Synthesis, Cytotoxicity and Antiplasmodial Activity of Neocryptolepine ...

Synthesis, Cytotoxicity and Antiplasmodial Activity of Neocryptolepine Derivatives S. Van Miert, T. Jonckers, L. Maes, A. Vlietinck, R. Dommisse, G. Lemière and L. Pieters Departments of Pharmaceutical Sciences and Chemistry University of Antwerp, Antwerp Belgium Keywords: Cryptolepis sanguinolenta, Periplocaceae, cryptolepine, neocryptolepine, Plasmodium, malaria Abstract Based on the original lead neocryptolepine, an alkaloid from Cryptolepis sanguinolenta, a series of substituted derivatives was prepared. All compounds were evaluated for their activity against chloroquine-resistant Plasmodium falciparum strains, and for their cytotoxicity on human MRC-5 cells. Mechanisms of action were investigated by testing inhibition of β-haematin formation and DNA interactions (DNA-methylgreen assay). Some neocryptolepine derivatives with a higher antiplasmodial activity and a lower cytotoxicity than the original lead have been obtained. This selective antiplasmodial activity was associated with inhibition of β-haematin formation. 2-Bromoneocryptolepine was the most selective compound with an IC50 value against chloroquine-resistant P. falciparum of 4.0 µM in the absence of cytotoxicity (LC50 > 32 µM). This compound was considered as the most promising lead from the present work for new antimalarial agents. INTRODUCTION Malaria is a parasitic disease caused by Plasmodium falciparum (the most severe form), P. vivax, P. ovale and P. malariae, confronting both developing and industrialised nations. Over 300 million new infections occur each year, many being the maligant form of malaria caused by Plasmodium falciparum. This disease causes more than 2.5 million deaths annually. The life cycle of the Plasmodium parasite has a sexual stage in the female anopheline mosquito, and an asexual stage in man. The problem with malaria is the increasing resistance of the mosquito to insecticides, and of the parasite, especially P. falciparum, to currently available drugs such as chloroquine. Therefore, there is a high need for new antimalarial drugs (O’Neill et al., 1998; Rang et al., 1999). The first antimalarial compound to be discovered, which also served as the lead compound for synthetic antimalarials of the chloroquine/mefloquine type, was the alkaloid quinine. A lot of other leads for potential new antimalarial agents have been characterized since then, and many of them have been isolated from medicinal plants (Newman et al., 2000). Aqueous decoctions or aqueous macerates of the root or the root bark of Cryptolepis sanguinolenta (Lindl.) Schlechter (Periplocaceae) are used in traditional medicine in Central and West Africa for the treatment of infectious diseases, amoebiasis and fevers including fever caused by malaria. Until 1990, only two alkaloids had been characterised from this plant: cryptolepine (Fig. 1a), the major alkaloid, and quindoline, which lacks the N-methyl group. Both alkaloids are indoloquinolines. Antiplasmodial activity in vitro (against chloroquine-resistant Plasmodium falciparum) as well as in vivo (in mice infected with Plasmodium berghei) has been reported for cryptolepine, but not for quindoline. This already demonstrated the importance of the N-methyl group for this biological activity (Kirby et al., 1995; Cimanga et al., 1997). In addition cryptolepine also showed antibacterial properties, and it was highly cytotoxic, due to DNA interactions and inhibition of topoisomerase II (Cimanga et al., 1996a; Bonjean et al., 1998). During the past decade, a whole series of minor alkaloids has been isolated from Cryptolepis sanguinolenta, including neocryptolepine (Fig. 1b) (Cimanga et al., 1996b). Simultaneously the same compound was reported by Sharaf et al. (1996) as

Proc. WOCMAP III, Vol. 3: Perspectives in Natural Product Chemistry Eds. K.H.C. Başer, G. Franz, S. Cañigueral, F. Demirci, L.E. Craker and Z.E. Gardner Acta Hort. 677, ISHS 2005

