January 1, 2018 (Vol. 38, No. 1)
Poulomi Acharya Ph.D. Global Product Manager Bio-Rad Laboratories
Six Best Practices for Dependable Data Quality
One of the most promising chemotherapy drug targets of the past decade was a sham.
Researchers believed that they found human erythropoietin receptor (EpoR) in tumors, and since EpoR normally contributes to the growth and differentiation of red blood cells, the researchers assumed it did the same for tumor cells.1 Based on this finding, researchers thought that they had an opportunity to stop cancer progression in its tracks.
After several preclinical studies2 and unsuccessful clinical trials,3 however, researchers took a closer look, and they found that the antibody originally used to detect EpoR wasn’t binding exclusively to its intended target. The antibody was also binding to a heat-shock protein found in all tissues, including tumor tissue. In truth, EpoR had no functional role in cancer cells whatsoever.4
The cross-reactivity of EpoR antibody is emblematic of the many challenges felt by many researchers who have worked with antibodies. In one survey, almost 50% of researchers reported that the antibodies that they had purchased did not work as advertised.5 Two in three researchers, in a Bio-Rad survey, found that their failed western blot was due to problems with the primary antibody.6
The false results in the EpoR case study could have been prevented if the researchers behind the original work used controls—especially negative controls—to detect nonspecific binding. Validating an antibody’s specificity is just one key best practice that researchers can use to ensure the success of their experiments.
There is, however, no universally accepted standard practice for validating antibodies. Consequently, antibody sellers may not always use best practices, and often, different researchers use different methods, which may yield incomparable results.
Current State of Antibody Validation Practices
One antibody can often have a range of applications in various species and tissues, but antibody manufacturers don’t have the resources to validate every antibody for every possible use. For instance, some antibodies work poorly when applied to fixed cells, whereas others may bind to a region of the target protein that interferes with its mechanism of action.
In most cases, these shortfalls can be managed, provided antibody companies present enough data for buyers to make informed decisions. Antibody companies, however, don’t always provide sufficient information, or worse, they may use improper controls. In most or all cases, vendors tend to show only the positive data, leaving out samples that didn’t show any expression. Furthermore, vendors may crop the blot to show just the region of interest, cutting out potentially valuable data.
The antibody supply chain can be destructive to information, too. Antibody shops don’t always update their labels when they switch manufacturers, leaving researchers unaware that they may be buying an antibody produced with vastly different practices. The new antibody may be diluted differently or have bovine serum albumin added to the solution, which can change how the product behaves. Despite this, according to a survey from the Global Biological Standards Institute (GBSI), only a third of junior researchers validate their antibodies.7
For these reasons, as a rule of thumb, researchers should always test their antibodies as routine procedure.
Validating Antibodies for Your Application
Similar to manufacturers and suppliers, researchers have no standard antibody validation process. Fortunately, organizations such as the GBSI are actively working to create and establish a set of standardized validation practices that can be applied across disciplines. It may be a while, however, before a universal set of validation practices is accepted. So, until then, here are six things researchers can do to ensure their antibodies perform optimally.
1. Optimize Antibody Protocols for Each Application
Researchers should optimize their experimental protocols and antibody dilutions for each new application. It’s especially important to record the final antibody concentration. This information can be calculated based on the stock concentration, which is available from the vendor upon request. A researcher can save time by using a vendor’s experimental protocols as a starting point.
2. Test the Specificity, Sensitivity, and Reproducibility of Each Antibody Used
Next, scientists should test antibodies for specificity and sensitivity, and determine whether the results are reproducible. It’s important to perform these tests in the context of a particular application. For instance, one may use antibodies to detect either native or denatured proteins, or to target a protein in either complex biological samples or purified ones. Each application will require different performance criteria to assess these antibody qualities.
