1. Volume 143, Issue 3, Pages 811-820.e15 (September 2012)
Histone Deacetylases Activate Hepatocyte Growth Factor Signaling by Repressing MicroRNA-449 in Hepatocellular Carcinoma Cells 
Reena Buurman, Engin Gürlevik, Vera Schäffer, Marlies Eilers, Maria Sandbothe, Hans Kreipe, Ludwig Wilkens, Brigitte Schlegelberger, Florian Kühnel, Britta Skawran 
Gastroenterology 
Volume 143, Issue 3, Pages e15 (September 2012)
DOI: /j.gastro
Copyright © 2012 AGA Institute Terms and Conditions

2. Figure 1
Histone acetylation, apoptosis, and proliferation after TSA treatment. Analysis of (A and C) acetylation, (B and D) apoptosis, and (E) proliferation after treatment for (A and B) 1 hour and (C–E) 24 hours of HCC cell lines HLE, HLF, Huh7, and HepG2, and immortalized cell lines THLE-2 and THLE-3 with 100 ng/mL TSA. Values of acetylation and apoptosis were normalized to proliferation. After 1 hour, acetylation already responded to inhibition of HDACs, but apoptosis remained almost unchanged as did proliferation (data not shown). After 24 hours of treatment, not only acetylation but also proliferation and apoptosis responded to TSA. Asterisks are related to the following P values: * =.01–.05 significant; ** =.001–.01 very significant; *** <.001 extremely significant. (F) Western blot analyses using antibodies against acetylated lysine 5, 8, 12, and 16 of histone H4 compared with actin. In agreement with the functional data, all cell lines showed a notable increase in acetylation after 1 hour of treatment and a more pronounced increase after 24 hours. The immortalized liver cell lines sporadically showed acetylation of H4K5, 8, 12, and 16 already in control lysates without treatment.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

3. Figure 2
Cluster analysis of miRNA microarrays after TSA treatment. Expression of miRNA as determined by microarray analyses in HCC cell lines HLE, HLF, Huh7, and HepG2, and immortalized liver cell lines THLE-2 and THLE-3 after treatment with 100 ng/mL TSA for 1 and 24 hours. Twenty-four hours of treatment revealed 3 miRNAs (hsa-mir-182*, hsa-mir-129, and hsa-mir-449) with strongly differing expression between treated and untreated cells, whereas after 1 hour of TSA treatment no miRNA was altered significantly. The analysis was performed by 2-way ANOVA of treated against untreated cells with a Benjamini–Hochberg correction (P = .05).
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

4. Figure 3
Identification of target genes of miR-449. (A) Venn diagram of potential targets of miR-449. The group of genes with decreased expression in mRNA microarray analyses was intersected first with putative targets from the Sanger database.24 The resulting 102 genes were again intersected with the databases MicroCosm ( and TargetScan.25 Only 8 genes were predicted by all 3 databases to be targets of miR-449 as well as down-regulated after 24-hour TSA treatment. qPCR using the ΔΔCT method confirmed (B) the induction of miR-449 expression and (C) the repression of c-MET after 24-hour TSA treatment. The observed effects were stronger in HCC cell lines (HLE, HLF, HepG2, Huh7) than in immortalized liver cell lines (THLE2, THLE3). Values were normalized to controls. Asterisks are related to the following P values: * =.01–.05 significant; ** =.001–.01 very significant; *** <.001 extremely significant. (D) Western blot analyses of c-MET, p-ERK1/2, and ERK 1/2. After 24-hour TSA treatment, c-MET expression is down-regulated. Downstream, phosphorylation of ERK1/2 clearly is reduced after TSA treatment as well. The amount of ERK 1/2 serves as a control. Again, the effects were stronger in HCC cell lines compared with immortalized liver cell lines.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

5. Figure 4
Histone acetylation, apoptosis, and proliferation as well as expression of miR-449 and c-MET after siRNA transfection against HDAC 1, 2, and 3. Values of (A) acetylation and (B) apoptosis were normalized to the cell count and control. Twelve hours after transfection of the HCC cell line, HLE showed only marginal changes in acetylation and apoptosis. Twenty-four hours of siRNA treatment led to increased acetylation and apoptosis. After 48 hours, acetylation and apoptosis increased substantially. During the first 24 hours, qPCR showed increasing expression of (C) miR-449 and, in parallel, a decreasing expression of (D) c-MET, always compared with controls. After 48 hours, the miR-449 expression was reduced, whereas the effect on the expression of c-MET was no longer detectable. Asterisks are related to the following P values: * =.01–.05 significant; ** =.001–.01 very significant; *** <.001 extremely significant. (E) Western blot analysis of c-MET, p-ERK 1/2, and ERK 1/2, compared with actin. After 24 and 48 hours, the expression of c-MET and the amount of p-ERK1/2 were reduced.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

