Fascaplysin-inspired diindolyls as selective inhibitors of CDK4/cyclin D1 Carine Aubry a , A. James Wilson a , Daniel Emmerson a , Emma Murphy a , Yu Yam Chan a , Michael P. Dickens a , Marcos D. García a, * , Paul R. Jenkins a, * , Sachin Mahale b,, Bhabatosh Chaudhuri b,a Department of Chemistry, University of Leicester, Leicester, LE1 7RH, UK b School of Pharmacy, De Montfort University, Leicester, LE1 9BH, UK article info Article history: Received 24 February 2009 Revised 15 June 2009 Accepted 18 June 2009 Available online 12 July 2009 Key words: CDK4 inhibitors Indoles Anti-cancer Fascaplysin abstract We present the design, synthesis and biological activity of a new series of substituted 3-(2-(1H-indol-1- yl)ethyl)-1H-indoles and 1,2-di(1H-indol-1-yl)alkanes as selective inhibitors of CDK4/cyclin D1. The com- pounds were designed to explore the relationship between the connection mode of the indolyl moieties and their CDK inhibitory activities. We found all the above-mentioned designed compounds to be selec- tive inhibitors of CDK4/cyclin D1 compared to the closely related CDK2/cyclin A, with IC 50 for the best compounds 10m and 13a being 39 and 37 lm, respectively. Ó 2009 Published by Elsevier Ltd. 1. Introduction The cyclin dependent kinases (CDKs) are a family of enzymes playing a key role as regulators of the eukaryotic cell division cy- cle, 1 the inhibition of these enzymes by small molecules is an area of major current interest in the anti-cancer field. 2 In particular, re- cent studies have shown that misregulation of CDK4 activity can lead to cancer, suggesting that inhibition of CDK4 could be of ther- apeutic benefice in cancer therapy. 2f,3 Conversely, inhibition of CDK2 which is often postulated as a cancer target, may not be as useful as initially proposed. 4 Consequently, the development of small molecule inhibitors of CDK4 which are specific compared to CDK2, is a very promising area of research. The oxidised diindolyl derivative Indirubin 1, is the active com- ponent from the traditional Chinese medicine ‘Danggui Longhui Wan’. 5 Compound 1 is a selective inhibitor of a range of CDKs com- pared to other protein kinases. 6 Anticancer activity is not only a property of the oxidised indole rings of 1 as shown by the fact that 3,3 0 -diindolylmethane (DIM) 2, a metabolic product of indole-3- carbinol, induces cell cycle arrest in the G 1 phase in human breast cancer cells by activation of the protein p21 (an endogenous inhib- itor of CDKs). 7 The more complex diindolyl compound stauro- sporine 3, 8 is an excellent unspecific inhibitor of many protein kinases, the lack of specificity is attributed to the binding mode of staurosporine closely mimicking that of ATP. 9 In contrast with the examples mentioned above of diindolyl containing compounds 1–3, molecules displaying a potent but non-specific CDK inhibitory activity, the marine natural product fascaplysin 4 10 is one of the few CDK4/cyclin D1-specific inhibitors known, 11 with an IC 50 = 0.55 lM. 12 Fascaplysin 4 has been consid- ered for therapeutic trials, 13 although ultimately its use as antican- cer drug is limited by its toxicity to normal cells, which is a result of its planar structure acting as a DNA intercalating agent. 14 In terms of chemical structure, fascaplysin 4 is a combination of an indole and an oxindole ring connected through a covalent link between, respectively, the 2 and 3 positions of one indole ring and 1 and 2 positions of the second modified indole moiety (a ‘2,2-linkage’ plus a ‘1,3-linkage’, Fig. 2). In contrast, compounds 1–3 show different connection modes (‘linkages’) between the diindolyl (or modified diindolyl) subunits (Fig. 1). In recent times our research group has been interested in the design, synthesis and biological evaluation of non-planar (non- toxic) analogues of fascaplysin 4 as CDK4/ cyclin D1-specific inhib- itors. 15 In our previous work, the structure of 4 was simplified and/ or modified in an effort to establish the key points in the mode of interaction between the designed fascaplysin-based compounds and the CDK4/cyclin D1 complex. Furthermore, the lack of toxicity of some of those fascaplysin-inspired derivatives (with tryptamine and tetrahydro-b-carboline derivatives was also reported, showing the potential of the designed strategy to obtain non-toxic and selective inhibitors of CDK4/cyclin D1. 15f,g Establishing the hypothesis that the mode of connection of the indolyl subunits in fascaplysin 4(and other diindolyl compounds) 0968-0896/$ - see front matter Ó 2009 Published by Elsevier Ltd. doi:10.1016/j.bmc.2009.06.070 * Corresponding authors. Present address: Departamento de Química Fundamen- tal, Universidade da Coruña, Campus da Zapateira, A Coruña 15071, Spain (M.D.G.). Tel.: +44 (0) 116 252 2124; fax: +44 (0) 116 252 3789. E-mail addresses: [email protected](M.D. García), [email protected](P.R. Jenkins), [email protected](B. Chaudhuri). Tel.: +44 (0) 116 250 7280; fax: +44 (0) 116 257 7284. Bioorganic & Medicinal Chemistry 17 (2009) 6073–6084 Contents lists available at ScienceDirect Bioorganic & Medicinal Chemistry journal homepage: www.elsevier.com/locate/bmc
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Fascaplysin-inspired diindolyls as selective inhibitors of CDK4/cyclin D1
Carine Aubry a, A. James Wilson a, Daniel Emmerson a, Emma Murphy a, Yu Yam Chan a, Michael P. Dickens a,Marcos D. García a,*, Paul R. Jenkins a,*, Sachin Mahale b,�, Bhabatosh Chaudhuri b,�
a Department of Chemistry, University of Leicester, Leicester, LE1 7RH, UKb School of Pharmacy, De Montfort University, Leicester, LE1 9BH, UK
a r t i c l e i n f o a b s t r a c t
Article history:Received 24 February 2009Revised 15 June 2009Accepted 18 June 2009Available online 12 July 2009
We present the design, synthesis and biological activity of a new series of substituted 3-(2-(1H-indol-1-yl)ethyl)-1H-indoles and 1,2-di(1H-indol-1-yl)alkanes as selective inhibitors of CDK4/cyclin D1. The com-pounds were designed to explore the relationship between the connection mode of the indolyl moietiesand their CDK inhibitory activities. We found all the above-mentioned designed compounds to be selec-tive inhibitors of CDK4/cyclin D1 compared to the closely related CDK2/cyclin A, with IC50 for the bestcompounds 10m and 13a being 39 and 37 lm, respectively.
