Rational Design of Bisubstrate-Type Analogues as Inhibitors of DNA

Article pubs.acs.org/jmc

Rational Design of Bisubstrate-Type Analogues as Inhibitors of DNA Methyltransferases in Cancer Cells Ludovic Halby,† Yoann Menon,† Elodie Rilova,† Dany Pechalrieu,† Véronique Masson,† Celine Faux,† Mohamed Amine Bouhlel,§ Marie-Hélène David-Cordonnier,§ Natacha Novosad,† Yannick Aussagues,† Arnaud Samson,† Laurent Lacroix,‡ Fréderic Ausseil,† Laurence Fleury,† Dominique Guianvarc’h,∥ Clotilde Ferroud,⊥ and Paola B. Arimondo*,†,# †

ETaC, Epigenetic Targeting of Cancer, CRDPF, CNRSPierre Fabre USR3388, 3 Avenue H. Curien, 31035 Toulouse cedex 01, France ‡ LBME, CNRS UMR 5099, 31062 Toulouse, France § UMR-S1172-JPARC (Jean-Pierre Aubert Research Center), INSERM, University of Lille, Centre Hospitalier Universitaire de Lille, IRCL, 59045 Lille, France ∥ Laboratoire des BioMolécules, UMR 7203, Université Pierre et Marie CurieParis 6ENSCNRS, 4, place Jussieu, 75252 Paris Cedex 05, France ⊥ Laboratoire de Chimie Moléculaire, CMGPCE, EA7341, Conservatoire National des Arts et Métiers, 2 rue Conté, 75003 Paris, France # Churchill College, CB3 0DS Cambridge, U.K. S Supporting Information *

ABSTRACT: Aberrant DNA hypermethylation of promoter of tumor suppressor genes is commonly observed in cancer, and its inhibition by small molecules is promising for their reactivation. Here we designed bisubstrate analogues-based inhibitors, by mimicking each substrate, the S-adenosyl-Lmethionine and the deoxycytidine, and linking them together. This approach resulted in quinazoline−quinoline derivatives as potent inhibitors of DNMT3A and DNMT1, some showing certain isoform selectivity. We highlighted the importance of (i) the nature and rigidity of the linker between the two moieties for inhibition, as (ii) the presence of the nitrogen on the quinoline group, and (iii) of a hydrophobic group on the quinazoline. The most potent inhibitors induced demethylation of CDKN2A promoter in colon carcinoma HCT116 cells and its reactivation after 7 days of treatment. Furthermore, in a leukemia cell model system, we found a correlation between demethylation of the promoter induced by the treatment, chromatin opening at the promoter, and the reactivation of a reporter gene.

■

using S-adenosyl-L-methionine (AdoMet) as a methyl donor.7 Thus, the enzymes present two substrates: the deoxycytine that will be methylated (in blue in Figure 1), and the AdoMet (in red in Figure 1), the cofactor that donates the methyl group. Alteration of DNA methylation patterns leads to various diseases such as cancer.8 Cancer cells often present aberrant DNA methylation; in particular, a specific hypermethylation of tumor suppressor genes is observed. Restoring their expression by inhibition of DNA methylation represents an attractive therapeutic strategy.9,10 Several DNMT inhibitors have been described.9,11 They are divided into two families: nucleoside and non-nucleoside analogues. The first are the most active ones. Two of them

INTRODUCTION Epigenetic modifications participate in controling gene expression. In mammals, methylation of deoxycytidines in DNA was shown to play a key role.1,2 It is the most stable epigenetic mark and occurs at CpG sites, which are grouped in islands and essentially located in promoters, repeated sequences, and CpG island shores.3 Hypermethylation of promoters induces gene silencing, while hypomethylation is associated with gene expression.4−6 Enzymes responsible for DNA methylation are the DNA methyltransferases (DNMTs). Two families of catalytically active DNMTs have been identified: DNMT1, mainly responsible for DNA methylation maintenance during replication, and DNMT3A and 3B, which perform de novo DNA methylation and support methylation maintenance. DNMTs add a methyl group on the carbon-5 position of the deoxycytidine at the CpG site in the DNA by © 2017 American Chemical Society

Received: February 21, 2017 Published: May 2, 2017 4665

DOI: 10.1021/acs.jmedchem.7b00176 J. Med. Chem. 2017, 60, 4665−4679

Journal of Medicinal Chemistry

Article

Figure 1. (A) Schematic representation of the catalytic pocket of the DNMT1 and DNMT3A and (B) compound 1 and the chemical modulation strategy. In red is depicted the methyl-donor AdoMet and its analogue 4-aminoquinazoline, in blue the deoxycytidine in the DNA and its analogues, and in green the linker between the two moieties.

Scheme 1. Synthesis of 1 and 14a

a (a) (i) SOCl2,110 °C, 0.5 h, (ii) phenylpropylamine, DMF, K2CO3, RT, 2 h, 85%; (b) phthalimide potassium salt, DMF, 90 °C, 6 h, 98%; (c) (i) CH3NHNH2, ethanol, RT, 18 h, 82%, (ii) NosCl, TEA, DMF, RT, 6 h, 96%; (d) 2-(N-Boc-amino)-ethyl bromide, TEA, DMF, 65 °C, 12 h, 92%; (e) TFA, RT, 0.5 h, 96%; (f) MOMCl, BSA, DCM, RT, 18 h, 93%. (g) (i) POCl3, 1,2,4-triazole, TEA, MeCN, 18 h, (ii) 8, TEA, DMF, RT, 6 h, 54% over 2 steps; (h) PhSH, K2CO3, DMF, RT, 12 h, 73%; (i) ethanolamine, 125°C, 4 h, quantitative yield; (j) SOCl2, DMF cat, flash boiling, 91%; (k) 6, K2CO3, KI, DMF, 90 °C, 12 h, 74%; (l) PhSH, K2CO3, DMF, RT, 24 h, 91%.

are FDA approved: 5-azacytidine (5azaC) and 5-aza-2′deoxycytidine (5azadC). Despite their high efficiency, poor

bioavailability, instability in physiologic media, and little selectivity restrict their use. Non-nucleoside analogues present 4666



Article

uracil was functionalized with methoxymethyl group to afford 9, converted to cytosine, and coupled to 8 upon use of triazole/ POCl3 in the presence of triethylamine to provide compound 10 in good yields. 10 was finally treated with thiophenol to give the desired compound 1. Compound 14 was synthesized by coupling quinazoline 6 and quinoline 12, obtained by reaction of 4-chloroquinoline with ethanolamine, to afford 4-(2hydroxyethylamino)quinoline 11, which was then treated with thionyl chloride by flash boiling. Derivative 19, in which a rigid linker arm was introduced, was synthesized starting from 4-aminoquinazoline 18 (Scheme 2).

various structures and mechanisms of action (reviewed in ref 12). Many of them were shown to target the DNMT enzymatic activity, but they lack both specificity and potency. A main goal is to achieve compounds that demethylate gene promoters in cancer cells in a selective manner and results in gene reexpression. Another goal is to develop inhibitors that show some selectivity for one isoform, DNMT1 or DNMT3s. Here, we synthesized bisubstrate-type inhibitors and evaluated their inhibition activity against human DNMT1 and DNMT3A and their activity in cancer cells. Interestingly, quinoline−quinazoline conjugates are able to demethylate gene promoters and induce gene expression. In addition, they constitute a promising new scaffold for the development of selective inhibitors of each isoform of DNMTs.

