[61]
IMMUNODIAGNOSIS OF INFECTIOUS DISEASES
529
values, Kaff = 2.7 × l09 M-L If 58% immunoreactivity of both tracer and inhibitor is assumed, then It = 411 x 10-12 M, and Kale = 4.6 × l09 M -1. The calculated affinity constants are lower than that shown for the intact antibody in Table I. This finding may be due to a variety of reasons, including the use of Fab fragments instead of intact antibodies, the presence of nonimmune IgG in the ascitic fluid, and/or the corrections made for the immunoreactive fraction of the radiolabeled Fab fragments or antigen. Acknowledgments This research was supported by grants from the Deutsche Forschungsgemeinschaft, Wa473/4-1, and from the National Large Bowel Program, National Cancer Institute, Grant CA 37808. We wish to thank Karen Rickard for expert technical assistance, and Dr. Y. H. Joy Yang for production of monoclonal antibodies.
[61] I m m u n o a s s a y s for D i a g n o s i s o f I n f e c t i o u s D i s e a s e s
By ROBERT H. YOLKEN Introduction Traditionally, the diagnoses of infectious diseases have been based on the isolation of the infecting microorganisms in pure culture. Recently, however, infectious disease diagnoses have relied more heavily on the direct identification of infecting organisms in blood and other body fluids of the ill individual. The diagnosis of infectious diseases by the direct detection of microorganisms has been particularly important for the identification of viral agents since these agents are particularly difficult to detect in short periods of time by means of standard cultivation methods. 1 Direct detection methods are also highly useful in the detection of bacterial, parasitic, and fungal pathogens that are fastidious or difficult to cultivate under standard laboratory conditions. 2 Most of these assays have relied on the measurement of the binding of antigens to defined antimicrobial antibodies. Immunoassays have a number of advantages for infectious disease diagnosis. These advantages are based on the sensitivi D. A. Fuccillo, I. C. Shekarchi, and J. L. Sever, in "Manual of Clinical Laboratory Immunology" (N. R. Rose, H. Friedman, and J. L. Fahey, eds.), 3rd ed., p. 489. American Society for Microbiology, Washington, D. C., 1986. 2 N. J. Schmidt, Med. Clin. North Am. 67, 953 (1983).
METHODS IN ENZYMOLOGY, VOL. 184
Copyright © 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
530
APPLICATIONS
[61 ]
ity and specificity inherent in antigen-antibody interactions. While there are a number of immunoassay systems that can be used for the direct detection of infectious antigens in human body fluids, solid-phase immunoassay techniques have attained widespread usage for this purpose.3 The use of enzymatic methods offers the possibility of high degrees of sensitivity owing to the amplificatory nature of the enzyme-substrate reaction. Furthermore, the stability and safety of many enzyme-substrate systems allow for the application of enzyme immunoassays to a wide range of clinical and laboratory environments/-7 Microorganisms are capable of infecting humans and causing disease in low concentrations. It is thus necessary that efficient assay systems for the diagnosis of infectious diseases attain a high rate of assay sensitivity. While there are a number of factors that determine the sensitivity of solidphase immunoassay systems, the most critical are related to the efficiency of the binding of labeled immunoreagents to microbial antigens. Initially, enzyme immunoassays for the detection of infecting antigens have been performed in "direct" formats in which the antimicrobial immunoglobulin is coupled to an enzyme. 8 One problem in this type of enzyme immunoassay system is that chemical linkage can result in a variable loss of the antigen-binding capacity of the antibody. In addition, the large mass of the enzyme-immunoglobulin complexes can result in slower rates of diffusion and thus in less favorable reaction kinetics. Furthermore, the variability inherent in the interactions of macromolecular species in solutions makes it difficult to control the rate and extent of immunoglobulin-enzyme coupling reactions. These problems can limit the development of immunoassay systems with consistent performance characteristics. There are a number of possible solutions to the problems inherent in the direct linkage of enzymes to immunoglobulins. For example, immunoassays can be performed in indirect formats in which an unlabeled primary antimicrobial immunoglobulin is reacted with immobilized antigens and the resultant complex is quantified by an enzyme-conjugated second antibody. 9 A single enzyme conjugate can thus be utilized to de3 R. H. Yolken, Rev. Infect. Dis. 4, 36 (1982). 4 S. Avrameas, T. Ternynck, and J. L. Guesdon, Scand. J. Immunol. 8, (Suppl. 7), 7 5 G. B. Wisdom, Clin. Chem. 22, 1243 (1976). 6 A. Voller, A. Bartlett, and D. E. Bidwell, Trans. R. Soc. Trop. Med. Hyg. 70, 98 (1976). 7 R. H. Yolken, Yale J. Biol. 59 (1), 25 (1986). g R. H. Yolken, H. W. Kim, T. Clem, R. G. Wyatt, A. R. Kalica, R. M. Chanock, and A. Z. Kapikian, Lancet 2, 263 (1977). 9 R. H. Yolken, R. Viscidi, F. Leister, C. Harris, and S-B. Wee, in "Manual of Clinical Laboratory Immunology" (N. R. Rose, H. Friedman, and J. L. Fahey, eds.), 3rd ed. p. 573. American Society for Microbiology, Washington, D. C., 1986.
