JOURNAL OF CLINICAL MICROBIOLOGY, Oct. 1979, p. 529-532 0095-1137/79/10-0529/04$02.00/0
Vol. 10, No. 4
Enzyme-Linked Protein A: an Enzyme-Linked Immunosorbent Assay Reagent for Detection of Human Immunoglobulin G and Virus-Specific Antibody H. PAUL MADORE' *
AND
A. BAUMGARTEN2
Department of Laboratory Medicine, Yale University, School of Medicine, New Haven, Connecticut 06520,2 and the Virology Laboratory, Veterans Administration Medical Center West Haven, Connecticut 065161 Received for publication 10 July 1979
A general-purpose reagent capable of reacting with immunoglobulin G in a modified enzyme-linked immunosorbent assay technique was prepared by using protein A coupled with horseradish peroxidase. The reagent detected low levels (0.003 to 1.0 ,ug/ml) of human immunoglobulin G and was also applied in an enzyme-linked immunosorbent assay for titration of antibody to human cytomegalovirus. The antibody titers to human cytomegalovirus determined by enzymelinked immunosorbent assay and by complement fixation were compared. The correlation coefficient between the two techniques was 0.85, but the enzymelinked immunosorbent assay was 10 times more sensitive than complement fixation in terms of antibody titers detected.
The enzyme-linked immunosorbent assay (ELISA) is widely applied in imnnunodiagnosis (6, 21, 22). Enzyme-linked antibody is the reagent employed to detect antibody-antigen reactions in the assay. Preparation of enzymelinked antibody involves antibody (immunoglobulin G [IgG]) purification and is associated with a marked loss of antibody as well as of enzyme activity during coupling (2, 19). As an altemative to enzyme-linked antibody in ELISA, enzyme-linked protein A was employed to detect antibody against mouse alpha fetoprotein (4). Protein A is a 40,000 molecular weight protein produced by certain strains of Staphylococcus aureus (7, 10, 20). The protein binds the Fc portion of immunoglobulins, chiefly IgG, of several mammalian species, including humans
(8,9).
This report describes the use of horseradish peroxidase-conjugated protein A as a reagent in ELISA for the quantification of human IgG and for the titration of antibody against human cytomegalovirus (HCMV). MATERIALS AND METHODS Horseradish peroxidase protein A conjugation. A modified version of the two-step glutaraldehyde procedure of Avrameas and Ternynck (1) was employed. A 5-mg amount of horseradish peroxidase (type VI; Sigma Chemical Co.) in 0.2 ml of 0.1 M phosphate buffer (pH 6.8), containing 1.25% glutaraldehyde (Sigma, grade I) was allowed to stand at room temperature for 18 h, dialyzed at 4°C against two 200ml changes of 0.05 M carbonate-bicarbonate buffer (pH 9.5), and then brought to 1 ml in the same buffer.
Protein A (Sigma) was dissolved in 0.15 M NaCl-0.1 M carbonate-bicarbonate buffer (pH 9.5) at 5 mg/ml. Activated peroxidase alone (500 ,tg) or the following mixtures of activated peroxidase and protein A, in micrograms, (1,000, 250, 500, 250, 250, and 250) were conjugated for 24 h at 4°C. Excess lysine (10 y1 of a 0.2 M solution) was added to the conjugates, the mixtures were kept for 2 h at 4°C, and then the preparations were dialyzed against three 100-ml changes of phosphate-buffered saline (pH 7.3) at 4°C. The preparations were centrifuged at 20,000 x g for 20 min at 4°C, and the supernatant fluids were removed, dispensed in working volumes, and stored at -20°C until used. Measurement of horseradish peroxidase. Enzyme activity was measured by the method of Mathiesen et al. (16) by adding 10 pg of enzyme in 100 p1 of 0.1 M citrate buffer (pH 5.0), to.l ml of fresh reaction mixture (50 mg of o-phenylenediamine per ml, 0.006% peroxide in 0.1 M sodium citrate buffer, pH 5.0), incubating at room temperature in the dark, and stopping the reaction with 0.55 ml of 2 M sulfuric acid at 5-, 10-, 15-, 30-, 45-, or 60-min intervals. Reaction mixture alone with 2 M sulfuric acid added at 0 min served as a blank. The reaction was linear for 45 min, so the optical density at 493 nmol (OD493) after 30 min was used as a measure of enzyme activity. ELISA procedures. Binding to human IgG was assayed in microtiter plates (polyvinyl tissue culture; Linbro) which were coated with either 100 t1 of human IgG (1 pg/ml; Hyland Laboratories, Inc., Costa Mesa, Calif.) per well, diluted in phosphate-buffered saline (PBS; pH 7.3), or 100 pl of PBS alone per well for 2 h at room temperature or for 24 h at 4°C. The wells were washed three times with 0.05% Tween 20-PBS (TP) and filled with 1% bovine serum albumin (BSA; fraction V) in PBS, and the plates were incubated for 1 h at room temperature. The wells were again washed three times with TP, appropriate dilutions (in 1% 529
