Targeting triple-negative breast cancer using cord-blood CD34⁺ HSPC-derived mesothelin-specific CAR-NKT cells with potent antitumor activity

Abstract

Background: Triple-negative breast cancer (TNBC) is an aggressive subtype of breast cancer characterized by the lack of ER, PR, and HER2 expression. Its aggressive behavior, high degree of tumor heterogeneity, and immunosuppressive tumor microenvironment (TME) are associated with poor clinical outcomes, rapid disease progression, and limited therapeutic options. Although chimeric antigen receptor (CAR)-engineered T cell therapy has shown certain promise, its applicability in TNBC is hindered by antigen escape, TME-mediated suppression, and the logistical constraints of autologous cell production.

Methods: In this study, we employed hematopoietic stem and progenitor cell (HSPC) gene engineering and a feeder-free HSPC differentiation culture to generate allogeneic IL-15-enhanced, mesothelin-specific CAR-engineered invariant natural killer T (^Allo15MCAR-NKT) cells.

Results: These cells demonstrated robust and multifaceted antitumor activity against TNBC, mediated by CAR- and NK receptor-dependent cytotoxicity, as well as selective targeting of CD1d⁺ TME immunosuppressive cells through their TCR. In both orthotopic and metastatic TNBC xenograft models, ^Allo15MCAR-NKT cells demonstrated potent antitumor activity, associated with robust effector and cytotoxic phenotypes, low exhaustion, and a favorable safety profile without inducing graft-versus-host disease.

Conclusions: Together, these results support ^Allo15MCAR-NKT cells as a next-generation, off-the-shelf immunotherapy with strong therapeutic potential for TNBC, particularly in the context of metastasis, immune evasion, and treatment resistance.

Keywords: Allogeneic CAR-NKT cells; Allogeneic cell therapy; Allorejection; CRISPR-Cas9 gene editing; HLA ablation; Mesothelin-targeting CAR (MCAR); Metastatic model; Multiple tumor targeting mechanism; Off-the-shelf; Orthotopic model; Potent antitumor activity; Triple-negative breast cancer (TNBC); Tumor microenvironment (TME); Tumor-associated macrophage (TAM).

Fig. 1

Primary metastatic TNBC patient sample profiling highlights the therapeutic potential of CAR-NKT cells. (A) Diagram showing the primary metastatic TNBC patient sample collection, the TME composition, and the tumor and TME targeting mechanisms of CAR-NKT cells. TNBC, triple negative breast cancer; TME, tumor microenvironment; TAM, tumor associated macrophages; MDSC, myeloid-derived suppressor cell; NKR, NK receptor; Ag, antigen. (B) FACS analyses of CAR targets (e.g., MSLN, TROP2, EGFR, and Nectin-4) and NKR targets (e.g., CD112, CD155, MICA/B, and ULBP-1) expression on tumor cells from TNBC patient samples. MSLN, mesothelin. (C) Quantification of (B) (n = 5). (D) Clinical data correlation studies. Kaplan-Meier plots are presented, showing the association between MSLN expression in tumor and survival of cancer patients in a breast cancer cohort (Prediction of Clinical Outcomes from Genomic Profiles [PRECOG]: GSE20486, n = 97). (E) Table showing the CAR antigen expressions on TNBC tumors and normal tissues. (F) Immune cell composition of 5 TNBC patients. (G) FACS analyses of CD1d expression on the indicated immune cells. (H) Quantification of (G) (n = 5). Representative of 5 experiments. Data are presented as the mean ± SEM. ns, not significant, ****p < 0.0001, by one-way ANOVA (H)

