📊 WESTERN BLOT ANALYSIS using ImageJ Software// Non-NIH method | RESULTS AND INTERPRETATION
HTML-код
- Опубликовано: 9 сен 2024
- Subscribe for a fun approach to learning lab techniques: / @adwoabiotech
ImageJ is a public domain program developed by Wayne Rasband while at the National Institutes of Health (NIH). The NIH has a recommend process for quantifying bands that differs from the one shown in this video. I made a tutorial of the NIH-based protocol here: • ⚖️ Quantifying Western...
.
.
In this video, we'll be looking at turning Western blot gel images into numerical data that we can do math on. i.e compare quantitatively.
The software that we'll b using is ImageJ, a free image processing program developed by the National Institutes of Health and the Laboratory for Optical and Computational Instrumentation.
A plugins-ready version of ImageJ called FIJI is also available for microscope-based images. FIJI has additional functionality already packaged with it.
During Western blot data analysis, reference proteins such as glyceraldehyde 3-phosphate dehydrogenase (GAPDH), B-Actin or alpha tubulin are used to adjust for uneven loading or transfer of proteins during the setup. This functions to eliminate reporting changes/differences that may be due to variability during the western blot procedure, without having biological significance. Note that there be may cellular expression differences in the chosen reference protein or the experimental conditions may influence the expression of the reference protein differently. Hence you may need to research to make sure that your reference protein is valid. In light of this, there is a move towards using total protein transferred as a reference for normalisation. Also, it's important that you're not overloading (saturating) your reference protein and load an amount that it is in the linear range of the protein.
In the example shown here, the protein lysate used for the Western blot was loaded into a single well, hence there was no need to adjust for uneven loading between wells/samples.
Western blotting involves separating proteins with the aid of the detergent sodium dodecyl sulphate (SDS) and polyacrylamide gel in a process known as polyacrylamide gel electrophoresis (PAGE). Following protein separation, the proteins in the gel are transferred to a membrane - such as nitrocellulose or PVDF - for easier/durable handling of the proteins.
The transferred protein of interest can be detected by exploiting a target-specific probe such as an antibody. Visualisation of the antibody itself is facilitated via coupling to a fluorophore or an enzyme. The most common enzymes coupled to antibodies are Horse Radish Peroxidase (HRP) and Alkaline Phosphatase. Radioisotopes were also widely used in the past.
To quantify the different amounts of protein detected via an antibody, software such as imageJ is used to convert the signal to a numerical value.
Once numerical values are obtained for the various amounts of proteins, direct comparisons can only be made after normalisation of the values.
The concept of normalisation of data can best be explained via an example:
suppose an adult jumped a distance of 150cm and a 10 year old jumped 120cm. Who jumped the furtherest?
Based on the information given, we may assume that the adult jumped the furtherest. However, if we take into consideration, the height of the child (90cm) versus that of the adult (171cm), the situation changes. The adult jumped 0.87x their height, while the child jumped 1.2x their height.
Thus, normalisation takes into account the starting values/state (baseline) before determining the increase or decrease as a result of the experimental conditions.
An alternative method to the one shown in this video is to use calculate the area.
This method is useful where you have multiple western blot bands per lane, that are distinct proteins and therefore cannot be calculated as an Integrated Density. I will do a separate video on that.
.
.
.
.
.
.
//
This video is about western blotting and gel electrophoresis, SDS-PAGE quantification, western blot theory and method, western blotting analysis step-by-step, western blot results and interpretation, western blot normalisation and western blot imagej.
Below are instructions for using the Area Under the Peak method for your quantitation.I found this method to be more accurate when the band intensities are faint. The Area Under the Peak method is also useful when you wish to quantify multiple western blot bands in the SAME lane that are distinct from each each. That is, when you cannot quantify them together as an Integrated Density:
1. select the rectangular tool
2. draw around the lane you wish to quantify
3. go to ANALYZE --> GELS---> SELECT FIRST LANE
4. Use the arrow keys to move the rectangular tool to the next band you wish to quantify.
5. go back to ANALYZE --> GELS---> SELECT NEXT LANE. repeat the process for all additional bands
6. go back to ANALYZE --> GELS---> PLOT LANES. This shows you individual peaks associated with each protein band in all the lanes selected.
