Chemistry and catalysis advances in organometallic chemistry and catalysis
Download 11.05 Mb. Pdf ko'rish
|
- Bu sahifa navigatsiya:
- Figure 45.1
- L2a– L2c and L3
- Figure 45.5
605 606 MODERATE CDK INHIBITORS TO POTENTIAL ANTITUMOR DRUGS Soon it became clear that DNA alone is possibly neither the only nor the best target for platinum (and possibly other) anticancer drugs. This becomes evident when one takes a closer look at DNA lesions caused by cisplatin and carboplatin. Although the platinum core after activation by hydrolysis in the cells and the application area are the same for both drugs, the ratio and the nature of DNA adducts formed is different. While cisplatin predominantly cross-links neighboring guanine bases present on the same DNA strand ( ∼65%), carboplatin generates by far more fragments that are believed to originate from two guanines that have another base in between ( ∼36%) [6]. As a consequence, carboplatin-caused DNA damages are processed differently by DNA repair enzymes than cisplatin-induced lesions, resulting in different active concentrations and (dose-limiting) side effects of the two closely related drugs. Moreover, focusing on DNA damage as the only mode of action does not entirely explain the varying sensitivity and resistance patterns of different cancer cell types. On the basis of those findings, scientists expanded their field of research. On the one hand, much effort was put into the development of platinum(II) complexes with different ligands [7], allowing for the fine-tuning of physicochemical properties and thereby affecting the biological distribution of the drugs in the organism. This led to the development of the third-generation platinum drug, oxaliplatin [8], which is approved worldwide against colon cancer. The next generation of platinum anticancer agents will presumably consist of octahedral platinum(IV) complexes, which are considered inert prodrugs in the healthy tissues and get activated in, or in the surroundings of, the tumors [9]. On the other hand, chemists remembered Ehrlich’s “magic bullet” concept and restarted the search for more selective chemotherapeutics. Nowadays, targeted chemotherapy is believed to be one of the most promising ways to fight cancer [10–12]. A variety of rewarding targets, mostly enzymes or proteins, have been identified. Among them are Ribonucleotide Reductases. This class of enzymes catalyzes the reduction of ribonucleotides to deoxyribonucleotides, the building blocks of DNA. Interference with this process leads to a depletion of the dNTP pool, slowing down DNA synthesis and reparation, two mechanisms affected in many types of cancer [13]. Thus inhibition of the enzyme can be regarded as a therapeutic option that can be achieved by using organic molecules [14] or gallium compounds [15, 16].
protein. (De)phosphorylation is a major on/off-switch for a vast array of proteins and particularly interesting regarding cancer, as tyrosine kinases were found to be overexpressed in many solid tumors. Chemotherapeutics that target these enzymes are erlotinib, an epidermal growth factor receptor (EGFR) inhibitor, and imatinib, which is also approved by the FDA (US Food and Drug Administration) as an anticancer agent [17].
an opportunity for correction of irregularly twisted DNA [18]. This enzyme is overexpressed in various cancer types, making it a major target for therapy, either by poisoning the DNA topoisomerase complex or by catalytic inhibition of the enzyme [19, 20]. Three drugs of this class are widely used in combination anticancer therapy, namely, doxorubicin, etoposide, and mitoxantrone [21, 22]. Within the past years, several organometallic complexes have been found to exhibit topoisomerase-inhibiting properties [23–25]. Hormone Receptors. Certain types of cancer, mainly breast cancer, endometrial cancer, and prostate cancer, show a growth dependency on sexual hormones such as estradiol and testosterone [26]. As those cells overexpress hormone receptors, down-regulation of the hormone concentration is a powerful tool in the fight against those cancer types. Among the chemotherapeutics approved for treatment is tamoxifen [27, 28]. In recent years, also related complexes, that is, the ferrocifenes [29] were developed for that purpose. For detailed information on this class of molecules, please see Chapter 42. Cyclin-Dependent Kinases (cdks). These enzymes play a crucial role in cell cycle progression. They facilitate the transfer of a phosphate group to threonine and serine residues of their respective target, strictly regulated by activators (cyclins) and different inhibitors. Activated or inhibited at certain cell cycle checkpoints, cdks trigger