Supplementary Figure 1. Characterization and high expression of Lnc-β-Catm in liver CSCs. (a) Heatmap of differently expressed lncrnas in Liver CSCs (CD13 + CD133 + ) and non-cscs (CD13 - CD133 - ) according to transcriptome analyses. (b) 3 and 5 RACE for full length of lnc-β-catm. The length of lnc-β-catm was 2281 nucleotides verified by sequencing. Black arrowhead denotes the full length of lnc-β-catm. (c) Histogram of lnc-β-catm coding potential analyzed by CPC (left panel), CPAT (middle panel) and PhyloCSF (right panel). HOX transcript antisense RNA (Hotair), X inactivation-specific transcript (XIST) and lnctcf7 serve as control non-coding RNAs. β-actin (ACTB) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) serve as control coding genes. For CPC and phylocsf, scores above 0 indicates coding potential, whereas scores below 0 represent no coding potential. CPAT scores indicate the possibility of coding. (d) Anti-Myc and anti-β-actin Western blots. Samples were shown in left panels. Full length of lnc-β-catm was cloned into an eukaryotic expression vector pcdna4-his-myc B with transcription initiating codon ATG in three expression patterns. T, thymine. (e) Scatter plots (means±s.d.) of nomorlized lnc-β-catm expression levels. ehcc, early HCC; ahcc, advanced HCC. (f, g) Histogram of lnc-β-catm expression levels in liver CSCs (CD13 + CD133 + ) and non-cscs (CD13 - CD133 - ) (f), or in oncospheres and non-spheres (g). Error bars, s.d. (n = 4 cell cultures). Two tailed Student s t-test was used for statistical analysis, **, P < 0.01; ***, P<0.001. (h) Western blots (upper panel) and realtime PCR (lower panel) of Nucleocytoplasmic separation fractions. Error bars, s.d. (n = 4 cell cultures). U1 RNA serves as a positive control for nuclear location. EEA1, endosome antigen 1; H3, histone 3. For b, d, h, uncropped blots and gels can be found in Supplementary Data Set 1.
Supplementary Figure 2 Lnc-β-Catm enhances self-renewal of liver CSCs. (a) Relative lnc-β-catm expression levels of lnc-β-catm and control cells. Error bars, s.d. (n = 3 cell cultures). Two tailed Student s t- test was used for statistical analysis, **, P < 0.01; ***, P<0.001. (b, c) Serial sphere formation (b) and tumor propagation (c) with lnc-β- Catm overexpressing and control cells. Error bars, s.d. (n = 3 cell cultures for b, n = 5 mice for c). Two tailed Student s t-test was used for statistical analysis, *, P < 0.05; **, P<0.01; ***, P<0.001. (d) Tumor-free mice ratios after 3 months tumor formation with lnc-β-catm overexpressing (oelnc) and control (oevec) cells. n = 6 mice for each group.
Supplementary Figure 3 Lnc-β-Catm associates with β-catenin and EZH2. (a) Relative mrna expression levels in lnc-β-catm silenced and control cells (lower panels), 16 nearby genes of lnc-β-catm locus (less
than 2 Mb) (upper panels) were analyzed. Error bars, s.d. (n = 4 cell cultures). Two tailed Student s t-test was used for statistical analysis, **, P < 0.01. (b, c) MS-MS profiles of β-catenin (b) and EZH2 (c), corresponding peptide sequences are listed on the top of the corresponding graphs. (d) Different regions of lnc-β-catm (left panel) were labeled with biotin and incubated with PLC sphere lysates, followed by RNA pulldown assays (right panel). (e) Stem-loop structures of full length (1-2281 nt), segment #6 (1181-1437 nt) and segment #9 (1938-2281 nt) of lnc-β-catm. Predictions were based on minimum free energy (MFE) and partition function. Color scales denote confidence of predictions for each base with shades of red indicating strong confidence (http://rna.tbi.univie.ac.at/). (f) Anti-β-catenin and anti-ezh2 Western blots. Samples were derived from co-immunoprecipitation (co-ip) with PLC spheres. (g, h) Antiβ-catenin, anti-ezh2, anti-β-actin (loading control) and anti-oct4 (serum treated control) Western blots. Samples were immunoprecipitates from spheres (S) and non-spheres (N) (g), or serum treated spheres (h). (i) Intensity profiles along the diagonal from upper left to lower right. Green profiles indicate lnc-β-catm gray value (intensity), red profiles indicate β-catenin intensity, and blue EZH2. (j) Flag-β-catenin truncates (upper panels) overexpressing spheres were established, followed by co-ip assays and Western blots (lower panels). (k) Anti-Myc and anti-his Western blots (right panels) with EZH2 truncates overepressing spheres, as in j. (l, m) Histogram of lnc-β-catm enrichment after RNA immunoprecipitation assays. β-catenin truncates (l) and EZH2 truncates (m) overexpressing PLC spheres were used. Error bars, s.d. (n = 4 independent experiments). Throughout figure, uncropped blots and gels can be found in Supplementary Data Set 1.