530
J. CLIN. MICROBIOL.
MADORE AND BAUMGARTEN
BSA-TP) of the conjugates (100 LI per well) were added, and the plates were incubated for 2 h at room temperature. The wells were washed three times with TP (at 3-min intervals), 250 IL of fresh reaction mixture per well was added, and the enzyme was assayed as described. Measurement was made by using pooled solutions from duplicate wells. Solutions from wells without conjugate were used as blanks. Determination of the inhibition curve of IgG was made by mixing predetermined amounts of conjugate (2:1) in a test tube with increasing amounts of freshly diluted IgG, incubating the mixture for 30 min at 37°C, and then assaying duplicate samples (100 Ld per well) for conjugate binding to IgG on microtiter plates as described above.
Cytomegalovirus antibody was titrated, using HCMV antigen (Flow Laboratories, Inc., Rockville, Md.) diluted 1:200 in PBS. The antigen was adsorbed (100 u1 per well) to microtiter plates for 2 h at room temperature, the supernatant fluids were removed, the wells were filled with 4% BSA-PBS, and the plates were incubated overnight in a moist box at 4°C. The plates were washed three times with TP, duplicate samples (100 1Ld per well) of twofold dilutions of serum in 4% BSA-TP were added, and the covered plates were incubated for 30 min at 35°C. The plates were then washed three times with TP (at 3-min intervals), duplicate samples (100 IlI per well) of conjugate (4:1) diluted 1:500 in 4% BSA-TP were added, and the covered plates were again incubated for 30 min at 35°C. The plates were washed with TP (at 3-min intervals), and enzyme activity was measured as described above. The complement fixation test was performed by standard procedures (14). Human sera. Sera used in this study were obtained from specimens submitted for routine determination of complement-fixing antibody to HCMV.
TABLE 1. Analysis of enzyme-protein A conjugation Sample
A representative standard curve for human IgG (Fig. 1) had a range for detecting 0.003 to 1.0 ,g of IgG per ml. This sensitivity was similar to quantitative ELISA for IgG using enzyme-linked antibody against IgG (5, 6). An estimate of the
PeroxiPeroxi- dase dase, bound (pg/ml)b to IgG
MY Peroxidase (untreated) 100 448 Peroxidase (treated)d 12 373 0 32 863 Conjugate 4:1e 8 2:1 11 430 26 1:1 18 142 19 'The enzyme assay was performed as described in the text. An activity of 100% is equivalent to 78.1 OD491/mg of peroxidase. b OD275 was measured and corrected for estimated amount of protein A in preparation (250 jig of protein A per ml = 0.044 OD275). 'The percent peroxidase (conjugated to protein A) bound to IgG was obtained by converting the OD493 values of peroxidase activity from the microtiter plate assays to micrograms of peroxidase per milliliter bound to IgG, using the peroxidase specific activities listed in the second column, and then dividing the micrograms of peroxidase bound to IgG per milliliter by the total peroxidase in the preparation (third column) and multiplying by 100. d Treated with glutaraldehyde. e Initially 1,000, 500, or 250 jig of peroxidase was mixed with 250 jg of protein A. .0
.8
RESULTS
Horseradish peroxidase protein A conjugation. The specific activity of peroxidase decreased 7- to 10-fold after glutaraldehyde treatment and coupling to protein A (Table 1). Subsequently, peroxidase activity of the conjugates remained stable for three months at -20 or 4°C. All three conjugate preparations were bound to human IgG, but neither peroxidase alone nor peroxidase mixed (but unconjugated) with protein A were bound. However, the conjugate preparations differed in their ability to bind IgG. A coupling ratio of 2:1 produced the highest percentage (26%) of functional conjugate (Table 1). Quantification of human IgG by ELISA. Human IgG was quantified in the inhibition assay, using the peroxidase-protein A conjugate.