91

cryptotackieine. Also for this compound antiplasmodial activity has been reported against chloroquine-resistant strains of P. falciparum (Cimanga et al., 1997). However, a direct comparison of the cytotoxicity of cryptolepine and neocryptolepine demonstrated that the latter was much less cytotoxic. DNA intercalating as well as topoisomerase II inhibiting properties were found to be more pronounced for cryptolepine than for neocryptolepine (Bailly et al., 2000). Therefore, neocryptolepine was selected in the present work as a lead for the development of new antimalarial agents. The antiplasmodial properties of some synthetic neocryptolepine derivatives are discussed, as well as their activity in functional assays related to possible mechanisms of action, in order to establish structure-activity relationships. MATERIALS AND METHODS Synthesis of Neocryptolepine Derivatives The organic synthesis of the neocryptolepine derivatives studied in this work has been reported before (Jonckers et al., 2002). The crucial step was a biradical cyclisation reaction of a series of carbodiimides. In this way a series of 2-substituted and 2,9disubstituted neocryptolepine derivatives could be obtained. If an appropriate carbodiimide precursor was used, a mixture of 1- and 3-substituted derivatives was obtained, which had to be separated by chromatographic means. Neocryptolepine derivatives substituted in position 4 could not be obtained, since N-methylation appeared to be impossible in the presence of a substituent in position 4. All compounds were biologically evaluated as their hydrochloride salts. In Vitro Activity against P. falciparum All compunds were evaluated for their antiplasmodial activity against a chloroquine-resistant Plasmodium falciparum strain. For the determination of the antiplasmodial activity the parasite lactate dehydrogenase assay was used, with slight modifications, as previously described by Makler and Hinrichs (1993). Briefly, the assay is based on the observation that the lactate dehydrogenase (LDH) enzyme of Plasmodium falciparum has the ability to rapidly use 3-acetyl pyridine NAD (APAD) as a coenzyme in the reaction leading to the formation of pyruvate from lactate. Test compounds were added in 2-fold serial dilutions in 96-well multiwell plates to Plasmodium falciparum (chloroquine-resistant W2 strain) cultures (1% parasitaemia, 2% HCT). After 48 h at 37°C and a gas mixture of 93% N2, 4% CO2, 3% O2, the parasite cultures were frozen at 20°C to lyse the erythrocytes and to store the plates until further processing. After thawing, 20 µl of the lysed culture was added to 100 µl Malstat reagent (Flow Inc., USA) and the formation of APADH was determined. Adding 40 µg nitroblue tetrazolium (NBT) and 2 µg phenazine ethosulfate (PES) to the Malstat reagent promoted the spectrophotometric assessment of LDH activity. As APADH is formed, the NBT is reduced and forms a blue formazan product that can be detected visually and measured spectrophotometrically at 650 nm. Determination of the IC50 values was performed in triplicate (mean ± SD). Cryptolepine (Arzel et al., 2001; Wright et al., 2001) and chloroquine were used as positive controls. The IC50 value of chloroquine was 0.09 µM (for cryptolepine, see Table 1). Cytotoxicty on MRC-5 Cells A human diploid embryonic lung cell line (MRC-5) was used to assess the cytotoxicity of the test compounds as described before (Girault et al., 2000). Briefly, MRC-5 cells were seeded at 5000 cells/well in 96-well microtiter plates. After 24 h, the cells were washed and 2-fold dilutions of the drug were added in 200 µl of standard culture medium (RPMI + 5% FCS). The final DMSO concentration in the culture remained below 0.5%. The cultures were incubated with different concentrations of test compounds at 37°C in 5% CO2 – 95% air for 7 days. Untreated cultures were included as controls. Cytotoxicity was determined using the colorimetric MTT assay and scored as a

