Expanding the cytokine receptor alphabet reprograms T cells into diverse states

Abstract

T cells respond to cytokines through receptor dimers that have been selected over the course of evolution to activate canonical JAK-STAT signalling and gene expression programs¹. However, the potential combinatorial diversity of JAK-STAT receptor pairings can be expanded by exploring the untapped biology of alternative non-natural pairings. Here we exploited the common γ chain (γ_c) receptor as a shared signalling hub on T cells and enforced the expression of both natural and non-natural heterodimeric JAK-STAT receptor pairings using an orthogonal cytokine receptor platform^2-4 to expand the γ_c signalling code. We tested receptors from γ_c cytokines as well as interferon, IL-10 and homodimeric receptor families that do not normally pair with γ_c or are not naturally expressed on T cells. These receptors simulated their natural counterparts but also induced contextually unique transcriptional programs. This led to distinct T cell fates in tumours, including myeloid-like T cells with phagocytic capacity driven by orthogonal GSCFR (oGCSFR), and type 2 cytotoxic T (T_C2) and helper T (T_H2) cell differentiation driven by orthogonal IL-4R (o4R). T cells with orthogonal IL-22R (o22R) and oGCSFR, neither of which are natively expressed on T cells, exhibited stem-like and exhaustion-resistant transcriptional and chromatin landscapes, enhancing anti-tumour properties. Non-native receptor pairings and their resultant JAK-STAT signals open a path to diversifying T cell states beyond those induced by natural cytokines.

Fig. 1. Signal activation and gene expression profiles induced by orthogonal chimeric receptors.

a, Schematic of orthogonal IL-2–IL-2Rβ pairs, whereby a mutant IL-2 interacts with a mutant IL-2Rβ that pairs with the common γ_c to trigger swappable ICD signalling. b, Natural or non-natural ICDs that pair with the γ_c. c–g, Orthogonal-receptor-expressing (YFP⁺) C57BL/6 WT CD3⁺ T cells were stimulated with MSA fusions of orthogonal mouse IL-2 (MSA–oIL-2) for 20 min. Cells were analysed for pSTAT. c, pSTAT3 signalling dose–response curves of o2R-, o10R-, o20R- and o22R-transduced WT T cells, plotted against the log₁₀-transformed concentration of MSA–oIL-2. d, The relative pSTAT1, pSTAT3, pSTAT4, pSTAT5 and pSTAT6 mean fluorescence intensity (MFI) in T cells transduced with o2R, o4R, o7R, o9R and o21R treated with MSA–oIL-2 (10 µM), normalized to the YFP⁻ controls. e, The relative pSTAT1, pSTAT3 and pSTAT5 MFI in T cells transduced with o2R, oIFNAR2, oIFNGR1 and oIFNLR1, treated with MSA–oIL-2 (10 µM), normalized to the YFP⁻ controls. f, The relative pSTAT1, pSTAT3, pSTAT4 and pSTAT5 MFI in T cells transduced with o2R, o10R, o20R and o22R and treated with MSA–oIL-2 (10 µM), normalized to the YFP⁻ controls. g, The relative pSTAT1, pSTAT3 and pSTAT5 MFI in T cells transduced with o2R, oEPOR and oGCSFR and treated with MSA–oIL-2 (10 µM), normalized to the YFP⁻ controls. h–j, C57BL/6 WT T cells transduced with orthogonal receptors were stimulated with MSA–oIL-2 (5 μM) or recombinant cytokines (10 nM) for 6 h, and then analysed using RNA-seq. n = 3 biologically independent samples. h, PCA of RNA-seq data. i, The expression of STAT-driven gene signatures. FC, fold change compared with the no-ICD controls. j, Comparison of the STAT gene signatures of orthogonal chimera-transduced T cells treated with MSA–oIL-2 versus those treated with the indicated native cytokines. Data are mean ± s.e.m. Source data

Fig. 2. Synthetic signalling drives enhanced anti-tumour efficacy of ACT immunotherapy.

