A microarray is an assay that enables the simultaneous detection of multiple molecules, such as nucleic acids, peptides, proteins, or antibodies. They are called “micro”-arrays because the analytes are measured within a small surface area; for example, thousands of molecules can be analyzed on a standard glass slide (75 mm x 25 mm). Such tools are invaluable in research and in the clinic because the data represent a large snapshot of biological processes occurring at one point in time. One type of microarray used to measure proteins is the antibody array. The antibody array provides a more affordable and cost-effective alternative with minimum sample volume requirements compared to “single-plex” assays that analyze only one protein at a time, such as enzyme-linked immunosorbent assays (ELISAs). In this blog, the different formats of antibody arrays for multiplex protein detection will be described. Choosing which format is appropriate for a specific research application is also discussed.
How Antibody Arrays Work
Antibody arrays immobilize a capture antibody onto a solid substrate, such as a glass slide, membrane, or microbead. For planar surface arrays, different capture antibodies with known binding specificities are spotted onto a standard microscope slide or a nitrocellulose membrane in an addressable format (Figure 1A). For bead-based arrays, capture antibodies are immobilized onto different beads with varying sizes and fluorescent properties. Since each specific antibody is immobilized onto a bead with unique characteristics, the protein target of each bead is known.
Prior to incubating a sample with the array, a blocking step is normally performed to inhibit nonspecific interactions that can result in high background and inaccurate data. The blocking solution is protein-based, such as bovine serum albumin (BSA) or non-fat milk, with a small amount of detergent (e.g., 1% Tween-20).
During sample incubation on the array, the capture antibodies bind to and immobilize their specific target proteins to the array. Unbound proteins are removed with a washing step.
With label-based antibody arrays, the sample proteins are labeled with biotin prior to incubating the sample on the array (Figure 2A). With sandwich-based arrays, a mixture of biotinylated detection antibodies is added to the array following sample incubation, thus “sandwiching” the protein-of-interest between two antibodies (Figure 2B). For both design principles, the sensitivity and specificity are dependent on the antibodies that are employed.
It is also important to point out that molecules other than the target protein or the detection antibody can be labeled instead. For example, a glycosylated protein could be sandwiched between the capture antibody and a biotinylated lectin (Figure 2C).
A streptavidin molecule conjugated to a fluorophore or horseradish peroxidase (HRP) is added to the array where it binds to the biotin moiety on the protein or detection antibody (Figure 1B). The streptavidin complex, via chemiluminescence or fluorescence, enables the detection of the immobilized proteins. Since the target protein of the spot or bead is known, the identity of the captured proteins on the array can be determined.
Chemiluminescence occurs when an HRP substrate is added to an HRP-streptavidin complex. Chemiluminescence is measured using a chemiluminescent blot documentation system, such as a CCD camera, X-ray film and a suitable film processor, gel documentation system, or other chemiluminescent detection system capable of imaging a western blot.
Fluorescence is produced by the fluorophore conjugated to the streptavidin molecule. Fluorescence is measured using a compatible laser scanner for glass slides or a flow cytometer for bead-based arrays.
Qualitative, Semi-Quantitative, & Quantitative Data
Antibody arrays can generate qualitative, semi-quantitative, or quantitative data for multiplex protein detection (Figure 3). An example of qualitative data would be the intensity of bands of a western blot that are assessed only by eye. However, accurate estimates of spot intensities are difficult to make with qualitative assessments; as such, semi-quantitative and quantitative data are preferred.
Semi-quantitative and quantitative data have fluorescent or chemiluminescent outputs that are assigned values; for example, pixel intensity. The primary difference between semi-quantitative and quantitative data is that quantitative data have a standard curve against which the sample data can be compared. Relative expression differences (i.e., fold changes) across samples can be obtained with semi-quantitative data. Exact protein concentrations are obtained with quantitative data.
The amount of sample required for analysis depends on the sample dilution, substrate (i.e., membrane, glass, bead), and design principle (i.e., label-based, sandwich-based).
Samples should be diluted at least 2-fold to avoid a phenomenon called “sample matrix effects” (SMEs), which are a common cause of non-linear dilution responses and inaccurate data. This can occur when proteins or other components within the sample affect the ability of an antibody to bind to its target molecule. For example, a protein may bind to the target molecule, thus blocking the antibody’s binding site (Figure 4A). SMEs are the result of known and unknown causes. However, matrix effects can become negligible with the appropriate sample dilution (Figure 4B). The optimal dilution will vary based on sample type, experiment, and target molecule; thus, it is important to optimize the experimental conditions with a few samples prior to running the entire experiment. One sample type that may not need dilution is urine because it contains very low protein content.
