Advances in the development of histone lysine demethylase inhibitors
Tamara Maes, Elena Carceller, Jordi Salas, Alberto Ortega and Carlos Buesa

The covalent modification of histones is closely associated with regulation of gene transcription. Chromatin modifications have been suggested to represent an epigenetic code that is dynamically ‘written’ and ‘erased’ by specialized proteins, and ‘read’, or interpreted, by proteins that translate the code into gene expression changes. Initially thought to be an irreversible process, histone methylation is now known to be reversed by demethylases, FAD dependent amineoxidases and by iron(II)- alpha-ketoglutarate dependent deoxygenases of the Jumonji family. Altered histone demethylase activities have been associated with human disease, including cancer. The first wave of novel investigational drugs directed against KDM1A has recently entered the clinic, and the first specific inhibitor targeting a Jumonji KDM is advancing in preclinical regulatory studies.
Address
Oryzon Genomics, S.A. Carrer Sant Ferran 74, 08940 Cornella` de Llobregat, Spain

Corresponding author: Maes, Tamara ([email protected])

Introduction
Histones are highly conserved proteins that play a dynamic role in modulating chromatin structure. Cova- lent modification of histones guides chromatin remo- deling and histone methylation [1,2] is among the most relevant modifications. Methylation of lysines or argi- nines by histone methyltransferases (HMTs) can be reversed by lysine or arginine demethylases (KDMs or RDMs).

The KDMs are divided into two families based on sequence conservation and catalytic mechanism. Ly- sine demethylation mediated by FAD dependent amine oxidases (KDM1s) starts with a hydride transfer from the N6-methylgroup of the methylated lysine to

the FAD cofactor, which generates an unstable imine intermediate that is hydrolyzed to liberate formalde- hyde (Figure 1). These enzymes are able to demethyl- ate secondary and tertiary but not quaternary amines, limiting the substrate to mono and dimethylated lysines. The KDM1 enzymes are structurally related to MAO-A and MAO-B, targeted by tranylcypromine (TCP, ParnateTM) and Rasagiline, drugs used to treat depression and Parkinson’s Disease. TCP inhibits the MAOs by binding irreversibly to the FAD co-factor, and compounds derived from TCP have been explored extensively as KDM1 inhibitors.

The second KDM family is formed by JmjC domain containing proteins. These Fe(II) dependent enzymes catalyze the demethylation of mono-, di- and trimethy- lated lysines using 2-oxoglutarate (2-OG) and oxygen, converting the methyl group in the methyllysine to a hydroxymethyl group, which is subsequently released as formaldehyde (Figure 2). This family contains over 30 members and includes the KDM2 to KDM8 subfa- milies, as well as JMJD6.

The JmjC domain containing KDMs have structural similarity to N-methyl DNA and RNA demethylases and to nucleic acid oxygenases. Broad-spectrum inhibi- tors of 2-OG oxygenase like N-oxalylglycine (NOG) and pyridine-2,4-dicarboxylate (2,4-PDCA) are also KDM inhibitors and have served as chemical starting points for the development more specific inhibitors [3,4].

Here, we will review the developmental status and thera- peutic potential of small molecule inhibitors of KDMs. This review does not pretend to be exhaustive, yet to illustrate the different compound classes explored for development and to preferentially focus on ‘chemical probes’ complying criteria as proposed by the Structural Genomics Consortium (URL: http://www.thesgc.org/).

The different KDM’s have been described to demeth- ylate mono-, di- or trimethylated H3K4, H3K9, H3K27, H3K26 and H4K20. Both methylation status and the position within or between genes can affect gene tran- scription levels. While H3K4me2/3 and H3K79me3 are preferentially associated with transcriptional activity, meth- ylation of H3K9me2/3, H3K27me2/3, and H4K20me3 correlate with repression. A general overview of the KDM families and their substrate specificities is provid- ed in Figure 3, substrate preferences at a given lysine are

Histone lysine demethylase inhibitors Maes et al. 53

Figure 1

Proposed mechanism for demethylation of H3K4me2 by KDM1A. (a) Hydride transfer from H3K4me2 onto FAD leads to the formation of iminium cation and FAD reduction, (b) hydration of unstable imine intermediate to the N,O-hemiacetal, (c) H3K4me1 and formaldehyde liberation, (d) FAD re-oxidation by one equivalent of molecular oxygen and (e) hydrogen peroxide liberation and subsequent H3K4me2 incorporation to the catalytic cycle. Center: view from crystal structure 2V1D of KDM1A (grey) with the FAD cofactor (yellow) in complex with Co-REST (magenta).

reflected in the order of appearance in the text, and a selection of inhibitors is represented in Figure 4.

KDM1A
KDM1A demethylates H3K4me2/1. More controversial is its capacity to demethylate H3K9me2/1: it cannot de- methylate H3K9me2 peptides in vitro although this ac- tivity was reported on native H3 in specific chromatin contexts in cells. Other KDMs, possibly JmjC proteins identified in KDM1A complexes, may be required for H3K9 demethylase activity. KDM1A is implicated in cancer and in neurodevelopment [5].