a–i, Female C57BL/6 WT mice bearing established subcutaneous (s.c.) B16F10 tumours were sublethally lymphodepleted by total body irradiation (5 Gy) on day 5. On day 6, the mice received intravenous (i.v.) ACT of untransduced (NT) pmel-1 (pmel) T cells (3 × 10⁶ cells) or pmel-1 T cells transduced with the indicated orthogonal chimeric receptor (γ_c receptor (b,c), IFN receptor (d,e), IL-10R (f,g) and homodimeric receptor (h,i); 3 × 10⁶ cells) followed by intraperitoneal (i.p.) administration of MSA–oIL-2 (2.5 × 10⁴ U per day) or PBS every other day until day 20. n = 8 (o4R, o7R, o21R, oIFNAR2, oIFNGR1, oIFNLR1, oEPOR and oGCSFR) and n = 9 (other groups) mice. a, The experimental timeline. b–i, The average tumour growth curves (b, d, f and h) and survival curves (c, e, g and i) of each treatment group. Data are mean ± s.e.m. Statistical analysis was performed using two-way analysis of variance (ANOVA) with Tukey’s post-test (b, d, f and h) or log-rank (Mantel–Cox) tests (c, e, g and i). NS, not significant. The diagram in a was created using BioRender. Zhao, Y. (2025) https://BioRender.com/t1qoa7d. Source data

Fig. 3. Divergent T cell fates induced by synthetic cytokine receptors.

a–j, Female C57BL/6 WT mice bearing established s.c. B16F10 tumours were lymphodepleted on day 7. On day 8, the mice received an i.v. adoptive transfer of 3 × 10⁶ pmel-1 T cells (NT or transduced with the indicated orthogonal chimeric receptor) followed by i.p. administration of MSA–oIL-2 (2.5 × 10⁴ U per day) or PBS every other day until day 20. On day 17, the mice were euthanized, and tumours, TDLNs and spleens were collected for flow cytometry analysis. For scRNA-seq analysis, Thy1.1⁺CD8⁺ TILs from ACT with NT pmel-1 T cells, and Thy1.1⁺CD8⁺ YFP⁺ TILs from ACT with o4R, o20R, o22R and oGSCFR pmel-1 T cells were sorted for scRNA-seq analysis. a, The experimental timeline. b, The gating strategy for TILs used in scRNA-seq analysis. c, The frequencies of Thy1.1⁺CD8⁺ cells for the live single-cell gate. n = 1 biologically independent sample. d, UMAP representation of NT pmel-1 T cells and those transduced with o4R, o20R, o22R and oGSCFR. e, UMAPs separated by the transduction conditions. f, The proportion of clusters in each transduction condition. g, The predicted activities of marker TFs. h, The expression of Tcf7 and Bach2 mRNA. i, The predicted activity of LEF1 and FOXO1. j, The frequencies of CD44^–CD62L⁺SCA1⁺ T_SCM cells among Thy1.1⁺CD8⁺ pmel-1 T cells in TDLNs. n = 8 (oEPOR and o9R), n = 9 (oGCSFR) and n = 10 (other groups) mice. Data are mean ± s.e.m. Statistical analysis was performed using one-way ANOVA with Tukey’s post-test (j). The diagram in a was created using BioRender. Zhao, Y. (2025) https://BioRender.com/t2vr9gg. Source data

Fig. 4. oGCSFR expression endows CD8⁺ T cells with myeloid cell features.

a, Mouse TIL scRNA-seq analysis (Fig. 3), showing the predicted activity of NFIC, MAF and MAFB. b, The top-ranked genes targeted by MAFB (top) and C/EBPb (bottom). c, Average expression (Avg. expr.) of myeloid-cell-associated surface-marker-encoding genes that are differentially upregulated in C6 cells. d, Expression of Cd14, Fcgr1, Cd86 and Plaur mRNA. e, Prediction of intercellular ligand–receptor interactions between T cells in each cluster. f, Predicted interactions between ligands expressed in C6 cells and receptors expressed in cells in the other clusters. g–k, NT pmel-1 CD8⁺ T cells and those transduced with Nfic, Maf, Mafb or oGSCFR were cultured for 48 h, with oGSCFR pmel-1 CD8⁺ T cells also stimulated with MSA–oIL-2 (500 nM) during this period. n = 4 biologically independent samples. g, Representative flow cytometry plots showing the frequencies of Mac-1⁺, CD14⁺, CD64⁺ and Gr-1⁺ cells among pmel-1 CD8⁺ T cells. h–k, The frequencies of Mac-1⁺ (h), CD14⁺ (i), CD64⁺ (j) and Gr-1⁺ (k) cells among pmel-1 CD8⁺ T cells. l, GFP⁺ L. monocytogenes was co-cultured with NT or oGSCFR pmel-1 CD8⁺ T cells in serial tenfold dilutions. The MFI of GFP⁺ L. monocytogenes is shown. n = 4 biologically independent samples. m, Representative fluorescence (top) and bright-field (bottom) images of GFP⁺ L. monocytogenes internalized by oGCSFR CD8⁺ T cells. Scale bar, 5 μm. n,o, FarRed-labelled A20 cells were pretreated with anti-mouse CD19 antibodies. Effector cells were co-cultured with the A20 cells for 2 h, then analysed using flow cytometry. n = 5 biologically independent samples. Representative flow cytometry plots (n) and frequencies (o) of FarRed⁺CD14⁺ cells (RAW 264.7 cells) or FarRed⁺CD8⁺ cells (T cells) are shown. p, The frequencies of SIRPα⁺ cells. n = 4 biologically independent samples. Data are mean ± s.e.m. Statistical analysis was performed using two-tailed Student’s t-tests (l and p) and one-way ANOVA with Tukey’s post-test (h–k and o). Source data