� 2009 Published by Elsevier Ltd.
1. Introduction
The cyclin dependent kinases (CDKs) are a family of enzymesplaying a key role as regulators of the eukaryotic cell division cy-cle,1 the inhibition of these enzymes by small molecules is an areaof major current interest in the anti-cancer field.2 In particular, re-cent studies have shown that misregulation of CDK4 activity canlead to cancer, suggesting that inhibition of CDK4 could be of ther-apeutic benefice in cancer therapy.2f,3 Conversely, inhibition ofCDK2 which is often postulated as a cancer target, may not be asuseful as initially proposed.4 Consequently, the development ofsmall molecule inhibitors of CDK4 which are specific comparedto CDK2, is a very promising area of research.
The oxidised diindolyl derivative Indirubin 1, is the active com-ponent from the traditional Chinese medicine ‘Danggui LonghuiWan’.5 Compound 1 is a selective inhibitor of a range of CDKs com-pared to other protein kinases.6 Anticancer activity is not only aproperty of the oxidised indole rings of 1 as shown by the fact that3,30-diindolylmethane (DIM) 2, a metabolic product of indole-3-carbinol, induces cell cycle arrest in the G1 phase in human breastcancer cells by activation of the protein p21 (an endogenous inhib-itor of CDKs).7 The more complex diindolyl compound stauro-sporine 3,8 is an excellent unspecific inhibitor of many protein
Elsevier Ltd.
ento de Química Fundamen-oruña 15071, Spain (M.D.G.).
kinases, the lack of specificity is attributed to the binding modeof staurosporine closely mimicking that of ATP.9
In contrast with the examples mentioned above of diindolylcontaining compounds 1–3, molecules displaying a potent butnon-specific CDK inhibitory activity, the marine natural productfascaplysin 410 is one of the few CDK4/cyclin D1-specific inhibitorsknown,11 with an IC50 = 0.55 lM.12 Fascaplysin 4 has been consid-ered for therapeutic trials,13 although ultimately its use as antican-cer drug is limited by its toxicity to normal cells, which is a resultof its planar structure acting as a DNA intercalating agent.14
In terms of chemical structure, fascaplysin 4 is a combination ofan indole and an oxindole ring connected through a covalent linkbetween, respectively, the 2 and 3 positions of one indole ringand 1 and 2 positions of the second modified indole moiety (a‘2,2-linkage’ plus a ‘1,3-linkage’, Fig. 2). In contrast, compounds1–3 show different connection modes (‘linkages’) between thediindolyl (or modified diindolyl) subunits (Fig. 1).
In recent times our research group has been interested in thedesign, synthesis and biological evaluation of non-planar (non-toxic) analogues of fascaplysin 4 as CDK4/ cyclin D1-specific inhib-itors.15 In our previous work, the structure of 4 was simplified and/or modified in an effort to establish the key points in the mode ofinteraction between the designed fascaplysin-based compoundsand the CDK4/cyclin D1 complex. Furthermore, the lack of toxicityof some of those fascaplysin-inspired derivatives (with tryptamineand tetrahydro-b-carboline derivatives was also reported, showingthe potential of the designed strategy to obtain non-toxic andselective inhibitors of CDK4/cyclin D1.15f,g
Establishing the hypothesis that the mode of connection of theindolyl subunits in fascaplysin 4 (and other diindolyl compounds)
Figure 1. Selected diindolyl-based non-specific inhibitors of CDKs. The diindolyl or dioxindolyl moieties, and their connection modes (‘linkages’), are highlighted in coloursfor clarity.
NH
N
O
Cl
"2,2-linkage"
"1,3-linkage"
4
Figure 2. The natural pigment fascaplysin 4. The modified diindolyl moiety, and itsconnection mode (‘linkage’), is highlighted in colours for clarity.
6074 C. Aubry et al. / Bioorg. Med. Chem. 17 (2009) 6073–6084
is correlated with its specificity among the different CDKs, we de-scribe in full our results on an investigation of this premise and itsapplication to the discovery of selective CDK4/cyclin D1 inhibitorsstructurally inspired in the model compound fascaplysin 4. Hence,two series of type I/II ‘1,3 or 1,1-linked’ diindolyl compounds(Fig. 3) were designed, synthesised and evaluated for their CDK4/cyclin D1 and CDK2/cyclin A activities in order to corroborate ourhypothesis.