[61]
IMMUNODIAGNOSIS OF INFECTIOUS DISEASES
531
t e c t a number of target antigens. The use of this format avoids the need to generate enzyme-labeled immunoglobulins for each target antigen. However, this format still relies on the direct conjugation of enzyme and immunoglobulin molecules and is thus subject to the above-mentioned disadvantages. More importantly, the species specificity of many antiimmunoglobulin conjugates is not absolute. Nonspecific binding can therefore generate high background levels or, in certain cases, false-positive reactions.~° The same problem is inherent in the use of other immunoglobulin-binding macromolecules such as staphylococcal protein A. H Another approach to this problem is to link the antimicrobial immunoglobulin to a low molecular weight molecule which can then be detected specifically with an enzyme-labeled macromolecular counterpart. Such a system makes it less likely that the antigen-binding capacity of the immunoglobulin will be altered by macromolecular cross-linking. In addition, low molecular weight agents can be added in great excess to immunoglobulins, thus ensuring that coupling reactions will proceed to completion in relatively short periods of time. While a number of low molecular weight agents can be bound to immunoglobulin, biotin has been most widely used for the detection of microbial antigens in solid-phase formats, hA3 The high affinity of avidin for biotin and biotin-substituted macromolecules ensures high degrees of assay sensitivity. There are a number of methods that can be utilized for the detection of biotin-labeled immunoreagents. For example, avidin can be covalently coupled with enzyme and reacted to the biotinylated antibody on the solid-phase surface. Chemically modified avidin or bacterial streptavidin can be used to minimize nonspecific binding which sometimes discourages the use of the system.~4 Another limitation, the need for precise chemical coupling of avidin to enzyme, can also be minimized by forming complexes between unlabeled avidin and biotin-substituted avidin. This approach can be optimized to generate more enzyme-catalyzed signal than methods using enzyme-linked avidin. J5 In terms of the labeling of immunoglobulins, we have found that standard biotinylation reagents such as biotin-N-hydroxysuccinimide ester can be used to label virtually all polyclonal antibodies over a wide range of ester and antibody concentrations (Fig. 1). We have also found that 10 R. H. Yolken and P. J. Stopa, J. Clin. Microbiol. 10, 703 (1979). n R. H. Yolken and P. J. Stopa, J. Clin. Microbiol. 11, 546 (1980). n E. A. Bayer, M. Wilchek, and E. Skutelsky, FEBS Lett. 68, 240 (1976). t3 j. L, Guesdon, T. Ternynck, and S. Avrameas, J. Histochern. Cytochem. 27, 1131 (1979). 14 L. S. Nerurkar, N. R. Miller, M. Namba, M, Monzon, G. Brashears, G. Scherba, and J. L. Sever, J. Clin. Microbiol. 25(1), 128 (1987). ~5R. H. Yolken, F. J. Leister, L. S. Whitcomb, and M. Santosham, J. lmmunol. Methods 56, 319 (1983).
532
APPLICATIONS
[6 1]
2.0
1.0
50'1
o
'\
o.>_.Q.0.5
'~,,
Fb--
o
_o iJ.
U.l m 0.1
0
"
|
~
50
5.0
It.
0.5
_ ~
~
~
-
04"1
At
Neg CTR
PRP (ng/rnl)
FIG. 1. Activity of biotin conjugates for the measurement of Haemophilus influenzae polyribitol phosphate (PRP). Numbers indicate the ratio (mg/mg) of biotin-N-hydroxysuccinimide ester to anti-PRP lgG utilized to formulate the conjugate. The specific activity was measured by means of a solid-phase assay performed in a manner similar to that described in the text. (From Ref. 15.)
monoclonal antibodies can be labeled with this method; h o w e v e r , some monoclonal antibodies a p p e a r to lose m u c h of their antigen-binding capacity. This p h e n o m e n o n is p r o b a b l y related to the linkage of biotin to an amino group located n e a r the antigen-combining site of the immunoglobulin molecule. As an alternative to linkage of biotin to amino groups, biotin hydrazide can be coupled to carboxyl groups of immunoglobulin molecules b y reaction with carbodiimide. ~6 Polyclonal chicken antirotavirus antibodies labeled with biotin by this method p e r f o r m e d as well in e n z y m e i m m u n o a s s a y s for detection of rotavirus antigens in stool specimens as the s a m e immunoglobulins labeled b y the N - h y d r o x y s u c c i n i m i d e ester method. Other m e a n s o f biotin linkage, such as reaction of biotin hydrazide with periodate-oxidized sugar residues, might also be utilized in 16D. J. O'Shannessy, M. J. Dobersen, and R. H. Quarles, lmmunol. Lett. 8, 273 (1984).
[61]
533
IMMUNODIAGNOSlS OF INFECTIOUS DISEASES
cases in which other linkages result in a diminution of antigen-binding activity. We have devised a number of biotin-based assays for the detection of microbial antigens in the body fluids of infected humans. ~5 We have found, for example, that assays utilizing biotinylated antibodies and a complex of avidin and biotinylated peroxidase can be used for the sensitive detection of cell wall polysaccharides of pathogenic bacteria such as Haemophilus influenzae and Streptococcus pneumoniae (Fig. 2). These assays were somewhat more sensitive than analogous assays that utilized immunoglobulins directly labeled with enzyme. We have also devised similar systems for the direct detection of viral antigens in body fluids. Furthermore, other enzymes can be used in addition to horseradish peroxidase to accomplish the enzyme-substrate indicator reaction. More
2.0-
_o I.IJ
0.1
*"~\~ %\
0
I
I
10.4
2.6
I
I
0.64 0 . 1 6 PRP (ng/ml)
I
I
0.04
Neg Control
FIG. 2. Sensitivity of enzyme immunoassays for the measurement of Haemophilus influenzae PRP. The conjugate systems utilized were as follows: 0 , biotin-labeled rabbit antiPRP, enzyme-labeled biotin-avidin complex; A, biotin-labeled rabbit anti-PRP, enzymelabeled avidin; IZ, unlabeled rabbit anti-PRP, fluorescein-labeled anti-IgG, enzyme-labeled antifluorescein; and O, unlabeled rabbit PRP, enzyme-labeled anti-rabbit IgG. (From Ref. 15.)
534
APPLICATIONS
[61]
900
"2
6oo
e-
~
300
,7
0
16
,
,
T
'I
64
256
1024
4096
Reciprocal Dilution FIG. 3. Detection of HIV antigens in serial dilutions of cell culture fluids. Samples from a cell culture inoculated with an HIV isolate ((3) and from a control uninoculated culture (0) were tested in an avidin-biotin-enhanced immunoassay for HIV antigens. Samples from the HIV cultures were also incubated with a monoclonal antibody to HIV-1 p24 protein and were then tested in the immunoassay (0). The assays utilized streptavidin labeled with/3galactosidase and a fluorescent substrate for that enzyme. (From Ref. 18.)
recently, we have utilized biotinylated immunoglobulins for the detection of rotavirus antigenic variants 17 and human immunodeficiency viruses. 18 In both cases the biotin-based immunoassays proved to be at least as sensitive as analogous immunoassays utilizing immunoreagents labeled with other markers (Fig. 3). One group of enzymes particularly useful for the diagnosis of infectious diseases are the bacterial fl-lactamases. 19,20 In addition to possessing 17 S. L. Vonderfecht, R. L. Miskuff, J. J. Eiden, and R. H. Yolken, J. Clin. Microbiol. 22, 726 (1985). 18 R. Viscidi, H. Farzadegan, F. Leister, M. L. Francisco, F. Polk, and R. Yolken, J. Clin. Microbiol. 26(3), 453 (1987). 19 U. Joshi, V. Rashavan, G. Zemse, A. Sheth, P. S. Borkar, and S. Ramachandran, in "Enzyme Labeled Immunoassay of Hormones and Drugs" (S. B. Pal, ed,), p. 233. de Gruyter, Berlin, 1978. 20 p. B. Geetha, A. A. Koshy, C. N. Dandawate, B. K. Shaikh, S. N. Ghosh, and P. S. Borkar, Hind. Antibiot. Bull. 24, 34 (1982).
[61]
535
IMMUNODIAGNOSIS OF INFECTIOUS DISEASES
100
---
80 60
. m
.>_ ,( t.-.t--,
0..
20 0
~, J 25
I 6.2
I l -'4 1.5 0.4 0.1 PRP (ng/well)
--~ 0.025
//]
O, Neg Ctr
FIG. 4. Comparison of enzyme immunoassays for PRP. II, Results with biotinylated/3lactamase; O, results with biotinylated horseradish peroxidase. Percent activity was calculated from duplicate determinations as described in the text. The values for the negative control indicate the mean and standard deviation of reactivity in quadruplicate wells to which btrffer was added in place of antigen dilution. (From Ref. 21.)
high turnover rates, these enzymes are stable, inexpensive, and can make use of a wide range of substrates. We have utilized biotinylated fl-lactamase from B a c i l l u s c e r e u s and a starch-iodine penicillin substrate solution to devise immunoassays for the detection of rotavirus and adenovirus antigens in human body fluids. 21 The sensitivity of these assays was at least as favorable as analogous ones utilizing biotinylated horseradish peroxidase (Fig. 4). In addition, biotin-labeled reagents have been utilized for the detection of a wide range of bacterial, viral, fungal, and parasitic antigens. 2z-28 As:l R. H. Yolken and S.-B. Wee, J. Clin. Microbiol. 19(3), 356 (1984). ~z j. C. Hierholzer, K. H. Johansson, L. J. Anderson, C. J. Tsou, and P. E. Halonen, J. Clin. Microbiol. 25(9), 1662 (1987). 23 S. Edwards, and G. C. Gitao, Vet. Microbiol. 13(2), 135 (1987). ~4 L. C. Shekarchi, D. A. Fuccillo, R. Strouse, and J. L. Sever, J. Clin. MicrobioL 25(2), 320 (1987). 25 V. P. Kurup, Zentralbl. Bakteriol. Mikrobiol. Hgy. Ser. A 261(4), 509 (1986). 26 A. Belmaaza, J. Hamel, S. Mousseau, S. Montplaisir, and B. R. Brodeur, J. Clin. Microbiol. 24(3), 440 (1986). 27 p. Vilja, H. J. Turunen, and P. O. Leinikki, J. Clin. MicrobioL 22(4), 637 (1985). 28 A. Sutton, W. F. Vann, A. B. Karpas, K. E. Stein, and R. Schneerson, J. Immunol. Methods 82, 215 (1985).
