7344
J. Med. Chem. 2008, 51, 7344–7347
Letters In Pursuit of Natural Product Leads: Synthesis and Biological Evaluation of 2-[3-hydroxy-2-[(3-hydroxypyridine-2carbonyl)amino]phenyl]benzoxazole4-carboxylic acid (A-33853) and Its Analogues: Discovery of N-(2-Benzoxazol-2-ylphenyl)benzamides as Novel Antileishmanial Chemotypes Suresh K. Tipparaju,†,§ Sipak Joyasawal,†,§ Marco Pieroni,† Marcel Kaiser,‡ Reto Brun,‡ and Alan P. Kozikowski*,† Drug DiscoVery Program, Department of Medicinal Chemistry and Pharmacognosy, UniVersity of Illinois at Chicago, 833 S. Wood Street, Chicago, Illinois 60612, and Parasite Chemotherapy, Swiss Tropical Institute, Socinstrasse, 57, P.O. Box CH-4002, Basel, Switzerland ReceiVed September 30, 2008 Abstract: The first synthesis and biological evaluation of antibiotic 31 (A-33853) and its analogues are reported. Initial screening for inhibition of L. donoVani, T. b. rhodesiense, T. cruzi, and P. falciparum cultures followed by determination of IC50 in L. donoVani and cytotoxicity on L6 cells revealed 31 to be 3-fold more active than miltefosine, a known antileishmanial drug. Compounds 14, 15, and 25 selectively inhibited L. donoVani at nanomolar concentrations and showed much lower cytotoxicity.
Leishmaniases are parasitic diseases caused by several protozoan parasites of the genus Leishmania. These parasitic agents are transmitted to humans by the bite of an insect vector, namely, the phlebotomine sand fly. Visceral leishmaniasis or Kala-azar is one of the most common pathological forms in which the disease occurs. It is mainly transmitted by L. donoVani and is lethal in 100% of the cases when left untreated.1 Among all parasitic infections, leishmaniases are the second most important from a socioeconomic point of view.2 They affect 12 million people in 88 countries,3 there are 1.5-2 million new cases and 70 000 deaths each year, and 350 million more are at risk of infection.4a Despite the increasing concerns about this disease, little effort has been made toward the development of new antileishmanial chemotherapeutics. In fact, antimonial derivatives were the firstline therapeutic option for more than 50 years.4b,c Only recently have novel antileishmanial agents such as amphotericin B, pentamidine, and the new oral drug miltefosine been added to the current therapeutic arsenal.5a However, all of these drugs suffer from several moderate to severe drawbacks. Antimonials may cause acute pancreatitis and cardiac arrhythmia and can even lead to death in extreme cases.5b Hypokalemia and nephrotoxicity are the most common side effects triggered by * To whom correspondence should be addressed. Phone: 312-996-7577. Fax: 312-413-0577. E-mail:
[email protected]. † University of Illinois at Chicago. § These authors contributed equally to this work. ‡ Swiss Tropical Institute.
amphotericin B, not to mention life-threatening first-dose anaphylaxis. Pentamidine, an aromatic diamidine, is not active orally and can lead to renal, pancreatic, and hepatic toxicity along with hypotension and dysglycemia.6 Miltefosine, a phosphocholine analogue, is the first oral antileishmanial agent used to cure both visceral and cutaneus leishmaniasis. Despite its great efficacy, miltefosine is limited by its extremely long half-life (6-8 days), low therapeutic index, and teratogenicity in animals.7a Moreover, given the unsatisfactory results of miltefosine when administered to HIV co-infected patients,7b the role of this drug for treatment of leishmaniases needs to be reconsidered. In view of the foregoing facts, there is an urgent need for the development of antileishmanial agents based on new molecular scaffolds endowed with improved efficacy and lacking toxicity. Unfortunately, our limited understanding of leishmanial biology complicates the rational design of antileishmanial agents. Thereby new drugs are often discovered serendipitously by testing large chemical libraries or by modifying structures already known to possess anti-infective activity. The latter is the case, for example, of the newly developed antileishmanial agents paromomycin8 and sitamaquine.9 Compound 3110 (Chart 1) is an antibiotic isolated a couple of decades ago from a culture broth of Streptomyces sp. NRRL 12068. The biological data revealed a high antibacterial activity for this natural product, leading to some speculation that this new benzoxazole based scaffold could be an attractive lead for the development of novel anti-infective agents. In spite of its interesting antibacterial profile, there are no reported efforts to chemically synthesize the