The MAO-inhibitor TCP was shown to be a time de- pendent, mechanism based inactivator of KDM1A [6]. Its clinical use as a KDM1A inhibitor has been proposed but the compound is essentially a MAO inhibitor and it is desirable to eliminate the MAO-A inhibitory capacity, responsible for unwanted and possibly life threatening

food–drug and drug–drug interactions. Many KDM1A inhibitors derived from TCP have been reported [6,7]. ORY-1001 is the most potent KDM1A inhibitor described till date. The compound has an IC50 of 18 nM, over 1000- fold selectivity over the MAOs and KDM1B and sub- nanomolar cellular activity in THP-1 cells. Cells with translocations in the Mixed Lineage Leukemia (MLL) gene are particularly sensitive to ORY-1001, but sensitiv- ity is not limited to these cells. ORY-1001 potently inhibits the leukemia initiation cell potential of acute myeloid leukemia (AML) cell lines and reduces AML tumor growth in vivo [Maes et al., abstract e13543, ASCO Annual meeting. Chicago, IL, May–June 2013]. ORY- 1001 is currently advancing in a Phase I study in acute leukemia. GSK2879552 is also selective although notably less potent in cells and in vivo than ORY-1001 [RG Kruger et al., abstract 3964, 5th ASH Annual Meeting and Expo- sition. New Orleans, LA, December 2013]. GSK2879552 is being tested in a Phase I study in AML and in small cell

54 Cancer

Figure 2

Proposed mechanism for demethylation of H3K9me3 by KDM4A. (a) Water displacement by trimethylated lysine, (b) dioxigen coordination with Fe cofactor, (c) Fe (IV)-intermediate formation and OG oxidative decarboxylation, (d) reaction with the Ne-methyl group, (e) fragmentation and demethylated lysine liberation, (f) succinate and formaldehyde production and (g) 2OG regeneration at the binding site. Center: view from crystal structure 3PDQ of 40-((2-aminoethyl)carbamoyl)-[2,20-bipyridine]-4-carboxylic acid inhibitor bound to KDM4A.

lung cancer. Several bona fide chemical probes derived from TCP are available for KDM1A, including com- pounds OG-L002 and RN-1 [8,9].

Apart from irreversible TCP derived inhibitors, a couple of reversible KDM1A inhibitors have been reported including the benzohydrazide SP2509 (IC50 = 13 nM, AML celI differentiation 500 nM) [10] and the tricyclic pyridine GSK690 (IC50 = 5 nM; cell 380 nM) [Dhanak
D. ASCO Annual Meeting Washington, DC, April 2013]. These compounds are reported to have low nanomolar in vitro activities yet so far, exhibit less cellular or in vivo potency than irreversible inhibitors.

KDM1B
Like KDM1A, KDM1B demethylates H3K4me2/1 [20]. Despite being the closest related homologue of KDM1A in the human genome, the above reported KDM1A inhibitors display excellent selectivity over KDM1B, a protein required to establish maternal imprinting during oogenesis. Vice versa, so far only (1R,2R)-2-phenylcyclopropyl-1-amine·HCl was

reported to have some selectivity for KDM1B (Ki 68 mM) over KDM1A (Ki 506 mM) [6].

KDM2/PHF
KDM2A demethylates H3K36me1/2 and KDM2B H3K36me2 as well as H3K4me3 [22]. KDM2A has been implicated in lung and gastric cancer [11,12]. KDM2B functions as a master regulator of miRNAs that target Polycomb Repressive Complexes (PRCs) and is involved in self-renewal of breast cancer stem cells and pancreatic cancer [13,14].

The catalytic domain of the Plant Homeodomain Finger (PHF) proteins is highly similar to that of the KDM2 proteins yet the PHF proteins recognize H3K4me3 and demethylate H3K9me1/2, H3K27me2 and nucleosomal H4K20me1. PHF8 is required for normal development and mutations in the gene cause X-linked mental retar- dation, but elevated expression of PHF8 has also been shown to be oncogenic in solid cancer [15,16]. PHF2, on the contrary, has been described as a potential tumor suppressor [17,18].

Histone lysine demethylase inhibitors Maes et al. 55

Figure 3

Substrate recognition and demethylation by Histone Lysine Demethylases. Left: Phylogenetic tree of histone lysine demethylases. Right: table reflecting binding and demethylation specificities reported for the different enzymes.

The plant growth regulator daminozide is a KDM inhibi- tor selective for the KDM2/7 subfamily (IC50 values are 0.55, 1.5 and 2.1 mM for PHF8, KDM2A and KDM7A
respectively and exhibits 60-fold selectivity for KDM2/ 7 demethylases over other histone demethylases and 2- oxoglurate oxygenases (IC50 > 100 mM) [19]. Treatment with 2 mM daminozide incremented H3K9 methylation in melanoma cells to similar levels as siRNA mediated KDM7 knockdown [20], reflecting good cell penetration. The substituted triazolopyridine, Compound 35, was identified as a KDM2A and likely KDM7 inhibitor with
a much superior biochemical potency IC50 of 63 nM, 30 selectivity over KDM5C and excellent selectivity over members of the KDM3, 4 and 6 subfamily [21], however cell potency was not reported.

