Voltar
Artigo Acesso aberto Revisado por pares
BIBTEX RIS

Lysine demethylase KDM1A promotes cell growth via FKBP8–BCL2 axis in hepatocellular carcinoma

2022; Elsevier BV; Volume: 298; Issue: 9 Linguagem: Inglês

10.1016/j.jbc.2022.102374

ISSN

1083-351X

Autores

Suli Lv, Xuefeng Zhao, Erlei Zhang, Yingying Yan, Xianyun Ma, Neng Li, Qingli Zou, Lidong Sun, Tanjing Song,

Tópico(s)

Cancer-related gene regulation

Resumo

Advanced hepatocellular carcinoma (HCC) has a dismal prognosis. KDM1A (lysine demethylase 1A), overexpressed in multiple cancer types, is a lysine demethylase that targets both histone and nonhistone proteins. However, it is unclear how KDM1A expression affects HCC etiology. Here, we show that KDM1A can interact with and demethylate FKBP8 (FKBP prolyl isomerase 8), a cytoplasmic protein that regulates cell survival through the antiapoptotic protein BCL2 (B-cell lymphoma-2). We show that demethylation of FKBP8 enhances its ability to stabilize BCL2. Consistently, we observed positive correlation between KDM1A and BCL2 protein levels in liver cancer patients. Functionally, we reveal that FKBP8 demethylation by KDM1A is critical for liver cancer cell growth in vitro and in vivo. We went on to explore the mechanisms that might regulate KDM1A cytoplasmic localization. We found that the cytoplasmic localization and protein stability of KDM1A were promoted by acetylation at lysine-117 by the acetyl transferase KAT8 (lysine acetyltransferase 8). In agreement with this, we show that KDM1A–K117 (lysine 117) acetylation promotes demethylation of FKBP8 and level of BCL2. Finally, it has been shown that the efficacy of sorafenib, a first-line treatment for advanced HCC, is limited by clinical resistance. We show that KDM1A and BCL2 protein levels are increased during acquired sorafenib resistance, whereas inhibiting KDM1A can antagonize sorafenib resistance. Collectively, these results define a functional KDM1A–FKBP8–BCL2 axis in HCC. Advanced hepatocellular carcinoma (HCC) has a dismal prognosis. KDM1A (lysine demethylase 1A), overexpressed in multiple cancer types, is a lysine demethylase that targets both histone and nonhistone proteins. However, it is unclear how KDM1A expression affects HCC etiology. Here, we show that KDM1A can interact with and demethylate FKBP8 (FKBP prolyl isomerase 8), a cytoplasmic protein that regulates cell survival through the antiapoptotic protein BCL2 (B-cell lymphoma-2). We show that demethylation of FKBP8 enhances its ability to stabilize BCL2. Consistently, we observed positive correlation between KDM1A and BCL2 protein levels in liver cancer patients. Functionally, we reveal that FKBP8 demethylation by KDM1A is critical for liver cancer cell growth in vitro and in vivo. We went on to explore the mechanisms that might regulate KDM1A cytoplasmic localization. We found that the cytoplasmic localization and protein stability of KDM1A were promoted by acetylation at lysine-117 by the acetyl transferase KAT8 (lysine acetyltransferase 8). In agreement with this, we show that KDM1A–K117 (lysine 117) acetylation promotes demethylation of FKBP8 and level of BCL2. Finally, it has been shown that the efficacy of sorafenib, a first-line treatment for advanced HCC, is limited by clinical resistance. We show that KDM1A and BCL2 protein levels are increased during acquired sorafenib resistance, whereas inhibiting KDM1A can antagonize sorafenib resistance. Collectively, these results define a functional KDM1A–FKBP8–BCL2 axis in HCC. Hepatocellular carcinoma (HCC) is one of the major causes to cancer-related mortality, with over 800,000 new cases and mortalities each year (1Sung H. Ferlay J. Siegel R.L. Laversanne M. Soerjomataram I. Jemal A. et al.Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.CA Cancer J. Clin. 2021; 71: 209-249Crossref PubMed Scopus (17955) Google Scholar). Advanced HCC has dismal prognosis with half-survival time at only about 1 year. HCC is notoriously refractory to most conventional chemotherapy. Sorafenib has been a first-line target therapy for advanced HCC (2Llovet J.M. Ricci S. Mazzaferro V. Hilgard P. Gane E. Blanc J.F. et al.Sorafenib in advanced hepatocellular carcinoma.N. Engl. J. Med. 2008; 359: 378-390Crossref PubMed Scopus (8751) Google Scholar, 3Demir T. Lee S.S. Kaseb A.O. Systemic therapy of liver cancer.Adv. Cancer Res. 2021; 149: 257-294Crossref PubMed Scopus (13) Google Scholar). Sorafenib works as a multitarget kinase inhibitor, which blocks important pathways including Raf (the rapidly accelerated fibrosarcoma)/mitogen-activated protein kinase, c-Kit, vascular endothelial growth factor receptor and platelet-derived growth factor receptor, leading to decreased tumor angiogenesis, tumor cell proliferation, and increased tumor cell apoptosis (4Zhu Y.J. Zheng B. Wang H.Y. Chen L. New knowledge of the mechanisms of sorafenib resistance in liver cancer.Acta Pharmacol. Sin. 2017; 38: 614-622Crossref PubMed Scopus (342) Google Scholar). Yet, efficacy of sorafenib is limited by clinical resistance because of primary or acquired mechanisms, such as activation of epidermal growth factor receptor, PI3K/AKT, hypoxia-inducible factors, stress-coping mechanisms, and increased cancer stem cell populations (4Zhu Y.J. Zheng B. Wang H.Y. Chen L. New knowledge of the mechanisms of sorafenib resistance in liver cancer.Acta Pharmacol. Sin. 2017; 38: 614-622Crossref PubMed Scopus (342) Google Scholar, 5Tang W. Chen Z. Zhang W. Cheng Y. Zhang B. Wu F. et al.The mechanisms of sorafenib resistance in hepatocellular carcinoma: theoretical basis and therapeutic aspects.Signal. Transduct. Target Ther. 2020; 5: 87Crossref PubMed Scopus (147) Google Scholar). Multiple epigenetic machineries also contribute to sorafenib resistance, including noncoding RNAs, DNA methylation, and histone modifications (5Tang W. Chen Z. Zhang W. Cheng Y. Zhang B. Wu F. et al.The mechanisms of sorafenib resistance in hepatocellular carcinoma: theoretical basis and therapeutic aspects.Signal. Transduct. Target Ther. 2020; 5: 87Crossref PubMed Scopus (147) Google Scholar, 6Huang M. Chen C. Geng J. Han D. Wang T. Xie T. et al.Targeting KDM1A attenuates Wnt/beta-catenin signaling pathway to eliminate sorafenib-resistant stem-like cells in hepatocellular carcinoma.Cancer Lett. 2017; 398: 12-21Crossref PubMed Scopus (62) Google Scholar). Investigation into HCC biology holds promise for new therapeutic strategies (7Huang A. Yang X.R. Chung W.Y. Dennison A.R. Zhou J. Targeted therapy for hepatocellular carcinoma.Signal. Transduct Target Ther. 2020; 5: 146Crossref PubMed Scopus (137) Google Scholar). KDM1A (lysine demethylase 1A), also known as LSD1 (lysine-specific histone demethylase 1), is the first identified lysine demethylase (8Shi Y. Lan F. Matson C. Mulligan P. Whetstine J.R. Cole P.A. et al.Histone demethylation mediated by the nuclear amine oxidase homolog LSD1.Cell. 2004; 119: 941-953Abstract Full Text Full Text PDF PubMed Scopus (3093) Google Scholar), which can regulate gene expression through demethylating histone (8Shi Y. Lan F. Matson C. Mulligan P. Whetstine J.R. Cole P.A. et al.Histone demethylation mediated by the nuclear amine oxidase homolog LSD1.Cell. 2004; 119: 941-953Abstract Full Text Full Text PDF PubMed Scopus (3093) Google Scholar, 9Metzger E. Wissmann M. Yin N. Muller J.M. Schneider R. Peters A.H. et al.LSD1 demethylates repressive histone marks to promote androgen-receptor-dependent transcription.Nature. 