Fig. 5. Human orthogonal chimeric IL-4R signalling redirects T cell differentiation into a type 2 phenotype and improves anti-tumour activity in the human melanoma xenograft model.

a, Mouse TIL scRNA-seq analysis (Fig. 3) of the expression levels of marker genes for C2. b, Expression of Il4, Il13 and Il5 mRNA. c, Representative flow cytometry plots with quadrant gating showing the IL-4⁺, IL-5⁺, IL-13⁺ and IFNγ⁺ cytokine subpopulations in CD4⁺ and CD8⁺ NT or human orthogonal chimeric IL-4R (ho4R)-transduced NY-ESO-1 TCR-T cells. d, The frequencies of the IL-4⁺, IL-5⁺ and IL-13⁺ subpopulations among CD4⁺ TCR-T cells. n = 4 biologically independent samples. e, The frequencies of the IL-4⁺, IL-5⁺ and IL-13⁺ subpopulations among CD8⁺ TCR-T cells. n = 4 biologically independent samples. f, The frequencies of the IFNγ⁺IL-4⁺, IFNγ⁺IL-5⁺ and IFNγ⁺IL-13⁺ subpopulations among CD4⁺ TCR-T cells. n = 4 biologically independent samples. g, The frequencies of the IFNγ⁺IL-4⁺, IFNγ⁺IL-5⁺ and IFNγ⁺IL-13⁺ subpopulations among CD8⁺ TCR-T cells. n = 4 biologically independent samples. h, The GATA3 MFI. n = 4 biologically independent samples. i,j, Female NSG mice bearing s.c. HLA*0201⁺NY-ESO-1⁺ A375 tumours received i.v. transfer of 3 × 10⁶ NT or ho4R-expressing CD3⁺ NY-ESO-1 TCR-T cells on day 6, followed by i.p. administration of MSA–hoIL-2 (1 × 10⁵ U per day) every other day. n = 5 mice. i, The experimental timeline. j, Tumour growth curves. k, The experimental setting is described in Extended Data Fig. 7d. Female NSG mice bearing s.c. A375 tumours received an i.v. transfer of 2 × 10⁶ NT or ho4R-expressing CD3⁺ NY-ESO-1 TCR-T cells on day 10, followed by i.p. administration of MSA–hoIL-2 (2.5 × 10⁴ U, 3 doses, 1 × 10⁵ U, 3 doses) every other day. On day 21, mice were euthanized for flow cytometry analyses. n = 5 mice. The frequencies of the IL-4⁺, IL-5⁺, IL-13⁺ subpopulations among CD3⁺ TCR-T cells in spleens are shown. Data are mean ± s.e.m. Statistical analysis was performed using two-tailed Student’s t-tests (d–h and k) and two-way ANOVA with Tukey’s post test (j). The diagram in i was created using BioRender. Zhao, Y. (2025) https://BioRender.com/98rylzr. Source data

Fig. 6. Enhanced potency of CD19 CAR T cells and NY-ESO-1 TCR-T cells through ho22R and hoGSCFR signalling in xenograft tumour models.

a,b, Female NSG mice bearing NALM6-Luc B cell acute lymphoblastic leukaemia (ALL) received an i.v. transfer of 4 × 10⁵ CD3⁺ CD19 CAR T cells on day 5, followed by i.p. administration of MSA–hoIL-2 (2.5 × 10⁴ U) every day. n = 5 mice. The blood samples were analysed on day 11. a, The experimental timeline. b, Individual radiance measured using bioluminescence imaging (BLI). c,d, Female NSG mice bearing Raji lymphoma received an i.v. transfer of 1 × 10⁶ CD3⁺CD19⁺ CAR T cells on day 7, followed by i.p. administration of MSA–hoIL-2 (2.5 × 10⁴ U) every other day. n = 5 mice. c, The experimental timeline. d, Survival curves. e–g, Female NSG mice bearing A375 melanoma received an i.v. transfer of 3 × 10⁶ CD3⁺ NY-ESO-1 TCR-T cells on day 6, followed by i.p. administration of MSA–hoIL-2 (1 × 10⁵ U) every other day. n = 5 mice. e, The experimental timeline. f, Tumour growth curves. g, Survival curves. h,i, The experimental setting is described in Extended Data Fig. 11c. n = 5 mice. h, Counts of CD3⁺EGFR⁺ TCR-T cells in tumours. i, The frequencies of T_SCM cells among EGFR⁺CD8⁺ TCR-T cells in tumours. j–n, The experimental setting is described in Extended Data Fig. 11c. n = 3 biologically independent samples. j, Overlap of differential chromatin accessibility regions between groups. False-discovery-rate (FDR)-adjusted P < 0.05. k, The number of significantly differentially accessible chromatin regions nearby versus their mean log₂-transformed fold change per gene. l, The log₂-transformed fold change in significantly differentially accessible chromatin regions nearby RORA, RUNX1 and TOX. m, Representative chromatin accessibility profiles of TOX. Chr., chromosome. n, The most differentially enriched TF motifs in open chromatin regions. Data are mean ± s.e.m. Statistical analysis was performed using one-way ANOVA (h and i), two-way ANOVA with Tukey’s post-test (b and f) or log-rank (Mantel–Cox) tests (d and g). The diagrams in a, c and e were created using BioRender. Zhao, Y. (2025). a, https://BioRender.com/giefd3b; c, https://BioRender.com/ulrtai9; e, https://BioRender.com/f6qd7eb. Source data

Extended Data Fig. 1. Characterization and signalling profiles of orthogonal receptor expressing WT mouse T cells.

a, Orthogonal chimeric receptor constructs were introduced into WT mouse T cells via YFP-encoding retroviral vectors. Expression levels of each receptor were assessed by YFP fluorescence using flow cytometry. b–e, Orthogonal receptor expressing (YFP⁺) C57BL/6 WT CD3⁺ T cells were stimulated with MSA fusions of MSA-oIL2 for 20 min. Cells were analysed for pSTAT using flow cytometry. Data are presented as the MFI of pSTAT from YFP⁺ T cells plotted against the log₁₀ concentration of MSA-oIL2. b, pSTAT-1, -3, -4, -5, and -6 signalling dose-response curves of orthogonal IL-2Rβ ECD–ICD from o4R, o7R, o9R, and o21R transduced T cells (n = 2 biologically independent samples). c, pSTAT-1, -3, and -5 signalling dose-response curves of orthogonal IL-2Rβ ECD–ICD from oIFNAR2, oIFNGR1, and oIFNLR1 transduced T cells (n = 2 biologically independent samples). d, pSTAT-1, -4, and -5 signalling dose-response curves of o10R, o20R, and o22R transduced T cells (n = 2 biologically independent samples). e, pSTAT-1, -3, and -5 signalling dose-response curves of orthogonal IL-2Rβ ECD–ICD from oGCSFR and oEPOR transduced T cells (n = 3 biologically independent samples). All data represent mean ± s.e.m. Source data

Extended Data Fig. 2. Orthogonal cytokine receptor system dependent on ligand-induced heterodimerization with γ_c.

a–f, Orthogonal chimeric receptors were introduced into YT-1 cells using YFP-encoding retroviral vectors. Cells were stimulated with MSA-human orthogonal IL2 (MSA-hoIL2) for 20 min and analysed for pSTAT via flow cytometry (n = 2 biological independent samples). Data are presented as the MFI of pSTAT from YFP⁺ T cells, plotted against the log₁₀ concentration of MSA-hoIL2. a–c, pSTAT1 (a), pSTAT3 (b), and pSTAT5 (c) signalling dose-response curves for human o2R (ho2R), ho10R, ho22R, and hoGCSFR-transduced YT-1 (CD25⁺) cells. d–f, pSTAT1 (d), pSTAT3 (e), and pSTAT5 (f) signalling dose-response curves for ho2R, ho10R, ho22R, and hoGCSFR-transduced γ_c knockout (KO) YT-1 cells. g, ho2R were transduced into CD25⁺ YT-1 cells. Cells were stimulated with orthogonal IL-2 (oIL-2) or oIL-2-TR (carrying the Q126T and S130R γ_c interface mutations) for 20 min and analysed for pSTAT via flow cytometry. Data are presented as the MFI of pSTAT from YFP⁺ T cells. h, CD19 CAR T cells transduced with the indicated CAR-encoding construct (n = 2 biological independent samples). CD19_28z: a murine CD19 CAR (1D3 clone); CD19_28z_YRHQ: CD19 CAR with a STAT3-binding motif (YRHQ) at the C-terminus; CD19_28z_22 R CTD: CD19 CAR with the C-terminal domain (CTD) of IL-22R fused at the C-terminus. CAR T cells were stimulated with CD19⁺ B cells or treated with IL-21 (50 ng/mL) served as a control. Shown are MFI of pSTAT3 over time. i, Normalized mRNA expression levels of indicated genes in mouse T cells transduced with o2R lacking ICD (NoICD) or oIFNGR1, following 6- or 48-hour stimulation (unstimulated [NS] or stimulated with MSA-oIL-2) (n = 1 biological independent sample). All data represent mean ± s.e.m. and are analysed by one-way ANOVA with Tukey’s post-test (h). Source data

