CAR‐NK cells from engineered pluripotent stem cells: Off‐the‐shelf therapeutics for all patients

Abstract Clinical success of adoptive cell therapy with chimeric antigen receptor (CAR) T cells for treating hematological malignancies has revolutionized the field of cellular immunotherapy. However, due to the nature of utilizing autologous T cells, affordability and availability are major hurdles, in addition to scientific challenges relating to CAR‐T therapy optimization. Natural killer (NK) cell is a specialized immune effector cell type that recognizes and kills targets without human leukocyte antigen (HLA) restriction and prior sensitization. CAR‐NK cells do not cause graft vs host disease and can be obtained from unrelated donors as well as pluripotent stem cells (PSC), representing an ideal off‐the‐shelf therapeutics readily available for patients. Furthermore, unlike cytotoxic T cells, NK cells specifically target and eliminate cancer stem cells, which are the cells causing relapse and metastasis. PSCs can be genetically manipulated and engineered with CARs at the pluripotent stage, which allows the establishment of permanent, stable, and clonal PSC‐CAR lines for the manufacture of unlimited homogenous CAR‐NK cells. Multiple master PSC‐CAR cell banks targeting a variety of antigens for cancer, viral infection, and autoimmune diseases provide inexhaustible cell sources for all patients. Development of a next‐generation 3D bioreactor platform for PSC expansion and NK cell production overcomes major barriers related to cost and scalability for CAR‐NK product.

1 | CAR-T THERAPY: PROMISE AND CHALLENGES

| The promise
Adoptive cell transfer (ACT) for the treatment of human diseases has advanced rapidly in the past few years, especially for cancer. Chimeric antigen receptor (CAR), artificially engineered to express on the surface of immune effector cells, functions as a GPS and redirects effector cells toward their targets such as tumors 1 and has revolutionized the ACT therapeutic field. The success of autologous CD19 CAR-T cell therapy against hematological malignancies, such as chronic lymphocytic leukemia (CLL), acute lymphocytic leukemia (ALL), and non-Hodgkin lymphoma (NHL), represents one of the most exceptional breakthroughs in cancer immunotherapy in the past decade. [2][3][4][5] This success and the approval of CD19 CAR-T cell therapy by the U.S. Food and Drug Administration (FDA) have prompted the exploration, all over the world, of CAR-T clinical trials targeting different antigens for different tumors. 6

| The challenges
The success of CD19 CAR-T therapy, however, is not accomplished without side effects. The eradication of tumors by CD19 CAR-T cells leads to the release of several inflammatory cytokines, including IL-6 and TNF-α, which is termed cytokine release syndrome (CRS). In most patients, CRS is associated with fever and hemodynamic compromise, which can be fatal without proper intervention. Neurotoxicity has also been observed concurrent with or following CRS in some patients treated with CD19 CAR T cells, the cause of which is not clear although there are reports that CD19 CAR-T cells may penetrate the blood-brain barrier and attack cells expressing CD19. [2][3][4][5]7 The translation of CAR-T therapy beyond B-cell malignancies to other tumors (including other blood malignancies) still faces several obstacles that need to be addressed [6][7][8][9] : First, specificity: most CARs are targeting tumor-associated antigens (TAAs) that are also expressed on normal tissues, which leads to severe toxicity. Although there are different approaches to minimize the scale of toxicity, on-target/off-tumor toxicity can be solved completely by identifying neoantigens, which are derived from tumorspecific gene mutations (drivers), as their formation and expression are restricted to malignant tumor cells. 7,10 Second, resistance/escape: as TAAs are not required for tumor cell survival, loss of TAA expression is the major cause of development of resistance to CAR-T therapies. A strategy to circumvent tumor resistance/escape is to target multiple TAAs simultaneously by constructing bivalent-or multivalent-specific CAR-T cells, such that only tumor cells that lack expression of all target molecules would escape CAR-T therapy. 11 Other approaches are also explored with some success. 7,12,13 Third, solid tumors: although some satisfactory efficacy results were observed in some leukemic patients, especially with CD19 CAR-T therapy for B-cell lymphoma and BCMA CAR-T for myeloid myeloma, comparable efficacy in solid tumors has not yet been achieved. These disappointing results can be attributed to several factors, including [6][7][8][9] the inability of the CAR T cell to infiltrate into solid tumors, immunosuppressive molecules, and cells in the tumor microenvironment (TME), heterogenous expression of target antigens in tumor cells, and intrinsic T cell dysfunction and exhaustion. Regional or intratumor injection instead of systemic administration has been shown to be superior at least in animal model studies. 14 Combination of CAR-T with modulators of the TME, such as checkpoint inhibitors, have shown efficacy in some cancers. Other strategies include (a) depleting Tregs and myeloid-derived suppressor cells with blocking antibodies, (b) engineering CAR T cells to secrete extracellular matrix-degrading enzymes that degrade cancer-associated fibroblasts, (c) altering CAR-T cell metabolic profiles to enhance their function in hostile TME, and (d) constructing bivalent/multivalent CAR-T cells. Some of these approaches are now undergoing clinical evaluation. 7,15,16 Fourth, manufacturing: all the above CAR T optimizations are biological challenges, and the manufacturing of CAR-T cells is another challenge. patients. 17 The allogenic off-the-shelf CAR T cells from healthy donors is an alternative strategy to address the complexities of manufacturing and high costs of individualized CAR-T cell products, but there is still the risk of GvHD caused by less than even 1% of TCR+ cells in TCR knock-out CAR-T cell preparation, and early-phase clinical trials have met some hurdles. 18,19 On July 6, 2020, the FDA issued a clinical hold for the phase I trial of a universal CAR-T product, UCARTCS1A be detected and destroyed by T cells. 22 Most importantly, adoptive NK cell transfer does not require strict HLA matching and lacks the potential to cause GvHD, a critical risk imposed by any T cell immunotherapy. 23 Therefore, NK cells can be off-the-shelf allogeneic therapeutics.

