Due to the amplifying potential of enzyme labels, immunoassays that use enzyme-conjugated antibodies have become increasingly popular because of their high specificity and sensitivity. 1 In 1971, Engvall and Perlmann 2 coined the term "enzyme-linked immunosorbent assay" which is perhaps better known by the acronym, "ELISA", to describe an enzyme-based immunoassay method which is useful for measuring antigen concentrations.
Cytokine sandwich ELISA are sensitive enzyme immunoassays that can specifically detect and quantitate the concentration of soluble cytokine and chemokine proteins. The basic cytokine sandwich ELISA method makes use of highly-purified anti-cytokine antibodies (capture antibodies) which are noncovalently adsorbed ("coated"—primarily as a result of hydrophobic interactions) onto plastic microwell plates. After plate washings, the immobilized antibodies serve to specifically capture soluble cytokine proteins present in samples which were applied to the plate. After washing away unbound material, the captured cytokine proteins are detected by biotin-conjugated anti-cytokine antibodies (detection antibodies) followed by an enzyme-labeled avidin or streptavidin stage. Following the addition of a chromogenic substrate, the level of colored product generated by the bound, enzyme-linked detection reagents can be conveniently measured spectrophotometrically using an ELISA-plate reader at an appropriate optical density (OD). Data storage and reanalysis are greatly simplified when the plate reader is connected to a computer.
A standard curve is incorporated into a sandwich ELISA assay by making serial dilutions of a standard cytokine protein solution of known concentration. Standard curves (aka "calibration curves") are generally plotted as the standard cytokine protein concentration (typically ng or pg of cytokine/ml) versus the corresponding mean OD value of replicates. The concentrations of the putative cytokine-containing samples can be interpolated from the standard curve. This process is made easier by using an ELISA computer software program. 3 Generally, it is useful to perform a dilution series of the unknown samples to be assured that the OD will fall within the linear portion of the standard curve. Depending on the nature of the ELISA reagents used, investigators may choose to apply different curve fit analysis to their data, including either linear-log, log-log, or four-parameter transformations. 1,4,5
Although opinions differ, one convention for determining the ELISA sensitivity is to choose the lowest cytokine concentration that gives a signal which is at least two or three standard deviations above the mean background signal value. 6,7 Because of the enzyme-mediated amplification of the detection antibody signal, the sandwich ELISA can measure physiologically relevant (ie, > 5-10 pg/ml) concentrations of specific cytokine and chemokine proteins, which are present in mixed cytokine milieus, e.g., from stimulated lymphocyte culture supernatants. Although many different types of enzymes have been used, horseradish peroxidase (HRP) and alkaline phosphatase (AKP) are the enzymes that are often employed in ELISA methods. 1,8
Cytokine sandwich ELISA are exquisitely specific because antibodies directed against two or more distinct epitopes are required. 9 Therefore, sandwich ELISA can discriminate between cytokines that can have overlapping biological functions which are not resolvable in a bioassay. Although cytokine sandwich ELISA are very useful for cytokine detection and measurement, several limitations for the interpretation of ELISA data must be mentioned. 9 For example, because test samples often come from tissue culture supernatants or biological fluids which are conditioned with cytokines produced by mixed cell populations, the ELISA data does not provide direct information on the identities and frequencies of individual cytokine producing cells. Techniques such as the "Immunofluorescent Staining of Intracellular Cytokines" are required for this latter type of analysis.