Fig. 2

Allogeneic stem cell-derived MCAR-NKT cells are generated using a clinically guided culture method and display prominent NK-like features and potent cytotoxic activity. (A) Schematics showing the generation of ^Allo15MCAR-NKT cells. CB, cord blood; HSPC, hematopoietic stem and progenitor cells; Lenti/iNKT-MCAR-IL-15, lentiviral vector encoding a pair of iNKT TCR α and β chains, a mesothelin-directed CAR, and a human soluble IL-15. (B) Schematics showing the design of Lenti/iNKT-MCAR-IL-15 lentivector. ΔLTR, self-inactivating long terminal repeats; MNDU3, internal promoter derived from the MND retroviral LTR U3 region; φ, packaging sequence; RRE, rev-responsive element; cPPT, central polypurine tract; WPRE, woodchuck hepatitis virus posttranscriptional regulatory element; F2A, foot-and-mouth disease virus 2 A; P2A, porcine teschovirus-1 2 A; T2A, thosea asigna virus 2 A. (C) FACS analyses of Lenti/iNKT-MCAR-IL-15 transduction rate on CD34⁺ HSPCs at 72 h after lentivector transduction (n = 3). 7 different cord blood donors were tested. (D) FACS monitoring of the generation of ^Allo15MCAR-NKT cells during the 6-week culture. iNKT TCR was stained using a 6B11 monoclonal antibody. (E) FACS detection of CAR expression on ^Allo15MCAR-NKT cells. (F) Quantification of ^Allo15MCAR-NKT cell purity (identified as 6B11⁺ and CD3⁺) (n = 3). 7 different cord blood donors were tested. (G) Quantification of CAR⁺ proportion of ^Allo15MCAR-NKT cell (n = 3). 7 different cord blood donors were tested. (H) Yield of ^Allo15MCAR-NKT cells (n = 3). 7 different cord blood donors were tested. (I) VCN measurement of ^Allo15MCAR-NKT cells (n = 7). VCN, vector copy number. (J) FACS detection of NK marker and NK receptor (NKR) expression, as well as intracellular cytokine and cytotoxic molecule production of ^Allo15MCAR-NKT and conventional MCAR-T cells. (K) Violin plots showing the expression distribution of the indicated gene signatures in ^Allo15MCAR-NKT and conventional MCAR-T cells. (L) Pathway analyses of differentiated expressed genes comparing ^Allo15MCAR-NKT with conventional MCAR-T cells. GO, Gene ontology ID. Representative of 3 (A-I) and 1 (J-L) experiments. Data are presented as the mean ± SEM

Fig. 3

Allogeneic MCAR-NKT cells kill TNBC tumor cells with superior antitumor efficacy and multifaceted killing mechanisms. (A-D) Studying the in vitro antitumor efficacy of ^Allo15MCAR-NKT cells against human MM and TNBC cell lines. MCAR-T cells and non-MCAR-engineered PBMC-derived T cells were included as therapeutic cell controls. (A) Schematics showing the indicated human MM and TNBC cell lines. MM.1 S-FG, MM.1 S cell engineered to overexpress FG; MM.1 S-MSLN-FG, MM.1 S-FG cell engineered to overexpress MSLN; HCC1806-FG, HCC1806 cell engineered to overexpress FG; MDA-MB-231-FG, MDA-MB-231 cell engineered to overexpress FG. (B) FACS detection of MSLN expression on the indicated tumor cells. (C) Experimental design. (D) Tumor cell killing data at 24 h (n = 4). (E-G) Studying the expression of effector molecules of ^Allo15MCAR-NKT cells. (E) FACS detection of surface CD69 as well as intracellular Perforin and Granzyme B in the therapeutic cells. (F) Quantification of (E) (n = 4). (G) ELISA analyses of IFN-γ production in the therapeutic cells (n = 4). (H-I) Studying the tumor cell killing mechanisms of ^Allo15MCAR-NKT cells mediated by NKRs (i.e., NKG2D and DNAM-1). (H) Experimental design. (I) Tumor cell killing data at 24 h (E: T ratio = 0.5:1; n = 4). (J) Studying the long-term tumor cell killing of ^Allo15MCAR-NKT cells after repeated tumor challenge (n = 4). HCC1806-FG tumor cells were added every 48 h for 5 times. (K) Radar plots showing the expression of immune checkpoint molecules on ^Allo15MCAR-NKT cells. Data represent percent-positive cells as measured by flow cytometry. (L-N) Studying the tumor antigen escape targeting of ^Allo15MCAR-NKT cells. (L) Experimental design. (M) FACS analyses of remaining live MSLN⁺ and MSLN⁻ tumor cells at 24 h. (N) Quantification of remaining live MDA-MB-231-MSLN-FG and MDA-MB-231-FG cells at 24 h. The percentages of MDA-MB-231-MSLN-FG and MDA-MB-231-FG cells were recorded, and the fold changes were calculated by normalizing to the NT group. (O-P) Studying the TNBC tumor cell killing capacity of ^Allo15MCAR-NKT cells against primary TNBC patient samples. (O) Experimental design. (P) Tumor cell killing data at 24 h (n = 4). Representative of 3 experiments. Data are presented as the mean ± SEM. ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by two-way ANOVA (D, J, and P), or by one-way ANOVA (F, G, and I)