Supplementary Figure 4 Characterization of β-catenin methylation. (a) Anti-Methylated lysine, anti-β-catenin, anti-h3 and anti-eea1 Western blots. Samples were nuclear (N) and cytoplasmic (C) fractions of the indicated spheres. EEA1, endosome antigen 1; H3, histone 3. (b) Western blots for β-catenin methylation signals in HCC tumor tissues (T) and peri-tumor tissues (P). (c) Methylation observation of β-catenin in peri-tumor and tumor tissues. β-catenin (green), methylated lysine (red), and EZH2 (blue). Scale bars, 20 μm. Uncropped blots in a and b can be found in Supplementary Data Set 1.
Supplementary Figure 5 β-catenin methylation promotes its stability. (a) Anti-phosphorylated β-catenin (p-β-catenin), anti-β-catenin, anti-ezh2 and anti-β-actin (control) Western blots using EZH2 overexpressing (oeezh2) and control (oevec) spheres. (b) Anti-K48 linkage ubiquitylation (K48-Ub), anti-β-catenin and anti-β-actin (control) Western blots. EZH2 inhibitors (GSK126 and GSK343) treated and control (DMSO) spheres were used for β-catenin immunoprecipitation. (c) Anti-β-catenin and anti-β-actin (control) Western blots of methylated and non-methylated β-catenin supplemented with sphere lysates (left panels). Relative β-catenin levels in the right panel. Error bars, s.d. (n = 3 cell cultures). Throughout figure, uncropped Western blot results can be found in Supplementary Data Set 1.
Supplementary Figure 6 Lnc-β-Catm promotes Wnt signaling by increasing β-catenin stability. (a) Relative expression levels of Wnt-β-catenin target genes in lnc-β-catm silenced and control spheres. (b) Relative β-catenin protein levels in lnc-β-catm depleted spheres and control spheres. Error bars, s.d. (n=3 cell cultures). Two tailed Student s t-test was used for statistical analysis, **, P < 0.01; ***, P < 0.001.(c) Anti-methylated lysine and anti-β-catenin (immunoprecipitation control) Western blots. Lnc-β-Catm overexpressing (oelnc) and control (oevec) HCC primary spheres were used. (d, e) Realtime PCR (d) and Western blots (e) of Wnt-β-catenin target genes in lnc-β-catm KO cells (lnc-β-catm KO) and rescued cells (lnc-β-catm KO+rescueLnc). Error bars, s.d. (n=3 cell cultures). (f-i) Anti-methylated lysine, anti-β-catenin, anti-ubiquitination and anti-actin (control) Western blots for β- catenin methylation (f), phosphorylation (g), ubiquitination (h) and stability (i). Samples were lnc-β-catm KO cells, rescued cells and control cells. For i, relative β-catenin protein levels were calculated and shown in the right panel. Error bars, s.d. (n = 3 cell cultures).
Two tailed Student s t-test was used for statistical analysis, *, P < 0.05; **, P < 0.01; ***, P < 0.001. (j, k) Sphere formation (j) and xenograft tumor growth (k). Samples were lnc-β-catm silenced cells rescued with Wnt-β-catenin target genes (c-myc, Ccnd1, and Pttg1). Scale bar, 500 μm. Error bars, s.d. (n = 3 independent experiments). Two tailed Student s t-test was used for statistical analysis, *, P < 0.05. (i) Sphere formation of lnc-β-catm overexpressed spheres supplemented with Wnt-β-catenin inhibitor WIKI4. Typical images were shown in left panels and sphere formation ratios were calculated (right panels). Error bars, s.d. (n = 3 independent experiments). Two tailed Student s t-test was used for statistical analysis, *, P < 0.05; **, P < 0.01. ns, not significant. Throughout figure, uncropped blots and gels can be found in Supplementary Data Set 1.
