Top 5 tips: Western Blots

21st May 2026

Top 5 Tips for Western Blotting

Western blotting is one of the most routine techniques in molecular biology. Most labs run blots weekly. Some run them daily. And yet inconsistent results, unexpected bands, and weak signal remain some of the most common complaints you'll hear from researchers at the bench. The technique itself isn't complicated. But there are five decisions that, if you get them wrong, will cause problems that are genuinely difficult to trace back to their source.

1. Quantify your protein before loading

This is the step that gets skipped most often, usually because it feels unnecessary when you're working to a deadline. It's also one of the most common reasons Western blot results can't be compared across lanes or across experiments. If you load inconsistent amounts of protein, you can't draw conclusions from band intensity. A band that looks stronger in one lane might simply reflect more protein loaded, not a real difference in expression. A Bradford or BCA assay takes less than an hour and gives you the numbers you need to load consistently, typically 20 to 50 micrograms for cell lysates depending on target abundance. Write it into your protocol as a fixed step, not an optional one.

2. Choose your loading control carefully

GAPDH and beta-actin have become default loading controls because they're reliable in most standard cell culture conditions. That reliability doesn't hold across every experimental model. In metabolically active cells or hypoxic conditions, housekeeping protein expression can shift in ways that make them unsuitable as stable references. If you haven't confirmed your chosen control is stable in your specific experimental model, your normalisation data may not mean what you think it does.

There is also a broader methodological shift worth knowing about. Total protein normalisation (TPN), using stains such as Ponceau S or fluorescent total protein stains applied directly to the membrane, is increasingly considered the more reliable approach. Rather than depending on a single reference protein that may saturate or vary, TPN normalises to the entire protein content of each lane, giving better linearity across the typical loading range of 10 to 50 µg. If your experimental conditions are likely to affect metabolic or cytoskeletal proteins, TPN is the stronger choice from the outset. For quantitative work in any model, it is worth evaluating as your default rather than a fallback.

3. Match your blocking buffer to your target

Non-fat milk works for most applications. For phosphorylated proteins, it doesn't, and this is a specific, known incompatibility that still catches people out regularly. Milk contains casein, which is itself a phosphoprotein. When you use milk to block for a phospho-specific antibody, casein competes directly with your target epitope. The result is suppressed or absent signal, which is easy to misread as a problem with expression level or antibody quality. For phospho-targets, switch to BSA in both your blocking and antibody diluent steps. It's a straightforward swap that removes a significant source of false negatives.

4. Optimise your primary antibody concentration

A 1:1000 dilution overnight at 4 degrees is a reasonable starting point for most primary antibodies. It isn't a guaranteed answer for every target, tissue, or cell type, and treating it as one is a common source of non-specific banding. Too high a concentration gives background and bands that don't reflect specific binding. Too low and your signal will be weak or absent. Neither tells you anything reliable about what's in your sample. If you're working with a new antibody or a new sample type, run a titration before your experimental samples. Find the dilution where your band is specific and clean, then lock it in. One extra day at the start saves considerably more in repeats.

5. Choose the right membrane for your protein

PVDF and nitrocellulose will both work for many standard applications. The differences matter more when you're working at the extremes. PVDF has higher protein binding capacity, making it better suited for low-abundance targets. It also holds up better through strip-and-reprobe workflows. If you need to reblot for multiple targets on the same membrane, PVDF is the more reliable choice. Nitrocellulose gives cleaner background for straightforward single-target detection and is slightly easier to handle day to day. PES (polyethersulfone) is also worth factoring into membrane selection. It offers comparable protein binding capacity and durability to PVDF, and as a non-fluoropolymer it sits outside the PFAS regulatory pressure that PVDF is increasingly subject to. It is gaining traction as the practical alternative for labs thinking ahead.

One more variable worth checking

Good reagents don't fix a bad protocol. But a good protocol with a poorly characterised antibody will still waste your time. At St John's Laboratory, our Western blot validated range comes with full validation data, representative blot images, and recommended conditions, so you can see what you're getting before it reaches the bench. Loading controls for GAPDH, beta-actin, beta-tubulin and more are also available.

Explore the loading control range →

Which of these five has given you the most grief at the bench? Let us know in the comments.