Scheme 2. Synthesis of 19a

■

RESULTS AND DISCUSSION Rational Design of Bisubstrate Analogues. On the basis of the schematic representation of the transition state of DNMT catalytic site (Figure 1), we first replaced the adenosine of the AdoMet (in red, Figure 1) by a base mimic: aminoquinazoline. This moiety was chosen inspired by the work of Hennequin et al., describing the design of tyrosine kinase inhibitors and their use of this group to mimic ATP.13 In addition, we chose to substitute the amino group in positionC4 of the quinazoline by phenylpropylamine because Wahhab et al.14,15 observed that a hydrophobic group on the amine at position-C4 of the adenosine moiety in AdoMet analogues was favorable for DNMT inhibition activity. This quinazoline moiety was coupled to the amine at position-C4 of a cytosine (in blue, Figure 1) to result in compound 1. Then we explored the different building blocks of the conjugates. (i) The cytidine was replaced by base mimics (in blue, Figure 1), with success by quinoline moieties. This latter substitution was guided by our docking studies of the known inhibitor 2 (SGI1027, structure in Table 1)16 in the DNMT catalytic pocket suggesting that the quinoline group fits within the cytidine pocket.17 (ii) Several hydrophobic groups (in black, Figure 1) were tested on the 4-aminoquinazoline. (iii) The linker arm (in green, Figure 1) was modulated. For example, we attempted to reduce the flexibility of the linker to form bioactive conformers. Indeed, in certain cases, constrained analogues are known to induce an increase in the inhibitory activity due to the restriction of the conformational flexibility that reduces entropy.18 This strategy was successfully used by Castellano et al. in the synthesis of constrained analogues of procaine, a DNMT inhibitor,19 and by us to develop potent derivatives of N-phthalyl-L-tryptophan (RG108),20 another DNMT inhibitor.21 Accordingly, constrained linkers containing piperidine, piperazine, and pyrrolidine moieties (in green, Figure 1) were synthesized. Chemical Synthesis. Compound 1 and 14 were obtained from key compound 9 that was synthesized from 4quinazolinone 3 as previously described22 (Scheme 1). The alcohol functionality and quinazolinone groups of 3 were chlorinated in a single step using hot thionyl chloride and coupled to phenylpropylamine to form 4-aminoquinazoline derivative 4 in 85% overall yield. Activation by phthalimide led to compound 5. Then methylhydrazine treatment followed by protection of the resulting amine by 2-nitrophenylsulfonyl chloride afforded 4-aminoquinazoline derivative 6 in good yields. After N-alkylation of 4-aminoquinazoline 6 with 2-(NBoc-amino)ethyl bromide, followed by TFA treatment, compound 8 was obtained. Then the N-1 nitrogen atom of

a (a) N-Boc-4-methanolpiperidine, NaH, DMF, 110 °C, 3 h, 67%; (b) POCl3, triazole, TEA, MeCN, RT, 18 h; (c) 3-3-phenylpropylamine, TEA, DMF, RT, 2 h, 80% over two steps; (d) TFA, RT, 1 h, 96%; (e) (i) 2-(N-Boc-amino)-ethyl bromide, TEA, DMF, 65 °C, 2 h, (ii) TFA, RT, 1 h; (f) (i) 9, POCl3, triazole, TEA, MeCN, 18 h, (ii) TEA, DMF, RT, 2 h, 23% over three steps.

7-Fluoroquinazolone 15, obtained as previously described by Hennequin et al.,13 was reacted with N-Boc-piperidinemethanol to afford 4-quinazolinone 16 that was converted to 4aminoquinazoline 17 in basic conditions according to a twostep procedure used in pyrimidine chemistry, implying the formation of a 4-triazolylquinazoline intermediate, followed by the substitution with phenylpropylamine.23 Then Boc-deprotection of compound 17 with TFA afforded derivative 18. Compound 18 was N-alkylated with 2-(N-Boc-amino)ethyl bromide, followed by TFA treatment, and then coupled to 9 converted to cytosine upon use of 1,2,4-triazole/POCl3 in the presence of triethylamine to afford compound 19. To explore the cytidine pocket, the pyrimidine moiety was replaced by a quinoline by using either a flexible linker as in 14 or the rigid linkers as in 20. Compound 20 was synthesized by coupling of compound 18 to quinoline 12 in the presence of K2CO3 (Scheme 3). Then each moiety of compound 1 was modulated (Figure 1) by (i) varying the linker between the quinazoline and quinoline moieties (compounds 21, 22, 25), (ii) replacing the quinazoline by a quinazolinone (compound 28), (iii) modulating the quinoline moiety mimicking the cytidine (Scheme 4), and last by (iv) varying or substituting the alkyl chain or the aromatic ring of the phenyl propyl group (Scheme 5). Synthesis of compound 21 was done by aromatic substitution of the chloride atom of 4-chloroquinoline by 18. Compound 22 and 25 were obtained from 15 by using, 4667



Article

Scheme 3. Synthesis of 20, 21, 22, 25, and 28a

(a) 12, K2CO3, KI, DMF, 65 °C, 12 h, 80%; (b) 4-chloroquinoline, K2CO3, DMF, 90 °C, 12 h, 12%; (c) (i) N-Boc-3-(hydroxymethyl)pyrrolidine, NaH, DMF, 110 °C, 3 h, 60%, (ii) POCl3, triazole, TEA, MeCN, RT, 18 h, (iii) 3-phenylpropylamine, TEA, DMF, RT, 2 h, 63% over two steps, (iv) TFA, RT, 1 h, 76%, (v) 12, K2CO3, KI, DMF, 65 °C, 12 h, 31%; (d) (i) 1-(N-Boc)-4-(2-hydroxyethyl)piperazine , NaH, DMF, 110 °C, 4 h, 61%, (ii) POCl3, triazole, TEA, MeCN, RT, 18 h, (iii) 3-phenylpropylamine, TEA, DMF, RT, 2 h, 21% over two steps; (e) TFA, RT, 1 h, 92%; (f) 12, K2CO3, KI, DMF, 65 °C, 12 h, 15%; (g) 1-chloro-3-phenylpropane, K2CO3, KI, DMF, 65 °C, 12 h, 95%; (h) TFA, RT, 1.5 h, 83%; (i) 12, K2CO3, KI, DMF, 65 °C, 12 h, 28%. a

respectively, N-Boc-3-(hydroxymethyl)pyrrolidine or piperazine derivative, obtained by N-alkylation of Boc-piperazine by 2-chloroethanol. Quinazolinone 28 was synthesized from intermediate 16. This latter was N-alkylated by 1-chloro-3phenylpropane to afford 26 that was deprotected to obtain 27. The N-alkylation of 27 by 12 in the presence of K2CO3 gave the desired product 28. Analogues 29−51 were synthesized following the same pathway used to afford compound 20 (Scheme 4) by N-alkylation of 18 with the corresponding 2chloroethylamine derivatives of quinoline, isoquinoline, indole, naphthalene, pyridine, or aniline in the presence of K2CO3. Compounds 47 and 49 were obtained from compounds 46 and 48, respectively, by action of BBr3. Finally, compounds 53−70 were also synthesized by following the same synthetic pathway

to obtain inhibitor 20 from quinazolinone 16, with the difference that the triazolyl intermediate 52 was isolated to be reacted specifically with the corresponding substituent (Scheme 5). Structure−Activity Relationship Studies. All compounds were tested for their ability to inhibit the methylation of a DNA duplex by human catalytic DNMT3A or human fulllength DNMT1 as previously described24,25 (Table 1). The quinoline DNMT inhibitor 2 was used as positive control. Compound 1 bearing a cytosine moiety showed no activity, while the constrained analogue 19 showed a weak inhibition of DNMT3A (44% at 32 μM). This inhibition activity increased with the replacement of cytosine by quinoline as in compound 14 (59% at 10 μM) and, more interestingly, in the constrained 4668



