Targeting ADAR1 with a small molecule for the treatment of prostate cancer

a-c, Comparison of overall RNA A-to-I editing (a), ADAR1 (b) and ADAR2 (c) expression level between matched tumor and normal tissues from 16 PCa patients. Statistical analysis was performed using two-tailed paired Student’s t test. d, Volcano plot of significantly affected genes (absolute fold change > 2, P value < 0.05) in high and low ADAR1 expression patient groups (from 498 TCGA patients, 50% cutoff). Statistical analysis was performed using two-tailed unpaired t-test. e, GSEA shows upregulated or downregulated pathways in low ADAR1 expression group. NES, normalized enrichment score. NES, normalized enrichment score. f, GSEA shows upregulated (left) or downregulated hallmark pathways (right) in tumors based on RNA-seq data of three pairs of PCa tissues. g, ADAR1 mRNA level in tumor and normal tissues in three PCa patients. Data represent the mean ± SD. n = 3 biologically independent replicates, assessed as one experiment. Statistical analysis was performed using two-tailed unpaired Student’s t-test. h, ADAR2 protein levels of tumor and normal tissues in three PCa patients (single experiment). i, Correlation analysis of ADAR1 expression and immune infiltration of cancer associated fibroblast (CAF), endothelial cell, CD4⁺ Th1 cell, CD8⁺ T cell, Treg cells, M1 macrophage, M2 macrophage in PCa. Spearman’s rho values and P values were calculated using two-sided Spearman correlation. These analyses were based on public TCGA database in PCa (n = 498 PCa patients) and were carried out by TIMER2.0 web tool. j, Correlation analysis between ADAR1 expression level and the level of eight immune check points in PCa (n = 498 patients, from TCGA). Statistical analysis was performed using two-sided Pearson correlation test.

a, Protein level of ADAR2 in DU-145 shADAR1 and shNC cells. b, Cell viability of VCaP shADAR1 and shNC cells during a 5-day course. n = 9 cell cultures from three independent experiments. c, A-to-I editing frequency of known ADAR1 targets GLI1 on DU-145 shADAR1 and shNC cells. d, Migration and invasion ability of VCaP shADAR1 and shNC cells. Representative images of migrated and invaded cells were shown. Scale bar, 200 μm. n = 9 cell cultures from three independent experiments. e, Cell cycle phase distribution of VCaP shADAR1 and shNC cells. n = 3 biological replicates, assessed as one experiment. The assay was repeated three times with similar results, and representative data was shown. The original data were shown in the Source Data. f, Apoptotic rates of VCaP shADAR1 and shNC cells. n = 9 cell cultures from three independent experiments. g, Cell viability of DU-145 sgADAR1 and sgNC cells during a 5-day course. n = 9 cell cultures from three independent experiments. h, Migration and invasion ability of DU-145 sgADAR1 and sgNC cells. Representative images of migrated and invaded cells were shown. Scale bar, 200 μm. n = 9 cell cultures from three independent experiments. i, Cell cycle phase distribution of DU-145 sgADAR1 and sgNC cells determined by flow cytometry. n = 9 cell cultures from three independent experiments. j, Apoptotic rates of DU-145 sgADAR1 and sgNC cells. n = 9 cell cultures from three independent experiments. k, Images of mice bearing tumor when mice were killed at the endpoint. l, A-to-I editing frequency of GLI1 on DU-145 cells that were overexpressed with empty vector (EV), Flag-tagged ADAR1 wild type (WT), and ADAR1 E912A mutant (Mut). m, Cell viability of VCaP ADAR1 WT, Mut, and control cells during a 5-day course. n = 9 cell cultures from three independent experiments. For b, d–j, and m, data represent the mean ± SD, and statistical analysis was performed using two-tailed unpaired t-test.

a, Protein level of ADAR2 in 22Rv1 cells that were infected with shNC or shADAR2 plasmid (single experiment). b, Volcano plot of significantly affected genes in 22Rv1 shADAR2 group relative to shNC group revealed by RNA-seq. Both shADAR1 and shNC groups contain two biological replicates, assessed as one experiment. P values were calculated using R v4.0.3 software. Statistical analysis was performed using two-tailed unpaired t-test. c, GSEA shows upregulated or downregulated pathways in shADAR2 cells compared to shNC. Statistical analysis was performed using two-sided permutation test, and P values were adjusted by Benjamini-Hochberg method. d, Histogram of number of various types of editing events in transcriptome-wide of 22Rv1 shADAR2 and shNC samples. n = 2 independent RNA-seq samples per group, assessed as one experiment. e, Expression level of ADAR1 and ADAR2 in DU-145 and 22Rv1 ADAR1 KD and control cells measured by RNA-seq. n = 2 independent RNA-seq samples per group, assessed as one experiment.

