Novel Targets in Acute Myeloid Leukaemia

Abstract

Background: Antibody based immunotherapies have revolutionised the treatment of haematological malignancies. Despite recent advances in Acute Myeloid Leukaemia (AML) most patients still have poor outcomes. Current surface targets in AML are not ideal and ongoing work is required to examine new antigens for meaningful clinical outcomes. Hypothesis: CD302 and CD300f have inherent properties that make them promising potential targets in AML. Preclinical work will establish these antigens as suitable targets in AML for further study. Methods: We looked at the distribution of CD302 on AML and Haematopoietic Stem and Progenitor Cells (HSPC) using flow cytometry from local patient cohorts and compared this data with AML gene expression profiling from public databases. The expression of CD302 from healthy organs was examined using PCR. The rate of internalisation of anti-CD302 antibodies was assessed using flow cytometry and fluorescent microscopy. The ability of unmodified antibodies to perform Antibody Dependant Cellular Cytotoxicity (ADCC) was examined. The impact of CD302 on AML cell migration was assessed using a transwell system. The influence of an anti-CD302 antibody upon AML cell line engraftment was tested in vivo. CD300f on AML and HSPC was analysed using flow cytometry from local patient cohorts and we compared the data to gene expression profiling from public databases. The distribution of CD300f isoforms across AML and HSPC was assessed by PCR and confirmed with TCGA RNA sequencing data. Alterations in antibody binding epitopes using multiple CD300f antibodies was assessed in AML cell lines as well as primary AML and HSPC. The cytotoxicity of an anti-CD300f based ADC was examined in vitro using cell lines. An anti-CD300f ADC was examined in vitro and in vivo against cell lines as well as primary AML and healthy HSPC. The rate of cytotoxicity and synergy of the ADC with Fludarabine was assessed in vitro with cell lines. Results: In a cohort of 33 AML patients, 88% were found to express CD302 on the surface of blasts and 80% on the surface of CD34+ CD38- population. Expression of CD302 was found on the surface of HSPC. A monoclonal antibody (mAb) targeting human CD302 was effective in mediating ADCC and was internalised, making it amenable to toxin conjugation. Targeting CD302 with this antibody limited in vivo engraftment of the leukaemic cell line HL-60 in NOD/SCID mice. While CD302 was expressed in a hepatic cell line, HepG2, this molecule was not detected on the surface of HepG2, nor could HepG2 be killed using a CD302 antibody-drug conjugate. CD300f antibodies bind to AML from 85% of patient samples. Transcriptomic analysis found that CD300f transcripts are expressed by healthy HSPC. Several CD300f protein isoforms exist as a result of alternative splicing. The extracellular region of CD300f can be present with or without the exon 4‐encoded sequence. This results in CD300f isoforms that are differentially bound by CD300f‐specific antibodies. Analysis of publicly available transcriptomic data indicated that CD34+ HSPC expressed fewer CD300f transcripts that expressed exon four compared to AML with monocytic differentiation. An anti-CD300f antibody, DCR‐2, to CD300f exposes a structural epitope recognized by a second CD300f mAb, UP‐D2. Analysis of a small cohort of AML cells revealed that this UP‐D2 conformational binding site could be induced in cells from AML patients with monocytic differentiation but not those from other AML or HSPC. CD300f is expressed evenly across HSPC subtypes. CD300f has equivalent transcription and protein expression as CD33 on AML. We have developed an anti-CD300f antibody which efficiently internalises into target cells and conjugated it with a PBD warhead that selectively depletes AML cell lines and AML Colony Forming Units (CFU) in vitro. CFU derived from healthy HSPC are depleted by the ADC. The ADC synergises with Fludarabine, which is often used in allogeneic Haematopoietic Stem Cell Transplant (allo-HSCT) conditioning. The ADC prolongs the survival of mice engrafted with human cell lines and depletes primary human AML engrafted with a single injection. In a humanised mouse model, a single injection of the ADC depletes CD34+ HSPC and CD34+ CD38- CD90+ HSC. Conclusions: CD302 is a potential target in AML. The hepatic expression may limit potential therapeutics but this could be mitigated as hepatocytes appear to express CD302 predominantly intracellularly, further work is required in this field. CD302 is expressed on HSPC and any future therapeutics would likely need to be part of a conditioning strategy. CD300f is a more promising target given the lack of non-haematopoietic expression. Certain isoforms or epitopes may be targeted to generate selective binding to AML with monocytic differentiation and avoid HSPC. An anti-CD300f ADC has promise as a targeted conditioning agent that may deplete residual AML and facilitate allo-HSCT.