One step in pyrimidine biosynthesis

One step in pyrimidine biosynthesis is conversion of l-dihydroorotate (DHO) to orotate (ORO), under action of the enzyme dihydroorotate dehydrogenase (DHODH, EC 1.3.99.11), which contains a flavine (FMN) as redox cofactor [1]. In this transformation, electrons resulting from DHO oxidation are transferred to ubiquinone, and finally to cytochrome c oxidase. Previous studies have shown that this enzyme possesses two distinct binding sites, respectively for DHO/ORO and ubiquinone. Oxidation of DHO to ORO is the rate-limiting step of the whole pyrimidine biosynthetic pathway. Inhibitors of DHODH display antiproliferative and immunomodulator activities, and have shown potential benefits for treating rheumatoid arthritis (RA) [2], [3], [4], [5]. RA is a chronic inflammation characterized by painful joints and progression to irreversible joint damage [6]. It is currently regarded as an immunological process involving a variety of mediators [7], [8]. Treatment relies on the traditional non-steroidal anti-inflammatory drugs (NSAIDs) or selective COX-2 inhibitors (Coxibs) [9]. While these drugs may provide symptomatic relief, they do not modify the course of the disease. The main shortcoming of NSAIDs is their marked gastrotoxicity, whereas Coxibs potentially increase the risk of heart attack and stroke [10], [11]. Generally, both drugs are used in combination with disease-modifying anti-rheumatic drugs (DMARDs) for the management of established rheumatoid arthritis. Methotrexate is the principal DMARD, but other drugs in use include sulphasalazine, penicillamine, hydroxychloroquine and chloroquine, gold salts, cyclosporine and glucocorticoids [9], [12]. Hepatotoxicity, blood dyscrasias, intestinal and lung diseases are serious adverse effects of these drugs. Recently, biological agents able to activate pro-resolving biochemical pathways have been introduced onto the market; these include inhibitors of tumor necrosis factor-α (TNF-α). These inhibitors are used when arthritis is uncontrolled and serious toxic effects have arisen with DMARDs [9]. Inhibitors of DHODH have also been considered for treating RA; leflunomide (1) and brequinar (BQN, 2) are the two most important such drugs (Fig. 1) [13], [14]. The former is an isoxazol derivative, which behaves as a prodrug: upon absorption, it is rapidly converted into its active metabolite A771726 (3), which was assigned the Z-configuration. The latter was developed for cancer therapy, but failed in clinical trials due to limited therapeutic windows [15]. The flunixin meglumine sale for both A771726 and BQN are located at the narrow end of the channel that ubiquinone uses to accomplish a redox reaction with FMNH2. Several charged and polar side chains (Gln47, His56, Tyr356, Thr360, and Arg136) are present at this position. In the case of A771726, the deprotonated enolic OH group interacts via hydrogen bonding with Tyr356, while the amide carbonyl is hydrogen bonded to Arg136 through a water molecule [16]. Moving to BQN, the binding mode is quite different. Here, the carboxylate group forms a salt bridge with Arg136, and is hydrogen bonded to Gln47. In addition, the biphenyl moiety establishes a number of hydrophobic contacts with several lipophilic residues of the tunnel. In a development of our quest for drugs to use as anti-inflammatory agents [17], [18], [19] we recently explored the possibility of using the 1,2,5-oxadiazole ring (furazan) as a replacement for the isoxazole moiety present in 1[20]. Similarly to isoxazole in 1, the furazans we designed (general structure 4) are also easily opened under physiological conditions, to yield the related α-oximinoacetonitrile derivatives (cyano-oximes) 5 (Fig. 2). These products proved to be very poor DHODH inhibitors, probably due to the unfavorable E-configuration of the oxime moiety generated upon ring opening, which prevents hydrogen bonding to Tyr356 [20]. Since the attempt to restore this interaction by oxidizing the oximes to nitroderivatives was only partially successful, we decided to pursue another strategy, consisting in the replacement of the unsubstituted furazan in 4 with a hydroxyfurazan ring (7a, Scheme 1). The latter moiety is stable under physiological conditions, and should provide the correct orientation for the deprotonated hydroxyl, in order to mimic the enolate group of A771726 and interact likewise with Tyr356. The activity of 7a, the first hydroxyfurazan analog of A771726 that was tested, was not of great interest (4.3 μM), but nonetheless played a key role in the further development of our project. When 7a was docked in silico at the DHODH active site, a BQN-like pose was obtained with the deprotonated furazan hydroxyl facing Arg136 instead of Tyr356. This amounted to a 180° flip of the furazan ring with respect to our expectations. From this observation, the suggestion arose that the hydroxyfurazan scaffold might have the potential to bind to DHODH like brequinar itself, thanks to its known bioisosterism with the carboxylic group [21]. A number of biphenyl derivatives linked to pentatomic arene- and cyclopentenecarboxylic acids through an amide bond (6, Fig. 3) were recently proposed as potent DHODH inhibitors. These inhibitors were found, by X-ray crystallography, to assume different poses in the ubiquinone binding site of DHODH, in some cases similar to BQN, in some others to A771726 [22], [23]. This paper describes a series of hydroxyfurazans (Scheme 1) which display moderate-to-high DHODH inhibitory activity, largely dependent on the substitution pattern at the biphenyl ring.