92

percent (%) reduction of absorption at 540 nm of treated cultures versus untreated control cultures. Determination of the LC50 values was performed out in triplicate (mean ± SD). Vinblastine was used as positive control (LC50 < 10 nM). Inhibition of β-Haematin Formation An in vitro method was used to measure the inhibition of β-haematin formation, based on the original method described by Egan et al. (1994) and modified by Wright et al. (2001). Haem refers to Fe(II)protoporphyrin IX, haemin is Fe(II)protoporphyrin IX chloride, and haematin is Fe(II)protoporphyrin IX hydroxide. Briefly, 0.1 M NaOH was added to Fe(II)haemin (Fluka; HPLC purity >98%) (typically 7.5 mg) which forms haematin. Polymerisation to β-haematin, which has been characterized as [Fe(III)protoporphyrin IX]2 (Pagola et al., 2000), proceeds after addition of a 9.9 M acetate buffer, pH 5. Test compounds (3 equivalents with respect to haemin) were added to the haematin solutions prior to acidification. After incubation for 40 min at 60°C, cooling and filtration, the precipitate was washed with water, dried, and FT-IR (PerkinElmer FT-IR 1760) was used to check the presence of β-haematin, which shows sharp bands at 1660 and 1207 cm-1. The spectra were examined on the presence or absence of the two typical peaks. When they were absent the compound had inhibited the formation of β-haematin. A sample containing haemin but without test compound, which was not incubated, was used as a negative control (no formation of β-haematin). A sample containing haemin but without test compound, which was incubated during 40 min at 60°C, was used as a positive control (formation of β-haematin). DNA-Methylgreen Assay The DNA-methylgreen assay is a simple microtiter assay for the detection of compounds that bind DNA. Agents that displace methylgreen from a DNA-methylgreen complex (Deoxyribonucleic acid Methylgreen; Sigma) are detected spectrophotometrically (Labsystems Multiscan MCC/340) by a decrease in absorbance at 620 nm (Burres et al., 1992). DNA-methylgreen was suspended in 100 ml of 0.05 M Tris-HCl buffer, pH 7.5, containing 7.5 mM MgSO4 and stirred at 37°C for 24 h. The dissolved samples were dispensed into wells of a 96-well microtiter plate. Solvent was removed under vacuum, and 200 µl of the DNA-methylgreen solution was added. The initial absorbance was compared with the final absorbance (after 24 h) in order to calculate the IC50 value (50% displacement of methylgreen from DNA). Determination of the IC50 values was performed in triplicate (mean ± SD). The decrease in absorbance observed represents the initial rapid displacement of methylgreen from DNA by the drug, followed by the slower reaction with water that yields the colorless carbinol. RESULTS AND DISCUSSION All compounds were evaluated for their antiplasmodial activity against a chloroquine-resistant Plasmodium falciparum strain, for cytotoxicity on a human cell line, and in two functional assays related to possible mechanisms of action or cytotoxicity, i.e. DNA interactions and inhibition of β-haematin formation. The discussion of the biological activity of these neocryptolepine derivatives will be limited here to some typical representatives, i.e. the neocryptolepine lead itself, 2-methyl-, 2-bromo- and 2methoxy-neocryptolepine, as well as 1-bromo- and 3-bromo-neocryptolepine. Cryptolepine was also included. Results are summarised in Table 1. With regard to the antiplasmodial activity, cryptolepine showed an IC50 of 2.0 µM, and neocryptolepine was less active (IC50 14.0 µM). The neocryptolepine derivatives substituted in position 2 with a methyl-, bromo- or methoxyl functionality, as well as 3bromo-neocryptolepine, were more active than neocryptolepine itself, but not as active as cryptolepine. The 1-bromo derivative was not active in the concentration range tested. The antiplasmodial activity of cryptolepine turned out not te be very selective, since the LC50 on the human cell line was comparable to the antiplasmodial IC50. Neocryptolepine showed a reduced cytotoxicity, which was in agreement with our earlier findings for these