Several key issues need to be considered when designing experiments that involve cytokine and chemokine protein measurements using sandwich ELISA. For instance, it is well known that cytokine production by stimulated cell populations is transient and that the kinetics of expression of different cytokine genes can vary. For these reasons, it may therefore be necessary to collect test samples at several time points to better characterize cytokine-production by an experimental animal or by a cultured cell population. As an example, in the case of stimulated mouse CD4 +T cell populations, the levels of IL-2 produced are detected relatively early after stimulation whereas the accumulated levels of IL-5 protein rise later in culture. 10 It should also be noted that cytokine production can be stimulus- and cell subset-dependent. For example, in the case of T cells, it is well known that naive T cells have a limited cytokine production capability (ie, primarily can produce IL-2) whereas memory T cells can produce high levels and different types of cytokine proteins including IFN-gamma and IL-4, as well as IL-2. 11,12 Moreover, T cell subsets have been found to produce cytokines differentially in response to different stimuli. 12,13 Another consideration is that cytokine protein concentrations, measured at any one time point, may reflect the concurrent processes of cytokine secretion, cytokine uptake by cells and cytokine protein degradation. Because of these processes, the measured level of cytokine protein may significantly underestimate the actual cytokine-producing potential of cells. In these cases, it may be necessary to use complementary techniques such as multi-probe ribonuclease protection assay analysis, immunofluorescent intracellular cytokine staining with flow cytometric analysis, or ELISPOT, to gauge the relative levels of cytokine expression by various test cell populations.
The levels of immunoreactive cytokine proteins detected by ELISA may or may not correlate directly with the levels of bioactive cytokine protein. 9,14 For example, an ELISA may utilize anti-cytokine antibodies that cannot discriminate between the precursor (inactive) and mature (bioactive) forms of a cytokine protein such as TGFb1. Moreover, an ELISA may detect partially-degraded cytokine proteins which have retained their immunoreactive properties (ie, at least two recognizable epitopes) but may have lost their bioactivity. In conclusion, cytokine sandwich ELISA are useful indicators of the presence and levels of cytokine and chemokine proteins but they do not actually provide information concerning the biological potency of the detected proteins.
With these caveats in mind, one can infer from the presence and amount of cytokine protein detected the potential mechanisms by which particular effector cell populations perform their functions. Moreover, sandwich ELISAs can detect soluble cytokine receptors which may be important for cytokine regulation. Soluble cytokine receptors may act as antagonists or as carrier proteins in vivo and may serve as disease markers in in vitro tests. 15 It should be noted that in addition to providing a rich source of information for clinical and basic science research studies, sandwich ELISA for measuring cytokines and their receptors have become increasingly important as diagnostic tools and for monitoring therapeutic regimens, 16 e.g., biological response modification regimens utilizing recombinant cytokine proteins. In the latter cases, highly optimized sandwich ELISA kits designed to minimize interference or nonspecific reactivities presented by patient samples is highly desirable.
ELISA Protocol General Procedure
- Dilute the purified anti-cytokine capture antibody to 1-4 µg/ml a in Binding Solution. Add 100 µl of diluted antibody to the wells of an enhanced protein-binding ELISA plate (e.g., Falcon cat. no. 353279 or Nunc Maxisorb cat. no. 446469).
- Seal plate to prevent evaporation. Incubate overnight at 4°C.
- Bring the plate to room temperature (RT), remove the capture antibody solution, and block non-specific binding by adding 200 µl of Blocking Buffer per well.
- Seal plate and incubate at RT for 1-2 hr.
- Wash ≥ 3 times d with PBS/Tween®.
Standards and Samples
- Add standards c and samples (diluted in Blocking Buffer/Tween® h) at 100 µl per well.
- Seal the plate and incubate it for 2-4 hrs at RT or overnight at 4°C. e
- Wash ≥ 4 times with PBS/Tween®.
- Dilute the biotinylated anti-cytokine detection antibody to 0.5-2 µg/ml in Blocking Buffer/Tween®. h Add 100 µl of diluted antibody to each well.
- Seal the plate and incubate it for 1 hr at RT.
- Wash ≥ 4 times d with PBS/Tween®.
Avidin-Horseradish Peroxidase (Av-HRP):
- Dilute the Av-HRP conjugate (cat. no. 554058) or streptavidin-HRP (cat. no. 554066) or other enzyme conjugate c,f to its pre-titered optimal concentration in Blocking Buffer/Tween®. Add 100 µl per well.
- Seal the plate and incubate it at RT for 30 min.
- Wash ≥ 5 times d with PBS/Tween®.
