Bradley Aaron Krasnick

Phenotypic analysis of blood NK cells from AM patients and healthy donors.

Enhanced ML responses are partially dependent on NKG2D and NKp46. A, Expression of the NKG2D ligand MICA/B in primary tumors. Patient-derived control and ML NK cells were stimulated for 6 hours with autologous tumor with or without αNKG2D (2.5 μg/mL) plus αNKp46 (2.5 μg/mL) blocking antibodies and the frequency of IFNγ was evaluated by flow cytometry. Representative flow cytometry dot plots (B) and summary data (C) showing reduction in IFNγ production by control and ML NK cells after NKG2D and NKp46 blockade. D, Patient-derived control and ML NK cells were stimulated in the presence of autologous tumor with or without blocking antibodies and specific killing was measured by 51Cr release assay. Bars represent mean ± SEM from all patients. Two-way ANOVA and paired t test *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. n = 7.

Purpose:Treatment of advanced melanoma is a clinical challenge. Natural killer (NK) cells are a promising cellular therapy for T cell–refractory cancers, but are frequently deficient or dysfunctional in patients with melanoma. Thus, new strategies are needed to enhance NK-cell antitumor responses. Cytokine-induced memory-like (ML) differentiation overcomes many barriers in the NK-cell therapeutics field, resulting in potent cytotoxicity and enhanced cytokine production against blood cancer targets. However, the preclinical activity of ML NK against solid tumors remains largely undefined.Experimental Design:Phenotypic and functional alterations of blood and advanced melanoma infiltrating NK cells were evaluated using mass cytometry. ML NK cells from healthy donors (HD) and patients with advanced melanoma were evaluated for their ability to produce IFNγ and kill melanoma targets in vitro and in vivo using a xenograft model.Results:NK cells in advanced melanoma exhibited a decreased cytotoxic potential compared with blood NK cells. ML NK cells differentiated from HD and patients with advanced melanoma displayed enhanced IFNγ production and cytotoxicity against melanoma targets. This included ML differentiation enhancing melanoma patients' NK-cell responses against autologous targets. The ML NK-cell response against melanoma was partially dependent on the NKG2D- and NKp46-activating receptors. Furthermore, in xenograft NSG mouse models, human ML NK cells demonstrated superior control of melanoma, compared with conventional NK cells.Conclusions:Blood NK cells from allogeneic HD or patients with advanced melanoma can be differentiated into ML NK cells for use as a novel immunotherapeutic treatment for advanced melanoma, which warrants testing in early-phase clinical trials.

Infiltrating NK cells exhibit a tissue resident phenotype.

Memory-like NK cells from patients with advanced melanoma exhibit enhanced ability to control autologous tumor. A, PBMC containing patient NK cells were preactivated as indicated in Fig. 2A. Melanoma cell lines generated from metastatic tissue were used as target cells for the functional assays. B, Representative flow cytometry plots of control and ML NK cells from a patient with advanced melanoma stimulated with autologous tumor. Numbers indicate percentage of positive cells. ML NK cells (green) from patients with melanoma patients display superior ability to produce IFNγ upon stimulation with autologous tumor (5:1 E:T ratio) compared with control NK cells (blue; C). D, ML NK cells from patients with advanced melanoma exhibit superior ability to kill autologous melanoma targets compared with control NK cells at different E:T ratios. Killing was evaluated in a standard 4-hours 51Cr release assay. Bars represent mean ± SEM from all patients. Two-way ANOVA with Tukey post hoc analysis. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. n = 7.

ML NK cells from normal donors exhibit enhanced ability to control melanoma compared with conventional NK cells. A, PBMC-derived NK cells from HD were activated with IL12/IL15/IL18 (or IL15 alone as control) for 16–18 hours (1). Activated NK cells were differentiated into ML NK cells in vitro during 7 days in presence of low-dose IL15 (2). Upon a second stimulation (3), ML NK cells exhibited enhanced responses against tumor target and cytokines, compared with control or conventional NK cells. B, Representative flow cytometry and gating strategy to evaluate cytokine secretion and degranulation (CD107a+ cells) of control and ML NK cells stimulated with melanoma cell lines (DM6, M14) and K562 cells. Numbers indicate percentage of positive cells. Summary data of IFNγ (C), TNF (D), and degranulation of control (blue) and ML (green) NK cells (E) restimulated 6 hours in vitro with DM6, M14, and K562 cells. Cells lines were used at a 10:1 E:T ratio. ML NK cell exhibited superior killing ability against DM6 compared with control NK cells in a standard 4-hours 51Cr release assay (F) and a 72-hour Incucyte assay (G). Purified NK cells were used for the cytotoxic assays. Bars represent mean ± SEM. Two-way ANOVA test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. n = 3 in C–E, n = 9 in F. G shows a representative experiment out of three independent experiments.

Cytokine production and degranulation of control and ML NK cells from AM patients and normal donors and effect of antibody blockade.

Immunohistochemistry staining of S100, Melan A and SOX10 in patient derived cell lines

Supplementary methods and tables

Mass cytometry analysis and main immune cell subsets in PBMC, Met and ULN from AM patient samples.