2. Results and discussion
2.1. Design
In the preliminary publication on this work,16 we reported thedesign and synthesis of non-planar fascaplysin analogues whichmight have good activity without the toxic side effects of the mod-
NH
N
"1,3-linked" diindolyltype I target compoundsR1, R2 = H, X, Me
R1
R2
NH
"1,3-linkage"
Simplif ication offascaplysin structure
Changeconnec
Figure 3. Strategy planned to convert fascaplysin 4 into ty
el compound. For that purpose, we considered the cleavage of bonda in structure 4, the removal of the ketone at b and changing thedouble bond at c into a single bond leading to type I targetcompounds (Fig. 3). A further modification in the connection modebetween the diindolyl moieties could lead us to the type II‘1,1-linked’ diindolyl target compounds.
In addition, concerning type I compounds, those were designedto explore the effect of a different pattern of substitution on posi-tion 5 and/or 50 of the diindole scaffold; whilst type II compoundswere intended to explore the influence of the size of the carbonchain linker on the CDK4/cyclin D1 inhibitory activity.
Before embarking on the synthesis of potential inhibitors wecarried out in silico docking of the proposed type I/II compounds10e–n (Scheme 1) and 13a–e (Scheme 2) in the active site of ourhomology model of CDK415a,b using the program Gold.17 All thecompounds were predicted to exhibit CDK4 activity binding inthe active site of the enzyme in a mode overlapping fascaplysin4. The predicted binding ‘energies’ (Goldscores) were in all thecases >40, demonstrating that the project could be, a priori,viable.18
2.2. Synthesis of substituted 3-(2-(1H-indol-1-yl)ethyl)-1H-indoles: type I target compounds
Type I target compounds 10e–n (Scheme 1) were synthesisedusing the methodology optimised in our previous work.16 In thismanner, the corresponding indole derivative 5a–d was reactedwith oxalyl chloride affording the consequent indolyl-oxo-acetyl
NNn
"1,1-linked" diindolyltype II target compoundsn = 1,2,3,4,5
N
O
Cl
"1,1-linkage"
4
in thetion mode
c
a
b
pe I/II ‘1,3 or 1,1-linked’ diindolyl target compounds.
N
N
Cl
OO
H
H
N
N
OO
H
R1
R1
R1
5-6 a : R1 = Hb : R1 = Fc : R1 = Cld : R1 = Me
8-10 e : R1 = H ; R2 = Hf : R1 = F ; R2 = Hg : R1 = Cl ; R2 = Hh : R1 = Me ; R2 = Hi : R1 = H ; R2 = Fj : R1 = H ; R2 = Clk : R1 = H ; R2 = Mel : R1 = F ; R2 = F
m : R1 = Cl ; R2 = Cln : R1 = Me ; R2 = Me
NH
R2
(c)
(d)
(e)
R2
N
N
H
R1
R2N
N
H
R1
R2
5a-d
6a-d
7a-d
8e-n
9e-n10e-n
(a)
(b)
7a : R2 = Hb : R2 = Fc : R2 = Cld : R2 = Me
Scheme 1. Synthesis of type I target compounds 10e–n: Reagents and conditions. (a) (COCl)2, anhydrous ether, N2, 0–5 �C, 1.5 h, 80–91%; (b) NaBH3CN 95%, glacial acetic acid,N2, 15–17 �C, 2 h, 85–90%; (c) 6 in anhydrous THF, K2CO3, N2, 7 was added at 0 �C, stir 2 h at rt, 78–90%; (d) LiAlH4, anhydrous THF, anhydrous ether, N2, reflux, 6 h, 57–99%;(e) activated MnO2, CHCl3, reflux, 60 h, 41–84%.
NH N
N
O
O
NN
NN
a
c
11a-e11-13a : n = 0b : n = 1c : n = 2d : n = 3e : n = 4
12a-e
b
13a-e
7a n
nn
Scheme 2. Synthesis of type II target compounds 13a–e. Reagents and conditions:(a) ClCO(CH2)nCOCl, K2CO3, anhydrous THF, N2, 0 �C, 1.5 h, 69–87%; (b) LiAlH4,anhydrous THF, anhydrous ether, N2, reflux, over night, 36–76%; (c) activated MnO2,CHCl3, reflux, 60 h, 30–75%.
C. Aubry et al. / Bioorg. Med. Chem. 17 (2009) 6073–6084 6075
derivative 6a–d in 80–91%.19 Conversely, reduction of 5a–d withcyanoborohydride in glacial acetic acid produced indolines 7a–din good yield,20 Coupling of 6a–d with 7a–d under basic conditionsafforded the indolyl-dihydro-indolyl-ethane diones 8e–n withexcellent yields.21 Reduction of the carbonyl groups,22 and furtheraromatization of the corresponding indoline subunit produced thedesired target compounds 10e–n in 41–84% yields.21
2.3. Synthesis of 1,2-di(1H-indol-1-yl)alkanes: type II targetcompounds
Synthesis of the type II target compounds 13a–e was achievedin a similar fashion than for type I molecules. Two equivalents of
the indoline 7a were reacted with the corresponding acyl chlorideto give the bis-dihydro-indolyl-ethane-diones (11a–e) in 69–87%yield. In some cases the yield could be improved by adding 3 equivof indoline instead of two. Reduction of 11a–e with lithium alu-minium hydride gave the respective bis-dihydro-indolyl-ethane12a–e in 36–76% yield.22 Final oxidation was performed using acti-vated manganese(IV) oxide.21 Under these conditions, the desiredbis-indolyl-alkanes 13a–e were obtained in 67–75% yield.
2.4. Biological evaluation. Selective inhibition of CDK4 versusCDK2
The type I/II target compounds 10e–n and 13a–e, as well as thesynthetic intermediates 8e–n and 12a–e, were assayed in vitro.IC50 values were measured for inhibition of CDK4/cyclin D1 andCDK2/cyclin A using previously reported methods (see Section 4for further details), and the results are shown in Table 1.
