Synthesis and characterization of novel bioactive 1,2,4-oxadiazole natural product analogs bearing the N-phenylmaleimide and N-phenylsuccinimide moieties

Taking into consideration the biological activity of the only natural products containing a 1,2,4-oxadiazole ring in their structure (quisqualic acid and phidianidines A and B), the natural product analogs 1-(4-(3-tert-butyl-1,2,4-oxadiazol-5-yl)phenyl)pyrrolidine-2,5-dione (4) and 1-(4-(3-tert-butyl-1,2,4-oxadiazol-5-yl)phenyl)-1H-pyrrole-2,5-dione (7) were synthesized starting from 4-(3-tert-butyl-1,2,4-oxadiazol-5-yl)aniline (1) in two steps by isolating the intermediates 4-(4-(3-tert-butyl-1,2,4-oxadiazol-5-yl)phenylamino)-4-oxobutanoic acid (3) and (Z)-4-(4-(3-tert-butyl-1,2,4-oxadiazol-5-yl)phenylamino)-4-oxobut-2-enoic acid (6). The two natural product analogs 4 and 7 were then tested for antitumor activity toward a panel of 11 cell lines in vitro by using a monolayer cell-survival and proliferation assay. Compound 7 was the most potent and exhibited a mean IC50 value of approximately 9.4 µM. Aniline 1 was synthesized by two routes in one-pot reactions starting from tert-butylamidoxime and 4-aminobenzoic acid or 4-nitrobenzonitrile. The structures of compounds 1, 2, 4, 5 and 6 were confirmed by X-ray crystallography.


Introduction
The five-membered heterocyclic 1,2,4-oxadiazole motif is of synthetic and pharmacological interest. It also forms an important constituent of biologically active compounds including natural products [1]. Sawyer et al. have described such compounds as bioisosteres for amides and esters [2], with the 1,2,4oxadiazoles showing higher hydrolytic and metabolic stability.
To the best of our knowledge, there are only a few examples of natural products with a 1,2,4-oxadiazole core or a structure based on it. The 3-substituted indole alkaloids, phidianidines A and B (Figure 1), have been isolated by Carbone et al. from the aeolid opisthobranch Phidiana militaris [3]. They are selective inhibitors of the dopamine transporter DAT and partial agonists of the μ opioid receptor [4]. Moreover, these selective molecules are attractive as CNS targets because neither phidianidine A nor B is cytotoxic. Another example of a natural product with a oxadiazole core is quisqualic acid (Figure 1). This metabolite was obtained from the seeds of Quisqualis indica and Q. fructus [5,6] and is a strong agonist for AMPA (α-amino-3-hydroxy-5methyl-4-isoxazolepropionic acid) receptors and group I metabotropic glutamate receptors [7]. Furthermore, 1,2,4-oxadiazoles are widely used in synthetic chemistry, e.g., in the search for antitumor agents. Cancer consists of more than one hundred different diseases, all of which are characterized by the uncontrolled growth and spread of abnormal cells. In this context, the identification of drugs acting as apoptosis inducers represents an attractive approach for the discovery of new anticancer agents. 1,2,4-oxadiazole A ( Figure 2) was found to act as an apoptosis agent by a highthroughput screening (HTS) assay [8]. A series of 1,2,4-oxadiazole-5-carboxamides B have been synthesized and tested as inhibitors of the glycogen synthase kinase 3 (GSK-3), a key regulator of both differentiation and cellular proliferation [9].
The maleimide motif is also a useful five-membered heterocycle in pharmacological chemistry. Kratz et al. synthesized maleimide derivatives of doxorubicin and camptothecin. After intravenous administration these designed anticancer drugs bind Scheme 1: Common synthetic strategies toward 1,2,4-oxadiazoles; (a) amidoxime route; (b) 1,3 dipolar cycloaddition route. rapidly to circulating albumin [17][18][19]. Endogenous albumin could be seen as a drug carrier, as it accumulates in solid tumors according to the pathophysiology of tumor tissue [20,21]. Therefore, designed prodrugs have a higher antitumor efficacy in vivo than drugs. Furthermore, maleimides possess strong antifungal activities against important human opportunistic pathogenic fungi. These antifungal drugs appear to be excellent candidates for further development [22][23][24][25][26][27]. Barrett et al. point out that the possibility of performing chemical modifications is a requirement for developing novel drugs, a strong activity is just the starting point [28].
Another moiety worth investigation is succinimide, because N-phenylsuccinimides are regarded as some of the most efficacious agricultural fungicides [29,30]. They have also been shown to be selective nephrotoxic compounds [31,32].
Considering natural products with the 1,2,4-oxadiazol moiety, such as phidianidines A and B (selective inhibitors of DAT), we decided to synthesize, isolate and characterize novel natural product analogs of 1,2,4-oxadiazole derivatives bearing N-phenylmaleimide or N-phenylsuccinimide functionalities in order to improve their biological activity. The new derivatives have been tested for in vitro antitumor activity toward a panel of 11 cell lines.

