Development of Novel Invariant TCR-Anchored Bispecific Engagers to Enhance Anti-Cancer Activity of iNKT Cells

iNKT are rare, CD1d-restricted, glycolipid-reactive T cells that bridge adaptive and innate immunity. iNKT cells are characterised by an invariant TCRVa24Ja18 chain nearly always pairing with a TCRVb11 chain (iTCR). The ability of allogeneic iNKT to protect from aGVHD makes them ideal candidates for ‘off-the-shelf’ immunotherapy without need for TCR deletion.

Indeed, CAR-iNKT show promise in pre-clinical models of blood cancer as well as early clinical development against B cell lymphoma and outperform CAR-T counterparts.

Here we aimed to enhance the modularity of adoptive iNKT cell immunotherapy for blood cancers by designing a bispecific iNKT cell engager (biNTe) in which one arm would engage the iTCR while the other arm would engage tumour antigen(s) of interest.

As a proof-of-principle we designed and synthesised a mono- and a bi-valent biNTe, one each, against the clinically validated myeloma target BCMA with purities of 90% and 70% respectively.

We focused on the bivalent biNTe which we first used to demonstrate its iNKT cell binding specificity. Indeed, in flow-cytometry assays, biNTe followed by fluorescent anti-human Fc Ab, stained purified iNKT and but not T cells while in the presence of the same iTCR-binding mAb used to generate the biNTes, binding of the bivalent biNTe on iNKT cells was abrogated. This shows the iTCR engaging specificity of the bivalent biNTe. To demonstrate BCMA specificity of the other arm of the biNTe, we used it to stain iNKT cells, followed by staining with biotinylated soluble BCMA and fluorescently labelled streptavidin. This, compared to staining without soluble BCMA, showed strong staining of iNKT cells thus validating the BCMA specificity of the bivalent biNTe.

Next, we set up co-cultures of pre-expanded iNKT against the BCMA-expressing and -knockout myeloma cell line MM1.S in the presence of varying concentrations of the biNTe. While, as assessed by a flow-cytometric cytotoxicity assay, only background killing of myeloma cells was observed in the absence of biNTe or when BCMA ko cell were used as targets, in the presence of biNTe, BCMA+ MM1S cells were killed, in a dose dependent manner (Figure 1A). Consistent with iNKT specificity of the biNTe similar co-cultures involving T cells showed only background killing of MM1S cells (Figure 1B). The same results were obtained using the BCMA-expressing myeloma cell line H929 and its corresponding BCMA knockout. The overall estimated EC50 for MM1.S and H929 cells were 6.56 and 3.4 nM respectively. Using the EC50 concentrations we show that biNTe very effectively kill and almost eliminate patient-derived bone marrow myeloma plasma cells (n=3).

To test whether engagement of the iTCR by biNTe would impact its ability to engage with CD1d, we co-cultured iNKT and MM1.S cells transduced with CD1D in the presence of biNTe and the activating, iNKT-specific, CD1d-binding glycolipid ligand a-galactosylceramide (aGalCer). We found that compared to no biNTe control and at biNTE concentrations lower than the EC50, killing of myeloma targets cells was consistently increased. This suggests that iTCR retains its ability to engage CD1d and the efficacy of future biNTe-based immunotherapy could be enhanced by aGalCer.

While in vivo validation experiments are ongoing, we conclude that biNTe constitute a novel therapeutic approach with notable pre-clinical anti-myeloma activity. We envisage that they can be used in conjunction with other therapeutic modalities such as CAR and iTCR to enhance the modularity and the efficacy of iNKT cells as an ‘off -the-shelf’ platform for the treatment of myeloma, other blood or solid tumour cancers.