The first striking observation is that, as we expected, we haveindeed achieved selectivity for CDK4 over CDK2 in almost all thecompounds tested. Unfortunately, none of the compounds de-signed achieved the nM CDK4 inhibition activity displayed by themodel compound fascaplysin 4 (Table 1, entry 1).
For type I target compounds 10e–n and their synthetic dicar-bonyl intermediates 8e–n (Scheme 2), an increased CDK4 activitycomes out for the 5,50-disubstituted compounds 8l–n/10l–n (Table1, entries 9–11 and 19–21) compared with the monosubstitutedanalogues 8e–k/10e–k (Table 1, entries 2–8 and 12–18). Figure 4shows the most active compound of the type I series, 10m(IC50 = 39 lM, Table 1, entry 20), docked in the active site of ourCDK4 homology model, this indicates a possible hydrogen bondbetween the indole NH and the carbonyl group of valine 96. Theother indole ring is close to the phenylalanines 93 and 159, wherethere is the strong possibility of p-stacking interactions. This fact issupported by our previous findings postulating the existence of a
a CDK4/cyclin D1 assay, using GST-RB152 fusion protein as substrate.b CDK2/cyclin A assay using Histone H1 as substrate (see Section 4 for further
details).
6076 C. Aubry et al. / Bioorg. Med. Chem. 17 (2009) 6073–6084
p-stacking pocket (‘Phe 93 pocket’) within the active site ofCDK4.15b,d,e The parent diindolyl compound 10e is essentially inac-tive, an IC50 of 44 lM is obtained in 10n by the introduction of twomethyl groups, the introduction of one fluorine in 10f and two flu-orines in 10l leads to IC50 values of 50 and 73 lM, respectively,while the introduction of 2 chlorine substituents in 10m gave anIC50 of 39 lM. The electronic effect of these halogens appears tobe markedly improving the p-stacking,23 and hence, IC50 forCDK4 inhibition. In the case of indole–indolinyl compound 9e theIC50 is 122 lM, substitution by methyl, fluoro and chloro groupslead to IC50’s in the range 38–48 lM. Both series of compoundshave broadly similar substituent effects.
Concerning the ‘1,1-linked’ type II target compounds 13a–e(Scheme 2), those presented a clear trend in its CDK4 inhibitory
Figure 4. Compound 10m docked into our CDK4 homology model.
activities. In essence, as the carbon chain increases in length, theinhibition decreases. For the synthetic dicarbonyl intermediates12a–e, due to the lack of solubility of those compounds in the con-ditions required for the inhibition tests, the inhibition activitiescould only be measured for compound 12a, that displayed thesame patron of strong CDK4 selectivity compared with CDK2(IC50’s of 66 and 1230 lM, respectively). The most active com-pound of the series, 13a showed an IC50 of 37 lM and its conforma-tion docked on the active site of the CDK4 is shown in Figure 5.
As for type I target compounds, a strong evidence of a p-stack-ing interaction into the ‘Phe 93 pocket’ was found.
3. Conclusions and further work
We have prepared a library of ‘1,1-linked’ and ‘1,3-linked’ diind-olyl compounds of type I/II (10e–n/13a–e) using a straightforwardsynthetic approach. These compounds (as well as their syntheticdicarbonyl intermediates 8e–n/12a), displaying a selectivity forCDK4/cyclin D1 over CDK2/cyclin A, corroborate our hypothesisthat the connection mode between the indole subunits for thesesimplified non-planar fascaplysin 4 derivatives is important forthe selectivity among the above-mentioned kinases.
Docking of the most active compounds of the series into the ac-tive site of the enzyme, 10m (IC50 = 39 lM) and 13a (IC50 = 37 lM),illustrate our previous findings of the existence of a p-stackingpocket (namely the ‘Phe 93 pocket’) within the active site ofCDK4.15b,d,e
The above-mentioned results could be considered a good start-ing point for the optimisation of the observed biological activity asCDK4 selective inhibitors of these fascaplysin-related non-planarcompounds. Further work is also planned with the aim of synthe-sise different kinds of analogues modifying the nature of the con-nection mode and/or the linker joining the indole subunits aswell as the pattern of substitution on the indolyl subunits.
4. Experimental
4.1. Bio assays
Expression and purification of CDK4/GST-CyclinD1, CDK2/GST-CyclinA and GST-RB152. Fusion proteins of human cyclins A andD1, covalently linked to glutathione s-transferase (GST), were co-expressed with the catalytic subunits CDK2 and CDK4 in Sf-9 insectcells as described previously.24–28
Active enzyme complexes, containing a catalytic subunit boundto GST-Cyclin, were bound to glutathione–agarose columns (Sig-ma, Cat. No. G3907) and were eluted from the columns with re-duced glutathione. The reduced glutathione was removed by
Figure 5. Compound 13a docked into our CDK4 homology model.
C. Aubry et al. / Bioorg. Med. Chem. 17 (2009) 6073–6084 6077
dialysing the enzymes in 10,000 MCO dialysis cassettes (Pierce, CatNo. 66830) with two buffer changes.
The GST-RB152 fusion construct was transformed into the Esch-erichia coli strain BL21(DE3)pLysS (Novagen Cat. No 69451-4). Forexpression of GST-RB152, the cells were induced in the presenceof a final concentration of 4 mM isopropyl-b-thiogalactopyrano-side (IPTG, Invitrogen Cat. No. 15529-091) and were allowed togrow for 4 h in a shaking incubator at 37 �C and 220 rpm. Purifica-tion of the GST-RB152 protein was carried out as described previ-ously.24 Protein estimation was performed using the Bradfordprotein assay (Bio-Rad Laboratories) with bovine serum albumin(BSA) as the standard and the purity of the fusion protein was as-sessed by SDS–PAGE analysis. Proteins were stained with Coomas-sie blue for visualisation.
