iPSC-derived hypoimmunogenic tissue resident memory T cells mediate robust anti-tumor activity against cervical cancer

Abstract

Functionally rejuvenated human papilloma virus-specific cytotoxic T lymphocytes (HPV-rejTs) generated from induced pluripotent stem cells robustly suppress cervical cancer. However, autologous rejT generation is time consuming, leading to difficulty in treating patients with advanced cancer. Although use of allogeneic HPV-rejTs can obviate this, the major obstacle is rejection by the patient immune system. To overcome this, we develop HLA-A24&-E dual integrated HPV-rejTs after erasing HLA class I antigens. These rejTs effectively suppress recipient immune rejection while maintaining more robust cytotoxicity than original cytotoxic T lymphocytes. Single-cell RNA sequencing performed to gain deeper insights reveal that HPV-rejTs are highly enriched with tissue resident memory T cells, which enhance cytotoxicity against cervical cancer through TGFβR signaling, with increased CD103 expression. Genes associated with the immunological synapse also are upregulated, suggesting that these features promote stronger activation of T cell receptor (TCR) and increased TCR-mediated target cell death. We believe that our work will contribute to feasible "off-the-shelf" T cell therapy with robust anti-cervical cancer effects.

Keywords: CRISPR-Cas9 scarless gene editing; HLA class I-edited iPSCs; cervical cancer; human papilloma virus-specific CTLs; hypoimmunogenic iPSC-derived CTLs; rejuvenated CTLs; single-cell RNA sequencing; tissue resident memory T cells; two-step gene editing; “off-the-shelf” CTL therapy.

Graphical abstract

Figure 1

Reduced T cell activity against HLA-edited rejTs in vitro (A) Schematic representation of donor DNAs and guide RNA (gRNA) for CRISPR-Cas9 “scarless” first- and second-step editing. Black bar, stop codon of B2M; red arrow, gRNA target site; black arrows, primers for genotyping. (B) Biallelic iPSC clone with bright EGFP expression identified by fluorescent imaging. Scale bar, 100 μm. (C) Marker integration confirmed using genotyping PCR; product 5.0 kb long. (D) HLA-A24 and HLA-E marker integration confirmed using genotyping PCR products 8.8 kb and 4.4 kb long, respectively. (E) Flow-cytometric analysis of HLA class I-edited iPSCs cocultured with IFN-γ. HLA expressions are shown in red, while isotype control tracings are shown in blue. The plots represent three independent experiments. (F) Schematic illustration of generation of HLA-edited HPV16 E6-rejTs. (G) Flow-cytometric E6_49–57 tetramer analysis of CD3⁺ gated HLA-edited rejTs. The plots represent three independent experiments. (H) Flow-cytometric analysis of HLA class I antigen expression on HLA class I-edited rejTs. The plots represent three independent experiments. (I–K) Flow-cytometry plots of proliferating T cells in carboxyfluorescein succinimidyl ester (CFSE)-diluted allogeneic CD8⁺/CD4⁺ T cell populations in CSFE-labeled allogeneic CD3⁺ T cells from HLA-A24⁺ donors cocultured with WT- and HLA-edited rejTs. The plots represent two independent experiments from each donor.

Figure 2

HLA-edited rejTs suppressed NK cell activity in vitro and in vivo (A) ⁵¹Cr release assay results of allogeneic NK cells from HLA-A24⁺ healthy donors (effector, E) and WT- and HLA-edited-rejTs (target, T). E/T ratios were 9:1, 3:1, 1:1, and 0.33:1. Positive controls were post-first-edit rejTs or K562. Negative controls were autologous T cells. Representative data of at least two independent experiments (n = 3). Error bars represent ±SD. ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, one-way ANOVA; ns, not significant. (B) NK cell activation measured by flow cytometry of CD107a expression for each donor. Positive controls were NK cells treated with stimulants. Negative controls were NK cells without target cells. Data represent means of at least two independent experiments (n = 3). Error bars represent ±SD. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, one-way ANOVA; ns, not significant. (C) Flow-cytometric analysis of NKG2A and KIR3DLI expression on NK cell subset (CD3⁻, CD56⁺) in PBMCs of healthy donors. Data represent means of three independent experiments. Error bars represent ±SD. (D) Representative flow cytometrograms of NK cells expressing NKG2A/KIR3DL1 for each donor (left) and total percentages of NK cells expressing NKG2A/KIR3DL1 for each donor (right). (E) Bioluminescence images of KI-A24&E-rejTs injected into mice. Mice were divided into three groups depending on NK cells injected: HLA-A24⁺ NK cells (n = 3), HLA-A24⁻ NK cells (n = 3), or no NK cells (n = 3) (control). Images of all mice from each independent experiment are shown. Data represent at least two independent experiments. (F) Bioluminescence signals of KI-A24&E-rejTs on day 7 for the three groups. Data from two independent experiments in HLA-A24⁺ NK cells (n = 4), HLA-A24⁻ NK cells (n = 6), and no NK cells (n = 4) are shown. ∗p < 0.05, ∗∗p < 0.01, one-way ANOVA; ns, not significant. (G) Kaplan-Meier survival curves of experimental groups. Data from two independent experiments in HLA-A24⁺ NK cells (n = 4), HLA-A24⁻ NK cells (n = 5), no NK cells (n = 4), and no treatment (n = 3) are shown. ∗p < 0.05 by log-rank testing; ns, not significant.

