Enzyme Immunoassays: The Details
Enzyme immunoassays (EIAs), also known as enzyme-linked immunosorbent assays (ELISAs), combine antibody binding with enzymatic detection to quantify molecules of interest. EIAs are easy to perform, require little specialized equipment, and both the experienced lab technician and the research lab novice can learn EIA skills quickly. However, a better understanding of the elements and design of EIAs is useful. This article delves into some of the specific details, as well as describes some properties unique to Cayman's EIAs.
Many commercial EIAs use either horseradish peroxidase (HRP) or alkaline phosphatase (AP) as the enzyme that drives color generation. Cayman's ACE™ EIAs use AChE, an enzyme derived from the electric organ of the electric eel, Electrophorus electricus. Acetylcholinesterase has an extremely high turnover number (64,000 per second), which allows for rapid color development in an immunoassay. Unlike HRP, AChE does not suicide inactivate, allowing assays to be redeveloped if they are accidently splashed or spilled. AChE is highly stable under assay conditions, is active over a wide pH range, and tolerates both phosphate and azide.
Competitive EIAs are most commonly used to measure small molecules including lipids, hormones, and small peptides, although larger molecules can also be measured using this format if they are present in high enough concentrations. This type of assay is based on the competition between the analyte of interest and an enzyme-conjugated version of the same analyte (referred to as the tracer) for a limited number of specific antibody binding sites (Figure 1). The concentration of tracer is held constant in all wells while the concentration of analyte varies from well-to-well. As a result, the amount of tracer that can bind to the antibody will be inversely proportional to the amount of analyte in the well – the presence of more analyte means less tracer will be able to bind to the specific antibody. The antibody-analyte (either free or tracer) complex is immobilized by binding to an anti-species IgG antibody coated to the wells of the plate. After sufficient equilibration time, unbound reagents are simply washed away. The developing solution which contains the substrate for AChE and Ellman's reagent is then added to the wells. The reaction product has a distinct yellow color which absorbs strongly at 412 nm.
Figure 1. Illustration of a competitive EIA without (above) or with competition (below)
When performing a competitive EIA, there are several types of wells that must be run on each plate in addition to the wells containing the samples. These types of wells, what they contain, and the information they provide is summarized in the table below.
While it is possible to plot standard curves for a competitive EIA as the concentration of standard (x-axis) versus absorbance (y-axis), it is common practice to instead plot the y-axis value as percent of maximal binding (%B/B0) (Figure 2). This value is obtained by dividing the absorbance for each well by the absorbance of the maximum binding well (B0) and multiplying by 100. This method offers the ability to easily compare results between assays performed on two different plates or two different days. While the absolute absorbance may differ from plate to plate or day to day, the %B/B0 values should be reasonably consistent from one plate to the next.
Figure 2. Typical standard curve for Cayman's PGE2 EIA Kit (Catalog No. 514010)
There are two curve fits that are routinely used for analysis of competitive assays, the four-parameter logistic fit and the log-logit fit. Both curve fits express the standard concentrations in logarithmic fashion on the x-axis. The four-parameter logistic fit uses %B/B0 on the y-axis; analysis of standard curve data using this curve fit results in a sigmoidal-shaped standard curve. The log-logit fit uses logit B/B0 on the y-axis and can be fit with a straight line. The log-logit fit is easy to perform and data can be analyzed in any spreadsheet program. Data analysis tools using the log-logit fit are available on Cayman's website. While the log-logit fit is very convenient, the line fit is sometimes hard to interpret and it is difficult to know where the highest variation in the assay occurs. The sigmoidal curve fit, on the other hand, gives a clear graphical representation of the data that is easily interpretable in terms of where samples can be most accurately analyzed (the linear center portion of the standard curve which is typically between 20-80% B/B0) and when sample concentrations should be evaluated with more caution (at the ends of the standard curve where it becomes non-linear). In order to provide as much information as possible about the performance of our assays, Cayman determines both the intra-assay and inter-assay Coefficient of Variation (%CV) at all concentrations of the standard curve. For easy reference, the region along the standard curve where the data can be evaluated with confidence is illustrated in blue. %CVs are typically lowest in the center of the curve and increase at the ends of the curve.