Kinase assays and IC50 determination. The assay measures thedepletion in ATP concentration as a result of phosphorylation ofretinoblastoma (GST-RB152) and Histone H1 (Upstate BiotechCat. No. 14-155) by CDK4 and CDK2, respectively. The assay wasrun in a 96 well format and all steps in one assay were carriedout in a single white polystyrene plate (Sarstedt, Catalogue No.DPS-134-050A). The compounds were dissolved in DMSO as10 mM stock solutions. Compounds were further serially dilutedin kinase buffer (40 mM Tris (pH 7.5), 20 mM MgCl2, 0.1 mg/mlBSA) in order to obtain the desired concentrations. The kinase as-say was performed in 50 lL of kinase buffer containing 2 lg ofpurified GST-RB152 (in case of Cdk4/GST-cyclin D1) or 3 lg ofHistone H1 (in case of Cdk2/GST-cyclin A) and 6 lM ATP. The phos-phatase and protease inhibitor cocktail containing b-glycerophos-phate, sodium fluoride and sodium orthovanadate in thepresence of reducing agent dithiothreitol was added at the finalconcentrations of 10 mM, 0.1 mM, 0.1 mM and 1 mM, respectively.The assay was initiated by adding 200 ng of active enzyme com-plexes and the plate was incubated for 30 min at 30 �C in a humid-ified incubator. The reaction was stopped by addition of equalvolume of the Kinase Glow ReagentTM (Promega Cat. No. V6711).The luminescence was measured using the Packard Luminometer(Fusion 3.50) and the rate of ATP depletion (rate of reaction) inthe control blank reactions (i.e., without substrate or enzyme)was calculated and used to determine the IC50 concentrations ofcompounds. In case of CDK4/cyclin D1 assay, the two compoundsfascaplysin and flavopiridol with known IC50 values were used tovalidate the assay. For the CDK2/cyclin A assay, roscovitine andflavopiridol were used as standards for the assay.
4.2. Chemistry
NMR spectra were recorded on Bruker DPX 300 (1H,300.13 MHz; 13C, 75.47 MHz; 19F 282.39 MHz) or DPX 400 (1H,400.13 MHz; 13C, 100.61 MHz) spectrometers as indicated. Chemi-cals shifts were measured relative to chloroform (1H d 7.26, 13C d77.0) or dimethylsulfoxide (1H d 2.50, 13C d 39.43) and are ex-pressed in ppm. Coupling constants J are expressed in hertz andthe measure values are corrected to one decimal place. Fast atombombardment (FAB) mass spectra were recorded on a Kratos Con-cept 1H using xenon and m-nitrobenzyl alcohol as the matrix. Elec-trospray (ES) mass spectra were recorded on a Micromass QuattroLC spectrometer. Accurate mass was measure on a Kratos Concept1H spectrometer using peak matching to stable reference peak.Flash column chromatography was carried out using Merck Kiese-gel 60 (230–400 mesh). Dry solvents were provided by a PURESOLVTM system of Innovative Technology Inc.
In some cases the starting materials used were not completelysoluble in the solvent specified; where this is the case the reactionwas carried out in suspension. Amide compounds showed rota-mers characteristics (broad singlets or double peaks) at room tem-perature, not all peaks were duplicated for the atoms of each
molecule. The rotamer ratios were measured from a clear dupli-cated signal in the 1H or 19F NMR. For suitable stable compounds,high temperatures NMR were run. In these conditions the coales-cent signals were described with the corresponding coalescencetemperature.
4.2.1. General procedure for the preparation of indolyl-oxo-acetyl chlorides 6a–d
To a solution of indole derivatives (17.10 mmol) in dry ether(34 mL) at 0 �C, under nitrogen flux, was added oxalyl chloride(19.80 mmol) over a period of 30 minutes. After addition, the mix-ture was stirred for 1 h at 0–5 �C. A yellow precipitate was formedand the mixture was filtered. The solid product was washed withdry ether and dried in vaccuo to give the expected compound.
4.2.2. General procedure for the preparation of dihydro indoles7b–d (7a commercially available)
To a solution of 5-substituted indole (3.71 mmol) in glacial ace-tic acid (9.7 mL) under N2 flux at 15–17 �C was added in one por-tion sodium cyanoborohydride 95% (11.50 mmol). After theaddition, the mixture was stirred for 2 h at 15–17 �C. Water(48.9 mL) was added. The mixture was cooled in an ice bath andNaOH in pellets were added slowly until a strongly basic pH wasobtained. The mixture was extracted with ether (3 � 25 mL). Thecombined ethered layers were washed with water (3 � 30 mL),brine (2 � 30 mL), dried over anhydrous potassium carbonate,and evaporated under reduced pressure to give the corresponding5-substituted dihydro indoles.
6078 C. Aubry et al. / Bioorg. Med. Chem. 17 (2009) 6073–6084
4.2.3. General procedure for the preparation of dihydro-indolylindolyl ethane diones 8e–n
To a solution of the dihydro indole derivative (7a–d)(2.95 mmol) in dry THF (6 mL) under N2 atmosphere, cooled inan ice bath, was added K2CO3 (5.89 mmol). The mixture was stirredunder N2 atmosphere and the indolyl-oxo-acetyl chloride deriva-tive (7a–d) (2.95 mmol) in solution in dry THF (9 mL) under N2
atmosphere was added slowly. The stirring was maintained 2 hat room temperature, and water (59 mL) was added. After 15 minsupplementary stirring the expected compound (9e–n) was fil-tered, washed with water and dried under vacuum many days.