Figure 3

Tumor-suppressive effect of intraperitoneally transferred KI-A24&E-rejTs in vivo (A) Real-time cell analyzer (RTCA) continuous graphical output of tumor proliferation index up to 70 h for SKGIIIa target cells (T) cocultured with WT-rejTs, KI-A24-rejTs, KI-E-rejTs, or KI-A24&E-rejTs effector cells (E). E/T ratios were 1:1. HLA-mismatched different epitope rejTs were used as control. The data shown represent three independent triplicate experiments. Mean values are plotted ±SD. (B) Similarly, RTCA continuous graphical output of the tumor proliferation index up to 70 h for SiHa cells (T) cocultured with WT-rejTs or KI-A24&E-rejTs (E). E/T ratios were 1:1. HLA-mismatched different epitope rejTs were used as control. The data shown represent three independent triplicate experiments. Mean values are plotted ±SD. (C) Bioluminescence imaging of mice engrafted with SiHa cells in four treatment groups: no-treatment group (control group, n = 5) and the three treatment groups of original CTL clone (n = 5), WT-rejTs (n = 4), and KI-A24&E-rejTs (n = 4). HPV16 E6-specific T cells (2.5 × 10⁶ cells) were injected intraperitoneally once a week (3 doses). Images of all mice from an independent experiment are shown. Data represent at least two independent experiments. (D) Quantified tumor burden on day 21. ∗∗p < 0.01 by one-way ANOVA; ns, not significant. (E) Kaplan-Meier curves representing percentage survival in experimental groups of no treatment (n = 5), original CTL clone (n = 5), WT-rejTs (n = 4), and KI-A24&E-rejTs (n = 4). ∗∗p < 0.01 by log-rank test; ns, not significant. (F) Photomicrographs of cervical cancer (1) as a control and of uterus (2), small intestine (3), lung (4), spleen (5), and liver (6) from mouse treated with KI-A24&E-rejTs (hematoxylin and eosin stain).

Figure 4

Phenotype characterization of KI-A24&E-rejTs and of original CTL clone (A) Histograms of fluorescence intensity of expression of granzyme B (left) and perforin (right) in HPV-CTL clone and HPV-rejTs ± SD (n > 3). Negative controls (fluorescence-minus-one control) are shown in gray. T cells were stained and analyzed approximately 10 days after stimulation. Error bars represent ±SD. ∗∗p < 0.01, unpaired Student’s t test (two-tailed). (B) Representative flow cytometrograms of memory-phenotype cells (stem cell memory: CD45RA⁺, CD62L⁺, CD95⁺, CD27⁺, and CD28⁺) in original CTL clone and HPV-rejTs. The plots represent three independent experiments. Mean percentages of each memory phenotype in HPV-CTL clone and HPV-rejTs ±SD are shown. ∗∗p < 0.01, unpaired Student’s t test (two-tailed). (C and D) ⁵¹Cr release assay with KI-A24&E-rejTs and original CTL clone (effectors, E) and SiHa (left) or SKGIIIa (right) (targets, T) with or without peptide pulse were examined at E/T ratios of 20:1, 10:1, 5:1, and 2.5:1 (SiHa) and 10:1, 5:1, and 2.5:1 (SKGIIIa). Representative data from at least three independent experiments (n = 3) are shown. (E) Real-time PCR of HPV16⁺ cervical cancer cells from patient samples 1–5 and cervical cell lines (Caski, SiHa, and SKGIIIa). Caski cells were used as a positive control, while HPV16-negative patient samples were used as negative controls for HPV16 E6 expression. Data represent three independent triplicate experiments.