Peroxidase sp. act., (% of original)'
.6
0
.4
.2
O 1o 0.001
0.01
0.1
IgG CONCENTRATION (pg/mi ) FIG. 1. Standard curve for human IgG. For the inhibition assay, equal volumes (0.15 ml each) of conjugate (2:1) at 10-4 dilution and human IgG at the indicated concentrations were mixed in a test tube and incubated for 30 min at 37°C, and duplicated portions (100 ,il per well) were assayed by the ELISA procedure described in the text.
precision of the assay was based on variation between two separate determinations on dilu-
VOL. 10, 1979
ENZYME-LINKED PROTEIN A IN ELISA
531
tions of the standard IgG solution (5). The esti- ELISA reagent (4) and have extended its applimated variation was ±10%. cation to the quantification of human IgG and Titration of antibody to HCMV by ELISA. the titration of virus-specific antibody. HorsePreliminary block titrations of the reagents were radish peroxidase was chosen because it was less made to define the dilutions where the OD493 expensive than alkaline phosphatase and could was in a reproducibly measurable range (0.1 to be used in cytochemistry (11). The two-step 1.5). The reagents tested were human cytomeg- glutaraldehyde procedure was easy to perform, alovirus antigen, control antigen, and peroxi- but considerable enzyme activity was lost (Table dase-protein A conjugate to be used in the 1) (2, 19). The periodate procedure of Nakane ELISA. Each ELISA included reference sera and Kawaoi (17) may yield higher levels of funcknown to be positive (titer, 1:128) or negative tional conjugate. Direct binding to IgG on mi(titer < 1:8) by complement fixation (CF) test to crotiter plates provided a simple, rapid method HCMV. A log-log plot of OD493 versus serum for assay of functional enzyme-protein A conjudilution yielded a straight line. The ELISA titer gate. The conjugate can readily be separated of a serum was defined as the highest serial from unconjugated peroxidase and protein A (3), twofold dilution with an OD493 of 0.2 or greater. with the possibility of a further increase in the Serum with an OD493 less than 0.2 at a dilution sensitivity of the reagents. of 1:100 was assigned a titer of 1:50. The positive Application of the quantitative inhibition test reference serum had an ELISA titer of 1:1,600, for the measurement of human IgG in body and the negative reference serum had an ELISA fluids must take into account the lack of an titer of 1:50. The variation in positive and neg- absolute specificity of protein A for IgG. Protein ative reference sera from assay to assay was no A binds to the Fc portion of human IgG subgreater than a twofold difference in titer. classes 1, 2, and 4, which comprise approxiA comparison of the antibody titer to HCMV mately 95% of the IgG (13). Subclasses of IgM in 31 sera by CF and ELISA (Fig. 2) resulted in and IgA may also bind protein A (18). The a correlation coefficient of 0.85, and correspond- enzyme-linked protein A ELISA shares advanence was 84% between the two techniques. The tages with enzyme-linked antibody ELISA over ELISA gave 10-fold higher titers than did the CF, including greater sensitivity and insusceptiCF test. bility to anticomplementary activity of serum. Several factors must be considered in interpretDISCUSSION ing the correlation (r = 0.85) between the ELISA The experiments have confirmed the feasibil- and the CF test in the measurement of antibody ity of using enzyme-linked protein A as an against HCMV. As noted above, protein A does not exclusively bind human IgG, but may also bind IgM and IgA. The CF test detects antibody IgG of subclass 1, 2, and 3 as well as IgM (12, 3200 15). The several serum titers showing deviations between CF and ELISA may reflect a difference 1: 1600 in the specificity of the two tests. Examination of the difference, especially with regard to IgM I:SOO . specificity, may provide a means for determining cr whether IgM antibody, perhaps of distinct subclasses, is synthesized in response to cytomegalovirus infection. Binding of protein A to IgG of several mammalian species (8, 9) allows application of en. zyme-linked protein A in the detection of antigens (viral and cellular), using antisera from 1: 50 several different species (3). This property may be particularly useful in a clinical virology laboratory where enzyme-linked antibody against 1:32 1:64 1:126 1:256 1:e 1:16 1:4 antisera from several species of animals would COMPLEMENT FIXATION TITER be required for the detection of different viral FIG. 2. Cytomegalovirus antibody titrated by antigens. *
4
ELISA and CF tests. The ELISA and CF tests were performed on 31 sera as described in the text. Sera with an ELISA titer of
Vol. 10, No. 4
Enzyme-Linked Protein A: an Enzyme-Linked Immunosorbent Assay Reagent for Detection of Human Immunoglobulin G and Virus-Specific Antibody H. PAUL MADORE' *
AND
A. BAUMGARTEN2
Department of Laboratory Medicine, Yale University, School of Medicine, New Haven, Connecticut 06520,2 and the Virology Laboratory, Veterans Administration Medical Center West Haven, Connecticut 065161 Received for publication 10 July 1979
A general-purpose reagent capable of reacting with immunoglobulin G in a modified enzyme-linked immunosorbent assay technique was prepared by using protein A coupled with horseradish peroxidase. The reagent detected low levels (0.003 to 1.0 ,ug/ml) of human immunoglobulin G and was also applied in an enzyme-linked immunosorbent assay for titration of antibody to human cytomegalovirus. The antibody titers to human cytomegalovirus determined by enzymelinked immunosorbent assay and by complement fixation were compared. The correlation coefficient between the two techniques was 0.85, but the enzymelinked immunosorbent assay was 10 times more sensitive than complement fixation in terms of antibody titers detected.