KDM3
KDM3A and KDM3B demethylate H3K9me1/me2 but not H3K9me3. KDM3A is induced by hypoxia and in- volved in tumor growth, migration, invasion and its inhi- bition has anti-angiogenic effects and reduces tumor- associated macrophages [22,23]. KDM3B acts either as a promoter of leukomogenesis or as a potential tumor suppressor in myeloid leukemia. The third member of the subfamily, JMJD1C, does not demethylate H3K9 peptides in vitro, but seems to do so in cells [24,25]. JMJD1C is required for MLL-AF9 leukemia mainte- nance, inhibits the neuronal differentiation of human embryonic stems cells and has been found mutated in

intracranial germline tumors [25–27]. No selective inhi- bitors are available.

KDM4
The KDM4 subfamily demethylates mainly H3K9me2/3 or H3K36me3. It is made up of 5 members that have been implicated in several types of cancer. KDM4A/KDM4B are key in androgen signaling and potential progression factors for prostate cancer [28,29]. KDM4A is a determi- nant for invasiveness and metastasis in squamous cell carcinoma and KDM4B is highly expressed in ER+ breast cancer [30]. KDM4C is amplified and over-expressed in numerous hematological and solid cancers [31]. Members of this family are also required for neural stem cell differentiation and have been identified as master reg- ulators of autophagy [32], a process suspected to play a role in neurodegenerative disorders.

The 6-substituted quinolin-8-ol analog ML324 inhibits KDM4E with an IC50 of 920 nM and reduces target gene expression in cellular model of HSV infection with an IC50 of 10 mM [33]. The selectivity of ML324 over other KDMs of the Jumonji family has not been reported. The 8-hydroxyquinoline derivative SD70 was identified as a KDM4C inhibitor through chemical affinity capture and massively parallel DNA sequencing [34●●]. The com- pound inhibits prostate cell growth in vitro (IC50 5–10 mM) and in vivo (10 mg/kg/day), but the in vitro selectivity profile has not been revealed.

56 Cancer

Figure 4

JmjC domain KDM inhibitors

O O
H N
HO OH
O
COOH

N OH

O
H N
HO 5 N
H
O

SAHA
Pan HDAC/KDM
O

N
OH
ML324
KDM4
O O

H O
N N N

C-70
KDM4,5,6?

H
N N

N N

N
Cl

GSK-J1
KDM 6 (5,4)

NOG
Pan KDM O
2,4-PDCA
Pan KDM
COOH

N

N
H N NH
OH OH O
O N
IOX-1
Pan-KDM SD70
KDM4C

O
H N
HO N
8 O
O N OH

N
O
Daminozide
KDM2/7 OH

Compound 9
KDM7(2)

HO
O
O OH

N
N N H
N N O
N N N N
Compound 35
KDM2(7) JIB-04
Pan KDM

Representation of a selection of histone lysine demethylase inhibitors and non-specific precursors. (a) Inhibitors of the FAD dependent KDM1A enzyme. OG-L002 [8], RN-1 [9], GSK2879552 and ORY-1001 are examples of potent selective irreversible inhibitors that were synthesized using the MAO inhibitor tranylcypromine as a starting point. GSK690 and SP-2509 [10] are selective reversible KDM1A inhibitors. 4SC-202 has dual HDAC and KDM1A inhibitory activity. Compound 3 [50] is a hybrid between tranylcypromine and 5-carboxy-8-hydroxyquinoline (IOX1). Clinical stage compounds and authorized drug are underlined. Compounds marked with * are trans-racemic mixtures. (b) Inhibitors of Iron(II)-alpha- ketoglutarate dependent demethylases of the Jumonji family. Chemical starting points for the design of inhibitors including the 2-OG analog NOG, the rigid 2-OG mimic 2,4-PDCA and hydroxamic acids like SAHA have given rise to different compound families like the 8-hydroxyquinoline derivatives IOX-1 [54], SD70 [34●●], ML324 [33]; the pyridine derivatives Compound 35 [21], JIB-04 [55●●], GSK-J1/4 [40●●]) and C-70 [WO 2014053491]); and the hydroxamic acid derivative Compound 9 [46]. The main targets for each compound are reflected, but currently only the selective KDM1A inhibitors meet the criteria for bona fide chemical probes.

KDM5
The KDM5 subfamily demethylates H3K4me2/3 at the transcription start site of actively transcribed genes. KDM5 proteins stimulate cancer cell proliferation, re- duce the expression of tumor suppressors, promote the development of drug resistance, the survival of tumor- initiating cells and relapse [35]. KDM5C and REST co- occupy the neuron-restrictive silencing elements and suppress the expression of REST target genes [36]. Loss-of-function of this gene causes mental retardation and affects memory in men and mice, but can also be beneficial in neurodegenerative disease with aberrant REST activity like Huntington’s disease [36,37].

EPT-103182 is likely the most advanced KDM5 inhibitor in preclinical development. It inhibits KDM5B in vitro with subnanomolar and a cellular IC50 of 1.8 nM for H3K4me3 accumulation in U2OS cells. EPT-103182

Exhibits 20–50 fold selectivity over KDM4 and 3000 fold over KDM6. Karpas-620 cells are exquisitely sensitive and displayed an IC50 for proliferation of 4 nM, and this potent in vitro activity translated into good in vivo efficacy in several multiple myeloma xenografts [P. Staller, per- sonal communication]. The structure of EPT103182 has not yet been disclosed.