Results and Discussion
Clapp reviewed the synthesis of 1,2,4-oxadiazoles [33]. He pointed out that two general methods dominate the practical preparation (≈95%): (a) The condensation of amidoximes with carboxylic acid derivatives.
(b) The dipolar cycloaddition of nitrile oxides to nitriles.
The general approach for the synthesis of 1,2,4-oxadiazoles is illustrated in Scheme 1.
Using route (a), the amidoxime route, the carboxylic acid has to be employed in an activated form. The activated carboxylic acid can be prepared beforehand or in situ by several methods [34], e.g., as an acyl chloride or by the use of N,N′-carbonyldiimidazole (CDI). In the first step the amidoxime is O-acylated with the activated derivative in a condensation reaction. The O-acylated amidoxime can be isolated or it can immediately undergo the cyclisation to the heterocyclic oxadiazole ring. This cyclodehydration reaction takes place by heating to temperatures above 100 °C [35,36]. Microwave techniques have also been employed in the synthesis of such heterocycles. The advantage of CDI is that it activates the carboxylic acid in situ and can be used for step 1 and step 2 in DMF.

Scheme 2:
One-pot synthesis of 4-(3-tert-butyl-1,2,4-oxadiazol-5-yl)aniline (1) by using the amidoxime route. i. 1.1 equiv CDI in DMF, 30 minutes; ii. 1.1 equiv CDI in DMF, 120 °C, 4 h.  The structure of compound 1 was confirmed by X-ray structure determination ( Figure 3 and Figure 4). It crystallizes with two molecules in the asymmetric unit, which differ in the relative orientation of the rings (interplanar angles 22° and 9°). Three of the four NH hydrogens are involved in hydrogen bonds, leading to ribbons of H-bonded rings parallel to the a axis.
amidoxime is activated by PTSA-ZnCl 2 ; resulting in the formation of a Lewis acid-ammonia complex as a leaving group, giving rise to the formation of the nitrile oxide. The 1,2,4oxadiazole moiety is established by the 1,3-dipolar cycloaddition of nitrile oxide to the 4-aminobenzonitrile. However, the Lewis acid might also be involved in the formation of the heterocycle via a Lewis acid catalyzed [3 + 2] cycloaddition reaction. Unfortunately, the yield for this reaction was very low (<20%).
In order to obtain compound 1 by the 1,3 dipolar cycloaddition route we changed the protocol. We synthesized the nitro derivative 2 (3-tert-butyl-5-(4-nitrophenyl)-1,2,4-oxadiazole) in situ by using the same catalyst pair as in the first synthesis route. Then, we attempted to hydrogenate nitro compound 2 to the corresponding amine 1 in the same reaction pot without having to isolate the intermediate. The overall yield in this case was 64%. The intermediate 2 was isolated to be fully charazterized. After a series of tests by using various acids as catalyst (p-toluensulfonic acid (PTSA), 2-mesitylenesulfonic acid (MSA) and methanesulfonic acid (MeSA) in combination with ZnCl 2 and ZnBr 2 ), MSA-ZnBr 2 in acetonitrile proved to be the best combination for the preparation of compound 2 from tert-butylamidoxime and 4-nitrobenzonitrile under mild conditions. The results are summarized in Table 1. The optimized yield was 93% (Scheme 3) which makes this route more practical than the amidoxime route presented in Scheme 2 with a yield of only 59%.
The structure of compound 2 was confirmed by X-ray structure determination ( Figure 5 and Figure 6). The interplanar angle in compound 2 is only 3° and the molecules are linked to ribbons parallel to the b axis by two C-H···O interactions.  Imide derivatives have been found to possess a broad spectrum of biological activities. A variety of methods have been reported for the preparation of this class of compounds. The synthesis follows a two-step protocol. First, it is necessary to synthesize the amide derivative. The synthesis of amide 3 was performed under inert conditions by mixing an equimolar amount of aniline 1 and succinic anhydride in a minimum volume of dichloromethane (Scheme 4). Compound 3 was obtained with a short reaction time and a high yield (91%). The ester derivative methyl 4-(4-(3-tert-butyl-1,2,4-oxadiazol-5-yl)phenylamino)-4-oxobutanoate (5) was prepared by the addition of a diethyl ether solution of diazomethane to a suspension of amide 3. The ester 5 was afforded in a quantitative amount.
For the synthesis of imide 4 the starting material, amide 3, was mixed with an equimolar amount of sodium acetate in acetic anhydride and the mixture was heated for 4 h at 80-85 °C; resulting in the corresponding N-arylsuccinimide 4 (87%).
The structures of compounds 4 and 5 were also confirmed by X-ray structure analysis ( Figure 7, Figure 8 and Figure 9). In compound 4 the oxazoline and phenyl rings are approximately coplanar (6°), but the pyrrolidine ring is rotated by 52° with respect to the phenyl ring. The main packing interaction is an  The synthesis of amide 6 was performed by mixing an equimolar amount of aniline 1 and maleic anhydride in dichloromethane (Scheme 5). The amide 6 was obtained with a good yield (84%). For the synthesis of imide 7 the amide 6 was mixed with an equimolar amount of sodium acetate in acetic anhydride and the mixture was heated for 4 h at 80-85 °C resulting in the corresponding N-arylmaleimide 7 (75%).
The structure of compound 6 was also confirmed by X-ray structure analysis ( Figure 10 and Figure 11). In compound 6 the

