Scientific paper
Design and Evaluation of Biological Activity of Diazenecarboxamide-extended Cisplatin and Carboplatin Analogues
Nikolina Stojanovic,1 Damijana Urankar,2 Anamaria Brozovic,1 Andreja Ambriovic-Ristov,1 Maja Osmak1* and Janez Kosmrlj2*
1 Division of Molecular Biology, Ru^r Bo{kovi} Institute, Zagreb, Croatia; 2 Faculty of Chemistry and Chemical Technology, University of Ljubljana, Slovenia * Corresponding author: E-mail: Maja.Osmak@irb.hr (M.O.), janez.kosmrlj@fkkt.uni-lj.si (J.K.)
Received: 29-01-2013 Dedicated to Professor Slovenko Polanc on his 65th birthday.
Abstract
Construction of a library of structurally diverse diazenecarboxamide-extended cis-[Pt(2-picolyl-1,2,3-triazole)Cl2] and cis-[Pt(propan-1,3-diamine)CBDCA] (CBDCA = 1,1-cyclobutanedicarboxylate) complexes 1—4 is described. These compounds retain oxidative properties of parent diazenecarboxamides against glutathione as demonstrated by NMR spectroscopy and high resolution mass spectrometry experiments. Cytotoxic activity of 1—4 was investigated against human cervical carcinoma HeLa cells. Four library members were found to possess moderate cytotoxic activity. Some model compounds were also examined, returning [PtCl2L2] (L = 1-(2-picolyl)-4-phenyl-1_ff-1,2,3-triazole) as the most potent under this investigation with IC50 of 19.05 |M, comparable to that of cisplatin (IC50 = 16.3 |M).
Keywords: Cisplatin, carboplatin, diazenecarboxamide, glutathione, anti-cancer activity
1. Introduction
Cisplatin, cis-[diamminedichloridoplatinum(II)] (Figure 1), has been clinically approved for many years to be used as a single agent or in combination with other anti-cancer drugs to treat solid tumors including lung carcinoma, cervical carcinoma, head and neck tumors, as well as carcinoma of ovaries, bladder and testis.1 Despite the fact that it is one of the most effective and commonly used agents, nephro-, neuro- and ototoxicities are the main side effects of this drug.2 Its major obstacle to successful chemotherapy, significantly reducing the effectiveness of the cancer treatment, is development of cisplatin-resistance in tumors. Cisplatin-resistance is multifactorial, i.e. it is based on several molecular mechanisms.3 Among others, it can emerge as a result of increased inactivation of cispla-tin by intracellular thiol-containing molecules, such as glutathione (GSH).3a,b Increased GSH level may cause resistance by binding/inactivating cisplatin, enhancing DNA repair or reducing cisplatin-induced oxidative stress.3c Glutathion-S-transferase (GST) may augment the
resistance by catalyzing GSH-cisplatin binding. The increased expression of GST along with elevated GSH level in resistant tumor cells suggests that the increased inacti-vation of cisplatin contributes significantly to the resistance phenotype.3ab
In so advanced diseases such as cancer it is unlikely that a single mono functional targeted drug would be sufficient to cure the cancer. Combined drugs (cocktails) that impact multiple targets simultaneously are better at controlling complex disease systems and are the standard of care in cancer treatment.4 In order to further improve the efficiency of using a two-drug cocktail, one approach involves the use of the so-called hybrid molecules, which comprises the incorporation of two drugs covalently bound into a single entity that combines pharmacological effects of independently acting drugs.4'5 For such molecules, it has been even demonstrated that superior synergistic effects can be achieved that is beyond mere simultaneous administration of two separate agents.6 An important example of hybrid molecules in
Carboplatin
Figure 1. The structure of cisplatin, carboplatin and selected hybrid molecules.
the field of metal-based anticancer drugs includes the development of platinum complexes having bioactive carrier ligands, which are capable to revert resistances that tumor cells acquire during chemotherapy. This has been exemplified by ethacraplatin (A, Figure 1) in which GST-inhibitory ethacrynic acid was attached to the platinum,7 as well as platinum complexes of nitrofuran car-bohydrazide (B, Figure 1),8 which is a human thioredo-xin reductase inhibitor.
