Natural-Based Indirubins Display Potent Cytotoxicity toward Wild-Type

Article pubs.acs.org/jnp

Natural-Based Indirubins Display Potent Cytotoxicity toward WildType and T315I-Resistant Leukemia Cell Lines Nicolas Gaboriaud-Kolar,† Vasillios Myrianthopoulos,‡ Konstantina Vougogiannopoulou,† Panagiotis Gerolymatos,† David A. Horne,§ Richard Jove,⊥ Emmanuel Mikros,‡ Sangkil Nam,*,§ and Alexios-Leandros Skaltsounis*,† †

Department of Pharmacognosy and Natural Products Chemistry and ‡Department of Pharmaceutical Chemistry, Faculty of Pharmacy, University of Athens, Panepistimiopolis Zografou, GR-15771 Athens, Greece § Molecular Medicine, Beckman Research Institute, City of Hope Comprehensive Cancer Center, 1500 East Duarte Road, Duarte, California 91010, United States ⊥ Cell Therapy Institute, Nova Southeastern University, 3301 College Avenue, Fort Lauderdale, Florida 33314, United States S Supporting Information *

ABSTRACT: Drug resistance in chronic myelogenous leukemia (CML) requires the development of new CML chemotherapeutic drugs. Indirubin, a well-known mutikinase inhibitor, is the major active component of “Danggui Longhui Wan”, a Chinese traditional medicine used for the treatment of CML symptoms. An in-house collection of indirubin derivatives was screened at 1 μM on wild-type and imatinibresistant T315I mutant CML cells. Herein are reported that only 15 analogues of the natural 6-bromoindirubin displayed potent cytotoxicity in the submicromolar range. Kinase assays in vitro show that eight out of the 15 active molecules strongly inhibited both c-Src and Abl oncogenic kinases in the nanomolar range. Most importantly, these eight molecules blocked the activity of T315I mutant Abl kinase at the submicromolar level and with analogue 22 exhibiting inhibitory activity at the low nanomolar range. Docking calculations suggested that active indirubins might inhibit T315I Abl kinase through an unprecedented binding to both active and Src-like inactive conformations. Analogue 22 is the first derivative of a natural product identified as an inhibitor of wild-type and imatinib-resistant T315I mutant Abl kinases. “Danggui Longhui Wan”, a traditional Chinese medicine (TCM), has been used traditionally to treat the symptoms of chronic myeloid leukemia (CML) and demonstrated efficacy during clinical trials.1 This remedy consists of a mixture of 11 ingredients from traditional Chinese medicinal herbs, including Indigofera tinctoria, Angelica sinensis, Aloe vera, Moschus moschiferus, Phellodendron chinense, Saussurea lappa, Coptis chinensis, Gardenia jasminoides, Scutellaria baicalensis, Rheum palmatum, and Gentiana scabra.2 Among these 11 ingredients, “Qing Dai” (Indigofera tinctoria) is the source of the CLL activity and contains the active constituent indirubin (1, Figure 1). Indirubin (1) appears to exhibit its activity through the inhibition of CDK2 and CDK5,2 two well-known kinases involved in cell cycle regulation, and has shown also efficacy in clinical trials,1 although its low bioavailability stopped its development. Although not used in TCM but being a part of the Mediterranean diet, mollusks in the family of the Muricidae such as Hexaplex trunculus3 or Bolinus brandaris4 accumulate natural bromoindirubins that have shown potent kinase inhibition notably toward GSK-3β kinase.5,6 Over the past decade, a number of structure−activity relationship (SAR) studies were reported confirming the versatility of the bis-indole scaffold to inhibit tyrosine and © 2016 American Chemical Society and American Society of Pharmacognosy

serine/threonine kinases with successful synthetic small molecules against CDKs,7−10 Aurora B and C,11 FLT3,12,13 GSK-3β,14,15 and DYRK2.16 These indirubins bind the kinase ATP binding pocket, thereby displaying substantial cytotoxicity toward various solid tumor cancer cell lines7,9,10,17−21 or even overcoming resistance.22 Findings with A2058 melanoma xenograft mice in vivo models have established E80423 (Figure 1) and 6-bromoindirubin-3′-oxime24 (6-BIO (7), Figure 1) to be efficient inhibitors of STAT5 and JAK/STAT3, respectively, three major tyrosine kinases involved in cell proliferation. The recent discovery of the 6-BIO analogue MLS-2384 (22; Figure 1) as a dual JAK/Src tyrosine kinases inhibitor25 in vivo has propelled this class of derivatives as promising Src inhibitors. However, only a limited number of studies have reported the cytotoxicity of indirubins toward leukemia cell lines.1,2,12,13 Despite the successful introduction of imatinib (Gleevec) for the treatment of CML, a highly efficient therapy for imatinibresistant CML remains to be discovered. Resistance to imatinib has been associated clinically with specific amino acid mutations in the Bcr-Abl kinase domain,26,27 leading to the Received: April 1, 2016 Published: October 11, 2016 2464

DOI: 10.1021/acs.jnatprod.6b00285 J. Nat. Prod. 2016, 79, 2464−2471

Journal of Natural Products

Article

Figure 1. Imatinib and indirubins.

