The advancement of western blotting using fluorescence has allowed detection of subtle changes in protein expression enabling quantitative analyses. Here we describe a robust methodology for detection of a range of proteins across a variety of species and tissue types. A strategy to overcome common technical problems is also provided.
The late 1970s saw the first publicly reported use of the western blot, a technique for assessing the presence and relative abundance of specific proteins within complex biological samples. Since then, western blotting methodology has become a common component of the molecular biologists experimental repertoire. A cursory search of PubMed using the term “western blot” suggests that in excess of two hundred and twenty thousand published manuscripts have made use of this technique by the year 2014. Importantly, the last ten years have seen technical imaging advances coupled with the development of sensitive fluorescent labels which have improved sensitivity and yielded even greater ranges of linear detection. The result is a now truly Quantifiable Fluorescence based Western Blot (QFWB) that allows biologists to carry out comparative expression analysis with greater sensitivity and accuracy than ever before. Many “optimized” western blotting methodologies exist and are utilized in different laboratories. These often prove difficult to implement due to the requirement of subtle but undocumented procedural amendments. This protocol provides a comprehensive description of an established and robust QFWB method, complete with troubleshooting strategies.
Western blotting (WB) is an analytical technique originally developed in the late 1970s to determine the presence or absence of a protein of interest in a complex biological sample, such as a tissue homogenate1. Commonly referred to as the protein immunoblot, due to the key antibody-antigen interaction, the methodology consists of 5 distinct steps: 1) electrophoretic separation of the proteins by their isoelectric point; 2) transfer to a nitrocellulose or polyvinylidene difluoride (PVDF) membrane; 3) labeling using a primary antibody specific to the protein of interest; 4) incubation with a secondary antibody directed against the primary antibody; and 5) visualization.
Visualization methodology has evolved with time to improve safety and sensitivity. Some of the first WBs were carried out using radio labeled tags which then progressed to colorimetric and then the more widely used chemilluminescent (ECL) methods. Radioactivity was directly labeled onto probes for specific antigens whereas colorimetric and ECL methodologies use an indirect labeling technique with an enzyme reporter such as alkaline phosphatase or streptavidin horseradish peroxidase (HRP)2. The intensity of labeling of the chromagen or luminescent product is measured using densitometry whereby the signal strength, strong or weak, indicates more or less the presence of the protein of interest in the sample. ECL is the more sensitive and therefore favored methodology2 but all 3 methods were initially developed on x-ray film with more sophisticated digital imaging techniques subsequently established3. The advancement of digital imaging of WB not only allowed a researcher to determine the presence or absence of their protein of interest, but also allowed an inference to the level of expression of a selected protein when compared against other samples and therefore can be referred to as “semi-quantitative”. Recently, a truly quantitative and more sensitive western blotting technology has been developed whereby the level of fluorescence measured is directly related to the quantity and expression of a single protein within a sample: quantitative fluorescent western blotting (QFWB).
When comparing QFWB with ECL labeling, the use of a fluorescent secondary antibody generates a linear detection profile4. This is in contrast to ECL techniques where signal linearity generally occurs with low protein loads below 5 μg and signal saturation directly related to protein expression, i.e., with ubiquitously expressed housekeeping genes5,6. This disparity is most likely caused by a greater number of binding sites available for an avidin ECL substrate to bind to a biotinylated secondary, resulting in a higher likelihood of potential signal saturation. This is one of the main reasons for ECL base immunoblotting being referred to as only “semi-quantitative”7. The saturation point of signal is of critical importance when measuring subtle differences in expression levels and can lead to inaccurate measurements. In recent years the advent of widespread proteomic techniques detailing ever increasing sensitivity and identification of subtle expression differentials has resulted in an ever increasing reliance on truly quantitative western blotting for validation experiments8,9. The application of sensitive, robust and truly quantitative methodology is therefore crucial.