Fig. 4

Allogeneic MCAR-NKT cells target the TNBC TME by killing the CD1d⁺ immunosuppressive cells. (A-B) Studying primary metastatic TNBC-derived TAM/MDSC targeting by ^Allo15MCAR-NKT cells. (A) Experimental design. (B) FACS analyses of live immune cells targeted by ^Allo15MCAR-NKT cells at 24 h (n = 3). The percentage of the indicated immune cells among total CD45⁺CAR⁻ immune cells was recorded, and the fold change was calculated by normalizing to the NT group. (C) Studying CD1d-mediated TAM/MDSC targeting by ^Allo15MCAR-NKT cells at 24 h (n = 3). Anti-CD1d antibody was added into the co-culture to block CD1d/NKT TCR recognition. (D-F) In vitro generation and polarization of human monocyte-derived M2 macrophages. (D) Experimental design. M-CSF, macrophage colony-stimulating factor; MDM, monocyte-derived macrophage; Mφ, macrophage. (E) FACS detection of CD1d on M2-polarized macrophages. Healthy donor PBMC-derived T and B cells were included as staining controls. (F) Quantification of (E) (n = 5). (G-H) Studying M2-polarized macrophage targeting by ^Allo15MCAR-NKT cells. (G) Experimental design. (H) FACS analyses of live macrophages at 24 h after co-culturing with ^Allo15MCAR-NKT cells (n = 4). Live cells were identified as e506⁻CD14⁺CD11b⁺ cells. An anti-CD1d antibody was added into the coculture to block CD1d/iNKT TCR recognition. Healthy donor PBMC-derived T and B cells were included as target cell controls. (I-L) Studying TAM targeting by ^Allo15MCAR-NKT cells using an in vitro 3D tumor organoid culture. (I) Experimental design. (J) Tumor killing data at 48 h (n = 4). (K) TAM killing data at 48 h (n = 4). (L) FACS analyses of surface markers (i.e., CD25) of ^Allo15MCAR-NKT cells. (M) Diagram showing the tumor-targeting and TAM-targeting mechanisms of ^Allo15MCAR-NKT cells. Representative of 3 experiments. Data are presented as the mean ± SEM. ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by Student’s t test (L left), or by one-way ANOVA (B, C, H, J, K, and L right)

Fig. 5

Allogeneic MCAR-NKT cells demonstrate superior antitumor efficacy and safety in a human TNBC orthotopic xenograft model. (A-E) Studying the in vivo antitumor efficacy of ^Allo15MCAR-NKT cells in an orthotopic HCC1806-FG human xenograft NSG mouse model. (A) Experimental design. BLI, live animal bioluminescence imaging. (B) BLI images measuring tumor loads in experimental mice over time. (C) Quantification of (B) (n = 4–5). (D) Tumor size measurements in experimental mice over time (n = 4–5). (E) Kaplan–Meier survival curves (n = 4–5). (F-I) Studying the in vivo safety of ^Allo15MCAR-NKT cells. (F) FACS analyses of CD3⁺CD45⁺ double positive T cells in mouse blood post injection of therapeutic cells. (G) Quantification of (F) (n = 4–5). (H) Clinical GvHD scores recorded on day 40 and 50 (n = 4–5). (I) Measurement of organ damage markers (Urea, ALT, AST, Bilirubin, and GLDH) in mouse serum. ALT, Alanine transaminase; AST, Aspartate Transaminase; GLDH, Glutamate Dehydrogenase. Representative of 3 experiments. Data are presented as the mean ± SEM. ns, not significant, **p < 0.01, ***p < 0.001, ****p < 0.0001 by Student’s t test (C right, and D right), by one-way ANOVA (C left, D left, G, and H) or by log rank (Mantel-Cox) test adjusted for multiple comparisons (E)

