Hi there,

I don't agree with what is written above: even if all the samples are on a single membrane you have first to take into account that the loading or even transfer efficiency may differ from one lane to another so first you have to quantify the loading control in each lane and then (assuming the detection is uniform) to normalize any band intensity you are interested in with the corresponding loading control. For evaluating single protein level variation, you can compare these ratio of intensities not the absolute intensities. If you want to compare protein stoechiometry from on lane to another, the ratio of normalized intensities for these proteins will tell you about the relative abundancy of one over the other but not about exact stoechiometry unless proteins have been revealed with the same antibodies and assuming these proteins react the same way to antibodies...

As far as you have all the samples to compare in the same blot (with same exposure and such), you are fine with either taking the absolute band intensity or ratio relative to housekeeping gene  or so to compare them. The ratio is more popular.

If you have have samples in different blot, you should have one common sample (e.g. control one) in all blots to normalize to, so the value for this common sample are same from each blot (or very close, at least). Then you can compare the samples from different blots. For that, you might take normalization to the whole membrane, or just normalize to the common sample, which will be 1 by normalizing to itself.

Hi there,

I don't agree with what is written above: even if all the samples are on a single membrane you have first to take into account that the loading or even transfer efficiency may differ from one lane to another so first you have to quantify the loading control in each lane and then (assuming the detection is uniform) to normalize any band intensity you are interested in with the corresponding loading control. For evaluating single protein level variation, you can compare these ratio of intensities not the absolute intensities. If you want to compare protein stoechiometry from on lane to another, the ratio of normalized intensities for these proteins will tell you about the relative abundancy of one over the other but not about exact stoechiometry unless proteins have been revealed with the same antibodies and assuming these proteins react the same way to antibodies...

I agree with Dominique. However, what ever approach you take to normalize your blot, if you are using a film to expose the blot, then the validity of that data is questionable. This is because, it is hard to control the exposure especially saturation which will vary from blot to blot if . It is always better to image the blot using a commercial imaging instrument. You can sequentially scan the image until the bands get saturated. So it is possible to choose a particular frame with unsaturated pixels all the time to compare multiple blots.

Hi Dieynaba,

You should measure peak areas of your bands versus a control band on each W-blot and calculate the percentage values. You can find a detailed protocol of how to do this here:

http://lukemiller.org/index.php/2010/11/analyzing-gels-and-western-blots-with-image-j/

I hope this helps. 

Thank you to all of you for this answers.

I can see clearly now the challenge to compare different blots by your different answers. And I am thinking of a new protocol. Would it be ok if I ran multiple gels (let's say 3) with only one lane of control sample in one of them. After that I would transfer the gels (previously cut in order to get the area I am interested in) all together on one membrane ? This way I have one control sample by membrane (so I can spare my sample) and I can multiply the wells I load. So I would analyze different membranes with the same control sample even if some of my gels would not be ran with this control sample. I don't know if I made myself clear... Would it be ok for the relevance of the analysis downstream ?

Hi

I think that we can normalize Western bloot using Image Lab

™ Software by 4 steps of acting:

1-Image your stain-free blot

2-Image your chemiluminescent blot

3-Creating a multichannel image

4-normalize your western blot

First of all, you need to make sure you have done some kind of BCA assay to quantify the amount of total protein in each sample. Then load in each well the same amount (in grams) of protein (this might mean different volumes of liquid).

Once the lanes have transferred to the membrane, I used to double blot my membranes, which you can do if your antibodies are specific. For me I had my target protein antibody and then one for B-actin. Go through the protocol for your target protein first, then wash the same membrane in block before completing the protocol for your housekeeping protein. If you use a relatively high concentration of your primary antibody you will not need to strip the membrane first. I always left my primary antibodies on overnight at 4C, so if you don't you might need to do an overnight wash at 4C to ensure the luminescence has faded from the first detection.

You can then normalise every lane perfectly as you have more specific data. It makes it more accurate and reduces error.