Many “optimized” western blotting methodologies have been utilized by independent laboratories, which frequently prove difficult to set up or replicate due to subtle methodological adjustments that may not be apparent in formal documented protocols. This is an established and robust protocol for QFWB and additionally provides valuable strategies for troubleshooting common problems that may arise during implementation.
This protocol was originally optimized for use with murine brain homogenates, but has since been used effectively across a broad range of tissue samples and species4,9,10. Potential protocol variations required for specific troubleshooting issues are included.
This protocol has been optimized using commercially produced buffers, gels and transfer stacks in order to reduce variability and improve consistency. Refer to Materials List for a complete list of consumables required.
Fluorescent WB protocol using I-Blot fast transfer and LI-COR Odyssey imaging system
1. Preparation of Sample
Figure 1. Positive control selection. The addition of positive controls to an experiment confirms the labeling detected is real. However, caution must be taken to ensure your control is working correctly prior to using experimental samples. A) Fusion protein TREM 2 was loaded at the manufacturer’s guidelines of 1 μg/ml, however the labeling observed at 110 kDa conflicts with the datasheet predicted molecular weight of 60-70 kDa. B) After addition and incubation with the reducing agent, the fusion protein labeling was detected at the predicted molecular weight. However, this meant a greater protein load was required as the reduction process decreased the signal of the protein.
2. Electrophoretic Separation of Proteins
3. Total Protein Stain of the Loading Control Gel
4. I-Blot Semi Dry Fast Protein Transfer
NOTE: All reagents required for protein transfer using the I-Blot machine are commercial products specifically designed for the I-Blot methodology and can be found in the Materials List.
Figure 2. Optimization of transfer using the I-Blot. A) A single gel loaded with ladder and 15 μg of murine whole brain homogenate in three tandem repeats was cut into three sections. One section was not transferred, one was transferred for 7.5 min (as per manufacturer's guidelines) and one for 8.5 min. The gel sections were then stained with Instant Blue protein stain, scanned on an infrared imager in the 680 channel and quantified. B) Graphical representation of the quantification values demonstrating the difference in residual protein content of each gel following 0, 7.5, and 8.5 min of transfer. Note that an additional minute of transfer time resulted in additional protein transfer of approximately 45%. Please click here to view a larger version of this figure.
5. Antibody Detection of Proteins
Figure 3. Optimization of secondary antibodies. A) A multi species comparison of secondary only non-specific labeling against 15 μg of murine (M), Ovine (O) and Equine (E) nervous tissue homogenates with a variety of fluorescent tagged secondary antibodies. L is the molecular weight ladder. Ovine tissue homogenate was the only sample to cross-react with the secondary labeling when using donkey anti-goat 800 antibody. B) It is also important to ascertain if non-specific labeling occurs when using a different tissue sample, i.e., murine gastrocnemius muscle (15 μg load). This sample cross reacts with the donkey anti-mouse 680 secondary antibody, however this did not occur when using mouse brain homogenate.
Figure 4. Troubleshooting secondary antibody specificity. Western blot of a range of tissue samples (15 μg protein per lane) — fat, muscle, liver and bone — incubated with ERK primary antibody and incubated with three different secondary antibodies. Top panel: WB labeled with LI-COR goat anti-rabbit 680 secondary antibody produced weak labeling of fat and bone and no signal was detected in the liver and muscle samples. Middle panel: Membrane (from top panel) was stripped and reprobed using ECL methodology and Goat anti-rabbit HRP linked secondary. Bands are now visible in muscle and liver samples and labeling appears more intense in the fat and bone samples. Bottom panel: Membrane (from top and middle panel) was stripped and reprobed using LI-COR Donkey anti-rabbit 680 secondary antibody which has shown a greater affinity for the ERK primary antibody. Labeling for muscle and liver is now visible with increased signal intensity from both fat and bone samples. Please click here to view a larger version of this figure.
6. Visualization
NOTE: All images are acquired using the LI-COR Odyssey Classic imager and associated Image Pro analysis software (version 3.1.4).