the cell to proceed with or cease the cycle, respectively [30]. There are various cdk/cyclin combinations known today, of which some appear to be essential for all cells, while others seem to regulate the cell cycle of highly specialized cells. In many human cancers, overactivity of cdks has been identified as one of the mechanisms underlying the pathological hyperproliferation. Therefore, inhibition of cdk activity may offer a therapeutic option for fighting cancer [31]. Flavopiridol (1, Fig. 45.1), a semisynthetic flavonoid, was the first cdk inhibitor that entered clinical trials as an anticancer drug [32]. This review is an account of the investigation of a class of compounds, namely, indolobenzazepines, which attracted our attention as likely cdk inhibitors. TOWARD METAL-BASED INDOLOBENZAZEPINES AS POTENTIAL ANTICANCER DRUGS 607 H N HN O R O O Cl OH HO N (a) (b)
(c) HO H H N HN R 1 2a: R = H 2b: R = OCH 3 O 3a: R = H 3b: R = Br 3c: R = NO 2 1 2 3 4 5 6 7 8 9 10 11 12
Flavopiridol (a); indolobenzazepinones described by MacPhillamy et al. (b); paullone core structure (3a), kenpaullone (3b) and alsterpaullone (3c) with atom numbering scheme (c). 45.2 INDOLOBENZAZEPINES AS MODERATE cdk INHIBITORS Substituted 7,12-dihydroindolo[3,2-d][1]benzazepin-6(5H)ones 2a and 2b were first reported in 1958 by MacPhillamy et al. [33] as the last intermediate in an 11-step synthesis of 4-ethyl-5,6,7,12-tetrahydro-2-methylindolo[3,2-d][1]benzazepine. In 1992, Kunick reported the preparation of six indolo[3,2-d][1]benzazepin-6(5H)ones [34] in a three-step procedure. The crucial step was the decarboxylation reaction of 2,3-dihydro-5-hydroxy-2-oxo-1H-benz[b]azepin-4-carboxylic acid ethyl ester into 3,4-dihydro-1H-benz[b]azepin-2,5-dione, which made possible the subsequent Fischer indole synthesis, with formation of the desired 7,12-dihydroindolo[3,2-d][1]benzazepin-6(5H)ones. Kenpaullone 1
Cancer Institute (NCI; NCI60 screen) with an antiproliferative activity profile similar to that of Flavopiridol [35]. Subsequent inhibitory experiments revealed that kenpaullone 3b is indeed a moderate ATP-competitive inhibitor of cdk1/cyclin B, cdk2/cyclin A and cdk5/p35 [36], with the lactam moiety as an important binding motif [37, 38]. Therefore, 3b is a good candidate for further structural optimization. Structure–activity relationship studies have shown that electron- withdrawing substituents in position 9 are favorable for the cdk inhibitory activity, provided that the lactam moiety remains underivatized. Alsterpaullone, the 9-nitro-derivative, was found to be the best inhibitor of the paullone family [39]. At the same time, a number of paullones prepared by modification of the lactam unit have shown high antiproliferative activity, which does not parallel their cdk inhibitory profile [40]. Therefore, other intracellular targets have been suggested for this class of compounds. In a screen on 28 commercially available kinases, kenpaullone and alsterpaullone selectively inhibited cdks, gsk-3 and Lck, the latter being a member of the Src family of tyrosine kinases (abbreviation derived from sarcoma) [40]. Gsk-3 enzymes belong to the family of serine/threonine kinases as well, with structural similarity to cdks. In screens for the gsk-3 inhibitory potential of paullones, alsterpaullone was again found to be the best performing derivative [40].
One of the limitations encountered in the development of paullones as antitumor drugs was, and still is, their low aqueous solubility and bioavailability. Given the experience of our group in the development of metal-based anticancer agents [41–45], we expected that coordination to metal ions can mitigate the imposed limitations, leading to improved pharmacokinetics and, maybe by synergistic effects with metal ions, to more favorable pharmacodynamic properties. Coordination to metal ions can result in (i) stabilization of certain, sometimes unusual, ligand geometries, (ii) redox activity, (iii) increased solubility, (iv) enhanced lipophilicity, (v) improved cellular uptake, (vi) different modes of action, and (vii) synergistic effects from metal and ligand(s) [46–48]. Biological effects exhibited by metal complexes, but unobserved for free ligands were also reported [49, 50]. Certain long-term side effects can be diminished or even avoided as metal complexes can break down and the metal ion can then interact with the organism [51–53]. Although the library of paullone derivatives prepared by Kunick et al. was large, metal-based derivatives were not known. This is primarily because the original paullones do not contain binding sites able to chelate metal ions. The only possibility is the binding via the lactam or thiolactam unit with 1 Named in honor of Dr. Kenneth Paull, contributor to the NCI60 screen. 