The enzyme-linked immunosorbent assay (ELISA) is widely applied in imnnunodiagnosis (6, 21, 22). Enzyme-linked antibody is the reagent employed to detect antibody-antigen reactions in the assay. Preparation of enzymelinked antibody involves antibody (immunoglobulin G [IgG]) purification and is associated with a marked loss of antibody as well as of enzyme activity during coupling (2, 19). As an altemative to enzyme-linked antibody in ELISA, enzyme-linked protein A was employed to detect antibody against mouse alpha fetoprotein (4). Protein A is a 40,000 molecular weight protein produced by certain strains of Staphylococcus aureus (7, 10, 20). The protein binds the Fc portion of immunoglobulins, chiefly IgG, of several mammalian species, including humans
(8,9).
This report describes the use of horseradish peroxidase-conjugated protein A as a reagent in ELISA for the quantification of human IgG and for the titration of antibody against human cytomegalovirus (HCMV). MATERIALS AND METHODS Horseradish peroxidase protein A conjugation. A modified version of the two-step glutaraldehyde procedure of Avrameas and Ternynck (1) was employed. A 5-mg amount of horseradish peroxidase (type VI; Sigma Chemical Co.) in 0.2 ml of 0.1 M phosphate buffer (pH 6.8), containing 1.25% glutaraldehyde (Sigma, grade I) was allowed to stand at room temperature for 18 h, dialyzed at 4°C against two 200ml changes of 0.05 M carbonate-bicarbonate buffer (pH 9.5), and then brought to 1 ml in the same buffer.
Protein A (Sigma) was dissolved in 0.15 M NaCl-0.1 M carbonate-bicarbonate buffer (pH 9.5) at 5 mg/ml. Activated peroxidase alone (500 ,tg) or the following mixtures of activated peroxidase and protein A, in micrograms, (1,000, 250, 500, 250, 250, and 250) were conjugated for 24 h at 4°C. Excess lysine (10 y1 of a 0.2 M solution) was added to the conjugates, the mixtures were kept for 2 h at 4°C, and then the preparations were dialyzed against three 100-ml changes of phosphate-buffered saline (pH 7.3) at 4°C. The preparations were centrifuged at 20,000 x g for 20 min at 4°C, and the supernatant fluids were removed, dispensed in working volumes, and stored at -20°C until used. Measurement of horseradish peroxidase. Enzyme activity was measured by the method of Mathiesen et al. (16) by adding 10 pg of enzyme in 100 p1 of 0.1 M citrate buffer (pH 5.0), to.l ml of fresh reaction mixture (50 mg of o-phenylenediamine per ml, 0.006% peroxide in 0.1 M sodium citrate buffer, pH 5.0), incubating at room temperature in the dark, and stopping the reaction with 0.55 ml of 2 M sulfuric acid at 5-, 10-, 15-, 30-, 45-, or 60-min intervals. Reaction mixture alone with 2 M sulfuric acid added at 0 min served as a blank. The reaction was linear for 45 min, so the optical density at 493 nmol (OD493) after 30 min was used as a measure of enzyme activity. ELISA procedures. Binding to human IgG was assayed in microtiter plates (polyvinyl tissue culture; Linbro) which were coated with either 100 t1 of human IgG (1 pg/ml; Hyland Laboratories, Inc., Costa Mesa, Calif.) per well, diluted in phosphate-buffered saline (PBS; pH 7.3), or 100 pl of PBS alone per well for 2 h at room temperature or for 24 h at 4°C. The wells were washed three times with 0.05% Tween 20-PBS (TP) and filled with 1% bovine serum albumin (BSA; fraction V) in PBS, and the plates were incubated for 1 h at room temperature. The wells were again washed three times with TP, appropriate dilutions (in 1% 529