We have shown previously that diazenecarboxami-des9 (Figure 2) are cytotoxic for different tumor cell lines and act synergistically with cisplatin.10 These compounds are powerful modulators of intracellular GSH concentration.10^0^ In a combined (cocktail) diazenecarboxami-de-cisplatin treatment, the GSH-depleting diazenecarbo-xamides led to the reversal of the acquired tumor resistances. The synergistic effect with cisplatin was shown on different tumor-cell lines as well as some cisplatin-resistant sublines.10c,1,n Besides the GSH-depleting activity, in selected examples an activation some alternative cell-death pathways has been observed, indicating that GSH may not be the only cellular target of diazenecarboxamides.10h
Figure 2. Diazenecarboxamides
The above results prompted design and screening an array of different diazenecarboxamide-extended platinum complexes. For these hybrid molecules a dual targeting that combines the advantages of diazenecarboxami-de and cisplatin (and carboplatin) structural motif is expected. However, the synthesis of a diverse library of li-gand-arm functionalized diazenecarboxamides that are capable of binding to platinum(II), as well as the corresponding complexes, is expected to be cumbersome and associated with redesign of synthetic routes for any specific type of the molecule. To increase the efficacy of the preparation of such a library of compounds we12-15and others16 developed strategies that are based on recently discovered copper-catalyzed azide-alkyne cycloaddition protocol,17,18 also known as "Click chemistry", as outlined in Scheme 1.19
Scheme 1. Schematic representations of the linear strategy versus the convergent strategy.
Tiioiii**» ^ F)in7pnpr,arl*ir»\'amirlp-p'ytpnrlprl nlntirnim mmnlpYPS
2. Results and Discussion
We have designed and prepared four types of diaze-necarboxamide-extended platinum complexes, which differ in i. functionality R, ii. ligand arm that coordinates to platinum and consequently iii. platinum(II) coordination environment, and iv. linker that joins the diazenecarboxa-mide and complex parts together (Figure 3). Briefly, it has been demonstrated that the oxidative properties of diaze-necarboxamides can be modulated through the choice of group R attached directly to the diazene -N=N- moiety,20 whereas specific hydrolytic and coordinative properties towards DNA of tumor cells can be tuned by ligand variation around platinum center. Finally, through the choice of the linker, one can in principle adjust the lipophilicity of the compound, an important pharmacological characteristic that correlates with cellular uptake of a drug.21
The syntheses of compounds 1-4 were accomplished through the strategies shown in Scheme 1.12-15 More specifically, the synthetic routes are outlined in Scheme 2.
Following the linear strategy shown in Scheme 1 propargyl functionalized diezenecarboxamides 5 were by copper-catalyzed azide-alkyne cycloaddition ("Click chemistry") with picolyl azide transformed into the corresponding picolyl-triazole derivatives 6 (Scheme 2).12,15 Although the power of this strategy lies in potential combinatorial approach to diazenecarboxamides, conjugated with different ligand-arms, by using different organoazides, at this point of the research we only used picolyl azide as a "click" partner. In the final stage, the ability of picolyl-tria-zole group at 6 to serve as bidentate ligand22,23 was employed to access the desired platinum-diazenecarboxamide conjugates 1 and 2 of cis-[Pt(diamine)Cl2] structure.13
Concerning the coordination chemistry, this is a "classical" strategy, which considers the ligand to metal coordination at the very last step. The major drawback of this linear strategy is a potential redesign of synthetic routes for any specific type of the ligand. Alternatively, convergent strategy relies on the post-functionalization of metal complexes with the introduction of specific features
Figure 3. Diazenecarboxamide-extended platinum complexes.
Scheme 2. Selected details showing the preparation of diazenecarboxamide-extended platinum complexes 1-4 using the strategies presented in Scheme 1.
at the very last step (Scheme 1). Following this convergent strategy, "click" reaction between propargyl functionali-zed diezenecarboxamides 5 and azide-tagged platinum(II) complex 7 allowed post-functionalization directly into the metalated "click" cycloadducts 3 and 4 of structure cis-[Pt(diamine)CBDCA] (CBDCA = 1,1-cyclobutanedicar-boxylate) (Scheme 2).14
The oxidative properties of diazenecarboxamide moiety in complexes 1-4 towards GSH is preserved as demonstrated by the following experiment. Compound 3b and slight excess of GSH (3.7 mol equiv.) were dissolved in a mixture of DMF-d7 and D2O at 23 °C. The progress of the reaction shown in Scheme 3 was monitored by 1H NMR spectroscopy. Figure 4 demonstrates clean and fast
transformation of diazenecarboxamide 3b and GSH into semicarbazide 3bRED and GSSG. The resulting reaction mixture after the NMR experiment was analyzed by ESI+ LC-MS spectrometry, which confirmed the presence of semicarbazide 3bRED (m/z found for C26H30F3N8O6Pt+ [M + H]+ = 802.1886) and GSSG (See Experimental).
To ascertain the oxidative properties of 3b under biological conditions, it was incubated with GSH (4.4 mol equiv.) in a mixture of DMSO and growth medium supplemented with 10% fetal serum (Experimental). Similar conditions were used for determination of 3b cytotoxicity (vide infra). Analysis of the reaction mixture by ESI+ LC-MS revealed nearly quantitative conversion of 3b into 3bRED within 3.5 hours.
Scheme 3. Oxidation of GSH with diazenecarboxamide 3b into GSSG and semicarbazide 3bRED.
Figure 4. Monitoring the course of the reaction from Scheme 3 between 3b (6.7 mM) and GSH (24.7 mM) in a mixture of DMF-^7 (0.50 mL) and D2O (0.25 mL) at 23 °C by 'H NMR spectroscopy. Spectra shown after a) 6 min, b) 11 h, and c) 43 h of the reaction. Well separated resonances of 3b (o), 3bRED (*), and GSSG (A) are indicated. Diagram d) shows the progress of the reaction as determined by integration of resonances of 3b and 3bRED.
Compounds 1-4 were tested using MTT assay for their in vitro anticancer activity against human cervical carcinoma HeLa cells (see Experimental section for details). The results for compounds 1 and 2 (Table 1) with cis-[diamminedichloroplatinum(II)] coordination sphere were compared to cisplatin (Table 2). In analogy, the results for compounds 3 and 4 with cis-[Pt(diamine)
In attempt to experimentally evaluate this idea of combining separate components into hybrid molecules 1-4, cytotoxic activities of compounds 6a, 6b, 7 and 8 were determined. The results are summarized in Table 2. cii-[Pt(2-azidopropan-1,3-diamine)CBDCA] (7) proved to be inactive. Unexpectedly, uncoordinated picolyl-tria-
Table 1. Cytotoxic activity of complexes 1-4 against cervical carcinoma HeLa cells expressed as IC50 values (|^M) obtained after 72 h incubation.
R	1	IC50	2	IC50	3	IC50	4	IC50
4-NO2-C6H4-	1a	>300	2a	87	3a	>1000		
4-CF3-C6H4-	1b	300	2b	277	3b	>1000		
C6F5-	1c	>300						
4-CI-C6H4-	1d	>300	2d	>300	3d	614 ± 161.9		
4-F-C6H4-			2e	>300			4e	649.6
4-CH3-C6H4-							4f	>1000
4-CH3O-C6H4-							4g	990
CBDCA] structure were compared to carboplatin. Compounds 2a and 2b having strongly electron-withdrawing p-nitrophenyl and p-(trifluoromethyl)phenyl substituents attached to the diazene -N=N- group were quite cytotoxic. Interestingly, this was not the case for 3a and 3b from the carboplatin-like series, where 3d and 4e were the most active compounds.
Table 2. Cytotoxic activity of cisplatin, carboplatin, 6a, 6b, 7, and 8 against cervical carcinoma HeLa cells expressed as IC50 values (|^M) obtained after 72 h incubation.
Compound
Cisplatin Carboplatin
IC50 (|M)
16.3 ± 3.6 441.1 ± 91.8
22.33 ± 3.19
162.40 ± 21.98
>1000
19.05
zole functionalized diazenecarboxamides 6a and 6b (IC50 = 22.33 pM and 162.40 pM, respectively) as well as model [Pt(picolyl-triazole)Cl2] complex 8 (IC50 = 19.05 pM) were all more active than the corresponding hybrids 2a and 2b (IC50 = 78 pM and 277 pM). The activity of 8 was in the range of cisplatin (IC50 = 16.3 pM).