Figure 2. General structure profile of the indirubin derivatives 1−95.

expression of the mutant protein T315I Abl. The secondgeneration CML drugs such as dasatinib and nilotinib that were subsequently discovered as Abl inhibitors were active toward most imatinib-resistant mutants of Bcr-Abl, with the exception of the T315I mutation.26−29 Currently, these drugs are used for these relapsed CML patients. Therefore, development of thirdgeneration CML drugs is still needed to address therapeutically the imatinib-resistant T315I Bcr-Abl mutant CML. The recognition of indirubin (1) as the active metabolite of a TCM remedy and its clinical utilization to assist the treatment of resistant leukemia1,30 have made it a scaffold of choice for the discovery of new active compounds against leukemia. During the past decade, our team has developed a large assembly of indirubin derivatives mainly inspired by 6-BIO (7). Moreover, the activity of 6-BIO derivatives toward the JAK/STAT signaling pathway and the discovery of the Src family as a valuable target for suppressing resistance in leukemia28,29 prompted the determination of the cytotoxic potential of our in-house compound library on leukemia cells. Considering the lack of insight concerning the structural prerequisite for the inhibition of T315I mutant Abl and the high affinity of

indirubins for the ATP binding site, it was expected to discover a new lead compound based on the indirubin scaffold, possibly acting as a dual c-Src/Abl and c-Src/T315I Abl inhibitor and thus able to overcome mutation-related resistance.

■

RESULTS AND DISCUSSION Selection of the Tested Compounds and Biological Screening. Based on previous findings on indirubins interacting with key kinases involved (or potentially involved) in cancer development (GSK-3, Aurora B and C, JAK, c-Src, Akt, and DYRK2), a number of derivatives were selected covering different substitutions such as halogens, nitro, amino, carboxylic acid, ester, aldehyde, cyano, or tetrazole groups on the 5-, 6-, 7-, 4′-, 5′-, or 6′-positions. Derivatives with substituents at the C-4 position have not been explored because of the steric hindrance,5 which prevents their synthesis. As the introduction of an oxime at the C-3′ position was observed to enhance vividly the affinity of indirubin for the kinase active site,2,5 a number of compounds with such substitution were selected. Finally, several derivatives carrying different ethylamine side chain substitutions on the oxime were 2465



Article

Figure 3. Graphical representation of the results of the biological screening.

Figure 4. Chemical structures of the active analogues.

selected due to their benefit in enhancing solubility.15 Figure 2 represents the general structure profile of the indirubin derivatives 1−95 (a full table of these compounds is given in Table S1 of the Supporting Information). In order to detect only the most active compounds, the selected derivatives were evaluated at a low threshold of 1 μM using two CML cell lines, namely, KCL-22 and T315I KCL-22, with the latter expressing the mutant T315I-Abl protein.31 Each test was conducted in triplicate, and the results are summarized in Figure 3 (a full table is given in Table S2 of the Supporting Information). A distinct activity profile emerged, as out of the 95 compounds assayed, 15 led to cell viability below 50% of at

least one cell line (Figure 3 and Table S2, Supporting Information). All the 15 active molecules possessed a bromine at the C-6 position (Figure 4 and Table S1, Supporting Information) and the key C-3′ oxime substituted with an ethylamino side chain (16−30), clearly indicating a chemotype for the exhibition of cytotoxicity against the selected cell lines. On the other hand, 6-BIO (7) and 6-brominated-3′-acetoximes or acetamide oximes exerted cell viability above 50% (6−14). These observations suggest that the combination of a flexible alkylamino side chain at the C-3′ position and a bromine atom at the C-6 position have the most potent effects on cytotoxicity. Bromine or trifluoromethyl groups at the C-7 position were less favorable for cytotoxic activity (36−56) even with the 3′-oxime 2466



Article

3′-ethylamino derivatives (16−22) each showed a very promising profile against the inhibition of leukemia cell viability. However, elongated carbon chains on the piperazine ring slightly decreased or suppressed antileukemic activity against KCL-22 cells (23, 24, 25, 27, 28, 29, 30) and KCL-22 T315I mutant cells (23, 24, 25, 26, 29), respectively, as shown in Table 1. Kinase Inhibition Assay. Active derivatives against KCL22 T315I mutant cells (17, 18, 19, 20, 22, 27, 28, 29) were further characterized using a kinase inhibition assay in vitro including wild-type Abl and T315I mutant Abl kinases. Alternatively, several protein kinases including the Src family of kinases (SFKs), Aurora kinases, and JAK2 have been validated as valuable targets for the development of chemotherapeutic agents for the treatment of T315I mutant CML. SFKs are still activated in imatinib-inhibited BCR-ABL cells.28,29 Thus, the SFKs may assist survival of leukemic cells and eventually allow the appearance of the T315I BCR-ABL mutation.29,32 Therefore, dual Src/Abl inhibitors (or combined administration of specific Abl and Src inhibitors) have emerged as a promising and valuable therapeutic strategy.28,33,34 Bcr-Abl activates JAK2,35 a signal transducer and activator of transcription (STAT) signaling pathway kinase identified as a promising molecular target for eliminating CML resistance.36−38 In addition, Bcr-Abl reportedly activates Aurora A in CML cells through the Akt signaling pathway.39 Aurora A has also become an attractive alternative target40 due to its critical involvement in mutation acquisition in resistance to CML drugs. All these alternative targets have been included in the screening, and the results are summarized in Table 2. All selected analogues showed strong c-Src inhibition with IC50 values ranging from nanomolar (19, 20, 27, 28) to the picomolar (17, 18, 22, 29) range. The identified active molecules represent a novel chemotype for c-Src kinase inhibition. This result was anticipated based on previous studies25 since all the analogues selected for the in vitro kinase assays are analogues of MLS-2384 (22), an already established Jak/Src dual inhibitor. This time, the larger panel of analogues tested allows for the first time to propose a substitution pattern for an enhanced inhibition of c-Src activity in vitro, reaching the subnanomolar range. Indeed, MLS-2438 (a 7-bromoindirubin analogue, namel, 38) has already appeared in a previous study to inhibit c-Src but in much lower inhibitory concentration (IC50 0.2 μM).20 These results display the need to possess bromine in position 6 to obtain a gain in affinity (100- to 400fold). In vitro kinase inhibition (Table 2) showed that those derivatives constitute also a new chemotype for the inhibition