Article

Scheme 4. Synthesis of 29−51a

a

(a) Corresponding halogenoalkane, K2CO3, KI, DMF, 65 °C, 12 h; (b) BBr3, DCM.

analogue 20 (81% inhibition at 10 μM). Derivative 20 also inhibited 91% of the activity of DNMT1 at 100 μM. The difference in concentration used in the two assays is due to the difference in sensitivity.25 Compared to the reference nonnucleoside DNMT inhibitor 2, the quinazoline−quinoline conjugate 20 inhibited the DNMT3A with an efficacy of 98% while 2 presented a maximum of inhibition efficacy of 72%. The antiproliferative activity of 20 on five cancer cell lines was measured, showing a micromolar range cytotoxicity on all cell lines (Supporting Information, Table S1). Because reference compound 2 is described as a weak DNA ligand,16,26 we tested whether our lead compound 20 could inhibit DNMTs by interaction with the DNA. The capacity of 20 to bind to DNA was evaluated by monitoring the melting temperature of three DNA double-stranded hairpins owning one (hp1), two (hp2), or no (hpctrl) CpG sites (Supporting Information, Figure S1). Compound 20 did not show any enthalpy interaction with DNA (Supporting Information, Figure S1). This was also confirmed by DNase I footprinting experiments in which no binding of compound 20 to 265 bp

and 117 bp DNA fragments was observed (Supporting Information, Figure S2). Because the presence of the constrained piperidine linker and the quinazoline moiety were associated with strong inhibition of the DNMTs, we explored compound 20 by modulating (i) the quinazoline moiety (in red in Figure 1 and Table 1), (ii) the linker between the quinoline and quinazoline moieties (depicted in green), (iii) the quinoline moiety (in blue), and (iv) the hydrophobic substituent (in black). First, the quinazoline scaffold was replaced by quinazolinone in compound 28 (Scheme 3), which induced a decrease in the DNMT inhibition activity underlining the importance of the quinazoline moiety. Role of the Linker. Second, the linker was explored by removing the aminoethyl linker to give the more rigid compound 21 (Scheme 3) or by replacing the piperidine by a pyrrolidine in compound 22 or by a piperazine in derivative 25. Analogue 21 completely lost the capacity to inhibit DNMTs (Table 1 and Figure 2), clearly suggesting that a short flexible linker between the quinoline moiety and the 4669



Article

Scheme 5. Synthesis of 53−70a

(a) POCl3, 1,2,4-triazole, TEA, MeCN, RT, 18 h, 98%; (b) Appropriate amine, TEA, DMF, RT or 65 °C, 2−6 h; (c) TFA, RT, 1 h; (d)12, K2CO3, KI, DMF, 65 °C, 12 h.

a

piperidine linker is crucial for activity. Compared to the lead compound 20, pyrrolidyl linker 22 and piperazinyl linker 25 induced a decrease in the DNMT inhibition activity, underlining the importance of the piperidine linker. Modulation of the Quinoline Moiety. Third, we modulated the quinoline moiety that mimics the cytosine. A naphthyl group was used to determine the importance of the nitrogen atom of the quinolone (α- and β-naphthyl, compounds 29 and 30). Then the nitrogen atom was moved or externalized (compounds 31−37). Monoaromatic rings were synthesized (38−40) together with various substitutions of the quinoline moiety (41−50). Finally the C4 amino group was Nmethylated (51). The naphthyl compounds 29 and 30 and isoquinoline 34 were inactive, while the quinoline and isoquinoline derivatives 31, 32, and 33 showed some inhibition of DNMT3A (Table 1 and Figure 2). These results stressed the importance of the nitrogen atom on the bicyclic aromatic group. More precisely, the nitrogen must be in “para” position of the linker either directly on the aromatic ring bearing the linker or external to the aromatic ring. Quinoline 20 bearing the nitrogen atom in “para” of the linker conserved the best activity. This nitrogen atom can also be positioned externally to the aromatic ring as demonstrated by the inhibitory activity of compound 37. This

was further confirmed by the monocyclic derivative 40, which maintained a fair inhibitory activity. Nevertheless, the monocyclic moiety (i.e. compounds 38, 39, and 40) is penalizing compared to a bicycle heteroaromatic ring, i.e., compound 20 compared to 28. Thus, we maintained the 4aminoquinoline attached and further studied its functionalization. Derivatives 41−50 showed a good activity except for trifluoromethyl-substituted 42 and 43 that resulted in being inactive, suggesting that electron withdrawing groups are less active. Interestingly, compounds 41 and 48 started to show an increase in the inhibition activity, especially against DNMT1, compared to parent compound 20. Finally, the decrease of activity observed with compound 51 attached through a methylamine to position C4 instead of an amine group as in compound 20 indicated the importance of a secondary amine group at this position Modification of the C4 Quinazoline Substituent. Fourth, the role of the hydrophobic group on position C4 of the quinazoline moiety (in black in Figure 1 and Table 1) was investigated by varying or substituting the alkyl chain or the aromatic ring of the phenylpropylamine group. The length of the linker between the amino group of the quinazoline and the phenyl was modified by two carbon atoms (compound 53) or one carbon atom (compound 54). Compound 53 with two 4670



Article

Table 1. Biological Activity: Inhibition Activity of Human DNMT1 and Human Catalytic DNMT3A Are Reported as Mean Value of the Percentage of Inhibition at the Indicated Concentration of Compounda

4671



Article

Table 1. continued

4672



Article

Table 1. continued

a

ND = not determined. 4673



Article

Figure 2. Percentage of inhibition of DNMT3A at 10 μM of compound (full bars) or DNMT1 at 32 μM of compound (hatched bars).