536
APPLICATIONS
[61]
says making use of avidin-biotin interactions have also been utilized for the measurement of the immune response to infection with microbial antigens. While the degrees of sensitivity and specificity obtained in such assays varied depending on the binding characteristics of the immunoreagents, the utilization of avidin-biotin interactions has generally resuited in the attainment of the maximal degree of sensitivity allowed by immunoreagents. The use of avidin-biotin interactions has thus allowed for the development of a number of important assays for the diagnosis of infectious diseases. The efficient application of practical assays for microbial detection should allow for the improved clinical management of patients with suspected infections and for the containment of disease transmission in susceptible populations. Preparation of Biotin-Labeled Immunoglobulins Immunoglobulins or immunoglobulin fragments are purified from the serum to be labeled by standard methods and diluted to a concentration of 1 mg/ml in 60 m M carbonate buffer (pH 9.0). Biotin-N-hydroxysuccinimide ester is dissolved in dimethyl sulfoxide to make a stock solution of I mg/ml. Aliquots of biotin are added to the immunoglobulin (generally at a biotin to immunoglobulin molar ratio of 10: 1), and the mixture is incubated for 3 hr at 25 °. After incubation, unreacted biotin is removed by extensive dialysis against PBS or by fractionation with Sephadex G-25. The biotinylated antibody is aliquoted and stored at - 2 0 °. Preservatives such as sodium azide (20 rag/liter) can be added to prevent bacterial contamination. Coat alternate rows of wells of the microtiter plate with a dilution of antiviral IgG and equal dilution of IgG from the same animal species that does not contain measurable antibody to rotavirus. Incubate the plate at least overnight at 4 °. If the plate is not used the next day, it should be covered with Parafilm and stored at 4 ° until used. Wash the plate 5 times with phosphate-buffered saline containing 0.05% Tween 20 (PBS-Tween). Add 50/zl of PBS-Tween to each of the wells. Buffer containing nonimmune animal sera can also be utilized to reduce the occurrence of nonspecific reactions. Add an equal amount of specimen to two wells coated with antiviral IgG and two wells coated with nonimmune IgG. Include a weakly positive control and four negative controls in each test. Incubate the plate for 2 hr at 37° or overnight at 4 °. Add antiviral antibody that has been labeled with biotin, and incubate the plate for 1 hr at 37°. Wash the plate 5 times with PBS-Tween. Prepare the avidin-biotin complex by adding predetermined amounts
[62]
ANTIPOLYSACCHARIDE ANTIBODIES
537
of avidin and biotin-linked enzyme to a tube containing PBS-Tween. Incubate for 10 min at 37 °, and add the complex to the wells. Incubate the plate for 30 min at 37°. Wash the plate 5 times with PBS-Tween. Add appropriate substrate. Incubate the plate at 37 ° or room temperature until the weakly positive control has visible color equivalent to an optical density of approximately 0.1 measured at the appropriate wavelength. Comments. For each sample calculate the virus-specific activity by subtracting the mean activity of the wells coated with the nonimmune IgG from that measured in wells coated with the antiviral IgG. To ensure accurate quantification, specimens giving readings of greater than 1.2 optical density units should be diluted 1 : 10 and retested. Calculate the mean and standard deviation of the virus activity of the negative controls. A specimen is considered positive if its mean activity is greater than 2 standard deviations above the mean activity of the negative controls. Alternatively, a specimen can be considered positive if its specific activity is greater than that of the weakly positive control. If qualitative visual determinations are used, a specimen is considered positive if the degree of color in the antiviral IgG wells is different than the amount of color generated in the nonimmune wells and in the wells containing the weakly positive control samples. Note that if the fl-lactamase starch-iodine system is utilized, a positive reaction will be manifested by a decrease in color, whereas in the case of most other enzymes, the binding of enzyme will result in an increase in color production. Acknowledgment This work was supported by contract N01-AI-52579from the National Institute of Allergies and Infectious Diseases.