93

compounds. Also the 2-methyl- and 2-methoxyl derivatives were not selective: their LC50 was lower than their antiplasmodial IC50. However, 3-bromo- but especially 2-bromoneocryptolepine showed some selectivity towards the Plasmodium parasite: The 2-bromo derivative had an IC50 against Plasmodium falciparum of 4.0 µM, whereas its LC50 was not in the concentration range tested. Because of the reported DNA intercalating properties of cryptolepine and to a lesser extent neocryptolepine, all compounds were also evaluated in the DNAmethylgreen assay. Cryptolepine exhibited the most potent activity as a DNA interacting agent, with an IC50 value of 65.7 µM. Also 2-methoxy-neocryptolepine showed DNAinteracting properties. This was in agreement with their pronounced cytotoxicity. When the DNA-interacting properties decreased, as for neocryptolepine, also the cytotoxicity decreased. Compounds showing no DNA interactions in the concentration range tested had a reduced cytotoxicity; the most notable compound in this regard was 2-bromoneocryptolepine. For cryptolepine, it has been shown that the antiplasmodial activity is not only due to its DNA-interacting properties, but also, at least in part, to inhibition of the haem detoxification process (Wright et al., 2001). It shares this mechanism of action with known antimalarials such as quinine, chloroquine or mefloquine (O’Neill et al., 1998). During the erythrocytic stage of the Plasmodium lifecycle in the human host, the parasites (merozoites) live in the red blood cells, and use haemoglobin for food. Haemoglobin is taken up in the acid food vacuole of the parasite. The globin part is used as a source of amino acids, but the haem part is useless; in fact it is even toxic for the parasite. Therefore toxic, free haem is converted by the parasite into a non-soluble, non-toxic haem polymer, which is called malaria pigment or haemozoin. This detoxification or polymerisation process is inhibited by the quinoline antimalarials. Haem is not detoxified, and will kill the parasite. The in vivo process of conversion of haem into haemozoin corresponds to an in vitro process where haemin [Fe(II)protoporphyrin IX Cl-] is polymerised to β-haematin (Egan et al., 1994). Quindoline antimalarials, as well as cryptolepine, are able to inhibit the formation of β-haematin in cell-free systems, indicating that inhibition of the haem detoxification process may be responsible, at least in part, for the antiplasmodial activity of these compounds (Wright et al., 2001). Results of the β-haematin assay are also included in Table 1. In addition to cryptolepine, also neocryptolepine, 2-bromo- and 3bromo-neocryptolepine were able to inhibit β-haematin formation. For cryptolepine and neocryptolepine the antiplasmodial activity may also be due in part to their DNAinteracting properties. Therefore it is non-selective, and associated with cytotoxicity to a human cell line. The 3-bromo- and especially the 2-bromo-derivative of neocryptolepine, however, exhibited a more selective antiplasmodial activity, since for the latter compound no DNA interactions were observed, and the cytotoxicity LC50 was not in the concentration range tested. On the other hand, 2-methoxy-neocryptolepine was not able to inhibit the formation of β-haematin in the in vitro assay, but nevertheless it showed a pronounced antiplasmodial activity. In this case it appeared that this activity was related to a non-specific mechanism of action, i.e. DNA interacting properties, and that therefore it was associated with cytotoxicity. In the neocryptolepine series, it appeared as if introduction of halo- and especially bromo-substituents on the neocryptolepine nucleus in position 2 improves the antiplasmodial activity, while reducing the cytotoxicity. Also for cryptolepine derivatives, it has been reported that introduction of bromo-substituents had a beneficial effect on the antiplasmodial activity. In a limited series of substituted cryptolepine derivatives, 2,7dibromo-cryptolepine was the most promising compound. It was 10 times more antiplasmodially active than cryptolepine, but still showed cytotoxic properties in vitro. However, it was also active in vivo in mice infected with Plasmodium berghei: 89% suppression of parasitaemia was observed without apparent systemic cytotoxicity (Wright et al., 2001).