substituted with an ethylamino side chain (39, 40, 42, 55, 56). The absence of substitutions led also to a loss of activity (1, 2). Substitution of the prime core like the natural 6′-bromoindirubin led also to inactive compounds (57 and 58). Finally, dual substitutions on the indole moieties (59−95) including compounds with a substituted C-3′ oxime (61, 62, 65, 71, 73, 78, 80, 88, 90, 93, 95) were not found to contribute favorably for displaying cytotoxicity. Overall, the cell viability results obtained highlight a series of 6-bromoindirubin-3′(ethylamine)oxime derivatives as potent cytotoxic agents against wild-type and resistant human leukemia cancer cells, highlighting the 6-bromo-3′-ethylamino substitution pattern for increased cytotoxicity (16−30, Figure 3). IC50 values were calculated for the compounds displaying cell viability below 50% against the two leukemia cell lines as shown in Table 1. Table 1. Active Indirubins against the Two Leukemia Cell Lines

a

KCL-22

T315l mutant KCL-22

compound

IC50 (μM)

IC50 (μM)

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

0.75 ± 0.05 0.68 ± 0.02 0.42 ± 0.05 0.67 ± 0.02 0.55 ± 0.01 0.78 ± 0.02 0.33 ± 0.02 0.9 ± 0.04 0.5 ± 0.01 0.83 ± 0.02 n.da n.da n.da n.da n.da

n.da 0.66 ± 0.02 0.47 ± 0.04 0.31 ± 0.02 0.23 ± 0.01 n.da 0.43 ± 0.02 n.da n.da n.da n.da 0.5 ± 0.03 0.64 ± 0.06 n.da 0.77 ± 0.04

n.d.: not determined, cell viability above 50%.

The majority of the selected derivatives possessed low IC50 values, and 10 out of the 15 molecules tested had an IC50 value below 0.83 μM for the KCL-22 cell line. Interestingly, eight (17, 18, 19, 20, 22, 27, 28, 30) out of 15 identified active derivatives reduced potently the viability of T315I mutant KCL-22 cells with IC50 values in the submicromolar range. This finding represents an interesting starting point for further development knowing the difficulty of the present marketed drugs to suppress viability of these resistant cells. The 6-bromo-

Table 2. Kinase Inhibitory Activity of Selected 6-Bromo-3′-ethylamino Indirubin Derivatives

a

Abl1

Abl1 T315I mutant

Aurora A

c-Src

JAK2

GSK-3βa

compound

IC50 (nM)

IC50 (nM)

IC50 (nM)

IC50 (nM)

IC50 (nM)

IC50 (nM)

17 18 19 20 22 27 28 29

7.2 10.2 106.5 66.9 0.87 468 3440 34.1

178 194 4564 5940 9.4 >10 000 >10 000 5160

9.7 41.5 79 175 7.1 330 >10 000 14.9

10 000 >10 000

67 23 26 54 3.3 110 200 14

Previously published.15 2467



Article

Figure 5. Different binding poses of 22 in T315I mutant Abl. The three distinct binding modes of 22 are shown in a ball-and-stick representation in an ESP-colored molecular surface of the protein binding site. The binding geometries were obtained by docking calculations using the three available conformational states of Abl kinase and show a differential consistency with experimental data. (A) The mode of the inhibitor bound to the active DFG-in conformation of Abl was reminiscent of the widely studied binding geometry of indirubins to kinases such as GSK3b and the CDKs, being in fair agreement with the experimental SAR data. (B) The mode of the inhibitor bound to a DFG-out inactive Abl conformation afforded a number of binding geometries that were neither converging to a single, consistent orientation nor in accordance with the available experimental binding affinity data. (C) The mode of 22 bound to the Src-like inactive Abl conformation afforded a binding geometry in good agreement with the available SAR data. Consideration of the aforementioned geometry along with the corresponding geometry obtained by the DFG-in conformation provided a more complete overview of the SAR landscape of indirubins inhibiting Abl in both its wild-type and mutant forms.