carbon atoms presented the same inhibition activity as compound 20, while a shorter linker induced a drastic loss of activity (Table 1 and Figure 2). Consequently, a propyl or ethyl linker between the quinazoline and the phenyl group must be preferably used. Compound 55, bearing a propyl group, and derivative 56, containing unsubstituted quinazoline, induced a decrease in the activity especially against DNMT3A, confirming that the hydrophobic group contributes to the inhibitory activity. This result is in agreement with the findings of Wahhab et al. on AdoMet analogues.15 N-Ethylaniline 57 showed a similar activity against DNMT3A to compound 20 and as light increase of activity against DNMT1 at 32 μM. Introduction of a cycle into the spacer as a piperidine (58) or piperazine (59) was not favorable. Next, we explored the substitutions on the aromatic ring. Ortho substitution by chloride in compound 60 induced a decrease in the inhibition of DNMT3A at 10 μM while it kept a good activity for DNMT1, suggesting that selectivity can be achieved. Meta substitution in derivative 61 resulted into a higher specific inhibition of DNMT3A and a loss of DNMT1 inhibition (Table 1 and Figure 2). The chloride atom in para position (compound 62) clearly enhanced DNMT1 inhibition at 32 μM and inhibited DNMT3A. Various substitutions in para position were tested. A methoxy group (63) or an amine (65) gave a good inhibition, while the presence of a sulfonamide group (66) induced a loss of DNMT3A inhibition. A nitro group (64) decreased the inhibition activity against both DNMT3A and DNMT1, confirming that electron-withdrawing groups are deleterious for DNMT inhibition at this position. Interestingly, isopropyl derivative 67 induced a weak loss of activity against DNMT3A, while it was completely inactive against DNMT1, comforting, as with compound 61, the possibility to design selective inhibitors of each DNMT isoform. Last, the size of the hydrophobic group was drastically enlarged by the use of biphenyl 68, propylnaphthalene 69 and N-ethyl-1-naphthylamine 70 substituents. Biphenyl 68 showed a good DNMT3A inhibition but decreased its ability to inhibit DNMT1 (Table 1). The naphthyl derivatives 69 and 70 were found to strongly inhibit DNMT3A and DNMT1, but 69 showed a problem of

solubility in the DNMT3A assay conditions, leading to a decrease of the inhibition activity at 32 μM. For the most potent compounds of this series, we determined the concentration at which they induced 50% of inhibition of the activity of the DNMT (Table 2) in comparison to parent compound 20. Table 2. EC50 Inhibition Values against Full-Length Human DNMT1 and Catalytic DNMT3Aa compd 20 41 48 53 57 61 62 63 68 69 70

DNMT3A

DNMT1

± ± ± ± ± ± ± ± ± ± ±

46 ± 2 24 ± 2 30 ± 0 32 ± 1 73 ± 2 ND 26 ± 1 27 ± 3 100 ± 3 16 ± 2 20 ± 4

4.4 2.1 1.4 4.9 6.8 1.0 1.2 3.4 1.1 1.9 0.3

0.3 0.7 1.1 1.2 2.8 0.4 0.3 0.5 1.2 1.2 0.2

a EC50 are reported as the median concentration, at which a compound inhibits 50% of the methylation efficacy and the standard error of at least two independent experiments for DNMT1 and three for DNMT3A (EC50 ± SE μM). ND = not determined.

EC50 confirmed that a methyl group at C2 position (compound 41) or dimethoxyl groups at C6 and 7 (compound 48) of the quinoline moiety induced a weak increase of DNMT1 and 3A inhibition (Figure 3). EC50 also showed that the presence of a spacer between the amino group of the quinazoline and aromatic of the hydrophobic group with two carbon atoms 53 as well as N-ethylmino moiety 57 presented the same inhibition activity as compound 20 (Table 2). The comparison of the EC50 for the substitution by a chloride on the aromatic ring in meta, compound 61, and in para, compound 62, revealed a high potency of the compounds. Compound 62 showed an EC50 of 1 μM against DNMT3A and 26 μM against DNMT1. The importance of the para 4674



Article

the expression of the luciferase after 24 h of treatment at 5 and 1 μM, respectively. Results obtained with compound 70 were disappointing given that it was the most potent inhibitor in the enzymatic assay. However, the poor solubility observed for 70 in cell media could explain the weaker activity in the cellular experiment. Compound 68 showed an interesting pattern because it induced a reactivation of the luciferase gene starting from 0.1 μM. Thus, we pursued the cellular investigation with compound 68 in comparison to parent compound 20. ii. DNA Demethylation and Chromatin Opening. We explored the methylation level of the CMV promoter upon treatment with compounds 20 and 68, together with nucleosome positioning on the promoter. Indeed, emerging data have revealed that nucleosome remodeling works in concert with DNA methylation and histone modifications to play a central role in tumor-specific gene silencing.27,28 DNA methylation at promoters induces distinct changes, resulting in nucleosome occupancy at transcription start sites.29,30 The reactivation of silenced genes using DNMT inhibitors should be accompanied by a loss of nucleosomes from the promoter region. Here, the impact of the treatment with the DNMT inhibitors at the CMV promoter was analyzed by using the NOMe-Seq method that follows both DNA methylation and DNA accessibility.31 CMV methylation and nucleosome occupancy was studied after 24 h of treatment with 5 μM of 20 and 68 that increased the luciferase signal of 4.7- and 4.2fold, respectively (Figure 4 and Supporting Information, Table S2); 1 μM of 5azadC was used as positive control. The concentrations were chosen in order to have no effect on cellular viability after 24 h of treatment (data not shown). CMV promoter was demethylated at 33% and 52%, respectively, upon treatment with compounds 20 and 68 (Figure 5A). The red and blue squares represent the methylated and unmethylated CpG sites, respectively, present on the CMV prometer. Each vertical line represents a CpG at the indicated position and each horizontal line an analyzed cell. In addition, 20 and 68 induced an increase in DNA accessibility by opening the chromatin of 77% and 87%, respectively (Figure 5B). The white and black square represent the nonaccessible and accessible DNA sites, respectively, on the CMV promoter. For comparison, nucleoside inhibitor 5azadC demethylated at 50% the CMV promoter and increased the accessibility of the DNA by 120%. iii. Tumor suppressor gene demethylation and reactivation. Finally, because the DNMT inhibitors were able to reactivate the luciferase gene, demethylate and open the chromatin at the CMV promoter, we studied if they were also able to demethylate an endogenous tumor suppressor gene in a solid cancer model, HCT116 colon cancer cells, wellstudied for the role of DNA methylation in tumor suppressor gene silencing.32 HCT116 cells were treated during 3 days and 7 days of treatments (treatment schedule reported in Figure 6A). This timing frame was chosen because we argued that several doubling cycles were needed to observe the demethylation effect because the inhibitors act by passive DNA demethylation.9 In addition, we chose low concentrations (100 nM) of compounds in order to distinguish the epigenetics effects from the cytotoxic one. 5AzadC was used as positive control in the assay. Cells treated with compounds 20 and 68, showed a 32% and 20% of demethylation of the well-studied CDKN2A promoter, respectively, after 3 days of consecutive treatment, and 42% and 71% of demethylation after 7 days of treatment (Figure 6B). Noteworthy, the decrease in the

Figure 3. Dose−response inhibition curves of hDNMT3Acat for selected inhibitors, run in technical duplicates in at least two independent experiments.

substitution to increase inhibitory potency against DNMT1 is shown by compound 62 because compound 61 is inactive on DNMT1 at 32 μM (Table 2), while the meta-methoxy derivative 63 presented the same EC50 as 62 against DNMT1. The evaluation of the EC50 for compound 68 confirmed that it preferentially inhibited DNMT3A. Derivative 70 was the most potent inhibitor with a submicromolar activity against DNMT3A (EC50 = 0.33 μM) (Table 2). Cellular Activity. i. Reactivation of a Luciferase Reporter Gene. The ability of the DNMT inhibitors to reactivate gene expression by promoter demethylation was tested on a gene reporter model containing the luciferase gene under the control of a partially methylated promoter, CMV, in a stable construction in leukemia KG-1 cells17 (Figure 4). Three

Figure 4. Luciferase reactivation fold (RF) in CMV-Luc KG-1 cells after 24 h treatment of cells by 5azadC (black circles), 20 (blue squares), 29 (gray lozenges), 68 (red triangles), and 70 (brown triangles). RF represents the ratio to nontreated cells.

inhibitors, parent compound 20, the most potent compound 70, and the more selective DNMT3A inhibitor compound 68, were chosen for further biological studies in cancer cell lines. In addition, inactive compound 29 was used as negative control. As positive control in cellular experiments and for comparison, nucleoside inhibitor 5azadC was used. As expected, negative compound 29 showed no induction of the luciferase expression, while compounds 20, 68, and 70 were able to induce by 4-fold 4675



Article

Figure 5. CMV methylation status (A) and chromatin accessibility (B) were monitored by NOMe-Seq after 24 h treatment of CMV-luc KG-1 cells by 1 μM of 5azadC (top) or by 5 μM of 20 (middle) and 68 (bottom). In red, methylated CpGs and in blue nonmethylated CpGs; in white, nonmethylated GpCs; in black, methylated GpCs corresponding to accessible DNA. Methylation status and chromatin access were represented as percentage compared to not-treated cells (NT).

