found (Fig. 1). The height of each peak was measured from the top of the well and expressed as a percentage of the height of the pooled normal serum. Each serum was run on duplicate plates and the results averaged. Discordant results were repeated. Radiolabeling with iodine-125 was performed Radioiodination of properdin. according to the iodine monochloride method of Helmkamp et al. (11). The
260
FIG. aliquots
SCHRAGER
ET AL
1. Typical electro-immunoassay plate showing properdin “rockets” of the normal pool (peaks 3. 9 and 14). The cathode is at the top.
of 13 sera
and three
radiolabeled protein was buffered in PBS containing 0.5% bovine serum albumin, sterilely filtered, and stored at refrigerator temperatures. Measurement of radioactivity. Radioactivity was measured by placing 12 x 75-mm Polystyrene assay tubes into 16 x 125-mm plastic carrier tubes (AmershamKearle Corporation, Arlington Heights, III.) and counting for gamma radiation at optimal instrument settings in a scintillation well counter (Model 4230. Nuclear-Chicago Corporation, Des Plaines, Ill.). Solid phase radioimmunoassay of human properdin. This was performed by modifications of the published method (8). Polystyrene tubes (12 x 75 mm, Falcon Plastics, Oxnard, Calif.) were coated at room temperature for 2 hr with 1 ml of monospecific rabbit antiserum to human properdin. diluted 1: 1000 in 0.06 M sodium barbital buffer, pH 9.6. Unbound antibody was aspirated and the tubes were then filled immediately with 1 ml of mixtures containing dilutions of unlabeled serum antigen (unknown serum) and a constant amount of 1251-labeled properdin antigen. PBS fortified with 1% bovine serum albumin (Fraction V powder, Sigma Chemical Company, St. Louis, MO.) and 0.02% sodium azide constituted the diluent. A control serum pool, assumed to contain a normal serum concentration of properdin (P) (25 wg of P/ml), was included in each assay to serve as a reference with which to compare patient sera. By using 16 x loo-mm disposable glass tubes and starting with a stock dilution of 1:27.8 (0.1 ml of serum + 2.68 ml of diluent), 10 serial 1: 1.5 dilutions were made by successive transfer of 2 ml of the preceding dilution into 1 ml of diluent. One milliliter of radiolabeled properdin at twice the standard concentration was added to the dilutions. One milliliter of the resultant solutions contained lz51-labeled P at the standard concentration (lo-20 pg of P/ml, - 800- 1800 cpm/ml) and decreasing amounts of serum
ELECTRO-IMMUNOASSAY
FOR
PROPERDIN
261
(11.99-0.31 ~1). Equilibrium conditions between antibody irreversibly bound to the tube wall and antigen were established after incubation overnight (16-24 hr) in a temperature-controlled (37 ? l”C), humidified incubator (double-chamber incubator, Model 17060-002, National Appliance Co., Portland, Oreg.). Additional tubes included in each assay that held only the standard concentration of lz51labeled P were placed directly into plastic carrier tubes and counted for 10 min for gamma emission; the average total count (T/ was expressed in counts per minute (cpm). For all remaining tubes, including several that contained only 125-i-labeled P, the free (unbound) antigen was separated from the bound antigen by aspirating and discarding the liquid contents (free antigen) prior to counting the tubes; the counts per minute of tubes that had contained mixtures of labeled and unlabeled antigen were designated bound counts (B), and the average counts of the tubes that had contained only lz51-labeled P at the standard concentration were referred to as maximum bound counts (B,). One criterion for an acceptable assay was met if the ratio (BJT) x 100 fell between 68 and 100%. The B values corresponding to the 10 dilutions of one serum was first converted to Y values by the formula Y = B/B,,, and then Y values were converted to the logit values Y/( 1- Y). The 10 pairs of points (microliters of serum/dilution, Y/( 1 -Y) when plotted on two-cycle log-log graph paper yielded a set of points through which a straight line could be interpolated. The value microliters of serum/dilution (ordinate) of the point corresponding to Y/( 1-Y) = 1 (determined from the straight line) and referred to as 50% BIB, was converted to the concentration of properdin in the serum sample (pg of P/ml) by the calculation: [(50% BIB, of the control serum)/(50% B/B,, of the serum sample)] x (25 pg of P/ml). RESULTS
In the electro-immunoassay, a single, well-defined cathodally migrating peak (“rocket”) was observed with all of the test sera and also with the pooled sera. This peak was sharply tapered, indicating complete electrophoretic development of the plate. Peak heights varied from 10 to 40 mm and were more readily measured if projected with an overhead projector. Serial dilutions of the test sera produced a straight line when rocket height was plotted against concentration, also indicating complete development (9). Electrophoretic runs of 6-8 hr under otherwise identical conditions, on the other hand, frequently produced peaks with blunt apices and nonlinear dilution curves in sera with high concentrations of properdin. Even though properdin has been shown to exist in two electrophoretic forms (12), a second peak was never observed. Most sera did produce background staining in both anodal and cathodal directions, but this did not have the welldefined borders or characteristic shape of the specific precipitin peak. Duplicate runs on different days did not vary by more than + 5%. Aliquots of a single serum were frozen immediately at -70°C. Individual aliquots were thawed and kept at 4°C for 0, 4, 5, 6, 8, 11, 12 and 14 days prior to being assayed. The results of duplicate determinations of each aliquot are shown in Fig. 2. There was no significant change in peak height with aging. The mean (*SD) properdin concentration of this serum was 112 ? 2.7%. The properdin concentrations of the 48 sera comprising the normal pool ranged
262
SCHRAGER
ET AL
Fit.. 2. Electra-immunoassay plate showing duplicate samples of aliquots cated number of days in addition to three aliquots of the standard pool (3.