- Use TMB (cat. no. 555214) according to directions or ABTS as a substrate. Thaw ABTS Substrate Solution c,f within 20 min of use. Add 100 µl of 3% H 2O 2 per 11 ml of substrate and vortex. Immediately dispense 100 µl into each well. Incubate at RT (5-80 min) for color development.
- Read the optical density (OD) for each well with a microplate reader set to 405 nm. g
Binding Solution: 0.1 M Na 2HPO 4, adjust to pH 9.0 or to pH 6.0 with 0.1 M NaH 2PO 4 (note: use pH 6.0 Binding Solution for mouse IL-10, mouse MCP-1, mouse TNF, rat GM-CSF ELISAs).
PBS Solution: 80.0 g NaCl, 11.6 g Na 2HPO 4, 2.0 g KH 2PO 4, 2.0 g KCl; q.s. to 10 L; pH to 7.0
PBS/Tween®: 0.5 ml of Tween®-20 in 1 L PBS.
Blocking Buffer: Prepare 10% fetal bovine serum (FBS), 10% newborn calf serum (NBCS) or 1% BSA (immunoassay grade) in PBS. The Blocking Buffer should be filtered to remove particulates before use.
Blocking Buffer/Tween®: Add 0.5 ml Tween®-20 to 1 L Blocking Buffer.
TMB Substrate Solution (cat. no. 555214): Prepare a working concentration of TMB substrate solution within 15 minutes prior to use by mixing equal volumes of Substrate Reagent A and Substrate Reagent B (e.g. for one 96-well plate, a 12 mL TMB substrate working solution can be prepared by mixing 6 mL of Substrate A with 6 mL of Substrate Reagent B).
ABTS Substrate Solution: Add 150 mg 2,2'-Azino-bis- (3-ethylbenzthiazoline-6-sulfonic acid) (e.g., Sigma, Cat. No. A-1888) to 500 ml of 0.1 M anhydrous citric acid (e.g., Fisher; Cat. No. A-940) in dd H 20; pH to 4.35 with NaOH. Aliquot 11 ml per vial and store at -20°C. Add 100 µl 3% H 2O 2 prior to use.
3% H 2O 2 Solution: Add 10 ml of 30% H 2O 2 to 90 ml of ddH 2O. Protect from prolonged exposure to light.
Cytokine ELISA Helpful Hints
- To determine the optimal signal and lowest background for the ELISA, the capture antibody (1-4 µg/ml) and detection antibody (0.25-2 µg/ml) should be titrated against each other in a preliminary experiment. An appropriate range of serial dilutions for the cytokine standard should be included. A suggested range is generally provided on the Technical Data Sheet (TDS) for ELISA reagents. Generally, use of the capture antibody at 2 µg/ml and the detecting antibody at 1 µg/ml provides strong ELISA signals with low back-ground.
- CYTOKINE STANDARD HANDLING: Please read the TDS for each recombinant cytokine carefully. Handling instructions are lot-specific. For maximum recovery of cytokine, the vial of cytokine should be quick-spun before opening. Lyophilized cytokines should be reconstituted as indicated in the lot-specific TDS. It is recommended to keep the cytokine solution in a concentrated form (e.g., ≥ 1 µg/ml) and in the presence of a protein carrier for long-term storage.
- The linear region of cytokine ELISA standard curves are generally obtainable in a series of eight two-fold dilutions of the cytokine standard, from 2000 pg/ml to 15 pg/ml. To increase sensitivity beyond that obtainable with the standard ELISA protocol, amplification kits, tertiary reagents, or alternate enzyme/substrate systems can be used.
- High backgrounds in blank wells (i.e., OD > 0.20) or poor consistency of replicates can be overcome by increasing the stringency of washes and optimizing the concentration of capture and detection antibodies. For example, during washes, the wells can be soaked for ~ 1 minute intervals. Moreover, lower concentrations of detecting antibody or more washes after the detecting antibody stage can reduce background.
- For optimal sensitivity, overnight incubation of standards and samples is recommended.