KDM6
KDM6A, KDM6B and KDM6C demethylate H3K27me2/3 [38]. These genes play a central role in posterior development by regulating HOX gene expres- sion [39]. KDM6B is involved in macrophage differentia- tion in inflammatory response, brainstem glioma, and T- ALL [40●●,41●●,42]. Mutations in KDM6A have been associated with Kabuki syndrome but the gene is also frequently mutated in cancer and appears to act as a tumor suppressor in T-ALL [43] and myeloid malignancies,

Histone lysine demethylase inhibitors Maes et al. 57

castration resistant prostate cancer, bladder transitional cell carcinoma and renal cell carcinoma. However, it is also over-expressed and associated with unfavorable prog- nosis in breast cancer [44].

GSK-J1 was the first potent KDM6 inhibitor described in the literature. The compound inhibits KDM6B in vitro with an IC50 of 60 nM but is not active in cells. Although crucial for in vitro binding, the highly polar carboxylate group of GSK-J1 restricts cellular permeability. Its ethyl ester GSK-J4 has improved cell penetrability and inhibits H3K27 demethylation in cells with an IC50 around 3 mM [40●●]. GSK-J4 has shown effects in glioma xenografts at a dose of 100 mg/kg/day [41●●]. However, an independent report revealed that GSK-J1 is only a 5–10-fold stronger inhibitor of KDM6A/B than KDM5B/C. In cells, the selectivity window is even smaller as GSK-J4 inhibited KDM6B, KDM5B and KDM4C with similar IC50s [45], raising a warning flag against the attribution of its effects to KDM6 inhibition.

KDM7
KDM7A is closely related to the KDM2 and PHF pro- teins and demethylates H3K9me2 and H3K27me2 as well as nucleosomal monomethylated H4K20me1. KDM7A is required for brain development and could act as a tumor suppressor and consequently, the protein is a potential anti-target. The HDAC inhibitor and metal chelator SAHA was the starting point for the synthesis of hydro- xamate-based inhibitors like Compound 9, which inhibits KDM7A and KDM2 with an IC50 of 0.2 and 6.8 mM in biochemical assays. The compound shows good selectivi- ty of the KDM2/7 over KDM4, 5 and 6 subfamilies. Dose dependent H3K27me2 accumulation was observed be- tween 10 and 100 mM and correlated with effects on proliferation in KYSE150 and HeLa cells [46]. While the pharmacological use of selective KDM7 inhibitors is still unclear, proper tool compounds may contribute to the analysis of gene function and definition of selectivity profiles for KDM inhibiting drugs.

KDM8
KDM8 shares homology with the asparaginyl and histidi- nyl hydroxylase FIH-1 and the lysyl hydroxylase JMJD6. Its role as a histone demethylase is still controversial since initial reports describing that KDM8 demethylates H3K36me2, could not be confirmed [47–49].

JMJD6
JMJD6 is not a histone lysine demethylase, instead it has been proposed to demethylate arginines on histones, to act as a lysine C-5 hydroxylase of splicing regulatory proteins [50], or to combine both activities on the ERa receptor.

The protein was reported to be a driver of cellular proliferation, motility and a marker of poor prognosis in

breast cancer, to predict unfavorable outcome in lung adenocarcinoma, and to promote colon carcinogenesis [51,52].

KDM8 and JMJD6 are bound by the broad spectrum inhibitor NOG, but no specific inhibitors are known.

Dual target KDM inhibitors
Finally, several compounds have been designed or found to inhibit multiple targets including at least one KDM in addition to other proteins.

Hybrid KDM1A/JmjC inhibitors were synthesized by coupling the skeleton of TCP with 2-OG competitors. Compound 3 inhibited KDM1A and KDM4C with IC50 < 1.0 and 1.2 mM, and showed 3–30-fold selec- tivity over other KDMs. Compound 3 increased H3K4/ 9 methylation in HeLa cells with IC50 below 10 mM and induced apoptosis in LnCAP cells with IC50 of 25 mM [53].

The HDAC1/2/3 inhibitor 4SC-202 was recently discov- ered to inhibit KDM1A with similar, low micromolar IC50 values as its cognate targets. The compound is currently being investigated in a Phase I trial in patients with hematological tumors, is well-tolerated and demonstrates indications of anti-tumor efficacy [B Tresokow et al., abstract 8559, ASCO Annual Meeting, Chicago, IL, May–June 2014]. Finally, taking advantage of the struc- tural similarity of the targets, ORY-2001 was designed to inhibit both KDM1A and MAOB and is envisaged for treatment of neurodegenerative disease. The compound inhibits both targets with equimolar IC50 values of ap- proximately 100 nM, has an IC50 of 20 nM in THP-1 cells and is currently in preclinical regulatory develop- ment [5].

Conclusions
Research on inhibitors targeting the FAD dependent amine-oxidase KDM1A has yielded two compounds, ORY-1001 and GSK2879552, that have proceeded swiftly to Phase I clinical studies.