530
J. CLIN. MICROBIOL.
MADORE AND BAUMGARTEN
BSA-TP) of the conjugates (100 LI per well) were added, and the plates were incubated for 2 h at room temperature. The wells were washed three times with TP (at 3-min intervals), 250 IL of fresh reaction mixture per well was added, and the enzyme was assayed as described. Measurement was made by using pooled solutions from duplicate wells. Solutions from wells without conjugate were used as blanks. Determination of the inhibition curve of IgG was made by mixing predetermined amounts of conjugate (2:1) in a test tube with increasing amounts of freshly diluted IgG, incubating the mixture for 30 min at 37°C, and then assaying duplicate samples (100 Ld per well) for conjugate binding to IgG on microtiter plates as described above.
Cytomegalovirus antibody was titrated, using HCMV antigen (Flow Laboratories, Inc., Rockville, Md.) diluted 1:200 in PBS. The antigen was adsorbed (100 u1 per well) to microtiter plates for 2 h at room temperature, the supernatant fluids were removed, the wells were filled with 4% BSA-PBS, and the plates were incubated overnight in a moist box at 4°C. The plates were washed three times with TP, duplicate samples (100 1Ld per well) of twofold dilutions of serum in 4% BSA-TP were added, and the covered plates were incubated for 30 min at 35°C. The plates were then washed three times with TP (at 3-min intervals), duplicate samples (100 IlI per well) of conjugate (4:1) diluted 1:500 in 4% BSA-TP were added, and the covered plates were again incubated for 30 min at 35°C. The plates were washed with TP (at 3-min intervals), and enzyme activity was measured as described above. The complement fixation test was performed by standard procedures (14). Human sera. Sera used in this study were obtained from specimens submitted for routine determination of complement-fixing antibody to HCMV.
TABLE 1. Analysis of enzyme-protein A conjugation Sample
A representative standard curve for human IgG (Fig. 1) had a range for detecting 0.003 to 1.0 ,g of IgG per ml. This sensitivity was similar to quantitative ELISA for IgG using enzyme-linked antibody against IgG (5, 6). An estimate of the
PeroxiPeroxi- dase dase, bound (pg/ml)b to IgG
MY Peroxidase (untreated) 100 448 Peroxidase (treated)d 12 373 0 32 863 Conjugate 4:1e 8 2:1 11 430 26 1:1 18 142 19 'The enzyme assay was performed as described in the text. An activity of 100% is equivalent to 78.1 OD491/mg of peroxidase. b OD275 was measured and corrected for estimated amount of protein A in preparation (250 jig of protein A per ml = 0.044 OD275). 'The percent peroxidase (conjugated to protein A) bound to IgG was obtained by converting the OD493 values of peroxidase activity from the microtiter plate assays to micrograms of peroxidase per milliliter bound to IgG, using the peroxidase specific activities listed in the second column, and then dividing the micrograms of peroxidase bound to IgG per milliliter by the total peroxidase in the preparation (third column) and multiplying by 100. d Treated with glutaraldehyde. e Initially 1,000, 500, or 250 jig of peroxidase was mixed with 250 jg of protein A. .0
.8
RESULTS
Horseradish peroxidase protein A conjugation. The specific activity of peroxidase decreased 7- to 10-fold after glutaraldehyde treatment and coupling to protein A (Table 1). Subsequently, peroxidase activity of the conjugates remained stable for three months at -20 or 4°C. All three conjugate preparations were bound to human IgG, but neither peroxidase alone nor peroxidase mixed (but unconjugated) with protein A were bound. However, the conjugate preparations differed in their ability to bind IgG. A coupling ratio of 2:1 produced the highest percentage (26%) of functional conjugate (Table 1). Quantification of human IgG by ELISA. Human IgG was quantified in the inhibition assay, using the peroxidase-protein A conjugate.
