Improving NK cell-based immunotherapy of cancer by exploration of migration and inhibitory receptor-ligand interactions

Sammanfattning: The possibility to utilize the immune system as a mean to target malignantly transformed cells has emerged as a novel treatment modality during the last decades and is a field that is under rapid development. Natural Killer (NK) cells are immune cells with an intrinsic ability to recognize and lyse tumor-transformed cells. Their role in immune surveillance of cancer is well established and clinical studies have also highlighted their treatment potential. However, the efficacy has so far been suboptimal and means to further improve NK cell-based immunotherapy against cancer is under intensive investigation. This thesis mainly focuses on how to enhance NK cell treatment efficacy against acute myeloid leukemia (AML), a disease with poor prognosis that is in urgent need for novel and more efficacious treatment options. The potential to better home NK cells to the tumor environment is a highly appealing, but so far relatively unexplored approach to improve treatment efficacy for adoptive NK cell transfer. In Papers I and II, we have utilized a clinically approved mRNA transfection technique to explore how ex vivo expanded NK cells can be engineered to express the gain-of-function bone marrow (BM) homing receptor variant CXCR4R334X. We have shown that CXCR4R334X-expressing NK cells display an increased in vitro migration potential towards the ligand SDF-1? without compromising functionality in any other aspect. We have further highlighted the importance of the CXCR4-SDF-1? axis for in vivo BM homing using both CXCR4R334X mRNA transfected cells and CXCR4 knocked-out cells generated using the CRISPR/Cas9 technology. Finally, we showed that CXCR4R334X-expressing NK cells can be utilized to improve leukemia clearance in vivo, and furthermore have the potential to mediate a survival benefit in this context. Education is a complicated process in which NK cells are dynamically tuned by surrounding HLA class I molecules facilitating a strong effector response against target cells lacking such molecules. Members of the killer cell immunoglobulin-like receptor (KIR) family and the natural killer group 2 member A (NKG2A) receptor have previously been shown to mediate education. In paper III, we have shown for the first time that the LIR-1 receptor can mediate education in ex vivo expanded NK cells. We have identified an educated LIR-1-expressing NK cell subset that display an elevated degranulation capacity towards HLA class I deficient or low tumor targets and further showed that this subset also has a phenotype which resembles that of educated cells. In addition, we have displayed that the subset can be detuned after blockade with specific LIR-1 monoclonal antibodies in line with the rheostat model. Finally, we revealed that this subset has a potent ADCC capacity that even can overcome inhibitory signals from cognate HLA class I ligands. We have postulated that these traits make this subset highly attractive for NK cell-based immunotherapy and could be utilized in various cancer settings, including AML. Isocitrate dehydrogenase (IDH) 1 and 2 are frequently mutated in AML. IDH inhibitors (IDHi) have emerged as novel therapeutic agents. However, only a portion of the patients respond to treatment and the relapse rate is high. In paper IV, we have employed the TF- 1 AML cell line overexpressing the hotspot mutation IDH2R140Q as a model system to study the epigenetical and transcriptional landscape before and after treatment with an IDHi. We have shown that TF-1 IDH2R140Q cells display a hypermethylated DNA profile leading to differential gene expression, which can be further linked to a dysregulated transcriptional network involving myeloid-related transcription factors. We further displayed that this, at least in part, can be reversed upon treatment with the IDHi. Moreover, we showed that both TF-1 IDH2R140Q cells and primary AML cells carrying IDH mutations have hypermethylated HLA class I genes leading to HLA class I downregulation. This was associated with an increased sensitivity to NK cell-mediated responses against the IDH2R140Q mutated TF-1 cells compared to wild-type (WT) TF-1 cells. Lastly, we showed that HLA class I genes in IDH mutated primary AML cells of patients that did not respond to treatment with IDHi remained hypermethylated, indicating that this patient group may benefit from treatment based on adoptive NK cell infusions. Overall, this thesis provides evidence that NK cell migration can be modulated to redirect NK cells to bone marrow compartments for improved leukemia clearance. Furthermore, it includes a paper where the inhibitory receptor LIR-1 for the first time is shown to mediate education of ex vivo expanded NK cells, and identifies a unique subset of highly educated NK cells with a robust ADCC capacity that have the potential to be utilized against various cancer types including AML. Finally, the thesis includes a manuscript which shows that mutations in IDH can confer an increased NK cell sensitivity, and that IDH mutated AML patients resistant to treatment with IDHi constitute a subgroup of AML that may specifically benefit from NK cell-based immunotherapy. In conclusion, the thesis provides novel insights on how NK cells can be utilized to improve treatment for AML and potentially also other types of BM residing malignancies.