Proximity ligation assay
Visualize two proteins sitting within about 40 nm of each other in cells.
Our ruleSimulations prepare you for the bench but don't replace hands-on practice.
1Turning proximity into a spot
A proximity ligation assay detects two proteins that sit close together in a fixed cell, and reports each occurrence as a single bright spot in the place where it happened. Two primary antibodies from different species bind the two targets.
2The single-antibody controls are mandatory
The obvious assumption is that one antibody alone cannot give a spot, because a spot needs two probes. In practice it can.
Primaries
Two species, titrated low enough that each alone gives almost no spots.
PLUS and MINUS probes
Species-specific secondaries carrying the oligonucleotides.
Ligation
Connectors hybridise across both probes and are closed into a circle, only within about 40 nm.
Amplification
Rolling-circle copying of the circle, with fluorescent detection oligonucleotides.
Image and count
Fixed exposure, fixed threshold, cells selected by nuclear mask, every slide alike.
3Proximity, not binding
Forty nanometres is a long way on the scale of a protein: two antibodies and two secondaries already span a good fraction of it, and two proteins that never touch can easily be closer than that. Two abundant proteins in the same membrane domain, or crowded into the same small compartment, produce spots by chance, and the count rises with their abundance whether or not they interact.
- PLUS and MINUS probes
- Secondary antibodies (or directly conjugated primaries) carrying the two different oligonucleotides. The pair, not either one, templates the ligation.
- Rolling-circle amplification
- A strand-displacing polymerase copying a closed DNA circle continuously, producing one long single-stranded product with hundreds of repeats that stays in place.
- Single-antibody control
- A slide with one primary only and both probes. Its spot count is the background that the pair has to exceed.
- Spots per cell
- The readout: detected objects inside each cell mask, averaged over many cells chosen without looking at the spots.
- Spot merging
- Two spots closer than the imaging resolution, or than their own size, counted as one. It flattens the count exactly where the signal is densest.
4Counting is part of the assay
The amplification products are not identical (the circle is copied for a fixed time, but each product ends up a different length) so spot brightness spreads over roughly a factor of two either way. The detection threshold therefore decides what is counted.
Key terms
Practise these as flashcards →Ligation of connector oligonucleotides into a circle, which is only possible when two probes are within about 40 nm of each other.
Two short DNA strands that hybridise to both probes at once and are joined by the ligase into a circular template.
The intensity a local maximum must exceed to be counted as a spot. It trades missed dim spots against counted noise.
Cells lacking one partner (knockdown or knockout), stained with both antibodies. Shows that the spots require the protein, not just the antibody.
Spots from two proteins that are near each other without binding, which rise with their abundance and crowding.
Cell outlines grown from the DAPI-stained nuclei, used to assign each spot to a cell without reference to the spots themselves.
Each spot is one ligation event, so the assay counts events rather than measuring an intensity, and is limited by counting statistics.
When to use it
- Showing where in the cell two proteins come close, and in which cells, when a lysate-based method would average over every compartment and every cell.
- Detecting an endogenous interaction in tissue or primary cells, where there is too little material for a co-IP and no possibility of expressing a fusion.
- Following how often two proteins are close across conditions, a time course or a drug treatment, with a count per cell that can be compared statistically.
- Detecting a post-translational modification on a specific protein, using one antibody against the protein and one against the modification.
Compared with alternatives
Works on lysates and can identify unknown partners, but loses location and sees only what survives the buffer; prefer it for discovery and biochemical follow-up.
Finds unknown neighbours of a bait in living cells; prefer it for discovery, and PLA for checking a named pair cell by cell.
FRET reports proximity under about 10 nm in living cells with fluorescent fusions; prefer it when the question is dynamic or needs tighter distance resolution.
Co-localisation at light-microscope resolution is about 200 nm, far coarser than PLA's 40 nm; overlapping signals are not evidence of proximity.
Localises both proteins to tens of nanometres directly, at much higher cost and complexity; prefer it when the spatial arrangement itself is the question.