insight into the structural aspects of Abl inhibition. While there are several structural studies describing the binding geometry of the indirubin scaffold to S/T kinases such as GSK-3β or the CDKs, there is a considerable lack of insight into the interaction of the bis-indole system with any member of the tyrosine kinase subfamily. The major difficulty in theoretically predicting such an interaction lies at the inherent high plasticity of the kinase catalytic domain,41 which among others involves an equilibrium between the active and at least one dominant inactive enzyme state, namely, a DFG-out conformation, facilitating the binding of type-II and -III inhibitors.42 An atypical binding mode of a 7-bromo-substituted indirubin determined for the dual specificity kinase DYRK216 is an excellent example, demonstrating the promiscuity of the kinase fold to translate subtle structural changes of the inhibitor into major induced-fit effects on the protein−inhibitor complex geometry. Moreover, it is well accepted that the equilibrium between active and inactive states does not depend solely on the apoenzyme dynamics or specific PTMs. It is also influenced to a great extent by the type of bound inhibitor, and this is in accordance with the observation that in several instances existing affinity data concerning a series of inhibitors cannot be explained adequately on the basis of a single kinase conformation.43,44 Indeed, the structural aspects of the balance between active and inactive kinase states have been studied principally by X-ray crystallography and, thus, consequently only in terms of determining “end points” or snapshots in an overall static point of view. While this is of use, efforts pursuing the computational simulation of the transition dynamics have been shown to be extremely challenging. These approaches are usually based on sophisticated MD or metadynamics simulations and protocols such as free energy surface mapping based on microsecond-scale trajectories.45,46 In the absence of protein X-ray crystallographic data describing binding of indirubins to Abl, it was attempted to assess the structural basis of the excellent activity of 22 by a computational approach. Extensive docking calculations were performed for a set of active indirubin analogues to both the wild-type and T315I mutant of Abl kinase. Representative X-ray structures of the DFG-in active and DFG-out inactive kinase states were utilized as docking templates. However, by

of wild-type Abl and T315I mutant Abl kinases. Several analogues displayed a potent inhibition of Abl (17, 18, 20, 22, 29), revealing for the first time the affinity of the indirubin scaffold for this target. Three derivatives were able to potently suppress the activity of the mutant T315I Abl kinase (17, 18, 22, Table 2). Besides its broad spectrum of activity, derivative 22 inhibited the mutant kinase at the low nanomolar level with an IC50 value of 9.4 nM. This is the first natural-productderived molecule to display such inhibitory activity, representing a new type of inhibitor for T315I mutant Abl kinase. Thus, several dual inhibitors (17, 18, 20, 22, 29) of c-Src/ wild-type and T315I mutant Abl were identified. Additionally, derivatives possessing a particular affinity for c-Src kinase (27, 28, Table 2) possessed good activities toward the mutant leukemia cell lines (Table 1) that were comparable to the dual inhibitors, suggesting that c-Src inhibitors could suppress survival or growth of imatinib-resistant leukemia cells. However, as a main drawback in the development of kinase inhibitors, our analogues were not clearly selective toward the targeted kinases. A larger screening could eventually provide more insights concerning their overall selectivity. However, the results are encouraging, and the indirubin backbone can now be used as a chemical template in order to develop new types of selective Abl inhibitors. Overall, the “fish and hit” strategy used aiming at the identification of new promising dual inhibitors, based on the chemical scaffold of natural antileukemic indirubin (1), has been fruitful. The ethylpiperazine moiety of 22 is a key substitution requirement to increase potency in kinase inhibition in vitro and reduce cell viability. Although additional experiments are needed to correlate target inhibition and effects on cells, this result is consistent with the strong observed cytotoxicity against T315I mutant KCL-22 CML cells. However, as no obvious structural differences between derivative 22 and the other analogues can explain such an activity, extensive calculations and in silico analysis have been performed. Docking Calculations. The promising activity profile demonstrated by the piperazine-substituted indirubin analogue 22 toward both the wild-type Abl and the T315I mutant raised questions concerning the structural aspects of these interesting effects. Docking-scoring calculations were implemented to gain 2468