The development of potent and selective inhibitors of KDMs of the large Jumonji family has proven more complicated. Targets and anti-targets coincide in the same subfamily, imposing strict selectivity criteria on development. Some analogs of the cofactor 2-OG have proven to be potent inhibitors but, inherent to their structure, present with poor cell penetration; the Achil- les heel of many compounds designed to target Jumonji KDMs. It is unclear if this drawback can be overcome by pro-drug preparation, especially in view of the in vivo behavior of the compounds. Summarized, a clear need remains for better chemical probes targeting the different members of this large enzyme family.

58 Cancer

Conflict of interest statement
Tamara Maes, Elena Carceller, Alberto Ortega, Jordi Salas, Carlos Buesa are employees of Oryzon genomics
S.A. Tamara Maes and Carlos Buesa are also directors and shareholders.

Acknowledgements
This work was partially financed by grants E8159-EUROSTAR EPILETH, FP7-278871-DDPDGENES and IPT-2012-0673-HEMAFARMA to
Oryzon Genomics S.A.

References and recommended reading
Papers of particular interest, published within the period of review, have been highlighted as:

● of special interest
●● of outstanding interest
1. Binda O: On your histone mark, SET, methylate! Epigenetics
2013, 8:457-463.
2. Molina-Serrano D, Schiza V, Kirmizis A: Cross-talk among epigenetic modifications: lessons from histone arginine methylation. Biochem Soc Trans 2013, 41:751-759.
3. Rose NR, Woon ECY, Kingham GL, King ONF, Mecinovic´ J, Clifton IJ, Ng SS, Talib-Hardy J, Oppermann U, McDonough MA et al.: Selective inhibitors of the JMJD2 histone demethylases: combined nondenaturing mass spectrometric screening and crystallographic approaches. J Med Chem 2010, 53:1810-1818.
4. Hopkinson RJ, Tumber A, Yapp C, Chowdhury R, Aik W, Hing- Che K, Li XS, Kristensen JBL, King ONF, Chan MC et al.: 5- Carboxy-8-hydroxyquinoline is a broad spectrum 2- oxoglutarate oxygenase inhibitor which causes iron translocation. Chem Sci 2013, 4:3110-3117.
5. Maes T, Mascaro´ Crusar C, Ortega A, Lunardi S, Ciceri F, Somervaille TCP, Buesa C: KDM1 histone lysine demethylases as targets for treatments of oncological and neurodegenerative disease. Epigenomics 2015. [in press].
6. Binda C, Valente S, Romanenghi M, Pilotto S, Cirilli R, Karytinos A, Ciossani G, Botrugno OA, Forneris F, Tardugno M et al.: Biochemical, structural, and biological evaluation of tranylcypromine derivatives as inhibitors of histone demethylases LSD1 and LSD2. J Am Chem Soc 2010, 132:6827-6833.
7. Benelkebir H, Hodgkinson C, Duriez PJ, Hayden AL, Bulleid RA, Crabb SJ, Packham G, Ganesan A: Enantioselective synthesis of tranylcypromine analogues as lysine demethylase (LSD1) inhibitors. Bioorg Med Chem 2011, 19:3709-3716.
8. Liang Y, Quenelle D, Vogel JL, Mascaro C, Ortega A, Kristie TM: A novel selective LSD1/KDM1A inhibitor epigenetically blocks herpes simplex virus lytic replication and reactivation from latency. mBio 2013, 4:e00558-e612.
9. Neelamegam R, Ricq EL, Malvaez M, Patnaik D, Norton S, Carlin SM, Hill IT, Wood MA, Haggarty SJ, Hooker JM: Brain- penetrant LSD1 inhibitors can block memory consolidation. ACS Chem Neurosci 2012, 3:120-128.
10. Fiskus W, Sharma S, Shah B, Portier BP, Devaraj SG, Liu K, Iyer SP, Bearss D, Bhalla KN: Highly effective combination of LSD1 (KDM1A) antagonist and pan-histone deacetylase inhibitor against human AML cells. Leukemia 2014:2155-2160.
11. Wagner KW, Alam H, Dhar SS, Giri U, Li N, Wei Y, Giri D, Cascone T, Kim JH, Ye Y et al.: KDM2A promotes lung tumorigenesis by epigenetically enhancing ERK1/2 signaling. Clin Invest 2013, 123:5231-5246.
12. Huang Y, Liu Y, Yu L, Chen J, Hou J, Cui L, Ma D, Lu W: Histone demethylase KDM2A promotes tumor cell growth and migration in gastric cancer. Tumour Biol 2015, 36:271-278.
13. Kottakis F, Foltopoulou P, Sanidas I, Keller P, Wronski A, Dake BT, Ezell SA, Shen Z, Naber SP, Hinds PW et al.: NDY1/KDM2B

functions as a master regulator of polycomb complexes and controls self-renewal of breast cancer stem cells. Cancer Res 2014, 74:3935-3946.

14. Tzatsos A, Paskaleva P, Ferrari F, Deshpande V, Stoykova S, Contino G, Wong KK, Lan F, Trojer P, Park PJ et al.: KDM2B promotes pancreatic cancer via polycomb-dependent and - independent transcriptional programs. J Clin Invest 2013, 123:727-739.

