Objective Dysregulated APRIL/BAFF signaling is implicated in the pathogenesis of multiple autoimmune diseases, including systemic lupus erythematosus and lupus nephritis. We undertook this study to develop and evaluate a high‐affinity APRIL/BAFF antagonist to overcome the clinical limitations of existing B cell inhibitors. Methods A variant of TACI‐Fc generated by directed evolution showed enhanced binding for both APRIL and BAFF and was designated povetacicept (ALPN‐303). Povetacicept was compared to wild‐type (WT) TACI‐Fc and related molecules in vitro and in vivo. Results Povetacicept inhibited APRIL and BAFF more effectively than all evaluated forms of WT TACI‐Fc and selective APRIL and BAFF inhibitors in cell‐based reporter assays and primary human B cell assays, mediating potent suppression of B cell proliferation, differentiation, and immunoglobulin (Ig) secretion. In mouse immunization models, povetacicept significantly reduced serum immunoglobulin titers and antibody‐secreting cells more effectively than anti‐CD20 monoclonal antibodies, WT TACI‐Fc, or APRIL and BAFF inhibitors. In the NZB × NZW mouse lupus nephritis model, povetacicept significantly enhanced survival and suppressed proteinuria, anti–double‐stranded DNA antibody titers, blood urea nitrogen, glomerulonephritis, and renal immunoglobulin deposition. In the bm12 mouse lupus model, povetacicept significantly reduced splenic plasmablasts, follicular helper T cells, and germinal center B cells. In non‐human primates, povetacicept was well tolerated, exhibited high serum exposure, and significantly decreased serum IgM, IgA, and IgG levels after a single dose. Conclusion Enhanced APRIL and BAFF inhibition by povetacicept led to greater inhibition of B cell populations critical for autoantibody production compared to WT TACI‐Fc and CD20‐, APRIL‐, or BAFF‐selective inhibitors. Potent, dual inhibition by povetacicept has the potential to significantly improve clinical outcomes in autoantibody‐related autoimmune diseases. image

ObjectiveCD28 and inducible T cell costimulator (ICOS) appear to have nonredundant roles in T cell activation and adaptive immunity. We undertook this study to characterize in vitro and in vivo the therapeutic potential of acazicolcept (ALPN‐101), an Fc fusion protein of a human variant ICOS ligand (ICOSL) domain designed to inhibit both CD28 and ICOS costimulation, in inflammatory arthritis.MethodsAcazicolcept was compared in vitro with inhibitors of either the CD28 or ICOS pathways (abatacept and belatacept [CTLA‐4Ig], prezalumab [anti‐ICOSL monoclonal antibody]) in receptor binding and signaling assays, and in a collagen‐induced arthritis (CIA) model. Acazicolcept was also compared in cytokine and gene expression assays of peripheral blood mononuclear cells (PBMCs) from healthy donors or rheumatoid arthritis (RA) or psoriatic arthritis (PsA) patients stimulated with artificial antigen‐presenting cells (APCs) expressing CD28 and ICOS ligands*.ResultsAcazicolcept bound CD28 and ICOS, prevented ligand binding, and inhibited human T cell functional interactions, matching or exceeding the activity of CD28 or ICOS costimulatory single‐pathway inhibitors tested individually or in combination. Acazicolcept administration significantly reduced disease in the CIA model and more potently than abatacept. Acazicolcept also inhibited proinflammatory cytokine production from stimulated PBMCs in cocultures with artificial APCs and demonstrated unique effects on gene expression distinct from those induced by abatacept, prezalumab, or a combination of both.ConclusionBoth CD28 and ICOS signaling play critical roles in inflammatory arthritis. Therapeutic agents such as acazicolcept that coinhibit both ICOS and CD28 signaling may mitigate inflammation and/or disease progression in RA and PsA more effectively than inhibitors of either pathway alone.image

Abstract Introduction: Checkpoint inhibition (CPI) has been shown to be an effective anti-tumor therapy, but CPI alone is frequently insufficient to control tumor growth, and costimulatory signals may also be required to produce clinically significant anti-tumor responses. PD1-PDL1 are established CPI targets and TMIGD2 is an inhibitory receptor expressed on T cells that is engaged by its cognate ligand HHLA2 on tumor cells, leading to inhibition of T cell responses. Novel biologics combining CD28 costimulation and CPI may yield promising tumor antigen-specific therapeutic candidates. Methods: Variants of CD86 with increased CD28 affinity were engineered using our directed evolution platform. PD1 and TMIGD2 variants were also engineered for increased affinity to PDL1 and HHLA2, respectively. Fusion proteins were generated including either PD1 or TMIGD2 domains, an effectorless Fc domain, and an engineered CD86 domain to generate proteins to provide target-dependent costimulation (TDC) and evaluated in multiple in vitro T cell stimulation assays. Costimulatory effects were confirmed by evaluating T cell proliferation, cytokine release, and tumor cell killing. Fusion proteins were also tested in vivo in a mouse tumor model using MC38 cells expressing human PDL1 or HHLA2, and in a humanized tumor model where an HPV+ squamous cell carcinoma cell line expressing either PDL1 or HHLA2 was implanted into NSG mice. Tumor-bearing NSG mice were given primary human T cells expressing an HPV peptide-specific TCR, with or without the corresponding fusion protein, and tumor growth was measured over time. Results: PD1-CD86 and TMIGD2-CD86 TDC proteins enhanced T cell costimulation in multiple in vitro T cell response assays, and costimulation was dependent on target cell lines expressing PDL1 or HHLA2, respectively. There was no effect on T cell responses if the target cells did not express the relevant target protein. PD1 and TMIGD2 domains alone had little effect on T cell responses, indicating these TDC proteins were driving enhanced responses. PD1-CD86 and TMIGD2-CD86 TDC proteins enhanced anti-tumor responses in vivo in a syngeneic MC38 implantation model only when MC38 cells expressed human PDL1 or HHLA2, respectively. Both proteins also enhanced antitumor responses in a humanized tumor implantation system using an HPV+ tumor cell line expressing the corresponding target antigen in the presence of human T cells expressing an HPV-specific TCR, but failed to control tumor growth when tumor cells lacked expression of the target protein. Conclusions: Tumor antigen-specific antitumor therapy can be achieved with fusion proteins that combine engineered CPI and CD28 costimulatory domains. Such novel biologics may provide promising approaches to enhancing the efficacy of CPI monotherapies and to address checkpoint inhibitor-resistant tumors. Citation Format: Steven D. Levin, Mark F. Maurer, Chelsea Gudgeon, Siddarth Chandrasekaran, Daniel Ardourel, Daniel Demonte, Joseph Kuijper, Martin Wolfson, Logan Garrett, Kayla N. Kleist, Sherri Mudri, Hieu Nguyen, Michelle Seaberg, Rachel Wang, Jing Yang, Katherine E. Lewis, Stacey R. Dillon, Mark Rixon, Stanford L. Peng. Engineered variant domain fusion proteins provide checkpoint inhibition and tumor antigen dependent CD28 costimulation resulting in potent anti-tumor immunity [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1740.