[62]
Enzyme-Linked
Immunosorbent
Assay
By WILLIE F. VANN, ANN SUTTON, and RACHEL SCHNEERSON Capsular polysaccharides are important in invasive bacterial disease as virulence factors and as immunogens. Since polysaccharides and oligosaccharides often do not adhere to a plastic solid phase as well as proteins, the measurement of antibodies to polysaccharides by enzymeMETHODS IN ENZYMOLOGY, VOL. 184
Copyright © 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
IMMUNODIAGNOSIS OF INFECTIOUS DISEASES
529
values, Kaff = 2.7 × l09 M-L If 58% immunoreactivity of both tracer and inhibitor is assumed, then It = 411 x 10-12 M, and Kale = 4.6 × l09 M -1. The calculated affinity constants are lower than that shown for the intact antibody in Table I. This finding may be due to a variety of reasons, including the use of Fab fragments instead of intact antibodies, the presence of nonimmune IgG in the ascitic fluid, and/or the corrections made for the immunoreactive fraction of the radiolabeled Fab fragments or antigen. Acknowledgments This research was supported by grants from the Deutsche Forschungsgemeinschaft, Wa473/4-1, and from the National Large Bowel Program, National Cancer Institute, Grant CA 37808. We wish to thank Karen Rickard for expert technical assistance, and Dr. Y. H. Joy Yang for production of monoclonal antibodies.
[61] I m m u n o a s s a y s for D i a g n o s i s o f I n f e c t i o u s D i s e a s e s
By ROBERT H. YOLKEN Introduction Traditionally, the diagnoses of infectious diseases have been based on the isolation of the infecting microorganisms in pure culture. Recently, however, infectious disease diagnoses have relied more heavily on the direct identification of infecting organisms in blood and other body fluids of the ill individual. The diagnosis of infectious diseases by the direct detection of microorganisms has been particularly important for the identification of viral agents since these agents are particularly difficult to detect in short periods of time by means of standard cultivation methods. 1 Direct detection methods are also highly useful in the detection of bacterial, parasitic, and fungal pathogens that are fastidious or difficult to cultivate under standard laboratory conditions. 2 Most of these assays have relied on the measurement of the binding of antigens to defined antimicrobial antibodies. Immunoassays have a number of advantages for infectious disease diagnosis. These advantages are based on the sensitivi D. A. Fuccillo, I. C. Shekarchi, and J. L. Sever, in "Manual of Clinical Laboratory Immunology" (N. R. Rose, H. Friedman, and J. L. Fahey, eds.), 3rd ed., p. 489. American Society for Microbiology, Washington, D. C., 1986. 2 N. J. Schmidt, Med. Clin. North Am. 67, 953 (1983).
METHODS IN ENZYMOLOGY, VOL. 184
Copyright © 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
530
APPLICATIONS
[61 ]
ity and specificity inherent in antigen-antibody interactions. While there are a number of immunoassay systems that can be used for the direct detection of infectious antigens in human body fluids, solid-phase immunoassay techniques have attained widespread usage for this purpose.3 The use of enzymatic methods offers the possibility of high degrees of sensitivity owing to the amplificatory nature of the enzyme-substrate reaction. Furthermore, the stability and safety of many enzyme-substrate systems allow for the application of enzyme immunoassays to a wide range of clinical and laboratory environments/-7 Microorganisms are capable of infecting humans and causing disease in low concentrations. It is thus necessary that efficient assay systems for the diagnosis of infectious diseases attain a high rate of assay sensitivity. While there are a number of factors that determine the sensitivity of solidphase immunoassay systems, the most critical are related to the efficiency of the binding of labeled immunoreagents to microbial antigens. Initially, enzyme immunoassays for the detection of infecting antigens have been performed in "direct" formats in which the antimicrobial immunoglobulin is coupled to an enzyme. 8 One problem in this type of enzyme immunoassay system is that chemical linkage can result in a variable loss of the antigen-binding capacity of the antibody. In addition, the large mass of the enzyme-immunoglobulin complexes can result in slower rates of diffusion and thus in less favorable reaction kinetics. Furthermore, the variability inherent in the interactions of macromolecular species in solutions makes it difficult to control the rate and extent of immunoglobulin-enzyme coupling reactions. These problems can limit the development of immunoassay systems with consistent performance characteristics. There are a number of possible solutions to the problems inherent in the direct linkage of enzymes to immunoglobulins. For example, immunoassays can be performed in indirect formats in which an unlabeled primary antimicrobial immunoglobulin is reacted with immobilized antigens and the resultant complex is quantified by an enzyme-conjugated second antibody. 9 A single enzyme conjugate can thus be utilized to de3 R. H. Yolken, Rev. Infect. Dis. 4, 36 (1982). 4 S. Avrameas, T. Ternynck, and J. L. Guesdon, Scand. J. Immunol. 8, (Suppl. 7), 7 5 G. B. Wisdom, Clin. Chem. 22, 1243 (1976). 6 A. Voller, A. Bartlett, and D. E. Bidwell, Trans. R. Soc. Trop. Med. Hyg. 70, 98 (1976). 7 R. H. Yolken, Yale J. Biol. 59 (1), 25 (1986). g R. H. Yolken, H. W. Kim, T. Clem, R. G. Wyatt, A. R. Kalica, R. M. Chanock, and A. Z. Kapikian, Lancet 2, 263 (1977). 9 R. H. Yolken, R. Viscidi, F. Leister, C. Harris, and S-B. Wee, in "Manual of Clinical Laboratory Immunology" (N. R. Rose, H. Friedman, and J. L. Fahey, eds.), 3rd ed. p. 573. American Society for Microbiology, Washington, D. C., 1986.