Similar to T cells, NK cells have been engineered with CAR expressed
on the surface to target a specific antigen, enhancing their natural cytotoxicity with specificity. 24,25 Recently, Liu et al 26 28,29 This is advantageous because cancer stem cells are resistant to both traditional chemotherapy and radiotherapy, as well as immunotherapy, which leads to relapse and metastasis.

| Donor-based CAR-NK: Challenges
The main sources of donor NK cells are peripheral blood (PB) and UCB, and their concentration of 1 Â 10 8 cells/L comprises $1-2% of total WBCs, or 0.01-0.02% of all cells in the PB. 30 Although in vitro exponential expansion with/without artificial antigen presenting cells (aAPC) from a single donor will be able to provide enough cells for a single or limited number of patients, the doses are still limited from a single-donor phlebotomy, [31][32][33] which is far less for an ideal off-theshelf allogeneic commercial therapeutic product. Functional and expansion variations among different donors add another burden for batch to batch validation as well as manufacture delay, which lead to inconsistent clinical outcome. Furthermore, transfection of NK cells with viral CAR-constructs was associated with low or variable levels of transgene expression and unfavorable effects on cell viability, limiting other genetic engineering to improve the efficacy of CAR-NK cells. 24 NK cell lines, such as NK9-2, have been pursued for clinical applications. These cells are cancerous in nature and need to be irradiated before infusion into patients, compromising their cancer killing capabilities. 34 As an ideal cellular immunotherapy product, cells should possess high specificity on targets with minimized off-target toxicities; be prepared from sustainable sources so every patient receives the same cell product; be able to be stored and ready for patient administration whenever needed, not be prepared for each individual patient; and be centrally manufactured in a current good manufacturing practice (cGMP) facility and be able to distribute worldwide. CAR-NK cells from donor sources fulfill most of above criteria but miss the most important property for an allogenic off-the-shelf cell product-sustainability. One potential source of CAR-NK cells is pluripotent stem cells T A B L E 1 Comparison of CAR-T vs donor-CAR-NK and PSC-CAR-NK cell products

| CAR-NK cells derived from PSCs: Advantages
PSCs provide a completely novel pathway for the generation of gene engineered cell products such as NK cells. There are several advantages of PSC-CAR-NK compared with donor based CAR-NK cells (Table 1). First, CAR construct can be inserted into undifferentiated PSCs to establish permanent, stable, and clonal PSC-CAR lines, which can be used for unlimited manufacture of homogenous CAR-NK cells, a true off-the-shelf and consistent product for all patients. 41

| 3D platform for industrial scale production of PSC-NK cells
We have developed a novel 3D-bioreactor platform to efficiently convert human PSCs into highly pure NK cells continuously in a single bioreactor (Figure 1). 48  NK cells, the increase of which is associated with disease regression in AIDS patients. 49 We have reproducibly generated $10 10 pure NK cells with a 300 mL bioreactor and this process has been repeated with multiple hESC and iPSC lines with/without gene editing or CAR insertion.
In vitro assays showed that PSC-NK cells produced by 3Dbioreactors were capable of secreting IFNγ, expressing CD107a (degranulation) in response to stimulation, and also displayed potent natural cytotoxicity against multiple tumor cells including leukemia cells as well as pancreatic, ovarian, colorectal and lung cancer cells.
We also demonstrated that PSC-NK cells harbored anti-virus activity, and specifically killed normal cells infected with influenza A, Dengue, Zika, and HIV viruses, but not uninfected normal cells. 48 This 3D PSC-NK manufacture platform uses defined materials without serum or feeder cells, and is scalable and reproducible for

| THE FUTURE PERSPECTIVE
Human PSCs promise to provide a potentially inexhaustible source of CAR-NK cells for immunotherapy. However, before PSC-CAR-NK cells can be used in the clinic, it is important to understand the risks and challenges associated with these therapeutic products.

One of the most important characteristics of undifferentiated
PSCs is teratoma formation. 50   F I G U R E 3 Genetically engineered PSC-CAR-NK cell therapeutic strategies for cancer, viral infection, and autoimmune diseases