608 MODERATE CDK INHIBITORS TO POTENTIAL ANTITUMOR DRUGS Figure 45.2 IC 50 (50% cell proliferation inhibition concentrations) of the hydrazino-paullone (HP), ligand L1a, and its gallium complex [Ga(L1a-H) 2 ]Cl on different cancer cell lines [54]. formation of thermodynamically nonstable four-membered metallocycles. Therefore, our primary aims were the design and synthesis of paullones, which contain binding sites able to accommodate specific metal ions and the study of the effect of metal coordination on the antiproliferative activity, cdk inhibitory activity, and cell cycle progression. The first indolobenzazepine ligand was designed for coordination to gallium(III). By reacting the methyl thioimidate [54] with hydrazine hydrate, the corresponding hydrazino-paullone (HP) was obtained, which was further converted into potentially tridentate ligands (L1a and L1b, Fig. 45.3) by condensation reaction with 2-hydroxybenzaldehyde [54] or 2-acetylpyridine [55], respectively. Given our experience with complexes of α-N-heterocyclic thio- and semicarbazones, gallium(III) complexes with paullone ligands, having N 2 O (L1a) [54] and N 3 (L1b) [55] donor sets, were prepared. Hydrazino-paullone HP, ligand L1a and its gallium(III) complex [Ga(L1a-H) 2 ]Cl, were tested for antiproliferative activity in a panel of 14 human cancer cell lines. The results, summarized in Fig. 45.2, clearly show a significantly improved cytotoxicity of L1a over its precursor HP by factors of 4–25. Upon complexation, activity against cancer cells could be further improved by a factor of 1.5–18, depending on the cell line (Fig. 45.2). However, the low aqueous solubility and hydrolytic instability of the positively charged cationic complex prompted us to search for other derivatives and, in particular, those able to coordinate to ruthenium(II). We expected that d 6 low spin ruthenium(II) species would be kinetically more resistant to hydrolysis. As a binding unit to be attached to the original paullone, α-picolylamine, an excellent bidentate chelator for ruthenium(II), was chosen. As we were interested in the elucidation of the influence of electron-withdrawing substituents in position 9 of the paullone scaffold on the antiproliferative activity, ligands with bromo (L2b) and nitro (L2c) substituents, as well as the unsubstituted ligand (L2a), were prepared by reaction of thiolactam derivatives with α-picolylamine in tetrahydrofuran (THF) (Fig. 45.3) in excellent yields and high purity without using mercury salts to promote the reaction or activation via thiolactam methylation. In addition, the location of the binding site was changed to position 9 by reducing the nitro group to the amino function, which was further reacted with 2-formylpyridine, to yield a paullone ligand with an intact lactam moiety (L3) [56]. Our attempts to develop a paullone ligand bearing a stable 2,2,6,6-tetramethylpiperidin-1-yl)oxyl (TEMPO) free-radical unit (L2d) were also successful [57]. Other potentially tridentate ligands were prepared by using pyridoxal (L1c) [58] and 2-formylpyridine (L1d) [59] instead of 2-hydroxybenzaldehyde or 2-acetylpyridine, respectively. The first ruthenium(II) complexes with paullone ligands were prepared by reaction of cis-[RuCl 2 (DMSO)
4 ] with L2a and L2b, respectively. Depending on the reaction conditions, complexes with metal-to-ligand ratio 1 : 1 and 1 : 2 were isolated. Bis-ligand complexes [Ru II Cl(DMSO)(L2a) 2 ]Cl and [Ru II Cl(DMSO)(L2b) 2 ]Cl were tested for antiproliferative activity and TOWARD METAL-BASED INDOLOBENZAZEPINES AS POTENTIAL ANTICANCER DRUGS 609 H N HN R 1 N NH 2 N HN R 1 S H N HN R 1 S H N HN R 1 O H N O O R 1 HN NH
2 EtO
O O OEt NH 2 O OEt H N O HO OEt O H N HN N H N HN N H N HN O N R 1 H N HN N N
R 1 N N R 2 X Y R 1 R 4 R 3
N(R 2
2 L4a L4b L4c R 1 H H Br R 2 H CH 3 CH 3 L2a L2b L2c L2d L2e R 1 H Br NO 2 C(O)NH-TEMPO C(O)NH-TMP L1a L1b L1c L1d L1e R 1 Br Br Br Br H R 2 H CH 3 H H CH 3
C-OH N
N N
3 H H CH 3 H H R 4 H H CH 2 OH H H
CH CH
CH CH
