Abstract
Natural killer (NK) cells are a key part in the innate immune system and have the ability to recognize diverse types of tumors and virally-infected targets. NK cells represent an attractive cell population for adoptive immunotherapy due to their ability to kill target cells in a human leukocyte antigen (HLA) non-restricted manner and without prior sensitization. Clinical studies using IL-2 activated NK cells demonstrate significant anti-tumor effects when adoptively transferred into patients with refractory leukemia (mainly AML). However, there has been a more limited response observed in clinical trials for the treatment of ovarian cancer and other solid malignancies. Chimeric antigen receptors (CARs) consist of an antigen-specific single chain antibody variable fragment fused to intracellular signaling domains derived from receptors involved in lymphocyte activation. CARs targeting various tumor-associated antigens have been developed and tested via expression in primary T cells with promising clinical results. However, engineering these T cells must be done on a patient-specific basis, thus limiting the number of patients who can be treated.
In order to produce a potential targeted, “off-the-shelf” product suitable to treat patients with diverse tumors or chronic infections, we have generated human pluripotent stem cells with stable CAR expression. Previous studies by our group demonstrate that human embryonic stem cells (hESCs) and induced pluripotent stem cells (iPSCs) provide an accessible, genetically tractable, and homogenous starting cell population to develop NK cells. We use a combined approach using “Spin-EB”- mediated differentiation of hESCs/iPSCs, followed by co-culture with artificial antigen presenting cells (aAPCs) that express mbIL-21. Using this strategy, we can generate 109 NK cells from a population of approximately 106 undifferentiated hESCs or iPSCs. This GMP compatible method is fully defined, without xenogeneic stromal cells or serum.
Here, we have expressed both an anti-CD19 (targeting B cell malignancies) and an anti-mesothelin CAR (targeting ovarian cancer cells and other adenocarcinomas) in both hESCs and iPSCs. Using the Sleeping Beauty transposon system, both hESCs and iPSCs have been genetically engineered to express 3rd generation CARs, which express a single chain antibody fragment recognizing either CD19 or mesothelin, a CD8α hinge region, the transmembrane protein CD28, a co-stimulatory protein 4-1BB, and the activating domain CD3ζ. NK cells derived from hESCs/iPSCs with or without CAR expression are phenotypically similar to NK cells isolated from peripheral blood. These NK cells are CD56+, CD94+/CD117-, Nkp44+, Nkp46+, NKG2A+, NKG2D+, and KIR+. In 51Cr release assays against tumor targets expressing either CD19 or mesothelin, NK cells expressing the corresponding CAR show an enhanced killing ability. In cell lines lacking CD19 or mesothelin expression, the engineered cell lines exhibit equal activity compared to their non-engineered counterparts. Specifically, at a 10:1 effector:target ratio, anti-CD19 CAR+ iPSC-NK cells kill 58% of Lax7R cells (a CD19+ ALL cell line) compared to just 5% cell killing by CAR- iPSC-NK cells. Anti-CD19 CAR+ iPSC-NK cells also killed 2 other CD19+ ALL cell lines (018Z and Raji) better than CAR- iPSC-NK cells killing 63% vs 18% and 61% vs 8%, respectively. Similar results are seen against the mesothelin+ ovarian tumor line A1847. Here, anti-mesothelin CAR+ iPSC-NK cells kill 39% vs 14% for CAR- iPSC-NK cells. Currently, CAR-expressing NK cells derived from hESCs and iPSCs are being tested in vivo against both mesothelin+ ovarian tumor lines and CD19+ leukemia cells. Together, these studies demonstrate engineering hESCs and iPSCs with tumor-specific receptors provides a novel strategy to produce targeted NK cells suitable for immune therapies against refractory malignancies.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.