Fig 2: TET2 and HDAC4 expression is positively correlated in high-risk MDS/AML. (A, B) Microarray expression data from total BM of MDS (A, n = 206) or mononuclear AML (B, n = 96) cells. P value was calculated by Pearson r correlation. (C) HDAC4 expression data from total BM from healthy donors (n = 73) or normal karyotype AML patients (n = 351). P value was calculated by unpaired t test. (D) Overall survival in 163 normal karyotype AML patients based on clustering of HDAC4 expression. Patients were separated into 3 expression groups—low (n = 61), middle (n = 69) and high (n = 33)—based on a standard K-means clustering approach. Kaplan-Meier curves were generated and log-rank p-values calculated to assess statistical significance.

Fig 3: Expression of mineralization-associated HDAC, HDAC4 and HDAC5, in the dentin–pulp complex increased during human-DPC mineralization and after MMP13-loss. (A) Time-course human-DPC-culture indicates a progressive significant increase in both Hdac4 and 5 expressions during mineralization. (B) IHC-HDAC4 expression decreases in adults compared with developing teeth (a–f). HDAC4 demonstrated high expression at post-natal 3-day WT samples in the odontoblast layer (a, b) compared with unstained control (ai—inset). In adult samples, reduced expression of HDAC4 (c–f) was demonstrated in pulp, odontoblast (e, f), and predentin, with increased expression in mineralized dentin. Higher expression in dentin (d, f—black arrows) and odontoblast (f) was evident in Mmp13 −/− samples (black arrows). High HDAC4 expression in adult alveolar bone (c- black star; ci inset—white star) was evident. ai and di represent unstained controls of a and d, while ci is a 10× section of a different area of the same tooth illustrated in C. Scale bars = (a, ai, c, ci, d, di) 100 μm (original mag. ×10) and (b, e, f) 5 μm (original mag. ×63). (C) HDAC5 expression increased in Mmp13 −/− teeth. HDAC5 expression was low at 3-week in WT and MMP13KO teeth but high in bony and mineralized pulpal deposits evident in Mmp13 −/− samples [(b), small white arrow] and alveolar bone [(b), large white arrow]. Dentin–pulp HDAC5 expression increased at 6-week (c, d) and considerably at 3-month in the pulp, odontoblast, and predentin area of Mmp13 −/− samples (d–f, black arrows). (b–f) represent higher magnifications of the indicated area in (bi–fi), respectively. Scale bars = (a, bi) 50 μm (original mag. ×10), (b, ci, di, ei, fi) 20 µm (original mag. ×20), and (c, d) 10 μm (e, f) 5 μm (original mag. x 63). Abbreviations: Am, ameloblast; Bn, alveolar bone; Dn, dentin; Pp, pulp. Sections were examined from five mice at 3-day, six mice at 10-day postnatal and nine mice at 3 months for both genotypes. Hdac5 expression is significantly increased at 3-month in DPC cultures (D) from MMP13KO molars by RT-qPCR. (E) WB analysis of HDAC4 expression in proliferating mouse DPC cells was similar in MMP13KO and WT but decreased in WT differentiating DPC samples, while HDAC5 increased in differentiating samples and all KO samples supporting data in (Ca–f). HDAC6 expression level was similar in WT and MMP13KO but reduced in differentiating DPCs. All qRT-PCR and WB data shown are based on at least three independent biological samples carried out in triplicate. qRT-PCR data are shown as mean ± s.d., *p < 0.05 vs. control.

Fig 4: HDAC4 is required to maintain TET2 activity. (A, B) HDAC4 was overexpressed in 293T cells and TET2 expression determined by RT-qPCR (A) or Western blot (B). (C, D) After HDAC4 knockdown using lentivirus to deliver shRNA to MDS-L cells, 5hmC levels were determined by ELISA (C) or dot blot (D). (E) hMeDIP-qPCR analysis of specific 5hmC enrichment on Chr15: 58979469-58979554, enhancer region of MTSS1 in c-kit+ cells from BM of NHD13/HDAC4 WT or NHD13/HDAC4 KO mice. Bars represent mean enrichment over input. (F) MTSS1 mRNA levels in c-kit+ cells from BM of NHD13/HDAC4 WT or NHD13/HDAC4 KO mice. (G) RT-qPCR detection of MTSS1 mRNA levels in c-kit+ cells from BM of NHD13 mice treated with SAHA or vehicle control. (H) Serial replating of NHD13/HDAC4 WT or NHD13/HDAC4 KO cells. Total BM cells from each these mice were seeded in methylcellulose medium, and colonies were counted on day 7.

Fig 5: circKEAP1 regulate KEAP1 signaling pathway by sponging miR-141-3p. (A) Luciferase activity in A549 cells co-transfected with miRNA mimics or circKEAP1 overepression plasmid and luciferase reporter plasmid which have putative binding site of KEAP1 to miR-141-3p. Luciferase activity was normalized by Renila luciferase activity. (B) Western blot analysis of KEAP1, NRF2 and HDAC4 levels in A549 cells co-transfected with mimic nc or miR-141-3p mimic or circKEAP1 overexpression plasmid. (C) The expression level of miR-1 and miR-206 in A549 cells co-transfected with mimic nc or miR-141-3p mimic or circKEAP1 overexpression plasmid. (D) Cell proliferation analysis for A549 cells co-transfected with mimic nc or miR-141-3p mimic or circKEAP1 overexpression plasmid by CCK-8 assay. (E) Cell proliferation analysis for A549 cells co-transfected with mimic nc or miR-141-3p mimic or circKEAP1 overexpression plasmid. (F) Cell migration analysis for A549 cells co-transfected with mimic nc or miR-141-3p mimic or circKEAP1 overexpression plasmid by Transwell assay. (G) Cell viability analysis for A549 cells co-transfected with mimic nc or miR-141-3p mimic or circKEAP1 overexpression plasmid MTT assay. Data are shown as the means ± standard error of the mean (n = 3), statistical analysis was performed by two-tailed Student’s t test; **P < 0.01, ***P < 0.001.