[61]
IMMUNODIAGNOSIS OF INFECTIOUS DISEASES
531
t e c t a number of target antigens. The use of this format avoids the need to generate enzyme-labeled immunoglobulins for each target antigen. However, this format still relies on the direct conjugation of enzyme and immunoglobulin molecules and is thus subject to the above-mentioned disadvantages. More importantly, the species specificity of many antiimmunoglobulin conjugates is not absolute. Nonspecific binding can therefore generate high background levels or, in certain cases, false-positive reactions.~° The same problem is inherent in the use of other immunoglobulin-binding macromolecules such as staphylococcal protein A. H Another approach to this problem is to link the antimicrobial immunoglobulin to a low molecular weight molecule which can then be detected specifically with an enzyme-labeled macromolecular counterpart. Such a system makes it less likely that the antigen-binding capacity of the immunoglobulin will be altered by macromolecular cross-linking. In addition, low molecular weight agents can be added in great excess to immunoglobulins, thus ensuring that coupling reactions will proceed to completion in relatively short periods of time. While a number of low molecular weight agents can be bound to immunoglobulin, biotin has been most widely used for the detection of microbial antigens in solid-phase formats, hA3 The high affinity of avidin for biotin and biotin-substituted macromolecules ensures high degrees of assay sensitivity. There are a number of methods that can be utilized for the detection of biotin-labeled immunoreagents. For example, avidin can be covalently coupled with enzyme and reacted to the biotinylated antibody on the solid-phase surface. Chemically modified avidin or bacterial streptavidin can be used to minimize nonspecific binding which sometimes discourages the use of the system.~4 Another limitation, the need for precise chemical coupling of avidin to enzyme, can also be minimized by forming complexes between unlabeled avidin and biotin-substituted avidin. This approach can be optimized to generate more enzyme-catalyzed signal than methods using enzyme-linked avidin. J5 In terms of the labeling of immunoglobulins, we have found that standard biotinylation reagents such as biotin-N-hydroxysuccinimide ester can be used to label virtually all polyclonal antibodies over a wide range of ester and antibody concentrations (Fig. 1). We have also found that 10 R. H. Yolken and P. J. Stopa, J. Clin. Microbiol. 10, 703 (1979). n R. H. Yolken and P. J. Stopa, J. Clin. Microbiol. 11, 546 (1980). n E. A. Bayer, M. Wilchek, and E. Skutelsky, FEBS Lett. 68, 240 (1976). t3 j. L, Guesdon, T. Ternynck, and S. Avrameas, J. Histochern. Cytochem. 27, 1131 (1979). 14 L. S. Nerurkar, N. R. Miller, M. Namba, M, Monzon, G. Brashears, G. Scherba, and J. L. Sever, J. Clin. Microbiol. 25(1), 128 (1987). ~5R. H. Yolken, F. J. Leister, L. S. Whitcomb, and M. Santosham, J. lmmunol. Methods 56, 319 (1983).
532
APPLICATIONS
[6 1]
2.0
1.0
50'1
o
'\
o.>_.Q.0.5
'~,,
Fb--
o
_o iJ.
U.l m 0.1
0
"
|
~
50
5.0
It.
0.5
_ ~
~
~
-
04"1
At
Neg CTR
PRP (ng/rnl)
FIG. 1. Activity of biotin conjugates for the measurement of Haemophilus influenzae polyribitol phosphate (PRP). Numbers indicate the ratio (mg/mg) of biotin-N-hydroxysuccinimide ester to anti-PRP lgG utilized to formulate the conjugate. The specific activity was measured by means of a solid-phase assay performed in a manner similar to that described in the text. (From Ref. 15.)
monoclonal antibodies can be labeled with this method; h o w e v e r , some monoclonal antibodies a p p e a r to lose m u c h of their antigen-binding capacity. This p h e n o m e n o n is p r o b a b l y related to the linkage of biotin to an amino group located n e a r the antigen-combining site of the immunoglobulin molecule. As an alternative to linkage of biotin to amino groups, biotin hydrazide can be coupled to carboxyl groups of immunoglobulin molecules b y reaction with carbodiimide. ~6 Polyclonal chicken antirotavirus antibodies labeled with biotin by this method p e r f o r m e d as well in e n z y m e i m m u n o a s s a y s for detection of rotavirus antigens in stool specimens as the s a m e immunoglobulins labeled b y the N - h y d r o x y s u c c i n i m i d e ester method. Other m e a n s o f biotin linkage, such as reaction of biotin hydrazide with periodate-oxidized sugar residues, might also be utilized in 16D. J. O'Shannessy, M. J. Dobersen, and R. H. Quarles, lmmunol. Lett. 8, 273 (1984).
