Fig 2: Aberrant RNA editing re-codes GLI1 transcripts. a Schematic representation of the GLI1 editing site in a putative SUFU binding domain. b Vienna RNA predicted secondary structure changes of GLI1 induced by A-to-I editing in exon 12. c RESSq-PCR analysis of GLI1 editing in primary MM total MNCs. Dots represent ratio of edit (G)/WT (A) GLI1 transcripts (mean values for individual patients ± S.E.M.; ctrl n = 3, smoldering MM n = 4, newly diagnosed MM n = 4, relapsed MM n = 7, PCL n = 4). *p < 0.05, **p < 0.01, by unpaired, two-tailed Student’s t-test. d Representative Sanger sequencing chromatograms for GLI1; the yellow box highlights the double peak A/G, labeled with the percentage of edited transcripts assessed as edit allele burden (%G/(G + A)). e Regression analysis of total ADAR1 mRNA expression and GLI1 editing levels in primary samples (n = 19). f Top 5 gene sets enriched in 1q amp (n = 40) from CoMMpass RNA-seq (IA7 data release) ranked by gene set size. g Venn diagram showing number of core enriched genes between KEGG_Pathways in Cancer and KEGG_Signaling Pathways Regulating Pluripotency of Stem Cells. h Enrichment of Hedgehog signaling in high versus low ADAR1 patients. i GLI1 editing after ADAR1 silencing in 1q-amplified cells (H929). Left: ADAR1 knockdown levels by qPCR (mean ± S.E.M. of three independent experiments). Right: GLI1 editing by RESSq-PCR. *p < 0.05, **p < 0.01 compared to Lenti-shCtrl by unpaired, two-tailed Student’s t-test. j GLI1 editing by RESSq-PCR in normal primary CD34+ cells (n = 3) transduced with lentiviral pCDH ADAR1-WT/ADAR1 Mutant, or pCDH backbone control. Histograms represent relative ratios of edit (G)/WT (A) GLI1 transcripts (mean ± S.E.M. of three independent experiments). **p < 0.01, by unpaired, two-tailed Student’s t-test. k GLI1 editing by RESSq-PCR after transient overexpression of GLI1 and WT or editase-deficient ADAR1 in HEK293T. Histograms represent mean ± S.E.M. of three independent experiments. **p < 0.01, by unpaired, two-tailed Student’s t-test. l Relative GLI1-Luciferase/Renilla reporter activity (mean ± S.E.M.). HEK293T cells were co-transfected with a dual reporter plasmid and GLI1 WT, GLI1 Edit or vector control (backbone) pCDH plasmids (n = 4). *p < 0.05, ***p < 0.001, by one way ANOVA plus Bonferroni post-test. Statistical significance was indicated when p < 0.05. See also Supplementary Fig. 2

Fig 3: Prognostic value of ADAR1 expression in HCC after tumor resection. (a) Representative images for scoring the ADAR1 IHC staining in HCC tumor tissues (left panel) and normal liver tissues (right panel). (b and c) Kaplan–Meier plots for total survival time of patients demonstrating ADAR1-negative (blue line in b; n=25), ADAR1-positive (green line in b; n=58), ADAR1 normal expression or downregulation (blue line in c; n=32) and ADAR1 overexpression (OE) (green line in c; n=51) in tumor tissues (log-rank test). (d and e) Kaplan–Meier plots for disease-free survival time of patients demonstrating ADAR1-negative (blue line in d; n=25), ADAR1-positive (green line in d; n=58), ADAR1 normal expression or downregulation (blue line in e; 32) and ADAR1 overexpression (OE) (green line in e; n=51) in tumor tissue (log-rank test). T, tumor; N, nontumor

Fig 4: Rapamycin partially blocked the effects of ADAR1 overexpression on GC cell growth and migration(A) The effects of ADAR1 overexpression on SGC7901 and MGC803 cell proliferation with or without rapamycin treatment (10 µM), *p < 0.05. (B) Representative photographs of migratory cells on the transwell membrane (magnification, 200×). The below panel is the average MGC803 cell number of triplicate, *p < 0.05.

Fig 5: High ADAR1-expressing cells serially transplant MM in primary patient-derived xenografts. a Schematic diagram showing primary high-risk MM (plasma-cell leukemia) patient-derived mouse model generation and analyses performed. b Representative images of primary MM10-engrafted mice, by in vivo bioluminescence assay. Mice were injected intra-peritoneally (i.p.) with luciferin (150 mg/kg) and luciferase signal was acquired by IVIS within 10 min of injection. c ELISA for human immunoglobulin light chain levels was used to monitor tumor cell growth in vivo over time. d Representative gating strategy on transplanted mouse BM. Upper dot plots show human malignant plasma cells gated out of total single live cells. Cell surface marker expression was determined as CD138+/CD319+/CD38high/CD45dim, as displayed in overlaid plots. Red dots show representative MM9-engrafted mouse BM, green and blue represent positive, and negative staining controls, respectively. e Human CD138/CD319 double positive cell engraftment in serially transplanted recipients (MM9, n = 6). f Human immunoglobulin kappa chain levels in mouse sera from M9 serially transplanted mice (n = 6). g qPCR analyses and GLI1 editing in serial transplants. Left panel: human cell engraftment, determined by human-specific Alu qPCR; center panel: ADAR1 RNA expression levels by qPCR. Right panel: GLI1 editing by RESSq-PCR. Histograms represent mean values ± S.E.M. from individual mice in spleen (SP, purple, n = 3), bone marrow (BM, red n = 4), and plasmacytomas (PC, green, n = 4) from MM9 serially transplanted animals). *p < 0.05, **p < 0.01 by one way ANOVA. h Engraftment of CD138/CD319 double positive cells in secondary recipients of Lentiviral (LV)-shADAR1 versus shCtrl-transduced MM9 cells (n = 7 LV-shCtrl; n = 8 LV-shADAR1). i ELISA for human immunoglobulin light chain levels in secondary recipients of LV-shADAR1 versus shCtrl-transduced MM9 cells (n = 7 LV-shCtrl, n = 8 LV-shADAR1; 7 weeks post intrahepatic transplant). p values by unpaired, two-tailed Student’s t-test. j ADAR1 and human Alu mRNA levels by qPCR in spleen tissue of MM9 secondary LV-shCtrl or shADAR1 mice (LV-shCtrl n = 7; LV-shADAR1 n = 8). **p < 0.01 by unpaired, two-tailed Student’s t-test. Statistical significance was indicated when p < 0.05. See also Supplementary Fig. 3