a, Histogram of various editing events in transcriptome-wide of DU-145 and 22Rv1 shADAR1 and shNC samples (left). The right graphs show Alu editing index (AEI) comparison between DU-145 and 22Rv1 shADAR1 and shNC samples. Data were calculated from two independent RNA-seq samples per group, assessed as one experiment. b, Distribution of A-to-I editing sites within Alu, L1 (LINE1), ERV1 (Endogenous retroviral sequence 1), other classes of repetitive element (other repeat), and non-repetitive element repeat (no repeat) was shown in DU-145 (top) and 22Rv1 (bottom) cells. Vertical axis indicates the percentage of A-to-I editing sites in each element class. Size of wedges indicate the percentage of A-to-I editing sites in each genomic region. c, Distribution of ADAR1-enriched RNA peaks across different mRNA regions in 22Rv1 as detected by RIP-sequencing. d, Venn gram showing intersection between RNA-immunoprecipitation (RIP) enriched RNA peaks, A-to-I editing genes (more than 0.1 in shNC relative to shADAR1). Shared genes were then intersected with A-to-I editing genes (more than 0.1 in tumor relative to normal tissue) in PCa tumors (three paired matched PCa tumor and normal tissues). e, Validation of potential targets by RIP-qPCR assays in DU-145 and 22Rv1 cells. Data represent the mean ± SD. n = 3 independent cell culture replicates, assessed as one experiment. Statistical analysis was performed using two-tailed unpaired t-test. f, Validation of editing sites (indicated by RNA-seq) in IFNAR1 by Sanger sequencing. g, Correlation of expression of ADAR1 and MTDH in a TCGA cohort of 498 PCa patients, analyzed and plotted by LinkedOmics (http://www.linkedomics.org/). Statistical analysis was performed using two-sided Pearson correlation test. h,i, Sequence chromatogram (h) of MTDH showing editing level difference between tumor and normal tissue from a patient and quantification (i) of editing level in single site of MTDH in three patients. Data represent the mean ± SD, obtained from three patients, assessed as one experiment. Statistical analysis was performed using two-tailed unpaired t-test. j, Comparison of mRNA levels of MTDH between tumor and matched normal tissues from three patients. Data represent the mean ± SD. n = 3 biologically independent replicates, assessed as one experiment. Statistical analysis was performed using two-tailed unpaired t-test. k, MTDH protein levels between tumor and normal tissues in three PCa patients (single experiment).

a, Validation of editing sites (suggested by RNA-seq) of MTDH by Sanger sequencing in 22Rv1. b, Protein and mRNA level of MTDH between 22Rv1 and VCaP shADAR1 and shNC cells. n = 9 cell cultures from three independent experiments. c, Protein level of MTDH in 22Rv1 shNC and shADAR2 cells. d, Sanger sequencing determines MTDH editing level in 22Rv1 shNC and shADAR2 cells. e, Relative luciferase activity of MTDH WT and MUT 3′ UTR reporters in DU-145 or 22Rv1 cells. Renilla activity was used as internal control. n = 9 cell cultures from three independent experiments. f, Immunoblotting analysis of indicated proteins on 22Rv1 shADAR1 and shNC cells treated with 1 μM C16 for 24 h (Representative pictures from 2 independent experiments). g, MTDH was overexpressed in VCaP shADAR1 cells as examined by immunoblot (single experiment). h, Cell viability of control cells and VCaP shADAR1 cells overexpressed EV or MTDH during a 5-day course. n = 9 cell cultures from three independent experiments. i, MTDH was overexpressed in 22Rv1 shADAR1 cells as examined by immunoblot (single experiment). j, Cell viability of control cells and 22Rv1 shADAR1 cells overexpressed EV or MTDH during a 5-day course. n = 9 cell cultures from three independent experiments. k,l, Representative images and quantification of migrated and invaded cells of control (shNC + EV) cells and VCaP (k) or 22Rv1 (l) shADAR1 cells overexpressed EV or MTDH. Scale bar, 200 μm. n = 9 cell cultures from three independent experiments. m, Effect of 8-Br-Ado on Inhibition of catalytic activity of human full-length ADAR1 protein (purchased from Euprotein, EP8230354). n = 3 independent biological replicates, assessed as one experiment. n, Cellular activity of 8-Br-Ado against DU-145 cell. n = 3 independent cell culture replicates, assessed as one experiment. o, Predicted binding mode of 8-Br-Ado with the homology modeling of ADAR1 protein. p, The change curve of RMSD value of docking model of ZYS-1 with ADAR1 in the process of 100 ns molecular dynamics. q, SPR sensorgram showing that ZYS-1 led to changes in resonance units (RU) in a concentration-dependent manner. r, Confirming ADAR1 K1003A mutation through Sanger sequencing and western blotting (single experiment). For b, e, h, and j–n, data represent the mean ± SD. For b, e, h, and j–l, statistical analysis was performed using two-tailed unpaired t-test.