For samples other than serum and plasma, the original total protein concentration prior to sample dilution should be at least 1 mg/ml. However, total protein concentrations above 2 mg/ml are recommended to improve signal. These concentrations exclude any serum (e.g., fetal bovine serum) that may be used with conditioned media samples. Importantly, high signal-to-noise ratios on the array may still be achieved with conditioned media samples with concentrations lower than 1 mg/ml because the supernatant has a lower protein content than cells and tissues. The range of protein concentrations in serum and plasma is narrow across individuals, with an average total protein concentration of ~70 mg/ml.
Different cells and tissues contain different amounts of protein. As such, the amount of cells, tissue and sample to load onto the array must be empirically determined. With that said, good starting points are: 500 µL of lysis buffer per 1 million cells or 10 mg tissue, and seeding ~1 million cells in 100 mm tissue culture plate and culture for 5 days for culture media samples.
Substrate & Design Principle
Membrane-based arrays, due to their large surface area, require more sample than glass- and bead-based arrays. However, membrane-based arrays are still a popular option for multiplex protein detection because they are easy to handle, their chemiluminescence can be detected with a common and inexpensive instrument, and their background noise is typically lower than glass-based arrays for cell and tissue lysates.
With label-based arrays, each sample protein can bind to numerous biotin molecules, thus amplifying the signal and increasing sensitivity. Because of this, very low sample volumes are required with label-based arrays. A comparison of volume requirements for a serum sample based on array substrate and design principle is provided in Table 1. The listed volumes do not account for pipetting error or sample loss on the side of the tubes.
Table 1. Serum Volume Requirements based on Solid Support and Design Principle for Antibody Arrays
Label-based versus Sandwich-based Antibody Arrays
Label-based antibody arrays use only one antibody per protein: the capture antibody (Figure 2A). Because of this, very high-density arrays are possible due to the large number of available antibodies. Since antibodies can crossreact with non-target proteins, any results obtained with label-based arrays should be validated using an orthogonal platform, such as western blots. RayBiotech offers high density label-based arrays (L-Series) for multiplex protein detection that can measure as many as 2,000 human proteins; 1308 murine proteins; 1500 rat proteins; or 500 rabbit proteins simultaneously.
Label-based array summary
- Highest density
- Lower specificity than sandwich-based arrays
- Very low sample volume required
- Semi-quantitative data
- Solid support: glass & membrane
- Excellent option for biomarker discovery
Sandwich-based arrays have very high specificity since two different antibodies per protein are required for the immobilized protein to be detected (Figure 2B). However, finding two antibodies that can be paired in a sandwich immunoassay may require extensive antibody screening. The antibodies not only have to bind to separate regions (or epitopes) on the same protein, but the epitopes must be accessible for binding on the platform. In some instances, two antibodies may not work as a capture-detection pair, but rather in a detection-capture configuration.
RayBiotech offers sandwich-based arrays on membrane (C-Series), bead (RayPlex), and glass (G-Series, Phosphorylation Arrays, Quantibody) substrates. These arrays can measure as many as 1,000 human proteins; 640 murine proteins; or 67 rat proteins simultaneously. Other species detected with these arrays include bovine, canine, feline, equine, porcine, rabbit, rhesus monkey, chicken, dolphin, and ovine. Both RayPlex and Quantibody arrays provide quantitative data.
Sandwich-based array summary
- Low to high density
- Highest specificity: Two antibodies (i.e., antibody pair) per target
- Semi-quantitative and quantitative data
- Solid support: glass, membrane, & bead
- Excellent option for clinical trials and biomarker validation
A comparison of label-based antibody arrays and sandwich-based arrays offered by RayBiotech is provided in Table 2. A decision tree to help decide between qualitative, semi-quantitative, or quantitative data for multiplex protein detection is depicted in Figure 5.
Learn how label- and sandwich-based antibody arrays are used in biomarker research in our blog, “A Comparison of Antibody Arrays and Mass Spectrometry in Protein Profiling and Biomarker Research.”
Table 2. A Comparison of Label-based and Sandwich-based Antibody Arrays
The antibody array is a powerful technique for protein profiling and multiplex protein detection. With different types of formats and data outputs, there is an antibody array to meet different research needs and capabilities. High density arrays that analyze hundreds to thousands of proteins simultaneously are ideal for large-scale protein profiling and biomarker discovery studies. There are also smaller panels focused on specific biological processes or pathways, such as inflammation, growth factors, angiogenesis, and obesity. Notably, the panels for antibody arrays can be customized, and full testing services are available for laboratories with limited personnel, time, or instrumentation.
Need help picking the right array? Please contact us at [email protected].