Introduction
Cancer immunotherapy has rapidly advanced in recent years and is now widely recognized as a major pillar of cancer treatment alongside surgery, chemotherapy, targeted therapy, and radiotherapy. The development of immune checkpoint inhibitors and adoptive cell therapies has significantly improved outcomes for patients with several malignancies (1–3). Immune cell-based cancer therapies have also shown remarkable clinical success in hematological cancers.
Autologous chimeric antigen receptor (CAR) T-cell therapies targeting CD19 and BCMA have produced durable responses in patients with relapsed or refractory B-cell and plasma cell malignancies and have transformed the treatment landscape for diseases such as acute lymphoblastic leukemia, diffuse large B-cell lymphoma, and multiple myeloma (2). These successes have established adoptive cell therapy as a powerful modality for cancer treatment and have stimulated intense interest in the development of next-generation cellular immunotherapies. Despite these advances, the broader clinical application of autologous CAR-T cell therapies remains limited by several practical and biological challenges (1, 3, 4).
Manufacturing of autologous products requires individualized production from each patient, involving complex and labor-intensive processes that contribute to high treatment costs. In addition, the manufacturing timeline for autologous therapies typically requires several weeks, which can delay treatment for patients with rapidly progressing diseases. Product variability resulting from differences in patient-derived starting material can also affect therapeutic consistency and efficacy.
Furthermore, autologous T-cell therapies are associated with potentially severe toxicities, including cytokine release syndrome and neurotoxicity. These limitations have stimulated growing interest in alternative immune cell platforms that enable scalable, standardized, and readily available cellular therapies (3).
Natural killer (NK) cells are innate immune lymphocytes that play a critical role in the immune surveillance against infected or transformed cells. Unlike T cells, NK cells do not require antigen-specific priming and can recognize malignant cells through a balance of activating and inhibitory receptor signals (5, 6). This mechanism enables NK cells to detect stressed or transformed cells that have downregulated major histocompatibility complex (MHC) molecules, a phenomenon commonly referred to as “missing-self” recognition (7).
Upon activation, NK cells exert potent cytotoxic effects through multiple mechanisms, including the release of perforin and granzymes, as well as death receptor–mediated pathways such as Fas ligand and TRAIL. In addition to direct tumor cell killing, NK cells produce cytokines such as interferon-γ (IFN-γ) and tumor necrosis factor-α (TNF-α), which can enhance antitumor immunity and promote interactions with other immune cell populations (8). NK cells can also eliminate antibody-coated tumor cells through antibody-dependent cellular cytotoxicity (ADCC) mediated by Fc receptors, allowing them to cooperate with therapeutic monoclonal antibodies (9, 10).
These diverse mechanisms make NK cells an attractive cellular platform for cancer immunotherapy. However, current NK cell–based therapies face several limitations that hinder their widespread clinical implementation. Primary NK cells derived from peripheral blood or cord blood often exhibit donor-to-donor variability and limited expansion potential, complicating large-scale manufacturing.
In addition, NK cells typically show limited persistence after infusion, which may restrict long-term therapeutic efficacy. NK cells are also sensitive to cryopreservation, which can impair their biological activity and antitumor efficacy post-thaw. Consequently, generating consistent, scalable, cryopreserved NK cell products suitable for clinical use remains a major challenge for the field (11–13).
Induced pluripotent stem cells (iPSCs) offer a promising strategy to address many of these limitations. iPSCs possess the capacity for unlimited self-renewal and can differentiate into a wide range of cell types, including NK cells (14, 15). Importantly, iPSC-based platforms enable the generation of large numbers of standardized cellular products derived from a well-characterized master cell bank, thereby facilitating scalable manufacturing of off-the-shelf cell therapies (15). In addition, iPSCs provide an ideal platform for extensive genetic engineering at the pluripotent stage, allowing the introduction of multiple genetic modifications designed to enhance the efficacy, persistence, and safety of derived immune cells.
These advantages make iPSC-derived immune cells an attractive approach for developing next-generation cellular immunotherapies (16–18). More recently, universal donor iPSC (UDC-iPSC) technologies using adeno-associated virus (AAV)–mediated gene editing have been developed to overcome immune rejection following allogeneic cell transplantation. By disrupting the B2M and RFXANK genes, which are required for surface expression of HLA class I and class II molecules, respectively, and introducing a single-chain HLA-E construct, UDCs and their differentiated derivatives can evade recognition by host T cells while avoiding NK cell–mediated rejection (19–21).
These strategies have the potential to generate broadly compatible allogeneic cellular therapeutics that can be administered across multiple patients without HLA matching. Consequently, UDC technologies represent a promising approach for enabling universal, off-the-shelf cell therapy platforms.
Recent advances in iPSC-derived NK cell technologies have further accelerated the development of next-generation, off-the-shelf cellular immunotherapies (16–18). Several groups have demonstrated the feasibility of generating large numbers of functional NK cells from iPSCs under clinically scalable manufacturing conditions, leading to the emergence of multiple clinical-stage iPSC-NK cell products (15, 17, 22, 23). In parallel, extensive genetic engineering strategies have been implemented to enhance the therapeutic performance of iPSC-derived NK cells.