An easy explanation for severely reduced activity of 1-2 when coming from their "parent" compounds 6 and 8 is impossible but it could be a result of change in lipophi-licity, molecular size, shape, etc.24 Before these factors are taken into the consideration, however, the issue of solubility has to be addressed first as compounds 1-4 are completely insoluble in water.
3. Conclusions
We have demonstrated a design of a library of diaze-necarboxamide-extended cisplatin and carboplatin analogues. The advantage of this strategy is modular, combinatorial approach, taking advantages of "Click chemistry" and offering timely preparation of organic-inorganic conjugates in general. The library was designed and evaluated for improved efficiency in antitumor therapy. Although several members were quite cytotoxic, in contrast to expectations, none performed better than carboplatin or cisplatin. This can be due to their limited solubility. Work is in progress to diazenecarboxamide-extended cisplatin and carboplatin analogues with improved water solubility.
4. Experimental
4. 1. General
Cisplatin, carboplatin and NMR solvents were used as obtained from commercial sources (Sigma-Aldrich,
USA). Compounds 1,13 2,13 3,14 4,14 5,15 6,22 and 714 and 822 were prepared as described in the literature. NMR spectra were measured on a Bruker Avance III 500 spectrometer, using Si(CH3)4 as internal standard. ESI+ HRMS spectra were recorded with Agilent 6224 Accurate Mass TOF LC/MS system.
4. 2. NMR and HRMS Experiments
a)	Monitoring the course of the reaction between 3b and GSH in DMF-^7/D2O:
A mixture of 3b (4.0 mg, 0.0014 mmol) and GSH (5.7 mg, 0.0185 mM) was dissolved in DMF-d7 (0.50 mL) and D2O (0.25 mL). The reaction mixture was kept at 23 °C and 1H NMR spectra were recorded at the same temperature at the times indicated in Figure 4d. Resonances of 3b,14 3bRED, GSH and GSSG were assigned on the basis of literature data and 2D 1H-1H COSY spectra of the reaction mixture. After 44 h a drop of the reaction mixture was dissolved in 0.1% formic acid in Milli-Q water (1 mL) and left for 24 h to allow complete deuterium exchange. HRMS spectrum was recorded, which revealed unreacted GSH, GSSG and 3bRED (m/z found: 802.1886. Calcd. for C26H30F3N8O6Pt+ [M + H]+ = 802.1890).
b)	Monitoring the course of the reaction between 3b and GSH in a mixture of DMSO and growth medium with fetal serum:
A mixture of 3b (1.1 mg, 0.0014 mmol) and GSH (1.9 mg, 0.00618 mmol) was dissolved in DMSO (0.50 mL), diluted with growth medium supplemented with 10% fetal serum (0.20 mL) and kept at 23 °C. The reaction was followed by dissolving an aliquot (one drop) in acetonitrile (1 mL) and measuring HRMS spectra.
4. 3. Cells
Human cervical carcinoma HeLa cells were obtained from cell culture bank (GIBCO BRL, Invitrogen, Grand Island, USA). They were cultured as a monolayer culture according in Dulbecco's medium supplemented with 10% fetal serum (Gibco BRL, Invitrogen, USA) and cultured in a humidified atmosphere of 5% CO2 at 37 °C.
4. 4. Compounds
Carboplatin and cisplatin were dissolved in water whereas compounds 1-8 were dissolved in DMSO. They were and stored at -20 °C, and diluted to the appropriate concentrations with growth medium just before use.
4. 5. Cytotoxicity Assay
Cytotoxic effect of compounds under investigation was determined by 3-(4,5-dimethylthiazol-2-yl)-2,5-dip-henyltetrazolium bromide (MTT) (Sigma Aldrich, USA)
assay, modified as described. Cells were seeded into 96-well tissue culture plates (3000 cells/0.18 mL medium/well). Different concentrations of new compounds were added (0.02 mL) to each well on the following day. Each concentration was tested in quadruplicate. Following 72 h incubation at 37 °C, the medium was aspirated, and 20 mg of MTT dye/0.04 mL medium/well was added. Four hours later, formazan crystals were dissolved in DMSO (0.17 mL/well), the plates were mechanically agitated for 5 min and the optical density at 545 nm was determined on a microtiter plate reader (Awareness Technology Inc, Palm City, FL). Each experiment was repeated three times.