7. use the 'line' tool to mark the base of each peak
8. Measure the size of each peak by clicking inside each peak with the Wand tool.
9. copy the values and analyse in excel (or other spreadsheet program) as show in this video
Original instructions from the NIH ImageJ website: imagej.nih.gov/ij/docs/menus/analyze.html#gels
Nice video! I have a tip on solving the issue of accidentally changing the size of the rectangle while trying to move it. Tip: After you draw the desired rectangle, just go to 'save' and save the rectangle selection as 'save selection'. If you lose the rectangle while moving it, just open the saved rectangle selection.
You are a legend for sharing this - thank you!
Nice video and comprehensive tutorial. You mention ROI sizing when making the boxes. Once you make the region, save it to the ROI manager and use that tool to make duplicates. I have a video on measuring intensity in a digital image that describes the process. Great channel. 👍
Fantastic. I appreciate the info.
PAPERS// FACTORS AFFECTING WESTERN BLOT DATA ANALYSIS
An Analysis of Critical Factors for Quantitative Immunobloting (Science Signaling, 2015): stke.sciencemag.org/content/8/371/rs2/tab-pdf
Stain-Free Technology as a Normalization Tool in Western Blot Analysis: journals.physiology.org/doi/pdf/10.1152/japplphysiol.00840.2014
A list of the most reliable house-keeping genes to use for making sure that the differences seen are biological: She, X, ., (2009) Definition, conservation and epigenetics of housekeeping and tissue-enriched genes. BMC Genomics 10:269.
Hi, great video. I am attempting to do this for SDS-PAGE gels rather than WB ones. I am trying to compare Rubisco content and purity from various treatments. However, unlike WB where you probe for a target protein (e.g Rubisco), and get one nice band, my SDS-PAGE gel has many bands other than my POI. In this case, where would I subtract the blank from as there are no 'empty' spaces in the lanes? I thought of taking three measurements at the top of the gel (not in the lanes) and taking an average of this as the background. Do you have any suggestions? Also, I am not sure about the case where there are multiple bands per lane and the fact that I want to find out relative purity and compare the relative quantity between treatments (using a standard curve generated from Rubisco quantitation standard) if I should use the integrated density method or the NIH (AUC) method? Please give me any tips and suggestions. Thanks!
Hi Jack, just to clarify: you have not transferred your protein unto a membrane (nitrocellulose or PVDF), you've simply stopped after gel electrophoresis (SDS-PAGE)? So did you do a commasie stain to visualise ALL the proteins?
In that case, you can use any part of the gel that is not rubisco (i.e. any part that is above or below the mw of rubisco - ~56kDa), as your background.
While Rubisco has a molecular mass of ∼560 kDa, it consists of eight small (∼14 kDa each) and eight large (∼56 kDa each) subunits arranged as eight heterodimers.
Here's a link to a journal article of the experiment you described, so that you can appreciate how they determined the rubisco band: www.ncbi.nlm.nih.gov/pmc/articles/PMC2689150/#:~:text=Rubisco%20has%20a%20molecular%20mass,Malkin%20and%20Niyogi%2C%202000).
All the best.
@@adwoabiotech Hi, no I haven't transferred my protein onto a membrane, I have just stained it with Coomassie (visualised all the proteins). I tried using the auto background subtraction function on ImageJ (w/ 250 pixels) and this seemed to do the job, but I'm not sure if this is appropriate. Also, what is your opinion on using software such as ImageLab? I tried using that instead and think it may give more consistent results and is more intuitive. For example tried three different methods, IntDen, NIH and ImageLab. For the same protein band in the same lane, they gave differing puritys of Rubisco of 30.93% (IntDen), 41.38% (NIH) & 57.07% (ImageLab). From this, I am not sure which technique to use going forward.
Also, thanks for the link to that article, it is very helpful!