Extended Data Fig. 3. Orthogonal cytokine receptor signalling enhances antitumor efficacy of pmel T cells in B16F10 tumour model without causing toxicity.

a,b, Experimental setting is described in Fig. 2a. Relative body weight of mice post indicated treatment. c, Experimental setting is described in Fig. 2a. Shown are individual tumour growth curves. Indicated are numbers of tumour-free mice among the total number of mice in the group. d, Experimental setting is described in Fig. 2a (n = 5 animals). e, WT mouse T cells transduced with the indicated orthogonal chimeric receptor were cultured in MSA-oIL2 (100 nM) for two days (n = 4 biologically independent samples). Shown are viable T cell counts. Shown are average tumour growth curves. All data represent mean ± s.e.m. and are analysed by one-way (e) or two-way ANOVA with Tukey’s post-test (d). Source data

Extended Data Fig. 4. o22R pmel T cells exhibited the most prominent cell expansion in the B16F10 tumours.

a–i, C57BL/6 mice bearing B16F10 tumours were lymphodepleted on day 7. On day 8, mice received an i.v. adoptive transfer of pmel T cells by i.p. administration of MSA-oIL2 every other day. Mice received i.p. administration of Bromodeoxyuridine (BrdU) on day 16. On day 17, mice were sacrificed, indicated organs were collected for flow cytometry analysis. a, The experimental timeline. b, Counts of Thy1.1⁺CD8⁺ pmel T cells per mg tumour (n = 8 for oEPOR, o9R, o20R; n = 9 for o10R; n = 10 for all other groups). c, Counts of Thy1.1⁺CD8⁺ pmel T cells per spleen (n = 8 for oEPOR; n = 9 for o9R, o10R, o20R and oGCSFR; n = 10 for all other groups). d, Counts of Thy1.1⁺CD8⁺ pmel T cells per TDLN (n = 8 for oEPOR; n = 9 for o9R, o10R, o20R, and o22R; n = 10 for all other groups). e, Counts of Thy1.1⁺CD8⁺ pmel T cells per 100 μl blood (n = 8 for oEPOR and o9R; n = 9 for o22R; n = 10 for all other groups). f, Frequencies of BrdU⁺ cells among Thy1.1⁺CD8⁺ pmel T cells in tumours (n = 10). g, Frequencies of Ki67⁺ cells among Thy1.1⁺CD8⁺ pmel T cells in tumours (n = 5). h, Frequencies of CD45.2⁺ cells in tumours (n = 8 for oEPOR; n = 9 for o9R, o10R, and o20R; n = 10 for all other groups). i, Spleen weight of mice from each treatment group (n = 8 for oEPOR; n = 9 for o9R, and o10R; n = 10 for all other groups). All data represent mean ± s.e.m. and are analysed by two-tailed Student’s t-test (f, g) or one-way ANOVA with Tukey’s post-test (b–e, h, i). The diagram in a was created using BioRender. Source data

Extended Data Fig. 5. TILs with orthogonal chimeras exhibited enhanced cytotoxicity.