94

CONCLUSIONS The neocryptolepine derivatives studied in this work may be classified in different groups (Fig. 2). A first group consists of compounds with a selective antiplasmodial activity, which are able to inhibit the formation of β-haematin; therefore their mechanism of action is most probably related to inhibition of the haem detoxification process. The most prominent example of this group is 2-bromo-neocryptolepine. On the other hand, a second group consists of compounds showing a non-selective antiplasmodial activity, which is related to DNA-interactions, but which are not capable of inhibiting the formation of β-haematin. A typical compound in this group is 2-methoxyneocryptolepine. Finally, a third group of compouds contains antiplasmodial agents with a mixed mechanism of action, such as neocryptolepine and cryptolepine. Also these compounds lack selectivity towards the Plasmodium parasite. Although 2-bromoneocryptolepine can be considered as the most promising lead at the moment in the series of the neocryptolepine derivatives, a further optimalisation of both the antiplasmodial activity and the selectivity is needed before a potentially clinically useful therapeutic agent may be obtained. ACKNOWLEDGEMENTS Financial support from the Fund for Scientific Research (FWO-Vlaanderen, Belgium) (project numbers G.0119.96, G.0082.98, G.0334.00 and 1.5.139.00), and from the Special Research Fund of the University of Antwerp (Concerted Research Project No. 99/3/34, fellowship for S. van Miert) is gratefully acknowledged. Dr. Colin Wright (University of Bradford, UK) is kindly acknowledged for assistance with the assay on inhibition of β-haematin formation. Literature Cited Arzel, E., Rocca, P., Grellier, P., Labaeïd, M., Frappier, F., Guéritte, F., Gaspard, C., Marsais, F., Godard, A. and Quéguiner, G. 2001. New Synthesis of Benzo-δ-carbolines, Cryptolepines, and Their Salts: In Vitro Cytotoxic, Antiplasmodial, and Antitrypanosomal Activities of δ-Carbolines, Benzo-δ-carbolines, and Cryptolepines. J. Med. Chem. 44:949-960. Bailly, C., Laine, W., Baldeyrou, B., De Pauw-Gillet, M.-C., Colson, P., Houssier, C., Cimanga, K., Van Miert, S., Vlietinck, A.J. and Pieters, L. 2000. DNA Intercalation, Topoisomerase II Inhibition and Cytotoxic Activity of the Plant Alkaloid Neocryptolepine. Anti-Cancer Drug Design 15:191-201. Bonjean, K., De Pauw-Gillet, M.-C., Defresne, M.-P., Colson, P., Houssier, C., Dassonneville, L., Bailly, C., Greimers, R., Wright, C., Quetin-Leclercq, J., Tits, M. and Angenot, L. 1998. The DNA Intercalating Alkaloid Cryptolepine Interferes With Topoisomerase II and Inhibits Primarily DNA Synthesis in B16 Melanoma Cells. Biochemistry 37:5136-5146. Burres, N.L., Frigo, A., Rasmussen, R.R. and McAlpine, J.B. 1992. A Colorimetric Microassay for the Detection of Agents That Interact with DNA. J. Nat. Prod. 55:1582-1587. Cimanga, K., De Bruyne, T., Lasure, A., Van Poel, B., Pieters, L., Claeys, M., Vanden Berghe, D., Kambu, K., Tona, L. and Vlietinck, A. 1996a. In Vitro Biological Activities of Alkaloids from Cryptolepis sanguinolenta. Planta Med. 62:22-27. Cimanga, K., De Bruyne, T., Pieters, L., Claeys, M. and Vlietinck, A. 1996b. New Alkaloids from Cryptolepis sanguinolenta. Tetrahedron Lett. 37:1703-1706. Cimanga, K., De Bruyne, T., Pieters, L., Vlietinck, A. and Turger, C.A. 1997. In Vitro and in Vivo Antiplasmodial Activity of Cryptolepine and Related Alkaloids from Cryptolepis sanguinolenta. J. Nat. Prod. 60:688-691. Egan, T.J., Ross, D.C. and Adams, P.A. 1994. Quinoline Anti-Malarial Drugs Inhibit Spontaneous Formation of β-Haematin (Malaria Pigment). FEBS Lett. 352:54-57. Girault, S., Grellier, P., Berecibar, A., Maes, L., Mouray, E., Lemière, P., Debreu, M.-A., Davioud-Charvet, E. and Sergheraert, C. 2000. Antimalarial, Antitrypanosomal, and

