Why Most Pulldowns Cost You the One Thing You Wanted to Keep

Team Dynamic Matrices | 26 November 2025

You pull down a protein because you want to know what it does: its activity, its partners, its state in the cell. Then you elute it by boiling in SDS or dropping the pH to 2.5, and what comes off the beads is a denatured chain. By the time the activity assay runs, you are measuring what survived purification, not what the protein was doing. We think this is the most under-discussed compromise in affinity capture, and it is the reason we built a release step that does not require it.

The tension built into affinity capture

Capture has to bind tightly enough to survive washes that remove contaminants. Release has to break that bond. The tighter the capture, the harsher the release. The numbers:

• Biotin-streptavidin has a dissociation constant around 10⁻¹⁴ to 10⁻¹⁵ M, among the strongest non-covalent interactions known (Green, Methods in Enzymology, 1990). Breaking it needs boiling in SDS, 8 M guanidine, or prolonged formamide. None is compatible with a folded protein.
• Antibody-antigen interactions run 10⁻⁸ to 10⁻¹² M. Elution is 0.1 M glycine at pH 2.5 to 3, or SDS sample buffer, or competition with a peptide when one exists. Low pH unfolds many proteins and dissociates multi-subunit complexes.
• Ni-NTA (His-tag) elutes with imidazole under mild conditions, which is why it is the go-to for recombinant work, but it requires an engineered tag and does not work on endogenous protein.
• FLAG and Strep-tag II elute by competition with FLAG peptide or desthiobiotin under native conditions; these are the gentle options, again only for tagged proteins.

For an endogenous protein, or a complex from a native lysate, the standard choices are denaturing.

What you lose

Three things, and they compound.

Catalytic activity. Enzymes assayed after denaturing elution and refolding show a fraction of native activity, and the fraction is unpredictable. Kinetic constants measured on such preparations are not the protein's.

Complex integrity. Multi-subunit enzymes and allosterically regulated proteins fall apart under SDS or low pH. Elute a proteasome or a polymerase holoenzyme by boiling and you have subunits.

Transient and weak interactions. The interactions with the most biological information are often the weakest. Stringent washes and harsh elution lose them first, and the inter-lab variation in "how stringent" is a large part of why interactome studies disagree.

How phase-separation capture avoids the trade-off

LASSO does not use a high-affinity bond between bait and support. A DNA-based capture element (an aptamer, or a DNA strand hybridised to a tag on an antibody or nanobody) is carried on a soluble polymer. When the capture element binds its target and neighbouring polymer chains crosslink through DNA hybridisation, the network undergoes a phase change and condenses into a pellet in solution. Contaminants stay in the supernatant. The target is held by the network, not by a 10⁻¹⁵ M bond.

Release is by addition of a complementary strand that displaces the crosslink, or by a small change in ionic strength or temperature that reverses the phase transition. Both happen in the same physiological buffer used for capture: no acid, no SDS, no denaturant. The protein comes out folded, in its complex, at the state it was captured (Dynamic Matrices, Angewandte Chemie, 2025; patent WO 2025/196123A1).

What this makes possible

• Kinetic assays straight after pulldown, with no refolding step and no recovery period, because activity was never lost.
• Intact complexes for structural work or for mapping subunit composition by native mass spectrometry.
• Reproducible interactome data, because release stringency is defined by a DNA sequence rather than by an operator's judgement of how long to wash.
• Reagent stability. The capture chemistry is DNA and polymer, stable for about a year at 4 °C, rather than an antibody batch.

Where it does not replace the alternatives

If you only need to identify what was bound (mass spec after denaturing elution), a biotin or antibody pulldown is validated and fine. If your target has a His tag, Ni-NTA is cheap and gentle already. LASSO earns its place when the protein has to come back alive.