Figure 5. Visualization and quantification of western blot. Scan of a western blot showing murine gastrocnemius muscle (30 μg load) probed with Annexin V primary antibody (36 kDa) and goat anti-rabbit 800 secondary antibody. The membrane was scanned and visualized in the 800 channel. To quantify the protein (Annexin V), a rectangular box is drawn around the band of interest from sample 1. This is then copied and pasted over the remaining sample lanes to ensure measurement of the same area. Background is automatically accounted for around the shape drawn but this can be altered to ensure the background measurement is accurately defined. The table below displays the quantified measurements of each shape drawn including total signal obtained, background and signal with background subtracted. This information can then be exported into a spreadsheet program to calculate expression ratios (as determined by relative fluorescence intensity) and allows subsequent statistical analyses to be performed. Please click here to view a larger version of this figure.
7. Post Visualization
As QFWB sensitivity and the linear range of detection is greater than conventional ECL detection, there are a number of control measures that are crucial to ensure that accurate data is collected, thereby aiding effective interpretation. Firstly, the inclusion of positive control samples as shown in Figure 1. Secondly, optimization of transfer to guarantee equivalent movement of high and low molecular weight proteins from the gel to the membrane as exhibited in Figure 2. Thirdly, optimization of antibodies, especially secondary antibodies whose optimization is often overlooked, but which can produce non-specific banding capable of interfering with correct interpretation of protein(s) of interest. See Figure 3. Fourthly, it may also be the case that when a protein appears undetectable but is expected to be present, this may also be a secondary antibody issue which can be corrected by simply using a secondary raised in a different species host. See Figure 4. Fifthly, total protein labeling and analysis is a far more robust and quantifiable method in comparison to the use of traditional single protein(s) that are ubiquitously expressed for internal reference standards3. Many of these single proteins have been found to be differentially expressed in models of neurodegenerative diseases as well as between different tissue samples and the uniformity of expression can alter within the same tissue3. Therefore, production of a loading control gel will confirm the uniformity of sample load when combined with a total protein analysis by comparing and quantifying the protein load in each lane at various molecular weights ranges measured against each sample to indicate standard error as demonstrated in Figure 6. Importantly, all of these troubleshooting techniques and controls are only as effective as the sensitivity and consistency of the analysis tools applied by the operator (Figure 5). Finally this technique lends itself to stripping and re-probing of membranes with more flexibility than ECL due to factors including but not limited to increased sensitivity, reduced background, dual color detection and membrane stability under long term storage conditions. See Figure 7.
Due consideration and planning is essential prior to any experiment and can ultimately determine the success of the technique used. The advancements in protein detection using WB can present a plethora of potential stumbling blocks when trying to choose the appropriate antibodies, transfer and visualization methods to use. Fortunately, using a careful checklist and appropriate control measures QFWB can be used routinely to determine protein presence and increasingly subtle expression differences between samples. This protocol provides a comprehensive guide to fluorescent quantitative western blotting as well as a few troubleshooting strategies to avoid and/or overcome some of the many common pitfalls associated with it.
The critical steps employed to maintain sensitivity and obtain truly quantifiable and comparable measurements include: 1) robust protein extraction from tissue samples; 2) sample preparation; 3) accurate protein loading determined by total protein analysis; 4) optimal transfer of proteins using I-Blot; 5) preparation of primary and secondary antibodies in blocking buffer containing 0.1% Tween20, and 6) correct visualization and analysis using an infrared imager and associated software.
Infrared fluorescent detection is truly quantitative and provides greater sensitivity compared with more traditional ECL detection techniques3,11. This detection system is multi faceted, and as such is not limited to QFWB. This system is capable of imaging of immunohistological labeling at low power allowing visualization and quantification of whole tissue sections12. This is one area of potential future development in terms of resolution which could see far red imaging rivaling conventional immunofluorescence capturing with conventional microscopes in terms of quantitative assessment.
However, with greater sensitivity to subtle changes in protein expression it is crucial to ensure variability is kept to a minimum and control measures are stringent with robust protocols. This begins with rigorous protein extraction from the tissue sample followed by production of total protein stained gels to provide assurance that sample loading is uniform, optimization of primary and secondary antibodies in order to determine if detection is real and testing the manufacturer's guidelines with regard to transfer times to obtain efficient protein transfer.
