English

Tekbiotech-Yeast and Phage Display CRO, Expert in Nano-body and Antibody Drug Development

Choosing an Affinity Assay for Nanobodies: SPR vs. BLI

The previous article in this series covered nanobody discovery platforms. During development, the affinity of candidate sequences is a key parameter, and there are two main ways to measure it: surface plasmon resonance (SPR) and biolayer interferometry (BLI). Both are biophysical tools widely used to obtain binding kinetics data for bimolecular interactions. Both platforms report association rate, dissociation rate, and affinity constant: by fitting the association time course, the association constant (ka) and dissociation constant (kd) are estimated, and from these the apparent dissociation constant (KD) is derived. But the two methods differ in detection principle, throughput, and sample requirements, so choosing between them matters. Here we compare SPR and BLI from four angles — principle and throughput, sample and buffer, parameter interpretation, and sample submission and pricing.

Principles and Throughput: How SPR and BLI Differ

How BLI works. In BLI, white light is shone down the optical fiber of a biosensor and crosses two interfaces separated by a thin film at the tip: the biocompatible surface at the sensor tip and an internal reference layer. Light reflected from both layers returns to the detector, and the two reflected beams interfere constructively or destructively at different wavelengths. When the sensor tip is immersed in a sample, analyte binds to ligand immobilized on the sensor surface. This binding builds a molecular layer, and as more analyte molecules attach, the layer grows thicker. As the molecular layer at the tip thickens, the surface layer moves effectively further from the internal reference layer and the detector, which shifts the interference pattern of the reflected light (Figure 1a). The change in the spectral pattern depends on the optical thickness of the molecular layer — that is, on the number of molecules adsorbed to the sensor surface. The detector tracks this change and reports it as a shift in wavelength (nm) on the sensorgram (Figure 1b). Multiple layers can bind sequentially to the surface, yielding accurate, comprehensive kinetic data on intermolecular interactions.

8807efdb-ff9c-47c5-89e4-f63184b45add.pngFigure 1 Schematic diagram of BLI affinity detection


3ffb643c-f456-47af-851b-17b912036732.png

Figure 2: Experimental principle of binding kinetics for BLI affinity detection based on streptavidin biosensor


How SPR works. SPR spectroscopy relies on total internal reflection of linearly polarized light at the interface between two media of different optical density — a thin metal film and a liquid, for example. At a specific angle (the resonance angle), incident light couples with free electrons oscillating within the thin metal film, generating a non-radiative evanescent electromagnetic wave parallel to the film surface and thereby exciting surface plasmons. The resonance angle is highly sensitive to the refractive index of the lower-optical-density medium — usually a liquid buffer — at the metal film–liquid interface. That refractive index is influenced by the properties of the buffer, including pH, salt concentration, and viscosity, as well as by changes in surface mass. By continuously monitoring shifts in the resonance angle, binding events between molecules in solution (analyte) and molecules immobilized on the metal film (ligand) can be observed directly, with the signal recorded in resonance units (RU).

The difference in practice. SPR depends on refractive index changes at a gold chip surface: ligand is immobilized on the chip, analyte flows through a microfluidic channel, binding events alter the surface refractive index, and the instrument records the change in real time in resonance units. BLI immerses the tip of a fiber-optic sensor in the sample; a change in the thickness of the biomolecular layer shifts the interference spectrum of the reflected light, recording the phase shift caused by mass migration. This difference shapes the two methods' throughput and their tolerance for dirty samples. SPR microfluidics are precise but delicate — a single particle or an air bubble can clog the channel, so sample clarity requirements are strict. BLI is dip-and-read with no flow path, and sensors can be shaken in parallel in batches, so throughput is inherently higher than SPR; crude supernatants and cell culture supernatants often give usable signals on BLI. But when it comes to precise measurement of candidate molecules and data destined for regulatory submission, SPR's sensitivity and data acceptance remain the gold standard — especially for very slow-dissociating, high-affinity molecules, where SPR's baseline stability and fitting confidence are more reliable.

Setting Sample Purity, Buffer, and Ligand Immobilization

Sample purity. SPR usually requires macromolecule purity of at least 90% at a concentration of 0.5 mg/mL or above; small molecules need 10 mM or more. BLI is far more forgiving on purity, but impurities occupy sites on the sensor surface, lowering the effective ligand density and distorting the kinetic constants. If a sample contains glycerol, sucrose, high concentrations of imidazole, or reducing agents, coupling efficiency and baseline stability on an SPR chip will be poor, and buffer exchange is mandatory.