These approaches include the introduction of high-affinity and non-cleavable CD16 variants to improve antibody-dependent cellular cytotoxicity (ADCC) (17, 21), expression of cytokine support systems such as membrane-bound IL-15 to enhance persistence and functionality (17, 22, 23), incorporation of chimeric antigen receptors (CARs) to increase tumor specificity (16, 17), and genetic modifications designed to improve resistance to the immunosuppressive tumor microenvironment (22, 23). More recently, immune-evasive engineering strategies, including modulation of HLA molecules and expression of inhibitory ligands, have been developed to reduce host immune rejection and prolong persistence following allogeneic administration (19, 23–25). Collectively, these advances have established genetically engineered iPSC-derived NK cells as a highly promising platform for the development of universal, off-the-shelf cancer immunotherapies capable of combining enhanced persistence, immune evasion, and potent antitumor activity.
In this study, we developed universal donor iPSC (UDC-iPSC)-derived NK cells (UDC-NK) designed to enhance immune evasion, safety, and antitumor efficacy. The engineered UDC clone incorporates multiple genetic modifications, including disruption of B2M and RFXANK to eliminate HLA class I and class II expression, expression of a single-chain HLA-E construct to prevent NK cell–mediated rejection, introduction of a high-affinity FCGR3A (F158V) variant to enhance ADCC (21), and integration of a herpes simplex virus thymidine kinase (HSV-TK) suicide gene as an inducible safety switch (26–28). We also established a feeder-free and scalable manufacturing process capable of generating large numbers of functional NK cells.
Using this platform, we efficiently generated NK cells from UDCs and comprehensively characterized their phenotype and function in vitro and in vivo. We evaluated surface marker expressions, cytotoxic effector functions including natural cytotoxicity and ADCC, cytokine production, and the immune-evasive properties of the engineered NK cells in the context of allogeneic immune recognition. In addition, we validated the functionality of the integrated HSV-TK suicide gene as an inducible safety switch in vitro.
Finally, the antitumor activity of UDC-NK cells was assessed in vitro and validated in vivo using xenograft tumor models to determine their therapeutic potential. Collectively, this study establishes a genetically engineered iPSC-derived NK cell platform that may enable the development of broadly compatible, off-the-shelf NK cell therapies for cancer.
Materials and methods
Cells and cell culture
The human WT iPSC line NIH3 c.C4 was established by single-cell cloning in-house from a clinical-grade iPSC bank (Lonza, Basel, Switzerland) and used for gene editing. All parental and gene-edited iPSC lines were maintained on Vitronectin XF™ (Stemcell Technologies, Vancouver, BC, Canada)–coated culture vessels in mTeSR1 or TeSR-E8 medium (Stemcell Technologies) in a 37 °C, 5% CO2, and 5% O2 incubator. Chronic myelogenous leukemia cell line K562, CD20+ non-Hodgkin lymphoma (NHL) cell lines Raji, Ramos, Daudi, Mino, SU-DHL-4, SU-DHL-5, SU-DHL-6, P3HR-1 cells were obtained from American Type Culture Collection (Manassas, VA, USA).
Granta-519 NHL cell line was purchased from Creative Bioarray (Shirley, NY, USA). K562 cells engineered to express HiBiT protein were purchased from Promega (Madison, WI, USA). All cells were cultured in RPMI 1640 media supplemented with 2 mM GlutaMAX™ (Thermo Fisher Scientific, Waltham, MA, USA) and with 10% heat-inactivated fetal bovine serum (FBS, Thermo Fisher Scientific).
For ADCC assays, NucLight™ Red-expressing Ramos cells were established by transducing cells with the IncuCyte® NucLight™ Red Lentivirus Reagent (Sartorius, Ann Arbor, MI, USA), followed by puromycin selection. For in vivo studies, Raji-Luc and Daudi-Luc cells were established by infection with IVISbrite™ Red F-luc-Puromycin Lentiviral Particles (Revvity, Waltham, MA, USA) to introduce a Red-Shifted luciferase and a GFP reporter gene, followed by single-cell cloning.
Gene editing to establish universal donor iPSC (UDC-iPSC)
Establishing universal donor iPSC (UDC-iPSC) engineered to express high affinity CD16A F158V variant for the manufacturing UDC-NK cells for clinical applications, has been described previously (29). Briefly, the NIH3 c.7C2 iPSC line, which was used for UDC-NK manufacturing, was generated from NIH3 c.C4 WT iPSC clone through the multiple rounds of gene-editing, which includes (1) knocking out the B2M and RFXANK genes to eliminate surface expression of HLA Class I and Class II molecules to escape allogeneic rejection mediated by T cells, (2) introduction of a single-chain HLA-E chimeric receptor to protect differentiated cells from allogeneic rejection mediated by CD94/NKG2A-expressing allogeneic NK cells (19), (3) ganciclovir-inducible thymidine kinase gene cassettes as a safety switch to reduce the risk of severe, unwanted adverse events following administration of NK cell product to cancer patients (26, 28), and (4) FCGR3A F158V gene coding naturally occurred high-affinity CD16A variant to improve ADCC in combination with therapeutic antibodies. iPSC line NIH2 c.2B9 was established by disrupting B2M genes. All gene-editing was performed using recombinant adeno-associated viruses.
Cell differentiation
NK cell differentiation from iPSCs was performed using a modified protocol described previously (15). Briefly, 1,000–2,000 Trypsin-adapted single iPSCs were seeded into each well of a 24-well AggreWell™ 800 plate (Stemcell Technologies) in APEL2 medium (Stemcell Technologies) supplemented with 40 ng/mL human stem cell factor (SCF; PeproTech, Rocky Hill, NJ, USA), 20 ng/mL human vascular endothelial growth factor (VEGF; R&D Systems, Minneapolis, MN, USA), and 20 ng/mL human bone morphogenetic protein 4 (BMP-4; R&D Systems) to promote embryoid body (EB) formation. Cells were cultured at 37 °C under 5% CO2 and 5% O2.