4.2.3.1. 1-(2,3-Dihydro-indol-1-yl)-2-(1H-indol-3-yl)-ethane-1,2-dione 8e. Prepared from 6a and 7a. Pale beige powder. Yield87%. Rotamers observed in ratio 4:1 (from the duplicated tripletsignal (1H) at 4.13 and 4.26 ppm). Melting point 237–238 �C. 1HNMR (300 MHz, DMSO) d (major rotamer) 3.20 (2H, t, J 8.3), 4.13(2H, t, J 8.3), 7.18 (1H, td, J 7.5 and 0.9), 7.31–7.39 (4H, m), 7.58–7.64 (1H, m), 8.20–8.24 (2H, m), 8.31 (1H, s), 12.42 (1H, br s). d(distinct peaks for minor rotamer) 4.26 (2H, t, J 8.6), 6.78–6.81(1H, m), 7.00–7.03 (2H, m), 8.28 (1H, s). 13C NMR (75 MHz, DMSO)d (major rotamer) 28.22 (CH2), 48.49 (CH2), 112.78 (Cq), 113.20(CH), 117.06 (CH), 121.43 (CH), 123.11 (CH), 124.11 (CH), 125.04(CH), 125.60 (CH), 125.71 (Cq), 127.61 (CH), 133.13 (Cq), 137.45(Cq), 138.34 (CH), 142.38 (Cq), 165.15 (Cq), 185.16 (Cq). d (distinctpeaks for minor rotamer) 26.68 (CH2), 47.26 (CH2), 113.38 (CH),123.24 (CH), 124.25 (CH), 126.32 (CH). m/z (FAB+) 291 (M+H)+
(found: C, 74.35; H, 4.69; N, 9.58; MH+ 291.11331. C18H14N2O2 re-quires C, 74.47; H, 4.86; N, 9.65; MH 291.11335).
4.2.3.2. 1-(2,3-Dihydro-indol-1-yl)-2-(5-fluoro-1H-indol-3-yl)-ethane-1,2-dione 8f. Prepared from 6b and 7a. Pale pink solid.Yield 90%. Rotamers observed in ratio 4:1 (from the duplicated sin-glet signal (1H) at 8.28 and 8.33 ppm). Melting point 242–246 �C.1H NMR (300 MHz, DMSO) d (major rotamer) 3.15 (2H, t, J 8.3),4.08 (2H, t, J 8.3), 7.10–7.17 (2H, m), 7.19–7.34 (2H, m), 7.57 (1H,dd, J 9.0 and 4.5), 7.84 (1H, dd, J 9.6 and 2.4), 8.15 (1H, d, 7.8),8.33 (1H, s), 12.41 (1H, br s). d (distinct peaks for minor rotamer)
4.2.4. General procedure for the preparation of dihydro-indolylethyl indoles 9e–n
To a suspension of the lithium aluminium hydride (5.51 mmol)in dry ether (12.5 mL) under N2 atmosphere, was added dropwise asolution of the dihydro-indolyl indolyl ethane dione (8e–n)(0.69 mmol) in dry THF (4.5 mL). The mixture was heated under re-flux 6 h, and cooled to room temperature then to 0 �C. A saturatedsolution of Na2SO4 was added slowly, until the mixture becamemilky white. After filtration the organic and aqueous layers wereevaporated. The crude was redissolve in dichloromethane, washedwith water (4 � 20 mL), dried (Na2SO4) and the organic layers con-centrated to give the title compound. A purification was needed for9g,i,j,l,m,n using a column chromatography of silica gel with thefollowing gradient of eluent: ethyl acetate/cyclohexane (1/9), ethylacetate/cyclohexane (2/8) and ethyl acetate.
4.2.5. General procedure for the preparation of indolyl ethylindoles 10e–n
To a solution of dihydro indolyl ethyl indole (6e–n) (2.05 mmol)in CHCl3 (12 mL) was added activated manganese dioxide(14.3 mmol). The mixture was heated under reflux for 60 h. Aftercooling to room temperature, the crude suspension was filteredthrough Celite and the filtrate was concentrated under vacuum.The crude product was purified by column chromatography on sil-ica gel with the following gradient of solvent ethyl acetate/cyclo-hexane (1/9), ethyl acetate/cyclohexane (2/8), then ethyl acetateto give the title compound (10e–n).
4.2.6. General procedure for the preparation of 1,2-bis-(2,3-dihydro-indol-1-yl)-1,2-dione derivatives 11a–e
To a solution of indoline (22.93 mmol) in dry THF (15 mL) underN2 atmosphere, cooled in an ice bath, was added K2CO3
(45.85 mmol). The mixture was stirred under N2 atmosphere andthe corresponding acyl chloride (11.46 mmol) in solution in dryTHF (16 mL) under N2 atmosphere was added slowly. The stirringwas maintained 2 hours at room temperature, and water(230 mL) was added. After 15 min. supplementary stirring the ex-pected compound (1a–e) was filtered, washed with water anddried with a freeze drier over night.