6. Figure 5
Effects of transfection of miR-449 in vitro and in vivo. qRT-PCR of (A) miR-449 and (B) c-MET after transfection of miR-449 into the HCC cell lines HLE and Huh 7. Efficient transfection of miR-449 was accompanied by c-MET repression. Asterisks are related to the following P values: * =.01–.05 significant; ** =.001–.01 very significant; *** <.001 extremely significant. (C) Western blot analysis confirmed a decreased expression of c-MET as well as the reduction of ERK 1/2 phosphorylation as opposed to AKT phosphorylation. The AKT phosphorylation is not reduced. The total amounts of ERK 1/2 and AKT were the same in all columns. (D) Functional assays showed simultaneously increasing apoptosis and (E) decreasing proliferation. Asterisks are related to the following P values: * =.01–.05 significant; ** =.001–.01 very significant; *** <.001 extremely significant. (F) Tumor size measurements in a Huh 7 xenograft model showed that miR-449 effectively attenuates tumor growth. Xenografts were established by subcutaneous injection of 5 × 106 tumor cells 72 hours after transfection with miR-449 or the negative control. The dashed line represents the growth of miR-449–treated tumors and the solid line represents negative control-treated tumors (mean ± standard deviation; n = 6 mice each group).
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

7. Figure 6
Luciferase assay of c-MET 3'UTR. (A) Two predicted binding sites of miR-449 in c-MET 3'UTR. (B) Analysis of luciferase activity, compared with Renilla activity, after cloning the c-MET 3'UTR into the luciferase reporter vector and cotransfecting it with miR-449 into HEK293 cells. (1) The luciferase activity of the empty vector was set to 100; (2) 34% reduction in luciferase activity by intact 3'UTR; (3 and 4) significantly increased luciferase activity after mutation of a single seed binding site, compared with the intact 3' UTR; and (5) luciferase activity after mutation of both seed binding sites is slightly higher than the initial activity of the empty vector.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

8. Figure 7
Expression of miR-449 and c-MET in primary HCC. qPCR analyses show (A) decreased miR-449 and (B) increased c-MET expression in primary HCC tumors. miR-449 showed minimal expression in tumors with Gleason grading G1. Coincidentally, these tumors show the highest values of expression of c-MET. Asterisks are related to the following P values: * =.01–.05 significant; ** =.001–.01 very significant; *** <.001 extremely significant.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

9. Supplementary Figure 1
Expression of HDAC1, HDAC2, and HDAC3 in primary HCC. qRT-PCR of (A) HDAC1, (B) HDAC2, and (C) HDAC3 of primary HCC differentiated by their grading normalized to TBP. Especially HDAC1 and HDAC3 show an increased expression compared with human reference. Progressing dedifferentiation enhances the expression of HDAC 1 and 2, whereas levels of HDAC3 remain almost unchanged.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

10. Supplementary Figure 2
qRT-PCR of HDAC in the cell lines used. qRT-PCR of (A) HDAC1, (B) HDAC2, and (C) HDAC3 in 4 HCC cell lines (HLE, HLF, Huh7, HepG2) and 2 immortalized liver cell lines (THLE-2, THLE-3) in contrast to a human reference. The HCC cell lines show an overexpression of the different HDACs in comparison with the human reference, whereas the immortalized liver cell lines show an expression level similar to the human reference.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

11. Supplementary Figure 3
Acetylation and apoptosis before treatment with HDAC inhibitors. Analysis of (A) acetylation and (B) apoptosis in the HCC cell lines HLE, HLF, Huh7, HepG2, and the immortalized liver cell lines THLE-2 and THLE-3. All analyses were normalized against the cell count as determined by WST-1 reagent. A 1-way ANOVA with the Tukey multiple comparison post-test was performed (GraphPad Prism version 5.00) to calculate the statistical relevance. Cell lines THLE-2 and THLE-3 show significantly heightened acetylation and apoptosis (P <.001) in comparison with the cancer cell lines examined.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