Half-medium changes were performed every other day. EBs were maintained under these conditions for 14 days to promote the generation of CD34+ hematopoietic progenitor cells (HPCs). On day 14, HPCs were harvested in DMEM/F-12 medium (Thermo Fisher Scientific) supplemented with 15% heat-inactivated human AB serum (BioIVT, Westbury, NY, USA), 5 ng/mL IL-3 (first week only), 10 ng/mL IL-15, 20 ng/mL IL-7, 20 ng/mL SCF, and 10 ng/mL Flt3 ligand (PeproTech).
In addition, 35 nM UM171 (Xcess Biosciences, San Diego, CA, USA), a pyrimidoindole-derived small molecule previously reported to promote the self-renewal and expansion of hematopoietic stem cells and progenitor cells, was added to stabilize HPCs and enhance NK cell differentiation efficiency (30–32). The HPCs were subsequently replated onto culture vessels coated with RetroNectin™ (Takara Bio, Shiga, Japan) and human Delta-like ligand 4–Fc protein (DLL4; ACRO Biosystems, Beijing, China) to facilitate Notch signaling–mediated NK lineage commitment. Cells were further cultured for 3 weeks to promote differentiation into immature CD56+ UDC-NK cells.
Culture medium was refreshed every 3–4 days throughout the differentiation period. Cells were maintained at 37 °C under 5% CO2 in air during this stage. For the assessment of immune evasion in UDC-NK cells, the B2M KO iPSC clone NIH2 c.2B9 was differentiated into CD14+ myeloid cells using the STEMdiff™ Hematopoietic Kit (STEMCELL Technologies) following previously published protocols (33).
Activation and expansion of UDC-NK cells
CD56+ immature UDC-NK cells were harvested and resuspended at a density of 2x105 cells/mL in CTS™ NK-Xpander™ Medium (Thermo Fisher Scientific) supplemented with 7.5% heat-inactivated human AB serum, 10 ng/mL IL-15, 10 ng/mL IL-18 (R&D Systems), and 10 ng/mL IL-21 (Peprotech). NK cells were seeded into culture vessels pre-coated with 0.89 μg/mL anti-human NKp30 antibody (BioLegend, San Diego, CA, USA), 0.89 μg/mL anti-DNAM-1 antibody (Bio-Rad Laboratories, Hercules, CA, USA), and 0.89 μg/mL RetroNectin™. Cells were cultured for 3 days at 37 °C under 5% CO2 in air to stimulate the maturation and proliferation of immature NK cells.
After 3 days of activation with anti-NKp30 and anti-DNAM-1 antibodies, NK cells were harvested and resuspended in CTS™ NK-Xpander™ Medium containing 7.5% heat-inactivated human AB serum, 10 ng/mL IL-15, and 10 ng/mL IL-18. The cells were then transferred at a density of 4x105 cells/mL to disposable spinner flasks (Corning, Corning, NY, USA) placed on a magnetic stirrer (Chemglass Life Science, Vineland, NJ, USA) for further expansion at 45–55 rpm in a 37 °C incubator with 5% CO2. Following initiation of spinner flask culture, an equal volume of fresh medium containing the same supplements was added to the culture.
NK cells were expanded in spinner flasks for 4 days. 3-day NK Activation and 4-day NK Expansion cycle was repeated three times. After three cycles of activation and expansion, NK cells were harvested and cryopreserved in CryoStor® CS-10 (Stemcell Technologies) for further characterization.
At harvest, NK cell purity was typically >99% among CD45+CD56+ cells. In selected experiments, the cycle was repeated up to seven times to evaluate the robustness and scalability of the manufacturing process.
Flow cytometry
Flow cytometry was performed using an Attune NxT flow cytometer (Thermo Fisher Scientific), and the data were analyzed using FlowJo™ software (FlowJo, Ashland, OR, USA). Details of antibodies used in this study are provided in Supplementary Table 1.
Ganciclovir sensitivity of UDC-NK cells
To assess the function of the HSV-TK-based safety switch in vitro, NK cells were seeded into a 96-well flat-bottom plate in the presence of 10 ng/mL IL-15 and a titrated concentration of ganciclovir (MilliporeSigma, Burlington, MA, USA) and cultured at 37 °C in a humidified incubator with 5% CO2. Four days after the initiation of ganciclovir treatment, NK cell viability was measured using CellTiter®-Glo 2.0 Cell Viability Assay (Promega).
Immune evasion of UDC-NK cells from allogeneic PBMC T cells and NK cells
To assess the activation of allogeneic T cells after co-culture with UDC-NK cells, CD4+ and CD8+ T cells isolated using magnetic-activated cell sorting (Miltenyi Biotec, Bergisch Gladbach, Germany) from the peripheral blood of healthy volunteers (Bioworks NW, Seattle, WA, USA; Charles River Laboratories, Wilmington, MA, USA) were co-cultured with UDC-NK cells in 96-well plates for 3 days. The purity of T cells and NK cells was confirmed by flow cytometry prior to co-culture. After co-culture, T cell proliferation was measured using CellTiter®-Glo 2.0 Cell Viability Assay (Promega).
