Induction of NK cell reactivity against acute myeloid leukemia by Fc-optimized CD276 (B7-H3) antibody

Abstract

Acute myeloid leukemia (AML) remains a therapeutic challenge despite recent therapeutic advances. Although monoclonal antibodies (mAbs) engaging natural killer (NK) cells via antibody-dependent cellular cytotoxicity (ADCC) hold promise in cancer therapy, almost none have received clinical approval for AML, so far. Recently, CD276 (B7-H3) has emerged as a promising target for AML immunotherapy, due to its high expression on leukemic blasts of AML patients. Here, we present the preclinical development of the Fc-optimized CD276 mAb 8H8_SDIE with enhanced CD16 affinity. We demonstrate that 8H8_SDIE specifically binds to CD276 on AML cell lines and primary AML cells and induces pronounced NK cell activation and degranulation as measured by CD69, CD25, and CD107a. Secretion of IFNγ, TNF, granzyme B, granulysin, and perforin, which mediate NK cell effector functions, was induced by 8H8_SDIE. A pronounced target cell-restricted lysis of AML cell lines and primary AML cells was observed in cytotoxicity assays using 8H8_SDIE. Finally, xenograft models with 8H8_SDIE did not cause off-target immune activation and effectively inhibited leukemia growth in vivo. We here present a novel attractive immunotherapeutic compound that potently induces anti-leukemic NK cell reactivity in vitro and in vivo as treatment option for AML.

Fig. 1. Characterization of CD276 expression in AML cell lines and patient-derived leukemia cells.

A CD276 mRNA expression in AML bone marrow (BM) (n = 350) and AML peripheral blood (PB) (n = 318) and healthy BM (n = 16) samples were analyzed using the public data sets of the BeatAML program [29]. B Flow cytometric analysis of surface CD276 expression on the indicated AML cell lines using chimeric mouse anti-human 8H8 mAb against CD276 and corresponding isotype control. Example histograms of one representative experiment out of three with similar results are shown. C The gating strategy for two representative patient-derived AML samples is outlined as follows: singlets, viable (7-AAD^-), leukemic blast marker (CD33⁺), and the percentage of CD276 expression. D CD276 expression on AML patient blasts (n = 68) is shown as percentage of CD276⁺ AML blasts and E as SFI levels. Expression levels above SFI of 1.5 and 10% for positive cells were considered as positive expression (indicated by the dotted line and shown by the corresponding pie chart). F Schematic representation of the generated CD276 antibody with a modified Fc part optimized for increased affinity to CD16 (8H8_SDIE) and the corresponding control mAb (MOPC_SDIE). G Titration of the 8H8_SDIE mAb indicated AML cell lines and H on AML patient sample analyzed by flow cytometry. MOPC_SDIE was used as an isotype control. The example data show the MFI values of one representative experiment out of a total of three with comparable results.

Fig. 2. Induction of NK cell reactivity by Fc-optimized CD276 antibody against AML cell lines.

PBMC from healthy donors were co-cultured (E:T 2.5:1) in the presence or absence of 8H8_SDIE antibody or the MOPC_SDIE control (both 1 µg/mL). A NK cell activation was analyzed by CD69 expression after 24 h. The left panels show representative flow cytometric results of THP-1 cells, the right panels show separate and pooled data of the indicated leukemic cell lines incubated with PBMC (n = 4). B NK cell activation was determined by CD25 expression after 72 h. The left panels show representative flow cytometric results of THP-1 cells, the right panels show individual and pooled data of the indicated leukemic cell lines incubated with PBMC (n = 4). C NK cell degranulation was analyzed by CD107a expression after 4 h. The left panels show representative flow cytometric results of THP-1 cells, the right panels show individual and pooled data of the indicated leukemic cell lines with PBMC (n = 4). D Intracellular IFNγ expression of NK cells within PBMC was characterized by CD3^-CD56⁺CD16⁺ counterstaining and determined by flow cytometry after 4 h. The left panel shows exemplary data from THP-1 cells, the right panels show individual and pooled data of the indicated leukemic cell lines incubated with PBMC (n = 4). E, F Supernatants of the respective co-cultures were analyzed after 24 h for the release of the immunoregulatory molecules TNF, IL-2, IFNγ and for the effector molecules granzyme A (GrzA), granzyme B (GrzB), perforin (PFN) and granulysin (Grly) by Legendplex assay. The heatmap diagrams show individual results for the leukemic cell lines indicated and different PBMC (n = 4).