Antitumor activity
The in vitro antitumor activity of six synthesized compounds toward a panel of 11 cell lines was assessed by using a monolayer cell survival and proliferation assay. By exhibiting a mean IC 50 value of 9.4 µM 7 was the most potent compound. Compounds 1, 5, 4, 6 and 3 have only marginal antitumor activity toward the 11 tested cell lines ( Figure 12).
An investigation of the activity of compounds in a cell line panel representing various tumor histo-types, as performed in this study, allows the analysis of potency and tumor selectivity and the identification of active compounds which qualify for further preclinical development. Tumor selectivity of the compounds is illustrated in Figure 13 by a heat-map presentation of the individual IC 50 values. Good tumor selectivity is reflected by cell lines exhibiting above-average sensitivity (i.e., an individual IC 50 value smaller than 0.5 of the mean IC 50 value;

Conclusion
Two novel 1,2,4-oxadiazol natural product-inspired derivatives bearing N-phenylmaleimide or N-phenylsuccinimide moieties were synthesized in two steps starting from 1,2,4-oxadiazole amine 1. All intermediate derivatives were isolated, characterized and tested in vitro for antitumor activity toward a panel of 11 cell lines by using a monolayer cell survival and proliferation assay. With an IC 50 value of 9.4 µM, compound 7 was the most active, probably because maleimide is able to rapidly bind to the cysteine-34 position of circulating albumin, transported as an albumin conjugate (macromolecular drug), and then released in the target tissue by acid related cleavage or enzymatic cleavage. Compounds 1, 3, 4, 5 and 6 have no or only marginal activity toward the 11 cell lines tested. Considering the bioactive natural products (quisqualic acid and phidianidines A and B), it is of great pharmacologic interest to synthesize new derivates of 1,2,4-oxadiazole with an efficient biological transport in cells by using natural transporters such as amino acids, peptides or sugars. Moreover, it is desirable to reduce, or even remove completely, the secondary effects to minimize the toxicity and increase the selectivity. All derivatives were obtained with a high purity (at least 97% based on 1 H NMR) and good to high yields (75-96%). The structural assignments were corroborated by X-ray structure analysis.

Experimental General
All reagents were purchased from commercial sources (Sigma-Aldrich or Acros) and used without further purification. Solvents were of analytical grade. 1

X-ray structure determinations
Data are summarized in Table 2. Intensities were determined at 100 K on Oxford Diffraction diffractometers by using monochromated Mo Kα or mirror-focussed Cu Kα radiation. Structures were refined on F 2 using the program SHELXL-97 [38]. Hydrogen atoms were refined either freely (NH), by using rigid methyl groups allowed to rotate but not tip or by using a riding model starting from calculated positions. Structure 5 was treated as a non-merohedral twin.

In vitro antitumor activity toward human tumor cell lines
Antitumor activity of the compounds was tested in a monolayer cell survival and proliferation assay with human tumor cell lines. Studies carried out with panels of human tumor cell lines of different origin and histotype allow for the analysis of potency and tumor selectivity of test compounds.
Ten out of the eleven tested cell lines were established at Oncotest from patient-derived human tumor xenografts passaged subcutaneously in nude mice [39]. The origin of the donor xenografts was described [40,41]. The cell line HT-29 was kindly provided by the National Cancer Institute (Bethesda, MA, USA). Cells were cultured in RPMI 1640 medium, supplemented with 10% fetal calf serum and 0.1 mg/mL gentamicin under standard conditions (37 °C, 5% CO 2 ). Authenticity of all cell lines was proven by STR analysis at the DSMZ (Braunschweig, Germany).
A modified propidium iodide assay was used to assess the compounds' activity toward human tumor cell lines [42]. Briefly, cells were harvested from exponential phase cultures by trypsinization, counted and plated in 96-well flat-bottom microtiter plates at a cell density dependent on the cell line (4.000-20.000 cells/well). After 24 h recovery period to allow the cells to adhere and resume exponential growth, compounds were added at 10 concentrations in half-log increments and incubated for 4 days. The inhibition of proliferation was determined by measuring the DNA content with an aqueous propidium iodide solution (7 μg/mL). Fluorescence was measured by using a Cytofluor micro-plate reader (excitation λ = 530 nm, emission λ = 620 nm) and thus establishing a direct relationship to the total viable cell number. In each experiment, all measurements were carried out twice.