Fig 2: NK cells maintain the expression of activating receptors after cryopreservation. PM21-NK cells expanded from T-cell-depleted PBMCs were obtained from multiple donors (N = 3–7). Cells were cryopreserved while donor-matched fresh PM21-NK cells (black squares) were maintained in culture. Frozen PM21-NK cells were thawed and rested 16 h the day before analysis (blue triangles) or thawed and rested for 1 h on the day of analysis (gray circles). Expression of NK cell activating receptors CD16, NKG2D, DNAM-1, NKp30, NKp46, NKp44, as well as CD69, granzyme B, and perforin was determined by flow cytometry. No statistically significant difference was seen in the expression of activating receptors either 1 h immediately post-thaw or after a 16-h rest in donor and expansion-matched PM21-NK cells (N = 5–7 donors). Data are presented as a scatter plot with each point representing the average of two technical replicates with donor-pair lines. No statistical significance was determined by multiple paired t-tests with a threshold for significance at p < 0.05.

Fig 3: HIOs promote proliferation and maturation of systemic human ILCPs(A) Gating strategy for PBMC-derived ILCPs, pregated on live, single, CD45+ cells (FMOs, blue and magenta).(B) Schematic of HIO + ILCP co-cultures, indicating the presence of mesenchymal cells and CD45+ ILCPs.(C) Representative plot overlays of EpCAM+ IECs, double-negative mesenchymal cells (M, magenta), and CD45+ ILCs (blue) after 14-day co-cultures, highlighting Matrigel debris in gray, and quantified in (D) as count of EpCAM-, CD45+, LINn- ILCs and (E) fold change expansion of ILCs after 14-day co-culture relative to the number of ILCPs seeded on day 1 (N = 3–15 across seven experiments).(F) Representative image of CD45+ ILCPs co-cultured with mesenchyme-depleted, EpCAM+ HIOs (scale bar: 25 µm). Error bars represent SEM; p values are from unpaired Student’s t tests.(G) Count of Live, EpCAM-CD45+LIN-ROR?t+ ILCs after 14-day co-culture with SD-HIOs or epithelial-depleted HIO-STROs expressing markers CCR6, NKp44, and/or T-bet (ILCPs from N = 3 PBMC donors).(H) Relative group 1 (Live, CD45+Lin-CRTH2-c-kit- [CD127+CD161+ in primary gut ILC only] and [EpCAM-ROR?t-GATA3- in SD-HIO and HIO-STRO only]), group 2 (Live, CD45+Lin-CRTH2+c-kit+/-, [CD127+CD161+/- in primary gut ILC only] and [EpCAM-ROR?t-GATA3+ in SD-HIO and HIO-STRO only]), group 3 (Live, CD45+Lin-CRTH2-c-kit+, NKp44+/- and [CD127+ CD161+ in primary gut ILC only] and [EpCAM-ROR?t+GATA3- in SD-HIO and HIO-STRO only]), and other LIN- ILCs (e.g., undifferentiated precursors; Live, CD45+, Lin-, CRTH2-, c-kit- and [CD127+ CD161+ in primary gut ILC only] and [EpCAM-, ROR?t-, GATA3- in SD-HIO and HIO-STRO only]) in unstimulated, primary human intestine (N = 13, adapted from Krämer et al., 2017), SD-HIOs (N = 13 from six experiments), and HIO-STROs (N = 7 from three experiments).

Fig 4: Human epithelial cells, not mesenchymal cells, drive proliferation and maturation of functional human ILCs(A) Frequency of Live, EpCAM-, CD45+, LIN-, ROR?t+ ILCs after 14-day co-culture expressing markers CCR6, NKp44, and/or T-bet (ILCPs from N = 3 donors).(B) Frequency of Live, EpCAM-, CD45+, Lineage-, ROR?t+ cells expressing IL-22+ and IL-17A+ in SD-HIO- and HIO-STRO-derived ILCs after 4-h stimulation with PMA/Ionomycin (ILCPs from N = 3 donors).(C) Representative flow plots corresponding to (B) (FMOs, blue and magenta).(D) Frequency of Live, EpCAM-, CD45+, LIN-, ROR?t- cells expressing T-bet and/or Eomes (ILCPs from N = 3 donors).(E) Representative flow plots and quantification of CD56 and IFN-? expression in T-bet+ and Eomes+ populations after PMA/Ionomycin 4-h stimulation (IFN-? FMO overlaid, magenta).(F) Total count of putative ILC2 (Live, EpCAM-, CD45+, Lineage-, ROR?t-, GATA3+, expressing either Klrg1, CD25, ST2, and/or CRTh2) after 14-day co-culture.(G) Expression of IL-5 and IL-13 in putative ILC2s after 4-h PMA/Ionomycin stimulation (FMOs, magenta and blue). Error bars represent SEM; p values are from unpaired Student’s t tests. All experiments were performed with ILCPs from N = 3 donors.

Fig 5: Gut-matured ILC2 upregulates ST2 on transfer to HLO culture(A) Representative image of SD-HIOs and SD-HLOs showing E-cadherin+ (E-CAD) epithelium, CD45+ ILCs, and nuclei (Hoechst) after 14-day co-culture (scale bars: 50 µm).(B) Count of EpCAM-, CD45+, LIN- ILCs after 14-day co-culture with SD-HIOs or SD-HLOs, with corresponding count of EpCAM-, CD45- mesenchyme.(C) Representative flow plots of ROR?t, CCR6, IL-22, IL-17A, IFN-?, and NKp44 after 14-day co-culture with SD-HIOs or SD-HLOs after 4-h PMA/Ionomycin stimulation (pregated population in gray, representative of N = 3).(D) IL-5 and IL-13 expression in putative GATA3+ ILC2s after 14-day co-culture and 4-h PMA/Ionomycin stimulation (FMOs, magenta and cyan).(E) Expression of CD25 and ST2 in putative GATA3+ ILC2s after 14-day co-culture with SD-HIOs or SD-HLOs.(F and G) Relative frequency of CD25+ and ST2+ putative ILC2s (F) and GeoMFI of ST2 in the same population with corresponding histogram overlay of ST2 (PE) after 14-day co-culture with SD-HIOs or SD-HLOs, followed SD-HIO-to-HIO, SD-HIO-to-HLO, or SD-HIO-to-HLO with 50 ng/mL hIL-33 neutralizing antibody (G). All experiments were performed with ILCPs from N = 3 donors; unpaired two-tailed Student’s t tests were used; and error bars represent SEM.