509 DSP216 — a novel anti HLA-GxCD47 bi-specific fusion protein for cancer immunotherapy

<h3>Background</h3> Therapeutic inhibition of CD47/SIRPα ‘don’t eat me’ signaling can reactivate innate immunity and CD47-based combination therapies demonstrated encouraging clinical efficacy.¹ An interesting target for combination is HLA-G, a cancer-expressed immune checkpoint that inhibits multiple immune cell subsets and recruits suppressive immune cells to the tumor microenvironment.² Here, we report on the preclinical characterization of Dual Signaling Protein 216 (DSP216), a heterodimeric Fc-fusion protein designed to simultaneously block both HLA-G/LILRB and CD47/SIRPα signaling (figure 1). DSP216 comprises sequence-optimized extracellular domains of SIRPα and LILRB2 fused to human IgG, designed to simultaneously bind and block its ligands, CD47 and HLA-G, both frequently overexpressed on cancer cells. Thus, DSP216 is designed to reactivate both innate and adaptive anticancer immunity. <h3>Methods</h3> Computational modeling was used to identify amino acid substitutions to increase DSP216 binding affinity to HLA-G. Target specificity of DSP216 was determined by incubating cancer cells with DSP216 preincubated with HLA-G and/or CD47 blocking antibodies. Further, binding was assessed in a mix of cancer cells, PBMCs and red blood cells (RBCs). Impact of DSP216 on macrophage polarization was assessed by flow cytometric staining for M1/M2 markers and cytokine secretion after co-cultures of cancer cells and M0 macrophages. Macrophage-mediated phagocytosis and granulocyte-mediated trogocytosis were evaluated using flow cytometry by gating on the CD11b-stained macrophages or on granulocytes that were positive for fluorescently-labelled cancer cells. Impact on NK cytotoxicity was assessed by staining of cancer cells with Annexin V. <h3>Results</h3> A modified sequence in the LILRB2 domain of DSP216 yielded the desired binding profile, with strong binding to cancer cells expressing both ligands and minimal binding to cells expressing one ligand, like PBMCs and RBCs. Binding was abrogated by pre-incubation with HLA-G or CD47 blocking antibodies. DSP216 treatment of co-cultures of M0-macrophages with cancer cells prevented HLA-G induced upregulation of M2-specific markers and downregulation of the M1-markers. Furthermore, DSP216 increased secretion of M1-cytokines TNFα and IL6. DSP216 treatment potentiated NK cytotoxicity towards cancer cells and significantly increased macrophage-mediated phagocytosis as well as granulocyte-mediated trogocytosis of cancer cells. <h3>Conclusions</h3> DSP216 selectively binds cancer cells expressing CD47 and HLA-G and not or minimally binds RBCs and PBMCs. DSP216 prevented HLA-G mediated M1 to M2 polarization of macrophages and promoted macrophage phagocytosis as well as NK cytotoxicity. DSP216 is a novel bifunctional therapeutic that has the potential to reverse immunosuppressive signaling and unleash anti-cancer immune responses. <h3>References</h3> Ye ZH, Yu WB, Huang MY, Chen J, Lu JJ. Building on the backbone of CD47-based therapy in cancer: combination strategies, mechanisms, and future perspectives.<i> Acta Pharm Sin B</i> 2023 Apr;<b>13</b>(4):1467–1487. Lin A, Yan WH. HLA-G/ILTs targeted solid cancer immunotherapy: opportunities and challenges. <i>Front Immunol</i> 2021 Jun 30;<b>12</b>:698677. <h3>Ethics Approval</h3> All procedures were done in accordance with International Ethical and Professional Guidelines (the Declaration of Helsinki and the International Conference on Harmonization Guidelines for Good Clinical Practice). Leftover blood samples were received from Hadassah blood bank, Hadassah Medical Center, Jerusalem, IL (Helsinki No. 0155-17-HMO). Samples were received anonymously, and consent was not required.