a–k, Experimental setting is described in Fig. 3a. a, Average expression levels of top-ranked genes across clusters. b, Tox expression over the UMAP. c, Violin plots showing expression levels of Tox cross clusters. d, Violin plots showing expression levels of Tox cross treatment groups. e, Gzma, Gzmb, Prf1, and Ifng expression over the UMAP. f, Violin plots showing expression levels of Gzma, Gzmb, Prf1, and Ifng cross clusters. g, Violin plots showing expression levels of Gzma, Gzmb, Prf1, and Ifng cross treatment groups. h, Representative flow cytometry plots showing IFN-γ⁺TNFα⁺ (top) and Granzyme B⁺ (bottom) subpopulations among Thy1.1⁺CD8⁺ pmel T cells in tumours. i, Counts of IFNγ⁺ TNFα⁺ Thy1.1⁺CD8⁺ T cells per mg tumour (n = 5 animals). j, Counts of Granzyme B⁺Thy1.1⁺CD8⁺ T cells per mg tumour (n = 5 animals). k, Frequencies of IL-7Rα⁺KLRG1⁻ cells among Thy1.1⁺CD8⁺ pmel T cells in spleens (n = 8 for oEPOR; n = 9 for o9R, o10R, and oGCSFR; n = 10 for all other groups). All data represent mean ± s.e.m. and are analysed by one-way ANOVA with Tukey’s post-test (i–k). Source data

Extended Data Fig. 6. Orthogonal GCSFR drives myeloid-like T cell states.

a, Proportion of each treatment group unmapped to the reference atlas of TILs. b, Gene Ontology enrichment analysis of oGCSFR-upregulated genes identified the top 20 enriched pathways (enrichment score ≥1.2, adjusted p < 0.05). c–f, NT pmel CD8⁺ T cells and those transduced with Nfic–Maf–MafB, or oGSCFR were cultured for 72 h, with oGSCFR pmel CD8⁺ T cells stimulated with MSA-oIL2 (500 nM) during this period, followed by flow cytometry analysis (n = 7 biological independent samples). Shown are the frequencies of Mac-1⁺ (c), CD14⁺ (d), CD64⁺ (e), and Gr-1⁺ (f) cells. g, MFI of pHrodo Red E. coli bioparticles (n = 3 biologically independent samples). h, Shown are the pHrodo Red confluences (n = 8 biologically independent samples). i, The experimental setup is described in Fig. 4l. Shown are the representative fluorescence images of GFP⁺ L. monocytogenes at the highest concentration. j, Experimental setting is described in Fig. 4n,o. Representative fluorescence images of CSFE⁺ (green) effector cells cocultured with FarRed⁺ (red) A20 target cells at a 1:1 E:T ratio. k, Pmel CD8⁺ T cells were cocultured for 2 h with FarRed⁺ A20 cells pretreated with mouse CD19 (mCD19) antibody, followed by flow cytometry analysis (n = 6 biologically independent samples). Shown are the frequencies of FarRed⁺CD8⁺ cells. l, oGCSFR pmel CD8⁺ T cells were preserved in MSA-oIL2 (500 nM) for 24 h. B16F10-mCD19 cells were co-cultured with pmel CD8⁺ T cells at a 1:1 E:T ratio with MSA-hoIL2 (200 nM), in the presence or absence of mCD19 antibodies for 48 h. Cells were collected for flow cytometry analysis (n = 4 biologically independent samples). Shown are the percentages of B16F10-mCD19 cell killing. All data represent mean ± s.e.m. and are analysed by two-tailed Student’s t-test (k) or one-way ANOVA with Tukey’s post-test (c–f, g, l). Source data

Extended Data Fig. 7. Human orthogonal IL-4R enhances type 2 function and antitumor activity.

a, NY-ESO-1 TCR and ho4R constructs were introduced via retroviral vectors. Expression levels of NY-ESO-1 TCR and ho4R were assessed by EGFR and YFP expression using flow cytometry, respectively. Shown are representative flow cytometry plots of transduction efficiency. b, Frequencies of IFN-γ⁺ subpopulation among CD4⁺ (left) and CD8⁺ (right) NY-ESO-1 TCR-T cells (n = 4 biologically independent samples). c, MFI of CXCR3 (left) and CCR4 (right) among CD3⁺ NY-ESO-1 TCR-T cells (n = 4 biologically independent samples). d, The experimental timeline for Fig. 5k. e, Experimental setting is described in Fig. 5i. Relative body weight of mice post indicated treatment. f, Representative flow cytometry histograms showing GATA-3 expression level. g, NT or ho4R NY-ESO-1 TCR-T cells (unmodified, GATA3 knockout, or treated with the indicated antibodies) were co-cultured with A375 cells at an E:T ratio of 1:2 for 48 h (n = 3 biologically independent samples). Shown are the counts of viable A375 tumour cells. All data represent mean ± s.e.m. and are analysed by two-tailed Student’s t-test (b,c) or one-way ANOVA with Tukey’s post-test (g). The diagram in d was created using BioRender. Source data

