|Home | About | Journals | Submit | Contact Us | Français|
Research into the anti-tumor properties of chalcones has received significant attention over the last few years Two novel large series of α-bromoacryloylamido chalcones 1a–m and 2a–k containing a pair of Michael acceptors in their structures, corresponding to the α-bromoacryloyl moiety and the α,β-unsaturated ketone system of the chalcone framework, were synthesized and evaluated for antiproliferative activity against five cancer cell lines. Such hybrid derivatives demonstrated significantly increased anti-tumor activity compared with the corresponding amino chalcones. The most promising lead molecules were 1k, 1m and 2j, which had the highest activity toward the five cell lines. Flow cytometry with K562 cells showed that the most active compounds resulted in a large proportion of the cells entering in the apoptotic sub-G0–G1 peak. Moreover, compound 1k induced apoptosis through the mitochondrial pathway and activated caspase-3.
Among currently identified anti-tumor agents, chalcones represent an important class of natural small molecules useful in cancer chemotherapy.1 Chalcones (1,3-diaryl-2-propen-1-ones, Chart 1) are known to exhibit antimitotic properties caused by inhibition of tubulin polymerization by binding to the colchicine-binding site.2 Chemically, they are open-chained molecules consisting of two aromatic rings linked by a three-carbon enone fragment. Several research groups have shown that the s-cis conformation of chalcones is important for their biological activity.3 The double bond of the enone system is the essential moiety for chalcones as anti-tumor agents.4 It has been reported that hydrogenation or bromination across the carbon–carbon double bond or its transformation into the corresponding epoxide dramatically reduces chalcone activity.5 Their simple structure and the ease of preparation make chalcones an attractive scaffold for structure–activity relationship (SAR) studies, and a wide number of substituted chalcones have been synthesized to evaluate effects of various functional groups on biological activity.1
The pyrroloiminoquinone cytotoxic alkaloids Discorhabdin A6 and Discorhabdin G7 are characterized by the presence of an α-bromoacryloyl alkylating moiety of low chemical reactivity, an unusual feature for cytotoxic compounds. In fact, α-bromoacrylic acid is not per se cytotoxic (IC50 for L1210 cells being greater than 120 μM).8 The same moiety is present in a series of potent anticancer distamycin-like minor groove binders, for example, PNU-166196 (brostallicin), which is currently undergoing Phase II clinical trials.9 PNU-166196 is an α-bromoacrylamido derivative of a four-pyrrole distamycin homologue ending with a guanidino moiety (Chart 1).9
The reactivity of the α-bromoacryoyl moiety has been hypothesized to be based on a first-step Michael-type nucleophilic attack, followed by a further reaction of the former vinylic bromo substituent alpha to the carbonyl, leading successively either to a second nucleophilic substitution or to beta elimination.10
Furthermore, several studies confirmed that the α,β-unsaturated ketone system of chalcones acts as a Michael acceptor, suggesting that alkylation of the β-position of the reactive enone system by biological nucleophiles may be one mechanism by which antiproliferative activity is exerted in vitro.11
The observations that both the chalcone and the α-bromoacryloyl group can act as trapping agents of cellular nucleophiles led us to prepare and evaluate two novel and unusual classes of synthetic conjugates with general formulae 1a–m and 2a–k (Chart 1), incorporating these two moieties within their structures. If such processes occur, the α-bromoacryloylamido chalcone derivatives 1a–m and 2a–k, characterized by the presence of two potential sites for electrophilic attack on cellular constituents, should be more potent than the corresponding compounds containing only one nucleophilic center.
While compounds 1a–m were designed to evaluate the SAR arising from different substitutions (both electron-releasing and electron-withdrawing groups), as well as different positions on the phenyl ring, for hybrids 2a–k we focused on the synthesis of methoxylated derivatives with one or more methoxy groups at different positions on the aryl moiety.
