LI-COR Biosciences
Principal Scientist
Reproducibility is essential to the scientific process, enabling other labs to verify and build upon published results. But it’s increasingly apparent that the scientific literature does not always meet those expectations. Much of this uncertainty arises from inconsistent reporting of methodological details, leading to incorrect interpretation of results. The methods themselves also may contribute to irreproducibility.
In 2013, Nature introduced new editorial standards to improve the consistency and quality of reporting in the life-science literature. These standards require authors to “describe methodological parameters that can introduce bias or influence robustness [1],” including “technical replicates [that] reflect the variation of the assay and/or sample preparation [2].”
Western blots are a method with inherent variability. Different antibody batches and clones can behave differently, making it hard to compare results between labs or even within the same lab. Gel loading, transfer efficiency and other factors are additional, recognized sources of variability.
Another key aspect of Western blotting has received less attention: detection and imaging. This aspect is critical for quantitative comparisons, and relative analysis of protein levels is increasingly common. Western blots are detected with labeled antibodies, using colorimetric, chemiluminescent or fluorescent methods. Chemiluminescent detection is the most common approach and also the most difficult to quantify accurately.
Scientists at LI-COR Biosciences recently examined sources of variability and error in chemiluminescent Western blot imaging. Here, we review their findings.
Limitations of chemiluminescence
Chemiluminescent Western blots use enzyme-conjugated antibodies to catalyze a reaction. Oxidation of a substrate molecule produces a transient light signal. Emitted light is imaged with x-ray film or a digital imager as the reaction occurs.
Variability and error in chemiluminescent Western blot imaging arise from two key sources: The kinetics of the enzymatic chemiluminescent reaction, and the challenge of finding the “best” exposure to capture both faint and strong signals.
The enzymatic nature of chemiluminescent detection enables high sensitivity but is dependent on substrate concentration. This a concern, because substrate concentration changes as the reaction continues—and changes at different rates in different areas of the blot. Substrate is consumed much more rapidly in areas with high antibody concentration (strong bands) than lower concentration (fainter bands).
As the reaction proceeds, changes in substrate availability can affect the relative signal intensity of strong and faint bands. Stronger signals become under-represented if less substrate remains available for oxidation. This unpredictable, silent source of variability is dependent on the timing of image capture and reagent distribution on the blot. The resulting variability may limit reproducibility from one experiment to the next.
Furthermore, most blot imaging methods require multiple exposures to document both faint and strong bands within the linear range of the method. But intense signals can saturate or “blow out” the capacity of many detection methods before weaker signals are even detected. A short exposure often is used to capture strong bands within the linear range, followed by a much longer exposure to detect fainter bands. It’s nearly impossible to accurately capture all signal intensities in a single exposure, and the dynamic nature of the reaction makes it more difficult to identify the “best” exposure times.
Imaging within the linear range
The linear range of your detection method determines the need for multiple exposures, and by extension, the potential impact of enzyme/substrate kinetics on your results. An ideal imaging method would capture all data in a single exposure, with both faint and strong bands captured in the linear range of detection.
We compared the linear detection ranges of x-ray film, a commercial CCD imaging system (“Imager B”) and the LI-COR Odyssey® Fc Imager. A Harta luminometer reference plate, with seven LEDs that produce consistent light over seven orders of magnitude, was used as a light source. (See Figure 1.)
Figure 1. Linear detection ranges of various methods
Film exhibited a very narrow linear range that is poorly suited to a wide range of signal intensities. Imager B displayed an improved linear range, although strong signals still became saturated during longer exposures. The LI-COR Odyssey Fc Imager was the best of the three, with a very wide linear range and no signal saturation observed. This system captured both faint and strong signals in a single image, at all exposure times.
Unlike other CCD-based imagers, the Odyssey Fc system’s unique optical technology brings out faint bands without saturating the stronger ones and does so across a very wide dynamic range (up to 6 logs). There’s no need to optimize image-capture settings, and a single digital image replaces a stack of multiple exposures.
Capturing all signals in a single exposure reduces the variability introduced by the kinetics of chemiluminescent detection. Faint and strong bands are collected during the same phase of the enzymatic reaction, for more consistent substrate availability and more reliable quantitative analysis—especially when band intensities vary. The system also detects near-infrared fluorescent signals at two wavelengths and images DNA gels.
Improving Western blot reproducibility
It is possible to obtain reliable results from film or standard digital imagers, of course. You’ll need to carefully calibrate the linear range for each experiment. Using an internal control for normalization can increase precision, but be sure that both the control and the target protein are detected within the linear range of your system. Increased precision may require dilution series, additional controls and a larger number of technical replicates.
The new Nature reporting requirements emphasize the importance of detection within the linear range and discourage use of short exposures that show only intense bands. For electrophoresis and gel data, the guidelines specify, “Appropriate reagents, controls and imaging methods with linear signal ranges must be used. Exposures should generally be such as to produce gray backgrounds. High-contrast gels and blots are discouraged, as overexposure may mask additional bands. Multiple exposures should be presented in supplementary information if high contrast is unavoidable [2].”
This study demonstrates that the LI-COR Odyssey Fc Imager can provide a more accurate, quantitative solution for chemiluminescent Western blots. Eliminating the need for multiple exposures simplifies your data analysis and minimizes the variability introduced by enzyme/substrate kinetics during chemiluminescent detection. Capturing the full range of data in a single image helps you get consistently reliable results—and more easily meet the new, increasingly rigorous requirements for publication of Western blot data.
The full study, “Film and CCD imaging of Western blots: exposure time, signal saturation, and linear dynamic range,” (LI-COR Biosciences, 2014) is available here.
References
[1] “Reducing our irreproducibility,” Nature, 496:398, 2013.
[2] “Reporting life sciences research,” Nature, May 2013.
Related Products from: LI-COR Biosciences