Synthesis and Evaluation of Febrifugine Analogues as Potential Antimalarial Agents

Abstract

Febrifugine is an alkaloid isolated from Dichroa febrifuga Lour as the active component against Plasmodium falciparum. Strong liver toxicity has precluded febrifugine as a potential clinical drug. In this study novel febrifugine analogues were designed and synthesized. Lower toxicity was achieved by reducing or eliminating the tendency of forming chemically reactive and toxic intermediates and metabolites. Synthesized compounds were evaluated in vitro against chloroquine sensitive (D6) and chloroquine resistant (W2) P. falciparum strains for efficacy and in freshly isolated rat hepatocytes for potential cytotoxicity. The IC₅₀’s of the best compounds were superior to their parent compound febrifugine. Noticeably, these compounds were also over 100 times less toxic than febrifugine. These compounds, as well as the underlying design rationale, may find usefulness in the discovery and development of new antimalarial drugs.

Malaria is one of the major parasitic infections in many tropical and subtropical regions. Worldwide, malaria infects 300–600 million people and kills about three million in a year.¹ The malaria parasite has now become resistant to the best anti-malarial drugs.^{2, 3} Therefore, medicinal agents based on novel mode of action are required to overcome the emergence of resistance and to control an ever-increasing number of epidemics caused by the malaria parasite.

Febrifugine (1) and isofebrifugine (2), as shown in Figure 1, were isolated as the active components against malaria in the Chinese herb Chang Shan (Dichroa febrifuga Lour),^{4, 5} which has been employed by the local people as medicine against fevers caused by malaria parasites for a long time. Febrifugine acts by impairing haemazoin formation required for maturation of the parasite at the trophozoite stage.⁶ Subsequent pre-clinical researches have found that febrifugine possesses adverse side effects. Strong liver toxicity has precluded febrifugine as a potential clinical drug.^{7, 8}

Figure 1.

Some febrifugine analogues have been synthesized; most of the structural modifications were focused on side chain alterations. It is known that the 4-quinazolinone moiety plays an essential role in the appearance of activity; and that the presence of a 1”-amino group and C-2’, C-3” O-functionalities is crucial in the anti-malarial activity of febrifugine.^{9, 10, 11} Structure-activity relationship studies have also demonstrated that attaching any electron-withdrawing group to the 1”-nitrogen decreases activity.¹⁰

Febrifugine could be metabolized to the corresponding arene oxide 3 by cytochrome P-450 enzymes (Scheme 1). When arene oxide 3 escapes deactivation process by certain enzymes such as epoxide hydrase or glutathione S-transferase, toxicity can result because this reactive electrophile will form covalent adducts with DNA, RNA, proteins, or other biomolecules of the host. Such binding can cause mutations and result in cell damage. In a recent metabolic study of febrifugine, Oshima and coworkers isolated metabolite 4.¹² This study indicates that arene oxide 3 is most probably a short-lived reactive metabolic intermediate because metabolite 4 would have been derived from intermediate 3 through the rearrangement known as NIH shift.¹³

Scheme 1.

There are two possible ways to make this process unfavorable: (I) Blocking the C-5 or C-6 position of the quinoline ring; and (II) Increasing the oxidation potential of the molecule. As part of an ongoing malaria chemotherapy project in our laboratory, we undertook the synthesis and antimalarial activity evaluation of the following febrifugine analogues (compounds 5–20, shown in Figure 2). Compounds 5–16 are designed to block the aforementioned unwarranted metabolic pathway. Introduction of an extra nitrogen atom on position 5 or 6 of the quinazolinone ring in febrifugine furnishes compounds 5 and 6. Compounds 7–11 each bear one or two electron-withdrawing group(s), while compounds 12 and 13 bear a bulky group, on C-5 and/or C-6 position of the quinazolinone ring. Compounds 14 and 15 have an extra nitrogen atom on position 7 or 8 of the aromatic ring, and compound 16 has an electron-withdrawing group (fluoride) attached to C-8 position of the quinazolinone ring. For compounds 14–16, the C-5 and C-6 positions are now exposed to oxidation, but such a process will be hampered because the oxidation potential of the molecule is increased due to the introduction of an extra nitrogen atom or an electron-withdrawing group on the quinoline ring. Overall, these compounds closely resemble febrifugine itself by possessing a planar aromatic ring, a 1”-amino group and C-2’, C-3” O-functionality and are therefore expected to possess same or similar mode of action. Meanwhile, they would be much less likely to produce toxic intermediates because of the blocking of C-5 and C-6 position (compounds 5–13) or the increase of the oxidation potential of the molecule (compounds 5–11, 14–16).

Figure 2.