2005; 437: 436-439Crossref PubMed Scopus (1337) Google Scholar). KDM1A can also function through demethylation of nonhistones or function independent of its enzymatic activity (10Gu F. Lin Y. Wang Z. Wu X. Ye Z. Wang Y. et al.Biological roles of LSD1 beyond its demethylase activity.Cell Mol. Life Sci. 2020; 77: 3341-3350Crossref PubMed Scopus (37) Google Scholar, 11Perillo B. Tramontano A. Pezone A. Migliaccio A. LSD1: More than demethylation of histone lysine residues.Exp. Mol. Med. 2020; 52: 1936-1947Crossref PubMed Scopus (34) Google Scholar). KDM1A plays pleiotropic roles in physiology and pathophysiology (12Majello B. Gorini F. Sacca C.D. Amente S. Expanding the role of the histone lysine-specific demethylase LSD1 in cancer.Cancers. 2019; 11: 324Crossref PubMed Scopus (71) Google Scholar, 13Kim D. Kim K.I. Baek S.H. Roles of lysine-specific demethylase 1 (LSD1) in homeostasis and diseases.J. Biomed. Sci. 2021; 28: 41Crossref PubMed Scopus (15) Google Scholar). It is overexpressed in multiple cancers and contributes to cancer malignancy (12Majello B. Gorini F. Sacca C.D. Amente S. Expanding the role of the histone lysine-specific demethylase LSD1 in cancer.Cancers. 2019; 11: 324Crossref PubMed Scopus (71) Google Scholar, 13Kim D. Kim K.I. Baek S.H. Roles of lysine-specific demethylase 1 (LSD1) in homeostasis and diseases.J. Biomed. Sci. 2021; 28: 41Crossref PubMed Scopus (15) Google Scholar), including liver cancer (6Huang M. Chen C. Geng J. Han D. Wang T. Xie T. et al.Targeting KDM1A attenuates Wnt/beta-catenin signaling pathway to eliminate sorafenib-resistant stem-like cells in hepatocellular carcinoma.Cancer Lett. 2017; 398: 12-21Crossref PubMed Scopus (62) Google Scholar, 14Lei Z.J. Wang J. Xiao H.L. Guo Y. Wang T. Li Q. et al.Lysine-specific demethylase 1 promotes the stemness and chemoresistance of Lgr5(+) liver cancer initiating cells by suppressing negative regulators of beta-catenin signaling.Oncogene. 2015; 34: 3188-3198Crossref PubMed Scopus (72) Google Scholar, 15Liu C. Liu L. Chen X. Cheng J. Zhang H. Zhang C. et al.LSD1 stimulates cancer-associated fibroblasts to drive Notch3-dependent self-renewal of liver cancer stem-like cells.Cancer Res. 2018; 78: 938-949Crossref PubMed Scopus (59) Google Scholar). Previous studies showed that inhibiting KDM1A could decrease cell survival through p53 (tumor protein p53) methylation (16Huang J. Sengupta R. Espejo A.B. Lee M.G. Dorsey J.A. Richter M. et al.p53 is regulated by the lysine demethylase LSD1.Nature. 2007; 449: 105-108Crossref PubMed Scopus (593) Google Scholar) and changes in gene transcription (17Sareddy G.R. Viswanadhapalli S. Surapaneni P. Suzuki T. Brenner A. Vadlamudi R.K. Novel KDM1A inhibitors induce differentiation and apoptosis of glioma stem cells via unfolded protein response pathway.Oncogene. 2017; 36: 2423-2434Crossref PubMed Scopus (52) Google Scholar, 18Harris W.J. Huang X. Lynch J.T. Spencer G.J. Hitchin J.R. Li Y. et al.The histone demethylase KDM1A sustains the oncogenic potential of MLL-AF9 leukemia stem cells.Cancer Cell. 2012; 21: 473-487Abstract Full Text Full Text PDF PubMed Scopus (345) Google Scholar, 19Zheng Y.C. Duan Y.C. Ma J.L. Xu R.M. Zi X. Lv W.L. et al.Triazole-dithiocarbamate based selective lysine specific demethylase 1 (LSD1) inactivators inhibit gastric cancer cell growth, invasion, and migration.J. Med. Chem. 2013; 56: 8543-8560Crossref PubMed Scopus (188) Google Scholar). Whether other mechanisms are involved remains elusive. KDM1A mainly localizes to the cell nucleus. N terminus of KDM1A, dispensable for its catalytic activity (20Forneris F. Binda C. Vanoni M.A. Battaglioli E. Mattevi A. Human histone demethylase LSD1 reads the histone code.J. Biol. Chem. 2005; 280: 41360-41365Abstract Full Text Full Text PDF PubMed Scopus (196) Google Scholar), harbors a nuclear localization signal (NLS), deletion of which causes translocation of KDM1A to cytoplasm (21Jin Y. Kim T.Y. Kim M.S. Kim M.A. Park S.H. Jang Y.K. Nuclear import of human histone lysine-specific demethylase LSD1.J. Biochem. 