Fig. 6

Allogeneic MCAR-NKT cells demonstrate superior antitumor efficacy with strong effector function and low exhaustion in a human metastatic TNBC xenograft model. (A-G) Studying the in vivo antitumor efficacy of ^Allo15MCAR-NKT cells in a metastatic HCC1806-FG human xenograft NSG mouse model. (A) Experimental design. (B) BLI images measuring tumor loads in experimental mice over time. (C) Quantification of (B) (n = 4–5). (D) Kaplan–Meier survival curves (n = 4–5). (E) BLI images showing the biodistribution of HCC1806-FG cells in a representative experimental mouse on day 39. (F) Quantification of tumor loads in the indicated organs (n = 4–5). (G) BLI images showing the biodistribution of HCC1806-FG tumor cells in the indicated organs. (H-I) Studying the in vivo distribution of ^Allo15MCAR-NKT cells on day 30. (H) FACS analyses of CD3⁺CD45⁺ double positive human T cells in different organs post injection of the therapeutic cells. (I) Quantification of (H) (n = 5). (J-L) Studying the in vivo profile of ^Allo15MCAR-NKT cells. (J) FACS analyses of activation and effector markers of the therapeutic cells collected from the lungs. (K-L) Radar plots showing the expression of activation and effector markers (K) and immune checkpoint molecules (L) on the therapeutic cells. Data represent percent-positive cells as measured by flow cytometry. Representative of 3 experiments. Data are presented as the mean ± SEM. **p < 0.01, ****p < 0.0001 by two-way ANOVA (C), or by log rank (Mantel-Cox) test adjusted for multiple comparisons (D)

Fig. 7

HLA-I/II-ablated universal MCAR-NKT cells resist host T cell-mediated allorejection while preserving potent antitumor efficacy. (A) Schematics showing the generation of HLA-ablated universal IL-15-enhanced MCAR-NKT (^U15MCAR-NKT) cells. (B) FACS analyses of NKT TCR and HLA-I/II expression in CD34⁺ HSPCs at 72 h after lentivector transduction. (C) FACS monitoring of the generation of ^U15MCAR-NKT cells during the 6-week culture. (D) FACS analyses of CAR expression on ^U15MCAR-NKT cells. (E) FACS analyses of HLA-I/II expression on ^U15MCAR-NKT cells pre- and post-purification. (F) Quantification of HLA-I/II double-negative proportion of ^U15MCAR-NKT cells (n = 5). (G) Quantification of CAR⁺ proportion of ^U15MCAR-NKT cells (n = 5). (H) Yield of ^U15MCAR-NKT cells (n = 5). (I) Quantification of HLA-I and HLA-II expression of ^U15MCAR-NKT cells (n = 4). ^Allo15MCAR-NKT and conventional MCAR-T cells were included as controls. (J-K) Studying the T cell-mediated allorejection of ^U15MCAR-NKT cells. (J) Experimental design. (K) ELISA analyses of IFN-γ production on day 4 (n = 4). (L) Illustration depicting the hypoimmunogenecity working model of ^Allo/U15MCAR-NKT cells. (M-N) Studying the in vitro antitumor efficacy of ^U15MCAR-NKT cells against human MM and TNBC cell lines. MCAR-T cells and non-MCAR-engineered T cells were included as therapeutic cell controls. (M) Experimental design. (N) Tumor cell killing data at 24 h (n = 4). (O-R) Studying the in vivo antitumor efficacy of ^U15MCAR-NKT cells in a metastatic HCC1806-FG human xenograft NSG mouse model. (O) Experimental design. (P) BLI images measuring tumor loads in experimental mice over time. (Q) Quantification of (P) (n = 5). (R) Kaplan–Meier survival curves (n = 5). Representative of 3 experiments. Data are presented as the mean ± SEM. ns, not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 by one-way ANOVA (I, K, and Q), two-way ANOVA (N), or log rank (Mantel-Cox) test adjusted for multiple comparisons (R)