Microfluidic Avidity-based Autologous Polyclonal T Cell Discovery in Triple Negative Breast Cancer

The identification and selection of anti-tumour T cells that recognise tumour-associated antigens present in patients is a challenge. This is particularly true for cancers that exhibit high degrees of inter- and intra-patient tumoural heterogeneity such as triple negative breast cancer. T cells therapies that are both autologous (derived from an individual's own cells) and polyclonal (express multiple T cell receptors (TCRs)) may be particularly effective because such therapies address the phenotypic diversity of tumoural cells amongst patients while minimising host rejection of transplanted cells such as through graft-vs-host disease. The greatest challenge in developing an autologous polyclonal T cell therapy is the recovery and selection of tumour reactive T cells or their TCRs. Commonly used techniques such as affinity readouts tend to rely on neoantigen prediction methods which can be inaccurate and focus on individual peptide/receptor interactions in isolation which do not accurately ascertain the complex interaction which exist between cells while multimer technology requires validated knowledge of antigens and provides false negative readouts of bulk populations. We have recently developed a microfluidic T cell selection platform that exploits TCR-neoantigen cellular avidity to identify and isolate tumour-reactive T cells from patient samples. This technology is capable of: a) screening the interaction of millions of interacting cancer and immune cells, b) ranking T cells based on their relative TCR avidities against multiple targets simultaneously, and c) recovering tumour reactive T cells that are viable, and suitable for downstream expansion or molecular interrogation (eg through TCR sequencing). This technology has been validated in patient and animal models where T cells were transduced with TCRs of known avidities and recovered by probing their avidity interactions against melanoma cancer cells. While promising, we have not yet demonstrated that this technology is capable of recovering rare tumour-reactive T cells by challenging the avidity of patient-derived lymphocytes against patient-derived cells. Therefore, in this project we propose to: 1) Develop Protocols for Integrating Patient Cells into Microfluidic Avidity. This will involve optimising patient-derived organoid dissociation protocols, device surface coatings, and validation of resistance to shear stress. 2) Develop pipeline for autologous polyclonal TCR discovery. This will be conducted by shear challenging patient matched peripheral blood mononuclear cells and tumour infiltrating leukocytes against tumour-seeded microfluidic devices as described above. T cells will be recovered into buckets, bar coded using cell-hashing techniques (10X Genomics), and their TCR repertoire sequenced. We will then combine the recovered TCR repertoire data and paired ranked affinity metrics to single-cell RNA and TCR sequencing of unsorted patient tumour infiltrating lymphocytes to determine whether the addition of affinity based metrics (generated using these platform) to sequenced based measures of T cell clonotype diversification and expansion can be used to rapidly and efficiently identify anti-tumour T cells. This development of a pipeline for selecting promising autologous polyclonal TCRs will lay the foundation for future work validating and refining a protocol for clinical studies. For instance, in follow on projects we will transduce 5-10 TCRs per patient into into patient derived CD8 T cells or HLA-matched T cell lines for in vitro toxicity testing and validation against banked patient organoids.This will prepare the technology for clinical trials featuring a autologous polyclonal T cells therapy using microfluidic avidity-based selection.