Quantcast

The 4 Important Steps for Western Blot Quantification

As scientists we love nothing more than quantitative data! Oh ya! But if you don’t quantify your Western blots correctly you’ll find yourself in an unpleasant, unrepeatable and totally meaningless place. And while some scientists are okay dwelling in a meaningless place, I hope you are not. Review these important concepts about how to correctly quantitate your next Western blot.

1. Find the Linear Range

For quantitate analysis of an image you must ensure your image was captured in a manner sensitive enough to detect change, in what we call the “linear range”. If you are not working within the linear range – e.g. if your detector or film can no longer absorb photons, it is saturated and you have hit your limit of detection – you are losing data. You don’t want this.

Lucky for you many digital capturing systems come with software designed to detect saturation and automatically correct the exposure thereby ensuring your data analysis is quantitative. So take the time to formally review your software and see if this is your case. However, film can easily become saturated.

To prevent saturation on film, you must empirically determine your linear range. To do this you need to serially dilute a known amount of your protein lysate, preform your Western, and plot the quantitated density of these Western blot bands against the amount you know you loaded. You should then find a linear line indicating where data is captured quantitatively. This is where you want to work. To fix any saturation problems, you can try loading less total protein, less primary antibody dilution, try a new antibody, or reduce the film exposure length.

And yes, you need to go through this process for each antibody separately.

2. Subtract Background

Sadly, most Western blots and image captures are infiltrated with random imperfections. For example, the left side of the blot may be a little darker (higher background) or your less abundant band might have more background or an annoying dark scratch. These differences can cause inconsistencies in your results. Many software packages can calculate background around your band of interest, using some variation of the “rolling ball” method (again, take time to understand your software). The background should be subtracted from both your bands of interest and the bands you are normalizing to. Perfection here is challenging; just do your best and let statistics estimate the real answer when you are all done (Step 4).

3. Normalize

Variability happens in Western blotting. You may have transferred unevenly, loaded too little in one lane, or maybe no one believes your data and they just want to see that you controlled for everything. This is why normalization exits. To control for variability we often normalize to another band in the blot, typically an abundant protein that we don’t expect to change in our experiment. These control proteins are often product from a “housekeeping gene” such as actin, beta-tubulin or a chaperone protein like Hsp70. However, as many of us have discovered, these proteins can unexpectedly change in our experimental conditions. And, due to their high abundance, they can also be challenging to acquire in the linear range. Sometimes choosing a random background band that doesn’t change is the best choice.

Steps to Normalize Your Protein Band of Interest:

Step 1: Determine the background-subtracted densities of your protein of interest (PI) and the normalizing control (NC).

Step 2: Identify the NC that has the highest density value.

Step 3: Divide all the NC values by the highest NC density value to get a relative NC value. If you do this correctly the highest density value will be 1, and the others a fraction of it (e.g., 0.97).

Step 4: Divide all of your PI values by the relative NC values in their respective lanes.

4. Graphs and Stats

Once you have obtained normalized values you are ready to crunch the numbers and view your results. Typically for quantitative experiments you should perform each condition in triplicate (preferably on the same blot). After you have determined your normalized values for each replicate, you can determine averages, p-values, fold changes and/or graph results. Then doing the entire experiment three independent times ensures that your results aren’t a fluke and are indeed repeatable.

In the end, never underestimate the power of getting quantitative – your real results might just surprise you. Good luck!

Leave a Comment





This site uses Akismet to reduce spam. Learn how your comment data is processed.