Figure 5

scRNA-seq analysis of original HPV-CTL clone and KI-A24&E-rejT (A) UMAP of scRNA-seq data showing original HPV-CTL clone in orange and KIA24&E-rejTs in blue (left). Cell colors reflect graph-based cluster assignments (right). Cluster gene characteristics determine cluster functional description. Cluster 0: CD8A^high KLRK1^high (IEL)/SELL^low CCR7^low(T_EM) No. 1; cluster 1: CD8A^high KLRK1^high (IEL)/ITGAE^high CD69^high TGFBR^high (T_RM) No. 1; cluster 2: CD8A^high KLRK1^high (IEL)/ITGAE^high CD69^high TGFBR^high (T_RM) No. 2; cluster 3: CD8A^high KLRK1^high (IEL)/ITGAE^high CD69^high TGFBR^high (T_RM)/SELL^high CCR7^high (T_CM); cluster 4: ITGAE^high CD69^high TGFBR^high (T_RM) No. 1; cluster 5: ITGAE^high CD69^high TGFBR^high (T_RM) No. 2; cluster 6: ITGAE^high CD69^high TGFBR^high (T_RM)/SELL^low CCR7^low(T_EM); cluster 7: CD8A^high KLRK1^high (IEL)/SELL^low CCR7^low (T_EM)/SELL^high CCR7^high(T_CM); cluster 8: CD27^high. (B–D) Cell colors reflect single-cell genotype group: (B) cytotoxicity (IFNG, PRF1, FASLG), (C) immunological synapse (PTPRC, ITGAM, PECAM1), and (D) intraepithelial lymphocytes (KLRK1, CD244, SLAMF7). (E) Tissue resident memory T cells (ITGAE, CD69, ITGA1). (F) Heatmap of genotype groups composed of unique signature genes for original CTL clone and KI-A24&ErejTs. (G) Representative flow cytometrograms of T_RM phenotype cells (CD3, CD8, CD62L, CCR7, CD69, and CD103 populations) in original CTL clone and KI-A24&E-rejTs. The plots represent three independent experiments. The mean percentages of each memory phenotype in original CTL clone and KI-A24&E-rejTs ±SD are shown. ∗∗∗∗p < 0.0001, unpaired Student’s t test (two-tailed). (H) Representative flow cytometrograms of T_RM phenotype cells (CD3, CD8, CD62L, CD69, CD103, CD44, and CD49a populations) in KI-A24&E-rejTs. The plots represent three independent experiments. (I) Histograms of E-cadherin expression in SiHa, SKGIIIa, and Caski cells for three independent experiments.

Figure 6

Cytotoxicity of T_RM and T_EM in original HPV-CTL clone (A) Representative flow cytometrograms of T_RM and T_EM phenotype cells (CD3, CD8, CD62L, CCR7, CD69, and CD103 populations) in original CTL clone. T_RM and T_EM phenotype cells were sorted for ⁵¹Cr release assay. The plots represent three independent experiments. (B) ⁵¹Cr release assay with T_RM (left) and T_EM (right) in original CTL clone (effectors, E) and SKGIIIa (targets, T) with or without peptide pulse were examined at E/T ratios of 20:1, 10:1, 5:1, and 2.5:1. Representative data for at least three independent experiments (n = 3) are shown. (C) RTCA continuous graphical output of the tumor proliferation index up to 70 h for SiHa cells (T) cocultured with KI-A24&E-rejTs (E) with/without anti-CD103 and/or anti-E-cadherin antibodies and/or TGFβR inhibitor. E/T ratios were 1:1. The data shown represent three independent triplicate experiments. Mean values are plotted ±SD. ∗∗p < 0.01, one-way ANOVA (n = 3). (D) Representative flow cytometrograms of CD103 expression in KI-A24&E-rejTs after coculturing with SiHa cells for 70 h in each condition of RTCA. The plots represent two independent experiments. (E) Bar graph representing CD103 expression on HPV-rejTs cocultured without SiHa cells, with SiHa cells, with SiHa cells + TGFβR inhibitor, with SiHa cells + all antibodies + TGFβR inhibitor. Error bars represent ±SD. ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, one-way ANOVA (n = 3). Ab, antibody. The data represent two independent triplicate experiments.