95

Antileishmanial Activities and Cytotoxicity of Bis(9-amino-6-chloro-2-methoxyacridines): Influence of the Linker. J. Med. Chem. 43:2646-2654. Kirby, G.C., Paine, A., Warhurst, D.C., Noamese, B.K. and Phillipson, J.D. 1995. In Vitro and in Vivo Antimalarial Activity of Cryptolepine, a Plant-derived Indoloquinoline. Phytother. Res. 9:359-363. Makler, M.T. and Hinrichs, D.J. 1993. Measurement of the Lactate Dehydrogenase Activity of Plasmodium falciparum as an Assessment of Parasitemia. Am. J. Trop. Med. Hyg. 48:205-210. Newman, D.J., Cragg, G.M. and Snader, K.M. 2000. The Influence of Natural Products upon Drug Discovery. Nat. Prod. Rep. 17:215-234. O’Neill, P.M., Bray, P.G., Hawley, S.R., Ward, S.A. and Park, B.K. 1998. 4Aminoquinolines – Past, Present, and Future: A Chemical Perspective. Pharmacol. Ther. 77:29-58. Pagola, S., Stephens, P.W., Bohle, D.S., Kosar, A.D. and Madsen, S.K. 2000. The Structure of Malaria Pigment β-Haematin. Nature 404:307-310. Rang, H.P., Dale, M.M. and Ritter, J.M. 1999. Chapt. 46: Antiprotozoal Drugs. In: Pharmacology, 4th ed., Churchill Livingstone, Edinburgh. Sharaf, H.M., Schiff, P.L., Tackie, A.N., Phoebe, C.H. and Martin, G.E. 1996. Two new indoloquinoline alkaloids from Cryptolepis sanguinolenta: Cryptosanguinolentine and cryptotackieine. J. Heterocyclic Chem. 33:239-243. Wright, C.W., Addae-Kyereme, J., Breen, A.G., Brown, J.E., Cox, M.F., Croft, S.L., Gökçek, Y., Kendrick, H., Phillips, R.M. and Pollet, P.L. 2001. Synthesis and Evaluation of Cryptolepine Analogues for Their Potential as New Antimalarial Agents. J. Med. Chem. 44:3187-3194.

Tables

Table 1. Antiplasmodial activity, cytotoxicity, DNA interactions (DNA methylgreen assay) and inhibition of β-haematin formation of neocryptolepine derivatives.

cryptolepine neocryptolepine 2-Me-neocryptolepine 2-Br-neocryptolepine 2-OMe-neocryptolepine 1-Br-neocryptolepine 3-Br-neocryptolepine

96

P. falciparum (W2) (IC50, µM) 2.0 ± 0.1 14.0 ± 1.7 2.3 ± 0.6 4.0 ± 0.1 4.7 ± 0.6 > 32 4.7 ± 0.6

Cytotoxicity (MRC-5 cells) (LC50, µM) 1.5 ± 0.7 11.0 ± 1.4 0.95 ± 0.07 > 32 4.0 ± 0.1 16.0 ± 0.1 18.5 ± 0.7

DNA interactions (IC50, µm) 65.7 ± 3.0 92.8 ± 9.7 (not tested) > 400 77.9 ± 4.4 > 400 > 400

Inhibition of β-haematin formation yes yes (not tested) yes no no yes

Figures

CH3

6

N

7 8 9

N 10

5

11

10

4 3

9

2

8

1

11

1 2

7

N

6

N5

3 4

CH3

(a)

(b)

Fig. 1. Structure of cryptolepine (a) and neocryptolepine (b).

DNA interactions (non-selective)

2-MeO-neocryptolepine

Inhibition of the haem detoxification process (selective)

cryptolepine neocryptolepine

2-Br-neocryptolepine

Fig. 2. Antiplasmodial activity: Mechanism of action of neocryptolepine derivatives.

97