12. Supplementary Figure 4
Functional assays after 12 hours of treatment with different concentrations of the HDAC inhibitor TSA. Analyses of acetylation, apoptosis, and proliferation in the 6 cell lines (Supplementary Figure 3) treated for 12 hours with (A) 100 ng/μL, (B) 200 ng/μL, (C) 500 ng/μL, (D) 800 ng/μL, and (E) 1000 ng/μL TSA. Acetylation and apoptosis were normalized against the cell count as determined by WST-1 reagent.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

13. Supplementary Figure 4
Functional assays after 12 hours of treatment with different concentrations of the HDAC inhibitor TSA. Analyses of acetylation, apoptosis, and proliferation in the 6 cell lines (Supplementary Figure 3) treated for 12 hours with (A) 100 ng/μL, (B) 200 ng/μL, (C) 500 ng/μL, (D) 800 ng/μL, and (E) 1000 ng/μL TSA. Acetylation and apoptosis were normalized against the cell count as determined by WST-1 reagent.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

14. Supplementary Figure 5
Functional assays after 24 hours of treatment with different concentrations of the HDAC inhibitor TSA. Analysis of acetylation, apoptosis, and proliferation in the 6 cell lines (Supplementary Figure 3) treated for 24 hours with (A) 100 ng/μL, (B) 200 ng/μL, (C) 500 ng/μL, (D) 800 ng/μL, and (E) 1000 ng/μL TSA. Acetylation and apoptosis were normalized against the cell count as determined by WST-1 reagent.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

15. Supplementary Figure 5
Functional assays after 24 hours of treatment with different concentrations of the HDAC inhibitor TSA. Analysis of acetylation, apoptosis, and proliferation in the 6 cell lines (Supplementary Figure 3) treated for 24 hours with (A) 100 ng/μL, (B) 200 ng/μL, (C) 500 ng/μL, (D) 800 ng/μL, and (E) 1000 ng/μL TSA. Acetylation and apoptosis were normalized against the cell count as determined by WST-1 reagent.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

16. Supplementary Figure 6
Functional assays after 36 hours of treatment with different concentrations of the HDAC inhibitor TSA. Analyses of acetylation, apoptosis, and proliferation in the 6 cell lines (Supplementary Figure 3) treated for 36 hours with (A) 100 ng/μL, (B) 200 ng/μL, (C) 500 ng/μL, (D) 800 ng/μL, and (E) 1000 ng/μL TSA. Acetylation and apoptosis were normalized against the cell count as determined by WST-1 reagent.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

17. Supplementary Figure 6
Functional assays after 36 hours of treatment with different concentrations of the HDAC inhibitor TSA. Analyses of acetylation, apoptosis, and proliferation in the 6 cell lines (Supplementary Figure 3) treated for 36 hours with (A) 100 ng/μL, (B) 200 ng/μL, (C) 500 ng/μL, (D) 800 ng/μL, and (E) 1000 ng/μL TSA. Acetylation and apoptosis were normalized against the cell count as determined by WST-1 reagent.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

18. Supplementary Figure 7
HDAC expression after siRNA treatment. (A) HDAC1, (B) HDAC2, and (C) HDAC3 expression in the HCC cell line HLE after 12, 24, and 48 hours after transfection with siRNA as determined by qRT-PCR. The qRT-PCR was calculated by the ΔΔ-CT method. TBP was used as housekeeping gene. The HDAC1–3 expression decreased (>70%) dramatically after siRNA treatment, compared with the control siRNA.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

19. Supplementary Figure 8
Western blot analysis of siRNA knockdown of HDAC1–3 in HLE cells, performed at 3 points in time. The HDAC1–3 expression decreased during the monitored period after siRNA treatment, whereas control siRNAs showed no effect. The amount of control protein β-actin remained constant.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

20. Supplementary Figure 9
Effects of transfection of siRNA against c-MET. qRT-PCR of (A) c-MET after transfection of siRNA against c-MET into the HCC cell line HLE. Efficient transfection of siRNA against c-MET was accompanied by c-MET repression. (B) Western blot analysis confirmed a decreased expression of c-MET as well as the reduction of ERK 1/2 and AKT phosphorylation. The total amounts of ERK 1/2 and AKT were the same in all columns. (C) Functional assays showed simultaneously decreasing apoptosis and (D) increasing proliferation.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions

21. Supplementary Figure 10
Huh7 xenograft mouse model showed that miR-449 effectively attenuates tumor growth. Tumors generated in nude mice after subcutaneous injection of Huh7 cells treated with (A) a negative control RNA and (B) miR-449.
Gastroenterology  , e15DOI: ( /j.gastro )
Copyright © 2012 AGA Institute Terms and Conditions