Article

state can adopt a conformation distinct from the frequently observed DFG-out inactive one. This fairly different yet relatively underexplored conformation of Abl has been characterized previously as a Src-like inactiveor alternatively “aC-Glu-out”conformation, due to its high resemblance to the corresponding form observed in the past only for members of the closely related Src family of enzymes.47,48 Taking into account this last structural representative of the Abl conformational space in docking calculations afforded reasonable poses that converged to a single binding geometry for the analogues investigated and were highly similar to the binding mode obtained using the active DFG-in template (Figure 5C). More importantly, inspection of those poses showed that they were in good agreement with the existing SAR data. Scoring of the individual compounds resulted in improved agreement with available IC50 data compared to ranking based on the DFG-in template. In part, this could be attributed to the fact that in the Src-like inactive structure of Abl the side chain of D381 is pointing away from the active site of the kinase, precluding any direct interactions with the charged 3′-substituents of the active indirubins and thus providing better consistency with the IC50 data obtained. Moreover, the docked poses were in good agreement with the type-I interaction mode, as was anticipated for these analogues, and, notably, with respect to binding affinities toward the T315I mutant, docking scores obtained by the Src-like template provided correct ranking of the tested compounds. On the basis of the consensus of these docking results, it was thus assumed that indirubins may retain binding affinity for the active and the Src-like inactive conformation of Abl as well, however not for the DFG-out inactive structure. In this study, docking-scoring calculations were performed using the highly efficient Glide SP algorithm49−51 (Schrodinger Inc.). In light of the SAR observations derived from experimental IC50 values as well as docking calculations based on three distinct conformations of Abl, it was concluded that 22 might inhibit the wild-type and the T315I-resistant forms via binding and stabilization of both the DFG-in active and the Srclike inactive conformations but not the DFG-out inactive enzyme state. In this relatively underexplored Src-like conformation, inactivation of the catalytic domain is achieved without flipping of the activation loop DFG motif. In this conformation, the dominant rearrangement observed is that of the N-terminal lobe helix aC, which demonstrates a major outward movement resulting in the disruption of the stabilizing salt bridge between the catalytic lysine (K271Abl) and the aC glutamate (E286Abl). Docking calculations combined with the observation of inhibition data showed that utilization of the Srclike inactive conformation was a prerequisite for obtaining convergence to experimental data. Given that most of the known kinase inhibitors including drugs such as nilotinib and dasatinib target the DFG-out state, the data presented herein suggest that development of indirubin-based inhibitors that target the Src-like state of the enzyme could possibly offer an alternative and elaborate way for addressing drug resistance issues. Even though the modeling results need to be further explored on an experimental basis, the existing data might open a possibility for development of inhibitors rationally targeting specific resistant mutants of the kinase based on the highly tractable bis-indole scaffold. Taken together, 22 holds promise as a member of a new molecular class for developing chemotherapeutic agents for targeting T315I mutant Bcr-Abl protein in the treatment of imatinib-resistant CML.

considering only solutions derived using those two conformations of the enzyme, the obtained results were inconclusive. Concerning the active conformation, reasonable poses were obtained converging to a single binding mode reminiscent of the well-known, experimentally determined kinase binding geometry of indirubin. In the type-I interaction, the inhibitor planar system was bound inside the ATP pocket, while the 3′substituent vector was directed toward the cavity entrance and stabilized by electrostatic interactions with residues of the pocket periphery (Figure 5A). On the other hand, the corresponding poses obtained for the DFG-out template were of high ambiguity. In most docking poses, the aromatic system of the indirubin analogue was packed inside the secondary hydrophobic cavity, which is adjacent to the hinge region being formed by the rearrangement of the activation loop in the DFG-out structures. While the bis-indole ring was buried in this pocket without any of its three pharmacophoric sites that accommodate obvious electrostatic or H-bonding interactions with the protein, the polar and bulky 3′-substituent was directed either toward the hinge region or toward the back side of the hydrophobic cavity, resembling a type-II or -III binding pose, respectively. In these cases, the 3′substituents formed several stabilizing interactions albeit of low consistency with experimental results (Figure 5B). Moreover, binding poses were not converging to specific interaction patterns with the protein and furthermore could not be supported by the available IC50 values. An inconsistency with the experimental data was especially true for docking results derived using the T315I mutant template of the DFG-out enzyme. Indeed, compounds with an extended alkyl chain such as 17, 29, or even 27 were predicted to bind with higher affinities to the T315I protein compared to 22, in direct contrast to experimental affinity data. Having several indications questioning the accuracy of the DFG-out derived docking poses, we tried to revisit and evaluate the corresponding docking results obtained using the active structure. It has been shown that the contribution of the piperazine N4 on binding affinity of the 3′-ethylaminosubstituted indirubins toward kinases such as GSK3β is fairly larger than that of the corresponding N1 heteroatom.15 This SAR notion can be attributed to the more favorable binding orientation that N4 can adopt compared to N1 with respect to the interaction partner, the aspartate residue of DFG motif (D200 in GSK-3β). To sustain a similar trend in affinities in the case of Abl as well, substituents of the C-3′ position should be in direct contact with D381, the corresponding DFG residue in the Abl enzyme. While the docking results were in agreement with this SAR prerequisite, inspection of the experimental binding affinities could not adequately support it. Indeed, compounds not having an isosteric atom to N4 of the heteroatom of 22, such as 17 or 19, retain high affinity for Abl, while 27, carrying an N4 heteroatom, demonstrates a markedly lower affinity than 29 or 22. These data indicate that the 3′-substituent groups are likely not in direct contact with D381 at the indirubin-Abl complexes. Taking into account the previously described docking results, it seems likely that neither the active nor the DFG-out inactive states can fully explain the available SAR of Abl inhibition. To account for this discrepancy, it was reasoned that a different docking template might provide the additional insight needed to better explain the observed SAR and afford a more consistent correlation with the existing experimental affinity data. It is known, not widely though, that Abl at its inactive 2469



Article

flexible mode. Scoring of the docked poses was performed using the default Glidescore scoring function, a modified version of Chemscore as implemented in Glide accounting for a number of energetic terms apart from VdW and electrostatic interactions (GScore = 0.05VdW + 0.15Coul + Lipo + Hbond + Metal + Rewards + RotB + Site).