parent natural product or to systematically explore the SAR of its analogues against various pathogens. Thus, as part of our continuing efforts toward the design and synthesis of novel anti-infective agents,11 we synthesized the parent natural product 31 and a number of its derivatives and evaluated them for their antiparasitic activity. The new structures were initially screened for their ability to inhibit the growth of L. donoVani and three other parasites, namely, T. b. rhodesiense, T. cruzi, and P. falciparum, at two different concentrations, i.e., 4.8 and 0.8 µg mL-1. Most of the synthesized derivatives showed selective inhibition against the amastigote forms of L. donoVani, and we thus ruled out the possibility of general cytotoxicity. Further determination of IC50 in amastigotes of L. donoVani and L6 cells furnished derivatives with activity comparable to that of miltefosine and a significantly improved toxicity profile. We envisaged that the antibiotic 31 and its analogues described in this paper could be synthesized from a common aminophenol intermediate 6. The synthesis of 6 is outlined in Scheme 1. Esterification of 2-nitro-3-hydroxybenzoic acid followed by protection of the phenolic hydroxyl group as a benzyl ether resulted in the intermediate 1. Saponification of the ester gave the benzoic acid derivative 2. Methyl benzoate intermediate 3 was obtained from readily available 3-hydroxyanthranilic acid. Intermediates 2 and 3 were coupled using carbonyldiimidazole to give the amide 4. Benzoxazole intermediate 5 was obtained by heating intermediate 4 to high temperatures. Hydrogenation of intermediate 5 in the presence of Pd/C resulted in the aminophenol intermediate 6 in good
10.1021/jm801241n CCC: $40.75 2008 American Chemical Society Published on Web 11/07/2008
Letters
Chart 1. Structure of inhibitors 14-31
yield. However, subsequent coupling of intermediate 6 with 3-hydroxypyridine-2-carbonyl chloride to give intermediate 7 proceeded in poor yield. The harsh reaction conditions required for the cyclization of intermediate 4 and the poor yields in the subsequent coupling reaction have prompted us to explore an alternative synthetic strategy for the synthesis of 31 and its analogues. An alternative synthetic route for the synthesis of 31 is shown in Scheme 2. The benzoxazole intermediate 9b was prepared by oxidative cyclization of the Schiff base obtained from commercially available 3-methoxy-2-nitrobenzaldehyde and aminophenol 312 or via cyclization-dehydration reaction of amide 8 using POCl3 in refluxing xylene. Reduction of the nitro functionality of intermediate 9b resulted in amine 10b. Selective O-benzylation of 3-hydroxypicolinic acid followed by hydrolysis of the benzyl ester led to the formation of the intermediate 11
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23 7345
in very good yield. Coupling of 3-benzoyloxypicolinic acid prepared from 11 with the amine 10b resulted in intermediate 13. Treatment of 13 with excess BBr3 gave antibiotic 31 in good overall yield. We found this synthetic strategy to be amenable for the ready scale-up of intermediates 10 and for the quick and easy synthesis of a small library of analogues of 31. Analogues 14-30 (Chart 1) of 31 were synthesized starting from the common intermediates 10a and 10b (Scheme 3). Intermediates 14a-30a were obtained by coupling intermediate 10a or 10b with an appropriate acid chloride. The acid chlorides were commercially available or were prepared from the readily available carboxylic acids by treating them with oxalyl chloride in the presence of catalytic DMF. Phenolic compounds 14-19, 21-23, 25, 26, and 28-30 were obtained upon prolonged reaction of the corresponding methoxy intermediates with excess BBr3. Compounds 20, 24, and 27 were obtained by hydrolysis of the corresponding intermediates 19a, 24a, and 26a with LiOH, respectively. Antibiotic 31 was initially tested for its ability to inhibit the growth of four protozoans, namely, T. b. rhodesiense, T. cruzi, P. falciparum, and L. donoVani, at two different concentrations (4.8 and 0.8 µg mL-1). It showed 100% inhibition of L. donoVani at both concentrations. However, 31 also showed notable activity against T. cruzi (96% inhibition) and P. falciparum (97% inhibition) at 4.8 µg mL-1. Thus, to rule out the possibility that this activity was due to general cytotoxicity, we determined its IC50 against axenic amastigote forms of L. donoVani MHOM/ET/67/L82,13 as well as toxicity to L6 cells.13 Miltefosine, a known antileishmanial agent, was included in the study as a control drug (Table 1). Compound 31 was found to be 3-fold more active (IC50 ) 80 nM) than miltefosine, but it also showed modest toxicity toward the L6 cells (IC50 ) 14 µM). Thus, to circumvent the cytotoxicity of 31, we undertook the synthesis of a variety of structural analogues in the hope of improving the activity against L. donoVani while reducing the cytotoxicity. The first set of structural changes were made by modifying the picolinic acid group (R1 group) while retaining the benzoxazole moiety. For all the analogues of 31 that showed reasonable activity against L. donoVani (>60% inhibition at 0.8 µg mL-1), along with good selectivity (absence of growth inhibition against the other parasites), IC50 values in the axenic amastigote form of L. donoVani and in L6 cells were determined. The results are shown in Table 1. Among the analogues made, 14, 15, and 25 displayed the most noticeable activity, with IC50 in the nanomolar range and comparable to that of miltefosine. We were pleased to note that 15, a 3-fluoropyridine analogue of 31, maintained good parasitic activity while showing lower toxicity toward the L6 cells than miltefosine itself. For 14, in which the hydroxyl group of the 3-hydroxypyridine moiety has been deleted, a drop in activities against the parasite and the L6 cells was observed. In contrast, replacement of the hydroxyl group with a fluorine atom led to a considerable reduction in cytotoxicity while maintaining good antiparasitic activity. A 10-fold reduction in antileishmanial activity was observed when the 3-hydroxypyridine group of the parent natural product was replaced by a 2-hydroxyphenyl group (16). A similar trend was also found upon comparing the activity of the pyridyl derivative 14 and the phenyl derivative 23, leading to the suggestion that a nitrogen containing ring imbues better activity than the phenyl ring as the R1. Further, it is interesting to note a considerable shift in the activity profile when the R1 group is changed from a 2-pyridyl unit (14) to a 3-pyridyl unit (18), indicating that the
7346 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 23
Letters
Scheme 1. Synthesis of Common Intermediate 6a
a Reagents and conditions: (a) (CH3)3SiCHN2, Et2O/MeOH (1:1), 0 °C, 2 h (96%); (b) BnBr, K2CO3, CH2Cl2/MeOH, reflux, 8 h (95%); (c) 1.4 N NaOH, dioxane, reflux, 3 h (94%); (d) CDI, THF, reflux, 18 h (94%); (e) 230 °C, 3 h (33%); (f) Pd/C, H2, MeOH, room temp, 12 h (76%); (g) 3-hydroxypyridine2-carbonyl chloride, pyridine, DMAP (cat.), CH2Cl2, room temp, 16 h (5%).
Scheme 2. Synthesis of Antibiotic 31a
a Reagents and conditions: (a) 3, CDI, THF, reflux, 18 h (60%); (b) POCl3, xylene, 140 °C, 3 h (58%); (c) 3, MeOH, 40 °C, 12 h; (d) DDQ, CH2Cl2, room temp, 12 h (37%); (e) Pd/C, H2, MeOH, room temp, 12 h (80%); (f) BnBr, Ag2O, CH2Cl2/DMSO (1:1), 4 Å, room temp, 12 h (70%); (g) LiOH, THF/ H2O/MeOH (3:1:1), room temp, 12 h (94%); (h) (COCl)2, room temp, 3 h; (i) pyridine, DMAP (cat.), CH2Cl2, room temp, 12 h (34%); (j) excess BBr3, CH2Cl2, -78 °C to room temp, 16 h (81%).
Scheme 3. Synthesis of Inhibitors 14-30a
a Reagents and conditions: (a) R1COCl, pyridine, DMAP (cat.), CH2Cl2, room temp, 12 h; (b) excess BBr3, CH2Cl2, -78 °C to room temp, 16 h; (c) LiOH, THF/H2O/MeOH (3:1:1), room temp, 12 h.
presence of a nitrogen atom ortho to the amide linkage is important for activity and for reducing the cytotoxicity of this series of compounds. If hydroxylation of the aromatic ring is considered to be an advantageous substitution based upon the activities of derivative 16 vs 23 as well as 31 vs 14, the lack of activity of 22 may be attributed to poor penetration through the protozoan membrane due to its greater hydrophilic character. Seemingly the presence of a fluorinated aromatic ring as the R1 group has a favorable effect on the cytotoxicity of these compounds. For example, 15 is 7-fold less toxic to L6 cells when compared to the unfluorinated 14. Although polyfluorinated derivative 21 is moderately active, it is considerably less toxic to L6 cells than miltefosine and 31. The presence of a chlorine substituent on
Table 1. Activity of 31 and Its Analogues in Axenic Amastigote Cultures of L. donoVani compd
IC50 (µM)
L6 cells, IC50 (µM)
SIc
miltefosine 31 14 15 16 17 18 19b 20b 21 22b 23 24 25 26 27 28 29 30b
0.26 0.08 0.31 0.52 0.85 1.93 4.36 NAa NAa 1.90 NAa 1.73 0.96 0.51 0.68 3.07 1.17 0.93 NAa
147.0 14.2 30.6 203.7 85.1 143.0 240.0 NAa NAa 210.0 NAa >240.0 114.4 94.4 212.3 211.0 108.0 141.0 NAa
565 185 99 392 100 74 55 NAa NAa 111 NAa >138 118 185 312 69 92 152 NAa
NA: not attainable. b Only