The limit of detection (or sensitivity) of competitive immunoassays can be defined in different ways. While some companies define the sensitivity of their assays using a mathematical formula to estimate a concentration of standard that would give an absorbance greater than that of the NSB, Cayman takes a more conservative approach. We believe that it is important to define sensitivity based on the concentration of analyte where a customer would be truly able to quantify this concentration using the standard curve. Since the majority of competitive immunoassays become non-linear around 80% B/B0, Cayman defines sensitivity as the concentration of analyte which would result in an absorbance of 80% B/B0. While this approach often makes Cayman's assays appear on paper to be less sensitive than those of the competition, we feel that it gives customers a more realistic estimation of quantities of analyte that they will be able to accurately measure with our kits.
The second type of solid phase immunoassay, and the most familiar to most researchers, is the immunometric or "sandwich" assay. As the name implies, this assay involves two antibodies which "sandwich" the analyte between them. Because the analyte must be relatively large to allow simultaneous binding of two antibodies, this type of assay is suitable only for proteins and peptides greater than 20 amino acids in length. In most immunometric assays, plates are coated with a capture antibody that is specific to the analyte. This antibody will bind to the analyte present in a sample or standard. A second antibody, called the detection antibody, which recognizes a different epitope on the analyte is also added to the well, resulting in the analyte being "sandwiched" between the two antibodies. This reaction is allowed to come to equilibrium and excess unbound reagents are then washed away. The presence of the detection antibody, which is directly proportional to the amount of analyte present, can be quantified by a few different methods. If the detection antibody was raised in a different species than the coating antibody, it is possible to use an anti-IgG antibody which was raised against IgG of the species that the detection antibody was raised in. If the coating antibody and detection antibody were raised in the same species, then the detection antibody must be tagged in some way. The two most common methods for tagging are biotinylation or direct conjugation of an enzyme to the detection antibody. The biotinylated detection antibody can be measured using a variety of streptavidin-based techniques. Cayman uses both of these approaches for our immunometric assays.
Standard curves for immunometric assays are plotted as the concentration of standard (x-axis) versus absorbance (y-axis) (Figure 3). Both axes are usually plotted on a linear scale. Data can sometimes be fit with a straight line, but are generally better fit using a quadratic equation. Sensitivity of Cayman's immunometric assays is defined as the point where the signal obtained with analyte is two times the signal of the non-specific binding.
Figure 3. Typical standard curve for Cayman's IL-1β (human) EIA Kit (Catalog No. 583311)
Regardless of the type of immunoassay, there are common aspects in designing and performing an EIA that will ensure the user gets the most from the assay. First, all immunoassays should contain blank wells. Blank wells receive only developing solution and measure the very low absorbance of that solution. Most plate readers will automatically subtract the blank values from the absorbance values obtained for all other wells on the plate. Second, an appropriate set of standards (as well as NSB wells and B0 wells if appropriate) should be run on each plate. One curve is not sufficient for use with samples on multiple assay plates, even if the plates are run at the same time. Third, all samples should be run at least in duplicate and preferably in triplicate. Additionally, samples should be run at a minimum of two dilutions that fall on the standard curve. The two dilutions should show good correlation in the final calculated concentration of analyte. If they do not show good correlation, it is an indication that something in the samples is interfering with the assay and purification of samples is recommended. Finally, accurate pipetting of small volumes is crucial to obtaining useful data. Cayman sells a practice kit (Catalog No. 10009658) that can used to become familiar with performing a typical competitive EIA before analyzing valuable samples.
|Non-Specific Binding (NSB)||Tracer and buffer||How much tracer binds to the wells of the plate in the absence of any specific antibody|
|Maximum Binding (B0)||Tracer, antibody, and buffer||Maximum signal obtainable by the antibody and tracer pair in the absence of any analyte|
|Standard Curve||Tracer, antibody, and a series of known amounts of analyte||An accurate reference across a broad analyte concentration range for determining analyte levels in the sample|
|Total Activity (TA)||Small amount of tracer (added at the development stage)||Enzyme is active|