As a compare study, compounds 17–20 were also designed. These compounds have a methyl group or an electron-donating methoxy group attached to the C-7 or C-8 of the aromatic ring. These compounds have comparable or even greater tendency to undergo oxidation. Biological data from these compounds should further validate, or nullify, the hypothesis that some oxidized febrifugine metabolites have contributed to the observed toxicity.

The synthesis of compound 5 is shown in Scheme 2. Previously published procedure was adopted to synthesize the aromatic moiety: 3H-Pyrido[3,2-d]pyrimidin-4-one (21).¹⁴ Compound 21 condensed with the known epoxide 22,¹⁵ to furnish corresponding alcohol 23, as a pair of diastereomers, this follows the methods of Ogasawara, et al.¹⁵ Compound 23 then underwent TPAP oxidation¹⁶ to give ketone 24 as a single enantiomer. Finally, compound 24 underwent hydrogenolysis to afford compound 5.^{17, 18} The synthesis of compounds 6–20¹⁸ follows similar route. The respective aromatic moieties are available from previously published procedures.^{19, 20, 21}

Scheme 2.

Synthesized compounds (5–20) were tested in vitro against two P. falciparum malaria parasite clones: W2 and D6, following the procedures of Desjardins et al.²² and Chulay et al.²³ Both strains are from the Malaria Research and Reference Reagent Resource Center (MR4). The W2 clone is susceptible to mefloquine but resistant to chloroquine, sulfadoxine, pyrimethamine, and quinine, whereas the D6 clone is naturally resistant to mefloquine but susceptible to chloroquine, sulfadoxine, pyrimethamine and quinine. For in vitro toxicological studies, synthesized compounds were evaluated in freshly isolated rat hepatocytes by the measuring of lysosomal (neutral red assay) functions as cytotoxicity parameter, as described by Fontaine et al.²⁴ Freshly isolated rat hepatocytes are widely used in drug metabolism and toxicity studies. Chloroquine and febrifugine were also screened in W2, D6, and rat hepatocytes as positive controls. Their IC₅₀ values were summarized in Table 1.

Table 1.

In Vitro Activity Against Malarial Parasite and Rat Hepatocytes [IC₅₀ (nM)]

Compound	D6 Clone	W2 Clone	Rat hepatocytes
5	1.2	1.3	1.8x104
6	1.5	1.8	1.7x104
7	4.7	5.0	1.3x104
8	1.3	1.3	1.9x104
9	0.33	0.39	1.7x104
10	14	16	1.5x104
11	0.43	0.41	1.9x104
12	132	160	7.0x104
13	165	188	9.8x104
14	63	73	1.9x104
15	70	109	2.2x104
16	103	100	1.6x104
17	2.0	2.2	117
18	1.7	1.9	140
19	1.4	1.3	15
20	1.2	1.3	16
Chloroquine	23	338	1.9x104
Febrifugine	1.6	1.9	169

Compounds 5 and 6 with an extra nitrogen atom on the position 5 or 6 of the aromatic ring and compound 8 with a fluoride substitution on C-5 possess antimalarial activity comparable to febrifugine. Compounds 9 and 11 with difluoride attached to C-5 and C-6 or trifluoromethyl substitution at C-5 have antimalarial activity superior to parent compound febrifugine. These compounds showed potency against both chloroquine sensitive malarial strain (D-6) and chloroquine resistant malarial strain (W-2). Noticeably, these compounds were over 100 times less toxic than febrifugine. On the other hand, compounds with an electron-withdrawing group on C-6 alone (7 and 10) or a bulky substitution (12 and 13) have both decreased antimalarial activity and toxicity.

Although compounds 14–16 were also much less toxic than febrifugine, these compounds were less active, suggesting that increasing the oxidation potential of the molecule alone reduces both toxicity and antimalarial activity.

Compounds 17–20 possess antimalarial activity comparable to parent compound febrifugine, suggesting that introducing a small alkyl or alkoxy substitution on C-7 or C-8 of the aromatic ring will not change the biological potency of the molecule. Nevertheless, compounds 19 and 20, with an electron-donating methoxy substitution at C-7 and C-8 of the aromatic ring and hence possess greater tendency to undergo oxidation, have become 10 times more toxic. A weak electron-donating methyl substitution (compounds 17 and 18) only increases the toxicity slightly.

In conclusion, new febrifugine analogues were designed and synthesized. Lower toxicity was achieved by reducing or eliminating the tendency of forming chemically reactive and toxic intermediates and metabolites. Some of these compounds (5, 6, 8, 9, 11) possess significant low cytotoxicity, while retaining potent antimalarial activity against both chloroquine sensitive and resistant strains. These compounds, as well as the underlying design rationale, may find usefulness in the discovery and development of new antimalarial drugs.