a, Histogram of various editing events in transcriptome-wide of DU-145 treated with DMSO or ZYS-1 (2 μM, 48 h). The right graphs show Alu editing index (AEI) comparison between DMSO- or ZYS-1-treated DU-145 samples. Data were calculated from two independent RNA-seq samples per group, assessed as one experiment). b, Distribution of A-to-I editing sites within Alu, L1, ERV1, other repeat, and no repeat was shown in ZYS-1-treated DU-145 samples. c, GSEA analysis of the shared pathways between ADAR1 KD and ZYS-1-treated DU-145 cells. Statistical analysis was performed using two-sided permutation test, and P values were adjusted by Benjamini-Hochberg method. d, Immunoblot of indicated proteins in VCaP cells after treatment with DMSO or ZYS-1 for 48 h (single experiment). e, IFN-β release in culture supernatant of 22Rv1 treated with DMSO, ZYS-1, or a combination of ZYS-1 (2 μM) and zVADfmk (50 μM) for 48 h. n = 9 cell cultures from three independent experiments. f, Cell viability (left) and IFN-β release (right) of 22Rv1 treated with ZYS-1 (2 μM) also with depletion of MDA5, RIG-I, or PKR. For IFN-β release assay, 22Rv1 was pre-treated with IFN-γ (100 ng/mL) for 24 h. n = 18 cell cultures from three independent experiments in cell viability assay. n = 9 cell cultures from three independent experiments in IFN-β release assay. g, ISG expression of 22Rv1 treated with ZYS-1 (2 μM) also with depletion of MDA5, RIG-I, or PKR. 22Rv1 was pre-treated with IFN-γ (100 ng/mL) for 24 h. n = 9 cell cultures from three independent experiments. h, MTDH protein and mRNA level changes in VCaP cells after treatment with DMSO or ZYS-1 for 48 h. n = 9 cell cultures from three independent experiments. i, Sanger sequencing showing editing level changes on IFNAR1 in DU-145 cells after treatment with 2 μM ZYS-1 for 48 h. j, Cell viability of DU-145 MTDH overexpressed cells after treatment with ZYS-1 for 48 h, detected by CCK-8 assay (OD450 nm). n = 9 cell cultures from three independent experiments. k, Apoptosis of HEK-293 cells treated with DMSO or ZYS-1 for 48 h determined by flow cytometry. n = 9 cell cultures from three independent experiments. l, GSEA shows the upregulated or downregulated pathways in HEK-293 cells upon treatment with ZYS-1 (2 μM, 48 h). m, Histogram of various editing events in transcriptome-wide of HEK-293 treated with DMSO or ZYS-1 (2 μM, 48 h). The right graph shows Alu editing index (AEI) comparison between DMSO- or ZYS-1-treated HEK-293 samples. Data were calculated from two independent RNA-seq samples per group, assessed as one experiment. n, Venn gram shows shared genes with decreased A-to-I editing level (difference > 0.1) after ZYS-1 treatment between DU-145 and HEK-293. For e–h, j, and k, data represent the mean ± SD, and statistical analysis was performed using two-tailed unpaired t-test.

a, b, Effect of ZYS-1 on inhibition of catalytic activity of ADA (a) and ADAR2 (b) protein. Data represent the mean ± SD. n = 3 independent biological replicates, assessed as one experiment. c, ADAR2 protein level changes in DU-145 and VCaP cells after treatment with DMSO or ZYS-1 for 48 h (single experiment). ADA protein level changes in DU-145, VCaP, and 22Rv1 cells after treatment with DMSO or ZYS-1 for 48 h (single experiment). d, Methyltransferase dendrogram showing the selectivity of ZYS-1 (1 μM) over the indicated DNA and protein methyltransferases. n = 2 independent biological replicates, assessed as one experiment. e, Bar graph showing the selectivity of ZYS-1 (1 μM) over the indicated DNA (blue bars) and protein (grey bars) methyltransferases. n = 2 independent biological replicates, assessed as one experiment. f, Kinase dendrograms showing the selectivity of ZYS-1 (1 μM) over a panel of lipid kinases and atypical kinases. n = 2 independent biological replicates, assessed as one experiment. g, Bar graphs showing the selectivity of ZYS-1 (1 μM) over a panel of lipid kinases and atypical kinases. n = 2 independent biological replicates, assessed as one experiment.