- If using peroxidase as the enzyme for color development, avoid sodium azide in wash buffers and diluents, as this is an inhibitor of peroxidase activity.
- If no signal is observed, check the following: a) verify that appropriate antibody clones were used; b) check the activity of the enzyme/substrate system: e.g., coat 1 µg/ml of biotinylated detecting antibody in several wells in binding buffer for a few hr. After blocking, wash several times then proceed with the cytokine ELISA protocol from Step 13. If the enzyme/substrate system is active, then a strong signal should be seen; c) verify the activity of cytokine standard or try a new sample of standard.
- When measuring cytokines in complex fluids, such as serum, sample diluents which include irrelevant Ig are suggested. 17
Crowther, J. R. 1995. ELISA. Theory and Practice. Methods Mol. Biol. 42:1-223.
Engvall, E., and P. Perlmann. 1971. Enzyme-linked immunosorbent assay (ELISA). Quantitative assay of immunoglobulin G. I mmunochem. 8:871-874.
Davies, C. 1994. Principles. In The Immunoassay Handbook. D. Wild, ed. Stockton Press, New York, p. 3-47.
Rogers, R. P. C. 1984. Data Analysis and Quality Control of Assays: A Practical Primer. In Practical Immuno Assay. W. R. Butt, ed. Marcel Dekker, Inc., New York.
Davies, C. 1994. Calibration curve fitting. In The Immunoassay Handbook. D. Wild, ed, New York, p. 118-123.
Davies, C. 1994. Concepts. In The Immunoassay Handbook. D. Wild, ed. Stockton Press, New York, p. 83-115.
Pathak, S. S., A. van Oudenaren, and H. F. J. Savelkoul. 1997. Quantification of immunoglobulin concentration by ELISA. In Immunology Methods Manual, vol. 2. I. Lefkovitz, ed. Academic Press, Inc., San Diego, p. 1056-1075.
Wild, D., and C. Davies. 1994. Components. In The Immunoassay Handbook. D. Wild, ed. Stockton Press, New York, p. 49-82.
Mosmann, T. R., and T. A. T. Fong. 1989. Specific assays for cytokine production by T cells. J. Immunol. Meth. 116:151-158.
Hobbs, M. V., W. O. Weigle, D. J. Noonan, B. E. Torbett, R. J. McEvilly, R. J. Koch, G. J. Cardenas, and D. N. Ernst. 1993. Patterns of cytokine gene expression by CD4 + T cells from young and old mice. J. Immunol. 150:3602-3614.
Ehlers, S., and K. A. Smith. 1991. Differentiation of T cell lymphokine gene expression: The in vitro acquisition of T cell memory. J. Exp. Med. 173:25-36.
Cerottini, J.C., and H. R. MacDonald. 1989. The cellular basis of T-cell memory. Annu. Rev. Immunol. 7:77-89.
Farber, D. L., M. Luqman, O. Acuto, and K. Bottomly. 1995. Control of memory CD4 T cell activation: MHC class II molecules on APCs and CD4 ligation inhibit memory but not naive CD4 T cells. Immunity 2:249-259.
Carter, L. L., and S. L. Swain. 1997. Single cell analyses of cytokine production. Curr. Opin. Immunol. 9:177-182.
Callard, R. E., and A. J. H. Gearing. 1994. Cytokine receptor superfamilies. In The Cytokine Facts Book. Academic Press Inc., San Diego, p. 18-27.
Rossio, J. L. 1997. Cytokines and immune cell products. In Weir's Handbook of Experimental Immunology. Fifth Edition. D. M. Weir, L. A. Herzenberg, L. A. Herzenberg, and C. Blackwell, eds. Blackwell Science, Inc., Cambridge, MA.
Abrams, J.S. 1995. Immunoenzymetric assay of mouse and human cytokines using NIP-labeled anti-cytokine antibodies. Current Protocols in Immunology (J. Coligan, A. Kruisbeek, D. Margulies, E. Shevach, W. Strober, eds). John Wiley and Sons, New York. Unit 6.20.