An NK cytotoxicity assay across four donors produces four different curves. The vials matched on CD56 purity and post-thaw viability, the operator was the same, the target line was the same. The usual conclusion is that something went wrong with handling. Occasionally it did. More often the answer was determined before the cells ever reached the lab, by which inhibitory receptors those donors happen to carry.
What KIR receptors do
Killer immunoglobulin-like receptors sit on the NK cell surface and read HLA class I on whatever the NK cell is inspecting. Some are inhibitory and some are activating, and the balance between the signals they generate determines whether the NK cell engages or moves on.
The inhibitory arm is the more consequential in practice. Healthy nucleated cells display class I, inhibitory KIR read it, and killing is restrained. When a cell loses or downregulates class I, which is what many transformed and virally infected cells do to hide from T cells, that restraint disappears. This is the missing-self principle, and it is the reason NK cells and T cells are complementary rather than redundant.
Why donors differ
KIR genes are polymorphic and, importantly, variable in gene content. Individuals do not all carry the same set of KIR genes. That is a different kind of variation from HLA, where everyone carries the same loci with different alleles, and it means two donors can differ not just in which version of a receptor they express but in whether they express it at all.
Layered on top is the interaction with the donor’s own HLA, since the education a developing NK cell receives depends on the class I ligands it encounters. The consequence is that donor NK function is a property of the KIR and HLA combination rather than of either alone.
Where this shows up
| Situation | What is usually blamed | What is worth checking |
|---|---|---|
| Donor-to-donor variance larger than expected | Thaw technique, rest period | KIR gene content and donor HLA |
| One donor kills a target line others ignore | Assay artefact | Class I expression on the target against donor inhibitory repertoire |
| Poor reproducibility across lots | Supplier consistency | Whether lots came from genotypically different donors |
| An engineered construct underperforms in some donors | Construct design | Whether the backbone donors were selected on genotype |
None of these are exotic. They are ordinary results, and the common feature is that the explanation was available in advance if characterization had been done up front rather than reconstructed afterwards.
Selecting on it
Once genotype is known across a donor pool, it becomes a selection criterion rather than a post-hoc explanation. A program can hold genotype constant to reduce variance, deliberately vary it to understand a construct’s sensitivity, or build a panel spanning a defined range so that a result generalises rather than describing one donor.
What is not possible is applying any of that retrospectively. Genotype determined after the experiment explains the result. Genotype determined before it designs the experiment.
What it changes for engineered programs
Hypoimmunogenic cell designs frequently remove class I to evade T cell recognition, and in doing so create a cell that presents as missing-self to NK cells. This is why such designs increasingly retain HLA-E or HLA-G, preserving an inhibitory signal so the cell escapes T cells without becoming an NK target. Evaluating whether that strategy works requires NK effectors whose inhibitory repertoire is known, otherwise the readout is uninterpretable.
For CAR-NK work the same logic applies to the backbone. A chimeric receptor supplies an activating signal, but the inhibitory arm remains active underneath it, and a donor whose repertoire is strongly inhibited against your target is a poor backbone regardless of construct quality.
Getting material with the genotype attached
OrganaBio includes KIR genotyping in its donor characterization program alongside high-resolution NGS HLA genotyping across HLA-A, HLA-B, HLA-C, HLA-DR, HLA-DQ and HLA-DP, applied at program level rather than per order. That is what allows donor selection by genotype, subject to availability, and repeat collection from eligible donors so a panel can be returned to.
NK material is documented as CD56+ NK cells from peripheral blood at 5 or 10 million cells per vial with post-thaw viability listed at 85% or above, and CD56+ NK cells from cord blood at 1 or 2.5 million cells per vial with viability and CD56+ purity each listed at 90% or above. Broader NK biology sits in the NK cell guide.
Frequently asked questions
What is KIR genotyping?
It determines which killer immunoglobulin-like receptor genes a donor carries. These receptors sit on NK cells and read HLA class I on target cells, with the balance of inhibitory and activating signals determining whether the NK cell engages.
Why do NK donors with identical purity and viability behave differently?
Because purity and viability describe the contents of the vial while KIR and HLA genotype describe how those contents behave. Donor NK function is a property of the KIR and HLA combination, and neither is visible in a viability figure.
How is KIR variation different from HLA variation?
HLA varies by allele at loci everyone carries. KIR varies in gene content as well, so donors can differ in whether they carry a given receptor at all, not merely in which version they express.
What is the missing-self principle?
Healthy nucleated cells display MHC class I, which inhibitory KIR read as a signal to hold back. When a cell loses or downregulates class I, as many transformed and virally infected cells do to evade T cells, that inhibition disappears and the NK cell is released to act.
Can KIR genotype be determined after an experiment?
It can, and it will explain the result. What it cannot do retrospectively is inform the design. Genotype known in advance lets you hold it constant, vary it deliberately, or build a panel spanning a defined range so a result generalises.
Why does KIR matter for CAR-NK development?
A chimeric receptor supplies an activating signal, but the inhibitory arm continues operating underneath it. A donor whose inhibitory repertoire is strongly engaged against your target makes a poor backbone regardless of how good the construct is.
Why do hypoimmunogenic designs retain HLA-E or HLA-G?
Removing class I evades T cell recognition but creates a cell that reads as missing-self to NK cells. Retaining HLA-E or HLA-G preserves an inhibitory signal so the cell escapes T cells without becoming an NK target. Evaluating that requires NK effectors with a known inhibitory repertoire.
Working through this on a live program?
The scientific team works through sourcing and specification questions with cell therapy and research groups directly, including donor characterization, format selection and documentation scope.
