A nanobody-stem cell platform targeting innate and adaptive immune axis in the tumour microenvironment.

[BACKGROUND] Tumour associated macrophages (TAMs) and exhausted T cells are dominant in the immuno-suppressive tumour microenvironment (TME) and pose a challenge to effective cancer immunotherapy in solid tumours.

[METHODS] In this study, we immunised llamas and generated monovalent and biparatopic nanobodies (Nbs) against colony stimulating factor 1 receptor (CSF-1R) and programmed death receptor 1 (PD-1) to re-educate TAMs and overcome T cell exhaustion, respectively, in the TME. To circumvent short systemic half-life and low peak concentrations of Nbs in the TME, we developed a platform of allogenic off-the-shelf stem cells (SC) releasing biparatopic PD-1 and CSF-1R Nbs and tested its efficacy in different mouse models.

[FINDINGS] Nbs targeting CSF-1R and PD-1 inhibited the CSF1/CSF-1R and PD-1/PD-L1 pathways, respectively, with biparatopic Nbs demonstrating superior efficacy and functionality compared with their monovalent counterparts. Locoregional SC mediated release of biparatopic CSF-1R and PD-1 Nbs, reduced tumour growth by increasing T cell numbers, enhancing T cell activation, and by shifting the macrophage polarisation towards a pro-inflammatory phenotype. Moreover, the presence of both Nbs improved dendritic cells (DCs) activation within the TME. Finally, we show that biocompatible gel encapsulated SC releasing Nbs against PD-1 and CSF-1R have therapeutic efficacy in a highly immunosuppressive glioblastoma model post-tumour resection.

[INTERPRETATION] Taken together, our findings establish a cell-based Nb platform targeting both innate and adaptive immune axis within the TME, which has the potential to facilitate treatment of solid cancers that are otherwise refractory to conventional immunotherapies.

[FUNDING] This study was mainly supported by Institutional Funds (K.S). A part of in vivo Nb characterisation was supported by NIH grant R01-CA285519 (K.S.) and the Nb dye conjugation and modelling was supported by NIH grant R01-AI165666 (M.R).