Designing a Foolproof Quantitative Western Blot Experiment
Complete the form below to unlock access to ALL audio articles.
Although western blotting is a well-established laboratory technique, it has recently come under fire as a quantitative method because extreme care must be taken when generating and interpreting the resulting data.
A recent methods article published in BioMed Research International and co-authored by Bio-Rad scientists, provides a rigorous and concise workflow with specific instructions on how to produce and analyze quantitative data using western blot experiments.
We spoke to one of the papers co-authors, Dr. Sean Taylor, Field Application Scientist at Bio-Rad, to understand more about the problems of reproducibility and the how the proposed techniques overcome them.
AB: How much of a problem is poor reproducibility in quantitative Western Blot experiments?
Sean Taylor (ST): There are now published commentaries [1-4], blog posts [Retraction Watch and Research Gate] and even an organization (Office of Research Integrity (ORI)) dedicated to assessing the quality of published data from western blots. Some papers contain data that has clearly been manipulated resulting in retraction but there are many articles that have been published with overloaded protein giving potentially misleading or false conclusions. These scientists have not purposely published artifactual data but have simply designed their western blot experiments as they have been taught by those before them. These methods that have been passed down from generation to generation of scientists come from the early days of the technique when western blotting was only meant to detect the presence or absence of a target protein in a sample. The same techniques are still being applied by many labs attempting to produce quantitative western blot data without adapting the appropriate procedures and validation steps required to obtain reproducible and quantitative results.
AB: Are these problems limited to reproducing work from journals or does this also apply to repeating protocols within the same laboratories?
ST: These problems apply both to reproducing previously published data and between repeats of the same experiment within a lab because the historically applied western blotting techniques from decades ago are simply not quantitative. It would be like asking two people with basic cooking skills to prepare a spaghetti sauce from the same ingredients but with no recipe and no measuring spoons with the expectation that the final product will taste the same. Even if one person attempted this task, it would be a major challenge without some guidelines to follow. This is not to say that the way people perform western blotting involves a lot of guesswork (although many frustrated scientists would say this is the case), it is simply that there has been no rigorous methodology released or published for researchers attempting to produce quantitative data from the technique. This has resulted in a widely varying number of approaches to designing western blotting experiments and analyzing the associated data as opposed to using an accepted norm that has been vetted and accepted by the scientific community at large.
AB: What do you believe are the major contributing factors to poor reproducibility?
ST: From an experimental design perspective, the two major contributing factors to poor reproducibility of western blot data are: 1) There is no standard curve generated to measure the protein of interest. Without this reference, it is not possible to assure a quantitative measurement in the linear dynamic range of detection and 2) The procedure is long and has many steps that are applied with various emphasis and techniques between labs and even between individuals within a given lab making it difficult or impossible to reproduce the data generated between experiments.
As with RT-qPCR, the principal overriding reason for poor reproducibility in western blotting is a term we have coined "historical biase" where scientists performing quantitative western blotting today have learned from those before them. This method of passing information from generation to generation is typically an excellent way to learn basic techniques that form a strong foundation of knowledge on which we can build to do research. However, when these basic techniques from the past are accepted as the gold standard and then applied to more recent or advanced equipment, reagents and data analysis methods without understanding the applicability and potential ramifications, it could lead to unexpected results. It is critical to use the appropriate image capture technique which for chemiluminescence is a camera-based imaging system such as the ChemiDoc MP. Film is definitely not recommended and we demonstrate the difference in data quality between film and camera-based detection in an article we published last year in the peer reviewed journal Molecular Biotechnology .
Also, many researchers randomly load a fixed amount of protein per lane on their gels for western blotting that typically falls between 10ug and 80ug which is in saturating amounts for typical housekeeping protein loading controls leading to improper and non-quantitative data normalization between samples. It may also result in the saturation of the target protein on the blot which would give the perception that there is no change in protein expression between the samples as an artifact of membrane saturation. All of this is explained in our recent article with detailed procedures to avoid this problem using Stain Free gel technology.
AB: How do your techniques help overcome these common pitfalls?
ST: The techniques summarized in the paper detail a stepwise approach to perform the numerous steps in a quantitative western blotting experiment from sample preparation through data analysis in a rigorous process. Essentially, we have written a recipe to help assure the resulting western blot data will be reproducible and quantitative between blots and experiments. For RT-qPCR, the MIQE Guidelines  and subsequent articles  were published to help scientists produce reliable and reproducible data from this technique and has since been referenced over 2000 times. For quantitative western blotting there have been no such official guidelines published so we have attempted to remedy the matter with this article.
AB: Do you believe that having researchers follow the proposed steps is a step towards a standardized Western Blot workflow?
ST: It is certainly the goal of writing this article in combination with our previous paper  to provide a definitive guide for Quantitative Western Blotting. The two articles together have been downloaded more than two thousand times and the video for the Defined Methodology paper (http://youtu.be/q0T_Af7khHw) combined with that of this article (http://youtu.be/iCgmeufGM2g) have been viewed more than three thousand times in only about nine months. This is certainly indicative of a need for this type of information for scientists globally who are attempting to produce solid and reproducible data from their western blots.
Sean Taylor was speaking to Ashley Board, Managing Editor for Technology Networks. You can find Ashley on Google+ and follow Technology Networks on Twitter.
1. Savla, U. 2004. When did everyone become so naughty? J. Clin. Invest. 113:1072.
2. Rossner, M. 2006. How to guard against image fraud. Scientist. 20:24.
3. Rossner, M., and Yamada, K.M. 2004. What’s in a picture? The temptation of image manipulation. J. Cell Biol. 166:11–15.
4. Adam Marcus and Ivan Oransky, “Can We Trust Western Blots?”, Lab Times, 2-2012, 41
5. Taylor SC, Berkelman T, Yadav G, Hammond M. A defined methodology for reliable quantification of Western blot data. Mol Biotechnol. 2013 Nov;55(3):217-26.
6. Bustin SA et al. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin Chem. 2009 Apr;55(4):611-22.
7. Taylor SC, Mrkusich EM. The state of RT-quantitative PCR: firsthand observations of implementation of minimum information for the publication of quantitative real-time PCR experiments (MIQE). J Mol Microbiol Biotechnol. 2014;24(1):46-52.