Fig 2: ILC2 are significantly increased in prostate cancer. a Quantification of PGD2 in supernatants of different human cancer cell lines (n = 2). b Representative example of flow cytometry analysis of expression of B7H6 in prostate cancer cell lines DU145 and PC3, or in a control cell line (721.221). c Relative frequencies of the ILC2 subset among total ILCs in healthy donors (HD) and prostate cancer patients (Prostate) (n = 21). d Relative frequencies of NKp30 expressing ILC2 in peripheral blood of HD and Prostate (n = 21). e Frequency of M-MDSC in peripheral blood of HD and Prostate (n = 21). f Relative frequencies of the ILC2 subset among total ILCs in healthy donors (HD, n = 21) and prostate cancer patients at stage (Gleason score) 6 (n = 8), 7 (n = 8), or 9/10 (n = 5) or in patients with benign prostate hypertrophy (BPH, n = 5). g Representative examples of flow cytometry analysis of innate lymphoid cell subsets in blood of FVB female mice (blood CTR, n = 5) and in blood (n = 29), spleen (n = 29) and prostate (n = 19) of TRAMP mice. h Relative frequencies of the ILC2 subset among total ILCs in blood of FVB female mice (blood CTR) and in blood, spleen and prostate of TRAMP mice. i Correlation of ILC2 and M-MDSCs in prostate tumours of TRAMP mice (n = 19) (three independent experiments). Error bars are s.e.m. Statistical analysis was performed using t test (c–e, h) and Kruskal-Wallis test (f)

Fig 3: ATRA reverses the PGD2-ILC2-IL-13-M-MDSC immunosuppressive axis. a Schematic representation of the hypothetical immunosuppressive chain established in APL patients at diagnosis. b Comparison of serum concentrations of PGD2 (APL = 11; Rem = 9), c frequencies of ILC2 NKp30+ (APL = 11, APL Rem = 9), d frequencies of total ILC2 (APL = 22; APL Rem n = 9), e serum concentrations of IL-13 (APL n = 11; APL Rem n = 9) and f frequencies of M-MDSC (APL n = 22; APL Rem = 9) in peripheral blood of APL patients at diagnosis (APL) or in remission after ATRA treatment (APL Rem). Dashed lines represent mean values of all the parameters in HD. g Schematic representation of APL establishment and ATRA treatment schedule in FVB/NJ mice upon injection of APL splenocytes (three independent experiments), as assessed by quantification of the normal composition of the bone marrow (Gr1+CD11b+ cells) h, as previously described27. i Survival curves of untreated (APL, n = 13) and ATRA-treated (ATRA) APL mice (n = 6). j Quantification of PGD2 concentrations (CTR n = 15, APL n = 13, ATRA day 14 n = 9, ATRA day 25 n = 6), k ILC2 frequencies and l M-MDSCs frequencies in control FVB/NJ mice (CTR), APL mice before (APL) and after ATRA treatment (ATRA day14; ATRA day25) (CTR n = 12, APL n = 10, ATRA day 14 n = 9, ATRA day 25 n = 6). m Schematic representation of APL establishment and ATRA treatment schedule in HIS mice upon injection of an APL cell line (two independent experiments). n Survival curves of untreated (APL, n = 5) and ATRA-treated (ATRA, n = 3) APL HIS mice. o Quantification of PGD2 concentrations (CTR n = 8, APL n = 8, ATRA n = 5), p ILC2 frequencies (CTR n = 11, APL n = 13, ATRA n = 9) and q M-MDSCs frequencies (CTR n = 8, APL n = 8, ATRA n = 5) in APL HIS mice before (APL) and after ATRA treatment at day 25 (ATRA) (up to two independent experiments). Error bars are s.e.m. Statistical analysis was performed using Mann-Whitney test (b–f), Kruskal-Wallis test (h, j–l, o–q) and Log-rank test (Mantel-Cox) (i, n)