15. Sun X, Qiu JJ, Zhu S, Cao B, Sun L, Li S, Li P, Zhang S, Dong S: Oncogenic features of PHF8 histone demethylase in esophageal squamous cell carcinoma. PLoS ONE 2013, 8:e77353 http://dx.doi.org/10.1371/journal.pone.0077353 [eCollection 2013].

16. Shen Y, Pan X, Zhao H: The histone demethylase PHF8 is an oncogenic protein in human non-small cell lung cancer. Biochem Biophys Res Commun 2014, 451:119-125.

17. Lee KH, Park JW, Sung HS, Choi YJ, Kim WH, Lee HS, Chung HJ, Shin HW, Cho CH, Kim TY et al.: PHF2 histone demethylase acts as a tumor suppressor in association with p53 in cancer. Oncogene 2014 http://dx.doi.org/10.1038/onc.2014.219.

18. Ghosh A, Maiti GP, Bandopadhyay MN, Chakraborty J, Biswas J, Roychoudhury S, Panda CK: Inactivation of 9q22.3 tumor suppressor genes predict outcome for patients with head and neck squamous cell carcinoma. Anticancer Res 2013, 33:1215-1220.

19. Rose NR, Woon EC, Tumber A, Walport LJ, Chowdhury R, Li XS, King ON, Lejeune C, Ng SS, Krojer T et al.: Plant growth regulator daminozide is a selective inhibitor of human KDM2/7 histone demethylases. J Med Chem 2012, 55:6639-6643.

20. Tan Y, Tajik A, Chen J, Jia Q, Chowdhury F, Wang L, Chen J, Zhang S, Hong Y, Yi H et al.: Matrix softness regulates plasticity of tumour-repopulating cells via H3K9 demethylation and Sox2 expression. Nat Commun 2014, 5:4619 http://dx.doi.org/ 10.1038/ncomms5619.

21. England KS, Tumber A, Krojer T, Scozzafava G, Ng SS, Daniel M, Szykowska A, Che K, Delft F et al.: Optimisation of a triazolopyridine based histone demethylase inhibitor yields a potent and selective KDM2A (FBXL11) inhibitor. Med Chem Commun 2014, 5:1879-1886.

22. Tee AE, Ling D, Nelson C, Atmadibrata B, Dinger ME, Xu N, Mizukami T, Liu PY, Liu B, Cheung B et al.: The histone demethylase JMJD1A induces cell migration and invasion by up-regulating the expression of the long noncoding RNA MALAT1. Oncotarget 2014, 5:1793-1804.

23. Osawa T, Tsuchida R, Muramatsu M, Shimamura T, Wang F, Suehiro J, Kanki Y, Wada Y, Yuasa Y, Aburatani H et al.: Inhibition of histone demethylase JMJD1A improves anti-angiogenic therapy and reduces tumor-associated macrophages. Cancer Res 2013, 73:3019-3028.

24. Brauchle M, Yao Z, Arora R, Thigale S, Clay I, Inverardi B, Fletcher J, Taslimi P, Acker MG, Gerrits B et al.: Protein complex interactor analysis and differential activity of KDM3 subfamily members towards H3K9 methylation. PLoS ONE 2013, 8:e60549 http://dx.doi.org/10.1371/journal.pone.0060549.

25. Wang J, Park JW, Drissi H, Wang X, Xu RH: Epigenetic regulation of miR-302 by JMJD1C inhibits neural differentiation of human embryonic stem cells. J Biol Chem 2014, 289:2384-2395.
26. Sroczynska P, Cruickshank VA, Bukowski JP, Miyagi S,
Bagger FO, Walfridsson J, Schuster MB, Porse B, Helin K: shRNA screening identifies JMJD1C as being required for leukemia maintenance. Blood 2014, 123:1870-1882.
27. Wang L, Yamaguchi S, Burstein MD, Terashima K, Chang K,
Ng HK, Nakamura H, He Z, Doddapaneni H, Lewis L et al.: Novel somatic and germline mutations in intracranial germ cell tumours. Nature 2014, 511:241-245.

28. Coffey K, Rogerson L, Ryan-Munden C, Alkharaif D, Stockley J, Heer R, Sahadevan K, O’Neill D, Jones D, Darby S et al.: The lysine demethylase, KDM4B, is a key molecule in androgen receptor signalling and turnover. Nucl Acids Res 2013, 41:4433-4446.

Histone lysine demethylase inhibitors Maes et al. 59

29. Chu CH, Wang LY, Hsu KC, Chen CC, Cheng HH, Wang SM, Wu CM, Chen TJ, Li LT, Liu R et al.: KDM4B as a target for prostate cancer: structural analysis and selective inhibition by a novel inhibitor. J Med Chem 2014, 57:5975-5985.

41.
●●

Hashizume R, Andor N, Ihara Y, Lerner R, Gan H, Chen X, Fang D, Huang X, Tom MW, Ngo V et al.: Pharmacologic inhibition of histone demethylation as a therapy for pediatric brainstem glioma. Nat Med 2014, 20:1394-1396.