Synthesis of derivatives 1a–m and 2a–k was carried out by the general methodology shown in Scheme 1. Nitrochalcones 5a–m and 6a–k were synthesized in high yields (81–95%) by the Claisen–Schmidt aldol condensation of 4-nitrobenzaldehyde or 4-nitroacetophenone with the corresponding appropriately functionalized acetophenones 3a–m or benzaldehydes 4a–k, respectively, in the presence of 50% w/v aqueous solution of sodium hydroxide.12 Coupling constants (J) from the proton nuclear magnetic resonance (1H NMR) spectra clearly indicated that derivatives 5a–m and 6a–k were both geometrically pure and were exclusively trans (E) isomers (JtransC=C = 15–16 Hz). The cluster of compounds 6a–k may be referred to as ‘reversed chalcones’, whereby the carbonyl and ethylene groups present in the series 5a–m are interchanged. Aminochalcones 7a–m and 8a–k were generated from the the corresponding nitro derivatives 5a–m and 6a–k by reduction with iron in a refluxing solution mixture of 37% HCl in water and ethanol (1:2.5 v/v).
Finally, the hybrid compounds 1a–m and 2a–k were prepared, in acceptable yields (54–68%), by the condensation of α-bromoacrylic acid with aminochalcones 7a–m and 8a–k, respectively. This condensation was performed using an excess (2 equiv) both of 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide hydrochloride (EDCl) and 1-hydroxy-1,2,3-benzotriazole (1-HOBt), in dry DMF as solvent at room temperature and with identical reaction times (18 h).
Tables 1 and and22 summarize the antiproliferative effects of α-bromoacryloylamido chalcones 1a–m and 2a–k, respectively, against the growth of murine leukemia (L1210), murine mammary carcinoma (FM3A), human T-lymphoblastoid (Molt/4 and CEM) and human cervix carcinoma (HeLa) cells, using aminochalcones 7a–m and 8a–k as reference compounds.
In general, the α-bromoacryloylamido chalcones 1a–m and 2a–k, were 10–100-fold more active than the corresponding amino chalcones 7a–m and 8a–k, respectively, demonstrating that the presence of an α-bromoacryloyl moiety significantly enhanced antiproliferative activity. The compounds displaying the greatest potency were 1k (p-EtO), 1m (2-thienyl) and 2j (o, m, p-3OMe), with IC50 values of 0.24–0.63, 0.52–0.68, 0.55–0.68, 0.73–0.84 and 0.34–0.75 μM against the L1210, FM3A, Molt4, CEM and Hela cell lines, respectively. The bioisosteric replacement of phenyl (1a) with a thiophene ring (1m) greatly increased activity, with IC50 values 0.25–0.73 μM versus 3.1–5.0 μM against the five tumor cell lines. A positive effect was also observed for the 1-naphthyl derivative 1l, which had IC50 values of 1.1–3.1 μM, as compared with 1a.
The data presented in Tables 1 and and22 demonstrate effects of different substituents on the phenyl rings linked to the carbonyl and at the β-position of the enone system, respectively. With compounds 1a–k (Table 1), the substitution pattern significantly affected potency. The introduction of either electron-releasing (ERG) or electron-withdrawing (EWG) substituents (derivatives 1b–k) enhanced antiproliferative activity as compared with the unsubstituted analogue 1a, and there was no clear difference between them. Ignoring the derivative with a p-I (1f), the greatest enhancement of activity occurred with the bulkier substituents, p-Br (1e) and p-EtO (1k).
Specifically with the para-halogen substituted 1bcef (Table 1), activity increased in the following order: Br (1e)>Cl (1c), F (1b)>I (1f). Insertion of a second chlorine atom, to yield the m,p-dichloro derivative 1d, resulted in about a twofold reduction in activity. Turning to the effects of an ERG on the phenyl moiety, we found that replacement of p-methyl (1g) with a p-methoxy (1h) group caused only a minor improvement in anti-proliferative activity. Moreover, a twofold reduction in activity was observed by shifting the methoxy group from the para to the meta position (1i). Replacement of the p-MeO group (1h) with a p-EtO moiety (1k) caused a twofold increase in potency, while the o,m,p-trimethoxy derivative 1j was only slightly more active than 1h.