Starting from natural indirubins and the literature information concerning the use of indirubin in the treatment of CML, we systematically and rationally assessed the bioactivity of a large assembly of indirubins derivatives. The “fish and hit” strategy implemented has been successful in the identification of several analogues based on the natural 6-bromoindirubin scaffold capable of suppressing the viability of wild-type and resistant leukemia cells. New types of dual c-Src/Abl inhibitors based on the natural scaffold have been identified directly in line with actual therapeutic needs. Docking calculations suggest that active indirubins might inhibit T315I Abl kinase through an unprecedented binding to both active and Src-like inactive conformations. It was discovered that indirubin analogue 22 may serve as a new scaffold for the development of Abl and T315I mutant Abl kinase inhibitors. Their selectivity will have to be definitely assessed using a larger panel of kinases. Nevertheless, analogue 22 is the first natural-product-derived molecule to be identified as a T315I-Abl inhibitor with such a binding mode and holds considerable promise for future development.

■

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.jnatprod.6b00285. Full table of substitution patterns; full table of results of the biological screening; IC50 curves of selected compounds (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*Tel: (+1) 626 218 1113. E-mail: [email protected] (S. Nam). *Tel: (+30) 2107274598. Fax: (+30) 2107274594. E-mail: [email protected] (A.-L. Skaltsounis).

EXPERIMENTAL SECTION

Notes

The authors declare no competing financial interest.

Preparation of Compounds 1−95. Compounds 1−95 were taken from our in-house chemical library. They were prepared as stock solutions at 10 or 50 mM in analytical grade DMSO (Carlo Erba) and kept at −27 °C. Cell Lines. KCL-22 CML cells were purchased from the German Collection of Cell Cultures (Braunschweig, Germany). Imatinibresistant human KCL-22 CML cells expressing the T315I mutant BcrAbl (T315I mutant KCL-22 CML) were a gift from Dr. WenYong Chen (City of Hope, CA, USA). KCL-22 and T315I mutant KCL-22 cells were cultured in RPMI-1640 medium containing 10% fetal bovine serum. Cell Viability Assays. MTS assays were performed to determine cell viability with minor modification as described by the supplier (Promega, Madison, WI, USA). Cells were seeded onto 96-well plates (5000 cells/well for solid tumor cells and 10 000 cells/well for blood tumor cells), incubated overnight at 37 °C in 5% CO2, and exposed to indirubins 1−95 for 48 h. Dimethyl sulfoxide (DMSO) was used as the vehicle control. Viable cell numbers were determined by tetrazolium conversion to its formazan dye, and absorbance was measured at 490 nm using an automated ELISA plate reader. Each experiment was conducted in quadruplicate. IC50 values are means. Kinase Assays in Vitro. Kinase assays in vitro were performed with recombinant Abl1, T315I mutant Abl1, Aurora A, c-Src, Jak2, and GSK-3β proteins. Briefly, proteins, freshly prepared substrates, and 33 P-ATP (specific activity 0.01 μCi/μL final) were mixed in reaction buffer (20 mM HEPES pH 7.5, 10 mM MgCl2, 1 mM EGTA, 0.02% Brij35, 0.02 mg/mL BSA, 0.1 mM Na3VO4, 2 mM DTT) in the presence of DMSO as control or indirubins 17−20, 22, or 27−29. The mixtures were reacted for 120 min at room temperature. Samples were transferred onto P81 ion-exchange paper, and filters were extensively washed with 0.75% phosphoric acid. The radioactivities were monitored. IC50 values were determined using GraphPad Prism software. Docking-Scoring Calculations. Protein structures of Abl kinase representative of the possible conformational states of the kinase (DFG-in, DFG-out, Src-like) were downloaded from PDB (pbd codes: 3cs9, 3kf4, 3kfa, 2g2f) and processed using the Protprep module for assignment of the correct bond orders, addition of hydrogens, deletion of water molecules, and the optimization of hydrogen bonds. After preparation, refinement of the structures was performed by a restrained minimization. For docking calculations, Glide software (Schrodinger Inc.) was implemented. Grids were generated for all docking templates using default settings for van der Waals radii scaling as well as a partial charge cutoff used to define the nonpolar atoms of the system. For ligand docking, the Glide-SP algorithm was used in a

■

ACKNOWLEDGMENTS This work was supported by the Commission of the European Community through the INsPiRE project (EU-FP7-REGPOT2011-1, proposal 284460). We thank Dr. W.-Y. Chen (City of Hope) for kindly providing KCL-22 and T315I mutant KCL-22 CML cells. Research reported in this publication included work performed in the Drug Discovery and Structural Biology Core of City of Hope Comprehensive Cancer Center supported by the National Cancer Institute of the National Institutes of Health under award number P30CA033572. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