Thank you Naguibou Mohamadou Djibril. My chief doesn't want me to normalize by stain free blot so I have to do it with a loading control.

Elizabeth C Naylor : Thank you for your answer. It's already in my routine to do a BCA assay but for that particular experiment I have to maximize the data. So following your protocol I would not be able to compare different membranes. So I made a control sample by lysing a pool of tissue and I will load it on each experiment in order to have the same control sample on all my membranes.

Thanks again to all of you ;-).

You can use actin , Beta Tubulin, GAPDH as control in your protein sample.of course, the BCA assay is important to make sure your samples are quite the same in each lane. After ECL revelation, you make a ration between dots area of your protein of interest versus these of actin or tubulin .or...

You can use '' image j'' or ''Totallab'' programs to determine your protein density and compare all your blot together. You should know the amont of your standard band that could be a sure amount of BSA or molecular weight marker.

I agree with Praveen that film is not the best “method of choice” for quantifying proteins. Instead, I recommend the LICOR Odyssey system. It offers a nice possibility to get rid of most problems you have with film based systems. It either works with chemiluminescence or infrared imaging (you need special secondary antibodies for the infrared based imaging (IR)). I strongly recommend the infrared system. Here the scanner uses two channels (2 wavelengths / colors). Thus, you can use one channel (red) for your strong signal and one channel (green) for your faint signal. You can use exactly the same scanner setting (intensity) for every blot, getting better reproducibility. The signal is directly proportional to the amount of target protein. Another advantage is that the fluorescent signal is stable indefinitely. For example, you can collect different time points and scan them together. For quantification, you can use the technique Dominique has mentioned before (normalize any band intensity you are interested in with the corresponding loading control). Another possibility is to absolutely quantify your blots but there for you have to use recombinant target protein as standard.

Good luck!

Hi  Dieynaba Ndiaye

Just in case I would suggest to prepare the pool for each set of your tissue sample, before rinning your experiment samples, for example Kidney, liver, brain poll etc as follow: add same equal amount of each sample (for example 25ul or more) to one eppendorf tube and making pool. You can use prepared pool first to test different antibody then use your original samples to save more of your samples.

Regarding the normalization, the B actine might be used for comparation.

Good luck!

Salman

Regular: anti-actin, anti-tubulin.

Best, particularly after checking the phosphorylation levels of a given protein: use an antibody against the protein you are looking the modifications (e.g. anti-Akt as loading control for anti-pAkt). Good luck!

@Dieynaba

It may be helpful to review, in consultation with your research mentor, current publication guidelines for Western Blot normalization. The guidelines vary from journal to journal, so identify the journal(s) in which you aspire to publish and then use techniques that adhere to the respective guidelines. Hope the following information helps!

• ASBMB (publishes J Biol Chem, J Lipid Res, & Mol Cell Proteomics) prefers normalization by total protein in lane. For single-parameter normalization (e.g., a "housekeeping protein"), ASBMB may ask for more evidence of appropriateness.

Author guidelines

http://www.jbc.org/site/misc/ifora.xhtml#blots

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4705984/

Recent seminar at the ASBMB National Meeting held this past April in San Diego

http://www.jbc.org/site/misc/meeting2016.xhtml

• NPG (publishes Nature, Sci Rep, etc.) allows use of traditional loading controls such as housekeeping proteins, but the linearity and proportional stability of their signals should be confirmed prior to making quantitative comparisons. 

Author guidelines (see under "Electrophoretic gels & blots")

http://www.nature.com/authors/policies/image.html

http://www.nature.com/srep/journal-policies/editorial-policies#digital-image

• Recent publication in Science Signaling comparing normalization by Total Protein vs. Single Proteins. 

"An Analysis of Critical Factors for Quantitative Immunoblotting"

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4401487/

• Free educational webinar with pertinent citations from the scientific literature:

"Western Blot Normalization What You Need to Know"

https://www.youtube.com/watch?v=3SqD4tm6fw4