A western blot (also called immunoblotting) is the standard lab method for detecting a specific protein in a complex sample and getting a rough read on how much of it is there relative to other samples. It combines two separate skills — separating proteins by size on a gel, then using an antibody to find the one protein you care about — and most of the technique’s notorious troubleshooting comes from a failure in one of those two halves rather than the whole process. This guide walks through the protocol end to end: what to prepare before you start, each step from lysing cells to imaging the final blot, and the most common places results go wrong.
What a Western Blot Tells You (and When to Use One)
A western blot confirms that a specific protein is present in a sample and gives a semi-quantitative estimate of its relative abundance compared to other lanes on the same gel — it is not a precise absolute-quantitation method the way an ELISA or mass spectrometry assay can be. Researchers reach for a western blot to verify a knockdown or knockout worked, to check whether a protein is expressed in a particular cell type or tissue, to track a post-translational modification (using a modification-specific antibody, for example against a phosphorylated residue), or to confirm the identity and size of a purified or overexpressed protein.
If the real question is “how much of this protein is in the sample, exactly” rather than “is it there, and roughly how much relative to my other lanes,” an ELISA or mass-spec-based approach is usually the better fit. If the question is about mRNA rather than protein, that’s a job for qPCR or RNA-seq, not a western blot.
What You Need Before You Start
Before running the protocol, have on hand: a method to lyse your sample and extract protein (a lysis buffer appropriate to your sample type, plus protease and, if relevant, phosphatase inhibitors), a protein quantification assay (BCA or Bradford) so every lane is loaded with an equal amount of total protein, a polyacrylamide gel (precast or hand-poured) sized to your target protein’s molecular weight, a transfer apparatus, a membrane (nitrocellulose or PVDF), blocking buffer, a validated primary antibody against your protein of interest, a matching secondary antibody conjugated to a detection enzyme or fluorophore, and a detection/imaging system suited to that conjugate. Antibody validation matters more than any single step in the wet-bench protocol: an antibody that hasn’t been validated in your species and application (western blot specifically, not just “works for IHC”) is the single most common root cause of an unpublishable or irreproducible blot, independent of technique.
Step 1: Sample Preparation and Protein Extraction
Lyse cells or homogenize tissue in a lysis buffer appropriate to the protein’s cellular location — a membrane-bound or nuclear protein needs a buffer formulated to solubilize that compartment, not a generic whole-cell lysis buffer. Add protease inhibitors immediately (and phosphatase inhibitors if you’re probing a phospho-epitope), since proteolytic degradation and dephosphorylation both start the moment cells are lysed and keep going during handling on ice. Clarify the lysate by centrifugation to remove insoluble debris, then quantify total protein concentration with a BCA or Bradford assay. Normalize every sample to the same total-protein amount before loading — loading unequal total protein is one of the most common causes of a result that looks like a real biological difference but is actually a loading artifact, which is why a loading control (a stably expressed housekeeping protein, or increasingly total-protein normalization via a stain-free gel or membrane stain) belongs in the plan from this step onward, not as an afterthought at the imaging stage.
Mix the normalized sample with SDS sample (Laemmli) buffer and heat — typically around 95–100°C for several minutes for most targets — to denature the protein and coat it with SDS, though some membrane proteins are prone to heat-induced aggregation and are better prepared at a lower temperature per the antibody manufacturer’s or protein-specific literature’s recommendation.
Step 2: SDS-PAGE — Separating Proteins by Size
Load the denatured samples, plus a molecular-weight marker/ladder lane, into an SDS-polyacrylamide gel and run it under an electric current. SDS coats proteins with a uniform negative charge proportional to their length, so the gel separates almost entirely by molecular weight rather than by native charge or shape — smaller proteins migrate faster through the gel matrix and travel further toward the positive electrode in a fixed run time. Gel percentage (the acrylamide concentration) is chosen based on the target protein’s molecular weight: a lower-percentage gel resolves large proteins better, a higher-percentage gel resolves small proteins better, and gradient gels are a common compromise when a blot needs to resolve targets across a wide size range on one gel.
Step 3: Transfer to a Membrane
Proteins separated in the gel are not yet accessible to an antibody — they have to be transferred out of the gel and onto a solid membrane (nitrocellulose or PVDF) that binds protein and can be probed and washed without the gel falling apart. This is done either by wet (tank) transfer, submerging a gel/membrane sandwich in transfer buffer under current, or semi-dry transfer, which uses less buffer and generally runs faster but can be less reliable for large proteins. PVDF membranes need to be pre-wetted in methanol before use; nitrocellulose does not. Under- or over-transfer is a frequent source of a blank or weak blot for large proteins in particular — large proteins transfer more slowly and are more prone to incomplete transfer than small ones, so transfer time and voltage often need adjusting by target size rather than using one fixed protocol for every experiment.
Step 4: Blocking
Incubate the membrane in a blocking buffer — commonly non-fat dry milk or bovine serum albumin (BSA) dissolved in a Tris- or phosphate-buffered saline with a small amount of detergent (TBST or PBST) — to saturate the remaining protein-binding sites on the membrane before adding antibody. Blocking prevents the primary and secondary antibodies from sticking to the membrane itself rather than to the target protein, which is what produces high, even background across the whole blot. Milk-based blockers are cheap and effective for most targets but can interfere with detection of phosphorylated proteins (casein in milk contains phosphoproteins that can cross-react with phospho-specific antibodies) — BSA is the standard substitute for phospho-blots.
Step 5: Primary Antibody Incubation
Dilute the primary antibody in blocking buffer (or a diluent recommended by the antibody’s datasheet) to the concentration validated for that specific antibody, lot, and application, and incubate the membrane with it — commonly overnight at 4°C, though some antibodies work at room temperature for one to two hours. There is no single correct dilution or incubation time that applies across antibodies; the manufacturer’s datasheet and any published validation for that specific antibody in western blot applications are the actual source of truth, and this is exactly why the “what you need before you start” antibody-validation step matters more than protocol mechanics.
