Western blot
An HRP-conjugated secondary oxidises luminol and the emitted light is collected on film or a cooled CCD. The cheapest and most sensitive routine detection, with picogram limits using high-sensitivity substrates, but the signal accumulates on the detector and the enzyme consumes its substrate, so the response is neither linear nor stable.
Our ruleSimulations prepare you for the bench but don't replace hands-on practice.
Chemiluminescent
An HRP-conjugated secondary oxidises luminol and the emitted light is collected on film or a cooled CCD. The cheapest and most sensitive routine detection, with picogram limits using high-sensitivity substrates, but the signal accumulates on the detector and the enzyme consumes its substrate, so the response is neither linear nor stable.
How it differs from the standard method
- Exposure is a parameter of the measurement: a long one compresses fold changes, a short one loses faint bands.
- Signal decays over minutes to hours as the substrate is used up, so two exposures of the same blot are not interchangeable.
- Only one target per probe, so a loading control needs stripping and reprobing, or a second membrane.
- A CCD imager with on-screen saturation warnings is far better than film for anything that will be quantified.
1The gel is upstream
A western blot has three parts, and the gel is the one that almost never fails. Proteins are separated by SDS-PAGE, driven sideways out of the gel onto a membrane that binds them where they land, and the membrane is then probed with an antibody.
2Ponceau S is the control that comes first
A blank blot has two explanations: nothing transferred, or the antibody did not work. Once the membrane is blocked, there is no way to tell them apart.
Transfer
Semi-dry for speed, wet tank for large proteins. Efficiency falls with size, and small proteins blow through.
Ponceau
Stain and photograph the membrane before blocking. The only record of what reached it.
Block
BSA, milk or a protein-free blocker. The choice interacts with the antibody.
Primary
Dilution trades signal against background; there is an optimum, not a maximum.
Secondary & detect
HRP with ECL, or a fluorescent conjugate. The detector decides the linear range.
3Blocking interacts with the antibody
Blocking covers the membrane's remaining protein-binding capacity so that the antibody can only bind what is on it. Non-fat milk is cheap and gives the lowest background for most antibodies, but milk is about 1% casein, and casein is a phosphoprotein.
4Signal, background, and why more antibody is not better
Specific binding saturates with antibody concentration: once the epitopes are occupied, more antibody adds nothing. Non-specific binding does not saturate: it rises roughly in proportion to concentration all the way up.
5Why most western blots are not quantitative
Chemiluminescent signal accumulates on the detector. A long exposure makes a beautiful dark blot by letting the brightest bands reach the ceiling, and once a band is at the ceiling it reports the detector rather than the protein.
- Linear range
- The range of protein in which measured density is proportional to amount. It is a property of the whole blot (target, antibody, detector and exposure) not of the technique, so it has to be established on each blot with a dilution series of one of the lysates.
- Loading control
- A band used to correct for unequal loading and transfer. It is only a control inside its own linear range, which an abundant housekeeping protein usually is not at the exposure that shows the target.
- Total-protein normalisation
- Imaging a fluorescent stain of all the protein on the membrane and normalising to it. It measures what actually reached the membrane and has no saturation problem of its own.
- Stripping and reprobing
- Removing antibody so the same membrane can be probed again. It always costs some antigen, so the second probe's signal is not comparable with the first's unless the loss is controlled for.
- Specificity control
- A lane of knockdown, knockout or antigen-blocked material showing that the band you measured disappears when the protein does. A band of about the right size is not an identification.
6What a loading control can and cannot do
The point of a loading control is that it rides the same transfer, the same blocker and the same membrane as the target, so it reports the things a protein assay cannot. But that only works if its band is in the same proportional region of the detector as the target's.
Key terms
Practise these as flashcards →The fraction of a protein that leaves the gel and sticks to the membrane. Falls with size for a fixed transfer, and falls for small proteins when the transfer is long.
A reversible red total-protein stain for membranes. Photographed before blocking, it separates a transfer failure from an antibody failure.
Protein or polymer that covers the membrane's remaining binding capacity. Its composition can compete with the antibody: milk casein with a phospho antibody, milk biotin with streptavidin detection.
Enhanced chemiluminescence: HRP oxidises luminol and the light is accumulated on film or a CCD. Sensitive, cheap, and saturating.
Integrating band intensity to get a number. Meaningful only inside the blot's linear range, and only after background subtraction.
A band at the detector's ceiling. Every ratio involving it is compressed towards 1, so a measured fold change is a lower bound.
Band density divided by the membrane background next to it. It peaks at an intermediate antibody dilution, not at the highest.
When to use it
- Confirming that a specific protein is present in a sample, at the expected apparent mass, in a way that a plate assay cannot show.
- Comparing the amount or the modification state of one protein between a handful of samples, with a dilution series on the same membrane to make the comparison quantitative.
- Distinguishing a protein from a cleaved, modified or differently spliced form, because the readout carries a size as well as an intensity.
- Validating an antibody before using it in an assay that gives no size information, such as an ELISA, an immunoassay panel or immunohistochemistry.
Compared with alternatives
Gives concentrations for many samples quickly and reproducibly, but no size information, so a cross-reacting or truncated species is invisible.
The upstream separation. A stained gel shows everything that is there; a blot shows only what the antibody binds.
Measures dozens of analytes in one small sample; prefer it over running dozens of blots, but validate the panel first.
Identifies and quantifies without an antibody, with far better specificity; prefer it when antibody availability or specificity is the limiting problem.
Shows where the protein is in the cell rather than how much is in the lysate; the two answer different questions and often use the same antibody.
Sub-techniques
Choosing one here switches the whole page, including the protocol and the lab, to it.
An HRP-conjugated secondary oxidises luminol and the emitted light is collected on film or a cooled CCD. The cheapest and most sensitive routine detection, with picogram limits using high-sensitivity substrates, but the signal accumulates on the detector and the enzyme consumes its substrate, so the response is neither linear nor stable.
- Exposure is a parameter of the measurement: a long one compresses fold changes, a short one loses faint bands.
- Signal decays over minutes to hours as the substrate is used up, so two exposures of the same blot are not interchangeable.
- Only one target per probe, so a loading control needs stripping and reprobing, or a second membrane.
- A CCD imager with on-screen saturation warnings is far better than film for anything that will be quantified.