Monoclonal antibodies against pathogenic S100A8/9 alarmins can restore hematopoiesis and result in immunomodulation in myelodysplasia and leukemia models

Abstract Introduction: Myelodysplastic syndromes (MDS); chronic myelomonocytic leukemias (CMML) and acute myeloid leukemia (AML) are characterized by altered hematopoiesis that leads to cytopenias, the major source of morbidity in these diseases. An immunosuppressive marrow microenvironment is also seen in these myeloid neoplasms and is characterized by expansion of myeloid-derived suppressor cells (MDSCs) and exhausted T cells. S100A8/9 heterodimers (also known as DAMPs, alarmins, calprotectin) are known drivers of disease pathogenesis, though no specific targeting therapeutic approaches have been tested in MDS/CMML. The S100A8/9 heterodimers bind to multiple receptors, including TLR4/MD2, CD33, and RAGE, and lead to leukocyte recruitment, inflammatory cytokine secretion, and formation of MDSCs, thus leading to immune suppression and chronic inflammation seen in MDS/AML patients Results: Analysis of a large dataset of MDS patient-derived bone marrow CD34+ hematopoietic stem and progenitor cells (N=183) revealed increased expression of S100A8/9 receptors, TLR4/MD2, CD33, and RAGE, when compared to healthy controls. Furthermore, within the MDS patient cohort, those with higher S100A8/9 expression were more dependent on RBC transfusions and had lower hemoglobin. Additionally, in a CMML dataset (N=39), scRNA-seq profiles showed significantly higher expression of both S100A8 and S100A9 in the disease-initiating, inflammatory (cluster 2) cells. Distinct neutralizing monoclonal antibodies (mAbs) targeting human S100A8/9 homo/heterodimers were generated by single B-cell cloning from antigen-immunized rabbits. Selected rabbit-derived, chimeric, and subsequently humanized antibodies demonstrated nano-molar affinity and neutralizing activity that effectively inhibits the interaction of S100A8/A9 with RAGE, Siglec-3, Siglec-9, CD147, CD146, CD300ld, and CD69. In functional assays, these neutralizing mAbs also suppressed the formation of MDSCs and T-regulatory cells from human PBMC cultures. Numerous primary MDS, CMML, and AML mononuclear cells (including those with TP53, TET2, DNMT3A, Splicing, and other mutations) were grown ex vivo in the presence of monoclonal S100A8/9 neutralizing antibodies. Neutralization of S100A8/9 led to increased erythroid colony formation and increased erythroid differentiation as evident from Glycophorin A and CD71 co-expression by flow cytometry. Conclusion: Overexpression of S100A8/9 and its receptors is seen in MDS and leukemia samples. We have developed potent neutralizing monoclonal antibodies against the S100A8/9 heterodimer that can deplete MDSCs and reduce T-reg cell formation, thus potentially reversing the immunosuppressive microenvironment. Importantly, neutralization of S100A8/9 with our monoclonal antibodies stimulates erythropoiesis in a broad range of primary MDS and myeloid leukemia patient specimens. These results provide compelling pre-clinical rationale for the clinical development of S100A8/9 antibodies for the treatment of myeloid malignancies and reversal of cytopenias.