Step 6: Secondary Antibody Incubation and Washing
Wash off unbound primary antibody with several changes of TBST or PBST, then incubate the membrane with a secondary antibody that recognizes the species and isotype of the primary antibody and carries a detection tag — typically horseradish peroxidase (HRP) for chemiluminescent detection or a fluorophore for fluorescent detection. Wash again thoroughly after the secondary incubation before detection; incomplete washing at either stage is one of the most common causes of high background across the whole membrane rather than a clean, target-sized band.
Step 7: Detection and Imaging
For chemiluminescent (HRP-based) detection, apply a chemiluminescent substrate and image the membrane on film or, more commonly now, a digital chemiluminescence imager. For fluorescent detection, image directly on a fluorescence imager without a substrate step, which allows more reliable multiplexing (probing for more than one target on the same membrane at once) and generally produces a wider linear dynamic range for quantification than chemiluminescence. Whichever method is used, compare the observed band’s size against the molecular-weight ladder to confirm it matches the target protein’s expected size — a band at the wrong size is a real result worth investigating (isoform, cleavage product, post-translational modification, or antibody cross-reactivity), not something to explain away.
Step 8: Stripping and Reprobing (Optional)
A membrane can often be stripped of bound antibody with a stripping buffer and reprobed for a second target — commonly used to probe a loading control (such as GAPDH, actin, or tubulin) on the same membrane after imaging the protein of interest, saving a second gel and transfer. Stripping is harsher on the membrane than a normal wash and can reduce signal on subsequent probes or, if pushed too far, degrade the membrane itself, so it works better for some antibody/membrane combinations than others and isn’t guaranteed to preserve enough signal for a second full antibody incubation cycle.
Troubleshooting Common Problems
- No band at all: check antibody validation for your species/application first, then work backward through the protocol — insufficient protein loaded, a transfer that didn’t work (a quick check: does the ladder show up cleanly on the membrane after transfer?), an expired or wrong-concentration primary or secondary antibody, or a target protein that is genuinely absent or below the assay’s detection limit in that sample.
- High background across the whole membrane: usually a blocking or washing problem — insufficient blocking time, antibody concentration too high, or incomplete washing after the primary or secondary incubation.
- Multiple or unexpected bands: can be a real biological result (splice isoforms, cleavage products, post-translational modifications, protein complexes that didn’t fully denature) or non-specific antibody binding — a knockout/knockdown control lane, where the band of interest should disappear or shrink, is the most reliable way to tell the two apart.
- Weak or inconsistent signal between lanes: re-check total protein quantification and loading normalization before assuming a biological difference; uneven loading is one of the most common causes of an apparent treatment effect that isn’t real.
- Smearing or distorted bands: often sample degradation (protease activity before or during lysis), overloading the gel, or old/degraded sample buffer.
Western Blot vs. Related Techniques
A western blot is sometimes confused with related methods that serve different purposes. An ELISA (enzyme-linked immunosorbent assay) also uses antibody-based detection but skips the size-separation step, is run in a plate format, and is better suited to precise quantification across many samples. A dot blot applies sample directly to a membrane without gel separation first, which is faster but gives up the ability to confirm protein identity by size and to distinguish a real signal from cross-reactivity at the wrong molecular weight. Immunohistochemistry (IHC) and immunofluorescence (IF) use a similar antibody-detection principle but on intact tissue or cells, to show where a protein is located rather than to measure how much of it is present in a lysate. Choosing among these is really a question of what the experiment needs to answer — identity and relative abundance (western blot), precise quantification (ELISA), speed at the cost of specificity (dot blot), or spatial localization (IHC/IF) — not which one is more sensitive in the abstract.
Frequently Asked Questions
How long does a western blot take from start to finish?
A full run — lysis and quantification, gel electrophoresis, transfer, blocking, primary incubation, secondary incubation, and detection — typically spans two to three working days when the primary antibody incubation is done overnight, which is the most common approach. Same-day protocols are possible with shorter incubation times, but usually at some cost to signal quality for lower-abundance targets.
Why is there no band on my western blot?
Most commonly an antibody validation issue, followed by transfer failure, insufficient protein loaded, or a target that is genuinely below the detection limit in that sample — see the troubleshooting section above for a systematic way to narrow it down.
What is the difference between western blot and SDS-PAGE?
SDS-PAGE is the gel-electrophoresis step that separates proteins by size; a western blot is the full technique that adds membrane transfer and antibody-based detection on top of that separation. Every western blot includes an SDS-PAGE step, but running SDS-PAGE alone (for example, with a general protein stain instead of an antibody) doesn’t identify a specific protein the way a full western blot does.
Can I reuse a membrane after stripping it?
Often yes, for a limited number of additional probes, but stripping is harsher than a normal wash step and signal quality on later probes can degrade, so it works better for some antibody/membrane/target combinations than others — treat a stripped-and-reprobed result as a bit less certain than a fresh membrane, especially for a low-abundance target.
Do I need a loading control?
Yes, in essentially all cases where you’re comparing relative protein abundance across lanes. A stably expressed housekeeping protein (GAPDH, actin, tubulin) probed on the same membrane, or total-protein normalization via a stain-free gel or total-protein membrane stain, corrects for small differences in loading or transfer efficiency between lanes that would otherwise be mistaken for a biological effect.
For the concentration and dilution math used throughout sample and buffer preparation — converting a stock antibody or reagent to a working dilution, or preparing a target-molarity buffer from a solid or concentrated stock — see CASRAI’s guide to molarity and solution calculations for the lab.