Buffer. In SPR, the analyte diluent and the running buffer must belong to the same system; otherwise the refractive index difference at injection creates a spurious bulk shift, visible on the sensorgram as a step at the start and end of injection. That forces extra parameters into the kinetic fit and undermines the rigor of the KD determination. For BLI, adding 0.02% Tween-20 to the buffer is almost mandatory to suppress non-specific binding — but too much Tween interferes with wetting of the fiber surface, so a pilot titration is needed.

Ligand immobilization. For ligands carrying an Fc or His tag, capture with Protein A, anti-Fc, or NTA chips is more convenient and provides orientation control. If the ligand is unstable or requires repeated regeneration, a biotinylated ligand on an SA chip or streptavidin sensor is the robust option.

Reading ka, kd, and KD — and What Counts as Good Data

Fitting a sensorgram yields three numbers: ka (association rate constant, M^-1 s^-1), kd (dissociation rate constant, s^-1), and KD (equilibrium dissociation constant, KD = kd/ka, in M).

ka reflects how fast binding occurs. For antibody–antigen interactions, a ka in the 10^5 M^-1 s^-1 range is normal to fast; above 10^6, check for non-specific adsorption or mass transport limitation.

kd reflects how fast the complex falls apart. The smaller the kd, the harder it is for the complex to dissociate. A molecule with KD = 1 nM built from ka = 10^5 and kd = 10^-4 binds slowly but dissociates extremely slowly — well suited to settings that require prolonged target occupancy. Conversely, ka = 10^6 with kd = 10^-3 also gives 1 nM, yet the dynamic profile is completely different. Which one to keep depends on the biological question.

What counts as good data. Do not look only at the absolute KD; look at the quality of the fits for both the association and the dissociation phase:

l Residuals between the fitted curve and the raw data are randomly distributed, with no systematic bias.

l Sensorgrams at different concentrations separate clearly in the dissociation phase instead of collapsing into one cluster.

l KD deviation across replicate concentration points (at least one concentration run in duplicate) stays within 30%.

l Mass transport limitation is ruled out: if ka falls systematically as ligand density rises, binding is diffusion-controlled rather than reflecting true kinetics, and the reported KD is unreliable.

Sample Submission and How Pricing Works

There is no standard price list for affinity measurement, but the logic behind a quote follows a pattern.

Where SPR costs go. The big items are chips and instrument time. An amine-coupling chip alone costs from several hundred to over a thousand RMB, and once method development, sample testing, and regeneration scouting are added, a complete kinetic characterization of one antibody–antigen interaction (pilot plus formal experiment) typically falls in the range of several thousand to tens of thousands of RMB at most service providers.

Where BLI wins on cost. BLI sensors cost less than SPR chips, and the throughput advantage of dip-and-read analysis means batch samples dilute the per-sample cost; at the early screening stage, BLI is usually the friendlier option when quotes are given per sample.

Before you ship samples, check:

l Whether the amount covers pilot-run losses (SPR: no less than 100 μg of macromolecule at >0.5 mg/mL; BLI can be relaxed somewhat but should not drop below 50 μg).

l Whether the buffer composition triggers a buffer-exchange surcharge.

l Whether active-concentration determination of the analyte is required (billed separately).

l If the project is at the regulatory submission stage, go straight to SPR and budget for reproducibility verification — that is more economical than starting with BLI to "save a little" and then having to redo the work.

About Tek Biotech (Tianjin) Co., Ltd.

Tek Biotech (Tianjin) Co., Ltd., working from its nanobody candidate drug development platform, offers a broad panel of assays for early developability assessment of candidate sequences: in vitro affinity validation (EC50 determination, SPR affinity measurement, BLI affinity measurement, flow cytometry binding validation), blockade validation (competitive ELISA, flow cytometry-based blockade), and in vivo validation (animal model studies, in vivo imaging). Together these provide the technical and data support behind our clients' drug development programs.

References

Li K, Huntwork RHC, Horn GQ, et al. TitrationAnalysis: a tool for high throughput binding kinetics data analysis for multiple label-free platforms. Gates Open Research. 2024;7:107.

Kumaraswamy S, Tobias R. Label-free kinetic analysis of an antibody–antigen interaction using biolayer interferometry. Methods in Molecular Biology. 2015;1278:165-182.

Kononenko A, Caroprese V, Duhoo Y, et al. Evolution of multivalent supramolecular assemblies of aptamers with target-defined spatial organization. Nature Nanotechnology. 2025;20:1087–1097.

Related recommendations

To experience the reliable service of Tekbiotech please subscribe: