Affinity of PD-1/PD-L1 interaction—a comparison between SPR, MST, ITC and MDS

Sebastian Fiedler1, Monika Piziorska1, Haris Choudhery1 & Sean Devenish1

The binding affinity of PD-1/PD-L1, determined by microfluidic diffusional sizing (MDS) on the Fluidity One-W, is in good agreement with values obtained by SPR, MST and ITC. Further, MDS analysis allowed the absolute size of PD-1 and PD-1/PD-L1 complex to be measured, from which the stoichiometry of the complex was inferred. Thus, the Fluidity One-W can determine binding affinity and stoichiometry of protein interactions without prior knowledge of the structure of the protein complex.


The PD‐1/PD-L1 interaction functions as an immune checkpoint that prevents autoimmune responses. This immunosuppressive effect of PD-1/PD-L1 is employed by many types of cancer cells to evade recognition and destruction by T-cells. Consequently, inhibition of this interaction is a major target in cancer immunotherapy, and several monoclonal antibodies are being used successfully in the clinic against a variety of cancers (1).

Programmed cell death protein 1 (PD-1) is expressed on many types of immune cells including T-cells, B-cells and macrophages. When programmed death ligand 1 (PD-L1) binds to PD-1, regulatory T-cell apoptosis is suppressed while apoptosis of antigen-specific T-cells is promoted.

Here, the binding affinity (KD) between PD-1/PD-L1 was measured using microfluidic diffusional sizing (MDS) on the Fluidity One-W. This KD was compared with literature values from other biophysical methods: surface plasmon resonance (SPR), microscale thermophoresis (MST) and isothermal titration calorimetry (ITC).

In addition to measuring binding affinity, the Fluidity One-W reports the absolute size of free PD-1 and PD-L1 as well as the PD-1/PD-L1 complex. These data were used to infer the stoichiometry of the protein complex.


Sample preparation

PD-1 (R&D Systems) was reconstituted into PBS at pH 7.4 resulting in a final concentration of 24 µM, diluted into labelling buffer (0.2 M NaHCO3, pH 8.3) and mixed with Alexa Fluor™ 488 NHS ester (Thermo Scientific) at a dye-to-protein ratio of 3:1. The protein–dye mixture was incubated at 4 °C overnight and purified with a 1 mL Pierce® Desalting Column (Thermo Scientific) using PBS-T (PBS with 0.05% Tween 20) at pH 7.4 as a buffer.

PD-L1 (R&D Systems) was reconstituted into PBS (pH 7.4) at a final concentration of 250 μM. To determine Rh of unbound PD-L1, PD-L1 carrying a C-terminal His tag was mixed with HIS-Lite™ OG488-Tris NTA-Ni complex (Stratech) in PBS-T buffer (pH 7.4) giving final concentrations of 2 μM and 1 μM, respectively. The mixture was incubated for 15 min at 4 °C and diluted to a PD-L1 concentration of 400 nM. Samples were measured on the Fluidity One-W at the medium flow rate in triplicate.

Experimental protocol

Alexa Fluor™ 488-labelled PD-1 was measured on the Fluidity One-W at 500 nM using the medium flow-rate setting to obtain the size of the unbound protein. To obtain binding curves, 500 nM Alexa Fluor™ 488-labelled PD1 was mixed with unlabelled PD-L1 samples between 0–225 μM and equilibrated at 4 °C overnight. Samples were then measured on the Fluidity One-W using the medium flow-rate setting and the KD was automatically determined by the Fluidity One-W utilizing non-linear least squares fitting to Equation 1 (see Appendix). 


The absolute sizes (hydrodynamic radii, Rh) of unbound PD-1, unbound PD-L1 and the PD-1/PD-L1 complex were measured on the Fluidity One-W, allowing the stoichiometry of the interaction to be inferred.

As depicted in Figure 1, the experimental Rh of the PD-1/PD-L1 complex is shown to be consistent with a hypothetical Rh derived from Rh values of the unbound binding partners—showing that the interaction has a 1:1 binding stoichiometry. Due to glycosylation, the individual Rh values of both unbound PD-1 and unbound PD-L1 are considerably larger than expected based on their nominal molecular weights (UniProt: Q15116 and Q9NZQ7) (2).

Figure 2 shows an equilibrium binding curve obtained by the addition of unlabelled PD-L1 to a constant concentration of Alexa Fluor™ 488-labelled PD-1. From this, the KD of the PD-1/PD-L1 interaction was determined to be 4 μM which agrees well with literature data obtained using other biophysical methods (Table 1).

ap14 fig 1

Figure 1: Absolute size measurement of PD-1, PD-L1 and complex. Experimentally determined hydrodynamic radii (Rh) of PD-1, PD-L1 and the PD-1/PD-L1 complex. Here, the PD-1/PD-L1 experimental Rh is consistent with the hypothetical Rh of a 1:1 complex derived from absolute size measurements of the individual proteins.

ap14 fig 2

Figure 2: Binding curve for unlabelled PD-L1 to 500 nM Alexa Fluor™ 488-labelled PD-1. All measurements were performed in triplicate. KD was determined by non-linear least squares fitting of the data with Equation 1 setting the stoichiometry parameter to n = 1.

Table 1: A comparison of binding affinity values for PD-1/PD-L1 from a range of biophysical techniques

Ap14 table 1


Using the Fluidity One-W, the KD of the PD-1/PD-L1 interaction was accurately determined and agrees with literature values derived from ITC, SPR and MST.

Crucially, because the Fluidity One-W performs absolute size measurements in solution, it provides key information on the quality of proteins as well as the stoichiometry of protein complexes. This type of data gives researchers vital additional information about the protein complexes they are studying without the need for detailed structural information.

Appendix notes available in PDF below.

Download PDF