Figure 6. CDKN2A methylation and mRNA expression in HCT116 cells. (A) Graphical schedule of treatments (white arrows) and analysis (gray arrows). (B) CDKN2A methylation status was monitored by COBRA, and results are represented as ratio to untreated cells. (C) mRNA expression levels of CDKN2A measured by RT qPCR are reported as fold change relatively to untreated cells. Cells were harvested after 3 and 7 days treatment with compounds 5azadC (black bars), 20 (hatched bars), and 68 (gray bars) at 100 nM for methylation analysis and 5azadC and 68 for mRNA expression. Each point represents average value from at least two independent experiments and error bars represent standard errors.

methylation induced by compounds 20 and 68 increased with the duration of the treatment, corroborating the passive demethylation mechanism expected for DNMT inhibitors. Furthermore, compound 68 was more active compared to compound 20, in agreement with the inhibition assay and the CMV-Luc KG-1 findings. Importantly, the progressive decrease of promoter methylation was also correlated to re-expression of CDKN2A after treatments with compound 68 (Figure 6C, gray bars). Similarly to the results obtained with 5azadC, no reexpression of CDKN2A was observed after 3 days of treatment with 68, when little demethylation was measured, whereas a 13fold increase of the expression was measured after 7 days treatment when the demethylation effect was high (Figure 6C). Our results are in agreement with the long-term effects described by Tsai et al., showing a re-expression of tumor suppressor genes after 7 to 14 days of treatment with low doses of 5azadC.33

DNMT1 at 46 μM. Consequently, a structure−activity relationship (SAR) study was undertaken to understand and improve the key features of quinoline−quinazoline conjugates. The different parts of the molecule were explored: the quinazoline and its substituents, the quinoline and the linker between the two moieties. SAR demonstrated the key roles of the quinoline group, the N-aminoethylpiperidyl part as linker and the presence of an aromatic hydrophobic substituent on the C4 amine of the quinazoline. Lead compound 20 and the most potent DNMT3A inhibitor 68 induced the reactivation of the luciferase reporter gene in KG-1 leukemia cells in correlation with the ability to open chromatin. Compounds 20 and 68 also demethylated the tumor suppressor gene CDKN2A promoter in HCT116 colon cancer cells after 7 days of treatment at 100 nM. Compound 68 is the most effective against endogenous CDKN2A methylation and was able to induce CDKN2A re-expression. Interestingly, non-nucleoside DNMT inhibitor 68 presented a similar profile as nucleoside inhibitor 5azadC. A major challenge is to increase the solubility of this family of compounds that was a major limit for the use of potent in vitro inhibitor 70 in cancer cells. Finally, certain compounds showed some selectivity for one isoform of DNMTs constituting an interesting feature to be further

■

CONCLUSIONS In conclusion, we designed bisubstrate analogues directed against the DNMTs. The adenine of the AdoMet substrate was replaced by a quinazoline, inspired by the ATP-mimicking kinase inhibitors, and was covalently attached to cytosine or quinoline. Cytosine conjugates resulted in being unattractive inhibitors of DNMT3A and DNMT1, while conjugate quinazoline−quinoline 20 inhibited DNMT3A at 4 μM and 4676



Article

previously described with minor modifications.27,31,37,38 The detailed protocol is reported in the Supporting Information. The treated DNA was treated by sodium bisulfite with the EZ DNA Methylation-Gold Kit (Zymo Research). Bisulfited DNA was eluted with 12 μL of water, and 8 μL were used for CMV amplification. PCR Amplification of Bisulfite-Treated DNA. The CMV promoter DNA amplification was set up with 8.3 μL of eluted bisulfited DNA in 20 μL of PCR reaction containing 1× KAPA2G buffer A, 2 mM MgCl2, 200 μM dNTPs, 125 nM each primer, and 1unit KAPA2G Robust HotStart DNA Polymerase (KapaBiosystems) on C1000 Touchthermal cycler 95 °C 3 min following by 95 °C for 20 s, 55 °C for 30 s, 72 °C for 30 s × 40 cycles, and final extension at 72 °C 1 min. PCR fragments were quality controlled by agarose gel electrophoresis. Primers used were: forward, 5′-GGGGTTATTAGTTTATAGTTTATATATGGA-3′; reverse, 5′-AATACCAAAACAAACTCCCATTAAC. Steps of cloning and sequencing were as the same were described previously except M13 PCR, which was performed to amplify cloned sequence in a 20 μL PCR reaction containing 1× KAPA2G buffer B, 2 mM MgCl2, 200 μM dNTPs, 125 nM each primer, and 1 unit KAPA2G Robust HotStart DNA polymerase (KapaBiosystems) on a C1000 Touchthermal cycler 95 °C 3 min followed by 95 °C for 20 s, 55 °C for 30 s, 72 °C for 30 s × 40 cycles, and final extension at 72 °C 1 min. Sequence Alignment and Analysis of CG and GC Methylation Levels. Genomic alignment and bisulfite sequence analysis was performed largely as previously described.39 Bisulfite Sequencing DNA Methylation Analysis (BISMA) software was used for CpG methylation analysis of primary bisulfite sequencing data from subcloned individual.40 Typically in such a lollipop-style map, red (CG) symbols are indicative of a methylated cytosine while blue (CG) symbols represent unmethylated cytosines. Methyl Viewer was used with parameters to obtain GC methylation results and nucleosome occupancy.41 With this software results, black (GC) symbols represented of a methylated cytosine and white (GC) symbols represent unmethylated cytosines. The ratio of methylation is representative of the chromatin accessibility by M.CviPI and nucleosome occupancy. The combined methylation status for each clone by CG (BISMA) or GC (Methyl Viewer) site has been assembled and represented as shown in Figure 3, where each row corresponds to one of the individual cloned alleles and each column represents an individual CG or GC site. The methylation status of CG and GC cytosines at a given position within the sequence, given as ratio of methylation, can be calculated as follows: R = [MCT/MCNT], where MCT is the percentage of methylated cytosines into analyzed DNA sequence in treated sample, and MCNT is the percentage of methylated cytosines in the same analyzed DNA sequence in not treated sample. Accuracy is improved by increasing the number of clones analyzed, provided there is not bias in the cloning or amplification steps. Methylation Analysis. Combined Bisulfite Restriction Analysis (COBRA). This technique combines bisulfite conversion-based polymerase chain reaction with restriction digestion. DNA Extraction and Bisulfite Treatment. DNA was isolated from cultured cells using the DNeasy Blood and Tissue Kit according to the manufacturer’s specifications (Qiagen). DNA bisulfite conversion was performed on 2 μg of DNA using the EZ DNA Methylation-Gold Kit according to the manufacturer’s specifications (Zymo Research) and bisulfited DNA was eluted with 10 μL of water. CDKN2A Promoter PCR Amplification of Bisulfite-Treated DNA. DNA amplification was set up with 100 ng of eluted bisulfited DNA in 50 μL of PCR reaction containing 1× PCR buffer, 2.5 mM MgCl2, 0.3 mM dNTPs, 400 nM each primer and 1.25 units EpiTaq HS (Takara) on C1000 Touchthermal cycler by 98 °C for 10 s, 55 °C for 30 s, and 72 °C for 30 s × 40 cycles. Primers were designed as described in the Supporting Information. Then 20 μL of PCR amplicons were digested in 30 μL by 2 units of BsiEI at 60 °C for 90 min. DNA fragments were migrated by 2% agarose gel electrophoresis, and each band was quantified by Image Lab Software v2.0 (BioRad Laboratories).