To evaluate the susceptibility of UDC-NK cells to lysis by allogeneic PBMC-derived NK cells, a DELFIA® Cytotoxicity Assay (Revvity) was performed. UDC-NK cells were labeled with BATDA reagent and used as target cells. BATDA-labeled UDC-NK cells were co-cultured with PBMC-derived allogeneic NK cells at titrated effector-to-target (E:T) ratios in 96-well round-bottom plates for 2 hours at 37 °C.
After co-culture, supernatants were harvested, and europium release from lysed UDC-NK cells was measured using time-resolved fluorescence with a Varioskan™ LUX Plate Reader (Thermo Fisher Scientific). Experiments were performed with 3 independent biological triplicates. Specific lysis was calculated as [(experimental release − spontaneous release)/(maximum release − spontaneous release)] × 100.
Cytotoxicity assays
The natural cytotoxicity of NK cells was evaluated by co-culturing NK cells with K562-HiBiT cells (Promega, Madison, WI, USA) at various effector-to-target (E:T) ratios in 96-well round-bottom plates. After 4 hours of incubation, HiBiT-tagged proteins released from lysed target cells were detected using Nano-Glo® HiBiT Extracellular Detection System (Promega) according to the manufacturer’s instructions. The luminescent signal was measured using a plate reader.
Experiments were performed with 3 independent biological triplicates. The percentage of specific lysis is calculated as follows: %Specific lysis = (Bioluminescence in the test samples − Background signal in the absence of effector cells)/(Bioluminescence in the presence of digitonin − Background signal in the absence of effector cells) × 100. To assess antibody-dependent cellular cytotoxicity (ADCC), NK cells were co-cultured with CD20+ NHL cell lines engineered to express NucLight™ Red (Sartorius) at titrated E:T ratios in the presence or absence of 1 µg/mL Rituximab biosimilar (Invivogen, San Diego, CA, USA) in 96-well flat-bottom plates.
NucLight Red+ target cells were monitored for 24 hours using an IncuCyte® S3 Live-Cell Analysis System (Sartorius).
Cytokine secretion
To assess cytokine secretion upon tumor cell stimulation, NK cells were co-cultured with K562 or Raji cells at an E:T ratio of 1:1 in 96-well flat-bottom plates, in the presence or absence of Rituximab (1 µg/mL). After 24 hours of incubation in a 37 °C with 5% CO2 incubator, the culture supernatants were harvested, and IFN-γ and TNF-α concentrations were quantified using AlphaLISA® Cytokine Assays (Revvity) according to the manufacturer’s instructions.
Murine xenograft tumor models
NOGF-IL15 (NOD.Cg-Prkdc<scid> Il2rg<tm1Sug> Fcer1g<tm1Rav> Fcgr2b<tm1Ttk> Tg(CMV-IL15)/Jic) mice, which lacks the surface expression of all Fc receptors and expresses human IL-15 under the control of the CMV promoter (32), was provided by Taconic Biosciences (Rensselaer, NY, USA) and maintained in-house under specific pathogen-free conditions. Mice were transplanted with 2x104 Raji-Luc cells or 1x105 Daudi-Luc cells via intravenous (i.v.) injection on day 0. The combination therapy with cryopreserved UDC-NK cells and Rituximab was initiated the next day and administered as indicated in figure legends.
Cryopreserved UDC-NK cells were thawed and immediately used for treatment. Tumor burden was monitored by bioluminescence imaging (BLI) using an IVIS Lumina III Imaging System (Revvity) under 2–4% inhaled isoflurane anesthesia. Mice were carefully monitored and euthanized upon reaching humane endpoints, defined as a body weight loss of ≥20%.
At the study endpoint, mice were humanely euthanized, either by cervical dislocation under deep anesthesia induced with 2–4% inhaled isoflurane or by gradual exposure to CO2 at a displacement rate of 30–70% of the cage or chamber volume per minute. All animal experiments were approved by the Institutional Animal Care and Use Committee of Omeros Corporation (Seattle, WA, USA) and conducted in an Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC International)-accredited facility.
Statistics
Dunnett’s multiple comparisons test was used for statistical analysis. Survival was analyzed using the log-rank test. All statistical analyses were performed using GraphPad Prism software (GraphPad Software, San Diego, CA, USA).
P value < 0.05 was considered statistically significant, and significance was indicated by asterisks.
Results
Establishing a universal donor iPSC (UDC) line with a safety switch and stable high-affinity FcR expression for NK cell manufacturing
We aimed to establish a robust platform for the manufacturing of NK cells from multiple gene-edited universal donor iPSC lines to support clinical applications of universal donor cell-derived NK (UDC-NK) therapies (Figure 1A). To this end, we first generated universal donor iPSCs (UDC-iPSCs) engineered to eliminate the surface expression of classical HLA class I and class II molecules and to introduce single-chain HLA-E molecules using AAV-based gene editing technologies (34). The UDC technology enables UDC-derived cell therapy products to escape from allogeneic rejection mediated by T cells and CD94/NKG2A+ NK cells after administration to patients, as previously described (19).