2014; 156: 305-313Crossref PubMed Scopus (11) Google Scholar). Besides, aberrant localization of KDM1A to cytoplasm is associated with pathology (22Engstrom A.K. Walker A.C. Moudgal R.A. Myrick D.A. Kyle S.M. Bai Y. et al.The inhibition of LSD1 via sequestration contributes to tau-mediated neurodegeneration.Proc. Natl. Acad. Sci. U. S. A. 2020; 117: 29133-29143Crossref PubMed Scopus (10) Google Scholar). Yet, it is currently unknown how the translocalization of KDM1A to cytoplasm and its function thereof are regulated.FKBP8 (FKBP prolyl isomerase 8), a member of the proline-isomerase family, plays a role in immunoregulation, cellular autophagy, and apoptosis (23Edlich F. Lucke C. From cell death to viral replication: the diverse functions of the membrane-associated FKBP38.Curr. Opin. Pharmacol. 2011; 11: 348-353Crossref PubMed Scopus (26) Google Scholar, 24Bhujabal Z. Birgisdottir A.B. Sjottem E. Brenne H.B. Overvatn A. Habisov S. et al.FKBP8 recruits LC3A to mediate Parkin-independent mitophagy.EMBO Rep. 2017; 18: 947-961Crossref PubMed Scopus (214) Google Scholar, 25Yoo S.M. Yamashita S.I. Kim H. Na D. Lee H. Kim S.J. et al.FKBP8 LIRL-dependent mitochondrial fragmentation facilitates mitophagy under stress conditions.FASEB J. 2020; 34: 2944-2957Crossref PubMed Scopus (18) Google Scholar). FKBP8 also regulates apoptosis in different cancer types (26Shirane M. Nakayama K.I. Inherent calcineurin inhibitor FKBP38 targets Bcl-2 to mitochondria and inhibits apoptosis.Nat. Cell Biol. 2003; 5: 28-37Crossref PubMed Scopus (253) Google Scholar, 27Choi B.H. Feng L. Yoon H.S. FKBP38 protects Bcl-2 from caspase-dependent degradation.J. Biol. Chem. 2010; 285: 9770-9779Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar). FKBP8 promotes resistance to apoptosis in cancers of epithelial origin (26Shirane M. Nakayama K.I. Inherent calcineurin inhibitor FKBP38 targets Bcl-2 to mitochondria and inhibits apoptosis.Nat. Cell Biol. 2003; 5: 28-37Crossref PubMed Scopus (253) Google Scholar, 28Edlich F. Weiwad M. Erdmann F. Fanghanel J. Jarczowski F. Rahfeld J.U. et al.Bcl-2 regulator FKBP38 is activated by Ca2+/calmodulin.EMBO J. 2005; 24: 2688-2699Crossref PubMed Scopus (113) Google Scholar). Antiapoptotic role of FKBP8 is also supported by in vivo study with genetically modified mouse models (29Wang H.Q. Nakaya Y. Du Z. Yamane T. Shirane M. Kudo T. et al.Interaction of presenilins with FKBP38 promotes apoptosis by reducing mitochondrial Bcl-2.Hum. Mol. Genet. 2005; 14: 1889-1902Crossref PubMed Scopus (77) Google Scholar, 30Wong R.L. Wlodarczyk B.J. Min K.S. Scott M.L. Kartiko S. Yu W. et al.Mouse Fkbp8 activity is required to inhibit cell death and establish dorso-ventral patterning in the posterior neural tube.Hum. Mol. Genet. 2008; 17: 587-601Crossref PubMed Scopus (38) Google Scholar). Mechanistic studies reveal that FKBP8 regulates apoptosis through regulating the stability and localization of BCL2 (B-cell lymphoma-2) (26Shirane M. Nakayama K.I. Inherent calcineurin inhibitor FKBP38 targets Bcl-2 to mitochondria and inhibits apoptosis.Nat. Cell Biol. 2003; 5: 28-37Crossref PubMed Scopus (253) Google Scholar, 27Choi B.H. Feng L. Yoon H.S. FKBP38 protects Bcl-2 from caspase-dependent degradation.J. Biol. Chem. 2010; 285: 9770-9779Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar). The effect on BCL2 relies on the