4.2.6.1. 1,2-Bis-(2,3-dihydro-indol-1-yl)-ethane-1,2-dione 11a.Pink powder. Yield 87%. Three rotamers designated A, B, C were
observed in ratio 15:4:1 (from the triplet signals (1H) at 3.97, 4.14and 4.21 ppm). Melting point 191–192 �C. 1H NMR (300 MHz,CDCl3) d (major rotamer A) 3.13 (4H, t, J 8.4), 4.14 (4H, t, J 8.4),6.84–7.24 (6H, m), 8.15 (2H, d, J 7.5). d (distinct peaks for the minorrotamer B) 3.97 (4H, t, J 8.4), 8.23 (2H, d, J 8.1). d (distinct peaks forthe minor rotamer C) 4.21 (4H, t, J 8.4). 13C NMR (75 MHz, CDCl3) d (major rotamer A) 28.23 (2CH2), 48.39 (2CH2), 117.52 (2CH), 124.90(2CH), 125.11 (2CH), 127.69 (2CH), 132.15 (2Cq), 141.69 (2Cq),161.08 (2Cq). d (distinct peaks for the minor rotamer B) 28.06(2CH2), 46.87 (2CH2), 112.61 (2CH), 124.64 (2CH), 125.91 (2CH),128.09 (2CH), 133.01 (2Cq), 141.45 (2Cq), 161.78 (2Cq). d (distinctpeaks for the minor rotamer C) 26.71 (2CH2), 47.74 (2CH2), 112.51(2CH), 124.74 (2CH), 125.25 (2CH), 127.84 (2CH), 132.08 (2Cq),139.40 (2Cq). m/z (ES+) 293 MH+, 585 (2 M+H)+; (ES�) 291 (M�H)�;(FAB+) 293 (M+H)+ (found: C, 74.08; H, 5.46; N, 9.47; MH+
293.12898. C18H16N2O2 requires C, 73.96; H, 5.52; N, 9.58; MH293.12900).
4.2.6.2. 1,3-Bis-(2,3-dihydro-indol-1-yl)-propane-1,3-dione11b. Very poor solubility in CDCl3. Not soluble in all other D-sol-vents. Tan powder. Yield 86%. Melting point 246–248 �C. 1H NMR(300 MHz, CDCl3) d 3.23 (4H, t, J 8.2); 3.74 (2H, br s); 4.28 (4H, t,
(found: C, 75.93; H, 6.84; N, 7.92; MH+ 349.19165. C22H24N2O2 re-quires C, 75.83; H, 6.94; N, 8.04; MH 349.19160).
4.2.7. General procedure for the preparation of 1,2-bis-(2,3-dihydro-indol-1-yl) derivatives 12a–e
To a suspension of the lithium aluminium hydride(37.45 mmol) in dry ether (85 mL) under N2 atmosphere, wasadded slowly a suspension of the 1,2-Bis-(2,3-dihydro-indol-1-yl)-1,2-dione derivatives 11a–e (4.68 mmol) in dry THF(30.5 mL). The mixture was heated under reflux at least over nightwith a calcium chloride exit, and cooled to room temperature thento 0 �C. A saturated solution of Na2SO4 was added slowly, until themixture became milky white. After filtration, the organic and aque-ous layers were evaporated. The crude was redissolve in dichloro-methane, washed with water (4 � 100 mL), dried (Na2SO4) and theorganic layer concentrated to give crude product. A purificationdone using a column chromatography of silica gel with the follow-ing gradient of eluent: ethyl acetate/petroleum ether (2.5:97.5),ethyl acetate/petroleum ether (5:95), ethyl acetate/petroleumether (10:90), ethyl acetate/petroleum ether (20:80), and ethylacetate gave the title compound 12a–e.
4.2.8. General procedure for the preparation of 1,2-bis-(indol-1-yl) derivatives 13a–e
To a solution of 1,2-Bis-(2,3-dihydro-indol-1-yl) derivatives(12a–e) (1.71 mmol) in CHCl3 (9.7 mL) was added activated man-ganese dioxide (23.94 mmol). The mixture was heated under refluxfor 60 h. After cooling to room temperature, the crude suspensionwas filtered through Celite and the filtrate was concentrated undervacuum. The crude product was purified by column chromatogra-phy on silica gel with the following gradient of solvent ethyl ace-tate/petroleum ether (10:90) and ethyl acetate to give the titlecompound 14a–e.
This work was made possible by generous support from CancerResearch UK. M.D.G. thanks the Xunta de Galicia for financial sup-port (‘programa Isidro Parga Pondal’ and ‘axudas para estadías noextranxeiro’).
6084 C. Aubry et al. / Bioorg. Med. Chem. 17 (2009) 6073–6084
References and notes
1. Alberts, B.; Johnson, A.; Lewis, J.; Raff, M.; Roberts K.; Walter, P. In Molecular CellBiology, 4th ed.; Garland, New York, 2002, Chapter 17.
2. (a) Huwe, A.; Mazitschek, R.; Giannis, A. Angew. Chem., Int. Ed. 2003, 42, 2122;(b) DePinto, W. et al. Mol. Cancer Ther. 2006, 5, 2644; (c) Vander Wel, S. N. et al.J. Med. Chem. 2005, 48, 2371; (d) Honma, T. et al. J. Med. Chem. 2001, 44, 4628;(e) Toogood, P. L. et al. J. Med. Chem. 2005, 48, 2388; (f) Malumbres, M.;Pevarello, P.; Barbacid, M.; Bischoff, J. R. Trends Pharmacol. Sci. 2007, 29, 16.
3. (a) Malumbres, M.; Barbacid, M. Cancer Cell 2006, 9, 2; (b) Yu, Q.; Sicinska, E.;Ahnstrom, M.; Zagozdzon, A.; Kong, Y.; Gardner, H.; Kiyokawa, H.; Harris, L. N.;Stal, O.; Sicinski, P. Cancer Cell 2006, 9, 23; (c) Landis, M. W.; Pawlyk, B. S.; Li, T.;Sicinski, P.; Hinds, P. W. Cancer Cell 2006, 9, 13.
4. Tetsu, O.; McCormick, E. Cancer Cell 2003, 3, 233.5. Hoessel, R.; Leclerc, S.; Endicott, J. A.; Nobel, M. E.; Lawrie, A.; Tunnah, P.; Leost,
M.; Damiens, E.; Marie, D.; Marko, D.; Niederberger, E.; Tang, W.; Eisenbrand,G.; Meijer, L. Nat. Cell Biol. 1999, 1, 60.