Using the UDC-iPSC line as the parental iPSCs, we performed additional gene editing to introduce a ganciclovir-inducible thymidine kinase safety switch, enabling elimination of UDC-derived cell therapy products in the event of severe adverse events following administration. Finally, to enhance antibody-dependent cellular cytotoxicity (ADCC), we introduced a high-affinity natural variant of the FCGR3A gene encoding CD16A, generating the iPSC line NIH3 c.7C2, which was used for NK cell manufacturing (Figure 1B). A GMP bank of NIH3 c.7C2 iPSCs uniformly expressed pluripotency markers OCT4, SOX2, and SSEA-3/4, indicating maintaining pluripotency, as confirmed by flow cytometry (Figure 1C).
A Highly-efficient, scalable, feeder-free NK manufacturing process from UDCs
Next, we established a highly efficient, scalable, and feeder-free cell differentiation process for the manufacturing of UDC-NK cells for clinical applications. The overall process consisted of three sequential steps: (1) a 14-day differentiation of CD34+ hematopoietic progenitor cells (HPCs) from iPSCs, (2) a 21-day UDC-NK cell differentiation phase, and (3) a 21-day NK activation and expansion phase consisting of three consecutive cycles of 3-day activation followed by 4-day expansion (Figure 2A). This 56-day process enabled the generation of sufficient numbers of UDC-NK cells for potential clinical use.
In the first step, UDC-iPSCs were seeded into AggreWell™ plates to promote embryoid body (EB) formation and cultured in the presence of a cocktail of cytokines and small molecules for 14 days to induce hematopoietic differentiation toward CD34+ HPCs. After HPC differentiation, CD34+CD43+ cells emerged, followed by the appearance of CD34+CD45+ cells, consistent with progression toward definitive hematopoiesis (Supplementary Figure 1A). In the second step, CD34+ HPCs were transferred to culture vessels pre-coated with RetroNectin™ and Delta-like ligand-4 (DLL4) and cultured in the presence of cytokine cocktails and the small molecule UM171.
DLL4-mediated Notch signaling has been reported to promote NK lineage commitment (35), while UM171 enhances HPC self-renewal and expansion, thereby improving NK cell yield (30, 31, 36). After 3 weeks of differentiation, the majority of cells (typically 70-90%) exhibited a CD45+CD56+ NK cell phenotype. In addition, consistent expression of activating receptors, including NKp30 and DNAM-1, was observed (Supplementary Figure 1B).
In the third step, CD56+ UDC-NK cells were harvested and transferred to culture vessels coated with anti-NKp30 and anti-DNAM-1 antibodies and cultured in medium containing IL-15, IL-18, and IL-21 for 3 days to induce activation. Activated UDC-NK cells were subsequently transferred to spinner flasks and cultured in the presence of IL-15 and IL-18, but in the absence of IL-21, for further expansion for 4 days. These activation and expansion cycles were repeated three times to promote NK cell maturation and maximize cell yield.
UDC-NK cells consistently expanded by approximately 10-fold during each 7-day activation/expansion cycle, and three consecutive cycles resulted in an overall expansion of >1,000-fold within 21 days (Figure 2B). Following three rounds of activation and expansion, the cells exhibited a uniform population with >99% CD45+CD3-CD56+ NK cell phenotype (Figure 2C). After cell harvest, UDC-NK cells were cryopreserved for further characterization in vitro and in vivo.
These results demonstrate the establishment of a robust, scalable manufacturing platform for generating large numbers of UDC-NK cells.
We further characterized the phenotype of UDC-NK cells by assessing the expression of key NK cell receptors and cytotoxicity-associated molecules, as well as HLA phenotype by flow cytometry. First, UDC-NK cells retained the engineered genotype introduced at the iPSC stage, as evidenced by sustained expression of HLA-E and the absence of HLA class I and class II molecules (Figure 2D). The cells expressed natural cytotoxicity receptors, including NKp30, NKp44, and NKp46, as well as activating receptors such as DNAM-1, NKG2D, CD2, and 2B4.
In addition, expression of killer cell immunoglobulin-like receptors (KIRs) was observed, indicating that the UDC-NK cells were matured. Notably, the cells exhibited high levels of cytotoxic effector molecules, including TRAIL, perforin, and granzyme B, consistent with a cytotoxic NK cell phenotype. In contrast, minimal expression of inhibitory receptors such as PD-1 and TIGIT was observed, suggesting a low level of cellular exhaustion (Figure 2E).
Collectively, these phenotypic data indicate that the established feeder-free manufacturing process enables the robust generation of UDC-NK cells with a characteristic activated immune phenotype.
Comprehensive functional characterization of cryopreserved UDC-NK cells
NK cells possess an intrinsic ability to distinguish tumor cells from healthy cells through a balance of activating and inhibitory receptors, a mechanism known as natural cytotoxicity (8). We first evaluated the natural cytotoxic activity of UDC-NK cells in vitro. UDC-NK cells were thawed, washed, and immediately co-cultured with K562 cells, which are highly susceptible to NK cell–mediated lysis and are widely used as target cells, to assess natural cytotoxicity (37).
UDC-NK cells exhibited potent cytotoxicity against K562 cells in vitro (Figure 3A). In addition to natural cytotoxicity, UDC-NK cells were engineered to express high and uniform levels of FCGR3A and were therefore expected to exhibit antibody-dependent cellular cytotoxicity (ADCC) (38). To evaluate ADCC, UDC-NK cells were co-cultured with CD20+ B-cell lymphoma Ramos cells in the presence of Rituximab, a therapeutic monoclonal antibody targeting CD20.