Peroxidase sp. act., (% of original)'
.6
0
.4
.2
O 1o 0.001
0.01
0.1
IgG CONCENTRATION (pg/mi ) FIG. 1. Standard curve for human IgG. For the inhibition assay, equal volumes (0.15 ml each) of conjugate (2:1) at 10-4 dilution and human IgG at the indicated concentrations were mixed in a test tube and incubated for 30 min at 37°C, and duplicated portions (100 ,il per well) were assayed by the ELISA procedure described in the text.
precision of the assay was based on variation between two separate determinations on dilu-
VOL. 10, 1979
ENZYME-LINKED PROTEIN A IN ELISA
531
tions of the standard IgG solution (5). The esti- ELISA reagent (4) and have extended its applimated variation was ±10%. cation to the quantification of human IgG and Titration of antibody to HCMV by ELISA. the titration of virus-specific antibody. HorsePreliminary block titrations of the reagents were radish peroxidase was chosen because it was less made to define the dilutions where the OD493 expensive than alkaline phosphatase and could was in a reproducibly measurable range (0.1 to be used in cytochemistry (11). The two-step 1.5). The reagents tested were human cytomeg- glutaraldehyde procedure was easy to perform, alovirus antigen, control antigen, and peroxi- but considerable enzyme activity was lost (Table dase-protein A conjugate to be used in the 1) (2, 19). The periodate procedure of Nakane ELISA. Each ELISA included reference sera and Kawaoi (17) may yield higher levels of funcknown to be positive (titer, 1:128) or negative tional conjugate. Direct binding to IgG on mi(titer < 1:8) by complement fixation (CF) test to crotiter plates provided a simple, rapid method HCMV. A log-log plot of OD493 versus serum for assay of functional enzyme-protein A conjudilution yielded a straight line. The ELISA titer gate. The conjugate can readily be separated of a serum was defined as the highest serial from unconjugated peroxidase and protein A (3), twofold dilution with an OD493 of 0.2 or greater. with the possibility of a further increase in the Serum with an OD493 less than 0.2 at a dilution sensitivity of the reagents. of 1:100 was assigned a titer of 1:50. The positive Application of the quantitative inhibition test reference serum had an ELISA titer of 1:1,600, for the measurement of human IgG in body and the negative reference serum had an ELISA fluids must take into account the lack of an titer of 1:50. The variation in positive and neg- absolute specificity of protein A for IgG. Protein ative reference sera from assay to assay was no A binds to the Fc portion of human IgG subgreater than a twofold difference in titer. classes 1, 2, and 4, which comprise approxiA comparison of the antibody titer to HCMV mately 95% of the IgG (13). Subclasses of IgM in 31 sera by CF and ELISA (Fig. 2) resulted in and IgA may also bind protein A (18). The a correlation coefficient of 0.85, and correspond- enzyme-linked protein A ELISA shares advanence was 84% between the two techniques. The tages with enzyme-linked antibody ELISA over ELISA gave 10-fold higher titers than did the CF, including greater sensitivity and insusceptiCF test. bility to anticomplementary activity of serum. Several factors must be considered in interpretDISCUSSION ing the correlation (r = 0.85) between the ELISA The experiments have confirmed the feasibil- and the CF test in the measurement of antibody ity of using enzyme-linked protein A as an against HCMV. As noted above, protein A does not exclusively bind human IgG, but may also bind IgM and IgA. The CF test detects antibody IgG of subclass 1, 2, and 3 as well as IgM (12, 3200 15). The several serum titers showing deviations between CF and ELISA may reflect a difference 1: 1600 in the specificity of the two tests. Examination of the difference, especially with regard to IgM I:SOO . specificity, may provide a means for determining cr whether IgM antibody, perhaps of distinct subclasses, is synthesized in response to cytomegalovirus infection. Binding of protein A to IgG of several mammalian species (8, 9) allows application of en. zyme-linked protein A in the detection of antigens (viral and cellular), using antisera from 1: 50 several different species (3). This property may be particularly useful in a clinical virology laboratory where enzyme-linked antibody against 1:32 1:64 1:126 1:256 1:e 1:16 1:4 antisera from several species of animals would COMPLEMENT FIXATION TITER be required for the detection of different viral FIG. 2. Cytomegalovirus antibody titrated by antigens. *
4
ELISA and CF tests. The ELISA and CF tests were performed on 31 sera as described in the text. Sera with an ELISA titer of
