The role of Eph proteins in haematopoiesis and leukaemia

Abstract

Eph and their membrane bound ephrin ligands represent the largest family of receptor tyrosine kinases (RTKs). Ephs interact with cell-surface ephrin ligands to direct cell migration and orchestrate developmental patterning during embryogenesis by modulating cell shape and adhesion. Members of the Eph/ephrin family of RTKs are expressed in adult life and have ongoing roles both in normal tissue homeostasis and in responses to pathological situations. Eph/ephrin proteins are also expressed on many tumours including leukaemia, prostate cancer, colorectal cancer and brain cancer. High expression in a number of cancers has been linked to progression through facilitation of invasiveness and metastatic spread.Some members of the Eph/ephrin family of proteins are expressed on normal and cancerous haematopoietic cells. Real time PCR analysis showed expression of all EphA proteins on the haematopoietic stem cell (HSC) population, except for EphA6 and EphA8. I have used EphA1, EphA2 and EphA7 knockout mice to investigate the effect of these individual Eph proteins in normal haematopoiesis, and in particular haematopoietic stem cells in vivo. These studies showed that mice lacking EphA1, EphA2 or EphA7 have normal haematopoiesis and no significant defects were observed in HSCs. This suggests that if Eph receptors have a role in haematopoietic differentiation the presence of other Eph receptors may compensate for the absence of individual Eph proteins.EphA2 expression has been investigated in many different solid tumours and expression of this member of the Eph/ephrin family has been detected in mixed lineage leukaemia (MLL). The role of EphA2 in leukaemogenesis was investigated in a syngeneic, retroviral model of MLL-AF9 induced acute myeloid leukaemia (AML). When HSCs from wild type mice were transduced with the MLL-AF9 retrovirus and engrafted into recipient mice, the subsequent leukaemias showed significant expression of EphA2 on leukemic blast cells. However when EphA2 null HSCs were engrafted, the absence of EphA2 appeared to have no effect on the leukemic process, demonstrating that EphA2 is not essential for leukaemia progression. Analysis showing increased expression of EphA7 and other Eph proteins demonstrate a possible compensatory role for the lack EphA2 expression, if Eph signalling was required for leukemic evolution in this model. Following these observations, the role of EphA2 as a therapeutic target for MLL-AF9 leukaemia was evaluated. Results showed that radiolabelled (beta particle-emitting lutetium) EphA2 was an effective therapeutic approach by significantly delaying the progression of MLL-AF9 leukaemia. Given the poor prognosis of MLL-driven paediatric acute lymphoblastic leukaemia (ALL) this data raises the prospect of effective EphA2-targeted therapy for this disease.EphA3 is another member of the Eph/ephrin family of RTKs, which was originally described in leukaemia and has subsequently been described in many other different cancers including sarcomas, lung cancer, melanoma and glioblastoma. The EphA3- specific monoclonal antibody, IIIA4, binds and activates both human and mouse EphA3 with similar affinities. High expression of EphA3 has been reported previously on LK63 pre-B ALL cell line which contrasts with lack of expression of EphA3 in the Reh, a similar pre-B ALL cell line. The effect of IIIA4 monoclonal antibody treatment on LK63 and Reh xenografts in SCID-NOD mice was examined, to explore both direct anti-tumour and potential effects of this antibody on the host cells. In this xenograft model of leukaemia, using the LK63 cell line, administration of the IIIA4 antibody led to inhibition of both tumour growth and spread from bone marrow to the spleen and other organs, hence increasing the latency of the disease. In contrast, there was no significant reduction in engraftment in the Reh xenograft model, suggesting that the anti-tumour effect was directed against the leukemic cells rather than the stromal and vascular elements.To further analyse the therapeutic targeting of EphA3, LK63 EphA3-knockdown and Reh EphA3-expressing cells were tested in the xenograft model. Similarly to the LK63 model of leukaemia, the EphA3-expressing Reh xenograft model showed a reduction in bone marrow engraftment and slowing of disease progression upon treatment with EphA3 monoclonal antibody. Consistent with the effect of EphA3 monoclonal antibody treatment on Reh xenograft model, the LK63 EphA3 knockdown xenograft, showed minimal differences between treated and control group but had a notably significant reduction in the splenic engraftment compared to the normal LK63 model. A more impressive anti-tumour effect was delivered by antibody-mediated targeting of a radio-active isotope to EphA3-positive leukemic cells, in this case radioactive bismuth (alpha particle emitting) isotope-linked to IIIA4. More importantly, no toxicity to normal tissues was observed, consistent with EphA3 being expressed at very low or being entirely absent on normal tissues.In summary, in this thesis I have shown that whilst individual Eph receptors do not have critical roles in normal and malignant haematopoietic cells, individual Eph proteins may be candidates for targeted, antibody-based therapies in blood cancers.