UDC-NK cells demonstrated strong cytotoxicity against Ramos cells, which was further enhanced in the presence of Rituximab. (Figure 3B). Proinflammatory cytokine secretion, including tumor necrosis factor-α (TNF-α) and interferon-γ (IFN-γ), is a key functional property of NK cells that contributes to stimulating both innate and adaptive immune responses.
To assess cytokine production, UDC-NK cells were stimulated with K562 cells or CD20+ Raji B-cell lymphoma cells in the presence or absence of Rituximab for 24 hours. UDC-NK cells produced minimal amounts of TNF-α and IFN-γ in the absence of stimulation, whereas cytokine secretion was significantly increased upon stimulation with K562 cells or Rituximab-coated Raji cells (Figures 3C, D). Serial cytotoxicity is an important characteristic of in vivo antitumor efficacy.
To assess this property, we established a co-culture system of UDC-NK cells and CD20+ Ramos cells in the presence of Rituximab. Notably, UDC-NK cells exhibited persistent serial cytotoxicity against Ramos cells even in the absence of Rituximab and demonstrated more robust serial ADCC activity upon Rituximab treatment during repeated tumor cell challenge (Figure 3E). We also assessed the ADCC activity of UDC-NK cells against a panel of CD20+ non-Hodgkin B-cell lymphoma cell lines.
All tested cell lines uniformly expressed CD20 on their surface with variable density (Supplementary Figure 2). Consistent with previous observations, UDC-NK cells exhibited significant cytotoxicity across a broad range of B-cell lymphoma cell lines, indicating robust and reproducible antitumor activity. Importantly, this cytotoxicity was further enhanced in the presence of Rituximab, supporting the ability of UDC-NK cells to mediate ADCC (Supplementary Figure 3).
Collectively, these findings demonstrate that cryopreserved UDC-NK cells exhibit robust natural cytotoxicity, potent and sustained ADCC activity, and inducible proinflammatory cytokine secretion, supporting their potential as an off-the-shelf cell therapy platform for CD20+ B-cell malignancies.
Phenotypic and functional change of UDC-NK cells during serial activation/expansion
Next, to assess the potential of our unique NK cell activation and expansion process, which combines anti-NKp30 and anti-DNAM-1 antibodies with a cytokine cocktail, UDC-NK cells were subjected to up to seven consecutive activation/expansion cycles over 49 days, followed by phenotypic and functional characterization. Surprisingly, UDC-NK cells continued expansion during seven rounds of serial expansion, resulting in approximately 109-fold expansion (Supplementary Figure 4). Flow cytometric analysis revealed increased expression of DNAM-1, NKG2D, and CD25, whereas TIGIT expression was not detectable throughout the expansion period.
Notably, CD2 expression was detected in a minor population after the first round of expansion; however, the frequency of CD2+ UDC-NK cells increased over time (Supplementary Figure 5). We next assessed the cytotoxicity of UDC-NK cells against K562 cells and found that their activity was maintained over the expansion period (Supplementary Figure 6A). Interestingly, serially expanded UDC-NK cells showed a trend toward increased IFN-γ production upon stimulation, while no clear change was observed in TNF-α production (Supplementary Figure 6B, C).
These results indicate that UDC-NK cells retain their functional competence even after prolonged serial expansion.
Immune evasion of UDC-NK cells from allogeneic T cells and NK cells
UDC-NK cells were engineered by knocking out the B2M and RFXANK genes to eliminate surface expression of HLA class I and class II molecules, thereby preventing allogeneic T cell–mediated recognition. In addition, HLA-E was knocked in to enable UDC-NK cells to evade allogeneic NK cell–mediated cytotoxicity¹8. To evaluate the immune evasion of UDC-NK cells, we assessed their susceptibility to allogeneic NK cell–mediated cytotoxicity.
Notably, UDC-NK cells exhibited complete resistance to NK cell–mediated lysis, whereas B2M-knockout cells lacking HLA-E expression remained highly susceptible (Figure 4A). Importantly, UDC-NK cells did not induce proliferation of either CD4+ or CD8+ T cells, indicating immune evasion capacity of UDC-NK cells, whereas NK cells differentiated from WT iPSCs stimulated robust proliferation of allogeneic T cells (Figure 4B). Collectively, these results demonstrate that UDC-NK cells effectively evade both allogeneic T cell– and NK cell–mediated immune responses in vitro, supporting their potential to improve persistence and enhance antitumor efficacy in vivo.
Introduction of a Herpes Simplex Virus Thymidine Kinase (HSV-TK) safety switch
Adoptive cell therapies, such as autologous CAR-T cells, have demonstrated remarkable efficacy in hematological malignancies and are now widely used in clinical practice. However, the management of severe adverse events, including cytokine release syndrome and immune effector cell–associated neurotoxicity syndrome, remains a major clinical challenge (39). To mitigate potential safety risks associated with excessive immune activation, UDC-NK cells were engineered to express two copies of the herpes simplex virus-derived thymidine kinase (HSV-TK) genes as a safety switch.
This system can be pharmacologically activated by ganciclovir, a small-molecule antiviral agent commonly used to treat cytomegalovirus infections. Upon ganciclovir administration, HSV-TK-expressing cells are selectively eliminated, enabling controlled removal of UDC-NK cells in the event of adverse reactions (37, 38, 40, 41). To evaluate the functionality of this safety switch, UDC-NK cells and control iPSC-derived NK cells differentiated from a wild-type iPSC clone were cultured in the presence of increasing concentrations of ganciclovir, and cell viability was assessed after 4 days of treatment.