In the case of the second series (Table 2), compounds 2b–j, with methoxy substituents on the phenyl ring, had antiproliferative activity that, in general, was comparable with that of the unsubstituted 2a. Similarly, the antiproliferative activity of the N,N′-dimethylamino derivative 2k was not greatly different from that of the unsubstituted 2a.
A comparison of the antiproliferative activity of compounds with the same substituent on the phenyl ring (1a vs 2a, 1h vs 2b, 1i vs 2c and 1j vs 2i), the ‘reversed’ α-bromoacryloylamido derivatives 2a, 2b, 2c and 2i generally had higher IC50 values than did the corresponding analogues 1a, 1h, 1i and 1j.
Chalcones are known to block cells in the G2-M phase of the cell cycle, which is consistent with their ability to inhibit tubulin assembly.2 We therefore determined whether these hybrid molecules behaved in a similar manner.
The most active compounds (1ekm and 2abdej) were evaluated for their inhibitory effects on tubulin polymerization and on the binding of [3H]colchicine to tubulin.13 With the exception of 2d, all tested compounds were ineffective as inhibitors of tubulin assembly (IC50 >40 μM). However, all compounds had poor solubility in the assay medium (all precipitated in the 40 μM assays). 2d showed relatively weak activity as a tubulin polymerization inhibitor, with an IC50 value of 8.3 ± 0.9 μM, and, at 50 μM, 2d weakly inhibited the binding of 5 μM [3H]colchicines to tubulin (37 ± 10% inhibition). These data make it unlikely that the antiproliferative activity of these hybrid compounds results from a direct interaction with tubulin and that it is unlikely that they act as microtubule depolymerizing agents.
To confirm this idea, we next examined the effects of the most active compounds on the cell cycle by performing a flow cytometric analysis of K562 human chronic myelogenous leukemia cells, which are usually employed by our research group to determine the alteration of cell cycle parameters following exposure to anti-tumor compounds.14 Cells were cultured for 72 h in the absence or presence of each compound at the concentrations summarized in Table 3, and the cells were then stained with propidium iodide (Fig. 1, panels A and B). Unlike chalcones, which induced a substantial recruitment of cells into the G2-M phase of the cell cycle, derivatives 1ekm and 2abdej uniformly caused a decrease in the proportion of cells in all phases of the cell cycle (G0–G1, S and G2-M), with a proportionate increase in the number of apoptotic cells in the sub-G1 phase region of the histogram. The absence of the typical increase in G2-M cell number, together with the minimal effect on tubulin assembly, makes it unlikely that microtubule disruption underlies the potent apoptotic effect caused by these hybrid compounds.
The induction of apoptosis was confirmed by the annexin V test15 with compounds 2j and 2k (Fig. 1, panels C and D). The increase in annexin V-positive cells demonstrated activation of the apoptotic pathway by these compounds. Figure 2C shows the extensive binding of annexin V to 2j-treated cells, and Figure 2D shows how time and concentration of 2k affect the proportion of cells that become positive for the marker. Similar observations were made with the other most active compounds.
Impairment of mitochondrial function, is an early event in the executory phase of programmed cell death in different cell types, and it occurs as the consequence of a preliminary reduction of the mitochondrial transmembrane potential (Δψmt).16,17 Early Δψmt disruption results from an opening of mitochondrial permeability transition pores, and this permeability transition triggers the release of apoptogenic factors, such as apoptosis inducing factor and cytochrome c, which in turn lead to later apoptotic events.16,17
We used the lipophilic cation 5,5′,6,6′-tetrachlo-1,1′.3,3′-tetra-ethylbenzimidazol-carbocyanine (JC-1) to monitor changes in Δψmt induced by 1k. The method is based on the ability of this fluorescent probe to enter selectively into the mitochondria, and its color changes reversibly from green to orange as membrane potential increases.18 This property is due to the reversible formation of JC-1 aggregates upon membrane polarization. Aggregation causes a shift in the emitted light from 530 nm (i.e., emission by JC-1 monomers) to 590 nm (emission by JC-1 aggregates) following excitation at 490 nm.