interaction of FKBP8 with and activation by Ca2+/calmodulin (CaM) (28Edlich F. Weiwad M. Erdmann F. Fanghanel J. Jarczowski F. Rahfeld J.U. et al.Bcl-2 regulator FKBP38 is activated by Ca2+/calmodulin.EMBO J. 2005; 24: 2688-2699Crossref PubMed Scopus (113) Google Scholar, 31Edlich F. Maestre-Martinez M. Jarczowski F. Weiwad M. Moutty M.C. Malesevic M. et al.A novel calmodulin-Ca2+ target recognition activates the Bcl-2 regulator FKBP38.J. Biol. Chem. 2007; 282: 36496-36504Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar). Thus, FKBP8 is considered as an important contributor to cellular resistance to chemotherapy (32Choi B.H. Yoon H.S. FKBP38-Bcl-2 interaction: a novel link to chemoresistance.Curr. Opin. Pharmacol. 2011; 11: 354-359Crossref PubMed Scopus (22) Google Scholar). Yet, how function of FKBP8–BCL2 axis is regulated in HCC is not clear.In this study, we found that KDM1A could be acetylated at lysine-117 in the NLS by lysine acetyltransferase 8 (KAT8), increasing cytoplasmic level of KDM1A. Cytoplasmic KDM1A could then interact with and demethylate FKBP8, which enhanced its ability to stabilize BCL2 in HCC cells. We further found that KDM1A–BCl2 axis was increased during acquired resistance to sorafenib, and inhibiting this signal axis could overcome sorafenib resistance.ResultsCytoplasmic KDM1A promotes HCC cell growthAnalysis of proteomic data from 159 pairs of HCC tumor and normal tissues (33Gao Q. Zhu H. Dong L. Shi W. Chen R. Song Z. et al.Integrated proteogenomic characterization of HBV-related hepatocellular carcinoma.Cell. 2019; 179: 561-577Abstract Full Text Full Text PDF PubMed Scopus (293) Google Scholar) showed that KDM1A protein level was increased in cancer compared with normal tissue (Fig. S1A). Survival analysis of the same dataset (Fig. S1B) and The Cancer Genome Atlas (TCGA) dataset (Fig. S1C) showed that higher KDM1A level correlated with poorer prognosis in HCC patients. To examine the role of KDM1A in HCC cell growth, we knocked down KDM1A in HLF human liver cancer cells and detected significant decrease in cell proliferation (Figs. 1A and S1D), which could be ameliorated by expressing exogenous KDM1A (Figs. 1B and S1E). Knocking down KDM1A decreased cell clonogenicity as well (Fig. 1C). Consistently, KDM1A inhibitor treatment also decreased cell proliferation significantly (Fig. S1F). These results showed that KDM1A was required for HCC cell proliferation. To explore potential connection between function of KDM1A and its subcellular localization, we examined localization of KDM1A. A small yet significant portion of KDM1A was detected in the cytoplasm by both subcellular fractionation and cell imaging (Fig. 1, D, E,and F). We next examined whether cytoplasmic KDM1A might affect cell proliferation. In KDM1A-KD (knockdown) cell, we rescue-expressed KDM1A–K114A/R115A mutant, which predominantly localized to the cytoplasm (Fig. S1, G and H) as reported (21Jin Y. Kim T.Y. Kim M.S. Kim M.A. Park S.H. Jang Y.K. Nuclear import of human histone lysine-specific demethylase LSD1.J. Biochem. 2014; 156: 305-313Crossref PubMed Scopus (11) Google Scholar). We found that KDM1A–K114A/R115A (referred to as KDM1A-AA hereafter) significantly rescued defect in cell proliferation (Figs. 1G and S1I) in enzymatic activity–dependent manner (Figs. 1H and S1J). In conclusion, these data showed that KDM1A partially localized to cytoplasm, and cytoplasmic KDM1A could promote HCC cell growth.FKBP8 interacts with KDM1A and mediates cytoplasmic KDM1A functionTo explore the mechanism underlying regulation of cell proliferation by cytoplasmic KDM1A, we performed mass spectrometry (MS) to identify the interactome of KDM1A (Table S2). Interestingly, FKBP8 was