Ning Liu Ph.D.

Find out what could be going wrong in your experiments.

Did you know that 41 percent of researchers say their Western blots fail at least 25 percent of the time? Ryan Jensen, a biologist at Yale University, has performed thousands of Western blots in his career. He and Aldrin Gomes, a biologist at UC Davis who recently wrote a review on the topic, say that often the issue is with the primary antibody. A recent survey showed that nearly two in three scientists are running into similar problems.

“Many problems encountered by researchers can be easily avoided if they’re taught theory,” Gomes says. If scientists understand how Western blotting works, they are more likely to be able to reason through the protocols and troubleshoot their own problems. For instance, one common problem is nonspecific-bands which is often caused by too much antibody or too much protein loaded in each well. Gomes says he primes all his students with his document, “Before You Start Western Blotting You Must Read This”, before they ever touch a membrane. Understanding the concept behind the techniques, Gomes says, “helps prevent many common mistakes from occurring.”

Problem #1: Nonspecific Bands

Expert diagnosis: Problems with antibody quality or protein concentration
Recommended solution: Try a different antibody or decrease protein concentration
Expert insights: Our experts agree that when they see nonspecific bands, they look at the antibody first

If you’ve wound up with a terrible antibody, you might be getting some unexpected bands in your Western blot. Jensen says sometimes this happens because companies provide low-quality primary antibody. Out of the 50 companies that make antibodies to BRCA2, Jensen says only one consistently works in his lab. He says when you’re looking for a good antibody, see what’s worked for other scientists. “I trust an antibody more when I see it in the literature by a real researcher working on an application similar to mine.” Jensen says. Outside of academic research journals, Jensen says to check university or online forums like ResearchGate’s Q&A board for your protein.

He also suggests purchasing $50 trial-sized vials of antibodies from companies such as that allow you to try different antibodies without the added investment. Once you’ve found the right antibody for your application and use it at the optimal concentration, Gomes says these extra bands tend to disappear. What if these bands still persist? The problem might be too much total protein loading or insufficient washing. While Gomes sees papers where labs load 100 to 150 micrograms in their Western blots, he’s found that the optimal amount of protein per lane is about a tenth that amount.

Once your Western blots produce a single band, you should confirm whether that band is your protein. Jensen says the best control is an siRNA knockdown or loading a knockout sample.

Problem #1

Problem #2: Weak (or no) signal detection

Expert diagnosis: Problems with antibody, insufficient antigen or even the buffer used
Recommended solution: Validate all new antibodies, increase exposure time or make fresh buffer
Expert insights: While you won’t see this problem in the literature because the blot wouldn’t be published, this problem happens a lot in the lab

When Jensen was purifying endogenous BRCA2, he found that it expressed at very low levels, which was a challenge he had to overcome. A weak sample begets a weak signal. Jensen says you might get a strong signal by loading more lysate or increasing incubation time. He transfers the gel to blot overnight and has it incubate with the primary antibody for an entire weekend.

He also points out that protein transfer can be compromised. A quick and dirty way to confirm protein transfer with a PVDF membrane is to let it dry for a second and then hold it up to a light at an angle. To know if your transfer was successful, you should see the bands. Remember to rehydrate the blot to prevent it from drying out.

Most of the time, Gomes says the problem is in the amount of antibody. If the concentration is too dilute, no detection will occur. He recommends using a positive control on all blots if possible to ensure the antibody is working. If the antibody is not working, then your buffer’s ionic concentration and pH might be off since these properties affect the antigen-antibody interaction. This summer, three top high school students in California who are part of the Young Scholars Program tried blotting in Gomes’ lab. “The first week, nothing they did worked,” says Gomes. So, he had them remake their buffers. Subsequently, every Western worked. Before his lab performs Westerns, they must first validate the antibody. This clears the antibody from fault if things do go wrong.

Problem #2

Problem #3: Protein Detection Saturated

Expert diagnosis: Protein overloaded, typically loading controls
Recommended solution: Try using a smaller amount of protein or use total protein to normalize for loading errors 
Expert insights: Both experts stated that this issue is prevalent, even in the literature

Saturated loading control is one problem you may not be watching out for. Jensen notes that he sees this problem all the time in papers and posters because “researchers don’t think about this issue often enough.” But, he says, “it should be taken seriously as it makes normalization pretty useless.” If you’re using film, it can be difficult to tell when your signal is saturated. Jensen used to expose the film for increments of time up to an hour to get a feel for the dynamic range. Now, he uses a CCD-based imaging system that automatically notifies him when a signal is saturated.

Jensen and Gomes have gotten around the saturation issue entirely by switching to stain-free total protein normalization. Using stain-free total protein for normalization can replace the use of housekeeping proteins like actin or tubulin, which may also be unreliable loading controls because their expression can vary due to experimental conditions. For best results, Gomes recommends loading 10 to 30 micrograms of total protein per well. To prevent signal saturation of his protein of interest, he also performs a standard curve.

Problem #3

Ning Liu, Ph.D., is senior product manager at Bio-Rad Laboratories. Have other problems such as blotchy background? Visit The Western Blot Doctor at for your diagnosis. Please email Aldrin Gomes at [email protected] if you would like a copy of the discussion guide he makes all his students read before performing Western blots

Previous articleNIH Publishes Finalized Policy on Genomic Data Sharing
Next articleAlcohol-Dependence Gene Linked to Neurotransmitter