K562 cells were treated with compound 1k for 24 and 48 h at different concentrations. As shown in Figure 2 (panels A and B), compound 1k induces substantial mitochondrial depolarization in a time- and concentration-dependent manner. The disruption of the Δψmt is associated with the appearance of sub-G1 cells and with the marked increase in the percentage of annexin V-positive cells.
Mitochondrial membrane depolarization is associated with mitochondrial production of reactive oxygen species (ROS).19,20 To investigate the effects of 1k on the production of oxygen species during apoptosis, we utilized the fluorescence indicator hydroethidine (HE), which fluoresces if reactive oxygen species are generated.21 As shown in Figure 2 (panel C), there was an increase in cells producing ROS that closely paralleled the increase in cells with low Δψmt, a function of both the concentration of 1k and treatment time.
We also evaluated the damage caused by ROS in mitochondria by assessing the oxidation state of cardiolipin, a phospholipid restricted to the inner mitochondrial membrane. We used 10 N-nonyl acridine orange (NAO) a fluorescent probe which is independent of mitochondrial permeability transition.22 The dye interacts stoichiometrically with intact, non-oxidized cardiolipin. Somewhat unexpectedly, cells did not show reduction in NAO fluorescence, but rather, a marked increase, especially after 48 h of treatment with 1k (Fig. 3 panels A and B). This effect suggests an increase in cardiolipin content as a consequence of increased mitochondrial mass. This has been observed in some tumor cell lines following treatment with herbimycin A,23 genistein,24 and the acronicyne derivative S23906-125 and following oxidative stress.26
Several caspases have been shown to be key executioners of apoptosis mediated by various inducers.27 Caspase-3, in particular, is essential to the propagation of the apoptotic signal after exposure of cells to many DNA-damaging agents and other anticancer drugs.27,28 We therefore determined whether caspase-3 is involved in the apoptosis induced by compound 1k. We used a monoclonal antibody specifically for the active fragment of caspase-3. As shown in Figure 4, with 0.6 and especially 1.25 μM 1k there was increased activated caspase-3 after 24 h of treatment, and a further increase after 48 h.
In summary, our plan to optimize the potency of a single Michael acceptor by providing multiple sites of reactivity towards cellular nucleophiles led us to synthesize two series of α-bromoacryloylamido chalcones 1a–m and 2a–k. These compounds were derived from the hybridization of two kinds of Michael acceptors, corresponding to the α,β-unsaturated ketone system of chalcone and the α-bromoacryloyl moiety. While the amino chalcones 7a–m and 8a–k showed weak or no antiproliferative activity against five different cancer cell lines, their conversion into the corresponding α-bromoacryloylamido derivatives 1a–m and 2a–k was accompanied by a 10–100-fold increase in potency. More noteworthy was that compounds 1k (52%), 2b (68%), 2d (83%) and 2j (57%) induced over half the cells to enter a sub-G1 population as compared with untreated control cells (11%). As demonstrated with 1k, their mechanism of action appears to induce apoptosis mediated by the involvement of mitochondria and by the activation of caspase-3. Thus, the mitochondrial apoptotic pathway plays a major role, in generation of the sub-G1 cell population. Further studies to clarify additional details of the molecular mechanism of action of these compounds and the selectivity to inhibit the growth of cancer cells are underway. On the other hand, the most active compounds should be analyzed for effects on normal primary normal human cells of different hystotype, in order to determine potential therapeutic windows supporting further studies on experimental tumor-bearing animals finalized to verify possible in vivo anticancer activity.
Financial support was provided by GOA (Krediet No. 05/19) of the K.U. Leuven. The technical assistance of Mrs. Lizette van Berckelaer was gratefully acknowledged. R.G. and C.L.C. are funded by AIRC and University of Granada, respectively.
Detailed biological protocols, synthesis and spectroscopic data for compounds α-bromoacryloylamido chalcones 1a–m and 2a–k. General procedures for the synthesis of nitrochalcones 5a–m/6a–k and aminochalcones 7a–m/8a–k are available. Supplementary data associated with this article can be found, in the online version, at doi: 10.1016/j.bmcl.2009.02.038.