To test an antibody’s specificity, scientists should compare its performance in cell lines that express the target protein versus ones that lack the target protein (generated using gene knockout technology or RNA interference). To measure sensitivity, scientists should use protein-specific index arrays8 containing cell lines that vary in the amount of expressed target protein. Alternatively, scientists should spike a sample free of the protein of interest with known quantities of that protein.
To measure reproducibility for immunohistochemistry, scientists should run their validated antibody on 20 to 40 tissue samples, represented either on a tissue microarray or as a whole-tissue section. For Western blotting, replicates should be run from lysates generated from the same batch of cells. Regardless of application, each experiment should be run three times, using the same antibody lot on different days and using different personnel. Scientists should also compare antibodies between lots to test lot-to-lot reproducibility. Finally, investigators can compare their results to previous data they trust.
3. Use Both Positive and Negative Controls for Each Experiment
Researchers should use both positive and negative controls for every experiment. Knockout or knockdown cells or samples that don’t express the target protein can be used as negative controls. Cells that overexpress the target protein can be used as positive controls.
4. Retest Antibodies before Applying Them to Especially Valuable Samples
Antibodies have limited shelf lives and are often commonly shared. Therefore, researchers should retest their antibodies, especially if the antibodies are to be used for an important experiment. Researchers do not need to perform a full validation protocol; rather, they only need to perform a quick experiment with relevant controls and pre-established conditions to ensure that the antibody’s performance has not changed significantly.
5. Store Antibodies according to Vendor Specifications
Researchers must follow the vendor’s storage specifications. They should write the date of first use on each vial, and if the antibody has expired, they should use it with caution. Researchers should note any changes in storage conditions, as these changes can reduce the product’s shelf life.
Additionally, researchers should not store working dilutions for later use because these dilutions will become less stable. Vendors place stabilizers in the solution, and as the antibody is diluted, so is the stabilizing agent. As stability decreases over time, researchers must repeatedly validate their antibodies and adjust antibody concentrations to maintain consistency.
6. Train New Lab Personnel
As a laboratory acquires new members, it is critical that consistent antibody validation practices remain a staff’s central focus. Each new member must learn the importance of validation, what controls to use for each application, and best practices as previously established by the lab.
1. S. Elliott et al., “Anti-Epo Receptor Antibodies Do Not Predict Epo Receptor Expression,” Blood 107(3), 454 (2006).
2. S.D. Patterson et al., “Functional EPOR Pathway Utilization Is Not Detected in Primary Tumor Cells Isolated from Human Breast, Non-Small Cell Lung, Colorectal, and Ovarian Tumor Tissues,” PLOS One 10, e0122149 (2015).
3. Z.H. Endre et al., “Early Intervention with Erythropoietin Does Not Affect the Outcome of Acute Kidney Injury (The EARLYARF Trial),” Kidney Int. 77(11), 1020–1030. doi: 10.1038/ki.2010.25 (2006).
4. S. Elliott et al., “Lack of Expression and Function of Erythropoietin Receptors in the Kidney,” Nephrol. Dial. Transplant. 27, 2733–2745, doi: 10.1093/ndt/gfr698 (2012).
5. A. Hodgson, “Guest Post: Finding the Perfect Antibody,” Science Exchange, blog.scienceexchange.com/2012/03/guest-post-alexandra-hodgson/, accessed December 6, 2017.
6. R. Bogoev, “The Antibody Challenge: Bio-Rad’s Precise Solution,” Bioradiations, accessed December 6, 2017.
7. L.P. Freedman et al., “Letter to the Editor: The Need for Improved Education and Training in Research Antibody Usage and Validation Practices,” Biotechniques 61, 16–18, doi: 10.2144/000114431 (2015).
8. A.W. Welsh et al., “Standardization of Estrogen Receptor Measurement in Breast Cancer Suggests False-Negative Results are a Function of Threshold Intensity Rather than Percentage of Positive Cells,” J. Clin. Oncol. 29(22), 2978–2984 (2011).