Extended Data Fig. 8. ho22R and hoGCSFR signalling enhance the antitumor potency of CD19 CAR-T cells and promote T cell stemness.

a, CD19 CAR, ho22R, hoGCSFR constructs were introduced via retroviral vectors. Expression levels of CD19 CAR were measured by Myc-tag fluorescence using flow cytometry. Expression levels of ho22R and hoGCSFR were measured by YFP fluorescence using flow cytometry. Shown are representative flow cytometry plots of transduction efficiency. b, pSTAT-1, -3, -4, and -5 signalling dose-response curves (n = 2 biologically independent samples). Data are presented as the MFI of pSTAT against the log₁₀ concentration of MSA-hoIL2. c,d, CD19 CAR T, ho22R CD19 CAR T, and hoGCSFR CD19 CAR T cells were cocultured with NALM6-Luciferase (NALM6-Luc) cells or Raji cells at different E:T ratios in the presence of MSA-hoIL2 (100 nM) for 24 h (n  =  3 biologically independent samples). Shown are the percentages of killing of NALM6-Luc cells (c) and Raji cells (d). e–g, CD19 CAR T, ho22R CD19 CAR T, and hoGCSFR CD19 CAR T cells were cultured in MSA-hoIL2 (100 nM) supplemented T cell medium for 3 days (n  =  3 biologically independent samples). e, Shown are MFI of CD62L and CD45RA among CD19 CAR T cells. f, Representative flow cytometry plots showing the gating strategy for CD45RA⁺ CD62L⁺ CCR7⁺ CD95⁺ subpopulations in NT, hoGCSFR, and ho22R CD3⁺ NY-ESO-1 TCR-T cells. g, Frequencies of CD45RA⁺ CD62L⁺ CCR7⁺ CD95⁺ cells. h–j, NT and oGCSFR CD8⁺ human T cells were stimulated with MSA-hoIL2 (500 nM) for 72 h, followed by flow cytometry analysis (n = 6 biologically independent samples). h, Representative flow cytometry plots of Mac-1⁺ and CD66b⁺ cells among CD8⁺ T cells. i,j, Shown are the frequencies of Mac-1⁺ (i) and CD66b⁺ (j) cells among CD8⁺ T cells. All data represent mean ± s.e.m. and are analysed by two-tailed Student’s t-test (i,j), one-way ANOVA (g), or two-way (c–e) ANOVA with Tukey’s post-test. Source data

Extended Data Fig. 9. ho22R and hoGCSFR signalling enhances the anti-tumour potency of CD19 CAR T cells against haematological malignances.

a, Experimental setting is described in Fig. 6a. Show are bioluminescence images. b, Experimental setting is described in Fig. 6a. Counts of CD19 CAR T cells per μl blood (n = 4 animals). c, Experimental setting is described in Fig. 6c. Individual tumour growth curves. d,e, NSG male mice were inoculated (s.c.) with Raji cells (1 × 10⁶) and received an i.v. adoptive transfer of 1 × 10⁶ NT, ho22R, or hoGCSFR CD3⁺ CD19 CAR T cells on Day 7, followed by i.p. administration of MSA-hoIL2 (2.5 × 10⁴ unit) or PBS every other day until day 29 (n = 5 animals). d, Average tumour growth curves. e, Survival curves. Indicated are the numbers of tumour-free mice per total number of mice in each group. f–j, NSG female mice were inoculated (s.c.) with Raji cells (2 × 10⁶) and received an i.v. adoptive transfer of 1 × 10⁶ NT, ho22R, or hoGCSFR CD3⁺ CD19 CAR T cells on Day 11, followed by i.p. administration of MSA-hoIL2 (2.5 × 10⁴ unit) or PBS every other day until day 25 (n = 6 animals). On day 26, the mice were sacrificed, and tumours were collected for flow cytometry analysis. f, The experimental timeline. g, Average tumour growth curves. h, Counts of CD3⁺ Myc⁺ CD19 CAR T cells in tumours. i, MFI of PD-1 (left) and LAG-3 (right) among CD3⁺ Myc⁺ CD19 CAR T cells in tumours. j, MFI of Granzyme B (left) and IFN-γ (right) among CD3⁺ Myc⁺ CD19 CAR T cells in tumours. All data represent mean ± s.e.m. and are analysed by one-way (b, h–j), or two-way (d) ANOVA with Tukey’s post-test, or log-rank (Mantel-Cox) test (e). The diagram in f created using BioRender.com. Source data

