Pull-down assays
A glutathione S-transferase fusion is immobilised on glutathione agarose, which is cheap, high-capacity and reversible: 10 mM reduced glutathione competes the bait straight off. GST is a 26 kDa dimerising enzyme with its own interactors, and the cell's own glutathione S-transferases bind the resin, so the GST-only and empty-resin controls both do real work here.
Our ruleSimulations prepare you for the bench but don't replace hands-on practice.
GST pull-down
A glutathione S-transferase fusion is immobilised on glutathione agarose, which is cheap, high-capacity and reversible: 10 mM reduced glutathione competes the bait straight off. GST is a 26 kDa dimerising enzyme with its own interactors, and the cell's own glutathione S-transferases bind the resin, so the GST-only and empty-resin controls both do real work here.
How it differs from the standard method
- Reversible capture means the bait can be eluted intact with free glutathione, or left on the resin and the prey released by cleaving the linker.
- GST dimerises, so a GST-fused domain is presented as a dimer, which can create or mask an interaction that needs a monomer.
- A degradation fragment of the fusion runs at 26 kDa, exactly where many small prey proteins run, and the empty-resin lane also carries endogenous glutathione S-transferases at that size.
- The tag is large enough to interfere with folding or activity, so a GST fusion that gives a negative result needs an independent check that the bait domain works.
1An antibody replaced by a bait
A pull-down is an immunoprecipitation with the antibody taken out. A recombinant protein, the bait, is immobilised on a resin through an affinity tag, and that surface is used to fish for whatever binds it: out of a lysate, or out of a tube containing nothing but purified protein.
2You built the surface, so you own its artefacts
The bait is at a concentration you chose, immobilised at a density you chose, in an orientation decided by where the tag is. None of that is physiological.
Charge the resin
Bind the bait, wash off what did not bind, and check the loading on a gel. Equal density in every tube, including the controls.
Bind
Lysate or purified prey, with detergent, carrier protein and a reductant in the buffer.
Wash
Stringency weakens the interaction faster than it removes background. The control resins are the better tool.
Elute
Boil, compete off with free ligand, or cleave the linker. Only cleaving leaves the tag behind.
Read
Stain for the bait and the abundant background, blot for the prey, and compare with empty and tag-only resin.
3The tag is a bait too
GST is a 26 kDa enzyme that dimerises and has interactors of its own; polyhistidine on a nickel resin captures the cell's metal-binding proteins; streptavidin beads capture every endogenously biotinylated carboxylase in a mammalian lysate, which is why a prominent 130 kDa band in a streptavidin eluate is usually pyruvate carboxylase rather than a discovery. So two controls are needed and they catch different things.
- Bait density
- Micrograms of tagged protein per unit of resin. It sets how much prey is captured, so it has to be matched between the pull-down and the tag-only control, and it should be as low as the signal allows.
- Bait leaching
- Bait released from the resin into the eluate. Boiling releases all of it, so micrograms of bait land in one lane; a competitive elution releases most of it; cleaving the linker leaves the tag on the beads.
- Tag-only control
- Resin carrying the tag alone at the same density as the fusion. The only lane that identifies a protein binding the tag rather than the bait.
- Flow-through
- What was not captured. It separates a prey that never bound from a prey that bound and was then washed off, two problems with opposite fixes.
- Direct binding
- A result obtained with purified bait and purified prey, where no third protein can bridge them. The thing a pull-down can show and a co-IP cannot.
4The bait on the gel
If the beads are boiled in sample buffer, everything on them goes into the lane, and that is mostly bait. Twenty micrograms of immobilised GST fusion against a few nanograms of captured prey means the stained gel is a picture of your bait with a smear beneath it where the degradation fragments run, and free GST at 26 kDa sits uncomfortably close to a small prey.
5What a positive and a negative each mean
A positive pull-down from purified protein means the two proteins can bind under those conditions, at those concentrations, with the bait on a surface. It says nothing about whether they ever meet in a cell, how much of either pool is bound, or whether the interaction is regulated.
Key terms
Practise these as flashcards →GST, polyhistidine, MBP, FLAG or a biotin acceptor peptide fused to the bait so it can be immobilised reproducibly. Each tag brings its own non-specific binding.
How much fusion protein a given bead volume can hold. Offering more than the capacity leaves free bait in solution rather than adding sites.
The tightest non-covalent pair in common use, with a dissociation constant near 10⁻¹⁴ M. Nothing comes off it by competition, so the elution has to be denaturing or the linker has to be cleavable.
Which end of the bait is tethered. A tag at the wrong terminus, or on too short a linker, can occlude the very surface the experiment is testing.
A third protein in the lysate that binds both bait and prey, producing a pull-down band with no direct contact between them.
In a pull-down for an active GTPase, the same prey loaded with GDP. A band that disappears when the nucleotide is swapped is recognition; a band that does not is surface binding.
BSA, detergent and DTT in the binding buffer. They block the bead surface, strip weak non-specific binding and keep a cysteine-containing bait from oxidising on the resin.
When to use it
- Testing whether two proteins bind each other directly, with purified components, after a co-IP has suggested they are in the same complex.
- Mapping the part of a protein that does the binding, by immobilising a series of domains or point mutants at matched density and comparing what each one captures.
- Measuring the activation state of a protein by capturing only the conformation a probe domain recognises, as in a GST–RBD pull-down of GTP-loaded Ras.
- Capturing a modified or tagged protein from a lysate for a downstream assay, when no good antibody exists for an immunoprecipitation.
Compared with alternatives
Uses an antibody against the endogenous protein, so it reports complexes as they existed in the cell; prefer it when physiological relevance matters more than directness.
Gives an affinity and on- and off-rates for the same purified pair instead of a yes-or-no band; the natural follow-up once a direct interaction is established.
Measures affinity and thermodynamics in solution with nothing immobilised, which removes the surface-density artefact entirely.
Works in living cells with no purification at all; prefer it when the question is which proteins are near the bait in vivo rather than what binds it in a tube.
Screens for binary interactions in cells without purified protein, and is a useful independent test of a candidate found on a resin.
The same chemistry used for preparative purification rather than for a binding test; the resin, tags and capacity arithmetic are shared.
Sub-techniques
Choosing one here switches the whole page, including the protocol and the lab, to it.
A glutathione S-transferase fusion is immobilised on glutathione agarose, which is cheap, high-capacity and reversible: 10 mM reduced glutathione competes the bait straight off. GST is a 26 kDa dimerising enzyme with its own interactors, and the cell's own glutathione S-transferases bind the resin, so the GST-only and empty-resin controls both do real work here.
- Reversible capture means the bait can be eluted intact with free glutathione, or left on the resin and the prey released by cleaving the linker.
- GST dimerises, so a GST-fused domain is presented as a dimer, which can create or mask an interaction that needs a monomer.
- A degradation fragment of the fusion runs at 26 kDa, exactly where many small prey proteins run, and the empty-resin lane also carries endogenous glutathione S-transferases at that size.
- The tag is large enough to interfere with folding or activity, so a GST fusion that gives a negative result needs an independent check that the bait domain works.