30. Ding X, Pan H, Li J, Zhong Q, Chen X, Dry SM, Wang CY: Epigenetic activation of AP1 promotes squamous cell carcinoma metastasis. Sci Signal 2013, 6 http://dx.doi.org/ 10.1126/scisignal.2003884 ra28.1-13, S0-15.

31. Uimonen K, Merikallio H, Pa¨ a¨ kko¨ P, Harju T, Mannermaa A, Palvimo J, Kosma VM, Soini Y: GASC1 expression in lung carcinoma is associated with smoking and prognosis of squamous cell carcinoma. Histol Histopathol 2014, 29:797-804.

32. Bernard A, Jin M, Gonza´ lez-Rodrı´guez P, Fu¨ llgrabe J, Delorme- Axford E, Backues SK, Joseph B, Klionsky DJ: Rph1/KDM4 mediates nutrient-limitation signaling that leads to the transcriptional induction of autophagy. Curr Biol 2015 http:// dx.doi.org/10.1016/j.cub.2014.12.049. pii: S0960- 9822(14)01652-2.

33. Rai G, Kawamura A, Tumber A, Liang Y, Vogel JL, Arbuckle JH, Rose NR, Dexheimer TS, Foley TL, King ON et al.: Discovery of ML324, a JMJD2 demethylase inhibitor with demonstrated antiviral activity. Probe reports from the NIH molecular libraries program [Internet] 2012 December 17 [Updated
2013 September 16]. Bethesda, MD: National Center for Biotechnology Information (US); 2010:. http://www.ncbi.nlm.nih. gov/books/NBK169450/.

The authors of this paper used GSK-J4 to illustrate the potential of pharmacological inhibition for treatment of brainstem glioma that harbor the K27M mutation in H3.3, a mutation that drives oncogenic transforma- tion in glioma. K27M inhibits the enhancer of zeste homologue 2 (EZH2) subunit of Polycomb repressive complex 2 (PRC2), leading to extensive global loss of H3K27me3, which apparently can be counteracted by treatment with GSK-J4. In spite of the conceptual fit, it is still premature to attribute all activity in vivo to KDM6 inhibition, and there is ample room for compound improvement.

42. Ntziachristos P, Tsirigos A, Welstead GG, Trimarchi T, Bakogianni S, Xu L, Loizou E, Holmfeldt L, Strikoudis A, King B et al.: Contrasting roles of histone 3 lysine 27 demethylases in acute lymphoblastic leukaemia. Nature 2014, 514:513-517.
43. Van der Meulen J, Sanghvi V, Mavrakis K, Durinck K, Fang F, Matthijssens F, Rondou P, Rosen M, Pieters T, Vandenberghe P et al.: The H3K27me3 demethylase UTX is a gender-specific tumor suppressor in T-cell acute lymphoblastic leukemia. Blood 2015, 125:13-21.
44. Kim JH, Sharma A, Dhar SS, Lee SH, Gu B, Chan CH, Lin HK, Lee MG: UTX and MLL4 coordinately regulate transcriptional programs for cell proliferation and invasiveness in breast cancer cells. Cancer Res 2014, 74:1705-1717.
45. Heinemann B, Nielsen JM, Hudlebusch HR, Lees MJ, Larsen DV,

34.
●●

Jin C, Yang L, Xie M, Lin C, Merkurjev D, Yang JC, Tanasa B, Oh S, Zhang J, Ohgi KA et al.: Chem-seq permits identification of genomic targets of drugs against androgen receptor regulation selected by functional phenotypic screens. Proc Natl Acad Sci U S A 2014, 111:9235-9240.

Boesen T, Labelle M, Gerlach LO, Birk P, Helin K: Inhibition of demethylases by GSK-J1/J4. Nature 2014, 514:E1-E2 http:// dx.doi.org/10.1038/nature13688.
46. Suzuki T, Ozasa H, Itoh Y, Zhan P, Sawada H, Mino K, Walport L, Ohkubo R, Kawamura A, Yonezawa M et al.: Identification of the

The assessment of the specificity of drugs targeting members of the large
Jumonji KDM family is tedious and complicated by the fact that not all activities can necessarily be reproduced by in vitro biochemical assays. Chem-Seq, was used to assess which chromatin binding targets are bound by tagged compound analogues and target genes are subse- quently identified by massive sequencing. This presents an alternative approach to address target identification in a relevant cellular context.

35. Stein J, Majores M, Rohde M, Lim S, Schneider S, Krappe E, Ellinger J, Dietel M, Stephan C, Jung K et al.: KDM5C is overexpressed in prostate cancer and is a prognostic marker for prostate-specific antigen-relapse following radical prostatectomy. Am J Pathol 2014, 184:2430-2437.

36. Aguilar-Valles A, Vaissie` re T, Griggs EM, Mikaelsson MA, Taka´ cs IF, Young EJ, Rumbaugh G, Miller CA:
Methamphetamine-associated memory is regulated by a writer and an eraser of permissive histone methylation. Biol Psychiatry 2014, 76:57-65.