exploited. Interestingly, Schramn’s group has recently resolved the transition state structure of DNMT1 that could guide further improvement.34 This will lead to conceive selective chemical probes for DNMT to understand the role of each isoform in cancer cells.

■

EXPERIMENTAL SECTION

Chemicals. AdoMet and 5-aza-2′-deoxycytidine (5azadC) were bought from Sigma-Aldrich (Saint-Quentin, France), aliquoted in H2O, and stored at −80 °C. All other compounds were prepared as 10−2 M aliquots in 100% DMSO, stored at −20 °C, and used as dilutions freshly prepared in culture medium. Commercial chemicals were from Sigma-Aldrich Alfa Aesar; oligonucleotides were bought from Eurogentec. The chemical synthesis, analysis of the compounds are detailed in the Supporting Information. The purity of all final compounds was verified to be higher than 95% using reversed-phase HPLC on an Xterra C18 MS column (3.9 100 mm; Waters) with a linear gradient acetonitrile in 0.01% TEA (0 to 95% CH3CN). HRMS-ESI were obtained on a Bruker MicroTOF. The NMR spectra were recorded on a Bruker Avance II spectrometer equipped with a 13C cryoprobe at 500 MHz for 1H and 125 MHz for 13C. DNA Methyltransferase Activity. DNMT3A and DNMT1 assays were previously described25,35 and are detailed in the Supporting Information. EC50 values were determined with curve fitting analysis method (nonlinear regression model with a sigmoid dose response, variable Hill slope coefficient) provided by the Prism Software (GraphPad). Results were expressed as average EC50 values (concentration of tested compound that inhibits 50% of the maximum effect for the considered compound). Cell Lines and Reagents. All cell media were supplemented with 10% fetal calf serum (Lonza) and cultivated at 37 °C and under 5% CO2. WM266-4 human metastatic melanoma cells and PANC-1 human pancreatic carcinoma cells were obtained respectively from the European Collection of Cell Cultures (ECACC, Salisbury, England) and from the American Type Culture Collection (ATCC, Manassas, VA, USA) and cultivated in DMEM media (Lonza, Basel, Switzerland). U-87 human primary glioblastoma cells and HCT116 human colon cancer cells were obtained from ATCC and cultivated in MEM containing 4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid (HEPES) and L-glutamine (Lonza). U-87 human primary glioblastoma cells, KG-1 human acute myeloid leukemia cells, and Karpass299 nonHodgkin’s Ki-positive large cells lymphoma were obtained from the ATCC and cultivated in RPMI-1640 medium (Lonza). Cell Proliferation Assay. For antiproliferative activity, cells were seeded (3 × 104 cells/mL for WM 266-4 and U-87, 2 × 104 cells/mL for PANC-1, KG-1, and Karpas299, and 1.5 × 104 cells/mL for HCT116) at day 0 in a 96-well plate. Cells were treated with test compound solutions at a dose range from 3.2 × 10−9 M to 1 × 10−5 M. Treatment was repeated on days 2 and 3, and on day 4 the antiproliferative activity of compounds was measured using the ATP quantification method “ATPlite One Step Luminescence Assay System” (PerkinElmer). Raw data were analyzed with Prism 4.03 to generate EC50 values corresponding to the compound concentration required to cause a 50% decrease in cell viability as compared with untreated controls. The values presented are the mean of at least two independent experiments run in duplicates. For other assays, cells were seeded at 2 × 105/mL for KG-1 and 1 × 105/mL for HCT116 at day 0 in 10 mL in T25 flasks, and cells were treated with different compounds of interest. Luciferase Induction. KG-1-Luc cells, containing the CMV-luc stable construction, were generated and treated as previously described17 in 96-well plates. After 24 h, the luciferase induction was quantified with the Britelite assay system (PerkinElmer). The mean of three experiments and the standard error is reported. CMV Nucleosome Footprinting. Nucleosome Occupancy and Methylome Sequencing (NOMe-Seq) is a modified version of the methylation-dependent single promoter assay described by Miranda et al. 2010,36 and methylase-based DNA assay was performed as 4677



Article

RNA Isolation and Quantitative PCR Analyses. Total RNA was extracted from the cultured HCT116 cells by using RNeasy Kit (Qiagen) following manufacturer’s protocols, digested by DNase I, and reverse transcribed (PrimeScript RT Master Mix, Takara). CDKN2A was amplified using the CFX384 real-time PCR detection system (BioRad Laboratories) and SYBR Premix Ex Taq (Takara) under the following conditions: 95 °C for 30 s, followed by 40 cycles of 95 °C for 5 s and then 60 °C for 30 s. A melt curve analysis was performed (60−95 °C, rising by 0.5 °C every 10 s). Primers used for amplification were as follows: CDKN2A, forward 5′-CATGGAGCCTTCGGCTGACT-3′, reverse 5′-CATCATCATGACCTGGATCG-3′; TBP, forward 5′-TTGACCTAAAGACCATTGCACTTCGT-3′, reverse 5′-TTACCGCAGCAAACCGCTTG-3′; YWHAZ, forward 5′-CCCTCAAACCTTGCTTCTAGGAGA-3′, reverse 5′-TCATATCGCTCAGCCTGCTCG-3′. All PCRs were performed in duplicates and analyses were conducted in parallel using human TBP and YWHAZ for normalization.

