NK cell function and metabolism in human cancer

Natural Killer (NK) cells are important anti-cancer innate immune cells. They are essential for cancer immunosurveillance and their activity is associated with better outcome in cancer patients. Some of their key functions include expression of cytotoxic molecules such as granzymes, production of the anti-tumour cytokine IFN?, and antibody dependent cellular cytotoxicity (ADCC), which involves the killing of tumour cells coated with antibodies. Highlighting the importance of this particular function, NK cell ADCC has been harnessed through the development of monoclonal antibody therapies such as Trastuzumab (anti-HER2, breast cancer) and Dinutuximab (anti-GD2, neuroblastoma), which boost NK cell mediated ADCC against patient tumours and have become mainstream in clinical practice. Despite their fundamental role in protection against cancer, the anti-tumour functions of NK cells often become impaired in patients who have cancer, and indeed this has been documented across a wide range of haematological and solid tumour cancers. This dysfunction likely impinges upon the efficacy of NK cell based immunotherapies, such as monoclonal antibody therapy and NK cell adoptive transfer therapy. Therefore, there is an urgent need to better understand the mechanisms that underlie NK cell dysfunction during cancer, so that we can identify novel therapeutic targets that can be used to reverse these defects. Several important studies from the past decade have revealed that cellular metabolism is essential for human NK cell responses. For example, cytokine stimulated NK cells upregulate both oxidative phosphorylation (oxphos) and glycolysis to support their functions. Mammalian target of rapamycin complex 1 (mTORC1) drives this metabolic shift towards glycolysis and is also required for the functionsof NK cells. Interestingly, NK cells shuttle their mitochondria towards the site of interaction with tumour cells, highlighting how fundamental metabolism is for cytotoxicity against tumour cells. While, the role that metabolism plays in driving NK cell functional responses is now well documented, whether metabolism is involved in NK cell dysfunction in cancer patients remains less clear. Therefore, we hypothesised that altered cellular metabolism might underlie NK cell dysfunction during cancer. In this thesis, we show that peripheral blood NK cells from metastatic breast cancer and paediatric neuroblastoma patients have impaired metabolism, which, amongst others,is associated with defective IFN? production. While a shift towards one form of metabolism was expected, both oxphos and glycolysis were severely reduced in patient NK cells, indicating metabolic paralysis. At the level of the organelle, NK cells from breast cancer and neuroblastoma patients had increased mitochondrial mass and mitochondrial membrane polarisation (MMP), as well as high levels of mitochondrial reactive oxygen species (mROS). Confocal microscopy analysis revealed that patient NK cells have modified mitochondrial morphology characterised by increased mitochondrial fragmentation, as seen by their round, punctate mitochondria, versus the long, elongated mitochondria generally observed in NK cells from healthy donors. Altered mTORC1 activity was also observed in NK cells from both patient groups, and is likely contributing towards the impaired metabolism and function. TGF? is highly implicated in cancer development and progression, and work from our lab recently showed that TGF? inhibits healthy donor NK cell metabolism and function. Therefore, we hypothesised that TGF? might play a role in driving some of the functional and metabolic defects observed. TGF? neutralising antibodies were added to overnight cell cultures of NK cells from metastatic breast cancer patients. This treatment strikingly improved NK cell mitochondrial metabolism, IFN? production and mTORC1 activity, indicating that TGF? contributes towards NK cell metabolic dysfunction during cancer. Importantly, we reveal for the first time that the GARP/TGF? axis, which involves the intrinsic anchoring of TGF? to the NK cell membrane, is a key driver of NK cell dysfunction during cancer. By adding monoclonal antibodies that block GARP/TGF? complexes to pure NK cell cultures overnight, we were able to recapitulate many of the effects of anti-TGF? neutralising antibodies, and improve IFN? production as well as granzyme B expression in patient NK cells. Finally, we explored the metabolic requirements of NK cell anti-GD2 mediated ADCC against neuroblastoma tumour cells, using pharmacological inhibitors against various metabolic pathways and regulators. The data clearly indicate that glycolysis and not oxphos is required for NK cell anti-GD2 mediated ADCC. Indeed, inhibition of glycolysis reduces ADCC and this is associated with reduced degranulation against anti-GD2 coated neuroblastoma tumour cells. Furthermore, we show that while mTORC1 is not required for this process, active amino acid transport though the SLC7A5 transporter is essential, and we suggest this could be required to sustain cMyc activity, which in turn may promote NK cell ADCC. Overall, we have shown in two different types of solid tumour cancers that reduced cellular metabolism and mTORC1 activity underlies peripheral NK cell dysfunction. Perhaps most significant is the finding that this altered metabolism is targetable and reversible, and that this can be achieved using monoclonal antibodies against TGF? alone as well as GARP/TGF? complexes. Furthermore, our data support a therapeutic strategy whereby NK cell glycolysis and amino acid transport is supported when neuroblastoma patients are receiving anti-GD2 therapy. These findings have significant implications for the design of future NK cell based immunotherapies against solid tumour cancers