Nevertheless, even when conditions for WB are optimized, there may still problems associated with running westerns that may not have been fully explored here. These include but are not limited to factors including protein solubilization and choice of extraction buffer. Some buffers can interfere with protein concentration assays, and some tissues are particularly difficult to solubilize, requiring more robust techniques such as the use of automated macerating sealed containers such as M tubes together with a Macs dissociator. In addition, simple control measures for storage of extracted and non-extracted material at -80 °C can be the difference between obtaining optimal labeling immediately after extraction and having poor results weeks later.
Modern QFWB methods are proven to be more sensitive for capturing subtle differences in protein expression and are more versatile allowing simultaneous dual labeling3 when compared to older techniques such as ECL. It is vital that western blotting protocols are robust and readily repeatable for accurate quantification and statistical analysis. This protocol is sensitive and robust enough to be used routinely for detection of proteins across a variety of different tissue samples and species3 and allows quantification of low and high abundance proteins within the same QFWB therefore reducing consumable usage as well as time per experiment11. In addition, the increased sensitivity of this technique allows validation of increasingly popular –omic studies9,14 however accuracy is crucial and inclusion of appropriate control measures must be adhered to thereby avoiding erroneous data acquisition.
The authors have nothing to disclose.
We would like to thanks the following for financial support: BBSRC Institute Strategic Programme Funding – CF & TMW; BBSRC East Bio DTP funding – LG; The Darwin Trust of Edinburgh – MLH. We would also like to thank Dr Barry McColl for permission to include the TREM2 optimization in this manuscript.
Name of Material/ Equipment | Company | Catalog Number | Comments/Description |
RIPA Buffer | Fisher Scientific UK Ltd | 10230544 | |
M Tubes | Miltenyi Biotec Inc. | 130-093-236 | |
iBlot Transfer Stack, PVDF Regular | Life technologies, UK | IB401001 | |
MagicMark XP Western Protein Standard (20-220 kDa) | Life technologies, UK | LC5602 | Use in gel 1 for Western blotting |
SeeBlue Pre-stained protein standard | Life technologies, UK | LC5625 | Use in gel 2 Total protein stained gel |
NuPAge LDS Sample buffer 4X | Life technologies, UK | NP0007 | |
NuPAGE MES SDS Running Buffer (for Bis-Tris Gels only) (20X) | Life technologies, UK | NP0002 | |
NuPAGE Novex 4-12% Bis-Tris Gel 1.0 mm, 12 well | Life technologies, UK | NP0322BOX | |
PHOSPHATE BUFFERED SALINE TABLET,*TRU-ME, PHOSPHATE BUFFERED SALINE TABLET | Sigma-Aldrich, UK | P4417-100TAB | |
Micro BCA, Protein Assay Kit | Fisher Scientific UK Ltd | 10249133 | |
Odyssey blocking buffer | Li-Cor Biosciences | P/N 927-40000 | |
IRDye 680RD Goat anti-Rabbit IgG (H+L), 0.5 mg | Li-Cor Biosciences | 926-68071 | |
IRDye 680RD Donkey anti-Mouse IgG (H+L), 0.5 mg | Li-Cor Biosciences | 926-68072 | |
IRDye 800CW Goat anti-Rabbit IgG (H + L), 0.5mg | Li-Cor Biosciences | 926-32211 | |
IRDye 800CW Donkey anti-Goat IgG (H + L), 0.5mg | Li-Cor Biosciences | 926-32214 | |
ODYSSEY CL Infra-red imager | Li-Cor Biosciences | Call for quotation | |
iBlot 7-Minute Blotting System | Life technologies, UK | This model is no longer in production | |
InstantBlue Protein stain | Expedeon, UK | ISB1L | |
Revitablot western blot stripping buffer | Rockland Immunochemicals Inc. | MB-085-0050 |