■

(5) Baylin, S. B.; Jones, P. A. A Decade of Exploring the Cancer Epigenome - Biological and Translational Implications. Nat. Rev. Cancer 2011, 11, 726−734. (6) Pandiyan, K.; You, J. S.; Yang, X.; Dai, C.; Zhou, X. J.; Baylin, S. B.; Jones, P. A.; Liang, G. Functional DNA Demethylation Is Accompanied by Chromatin Accessibility. Nucleic Acids Res. 2013, 41, 3973−3985. (7) Jurkowska, R. Z.; Jurkowski, T. P.; Jeltsch, A. Structure and Function of Mammalian DNA Methyltransferases. ChemBioChem 2011, 12, 206−222. (8) Subramaniam, D.; Thombre, R.; Dhar, A.; Anant, S. DNA Methyltransferases: A Novel Target for Prevention and Therapy. Front. Oncol. 2014, 4, 80. (9) Erdmann, A.; Halby, L.; Fahy, J.; Arimondo, P. B. Targeting DNA Methylation with Small Molecules: What’s Next? J. Med. Chem. 2015, 58, 2569−2583. (10) Lund, K.; Cole, J. J.; VanderKraats, N. D.; McBryan, T.; Pchelintsev, N. A.; Clark, W.; Copland, M.; Edwards, J. R.; Adams, P. D. Dnmt Inhibitors Reverse a Specific Signature of Aberrant Promoter DNA Methylation and Associated Gene Silencing in Aml. Genome Biol. 2014, 15, 406. (11) Fahy, J.; Jeltsch, A.; Arimondo, P. B. DNA Methyltransferase Inhibitors in Cancer: A Chemical and Therapeutic Patent Overview and Selected Clinical Studies. Expert Opin. Ther. Pat. 2012, 22, 1427− 1442. (12) Lopez, M.; Halby, L.; Arimondo, P. B. DNA Methyltransferase Inhibitors: Development and Applications. Adv. Exp. Med. Biol. 2016, 945, 431−473. (13) Hennequin, L. F.; Thomas, A. P.; Johnstone, C.; Stokes, E. S.; Ple, P. A.; Lohmann, J. J.; Ogilvie, D. J.; Dukes, M.; Wedge, S. R.; Curwen, J. O.; Kendrew, J.; Lambert-van der Brempt, C. Design and Structure-Activity Relationship of a New Class of Potent Vegf Receptor Tyrosine Kinase Inhibitors. J. Med. Chem. 1999, 42, 5369− 5389. (14) Isakovic, L.; Saavedra, O. M.; Llewellyn, D. B.; Claridge, S.; Zhan, L.; Bernstein, N.; Vaisburg, A.; Elowe, N.; Petschner, A. J.; Rahil, J.; Beaulieu, N.; Gauthier, F.; MacLeod, A. R.; Delorme, D.; Besterman, J. M.; Wahhab, A. Constrained (L-)-S-Adenosyl-LHomocysteine (SAH) Analogues as DNA Methyltransferase Inhibitors. Bioorg. Med. Chem. Lett. 2009, 19, 2742−2746. (15) Saavedra, O. M.; Isakovic, L.; Llewellyn, D. B.; Zhan, L.; Bernstein, N.; Claridge, S.; Raeppel, F.; Vaisburg, A.; Elowe, N.; Petschner, A. J.; Rahil, J.; Beaulieu, N.; MacLeod, A. R.; Delorme, D.; Besterman, J. M.; Wahhab, A. Sar around (L)-S-Adenosyl-LHomocysteine, an Inhibitor of Human DNA Methyltransferase (DNMT) Enzymes. Bioorg. Med. Chem. Lett. 2009, 19, 2747−2751. (16) Datta, J.; Ghoshal, K.; Denny, W. A.; Gamage, S. A.; Brooke, D. G.; Phiasivongsa, P.; Redkar, S.; Jacob, S. T. A New Class of Quinoline-Based DNA Hypomethylating Agents Reactivates Tumor Suppressor Genes by Blocking DNA Methyltransferase 1 Activity and Inducing Its Degradation. Cancer Res. 2009, 69, 4277−4285. (17) Rilova, E.; Erdmann, A.; Gros, C.; Masson, V.; Aussagues, Y.; Poughon-Cassabois, V.; Rajavelu, A.; Jeltsch, A.; Menon, Y.; Novosad, N.; Gregoire, J. M.; Vispe, S.; Schambel, P.; Ausseil, F.; Sautel, F.; Arimondo, P. B.; Cantagrel, F. Design, Synthesis and Biological Evaluation of 4-Amino-N- (4-Aminophenyl)Benzamide Analogues of Quinoline-Based SGI-1027 as Inhibitors of DNA Methylation. ChemMedChem 2014, 9, 590−601. (18) Mann, A.; Camille Georges, W. Conformational Restriction and/or Steric Hindrance in Medicinal Chemistry. In The Practice of Medicinal Chemistry 3rd ed.; Academic Press: New York, 2008; Chapter 17, pp 363−379. (19) Castellano, S.; Kuck, D.; Sala, M.; Novellino, E.; Lyko, F.; Sbardella, G. Constrained Analogues of Procaine as Novel Small Molecule Inhibitors of DNA Methyltransferase-1. J. Med. Chem. 2008, 51, 2321−2325. (20) Siedlecki, P.; Garcia Boy, R.; Musch, T.; Brueckner, B.; Suhai, S.; Lyko, F.; Zielenkiewicz, P. Discovery of Two Novel, Small-Molecule Inhibitors of DNA Methylation. J. Med. Chem. 2006, 49, 678−683.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jmedchem.7b00176. Description of the synthesis and analytical characterization of all compounds and additional biological data (PDF) Molecular formula strings (CSV)

■

AUTHOR INFORMATION

Corresponding Author

*Phone: +33667435554. E-mail: [email protected]. ORCID

Paola B. Arimondo: 0000-0001-5175-4396 Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported for P.B.A. by Région Midi Pyrenées (Equipe d’Excellence and FEDER CNRS/Région Midi Pyrenées); Fondation InNaBioSanté, PlanCancer2014 (no. EPIG201401), the National Research Agency (Agence Nationale de la Recherche (ANR)) “Investissement d’avenir” (ANR-11-PHUC-001, CAPTOR research program).

■

ABBREVIATIONS USED DNMT, DNA methyltransferase; 5azadC, 5-aza-2′-deoxycytidine; 5azaC, 5-azacytidine; AdoMet, S-adenosyl-L-methionine; AdoHcy, S-adenosyl-L-homocysteine; SAR, structure−activity relationships

■

REFERENCES

(1) Berger, S. L.; Kouzarides, T.; Shiekhattar, R.; Shilatifard, A. An Operational Definition of Epigenetics. Genes Dev. 2009, 23, 781−783. (2) Kelly, T. K.; De Carvalho, D. D.; Jones, P. A. Epigenetic Modifications as Therapeutic Targets. Nat. Biotechnol. 2010, 28, 1069−1078. (3) Gros, C.; Fahy, J.; Halby, L.; Dufau, I.; Erdmann, A.; Gregoire, J. M.; Ausseil, F.; Vispe, S.; Arimondo, P. B. DNA Methylation Inhibitors in Cancer: Recent and Future Approaches. Biochimie 2012, 94, 2280− 2296. (4) Jones, P. A. Functions of DNA Methylation: Islands, Start Sites, Gene Bodies and Beyond. Nat. Rev. Genet. 2012, 13, 484−492. 4678