identified, which was reported to localize to cytoplasm and regulate apoptosis. Immunostaining of myc-FKBP8 in AD293 cell and HLF confirmed that FKBP8 mainly localized to the cytoplasm (Fig. S2, A and B). Proteomic study (33Gao Q. Zhu H. Dong L. Shi W. Chen R. Song Z. et al.Integrated proteogenomic characterization of HBV-related hepatocellular carcinoma.Cell. 2019; 179: 561-577Abstract Full Text Full Text PDF PubMed Scopus (293) Google Scholar) showed that FKBP8 protein level was also increased in HCC tissue compared with normal liver (Fig. S2C). Meanwhile, TCGA dataset suggested that higher FKBP8 level correlates with poorer survival in liver cancer patients (Fig. S2D). To examine the role of FKBP8 in HCC, we knocked down FKBP8 in HLF and detected significant decrease in cell growth and clonogenesis (Figs. 2A, S2E and 2B) accompanied with apoptosis (Fig. 2C). Next, we confirmed the MS result with coimmunoprecipitation (IP) and found that both exogenous and endogenous FKBP8 interacted with KDM1A (Figs. 2D, S2F and 2E). Domain mapping showed that the C-terminal part of FKBP8 cytoplasmic domain mediated interaction with KDM1A (Figs. 2F and S2G). Importantly, while KDM1A-AA increased cell proliferation in KDM1A-KD cell, such effect was largely abolished by FKBP8-KD (Figs. 2G and S2H). Collectively, these data showed that KDM1A interacted with FKBP8 and indicated FKBP8 was a critical cytoplasmic effector of KDM1A.Figure 2FKBP8 interacts with KDM1A and mediates cytoplasmic KDM1A function. A, same number of HLF-control or FKBP8-KD cells were seeded. Cell proliferation over 5 days was measured with CCK-8. Fold of growth was normalized to that in control cells. Error bars denote standard deviation of six biological replicates. p Value was calculated with one-way ANOVA followed by pair-wise comparison as indicated. B, 500 HLF-control or FKBP8-KD cells were seeded. About 14 days later, cell colonies were stained with crystal violet. C, shown are Western blot (WB) of HLF control and FKBP8-KD cell lysates. D, Myc-FKBP8 and FLAG-KDM1A were cotransfected into 293T cells. Shown are results of co-IP–WB. E, Co-IP–WB for endogenous FKBP8 and KDM1A in HLF cells. F, GST-FKBP8 fragments and FLAG-KDM1A were cotransfected into 293T cells. Shown are results of WB for samples from co-IP and GST-pulldown. Asterisk denotes band from immunoglobulin G (IgG) heavy chain. G, HLF KDM1A-KD cells were treated as indicated. Left, same number of indicated HLF cells were seeded. Cell proliferation over 5 days was measured with CCK-8. Fold of growth was normalized to that in control cells. Error bars denote standard deviation of six biological replicates. Right, WB result for the whole cell lysates. p Value was calculated with one-way ANOVA followed by pair-wise comparison as indicated. CCK-8, Cell Counting Kit-8; co-IP, coimmunoprecipitation; FKBP8, FKBP prolyl isomerase 8; GST, glutathione-S-transferase; KD, knockdown; KDM1A, lysine demethylase 1A.View Large Image Figure ViewerDownload Hi-res image Download (PPT)FKBP8 is demethylated by KDM1AAs the ability of KDM1A to promote cell growth in cytosol was dependent on its enzymatic activity (Fig. 1), we next explored whether FKBP8 was subject to lysine methylation. We tested potential activity of a panel of methyltransferases on FKBP8 and found that SMYD3 (SET and MYND domain–containing 3) increased methylation of FKBP8 (Fig. 3A). Consistently, SMYD3 was partially localized to cytoplasm in both AD293 and HLF cells (Fig. S3, A and B). Only SMYD3 WT but not catalytically inactive mutant could increase FKBP8 methylation (Fig. 3B) in a dosage-dependent manner (Fig. S3C). With in vitro methylation assay, we showed that SMYD3 directly methylated FKBP8 (Fig. 3C). Next, we set out to identify the methylated site. With deletion mutants, we narrowed down the methylation site to region 365 to 388 amino acids (Fig. S3D). Consistently, MS identified K377 to be methylated (Fig. S3E). Indeed, mutation of K377 but not adjacent lysines of FKBP8 to arginine abolished its methylation by SMYD3 (Figs. 3D and S3F). Next, we examined whether KDM1A could demethylate FKBP8. Coexpression with KDM1A-WT but not inactive mutant decreased FKBP8 methylation (Fig. 3E).Consistently, treatment with KDM1A selective inhibitor ORY1001 (TargetMol; catalog no.: T6922) increased FKBP8 methylation (Fig. 3F). In addition, KDM1A-AA demethylated FKBP8 more effectively than its WT counterpart, consistent with the cytoplasmic localization of FKBP8 (Fig. 3G). We then performed in vitro demethylation assay and further confirmed that KDM1A directly demethylated FKBP8 (Fig. 3H). Collectively, these data showed that KDM1A demethylated FKBP8–K377 in the cytoplasm.Figure 3FKBP8 is demethylated by KDM1A. A, FLAG-FKBP8 was cotransfected with Myc-tag methyltransferases into 293T cells. Methylation of FKBP8 was analyzed with IP–WB. (Me-K means methyl-lysine). B, GST-FKBP8 (110–389) was cotransfected with Myc-SMYD3-WT or Myc-SMYD3-DN (Y239F inactive mutant) into 293T cells. Shown are results of WB for GST-pulldown (PD) samples. C, recombinant GST-FKBP8 (110–389) was in vitro methylated with immunopurified FLAG-SMYD3. Reactant was analyzed with WB. D, GST-FKBP8 (110–389), -WT, or -K377R was cotransfected with Myc-SMYD3 into 293T cells. Shown are results of WB for GST-PD samples. E, GST-FKBP8 (110–389), Myc-SMYD3, and Myc-KDM1A were cotransfected into 293T cells. Shown are results of WB for GST-PD samples. F, GST-FKBP8 (110–389) was coexpressed with Myc-SMYD3 in 293T cells. Cells were then treated with 10 μM ORY1001 for 48 h before harvest. Shown are results of IP–WB. G, GST-FKBP8 (110–389) was coexpressed with Myc-SMYD3 and Myc-KDM1A in 293T cells. Cells were analyzed with IP–WB as indicated. H, methylated GST-FKBP8 (110–389) was affinity purified from 293T cells cotransfected with GST-FKBP8 and Myc-SMYD3. GST-FKBP8 was then demethylated in vitro by immunopurified FLAG-KDM1A. Reactant was analyzed with WB. FKBP8, FKBP prolyl isomerase 8; GST, glutathione-S-transferase; IP, immunoprecipitation; KDM1A, lysine demethylase 1A; SMYD3, SET and MYND domain–containing 3; WB, Western blot.View Large Image Figure ViewerDownload Hi-res image Download (PPT)KDM1A regulates BCL2 protein level via FKBP8 demethylationFKBP8 was previously reported to regulate BCL2 protein level (27Choi B.H. Feng L. Yoon H.S. FKBP38 protects Bcl-2 from caspase-dependent degradation.J. Biol. Chem. 2010; 285: 9770-9779Abstract Full Text Full Text PDF PubMed Scopus (34) Google Scholar). Consistently, FKBP8-KD also decreased BCL2 protein level in HLF cells (Fig. S4A). We then examined whether KDM1A might affect FKBP8–BCL2 axis. KDM1A knocking-down or inhibitor treatment decreased the protein level of BCL2 but not FKBP8 (Fig. 4, A and B). We found that expressing KDM1A-AA in KDM1A-KD cell restored BCL2 protein level (Fig. 4C) but not mRNA level (Fig. S4B). To corroborate the connection between KDM1A and BCL2, we analyzed the protein level of KDM1A and BCL2 in 24 pairs of HCC patient samples with Western blot (WB) (Fig. S4C). Besides that BCL2, FKBP8, and KDM1A were increased in tumor compared with normal tissues (Fig. S4, C–F), we also observed significant correlation between KDM1A and BCL2 protein level (Fig. 4D). Next, we examined whether FKBP8 was involved in regulation of BCL2 by KDM1A. Knocking down FKBP8 abolished the effect of KDM1A-AA on BCL2, suggesting that cytoplasmic KDM1A regulated BCL2 