Extended Data Fig. 10. ho22R and hoGCSFR NY-ESO-1 TCR-T cells exhibit enhanced antitumor efficacy in HLA*0201⁺ NY-ESO-1⁺ A375 xenograft tumour model.

a, NY-ESO-1 TCR, ho2R, ho22R, and hoGCSFR constructs were introduced via retroviral vectors. Expression levels were measured by EGFR (NY-ESO-1 TCR) or YFP fluorescence (ho2R, ho22R, and hoGCSFR) using flow cytometry. Shown are representative flow cytometry plots of transduction efficiency. b, pSTAT-1, -3, -4, and -5 signalling dose-response curves (n = 2 biologically independent samples). Data are presented as the MFI of pSTAT against the log₁₀ concentration of MSA-hoIL2. c, Fold growth of CD3⁺ NY-ESO-1 TCR-T cells post 3-day culture in the presence of MSA-hoIL2 (100 nM) (n = 4 biologically independent samples). d–f, NT T cells, NY-ESO-1 TCR-T cells, or NY-ESO-1 TCR-T cells transduced with ho2R, ho22R, or hoGCSFR were cocultured at a 1:1 E:T ratio with the HLA*0201⁺ NY-ESO-1⁺ melanoma cell line (nRFP-M407) in the presence of MSA-hoIL2 (100 nM). nRFP-M407 cells were reintroduced after 48 h of coculture (blue arrows) in the presence of MSA-hoIL2 (100 nM). After three rounds of tumour challenge, cells were collected for flow cytometry analyses. d, Shown is the normalized tumour cell confluence (n = 4 biologically independent samples). e, Frequencies of Granzyme B⁺ (left) and IFN-γ⁺ (right) among CD3⁺ NY-ESO-1 TCR-T cells (n = 3 biologically independent samples). f, Frequencies of CD45RA⁺CD62L⁺CCR7⁺CD95⁺ among CD3⁺ NY-ESO-1 TCR-T cells (n = 3 biologically independent samples). g, Experimental setting is described in Fig. 6f. Shown are the individual tumour growth curves. Indicated are the numbers of tumour-free mice per total number of mice in each group. All data represent mean ± s.e.m. and are analysed by one-way ANOVA (e, f) or two-way (c, d) ANOVA with Tukey’s post-test. Source data

Extended Data Fig. 11. ho22R and hoGCSFR epigenetically rewire TILs in A375 xenograft tumour model.

a,b, Male NSG mice bearing A375 tumours e received an i.v. adoptive transfer of 3 × 10⁶ CD3⁺ NY-ESO-1 TCR-T cells on Day 6, followed by i.p. administration of MSA-hoIL2 every other day (n = 5 animals). a, Tumour growth curves. b, Survival curves. Indicated are the numbers of tumour-free mice per total number of mice. c, The experimental timeline for Fig. 6(h–n) and Extended Data Fig. 11d–h. Female NSG mice bearing A375 melanoma received an i.v. adoptive transfer of 3 × 10⁶ CD3⁺ NY-ESO-1 TCR-T cells on day 10, followed by i.p. administration of MSA-hoIL2 every other day. On day 21, the tumours were analysed by flow cytometry (n = 5 animals). On day 22, TCR-T cells in tumours were sorted for ATAC-seq analysis (n = 3 biological independent samples). d, MFI of PD-1, LAG−3, and TIM-3 among EGFR⁺CD8⁺ TCR-T cells in tumours (n = 5 animals). e, Principal component analysis (PCA). f, Proportions of differential or non-significant chromatin regions. g, KEGG pathway enrichment was performed on accessible promoter regions (TSS ± 1 kb) using one-tailed hypergeometric tests, with FDR-adjusted q values (Benjamini-Hochberg). h, Genes underlying the enrichment for the “KEGG cytokine-cytokine receptor interaction” term (differential chromatin regions within transcription start sites +/−1 kb of known genes). i–m, NT, ho22R, and ho22R (BACH2 knockout) CD3⁺ human T cells were restimulated with soluble CD3 antibodies (1 μg/mL) in the presence of MSA-hoIL2 (200 nM) for 48 h (n = 4). Cells were then collected for quantitative PCR and flow cytometry analysis. i, Relative expression of BACH2 (n = 4 biologically independent samples). j–m, MFI of CD62L (j), CD44 (k), TCF-1 (l), FOXO1 (m). All data represent mean ± s.e.m. and are analysed by one-way ANOVA (d, j–m), or two-way (a) ANOVA with Tukey’s post-test, or log-rank (Mantel-Cox) test (b). The diagram in c created using BioRender.com. Source data