No significant cytotoxicity was observed in the absence of ganciclovir. UDC-NK cells expressing HSV-TK were efficiently eliminated in a dose-dependent manner upon ganciclovir treatment, whereas wild-type iPSC-NK cells showed no significant reduction in viability (Figure 5). Notably, UDC-NK cells were completely eliminated at clinically relevant ganciclovir concentrations (1-100 µM) (39).
These results demonstrate that the HSV-TK safety switch is functional in UDC-NK cells and provides a mechanism for selective cell ablation, thereby supporting the safety of the UDC-NK cell platform.
Efficacy of UDC-NK cells in combination with rituximab in B-cell lymphoma xenograft models
Following our in vitro characterization, we next investigated the in vivo persistence and safety profile of UDC-NK cells using NOG-IL-15 mice, an immunodeficient mouse strain transgenic for human IL-15 to support human NK cell persistence. We thawed cryopreserved UDC-NK cells and immediately administered 1.5x107 cells per mouse. In the flow cytometric analysis, UDC-NK cells were detected most frequently on day 3, and their frequency decreased by day 7 after administration (Supplementary Figure 7A).
The mice exhibited transient body weight loss but fully recovered by day 7, with no obvious clinical signs of deterioration (Supplementary Figure 7B).
Finally, to evaluate the antitumor activity of cryopreserved UDC-NK cells in vivo, we established disseminated lymphoma xenograft models in NOGF-IL-15 mice, an immunodeficient strain lacking surface Fc receptor expression and constitutively expressing human IL-15. The absence of murine Fc receptors eliminates ADCC and antibody-dependent cellular phagocytosis (ADCP) mediated by mouse effector cells, thereby enabling selective evaluation of ADCC mediated by UDC-NK cells in vivo. In addition, systemic expression of human IL-15 supports NK cell persistence (32, 42).
CD20+ B lymphoma Raji cells were intravenously injected into NOGF-IL-15 mice to establish disseminated lymphoma. One day after tumor inoculation, mice were treated with Rituximab alone or in combination with cryopreserved UDC-NK cells administered twice weekly for three weeks (Figure 6A). Rituximab monotherapy resulted in moderate inhibition of tumor progression compared with untreated controls.
In contrast, the combination of UDC-NK cells and Rituximab showed approximately 99% suppression of tumor progression (Figure 6B). Consistent with this marked antitumor activity, mice treated with the combination of UDC-NK cells and Rituximab exhibited a significant survival advantage compared with all other treatment groups (Figure 6C). Representative IVIS images demonstrated effective suppression of tumor growth in the UDC-NK plus Rituximab group, consistent with the reduced tumor burden and improved survival observed in treated mice.
(Figure 6D).
We also employed a disseminated B-cell lymphoma model using CD20+ Daudi cells to evaluate the in vivo efficacy of cryopreserved UDC-NK cells (Figure 6E). In this context, repeated administration of UDC-NK cells in combination with Rituximab resulted in robust antitumor activity, achieving 99% tumor growth inhibition (Figure 6F). Furthermore, the combination significantly improved overall survival in treated mice (Figure 6G).
Representative IVIS imaging further confirmed effective suppression of tumor progression in the UDC-NK plus rituximab group, consistent with the tumor burden and survival data (Figure 6H). Collectively, these results demonstrate that UDC-NK cells exert potent antitumor activity in combination with Rituximab through the ADCC mechanism in vivo. These findings support the potential of UDC-NK cells as an effective off-the-shelf cellular immunotherapy in combination with therapeutic antibodies for the treatment of hematological malignancies (43).
Discussion
Allogeneic NK cell therapies are emerging as a promising alternative to T cell-based immunotherapies due to their intrinsic ability to recognize and eliminate malignant cells without prior antigen sensitization, as well as their favorable safety profile (8, 44, 45). In this study, we developed a universal donor iPSC-derived NK cell (UDC-NK) cell platform that integrates multiple genetic modifications to enhance both therapeutic efficacy and safety, establishing a novel approach to generating immune-evasive NK cells with a favorable safety profile. Using a novel feeder-free and scalable manufacturing process, we generated highly functional NK cells exhibiting canonical NK cell phenotypes, potent natural cytotoxicity, and robust antibody-dependent cellular cytotoxicity (ADCC).
Notably, UDC-NK cells demonstrated strong antitumor activity in xenograft lymphoma models, particularly in combination with Rituximab, leading to near-complete suppression of tumor progression and significantly prolonged survival.
A notable aspect of this study is the development of a feeder-free manufacturing process that enables robust expansion of genetically engineered iPSC-derived NK cells while maintaining NK-cell purity and functionality. Many previously reported iPSC-NK manufacturing approaches have relied on feeder-cell-based expansion systems, which may add complexity to large-scale clinical manufacturing. In contrast, our novel feeder-free activation and expansion process utilizing agonists of NK-cell activating receptors together with a cytokine cocktail supported extensive expansion and generated large numbers of UDC-NK cells while preserving NK-cell phenotype and cytotoxic function.
These findings suggest that this platform may facilitate scalable and standardized manufacturing of off-the-shelf NK-cell products. Furthermore, the sequential emergence of CD34+CD43+/CD34+CD45+ hematopoietic progenitors and CD45+CD56+ NK cells may provide useful in-process checkpoints for monitoring differentiation progression and manufacturing consistency. Future studies should evaluate whether these markers can serve as predictive quality attributes for clinical-scale production.