■

REFERENCES

(1) Xiao, Z.; Hao, Y.; Liu, B.; Qian, L. Leuk. Lymphoma 2002, 43, 1763−1768. (2) Hoessel, R.; Leclerc, S.; Endicott, J. A.; Nobel, M. E. M.; 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, 60−67. (3) Meijer, L.; Skaltsounis, A.-L.; Magiatis, P.; Polychronopoulos, P.; Knockaert, M.; Leost, M.; Ryan, X. P.; Vonica, C. A.; Brivanlou, A.; Dajani, R.; Crovace, C.; Tarricone, C.; Musacchio, A.; Roe, S. M.; Pearl, L.; Greengard, P. Chem. Biol. 2003, 10, 1255−1266. (4) Clark, R. J. H.; Cooksey, C. J. J. Soc. Dyers Colour. 1997, 113, 316−321. (5) Polychronopoulos, P.; Magiatis, P.; Skaltsounis, A.-L.; Myrianthopoulos, V.; Mikros, E.; Tarricone, A.; Musacchio, A.; Roe, S. M.; Pearl, L.; Leost, M.; Greengard, P.; Meijer, L. J. Med. Chem. 2004, 47, 935−946. (6) Vougogiannopoulou, K.; Skaltsounis, A. L. Planta Med. 2012, 78, 1515−1528. (7) Choi, S.-J.; Lee, J.-E.; Jeong, S.-Y.; Im, I.; Lee, S.-D.; Lee, E.-J.; Lee, S. K.; Kwon, S.-M.; Ahn, S.-G.; Yoon, J.-H.; Han, S.-Y.; Kim, J.-I.; Kim, Y.-C. J. Med. Chem. 2010, 53, 3696−3706. (8) Beauchard, A.; Ferandin, Y.; Frere, S.; Lozach, O.; Blairvacq, M.; Meijer, L.; Thiery, V.; Besson, T. Bioorg. Med. Chem. 2006, 14, 6434− 6443. (9) Damiens, E.; Baratte, B.; Marie, D.; Eisenbrand, G.; Meijer, L. Oncogene 2001, 20, 3786−3797. 2470



Article

(10) Lee, J.-W.; Moon, M. J.; Min, H.-Y.; Chung, H.-J.; Park, E.-J.; Park, H. J.; Hong, J.-Y.; Kim, Y.-C.; Lee, S. K. Bioorg. Med. Chem. Lett. 2005, 15, 3948−3952. (11) Myrianthopoulos, V.; Magiatis, P.; Ferandin, Y.; Skaltsounis, A.L.; Meijer, L.; Mikros, E. J. Med. Chem. 2007, 50, 4027−4037. (12) Choi, S. J.; Moon, M. J.; Lee, S. D.; Choi, S.-U.; Han, S.-Y.; Kim, Y.-C. Bioorg. Med. Chem. Lett. 2010, 20, 2033−2037. (13) Han, S.-Y.; Ahn, J. H.; Shin, C. Y.; Choi, S.-U. Drug Dev. Res. 2010, 71, 221−227. (14) Ginzinger, W.; Egger, A.; Mühlgassner, G.; Arion, V. B.; Jakupec, M. A.; Galanski, M.; Berger, W.; Keppler, B. K. Chem. Biodiversity 2012, 9, 2175−2185. (15) Vougogiannopoulou, K.; Ferandin, Y.; Bettayeb, K.; Myrianthopoulos, V.; Lozach, O.; Fan, Y.; Johnson, C. H.; Magiatis, P.; Skaltsounis, A.-L.; Mikros, E.; Meijer, L. J. Med. Chem. 2008, 51, 6421−6431. (16) Myrianthopoulos, V.; Kritsanida, M.; Gaboriaud-Kolar, N.; Magiatis, P.; Ferandin, Y.; Durieu, E.; Lozach, O.; Cappel, D.; Soundararajan, M.; Filippakopoulos, P.; Sherman, W.; Knapp, S.; Meijer, L.; Mikros, E.; Skaltsounis, A.-L. ACS Med. Chem. Lett. 2013, 4, 22−26. (17) Ferandin, Y.; Bettayeb, K.; Kritsanida, M.; Lozach, O.; Polychronopoulos, P.; Magiatis, P.; Skaltsounis, A.-L.; Meijer, L. J. Med. Chem. 2006, 49, 4638−4649. (18) Kim, S.-A.; Kim, Y.-C.; Kim, S.-W.; Lee, S.-H.; Min, J.-J.; Ahn, S.-G.; Yoon, J.-H. Clin. Cancer Res. 2007, 13, 253−259. (19) Kritsanida, M.; Magiatis, P.; Skaltsounis, A.-L.; Peng, Y.; Li, P.; Wennogle, L. P. J. Nat. Prod. 2009, 72, 2199−2202. (20) Liu, L.; Kritsanida, M.; Magiatis, P.; Gaboriaud, N.; Wang, Y.; Wu, J.; Buettner, R.; Yang, F.; Nam, S.; Skaltsounis, L.; Jove, R. Cancer Biol. Ther. 2012, 13, 1255−1261. (21) Braig, S.; Kressirer, C. A.; Liebl, J.; Bischoff, F.; Zahler, S.; Meijer, L.; Vollmar, A. M. Cancer Res. 2013, 73, 6004−6012. (22) Braig, S.; Bischoff, F.; Abhari, B. A.; Meijer, L.; Fulda, S.; Skaltsounis, L.; Vollmar, A. M. Biochem. Pharmacol. 2014, 91, 157− 167. (23) Nam, S.; Buettner, R.; Turkson, J.; Kim, D.; Cheng, J. Q.; Muehlbeyer, S.; Hippe, F.; Vatter, S.; Merz, K.-H.; Eisenbrand, G.; Jove, R. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 5998−6003. (24) Liu, L.; Nam, S.; Tian, Y.; Yang, F.; Wu, J.; Wang, Y.; Scuto, A.; Polychronopoulos, P.; Magiatis, P.; Skaltsounis, L.; Jove, R. Cancer Res. 2011, 71, 3972−3979. (25) Liu, L.; Gaboriaud, N.; Vougogianopoulou, K.; Tian, Y.; Wu, J.; Wen, W.; Skaltsounis, A.-L.; Jove, R. Cancer Biol. Ther. 2014, 15, 178− 184. (26) O’Hare, T.; Walters, D. K.; Stoffregen, E. P.; Jia, T.; Manley, P. W.; Mestan, J.; Cowan-Jacob, S. W.; Lee, F. Y.; Heinrich, M. C.; Deininger, M. W.; Druker, B. J. Cancer Res. 2005, 65, 4500−4505. (27) Shah, N. P.; Tran, C.; Lee, F. Y.; Chen, P.; Norris, D.; Sawyers, C. L. Science 2004, 305, 399−401. (28) Martinelli, G.; Soverini, S.; Rosti, G.; Baccarani, M. Leukemia 2005, 19, 1872−1879. (29) Li, S. Int. J. Biochem. Cell Biol. 2007, 39, 1483−1488. (30) Chen, F.; Li, L.; Ma, D.; Yan, S.; Sun, J.; Zhang, M.; Ji, C.; Hou, M. Leuk. Res. 2010, 34, e75−e77. (31) Yuan, H.; Wang, Z.; Gao, C.; Chen, W.; Huang, Q.; Yee, J.-K.; Bhatia, R.; Chen, W. J. Biol. Chem. 2010, 285, 5085−5096. (32) Pene-Dumitrescu, T.; Smithgall, T. E. J. Biol. Chem. 2010, 285, 21446−21457. (33) Crespan, E.; Radi, M.; Zanoli, S.; Schenone, S.; Botta, M.; Maga, G. Bioorg. Med. Chem. 2010, 18, 3999−4008. (34) Azam, M.; Nardi, V.; Shakespeare, W. C.; Metcalf, C. A.; Bohacek, R. S.; Wang, Y.; Sundaramoorthi, R.; Sliz, P.; Veach, D. R.; Bornmann, W. G.; Clarkson, B.; Dalgarno, D. C.; Sawyer, T. K.; Daley, G. Q. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 9244−9249. (35) Danial, N. N.; Rothman, P. Oncogene 2000, 19, 2523−2531. (36) Deininger, M. W. N.; Goldman, J. M.; Melo, J. V. Blood 2000, 96, 3343−3356.