Article

(21) Asgatay, S.; Champion, C.; Marloie, G.; Drujon, T.; SenamaudBeaufort, C.; Ceccaldi, A.; Erdmann, A.; Rajavelu, A.; Schambel, P.; Jeltsch, A.; Lequin, O.; Karoyan, P.; Arimondo, P. B.; Guianvarc’h, D. Synthesis and Evaluation of Analogues of N-Phthaloyl-L-Tryptophan (RG108) as Inhibitors of DNA Methyltransferase 1. J. Med. Chem. 2014, 57, 421−434. (22) Mortlock, A. A.; Foote, K. M.; Heron, N. M.; Jung, F. H.; Pasquet, G.; Lohmann, J. J.; Warin, N.; Renaud, F.; De Savi, C.; Roberts, N. J.; Johnson, T.; Dousson, C. B.; Hill, G. B.; Perkins, D.; Hatter, G.; Wilkinson, R. W.; Wedge, S. R.; Heaton, S. P.; Odedra, R.; Keen, N. J.; Crafter, C.; Brown, E.; Thompson, K.; Brightwell, S.; Khatri, L.; Brady, M. C.; Kearney, S.; McKillop, D.; Rhead, S.; Parry, T.; Green, S. Discovery, Synthesis, and in Vivo Activity of a New Class of Pyrazoloquinazolines as Selective Inhibitors of Aurora B Kinase. J. Med. Chem. 2007, 50, 2213−2224. (23) Buchini, S.; Leumann, C. J. 2′-O-Aminoethyl Oligoribonucleotides Containing Novel Base Analogues: Synthesis and Triple-Helix Formation at Pyrimidine/Purine Inversion Sites. Eur. J. Org. Chem. 2006, 2006, 3152−3168. (24) Ceccaldi, A.; Rajavelu, A.; Champion, C.; Rampon, C.; Jurkowska, R.; Jankevicius, G.; Senamaud-Beaufort, C.; Ponger, L.; Gagey, N.; Ali, H. D.; Tost, J.; Vriz, S.; Ros, S.; Dauzonne, D.; Jeltsch, A.; Guianvarc’h, D.; Arimondo, P. B. C5-DNA Methyltransferase Inhibitors: From Screening to Effects on Zebrafish Embryo Development. ChemBioChem 2011, 12, 1337−1345. (25) Gros, C.; Chauvigne, L.; Poulet, A.; Menon, Y.; Ausseil, F.; Dufau, I.; Arimondo, P. B. Development of a Universal Radioactive DNA Methyltransferase Inhibition Test for High-Throughput Screening and Mechanistic Studies. Nucleic Acids Res. 2013, 41, e185. (26) Gros, C.; Fleury, L.; Nahoum, V.; Faux, C.; Valente, S.; Labella, D.; Cantagrel, F.; Rilova, E.; Bouhlel, M. A.; David-Cordonnier, M. H.; Dufau, I.; Ausseil, F.; Mai, A.; Mourey, L.; Lacroix, L.; Arimondo, P. B. New Insights on the Mechanism of Quinoline-Based DNA Methyltransferase Inhibitors. J. Biol. Chem. 2015, 290, 6293−302. (27) Sharma, S.; Kelly, T. K.; Jones, P. A. Epigenetics in Cancer. Carcinogenesis 2010, 31, 27−36. (28) Ahuja, N.; Easwaran, H.; Baylin, S. B. Harnessing the Potential of Epigenetic Therapy to Target Solid Tumors. J. Clin. Invest. 2014, 124, 56−63. (29) Lin, J. C.; Jeong, S.; Liang, G.; Takai, D.; Fatemi, M.; Tsai, Y. C.; Egger, G.; Gal-Yam, E. N.; Jones, P. A. Role of Nucleosomal Occupancy in the Epigenetic Silencing of the Mlh1 CpG Island. Cancer Cell 2007, 12, 432−444. (30) Portela, A.; Liz, J.; Nogales, V.; Setien, F.; Villanueva, A.; Esteller, M. DNA Methylation Determines Nucleosome Occupancy in the 5′-CpG Islands of Tumor Suppressor Genes. Oncogene 2013, 32, 5421−5428. (31) Kelly, T. K.; Liu, Y.; Lay, F. D.; Liang, G.; Berman, B. P.; Jones, P. A. Genome-Wide Mapping of Nucleosome Positioning and DNA Methylation within Individual DNA Molecules. Genome Res. 2012, 22, 2497−2506. (32) Burri, N.; Shaw, P.; Bouzourene, H.; Sordat, I.; Sordat, B.; Gillet, M.; Schorderet, D.; Bosman, F. T.; Chaubert, P. Methylation Silencing and Mutations of the P14ARF and P16INK4a Genes in Colon Cancer. Lab. Invest. 2001, 81, 217−229. (33) Tsai, H. C.; Li, H.; Van Neste, L.; Cai, Y.; Robert, C.; Rassool, F. V.; Shin, J. J.; Harbom, K. M.; Beaty, R.; Pappou, E.; Harris, J.; Yen, R. W.; Ahuja, N.; Brock, M. V.; Stearns, V.; Feller-Kopman, D.; Yarmus, L. B.; Lin, Y. C.; Welm, A. L.; Issa, J. P.; Minn, I.; Matsui, W.; Jang, Y. Y.; Sharkis, S. J.; Baylin, S. B.; Zahnow, C. A. Transient Low Doses of DNA-Demethylating Agents Exert Durable Antitumor Effects on Hematological and Epithelial Tumor Cells. Cancer Cell 2012, 21, 430− 446. (34) Du, Q.; Wang, Z.; Schramm, V. L. Human DNMT1 Transition State Structure. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 2916−2921. (35) Jia, D.; Jurkowska, R. Z.; Zhang, X.; Jeltsch, A.; Cheng, X. Structure of Dnmt3a Bound to Dnmt3l Suggests a Model for De Novo DNA Methylation. Nature 2007, 449, 248−251.

(36) Miranda, T. B.; Kelly, T. K.; Bouazoune, K.; Jones, P. A. Methylation-Sensitive Single-Molecule Analysis of Chromatin Structure. In Current Protocols in Molecular Biology;John Wiley & Sons: Hoboken, NJ, 2010; Vol. 89, pp 21.17.1−21.17.16, DOI 10.1002/ 0471142727.mb2117s89. (37) Hochedlinger, K.; Plath, K. Epigenetic Reprogramming and Induced Pluripotency. Development 2009, 136, 509−523. (38) You, J. S.; Kelly, T. K.; De Carvalho, D. D.; Taberlay, P. C.; Liang, G.; Jones, P. A. Oct4 Establishes and Maintains NucleosomeDepleted Regions That Provide Additional Layers of Epigenetic Regulation of Its Target Genes. Proc. Natl. Acad. Sci. U. S. A. 2011, 108, 14497−14502. (39) Berman, B. P.; Weisenberger, D. J.; Aman, J. F.; Hinoue, T.; Ramjan, Z.; Liu, Y.; Noushmehr, H.; Lange, C. P.; van Dijk, C. M.; Tollenaar, R. A.; Van Den Berg, D.; Laird, P. W. Regions of Focal DNA Hypermethylation and Long-Range Hypomethylation in Colorectal Cancer Coincide with Nuclear Lamina-Associated Domains. Nat. Genet. 2012, 44, 40−46. (40) Rohde, C.; Zhang, Y.; Reinhardt, R.; Jeltsch, A. BISMA–Fast and Accurate Bisulfite Sequencing Data Analysis of Individual Clones from Unique and Repetitive Sequences. BMC Bioinf. 2010, 11, 230. (41) Pardo, C. E.; Carr, I. M.; Hoffman, C. J.; Darst, R. P.; Markham, A. F.; Bonthron, D. T.; Kladde, M. P. Methylviewer: Computational Analysis and Editing for Bisulfite Sequencing and Methyltransferase Accessibility Protocol for Individual Templates (MAPIT) Projects. Nucleic Acids Res. 2011, 39, e5.

4679


Rational Design of Bisubstrate-Type Analogues as Inhibitors of DNA

Recommend Documents