AUTHORS
Jisu Chae, PhD, Baskar Sivasubramanian, PhD2, Christian Combs, PhD3, Pradeep Dagur, PhD4, Ronald Taylor, PhD5, Christoph Rader, PhD6, Adrian Wiestner, MD, PhD7
AFFILIATIONS
1 Hematology Branch, National Heart, Lung, and Blood Institute, NIH, Bethesda, Maryland, United States
2 Hematology Branch, National Heart, Lung, and Blood Institute, NIH, Bethesda, Maryland, United States
3 Light Microscopy Core, National Heart, Lung, and Blood Institute, NIH, Bethesda, Maryland, United States
4 Flow Cytometry Core, National Heart, Lung, and Blood Institute, NIH, Bethesda, Maryland, United States
5 Department of Biochemistry and Molecular Genetics, University of Virginia School of Medicine, Charlottesville, Virginia, United States
6 Department of Immunology and Microbiology, The Herbert Wertheim UF Scripps Institute for Biomedical innovation & Technology, Jupiter, Florida, United States
7 Hematology Branch, National Heart, Lung, and Blood Institute, NIH, Bethesda, Maryland, United States
ABSTRACT
Monoclonal antibodies (mAbs) targeting CD20 or CD38 can induce deep remissions in patients with B-cell malignancies. Internalization or trogocytosis can lead to loss of target antigens and limit the efficacy of mAb therapy. Through complement activation these mAbs also deposit complement fragments, including C3b which decays to the terminal product C3d, on the cell surface. C3d remains covalently bound and can serve as a neoantigen on the tumor cell surface. We previously generated mAbs targeting C3d and demonstrated that a combination of C3d mAbs together with CD20 or CD38 mAbs significantly improved anti-tumor activity in lymphoma and myeloma xenograft models. Thus, C3d is an actionable target to enhance mAb activity against B-cell malignancies.
C3d has properties distinct from classic cell surface mAb targets. C3d is bound covalently to cell surface molecules via a thio-ester bond. Even weeks after administration of a complement fixing mAb, residual tumor cells remained C3d opsonized. We first sought to identify C3d bound cell surface proteins. Using anti-C3d mAbs, we co-immunoprecipitated proteins from C3d-opsonized human lymphoma and myeloma cell lines. Mass spectrometry (LC-MS/MS) identified several C3d-bound proteins including CD19, CD44, CD47, in addition to the original target antigens CD20 or CD38, respectively.
Next, we studied the physical properties of C3d deposits on tumor cells and their interactions with macrophages. STED microscopy demonstrated that mAb-bound C3d formed a broad, homogeneous membrane distribution. In contrast, mAb-bound CD20 formed discrete membrane clusters. We observed markedly different responses to macrophage-mediated trogocytosis likely due to how C3d distributes across the cell membrane. For example, after exposure to activated THP-1 monocyte/macrophage cells, at 1:1 ratio, target lymphoma cell lines lost a median of 84% - 92% of their surface CD20. In contrast, C3d opsonized target cells retained a median of 42% - 80% of their surface C3d, indicating substantially less trogocytosis and thereby less loss of C3d than of CD20.
Repeated macrophage encounters progressively depleted C3d (~90% loss after 3 rounds). Nevertheless, C3d levels remained higher for longer than CD20 levels. Despite the higher antigen retention of C3d during trogocytosis, imaging flow cytometry demonstrated approximately 6 times more transfer of tumor membrane fragments during C3d-directed (19.1%) than CD20-directed trogocytosis (2.99%). Intracellular uptake of membrane fragments and bound antibodies within macrophages was confirmed with confocal microscopy.
To study antibody-dependent cellular phagocytosis (ADCP) further, we obtained chronic lymphocytic leukemia (CLL) cells from patients enrolled on our natural history study (NCT00923507), who were treated with the anti-CD20 mAb ofatumumab (OFA) (NCT01145209). In-vitro opsonization of treatment-naïve CLL cells with C3d increased ADCP from 35% with anti-CD20 mAb alone to 48% with anti-C3d. We previously reported that CLL cells obtained the day after ofatumumab administration are devoid of all CD20 but remain C3d opsonized. Consistently, anti-C3d mAb but not anti-CD20 mAb induced effective phagocytosis of these cells. ADCP was further enhanced by combining two anti-C3d mAbs that target non-overlapping epitopes, suggesting additional benefits of combining non-competing antibodies against a shared target to maximize ADCP.
Trogocytosis removes entire sections of the cell membrane, which can lead to cell death, referred to as trogoptosis. Compared to anti-CD20, the fraction of in vitro C3d opsonized CLL cells undergoing trogoptosis in response to anti-C3d increased 500%. CLL cells obtained from patients 24 hours after OFA infusion showed a similar increase in trogoptosis in response to anti-C3d. Loss of larger membrane areas likely increased mechanical disruption, resulted in net energy loss, and might also remove pro-survival cell surface proteins. All these factors combined may lead to increased target cell death. Collectively, these findings establish complement-deposited C3d as a treatment-induced therapeutic target whose unique membrane organization and persistence enhance macrophage effector function. These results support anti-C3d antibodies as a rational strategy to augment complement-fixing antibody therapies.
DISCLOSURES
Jisu Chae, PhD: No relevant conflicts of interest to declare.