Interview with Dr. Fabian Vandermeers and Dr. Arnaud Delobel
"Affinity determination for TCR-based cell products"
Immudex and Quality Assistance S.A. have partnered to provide T cell receptor (TCR) affinity measurement services to the pharmaceutical industry, combining Immudex’s expertise in manufacturing high-quality TCR and peptide-MHC monomers with Quality Assistance’s advanced analytical capabilities.
This collaboration supports the development of T cell-based immunotherapies by offering reliable assays to determine TCR affinity and kinetics.
In this interview, we speak with R&D and Innovation Director Dr. Arnaud Delobel and R&D Technical Leader Dr. Fabian Vandermeers, who spearheaded a study of methods to evaluate TCR affinity at Quality Assistance. Together, they have nearly 30 years of experience in analytics development at the company.
What gap in cell therapy analytics led you to examine methods to quantify the affinity of TCR-engineered T cells?
Our project aimed to study the interaction between the TCR and the MHC-peptide complex. As the discovery and development of novel TCR-based cell therapies gains traction, especially in the field of cancer research, we anticipate mounting interest in a tractable and reliable way to assess and demonstrate the affinity of this interaction.
Currently, the analyses required in the dossier of a cell therapy application to regulatory bodies are loosely defined. Guidelines are very high level, and the battery of tests to include in an analytical package emerges from an exchange with authorities about the specific product. However, the potency of a TCR-based cell therapy relies on the affinity of the product. If it’s too low, you have no potency; too high can also be an issue.
Thus, a method to measure that affinity and ensure it remains within a defined window may prove critical. Also, potency assays are not always easy to develop for products with complex gene engineering. Having a surrogate test based on affinity can be useful.
You evaluated surface plasmon resonance (SPR) and biolayer interferometry (BLI) as methods to measure the affinity of TCR-pMHC binding. Why those two technologies?
For one, these are the main techniques used today to measure affinity. But also, as a leader in analytical sciences, Quality Assistance has over four decades of experience providing FDA and EMA-required evidence for innovative human medicinal products. So, our teams are skilled in these methods and understand how to optimize assay conditions and tweak assay design.
Our goal was to determine the kinetics of the interaction – measure the association and dissociation constants – to then calculate the affinity as the KD value. We chose to work with a TCR targeting NY-ESO-1. In itself, the exact target was less important to our project since we aimed for proof of concept of our methods. But NY-ESO-1 proved fortuitous for three reasons. First, it is a target of interest in cancer research.
Being immunogenic and restricted to cancer and germ cells, NY-ESO-1 is being pursued as a high-potential target in cancer therapies. Second, the TCR and corresponding MHC complex were available from Immudex. So sourcing was easy. Third, the affinity of the TCR was in a range that allowed cost-efficient reagent use. In SPR and BLI, the reagent concentration used in the assay is related to the expected affinity. Typically, we analyze samples between 0.1 and ten times the KD in SPR and BLI experiments.
So, using a TCR with a nanomolar-range affinity means working with nanomolar reagent concentrations, whereas a millimolar-range affinity requires large amounts of TCR and MHC complex. Not only is the sizeable reagent consumption very expensive, but high concentrations are also difficult to produce. For example, high concentrations can lead to formulation issues as protein precipitation.
You mentioned adjustments to the assay. Could you provide some details?
We encountered some issues during the project. One was the correct design of the assay. In SPR and BLI, one of the two interactants is immobilized, and the second is applied in solution. So, two options exist: to immobilize the TCR or the MHC complex.
First, we tried immobilizing the MHC complex on the SPR chip. Upon injecting the ligand MHC, however, we observed dissociation of the MHC complex. The complex is a heterodimer, and the two subunits are not covalently bound. As a result of their dissociation, there was a weak to no signal when TCR was added. So, we reversed the design to immobilize the TCR instead. With that architecture, we faced another issue.
The TCR and MHC complex we used were both tagged with biotin. We used the biotin to bind the TCR to the streptavidin-coated sensor chip or biosensor. But then, of course, the biotin-tagged MHC complex bound directly to the chip. We tried different options to solve the problem, like saturating the chip with the biotin analog biocytin.
In the end, we used a more elegant solution. We obtained a histidine-tagged version of the MHC complex, and that final setup provided the best data.
After all the testing and setup permutations, did one method outperform the other?
If you compare the SPR and BLI results, BLI performed better. We observed very good association and dissociation curves, and the data fit our mathematical model very well (Fig. 1). Outcomes with SPR were slightly less reliable.
At this point, BLI with an immobilized TCR and an MHC complex analyte using a non-biotin tag is our method of choice. Of course, each product analyzed with this design will call for method optimization. The amount of TCR immobilized on the chip or biosensor must be evaluated, as well as the regeneration conditions and concentration of the analyte injected.
Also, the time of injection and time of dissociation may need adjustment depending on association and dissociation rates. If we have a TCR that associates very fast, we don't need to inject for 400 seconds, for example. Twenty or 30 seconds will suffice.
In contrast, injection and dissociation times will need to be longer for TCRs with a low association or dissociation phases. These parameters may need optimization for a new TCR, but the main assay design is set.

Figure 1. TCR-pMHC interaction assessment using (A) SPR and (B) BLI with a 1:1 fitting model. The fit is closer and more reliable for the BLI-based assay.
What advice would you give researchers considering the assay to measure the affinity of a TCR-based therapy?
As with any other assay, SPR and BLI have upper and lower bounds in measurement capacity. That means they are methods that work best with TCR affinities in the nanomolar to micromolar range. With a higher affinity product, like in the picomolar range, no dissociation occurs within the time limits of the experiment.
The interaction is too strong. As a result, a dissociation constant and, thus, KD value cannot be determined. We briefly discussed the other end of the spectrum. If the affinity of the TCR is too low, a high concentration is needed for each experiment, which can be prohibitively expensive or may result in formulation issues.
We advise developers to have a rough idea of the KD value of their product to serve as a starting point for assay optimization.
In our experience, most cell therapy developers are interested in evaluating high- to medium-affinity TCRs.
Did you assess the specificity of the assay?
We evaluated assay performance using an MHC complex with a completely divergent peptide sequence. Of course, there was no measurable interaction. Our next experiment will use an MHC complex with a peptide sequence that differs by only one or two amino acids and a slightly different TCR.
Could the assay be adapted to evaluate the affinity of other constructs of interest, like TCRs that recognize and bind MR1?
As long as the interaction is similar and the molecular weight of the interactants is within a comparable range, it should be possible to adapt the assay quite easily. That would be interesting to test.
Further Resources
For more details about the study:
Learn more about our Partnership with Quality Assistance on TCR Affinity Measurement Services: https://www.immudex.com/about-us/news/partnership-on-tcr-affinity-measurement-services-announced-by-immudex-and-quality-assistance/
Learn more about Quality Assistance: https://www.quality-assistance.com/
Explore Soluble TCRs from Immudex: https://www.immudex.com/services/custom-solutions-and-services/tcr-solutions/
Request non-biotinylated pMHC Monomers from Immudex: https://www.immudex.com/services/custom-solutions-and-services/