Augmenting anti-tumour T cell responses using innate-like T cell ligands

Abstract

Immunosurveillance is critical for the detection and elimination of malignant cells. Activated CD8+ T cells are fundamental to anti-tumour immune responses due to their ability to specifically target peptide antigens, including neoantigens, derived from tumour cells, and eliminate tumour cells in response.

The activation of naïve CD8+ T cells by antigen presenting cells (APCs) requires presentation of peptide antigens alongside co-stimulatory molecules. APCs upregulate co-stimulatory molecules in response to additional immunostimulatory signals that are not provided by peptide alone. Therapeutic peptide vaccines often attempt to provide these signals by incorporating adjuvants, molecules that enhance an immune response, which activate the APC and lead to enhanced cross-presentation of the exogenously-applied peptide to CD8+ T cells. However, two large phase III cancer vaccine trials failed to demonstrate clinical efficacy, raising the possibility that current adjuvant strategies are insufficient to drive potent anti-tumour CD8+ T cell responses.

This thesis examines whether molecules that activate innate-like T cells can serve as adjuvants to stimulate CD8+ T cell responses with anti-tumour activity, when chemically conjugated to antigenic peptides. In particular, the anti-tumour activity of a new aminobisphosphonate (ABP)-peptide conjugate vaccine was assessed. ABPs inhibit the mevalonate pathway, leading to the accumulation of a compound that activates a subset of human innate-like T cells called, Vγ9Vδ2 T cells. These vaccines enhanced both Vγ9Vδ2 T cell, and peptide-specific CD8+ T cell responses in human cell cultures in vitro. However, the CD8+ T cell response was independent of Vγ9Vδ2 T cell activity, suggesting an alternative mechanism was involved in the adjuvant activity. In support of this, these vaccines induced T cell responses in mice, which completely lack Vγ9Vδ2 T cells, and these responses were sufficient to delay tumour progression in murine models of melanoma and lung carcinoma.

Investigation into the cell-types responsible for the anti-tumour activity revealed responses were independent of the classical cross-presenting APC subset, conventional type 1 dendritic cells (cDC1s), and required a phagocytic CD11c+ APC subset outside of the spleen. Unexpectedly, vaccine activity was also dependent on CD1d and activated NKT cells in vivo. This was not a direct effect of ABP activity, as free ABPs failed to activate NKT cells in vivo. Further assessment raised the possibility that a contaminating particle or macromolecule was responsible for the NKT cell stimulatory activity. Interestingly, the required APC subsets differed from vaccines designed to rely on NKT cell activity, and NKT cells were neither activated nor required for the activity of the ABP-peptide vaccines in human peripheral blood cell assays. Further dissection of the mechanism of these vaccines was confounded by batch-to-batch variability of ABP-peptide conjugate vaccines, requiring modification to the manufacture and testing strategy in the partner chemistry laboratory.

Another class of conjugate vaccines that activate NKT cells had already been assessed in models of melanoma and lung cancer, using model or viral peptides. To extend the applicability of these findings, I sought to assess this concept in a model of HER2+ breast cancer (4T1.2-HER2). This model extends the NKT cell activating conjugate work to a different mouse strain (with lower NKT cell frequency and different NKT cell phenotype), a relevant human tumour-associated antigen (TAA), and to a different tumour model. I demonstrated that NKT cell activation can enhance CD8+ T cell responses against HER2 in vivo, and showed that this glycolipid-HER2 vaccine prevented lung colonisation by 4T1.2-HER2 tumour cells in a metastatic model of breast cancer, providing evidence that NKT cells can adjuvant immune responses against an antigen of translational relevance in breast cancer.

In summary, this thesis first demonstrates that a new class of ABP-peptide conjugate vaccines can drive peptide-specific CD8+ T cell responses independently of the innate-like T-cells it was designed to stimulate, and likely via mevalonate pathway blockade. I show that ABP-peptide vaccines can elicit anti-tumour efficacy in vivo independently of classical cross-presenting APCs, although additional mechanistic work was hindered by manufacturing issues with the ABP-peptide vaccines. Alternative vaccine design strategies are proposed. Second, I show that an existing class of innate-like T-cell-stimulating conjugate vaccines can be successfully employed in a mouse strain that had not previously been used, against a clinically-relevant human TAA, and in a breast cancer model, widening the scope of disease models for conjugate vaccine investigation. I conclude that the concept of conjugate vaccines that target the mevalonate pathway is a promising means of stimulating immune responses against cancer.