a, Weights of DU-145 xenograft mice treated with vehicle, 20 or 40 mg/kg ZYS-1 were measured every two days. Data represent the mean ± SD. n = 8 mice in vehicle group, and n = 12 mice in 20 or 40 mg/kg ZYS-1 treatment group. b, Representative images of H&E staining of key organs of DU-145 xenograft model. Scale bar, 500 μm. c, Quantitative PCR analysis of ISGs on DU-145 xenograft tumor. Data represent the mean ± SD (n = 3 tumor samples from one experiment). Statistical analysis was performed using two-tailed unpaired t-test. d, Immunoblot analysis of indicated proteins on DU-145 xenograft tumor. Data were from three mice. e, Images of 22Rv1 xenograft tumor at the end point of experiment. f, Weights of 22Rv1 xenograft tumor at the end point of experiment. Data represent the mean ± SD (n = 5 mice in each group). Statistical analysis was performed using two-tailed unpaired t-test. g, Representative images of H&E and Ki67 staining of tumors from three mice in 22Rv1 xenograft model, and quantification of Ki67 staining positive cells. Scale bar, 100 μm. Data represent the mean ± SD (n = 3 tumor samples from one experiment). Statistical analysis was performed using two-tailed unpaired t-test. h, Quantitative PCR analysis of ISGs in 22Rv1 xenograft tumor. Data represent the mean ± SD (n = 3 tumor samples from one experiment). Statistical analysis was performed using two-tailed unpaired t-test. i, Immunoblot analysis of indicated proteins on 22Rv1 xenograft tumor. Data were from three mice. j, Body weights of 22Rv1 xenograft mouse during treatment. Data represent the mean ± SD (n = 5 mice in each group). k, Representative images of H&E staining of key organs in 22Rv1 xenograft model. Scale bar, 100 μm. l, Quantitative PCR analysis of ISGs in blood, muscle, cerebral cortex and skin from 22Rv1 xenograft mice. Data represent the mean ± SD (n = 3 tumor samples from one experiment). Statistical analysis was performed using two-tailed unpaired t-test.

a, Images of PDX mice tumor at the end point of experiment. b, Weights of PDX xenograft tumor at the end point of experiment. Data represent the mean ± SD (n = 5 mice in each group). Statistical analysis was performed using two-tailed unpaired t-test. c, Representative images of H&E and Ki67 staining of paraffin section of tumor from three PDX mice, and quantification of Ki67 staining positive cells. Scale bar, 100 μm. Data represent the mean ± SD (n = 3 tumor samples from one experiment). Statistical analysis was performed using two-tailed unpaired t-test. d, Quantitative PCR analysis of ISGs in PDX tumor. Data represent the mean ± SD (n = 3 tumor samples from one experiment). Statistical analysis was performed using two-tailed unpaired t-test. e, Immunoblot analysis of indicated proteins in PDX tumor. Data were from three mice. f, Representative images of H&E staining of key organs of PDX mice. Scale bar, 100 μm. g, Quantitative PCR analysis of ISGs in blood, muscle, cerebral cortex, and skin from PDX model mouse. Data represent the mean ± SD (n = 3 tumor samples from one experiment). Statistical analysis was performed using two-tailed unpaired t-test.

a, b, Wild-type ICR mice were intraperitoneally administrated with various doses of ZYS-1 for observing acute toxicities in two weeks. Each dose group contained 10 mice. Two mice died in 105 mg/kg group, and 5 mice died in 205 mg/kg group. a, Mice were weighed every two days and organs were weighted after euthanizing mice. Data represent the mean ± SD (n = 10 mice in vehicle group, n = 8 mice in 105 mg/kg group, n = 5 mice in 205 mg/kg group). b, Representative images of H&E staining of key organs of ICR mice. Scale bar, 100 μm. c, IC₅₀ value of ZYS-1 against cell viability of RM-1. Data represent the mean ± SD. n = 3 independent cell culture replicates, assessed as one experiment. d, ADAR1 protein level changes in RM-1 cells after treatment with ZYS-1 for 48 h (single experiment). e, Sanger sequencing showing A-to-I editing level changes on Fam193a in RM-1 cells after treatment with 1 or 2 μM ZYS-1 for 48 h. f-h, Male C57BL/6 J mice were implanted subcutaneously with RM-1 cells. When tumors grew about 80-100 mm³, the mice were injected intraperitoneally with vehicle (DMSO and IgG2a isotype control), anti-PD-1 antibody (dosing on days 4, 8, and 12), ZYS-1 (80 mg/kg/day), or a combination of anti-PD-1 and ZYS-1. Each group has 8 mice. f,g, Tumor volumes and body weights were measured every two days. At the end point of experiment, mice were euthanized, then tumors were stripped and weighed. Data represent the mean ± SD (n = 8 mice in each group). h, Representative images of H&E staining of key organs of C57BL/6 J mice. Scale bar, 200 μm. i, Quantitative PCR analysis of ISGs in RM-1 xenograft tumor. Data represent the mean ± SD (n = 3 tumor samples from one experiment). Statistical analysis was performed using two-tailed unpaired t-test.