5. Acknowledgments
This study was supported by the Ministry of Science, Education and Sport of the Republic of Croatia (Projects 098-0982913-2748; 098-0982913-2850) and the Slovenian Research Agency (Project P1-0230-0103 and Joint Project BI-HR/12-13-028).
6. References
1.	T. Boulikas, M. Vougiouka, Oncol. Rep. 2003, 10, 16631682.
2.	L. Kelland, Nature Rev. Cancer 2007, 7, 573-584.
3.	(a) D. J. Stewart, Crit. Rev. Oncol./Hematol. 2007, 63, 12-31. (b) B. Köberle, M. T. Tomicic, S. Usanova, B. Kaina, Biochim. Biophys. Acta 2010, 1806, 172-182. (c) A. Brozo-vic, A. Ambriovic-Ristov, M. Osmak, Crit. Rev. Toxicol. 2010, 40, 347-359. (d) H. H. W. Chen, M. T. Kuo, Metal Based Drugs 2010, 1-6.
4.	L. K. Gediya, V. C. O. Njar, Expert Opin. Drug Discov. 2009, 4, 1099-1111.
5.	(a) E. Monti, M. Gariboldi, A. Maiocchi, E. Marengo, C. Cassino, E. Gabano, D. Osella, J. Med. Chem. 2005, 48, 857-866. (b) B. Meunier, Acc. Chem. Res. 2008, 41, 69-77. (c) N. Das, M. Dhanawat, B. Dash, R. C. Nagarwal, S. K. Shrivastava, Eur. J. Pharm. Sci. 2010, 41, 571-588.
6.	A. Müller-Schiffmann, J. März-Berberich, A. Andreyeva, R. Rönicke, D. Bartnik, O. Brener, J. Kutzsche, A. H. C. Horn, M. Hellmert, J. Polkowska, K. Gottmann, K. G. Reymann, S. A. Funke, L. Nagel-Steger, C. Moriscot, G. Schoehn, H. Sticht, D. Willbold, T. Schrader, C. Korth, Angew. Chem. Int. Ed. 2010, 49, 8743-8746.
7.	W. H. Ang, I. Khalaila, C. S. Allardyce, L. Juillerat-Jeanne-ret, P. J. Dyson, J. Am. Chem. Soc. 2005, 127, 1382-1383.
8.	R. Millet, S. Urig, J. Jacob, E. Amtmann, J. -P. Moulinoux, S. Gromer, K. Becker, E. Davioud-Charvet, J. Med. Chem. 2005, 48, 7024-7039.
9.	J. Kosmrlj, M. Kocevar, S. Polanc, Synlett 2009, 2217-2235. 10. (a) M. Osmak, T. Bordukalo, J. Kosmrlj, M. Kvajo, Z. Mari-
janovic, D. Eljuga, S. Polanc, Neoplasma 1999, 46,
201-206. (b) M. Osmak, T. Bordukalo, B. Jernej, J. Košmrlj, S. Polanc, Anti-Cancer Drugs 1999, 10, 853-859. (c) M. Osmak, T. Bordukalo, A. Ambriovic Ristov, B. Jernej, J. Košmrlj, S. Polanc, Neoplasma 2000, 47, 390-395. (d) D. Moskatelo, A. Benjak, V. Laketa, S. Polanc, J. Košmrlj, M. Osmak, Chemother. 2002, 48, 36-41. (e) D. Moskatelo, S. Polanc, J. Košmrlj, L. Vukovic, M. Osmak, Pharmacol. Toxicol. 2002, 91, 258-263. (f) T. Čimbora, S. Bombek, S. Polanc, M. Osmak, Toxicol. in Vitro 2003, 17, 159-164. (g) T. Čimbora-Zovko, S. Bombek, J. Košmrlj, L. Kovacic, S. Polanc, A. Katalinic, M. Osmak, Drug Devel. Res. 2004, 61, 95-100. (h) S. Jakopec, K. Dubravcic, S. Polanc, J. Košmrlj, M. Osmak, Toxicol. in Vitro 2006, 20, 217-226. (i) S. Jakopec, K. Dubravcic, A. Brozovic, S. Polanc, M. Osmak, Cell Biol. Toxicol. 2006, 22, 61-71. (i) I. Martin-Kleiner, S. Bombek, J. Košmrlj; B. Čupic, T. Čimbora-Zovko, S. Jakopec, S. Polanc, M. Osmak, J. Gabrilovac, Toxicol. in Vitro 2007, 21, 1453-1459.