(37) Samanta, A. K.; Lin, H.; Sun, T.; Kantarjian, H.; Arlinghaus, R. B. Cancer Res. 2006, 66, 6468−6472. (38) Samanta, A.; Perazzona, B.; Chakraborty, S.; Sun, X.; Modi, H.; Bhatia, R.; Priebe, W.; Arlinghaus, R. Leukemia 2011, 25, 463−472. (39) Yang, J.; Ikezoe, T.; Nishioka, C.; Udaka, K.; Yokoyama, A. Int. J. Cancer 2014, 134, 1183−1194. (40) Yuan, H.; Wang, Z.; Zhang, H.; Roth, M.; Bhatia, R.; Chen, W. Y. Carcinogenesis 2012, 33, 285−293. (41) Huse, M.; Kuriyan, J. Cell 2002, 109, 275−282. (42) Gavrin, L. K.; Saiah, E. MedChemComm 2013, 4, 41−51. (43) Mol, C. D.; Fabbro, D.; Hosfield, D. J. Current Opin. Drug Discovery Dev. 2004, 7, 639−648. (44) Zhao, Z.; Wu, H.; Wang, L.; Liu, Y.; Knapp, S.; Liu, Q.; Gray, N. S. ACS Chem. Biol. 2014, 9, 1230−1241. (45) Shan, Y.; Arkhipov, A.; Kim, E. T.; Pan, A. C.; Shaw, D. E. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 7270−7275. (46) Sutto, L.; Gervasio, F. L. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 10616−10621. (47) Levinson, N. M.; Kuchment, O.; Shen, K.; Young, M. A.; Koldobskiy, M.; Karplus, M.; Cole, P. A.; Kuriyan, J. PLoS Biol. 2006, 4, e144. (48) Lovera, S.; Sutto, L.; Boubeva, R.; Scapozza, L.; Dölker, N.; Gervasio, F. L. J. Am. Chem. Soc. 2012, 134, 2496−2499. (49) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.; Klicic, J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.; Perry, J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. J. Med. Chem. 2004, 47, 1739−1749. (50) Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye, L. L.; Pollard, W. T.; Banks, J. L. J. Med. Chem. 2004, 47, 1750− 1759. (51) Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.; Greenwood, J. R.; Halgren, T. A.; Sanschagrin, P. C.; Mainz, D. T. J. Med. Chem. 2006, 49, 6177−6196.

2471


Natural-Based Indirubins Display Potent Cytotoxicity toward Wild-Type

Recommend Documents