through FKBP8 (Fig. 4E). To examine whether KDM1A regulated BCL2 through demethylating FKBP8, we rescue-expressed FKBP8-WT or FKBP8–K377R (referred to as FKBP8–K/R hereafter) in cells with both FKBP8-KD and KDM1A-KD. The result showed FKBP8–K/R more potently restored BCL2 protein level than FKBP8-WT (Fig. 4F). Furthermore, while overexpressing KDM1A-AA increased BCL2 in FKBP8-WT cells, little effect was seen in FKBP8–K/R cells (Fig. 4G). Indeed, FKBP8–K/R was more potent than WT in promoting cell proliferation in vitro in multiple HCC cell lines (Figs. 4H and S4, G–J). We further confirmed that FKBP–K/R was more potent in promoting xenograft tumor growth in nude mice model (Fig. 4I). These results showed that KDM1A upregulated BCL2 through FKBP8 demethylation. We next explored how FKBP8–K377 methylation affected FKBP8 function. Regulation of BCL2 by FKBP8 was reported to depend on its interaction with CaM. We found FKBP8–K/R bound CaM more efficiently than WT, and methylation indeed reduced FKBP8 binding with CaM (Figs. 4J and S4K). Collectively, these data showed that KDM1A increased BCL2 protein level through demethylating FKBP8.Figure 4KDM1A regulates BCL2 protein level via FKBP8 demethylation. A, HLF control or KDM1A-KD cells were analyzed with WB. B, HLF cells were treated with 10 μM GSK2879552 (GSK) or 10 μM ORY1001 for 5 days. Cell lysates were analyzed with WB. C, HLF KDM1A-KD and rescue cells were analyzed with WB. D, scatter plot of KDM1A and BCL2 protein levels in liver cancer patient samples. Protein levels were calculated as log2(density/density of GAPDH). Density of each lane was normalized to HLF on the same blot. E, HLF KDM1A-KD, FKBP8-KD, and rescue cells were analyzed with WB. F, FKBP8 was knocked down in KDM1A-KD HLF cells and then FKBP8-WT or FKBP8–K377R was rescue expressed. Cell lysates were analyzed with WB. G, FKBP8-WT or FKBP8–K377R and/or KDM1A-AA were expressed in KDM1A-KD HLF cells. Cell lysates were analyzed with WB. H, FKBP8 was knocked down in KDM1A-KD HLF cells and then FKBP8-WT or FKBP8–K377R was rescue expressed. Same number of cells were then seeded. Cell proliferation over 5 days were measured with CCK-8. Fold of growth was normalized to that in control cells. Error bars denote standard deviation of six biological replicates. p Value was calculated with one-way ANOVA followed by pair-wise comparison as indicated. I, same cells as in (H) were injected subcutaneously into nude mice. Shown is the growth curve of xenograft. p < 0.05 denotes the comparison between FKBP8-WT and FKBP8–K/R rescue cells. p Value was calculated with two-way ANOVA followed by pair-wise comparison as indicated. J, Myc-FKBP8 was coexpressed with FLAG-CaM in 293T KDM1A-KD cells. Cells were harvested for IP–WB analysis as indicated. BCL2, B-cell lymphoma-2; CaM, calmodulin; CCK-8, Cell Counting Kit-8; FKBP8, FKBP prolyl isomerase 8; IP, immunoprecipitation; KD, knockdown; KDM1A, lysine demethylase 1A; WB, Western blot.View Large Image Figure ViewerDownload Hi-res image Download (PPT)KDM1A–lysine 117 is acetylated by KAT8As FKBP8 was regulated by cytoplasmic KDM1A (Fig. 3G), we next explored what might regulate KDM1A cytoplasmic localization. KDM1A was previously shown to be acetylated (34Nalawansha D.A. Pflum M.K. LSD1 substrate binding and gene expression are affected by HDAC1-mediated deacetylation.ACS Chem. Biol. 2017; 12: 254-264Crossref PubMed Scopus (18) Google Scholar, 35Luo H. Shenoy A.K. Li X. Jin Y. Jin L. Cai Q. et al.MOF acetylates the histone demethylase LSD1 to suppress epithelial-to-mesenchymal transition.Cell Rep. 2016; 15: 2665-2678Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar). We tested the activity of a panel of

Referência(s)