Targeting the Membrane-Proximal Domain of CD33 to Maximize the Efficacy of Natural Killer Cell-Based Immunotherapies

Background: Improved survival of some patients with AML with the antibody-drug conjugate gemtuzumab ozogamicin (GO) validates CD33 as a therapeutic target, but GO is often ineffective. As one limitation, hP67.6 (used in GO) and almost all other CD33 antibodies recognize the membrane-distal V-set domain. This may be problematic given existence of a CD33 variant lacking this domain. Moreover, we have shown that CD33/CD3 bispecific antibodies (BiAbs) and chimeric antigen receptor (CAR)-modified T cells binding CD33 closer to the cell membrane exert greater T-cell mediated cytotoxicity than those binding distally. Here, we investigated whether this principle applies to CD33-targeted therapies harnessing natural killer (NK) cells.

Methods: We compared NK cell-mediated cytotoxicity against human leukemia cells expressing full-length CD33 (CD33^FL) vs. isogenic cells expressing similar levels of a CD33 molecule lacking the membrane proximal C2-set domain (CD33^ΔE3-4), bringing the V-set domain into immediate cell membrane proximity. We studied 3 types of therapeutics: 1) unconjugated V-set-directed CD33 IgG1 antibodies; 2) CD33^Vset/CD16a BiAbs; and 3) CD33^V-set-directed CAR-NK cells. NK cell-mediated cytotoxicity was determined flow cytometrically in co-culture assays by quantifying cell numbers and proportion of non-viable target leukemia cells.

Results: First, we investigated antibody-dependent cell-mediated cytotoxicity (ADCC) of CD33^V-set antibodies with NK-92 cells transduced with high-affinity CD16a (NK-92^CD16a) and primary human NK cells as effectors. Three different CD33 antibodies (including lintuzumab) elicited dose-dependent cytotoxicity against human leukemia cells overexpressing CD33^FL or CD33^∆E3-4 but not against cells lacking CD33. Importantly, the CD33^V-set antibodies exhibited greater ADCC against cells expressing CD33^∆E3-4 than cells expressing CD33^FL. We then tested the effect of membrane proximity on the efficacy of CD33^Vset/CD16a BiAbs in IgG4-scFv format. Like CD33^V-set antibodies, the CD33^Vset/CD16a BiAb (using sequences from lintuzumab) induced greater cytotoxicity against cells expressing CD33^∆E3-4 than cells expressing CD33^FL. In a third series of experiments, we investigated the effect of membrane proximity on the efficacy of human NK cells (KHYG-1 cells) transduced with CD33^V-set-directed CARs. Both the lintuzumab- and hP67.6-based CAR-NK cells demonstrated dose-dependent, CD33-specific cytotoxicity. Both CAR-NK cell products exhibited greater cytotoxicity against cells expressing CD33^∆E3-4 than cells expressing CD33^FL. Consistent with this notion of greater activation upon membrane proximal engagement, CD33^V-set CAR-NK cells showed higher intracellular TNFα and IFNγ levels after co-culture with CD33^∆E3-4-expressing cells compared to CD33^FL-expressing cells.

Our observations provided the rationale to explore CD33^C2-set-directed therapeutics. We previously generated a panel of murine and human antibodies that bind the C2-set domain regardless of the presence or absence of the V-set domain (“CD33^PAN” antibodies). In the presence of NK-92^CD16a cells, all 4 CD33^PAN antibodies we tested induced CD33-specific ADCC against human AML cell lines. Finally, we generated CD33^C2-set-directed CAR-NK cells using the sequences from 5 CD33^PAN antibodies. Compared to non-targeting CAR-NK cells, CD33^PAN CAR-NK cells induced substantially greater cytotoxicity against AML cell lines as well as primary human AML cells with a range of CD33 molecules expressed on leukemic cells.

Conclusions: Our data indicate decreasing the distance between CD33 binding epitope and leukemia cell membrane enhances the efficacy of CD33-directed NK cell therapies. These findings are the first to show that membrane proximity may matter for ADCC efficacy even for a relatively small target antigen like CD33, with a change in epitope position by ~4 nm sufficing to yield differences. Our studies are also the first to show an advantage of targeting CD33 membrane-proximally to enhance CAR-NK efficacy. Together, our findings support the further development of CD33^C2-set-directed NK cell-based therapies for AML and other CD33-expressing neoplasms.