6. (a) Hoessel, R.; Leclerc, S.; Endicott, J. A.; Nobel, M. E.; Lawrie, A.; Tunnah, P.;Leost, M.; Damiens, E.; Marie, D.; Marko, D., et al. Nat. Cell Biol. 1999, 1, 60; (b)Marko, D.; Schätzle, S.; Friedel, A.; Genzlinger, A.; Zankl, H.; Eisenbrand, G. Br. J.Cancer 2001, 84, 283; (c) Leclerc, S.; Garnier, M.; Hoessel, R.; Marko, D.; Bibb, J.A.; Snyder, G. L.; Greengard, P.; Biernat, J.; Wu, Y. Z.; Mandelkow, E. M., et al. J.Biol. Chem. 2001, 276, 251.
7. Hong, C.; Kim, H.-A.; Firestone, G. L.; Bjeldanes, L. F. Carcinogenesis 2002, 23,1297.
9. Rialet, V.; Meijer, L. Anticancer Res. 1991, 11, 1581.10. Roll, D. M.; Ireland, C. M.; Lu, H. S. M.; Clardy, J. J. Org. Chem. 1988, 53, 3276.11. Fischer, P. M.; Endicott, J.; Meijer, L. Prog. Cell Cycle Res. 2003, 5, 235.12. Soni, R.; Muller, L.; Ferut, P.; Schoepfer, J.; Stephan, C.; Zumstein-Mecker, S.;
Fretz, H.; Chaudhuri, B. Biochem. Biophys. Res. 2000, 275, 877.13. Subramanian, B.; Nakeff, A.; Tenney, K.; Crews, P.; Gunatilaka, L.; Valeriote, F. J.
Exp. Ther. Oncol. 2006, 5, 195.
14. Hormann, A.; Chaudhuri, B.; Fretz, H. Bioorg. Med. Chem. 2001, 9, 917.15. (a) Aubry, C. A.; Jenkins, P. R.; Mahale, S.; Chaudhuri, B.; Marechal, J.-D.;
Sutcliffe, M. J. Chem. Commun. 2004, 15, 1696; (b) Aubry, C. A.; Wilson, A. J.;Jenkins, P. R.; Mahale, S.; Chaudhuri, B.; Maréchal, J.-D.; Sutcliffe, M. J. Org.Biomol. Chem. 2006, 4, 787; (c) Garcia, M. D.; Wilson, A. J.; Emmerson, D. P. G.;Jenkins, P. R. Chem. Commun. 2006, 24, 2586; (d) Garcia, M. D.; Wilson, A. J.;Emmerson, D. P. G.; Jenkins, P. R.; Mahale, S.; Chaudhuri, B. Org. Biomol. Chem.2006, 424, 4484; (e) Jenkins, P. R.; Wilson, J.; Emmerson, D.; García, M. D.;Smith, M. R.; Gray, S. J.; Britton, R. G.; Mahale, S.; Chaudhuri, B. Bioorg. Med.Chem. 2008, 16, 7728; (f) Mahale, S.; Aubry, C.; Jenkins, P. R.; Maréchal, J.-D.;Sutcliffe, M. J.; Chaudhuri, B. Bioorg. Chem. 2006, 34, 287; (g) Mahale, S.; Aubry,C.; Wilson, A. J.; Jenkins, P. R.; Maréchal, J.-D.; Sutcliffe, M. J.; Chaudhuri, B.Bioorg. Med. Chem. Lett. 2006, 16, 4272.
16. Aubry, C. A.; Patel, A.; Mahale, S.; Chaudhuri, B.; Maréchal, J.-D.; Sutcliffe, M. J.;Jenkins, P. R. Tetrahedron Lett. 2005, 46, 1423.
17. Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R. J. Mol. Biol. 1997, 267,727.
18. In our experience a Goldscore predicted >40 is indicative of a significantbinding affinity; see for instance Refs. 16,17.
19. Shaw, K. N. F.; McMillan, A.; Gudmundson, A. G.; Armstrong, M. D. J. Org. Chem.1958, 23, 1171.
20. Gribble, G. W.; Hoffman, J. H. Synthesis 1977, 859.21. Gribble, G. W.; Pelckman, B. J. J. Org. Chem. 1992, 57, 3636.22. Sadanandan, E. V.; Pillai, S. K.; Lakshimikantham, M. V.; Billimoria, A. D.;
Culpepper, J. S.; Cava, M. P. J. J. Org. Chem. 1995, 60, 1800.23. Cockroft, S. L.; Perkins, J.; Zonta, C.; Adams, H.; Spey, S. E.; Low, C. M. R.; Vinter,
J. G.; Lawson, K. R.; Urch, C. J.; Hunter, C. A. Org. Biomol. Chem. 2007, 5, 1062.24. Phelps, D.; Xiong, Y. Methods Enzymol. 1996, 283, 194.25. Meijer, L.; Borgne, A.; Mulner, O.; Chong, J. P. J.; Blow, J. J.; Inagaki, N.; Inagaki,
M.; Delcros, J. G.; Moulinoux, J. P. Eur. J. Biochem. 1997, 243, 527.26. Boris, S.; Richard, L.; Robert, L. S. J. Biol. Chem. 1997, 272, 33327.27. Grisar, M.; Petty, M. A.; Bolkenius, F. N.; Dow, J.; Wagner, J.; Wagner, E. R.;
Haegele, K. D.; Dejong, W. J. Med. Chem. 1991, 34, 257.28. Biniecki, S.; Modrzejewska, W. Acta. Pol. Pharm. 1981, 38, 407.