Supplementary Figure 7 Lnc-β-Catm plays a predominant role in HCC and liver CSCs. (a, b) Expression levels of EZH2 and Wnt-β-catenin target genes in HCC tumors (a) and metastasis patients (b) derived from Wang s cohort. Data are shown as box-and-whisker plots. Whiskers: 5th and 95th percentiles; Horizontal lines: median levels; Boxes: interquartile range (IQR); upper and lower edges: 75th and 25th percentiles. (c) Kaplan-Meier survival analysis of Wnt-β-catenin target genes. HCC samples were divided into 2 groups according to the indicated gene expression levels. (d) Sphere formation of lnc-β-catm
and lnctcf7 silenced and control HCC primary cells. 31 HCC primary cells were used. *, **, ***, lncrna shrna versus control shrna. #, ##, lnc-β-catm shrna versus lnctcf7 shrna. Error bars, s.d. (n = 3 cell cultures). Two tailed Student s t-test was used for statistical analysis, *, P < 0.05; **, P < 0.01; ***, P < 0.001; #, P < 0.05; ##, P < 0.01. (e, f) Confocal observation with CD133 antibody (e), Oct4 antibody and c-myc antibody (f). Control, lnc-β-catm silenced and lnctcf7 silenced spheres were used. Scale bar, 20 μm.
Supplementary Table 1. Table 1a. Frequencies of tumor initiating cells of lnc-β-catm depletion Cell CSC ratio (95% CI) P value Control shrna (A) 1/3190 (1/8046-1/1264) Lnc shrna#1 (B) 1/31905 (1/80461-1/12651) 7.0E-5(B vs A) Lnc shrna#2 (C) 1/12241 (1/33426-1/4483) 0.029(C vs A) Table 1b. Frequencies of tumor initiating cells of lnc-β-catm overexpression Cell CSC ratio (95% CI) P value oevec (A) 1/3190 (1/8046-1/1264) oelnc-β-catm (B) 1/318(1/804-1/126) 6.9E-5(B vs A) Supplementary Table 1. Tumor initiating cell ratios of lnc-β-catm silenced (a) and overexpression (b) cells. 1 10 5, 1 10 4, 1 10 3, 1 10 2 and 10 sphere cells were subcutaneously implanted into BALB/c nude mice for tumor initiation. Frequencies of tumor initiating cells were calculated using extreme limiting dilution analysis. 95% CI, 95% confidence interval of the estimation; vs, versus. P value less than 0.05 was considered significant.
Supplementary Table 2. Realtime PCR primers used in this study Primers 18S (Forward) 18S (Reverse) actin (Forward) actin (Reverse) Lnc-β-Catm (Forward) Lnc-β-Catm (Reverse) EZH2 (Forward) EZH2 (Reverse) CD13 (Forward) CD13 (Reverse) c-myc (Forward) c-myc (Reverse) Ccnd1 (Forward) Ccnd1 (Reverse) n-myc (Forward) Sequences 5 -AACCCGTTGAACCCCATT-3 5 -CCATCCAATCGGTAGTAGCG-3 5 -TCCATCATGAAGTGTGACGT-3 5 -GAGCAATGATCTTGATCTTCAT-3 5 -GGACAGTGAGCGGTCCAAAT-3 5 -TCCTTGTTCCAGTGAAGCGG-3 5 -AATCAGAGTACATGCGACTGAGA-3 5 -GCTGTATCCTTCGCTGTTTCC-3 5 -GACCAAAGTAAAGCGTGGAATCG-3 5 -TCTCAGCGTCACCCGGTAG-3 5 -GGCTCCTGGCAAAAGGTCA-3 5 -CTGCGTAGTTGTGCTGATGT-3 5 -GCTGCGAAGTGGAAACCATC-3 5 -CCTCCTTCTGCACACATTTGAA-3 5 -TGATCCTCAAACGATGCCTTC-3 n-myc (Reverse) 5 -GGACGCCTCGCTCTTTATCT-3 Pparγ (Forward) 5 -ACCAAAGTGCAATCAAAGTGGA -3 Pparγ (Reverse) 5 -ATGAGGGAGTTGGAAGGCTCT-3 Pttg1 (Forward) Pttg1 (Reverse) DKK1 (Forward) DKK1 (Reverse) PCNXL2 (Forward) PCNXL2 (Reverse) 5 -ACCCGTGTGGTTGCTAAGG-3 5 -ACGTGGTGTTGAAACTTGAGAT-3 5 -CCTTGAACTCGGTTCTCAATTCC-3 5 -CAATGGTCTGGTACTTATTCCCG-3 5 -CCATCCCATAGCACCAGTGTC-3 5 -GGTATTACTGACTTGGTCGTGG-3 NTPCR (Forward) 5 -GATGTCGTCACGTTGTCCG -3 NTPCR (Reverse) 5 -GGCATTCACGTTTTCCAGGT -3 KCNK1 (Forward) 5 -TTCTGGAAACCTTCTGTGAACTC-3 KCNK1 (Reverse) IRF2BP2 (Forward) IRF2BP2 (Reverse) TOMM20 (Forward) TOMM20 (Reverse) RBM34 (Forward) 5 -AGTTGGTCATGCTCTATGATGTG-3 5 -ACACCCATTTTGTGCAGTGC-3 5 -ACTGGGACAATAGACCTCTCC-3 5 -GGTACTGCATCTACTTCGACCG-3 5 -TGGTCTACGCCCTTCTCATATTC-3 5 -TACAGGCTTGGACAGGTCG-3 RBM34 (Reverse) 5 -CGTACACGGGTTGAATCTGGG-3 ARID4B (Forward) 5 -ATGAGCCTCCCTATTTGACAGT-3 ARID4B (Reverse) 5 -GGCCCTTTATGTGGTCATCCT -3 GGPS1 (Forward) GGPS1 (Reverse) TBCE (Forward) TBCE (Reverse) GNG4 (Forward) GNG4 (Reverse) LYST (Forward) LYST (Reverse) NID1 (Forward) 5 -ACAGCATCTATGGAATCCCATCT-3 5 -CAAAAGCTGGCGGGTAAAAAG-3 5 -CAGCGGATGTCATTGGTCGAA-3 5 -TCTACTCCTAACCAGGGTCCT-3 5 -GAGGGCATGTCTAATAACAGCAC-3 5 -AGACCTTGACCCTGTCCATAC-3 5 -GCCTCAGAGCATTTGAAAGCC-3 5 -TTCACATCGTCTGTGCCTTTT-3 5 -TCTACGTCACCACAAATGGCA-3 NID1 (Reverse) 5 -GCGACTGCACCGAATGTTG-3 ERO1LB (Forward) 5 -TTCTGGATGATTGCTTGTGTGA -3 ERO1LB (Reverse) 5 -GGTCGCTTCAGATTAACCTTGT-3
Supplementary Table 3. shrna sequences used in this study Primers shlnc-β-catm#1 shlnc-β-catm#2 shttty16#1 shttty16#2 shlinc00211#1 shlinc00211#2 shlinc00244#1 shlinc00244#2 shflvcr1-as1#1 shflvcr1-as1#2 shncrupar#1 shncrupar#2 shhotair#1 shhotair#2 shcrym-as1#1 shcrym-as1#2 shlinc00221#1 shlinc00221#2 shlinctcf7#1 shlinctcf7#2 shβ-catenin Sequences 5 -GAAGCAAGGAAGAGACATA-3 5 -GGCTGAGATATTCCGAAGA-3 5 -GAGTCACTTTGAGGAGCAT-3 5 -GTGCAAGTGGGAATCTTGA-3 5 -GGGCTATGGAACTATGAGA-3 5 -GCAGCATTTCTGTCCTAAA-3 5 -GCAGAAAGTCAGAAGCACA-3 5 -GGGACAAGCTAAGACATGA-3 5 -GCACCAACACATGTAGTTA-3 5 -GGACCACGTTTCATAAGTT-3 5 -GCAGTAGAATGGCGTAAAC-3 5 -GGATCATGAGGTCAGGAGA-3 5 -GAACGGGAGTACAGAGAGA-3 5 -GAGGAAAAGGGAAAATCTA-3 5 -GGAGCAAGCATTATAGAAC-3 5 -GGATCAGTTTCTTACCTCT-3 5 -GCAGGAGAATCACTTGAAC-3 5 -GGATGAATCTCTGGAGAAG-3 5 -AGCCAACATTGTTGGTTAT-3 5 -CACCTAGGTGCTCACTGAA-3 5 -GCAAGCTCATCATACTGGC-3