37. Vashishtha M, Ng CW, Yildirim F, Gipson TA, Kratter IH, Bodai L, Song W, Lau A, Labadorf A, Vogel-Ciernia A et al.: Targeting H3K4 trimethylation in Huntington disease. Proc Natl Acad Sci U S A 2013, 110:E3027-E3036 http://dx.doi.org/10.1073/ pnas.1311323110.

38. Walport LJ, Hopkinson RJ, Vollmar M, Madden SK, Gileadi C, Oppermann U, Schofield CJ, Johansson C: Human UTY(KDM6C) is a male-specific Ne-methyl lysyl demethylase. J Biol Chem 2014, 289:18302-18313.

39. Shpargel KB, Starmer J, Yee D, Pohlers M, Magnuson T: KDM6 demethylase independent loss of histone H3 lysine
27 trimethylation during early embryonic development. PLOS Genet 2014, 10:e1004507 http://dx.doi.org/10.1371/ journal.pgen.1004507.

KDM2/7 histone lysine demethylase subfamily inhibitor and its antiproliferative activity. J Med Chem 2013, 56:7222-7231.
47. Hsia DA, Tepper CG, Pochampalli MR, Hsia EY, Izumiya C, Huerta SB, Wright ME, Chen HW, Kung HJ, Izumiya Y: KDM8, a H3K36me2 histone demethylase that acts in the cyclin A1 coding region to regulate cancer cell proliferation. Proc Natl Acad Sci U S A 2010, 107:9671-9676.
48. Ishimura A, Minehata K, Terashima M, Kondoh G, Hara T, Suzuki T: Jmjd5, an H3K36me2 histone demethylase, modulates embryonic cell proliferation through the regulation of Cdkn1a expression. Development 2012, 139:749-759.
49. Del Rizzo PA, Krishnan S, Trievel RC: Crystal structure and functional analysis of JMJD5 indicate an alternate specificity and function. Mol Cell Biol 2012, 32:4044-4052.
50. Webby CJ, Wolf A, Gromak N, Dreger M, Kramer H, Kessler B, Nielsen ML, Schmitz C, Butler DS, Yates et al.: Jmjd6 catalyses lysyl-hydroxylation of U2AF65, a protein associated with RNA splicing. Science 2009, 325:90-93.
51. Zhang J, Ni SS, Zhao WL, Dong XC, Wang JL: High expression of JMJD6 predicts unfavorable survival in lung adenocarcinoma. Tumour Biol 2013, 34:2397-2401.
52. Wang F, He L, Huangyang P, Liang J, Si W, Yan R, Han X, Liu S, Gui B, Li W et al.: JMJD6 promotes colon carcinogenesis through negative regulation of p53 by hydroxylation. PLOS Biol 2014, 12:e1001819 http://dx.doi.org/10.1371/ journal.pbio.1001819.
53. Rotili D, Tomassi S, Conte M, Benedetti R, Tortorici M, Ciossani G, Valente S, Marrocco B, Labella D, Novellino E et al.: Pan-histone demethylase inhibitors simultaneously targeting Jumonji C and lysine-specific demethylases display high anticancer

40.
●●

Kruidenier L, Chung CW, Cheng Z, Liddle J, Che K, Joberty G, Bantscheff M, Bountra C, Bridges A, Diallo H et al.: A selective jumonji H3K27 demethylase inhibitor modulates the proinflammatory macrophage response. Nature 2012, 488:404-408.

activities. J Med Chem 2014, 57:42-55.
54. Rose NR, Ng SS, Mecinovic´ J, Lie´ nard BM, Bello SH, Sun Z, McDonough MA, Oppermann U, Schofield CJ: Inhibitor scaffolds for 2-oxoglutarate-dependent histone lysine demethylases. J Med Chem 2008, 51:7053-7056.

The authors of this paper sought to explore structure-guided small-
molecule and chemoproteomics approach to elucidate the functional role of the H3K27me3-specific demethylase subfamily KDM6 and to generate GSK-J1/J4. Although selectivity was later revealed to be lower than described in this paper, GSK-J4 is still one of the few inhibitors revealed to have cell activity.

55.
●●

Wang L, Chang J, Varghese D, Dellinger M, Kumar S, Best AM, Ruiz J, Bruick R, Pen˜ a-Llopis S, Xu J et al.: A small molecule modulates Jumonji histone demethylase activity and selectively inhibits cancer growth. Nat Commun 2013, 4:2035 http://dx.doi.org/10.1038/ncomms3035.

60 Cancer

The authors identified JIB-04 as an inhibitor of the Jumonji family of histone demethylases JIB-04 in a cell screen. Although the compound is a pan-KDM inhibitor it is interesting because it is not a competitive inhibitor of a-ketoglutarate, because it enters relatively well in cells, and displays selectivity in killing certain tumor cells vs normal cells. The compound was also active in vivo in H358 an A549 xenograft models when administered

i.p. at 110 mg/kg. Especially relevant but not available in the form of a published manuscript: **The most advanced efforts on KDM inhibitors have been reported by Oryzon (ORY-1001), GSK (GSK2879552) and Epitherapeutics (EPT-103182). These advances are cited in the text but they are not yet available in the form of a regular manuscript on the development of the cited compounds.