[61]
533
IMMUNODIAGNOSlS OF INFECTIOUS DISEASES
cases in which other linkages result in a diminution of antigen-binding activity. We have devised a number of biotin-based assays for the detection of microbial antigens in the body fluids of infected humans. ~5 We have found, for example, that assays utilizing biotinylated antibodies and a complex of avidin and biotinylated peroxidase can be used for the sensitive detection of cell wall polysaccharides of pathogenic bacteria such as Haemophilus influenzae and Streptococcus pneumoniae (Fig. 2). These assays were somewhat more sensitive than analogous assays that utilized immunoglobulins directly labeled with enzyme. We have also devised similar systems for the direct detection of viral antigens in body fluids. Furthermore, other enzymes can be used in addition to horseradish peroxidase to accomplish the enzyme-substrate indicator reaction. More
2.0-
_o I.IJ
0.1
*"~\~ %\
0
I
I
10.4
2.6
I
I
0.64 0 . 1 6 PRP (ng/ml)
I
I
0.04
Neg Control
FIG. 2. Sensitivity of enzyme immunoassays for the measurement of Haemophilus influenzae PRP. The conjugate systems utilized were as follows: 0 , biotin-labeled rabbit antiPRP, enzyme-labeled biotin-avidin complex; A, biotin-labeled rabbit anti-PRP, enzymelabeled avidin; IZ, unlabeled rabbit anti-PRP, fluorescein-labeled anti-IgG, enzyme-labeled antifluorescein; and O, unlabeled rabbit PRP, enzyme-labeled anti-rabbit IgG. (From Ref. 15.)
534
APPLICATIONS
[61]
900
"2
6oo
e-
~
300
,7
0
16
,
,
T
'I
64
256
1024
4096
Reciprocal Dilution FIG. 3. Detection of HIV antigens in serial dilutions of cell culture fluids. Samples from a cell culture inoculated with an HIV isolate ((3) and from a control uninoculated culture (0) were tested in an avidin-biotin-enhanced immunoassay for HIV antigens. Samples from the HIV cultures were also incubated with a monoclonal antibody to HIV-1 p24 protein and were then tested in the immunoassay (0). The assays utilized streptavidin labeled with/3galactosidase and a fluorescent substrate for that enzyme. (From Ref. 18.)
recently, we have utilized biotinylated immunoglobulins for the detection of rotavirus antigenic variants 17 and human immunodeficiency viruses. 18 In both cases the biotin-based immunoassays proved to be at least as sensitive as analogous immunoassays utilizing immunoreagents labeled with other markers (Fig. 3). One group of enzymes particularly useful for the diagnosis of infectious diseases are the bacterial fl-lactamases. 19,20 In addition to possessing 17 S. L. Vonderfecht, R. L. Miskuff, J. J. Eiden, and R. H. Yolken, J. Clin. Microbiol. 22, 726 (1985). 18 R. Viscidi, H. Farzadegan, F. Leister, M. L. Francisco, F. Polk, and R. Yolken, J. Clin. Microbiol. 26(3), 453 (1987). 19 U. Joshi, V. Rashavan, G. Zemse, A. Sheth, P. S. Borkar, and S. Ramachandran, in "Enzyme Labeled Immunoassay of Hormones and Drugs" (S. B. Pal, ed,), p. 233. de Gruyter, Berlin, 1978. 20 p. B. Geetha, A. A. Koshy, C. N. Dandawate, B. K. Shaikh, S. N. Ghosh, and P. S. Borkar, Hind. Antibiot. Bull. 24, 34 (1982).
[61]
535
IMMUNODIAGNOSIS OF INFECTIOUS DISEASES
100
---
80 60
. m
.>_ ,( t.-.t--,
0..
20 0
~, J 25
I 6.2
I l -'4 1.5 0.4 0.1 PRP (ng/well)
--~ 0.025
//]
O, Neg Ctr
FIG. 4. Comparison of enzyme immunoassays for PRP. II, Results with biotinylated/3lactamase; O, results with biotinylated horseradish peroxidase. Percent activity was calculated from duplicate determinations as described in the text. The values for the negative control indicate the mean and standard deviation of reactivity in quadruplicate wells to which btrffer was added in place of antigen dilution. (From Ref. 21.)
high turnover rates, these enzymes are stable, inexpensive, and can make use of a wide range of substrates. We have utilized biotinylated fl-lactamase from B a c i l l u s c e r e u s and a starch-iodine penicillin substrate solution to devise immunoassays for the detection of rotavirus and adenovirus antigens in human body fluids. 21 The sensitivity of these assays was at least as favorable as analogous ones utilizing biotinylated horseradish peroxidase (Fig. 4). In addition, biotin-labeled reagents have been utilized for the detection of a wide range of bacterial, viral, fungal, and parasitic antigens. 2z-28 As:l R. H. Yolken and S.-B. Wee, J. Clin. Microbiol. 19(3), 356 (1984). ~z j. C. Hierholzer, K. H. Johansson, L. J. Anderson, C. J. Tsou, and P. E. Halonen, J. Clin. Microbiol. 25(9), 1662 (1987). 23 S. Edwards, and G. C. Gitao, Vet. Microbiol. 13(2), 135 (1987). ~4 L. C. Shekarchi, D. A. Fuccillo, R. Strouse, and J. L. Sever, J. Clin. MicrobioL 25(2), 320 (1987). 25 V. P. Kurup, Zentralbl. Bakteriol. Mikrobiol. Hgy. Ser. A 261(4), 509 (1986). 26 A. Belmaaza, J. Hamel, S. Mousseau, S. Montplaisir, and B. R. Brodeur, J. Clin. Microbiol. 24(3), 440 (1986). 27 p. Vilja, H. J. Turunen, and P. O. Leinikki, J. Clin. MicrobioL 22(4), 637 (1985). 28 A. Sutton, W. F. Vann, A. B. Karpas, K. E. Stein, and R. Schneerson, J. Immunol. Methods 82, 215 (1985).