Fig. 3. Induction of targeted cell lysis in AML by 8H8_SDIE.

PBMC from healthy donors were co-cultured with the indicated leukemic cells in the presence or absence of 8H8_SDIE mAb or MOPC_SDIE control (both 1 µg/mL). A Targeted lysis of AML cells was determined by Europium-based cytotoxicity assays after 2 h of incubation. The left panel shows example data with THP-1 cells and a PBMC donor at different E:T ratios. The right panel shows pooled data from EOL-1, U937 and THP-1 cell lines with different PBMC donors (n = 6-9) (E:T 80:1). B Lysis of AML cell lines with different PBMC donors (n = 3-4) was analyzed by flow cytometry after 24 h (E:T 20:1). C Lysis of the different AML cell lines with PBMC from healthy donors (n = 3-4) was analyzed by flow cytometry after 72 h (E:T 20:1). D Survival of AML cell lines was determined using a live cell imaging system. THP-1 cells were cultured with PBMC from healthy donors (n = 4) (E:T 80:1) for 96 h. Dead target cell areas were normalized to the initial target cell area’s initial measurement at T = 0 h.

Fig. 4. Induction of NK cell reactivity and cytotoxicity by Fc-optimized CD276 antibody against primary AML cells.

PBMC from healthy donors were co-cultured with primary AML cells in the presence or absence of 8H8_SDIE or MOPC_SDIE control (both 1 µg/mL) (E:T 2.5:1). A–D The left panels show representative flow cytometry results obtained from an AML patient sample and a PBMC donor, the right panel shows combined data from AML patients (n = 6) and PBMC from healthy donors (n = 4). A NK cell activation was analyzed based on the CD69 expression after 24 h. B NK cell activation was determined by the CD25 expression after 72 h. C NK cell degranulation was determined by the CD107a expression after 4 h. D Intracellular IFNγ expression in NK cells within PBMC, characterized by CD3^-CD56⁺CD16⁺ counterstaining, was determined by FACS after 4 h. E Culture supernatants were analyzed after 4 h for release of immunoregulatory molecules TNF, IL-2, IFNγ and for the effector molecules granzyme A (GrzA), granzyme B (GrzB), perforin (PFN), granulysin (Grly) by Legendplex assays. The heatmap plots show results for AML patients (n = 2) with PBMC from healthy donors (n = 4). F Targeted lysis of primary AML cells was determined by Europium cytotoxicity assays after 2 h of incubation. In the left panel exemplary data obtained from a healthy PBMC donor and a primary AML sample at different E:T ratios is shown. In the right panel, pooled data for primary AML cells (n = 3) were obtained with PBMC from healthy donors (n = 3) (E:T 80:1). G Survival of primary AML cells was determined using a live cell imaging system. The primary AML sample was cultured with PBMC from 4 healthy donors (E:T 80:1) for 140 h. The live target cell areas were normalized to the respective initial target cell area’s initial measurement at T = 4 h. H Lysis of primary AML cells with PBMC from healthy donors (n = 4) was analyzed by flow cytometry after 72 h (E:T 20:1). In the left panel, separate data for each AML patient with PBMC from healthy donors (n = 4), in the right panel pooled data for all AML patients (n = 4) (E:T 20:1) are shown.

Fig. 5. Determination of toxicity and efficacy of the 8H8_SDIE mAb in vivo.

A Immunocompromised NSG mice (n = 4/group) were injected with human PBMC (20 × 10⁶ cells, i.v.) and treated with an CD3 antibody (clone UCHT1, 20 µg, i.v.) as control or 8H8_SDIE (20 µg, i.v.). Blood was collected 24 h after injection and serum levels of IFNγ, IL-2, TNF, IL-6 were measured by Legendplex. B Luciferase-expressing U937 cells (1 × 10⁶) were injected intravenously into NSG mice (n = 5/group) followed by PBMC injection (40 × 10⁶, i.v.) and 8H8_SDIE (20 µg, i.v.) on the next day. AML tumor burden was monitored over time by bioluminescence imaging. The left panel shows exemplary bioluminescence images of mice from the indicated treatment groups on days 7 and 14. The middle panel shows the quantitative analysis of AML tumor burden measured by luminescence for all treatment groups on days 7 and 14. The right panel shows the AML tumor burden measured twice a week over the 2-week period.