Investigating the role of MHC class I molecules in immune evasion by transmissible tumour cells

Abstract

Downregulation of Major Histocompatibility Complex (MHC) molecules is often key to evasion of the immune system by tumours and viruses. Transmissible cancers transmit between individuals, in a manner akin to a metastatic event, providing a unique opportunity to study the evolution of MHC loss in the face of selective pressure from the immune system. Tasmanian devils are infected with two genetically distinct transmissible cancers which transmit via biting behaviours. Devil Facial Tumour Disease (DFT1) emerged over 25 years ago and has spread across most of Tasmania, while DFT2 was identified in 2014 and is still limited in geographical range. In contrast to DFT1, which has epigenetically downregulated MHC class I expression, DFT2 tumours express MHC class I molecules, but recent evidence shows some DFT2 tumours have lower MHC class I expression, suggestive of evolving immune escape by this cancer. In this thesis, I investigate how MHC expression is changing in DFT2 as it spreads through the population and encounter hosts with disparate MHC class I genotypes, and how non-classical MHC class I molecules may be modulating immune escape by DFT1. DFT1 (n=76) and DFT2 (n=34) tumour biopsies were stained by immunohistochemistry (IHC) for classical MHC class I molecules, and non-classical, Saha-UD and Saha-UK. An expression score was generated using semi-automated image analysis to quantify IHC staining. To investigate host immune responses to MHC class I expression, tumours were also stained for IFNγ (DFT1, n=51; DFT2, n=25) and CD3 (DFT1, n=42; DFT2, n= 20). Two antibodies previously generated in our lab were validated for use in this thesis, Saha-UD (α-UD_14-37-3) and IFNγ (α-IFNγ_13-44-6). Hosts (n=10) were genotyped using deep sequencing at three classical MHC class I loci to study whether mismatch between host and tumour drives MHC class I loss in DFT2. In this thesis I have confirmed classical MHC class I expression is highly variable among DFT1 and DFT2 tumours. In DFT2, classical MHC class I expression correlated with IFNγ expression, therefore it’s expression in DFT2 may be immunogenic. However, there was no correlation between classical MHC class I expression and host-tumour mismatch. A ubiquitous and likely fixed allele, SahaI*27, is expressed within the population and DFT2 cells, which may assist DFT2 immune evasion. In DFT1, the finding of classical MHC class I expression in primary tumours is surprising and challenges our understanding of immune escape by DFT1, as it was presumed all DFT1 tumours had downregulated MHC class I from the cell surface. For the first time, non-classical, Saha-UD expression has been confirmed in DFT1 tumours. Interestingly, DFT1 tumours are heterogeneous for expression of non-classical, Saha-UD and Saha-UK, and their expression is positively correlated with classical MHC class I expression. Therefore, their expression may be immunosuppressive, to prevent host immune activation. These results demonstrate that DFT2 is evolving immune evasion mechanisms as it transmits between individuals in the population, with the potential for more rapid dispersion if MHC-negative subclones gain dominance. Expression of a dominant MHC class I in the population likely facilitates DFT2 spread in this area. While non-classical Saha-UD expression by DFT1 tumours in vivo may be a mechanism for immunosuppression, further work is needed to characterise its ligand on immune cells. This data can be used to inform more effective management of the population and vaccine design. Further, this study provides a platform to investigate the mechanisms behind MHC loss in a cancer under sustained pressure from the immune system.

More information

Identifiers

ORCID for Kathryn Anne Hussey: orcid.org/0000-0002-8882-1502

ORCID for Hannah Siddle: orcid.org/0000-0003-2906-4385

ORCID for Edd James: orcid.org/0000-0001-8638-7928

Catalogue record

Export record