Providing a Homing Receptor for CAR Engineered NK Cells - Improving Cellular Immunotherapy for B-Cell Lymphoma

Background

Despite some robust initial responses, anti-CD19 Chimeric Antigen Receptor (CAR) T-cell therapy can be associated with significant short-term (cytokine release syndrome) and long-term (B-cell deficiency) toxicities. CAR-engineered natural killer (NK) cells potentially provide a safer alternative while maintaining efficacy.

Activated Natural Killer (aNK^TM) cells are a clinical grade cell line derived from the NK-92^R cell line that has demonstrated potent cytotoxicity towards a broad spectrum of malignant cell lines as well as safety and efficacy in phase I trials. Variants of the aNKcell line are currently in Phase I/II clinical trials: a CAR-expressing aNK cell line and the haNK^TM cell line, which have been engineered to carry a high-affinity version of the CD16/FcγRIII receptor to allow for combination therapy with monoclonal antibodies. haNK cells have also been genetically modified to express an endoplasmic reticulum-retained version of IL-2 (ERIL-2), which provides IL-2 independence and limits IL-2 secretion to sub-physiological, safe levels.

A key factor for the efficacy of cellular immunotherapies against a given target is biodistribution, which affects the local effector to target ratio. Inability to reach the tumor cells, either by lack of homing or by the accumulation of extracellular matrix (ECM) surrounding a tumor, can be responsible for the clinical failure of even the most effective CAR. The chemokines CCL19 and CCL21 drive recruitment of CCR7-expressing immune cells to secondary lymphoid organs. Engineering aNK cells to express the CCR7 receptor is likely to improve their efficacy by increasing their targeted migration to lymphoma tumor sites.

Methods and Results

Clinical grade aNK cells were electroporated with a non-viral vector containing the CCR7 receptor, an anti-CD19 CAR, and a high affinity CD16 receptor. To assay the migration of these engineered cell lines, a modified Boyden Chamber assay was performed using Matrigel coated Transwells. K562 cells or modified K562 cells engineered to express CCL19 (K-19) were placed in the destination chamber and CFSE-stained effector cells were placed in the top well. After 24 hours, cells in the bottom well were analyzed by flow cytometry to measure the number of effectors which had migrated through the Matrigel (Fig 1a). The excellent activity of the CAR in stably transfected cells was confirmed against SUP-B15 cells (aNK-resistant), while the ADCC activity was tested against a SUP-B15 variant expressing CD20, but engineered to lack the CD19 antigen (Sup-B15 CD19-, CD20+).

Migration towards human lymph node chemokine CCL19 was also tested in vivo in NSG mice with bilateral subcutaneous tumors - with parental K562 in one flank and K-19 tumors on the contralateral flank. CFSE-stained effector cells were delivered via tail vein injection once average tumor size reached 100mm³ and following randomization. Tumors were then harvested at multiple time points, dissociated, and the number of infiltrating effectors in each tumor compared by flow cytometric analysis. (Fig 1b).

In testing, monoclonal cell lines expressing all components of the polycistronic system displayed preferential migration towards CCR7 chemokines both in vitro and in vivo, as well as robust cytotoxicity vs. K562 (92.4% +/- 2.4% at 5:1 E:T), Sup-B15(97% +/- 0.6% at 5:1 E:T), and Sup-B15(CD19-, CD20+) when pre-incubated with Rituximab(83.2% +/- 2.8% at 5:1 E:T) but not with control antibody Trastuzumab (22.3% +/- 1.1% at 5:1 E:T) in standard cytotoxicity and ADCC assays.

Conclusion

We show here that the incorporation of a CCR7 receptor into an off the shelf CAR engineered NK cell line improves their homing towards lymph node chemokines both in vitro and in vivo. This improved homing should result in a greater ratio of effector to target in lymphoid tissue, and maximize the immunogenic cell death.

View largeDownload slide

View largeDownload slide

Close modal