Synthesis of paullone ligands with different location of their binding sites. TEMPO = (2,2,6,6-tetramethylpiperidine-1-yl)- oxyl. TMP, 2,2,6,6-tetramethyl-piperidine. showed IC 50 values in the low micromolar concentration range [46]. Electrophoretic mass titration studies revealed both compounds to alter the cDNA secondary structure, leading to relaxation of the supercoiled form of the plasmid. UV–vis and NMR investigations of ruthenium(II) bis-ligand complexes did not show any evidence for binding to 5 -GMP, a model system for nucleotide binding. In addition, ethidium bromide was displaced with increasing amounts of the complexes. Thus, intercalation rather than monofunctional binding to nucleotides was assumed to be responsible for the alteration of DNA secondary structure. Another paullone metal coordination approach was based on using metal-arene compounds as an ideal metal scaffold to deliver biologically active compounds into the cell, as they combine proper lipophilicity to cross the cell membrane with sufficient aqueous solubility. In addition, binding to the arene stabilizes ruthenium and osmium in oxidation state 2 + [56]. As ligands to be attached to the metal-arene fragment, L2a–L2c and L3 with similar but differently placed binding sites (Fig. 45.3) were chosen. By exploiting the μ-chlorido bridge-splitting reaction of [M II (p-cymene)( μ-Cl)Cl] 2 (M = Ru, Os) with L2a– L2c and L3 complexes [M II (p-cymene)(L2a– L2c)Cl]Cl and [M II (p-cymene)(L3)Cl]Cl were prepared. Interestingly, two types of differently shaped crystals were discovered by inspection of the product, [Ru II (p-cymene)(L2c)Cl]Cl, under the microscope. They were large enough to be separated mechanically and studied by NMR spectroscopy. In addition, the crystals were suitable for X-ray diffraction analysis, which revealed the presence of two different isomers: E (block-shaped crystals) and Z (stick-shaped crystals). The structures of both isomers are shown in Fig. 45.4. The E/Z isomerization occurs at the exocyclic C 6 =N
double bond. This isomerization was found to be solvent and concentration dependent. By investigation of the process at different temperatures and usage of the Eyring plot, activation parameters were determined [56]. Cell culture experiments showed IC 50 values in the micromolar concentration range, whereas binding to 5 -GMP remained rather poor (10%). Ruthenium(II)- and osmium(II)-arene complexes with L2a, a ligand modified at the lactam unit, were 610 MODERATE CDK INHIBITORS TO POTENTIAL ANTITUMOR DRUGS Figure 45.4 Structures of the cationic E- (a) and Z-isomers (b) of [Ru II (p-cymene)(L2c)Cl] + . Non-labeled atoms represent carbons. Hydrogen atoms were omitted for clarity. (See insert for color representation of the figure.)
Structures of the protonated paullone ligand HL1a + with an intramolecular hydrogen bond O–H ···N (a) and ruthenium- arene complex cation [Ru(p-cymene)(L1a)Cl] + (b). Non-labeled atoms represent carbon atoms. (See insert for color representation of the figure.) found to reduce the rate of DNA synthesis more efficiently than the corresponding complexes with L3, the ligand modified in position 9 of the original paullone. In addition, cell cycle perturbations were more pronounced in the case of the L2a-based complexes. These results indicate that other intracellular targets might be responsible for the observed cytotoxicity [56]. The next type of complex [M II (p-cymene)(L1a,c)Cl]Cl was synthesized by reacting [M II (p-cymene)( μ-Cl)Cl] 2 (M = Ru, Os) with potentially tridentate ligands L1a and L1c, originally designed for coordination to gallium(III). Both compounds act as bidentate ligands in [M II (p-cymene)(L1a,c)Cl]Cl with the phenolic oxygen not involved in the coordination to the central metal ion, as shown in Fig. 45.5. All four complexes exhibited similar cytotoxicity in the low micromolar concentration range. Both ruthenium complexes showed 5 -GMP binding behavior remarkably different from that for the previously described ruthenium–arene complexes with L2a and L3 ligands, forming several adducts, with the most abundant reaching 47%. Osmium(II)–arene complexes with L1a and L1c ligands did not react with 5 -GMP [58]. Biological studies have shown that the complexes are able to induce apoptosis and, to a lower extent, also necrosis. In addition, inhibition of cdk2/cyclin E was found in a cell-free assay, but this is rather unlikely to be decisive for the antiproliferative activity of the complexes, given weak effects on cell cycle progression [60]. Thus, it appears that induction of apoptosis constitutes a new mode of action for paullone-based metal complexes. Attachment of a TEMPO free-radical unit to a paullone ligand with a carboxylic acid group in position 9, by using a well-developed active ester coupling protocol from peptide synthesis, afforded ligand L2d, with enhanced aqueous solubility [57]. Its molecular structure is shown in Fig. 45.6. The TEMPO-bearing ligand showed high cytotoxicity, with IC 50 values in either the nanomolar concentration range in A549 (non-small-cell lung carcinoma), CH1 (ovarian carcinoma), SW480 (colon carcinoma), N87 (gastric cancer), and SK-Mel-28 (skin melanoma) cell lines, or in the low micromolar concentration range |
ma'muriyatiga murojaat qiling