There is no single best method for measuring peptide binding. The right choice depends on what you need to know, how much material you have, whether you can label or immobilize a partner, and how tight the interaction is expected to be. This page compares the methods researchers most often use for peptides, with the practical trade-offs that are easy to discover too late.
| Method | What you get | Label or surface needed | Typical affinity window | Material needed |
|---|---|---|---|---|
| SPR Surface plasmon resonance |
ka, kd and KD in real time | One partner immobilized on a chip | Very wide, pM to mM with care | Low; microgram quantities |
| BLI Biolayer interferometry |
ka, kd and KD in real time | One partner on a biosensor tip | Roughly nM to high µM | Low, but higher sample volume per well |
| ITC Isothermal titration calorimetry |
KD, stoichiometry and binding enthalpy | None; free solution | Roughly nM to tens of µM | High; the main limitation |
| FP / FA Fluorescence polarization |
KD, and IC50 in competition mode | Small fluorescent tracer peptide | Bounded below by the tracer concentration | Low; plate format |
| MST Microscale thermophoresis |
KD | Fluorescent label, or intrinsic fluorescence | Wide, pM to mM | Very low volumes |
| ELISA Peptide-coated plate |
Relative binding, EC50 | Peptide coated or captured; detection antibody | Semi-quantitative | Low; very high throughput |
| Pull-down Biotinylated peptide and beads |
Which proteins bind; usually qualitative | Biotin tag and streptavidin support | Not an affinity measurement | Moderate |
| Ligand-observed NMR STD, CPMG, WaterLOGSY |
Binding yes/no, and binding epitope on the ligand | None, but needs a spectrometer and expertise | Best for weak binding, µM to mM | High peptide, moderate protein |
If the question is whether a peptide binds fast and lets go slowly — which matters for inhibitor work — you need a real-time method: SPR or BLI. If a single equilibrium constant answers the question, solution methods such as ITC, FP or MST avoid the complications of a surface.
SPR, BLI, ELISA and pull-downs all require one partner attached to something. FP and MST require a fluorophore. ITC and NMR require neither, which is why they are so useful as orthogonal confirmation.
ITC is the most informative single experiment and the most demanding: it typically needs hundreds of microliters at micromolar concentrations, and a matched buffer for both partners. If protein is scarce, it is usually the method you drop.
Many isolated peptides bind in the high micromolar to millimolar range. FP struggles there, ITC runs out of window, and surface methods suffer from bulk refractive index effects. Ligand-observed NMR is built for exactly this range.
For peptides, which partner you immobilize matters more than anything else. SPR responds to mass arriving at the surface, so flowing a 2 kDa peptide over an immobilized protein gives a small signal. Immobilizing the peptide and flowing the protein usually gives a far better response.
The common artifacts are avidity and non-specific binding. If a dimeric or multivalent protein binds a densely coated peptide surface, the apparent affinity will be tighter than the true 1:1 value and the dissociation will look artificially slow. Keeping surface density low and checking that the measured KD does not change with density is the standard control. Peptides rich in arginine and lysine also bind carboxymethylated dextran surfaces non-specifically; a reference surface and a buffer with modest salt help.
BLI is more tolerant of crude or coloured samples than SPR and is easier to run in plate format, which makes it attractive for screening panels of peptides. It is generally less sensitive for small analytes and less able to resolve very fast kinetics. The same avidity considerations as SPR apply. Because the sensor moves between wells, evaporation and well-to-well variation deserve attention in long runs.
ITC is the closest thing to a direct measurement: no label, no surface, and it reports stoichiometry and enthalpy alongside affinity. Stoichiometry is a useful sanity check that is unavailable in most other formats. Its weaknesses are appetite for material and sensitivity to buffer mismatch; dialysing both partners into the same buffer is not optional, and heats of dilution must be measured and subtracted. Peptides stored as TFA salts can produce a buffer artifact, so counter-ion exchange is worth considering.
FP works when a small labeled peptide tumbles quickly in free solution and much more slowly once bound to a larger partner, so the size difference between tracer and target needs to be substantial. It shines in competition mode: establish a labeled tracer and target pair once, then rank unlabeled competitors cheaply in a plate.
The measurable affinity is limited by the tracer concentration, so a direct KD far below the tracer concentration cannot be determined accurately. Fluorophore choice is not neutral either: a hydrophobic dye can contribute to binding or cause the tracer to stick to surfaces.
MST needs very little material and tolerates complex buffers, including serum and lysate in many cases. It is a good choice when the sample is precious or messy. Results are sensitive to aggregation and to adsorption on the capillary, so the controls that matter are a check that the fluorescence of the labeled partner does not change on binding, and visual inspection of the capillary scan.
These are discovery and screening tools rather than affinity measurements. A peptide ELISA is excellent for ranking a library or checking that an antibody recognizes its antigen, but a plate-bound peptide is multivalent and the readout usually involves an avidity-prone secondary antibody, so numbers should be treated as relative. A biotinylated-peptide pull-down from lysate answers "what binds this motif" and is a good partner to mass spectrometry; it says little about affinity.
Confirm in a second format
Methods fail in different ways. A surface method can report avidity as affinity; a fluorescence method can report dye behaviour as binding; a calorimetry experiment can report a buffer mismatch as heat. Two methods that disagree are telling you something useful. Two that agree are worth far more than one that is very precise.
Ordering the controls in the same batch as the test peptides is worth doing. They are made with the same chemistry and analyzed the same way, which removes one source of doubt when the result is surprising.
Paste a peptide or protein sequence to review length, charge, hydrophobicity and other properties before you order.
Research use only. LifeTein peptides and proteins are supplied for laboratory research use. They are not for human or veterinary use, and not for diagnostic or therapeutic use. See our research use policy.
Tell us which assay you plan to run and we will advise on tag position, spacer, purity and controls before synthesis. Email sales@lifetein.com or use our quotation form.