A central component of this platform is the use of a multi-gene-edited universal donor iPSC line engineered to evade host immune rejection (19, 24, 44). Disruption of B2M and RFXANK eliminates surface expression of HLA class I and II molecules, thereby preventing recognition by allogeneic T cells, while the introduction of a single-chain HLA-E construct provides protection against NK cell–mediated cytotoxicity (19). Consistent with this design, UDC-NK cells did not activate allogeneic CD4+ or CD8+ T cells and were resistant to cytotoxicity mediated by allogeneic PBMC-derived NK cells in vitro.
Consequently, UDC-NK cells represent a novel class of engineered cellular therapeutics that integrate immune evasion strategies with a built-in safety switch system. Recently, novel UDC technologies utilizing CD47 overexpression and Autoimmune Defense Receptors (ADRs) targeting 4-1BB have been reported to further enhance immune evasion (24, 25). Accordingly, the UDC platform described here could be further enhanced by combining HLA editing with these additional gene-engineering strategies.
Although the immune-evasive properties of UDC-NK cells have so far been evaluated only in vitro, further validation in robust in vivo models will be necessary to confirm their translational potential. UDC-NK cells were engineered to express herpes simplex virus thymidine kinase (HSV-TK) safety switch (39). The safety switch is a well-characterized and clinically validated system that enables the selective pharmacological elimination of engineered cells upon ganciclovir administration (37, 43).
Activation of this safety switch by ganciclovir resulted in efficient elimination of UDC-NK cells in vitro, providing an additional layer of control and may help manage potential adverse events associated with cell therapy. Importantly, the functionality of this safety switch has only been demonstrated in vitro. Future studies will be required to assess the kinetics, completeness, and durability of UDC-NK cell elimination following ganciclovir administration under clinically relevant in vivo conditions.
Such studies will be important to further establish the translational utility of the HSV-TK safety switch for the management of potential cell therapy-related toxicities.
Another important feature of UDC-NK cells is the uniform expression of the high-affinity FCGR3A (F158V) variant, which enhances ADCC (21). NK cell-mediated ADCC plays a critical role in the therapeutic activity of many clinically approved monoclonal antibodies, including Rituximab (43). Consistent with this mechanism, UDC-NK cells demonstrated robust ADCC against multiple CD20-positive lymphoma cell lines in the presence of Rituximab, while retaining natural cytotoxic activity in the absence of antibody stimulation.
The ability to mediate both natural cytotoxicity and antibody-dependent killing may provide complementary mechanisms of tumor control and enhance the therapeutic potential of this engineered NK cell platform. Importantly, the enhanced cytotoxic activity observed in vitro translated into substantial therapeutic benefits in vivo. In Fc receptor-deficient NOGF-IL-15 mice, which enable selective evaluation of ADCC mediated by administered NK cells, the combination of UDC-NK cells and Rituximab resulted in near-complete suppression of tumor progression and significantly improved survival compared with Rituximab monotherapy in Raji and Daudi B cell lymphoma disseminated mouse models.
These findings support the concept that engineering NK cells to express high-affinity Fc receptors can enhance the efficacy of antibody-based cancer therapies (43). Such combination approaches may represent a promising strategy to improve the clinical activity of monoclonal antibodies in hematological malignancies. However, a limitation of the present study is that ADCC activity was not directly compared with that of CD16-unmodified iPSC-derived NK cells.
Therefore, although robust ADCC activity was demonstrated, the extent to which FCGR3A (F158V) engineering contributed to the enhanced ADCC phenotype could not be precisely determined. Furthermore, the relative contributions of natural cytotoxicity and ADCC to the observed antitumor efficacy remain unclear because UDC-NK monotherapy was not systematically evaluated in vivo. Future studies incorporating appropriate comparator NK-cell products and UDC-NK monotherapy groups will be necessary to address these questions.
More broadly, the present study did not include direct side-by-side comparisons between UDC-NK cells and primary NK cells or unedited iPSC-derived NK cells under identical manufacturing and assay conditions. Future comparative studies performed under standardized manufacturing and assay conditions will be important to further define the relative advantages of the UDC-NK platform.
In addition, further optimization of this platform may enhance its therapeutic potential. For example, combining UDC-NK cells with additional antibody-based therapeutics, such as PD-1 immune checkpoint inhibitors, may further improve antitumor responses. Moreover, the gene-editing strategy employed in this study could be extended to incorporate additional modifications to enhance NK cell cytotoxicity, persistence, trafficking, and resistance to immunosuppressive signals within the tumor microenvironment (12, 22, 23, 42, 46).
Such next-generation engineering approaches may include the introduction of chimeric antigen receptors (CARs), expression of membrane-bound cytokines such as IL-15 to promote persistence (17, 22, 23), or disruption of inhibitory pathways such as CISH to augment NK cell activation (47).
In conclusion, we developed a universal donor iPSC-derived NK cell platform that integrates immune evasion, enhanced ADCC capability, and an inducible safety switch within a feeder-free, scalable manufacturing framework. UDC-NK cells manufactured from multiple gene-edited UDC-iPSCs exhibit potent antitumor activity both in vitro and in vivo, particularly in combination with antibody-based therapies such as Rituximab. These findings support the potential of UDC-NK cells as a next-generation, off-the-shelf NK cell therapy for hematologic malignancies and potentially solid tumors.