kept
at 4°C for the indi-
from 66 to 146% of the pool with a mean (2 1 SD) of 103 + 18%. Hence, the normal range (mean + 2 SD) was 66138%. The values for the normal pool conformed to a Gaussian distribution. Values obtained for each serum by RIA and electro-immunoassay. compared by linear regression analysis, are shown in Fig. 3. Pearson’s r for this comparison was 0.8945, indicating a close correspondence between the two methods. A still unexplained phenomenon in the electro-immunoassay was the apparent increase in height of the pooled serum peak when the pool was kept at 4°C for several days. Comparison of freshly thawed aliquots of the pool with those aged at 4°C for several days showed a net increase in height of nearly 25% during a 5-day period after which the height stabilized. Therefore. only such aged aliquots were used. This phenomenon did not occur with individual test sera. CONCLUSION
Electra-immunoassay is a simple and accurate technique which is applicable to the measurement of serum properdin levels. The method is reproducible and agrees well with the radioimmunoassay. The latter technique is obviously much more sensitive, but since the normal concentration of properdin in serum is estimated to be approximately 25 &ml, this sensitivity is not necessary for clinical purposes. In our hands, the results are more reproducible than those reported for radial immunodiffusion . The long period necessary for the completion of the electrophoretic run and the cathodal orientation of the peak are due to the fact that at pH 8.6 properdin exists with a slight positive charge in contrast to other plasma proteins which bear a net negative charge. It is possible that native properdin may be converted to the activated form under the assay conditions described. Because of the close agree-
ELECTRO-IMMUNOASSAY
FOR
263
PROPERDIN
601 IO
I I5 PROPERDIN
1 20 CONCENTRATION
I 25 BY RIA
,
(w~/ml)
3. Scattergram of properdin concentrations in 22 individual sera determined independently by electro-immunoassay (ordinate) and RIA (abscissa). There are 12 SLE sera (0) and 10 normal control sera (a). Regression coefficient (r) is 0.8945. FIG.
ment with the radioimmunoassay, however, such a phenomenon would not appear to affect the quantitative estimation of properdin protein concentration. ACKNOWLEDGMENTS We would like to thank Marilyn L. Cowles for typing the manuscript. We are grateful for the expert technical assistance of Bonnie J. Bordwell in performing the electro-immunoassays.
REFERENCES 1. Pillemer, L., Blum, L., Lepow, I. H., Ross, 0. A., Todd, E. W., and Wardlaw, A. C., Science 120, 279, 1954. 2. Gotze, O., and Mtiller-Eberhard. H. J., J. Exp. Med. 139, 44, 1974. 3. Fearon, D., and Austin, K. F., Fed. Proc. 34, 981, 1975. 4. Michael, A. F., Westberg, N. G., Fish, A. J., and Vernier, R. L.,J. Exp. Med. 134, 208s, 1971. 5. Penin, L. H., Lambert, P. H., and Miescher, P. A., Clin. Exp. Immunol. 16, 575, 1974. 6. McLean, R. H., and Michael, A. F., J. Clin. Invest. 52, 634, 1973. 7. Westberg, N. G., McLean, R. H., and Michael, A. F. Int. Arch. Allergy 44, 155, 1973. 8. Minta, J. O., Goodkofsky, I., and Lepow, I. H., Immunochemistry 10, 341, 1973. 9. Laurel], C. B., Stand. J. C/in. Lab. Invest. 29, Suppl. 124, 21, 1972. IO. Pensky, J., Hinz, C. F., Jr., Todd, E. W., Wedgewood, R. J., Boyer, J. T., and Lepow, I. H.,J. lmmunol. 100, 142, 1968. II. Helmkamp, R. W., Goodland, R. L., Bale, W. F., Spar, I. L., and Mutschler, L. E., Cancer Res. 20, 1495, 1960. 12. McLean, R. H., and Michael, A. F., Proc. Sot. Exp. Biol. Med. 141, 403, 1972.