536
APPLICATIONS
[61]
says making use of avidin-biotin interactions have also been utilized for the measurement of the immune response to infection with microbial antigens. While the degrees of sensitivity and specificity obtained in such assays varied depending on the binding characteristics of the immunoreagents, the utilization of avidin-biotin interactions has generally resuited in the attainment of the maximal degree of sensitivity allowed by immunoreagents. The use of avidin-biotin interactions has thus allowed for the development of a number of important assays for the diagnosis of infectious diseases. The efficient application of practical assays for microbial detection should allow for the improved clinical management of patients with suspected infections and for the containment of disease transmission in susceptible populations. Preparation of Biotin-Labeled Immunoglobulins Immunoglobulins or immunoglobulin fragments are purified from the serum to be labeled by standard methods and diluted to a concentration of 1 mg/ml in 60 m M carbonate buffer (pH 9.0). Biotin-N-hydroxysuccinimide ester is dissolved in dimethyl sulfoxide to make a stock solution of I mg/ml. Aliquots of biotin are added to the immunoglobulin (generally at a biotin to immunoglobulin molar ratio of 10: 1), and the mixture is incubated for 3 hr at 25 °. After incubation, unreacted biotin is removed by extensive dialysis against PBS or by fractionation with Sephadex G-25. The biotinylated antibody is aliquoted and stored at - 2 0 °. Preservatives such as sodium azide (20 rag/liter) can be added to prevent bacterial contamination. Coat alternate rows of wells of the microtiter plate with a dilution of antiviral IgG and equal dilution of IgG from the same animal species that does not contain measurable antibody to rotavirus. Incubate the plate at least overnight at 4 °. If the plate is not used the next day, it should be covered with Parafilm and stored at 4 ° until used. Wash the plate 5 times with phosphate-buffered saline containing 0.05% Tween 20 (PBS-Tween). Add 50/zl of PBS-Tween to each of the wells. Buffer containing nonimmune animal sera can also be utilized to reduce the occurrence of nonspecific reactions. Add an equal amount of specimen to two wells coated with antiviral IgG and two wells coated with nonimmune IgG. Include a weakly positive control and four negative controls in each test. Incubate the plate for 2 hr at 37° or overnight at 4 °. Add antiviral antibody that has been labeled with biotin, and incubate the plate for 1 hr at 37°. Wash the plate 5 times with PBS-Tween. Prepare the avidin-biotin complex by adding predetermined amounts
[62]
ANTIPOLYSACCHARIDE ANTIBODIES
537
of avidin and biotin-linked enzyme to a tube containing PBS-Tween. Incubate for 10 min at 37 °, and add the complex to the wells. Incubate the plate for 30 min at 37°. Wash the plate 5 times with PBS-Tween. Add appropriate substrate. Incubate the plate at 37 ° or room temperature until the weakly positive control has visible color equivalent to an optical density of approximately 0.1 measured at the appropriate wavelength. Comments. For each sample calculate the virus-specific activity by subtracting the mean activity of the wells coated with the nonimmune IgG from that measured in wells coated with the antiviral IgG. To ensure accurate quantification, specimens giving readings of greater than 1.2 optical density units should be diluted 1 : 10 and retested. Calculate the mean and standard deviation of the virus activity of the negative controls. A specimen is considered positive if its mean activity is greater than 2 standard deviations above the mean activity of the negative controls. Alternatively, a specimen can be considered positive if its specific activity is greater than that of the weakly positive control. If qualitative visual determinations are used, a specimen is considered positive if the degree of color in the antiviral IgG wells is different than the amount of color generated in the nonimmune wells and in the wells containing the weakly positive control samples. Note that if the fl-lactamase starch-iodine system is utilized, a positive reaction will be manifested by a decrease in color, whereas in the case of most other enzymes, the binding of enzyme will result in an increase in color production. Acknowledgment This work was supported by contract N01-AI-52579from the National Institute of Allergies and Infectious Diseases.
[62]
Enzyme-Linked
Immunosorbent
Assay
By WILLIE F. VANN, ANN SUTTON, and RACHEL SCHNEERSON Capsular polysaccharides are important in invasive bacterial disease as virulence factors and as immunogens. Since polysaccharides and oligosaccharides often do not adhere to a plastic solid phase as well as proteins, the measurement of antibodies to polysaccharides by enzymeMETHODS IN ENZYMOLOGY, VOL. 184
Copyright © 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