11.	S. Grabner, J. Košmrlj, N. Bukovec, M. Čemažar, J. Inorg. Biochem. 2003, 95, 105-112.
12.	D. Urankar, J. Košmrlj, J. Comb. Chem. 2008, 10, 981-985.
13.	D. Urankar, A. Pevec, J. Košmrlj, Eur. J. Inorg. Chem. 2011, 1921-1929.
14.	D. Urankar, J. Košmrlj, Inorg. Chim. Acta 2010, 363, 3817-3822.
15.	D. Urankar, M. Steinbücher, J. Kosjek, J. Košmrlj, Tetrahedron 2010, 66, 2602-2613.
16.	For examples of "click" post-postfunction of platinum com-
plexes, see: (a) A. R. McDonald, H. P. Dijkstra, B. M. J. M. Suijkerbuijk, G. P. M. van Klink, G. van Koten, Organome-tallics 2009, 28, 4689-4699. (b) M. C. Clough, P. D. Zeits, N. Bhuvanesh, J. A. Gladysz, Organometallics 2012, 31, 5231-5234. (c) C. Gibard, D. Avignant, F. Cisnetti, A. Gautier, Organometallics 2012, 31, 7902-7908. (d) E. Chardon, G. L. Puleo, G. Dahm, S. Fournel, G. Guichard, S. Bellemin-Laponnaz, ChemPlusChem 2012, 77, 1028-1038.
17.	For selected comprehensive review, see: M. Meldal, C. W. Torn0e, Chem. Rev. 2008, 108, 2952-3015.
18.	For recent book, see: J. Košmrlj (Ed.), Click Triazoles (Topics in heterocyclic chemistry, vol. 28), Springer-Verlag Berlin and Heidelberg Gmbh & Co. Kg (Germany), 2012.
19.	H. C. Kolb, M. G. Finn, K. B. Sharpless, Angew. Chem. Int. Ed. 2001, 40, 2004-2021.
20.	J. Košmrlj, M. Kocevar, S. Polanc, J. Chem. Soc., Perkin Trans. 1 1998, 3917-3919.
21.	(a) S. Y. Loh, P. Mistry, L. R. Kelland, G. Abel, K. R. Harrap, Br. J. Cancer 1992, 66, 1109-1115. (b) J. A. Platts, D. E. Hibbs, T. W. Hambley, M. D. Hall, J. Med. Chem. 2001, 44, 472-474. (c) X. Liu, B. Testa, A. Fahr, Pharm. Res. 2011, 28, 962-977.
22.	D. Urankar, B. Pinter, A. Pevec, F. De Proft, I. Turel, J. Košmrlj, Inorg. Chem. 2010, 49, 4820-4829.
23.	J. D. Crowley, D. A. McMorran, Topics in Heterocyclic Chemistry 2012, 28, 31-83.
24.	A. Nadin, C. Hattotuwagama, I. Churcher, Angew. Chem., Int. Ed. 2012, 51, 1114-1122.
Povzetek
Opisali smo knjižnico strukturno različnih c/s-[Pt(2-pikolil-1,2,3-triazol)Cl2] in c/s-[Pt(propan-1,3-diamin)CBDCA] (CBDCA = 1,1-ciklobutandikarboksilat) kompleksov 1—4, funkcionaliziranih z diazenkarboksamidno skupino. Te spojine ohranjajo oksidativne sposobnosti diazenkarboksamidov do glutationa, kar smo dokazali na osnovi NMR in HRMS eksperimentov. Citotoksičnost spojin 1—4 smo testirali na humanih tumorskih celicah materničnega vratu HeLa. Štiri spojine so pokazale zmerno aktivnost. Prav tako smo testirali nekatere modelne spojine, od katerih je bil [PtCl2L2] (L = 4-fenil-1-(2-pikolil)-1_ff-1,2,3-triazol) najbolj aktiven z IC50 = 19.05 |M, kar je primerljivo z aktivnostjo cisplatina (IC50 = 16.3 |M).