Nonsteroidal Anti-Inflammatory Drugs (NSAIDs): Progress in Small Molecule Drug Development

Ever since the discovery of aspirin, small molecule therapeutics have been widely prescribed to treat inflammation and pain. Aspirin and several small molecule NSAIDs are known to inhibit the enzymes cyclooxygenase-1 (COX-1) and -2 (COX-2). Despite the success of NSAIDs to treat inflammatory disorders, the development of a clinically useful small molecule NSAIDs with decreased side effect profiles is an ongoing effort. The recent discovery and development of selective COX-2 inhibitors was a step toward this direction. Emerging trends are represented by the progress in the development of hybrid agents such as nitric oxide donor-NSAIDs (NO-NSAIDs) and dual COX/lipoxygenase (LOX) inhibitors. This review focuses on the recent advances in the rational design of small molecule NSAIDs in therapy.


Introduction
The history of treating fever, pain and inflammation is a fascinating tale of human adventure that goes back centuries [1]. Since the discovery and isolation of salicin from willow bark in the early 18th century to the development of selective COX-2 inhibitors in the 1990s, small molecule therapies to treat fever, pain and inflammation have evolved [1,2]. Traditional NSAIDs such as aspirin (1), ibuprofen (2) and diclofenac (3) that exhibit nonselective COX inhibition represent some of the most widely prescribed NSAIDs to relieve short term fever, pain and inflammation [3,4]. The characteristic OPEN ACCESS feature of these traditional nonselective COX inhibitor NSAIDs was the presence of a carboxylic acid (COOH) functional group. In the early 1990s the second isoform of COX was discovered, providing a novel target to develop anti-inflammatory agents with superior safety profiles compared to traditional NSAIDs [5,6]. Consequently, selective COX-2 inhibitors (coxibs) based on a diarylheterocyclic ring template as in celecoxib (4) and rofecoxib (5) were developed [7,8]. These agents were characterized by the presence of a para-sulfonamide (SO 2 NH 2 ) or a paramethanesulfonyl (SO 2 Me) pharmacophore present on one of the aryl rings. Crystal structure studies supported the hypothesis that the p-SO 2 NH 2 or p-SO 2 Me pharmacophore was conferring COX-2 selectivity by orienting in a secondary pocket accessible only in the COX-2 active site [9,10]. The initial euphoria surrounding the selective COX-2 inhibitors, was short lived as studies indicated serious risks of cardiovascular complications in susceptible population during therapy [11,12]. Therefore, developing novel orally active small molecule anti-inflammatory agents with superior safety profile presents a significant challenge. The inflammatory pathway is a complex event involving multiple effectors ( Figure 2). Inflammatory mediators such as prostaglandins (PGs), leukotrienes (LTs) and tumor necrosis factor-alpha (TNF-α) are implicated in a wide variety of diseases such as rheumatoid arthritis (RA), osteoarthritis (OA), asthma, atherosclerosis, different types of cancers and diseases of the central nervous system [13][14][15][16][17][18][19][20][21][22]. Traditional NSAIDs targeted COX isozymes, whereas later studies investigated nitric oxide (NO) donating NSAIDs (NO-NSAIDs), dual COX/LOX inhibitors, leukotriene receptor antagonists and selective COX-2 inhibitors in an effort to develop antiinflammatory agents with superior safety profile [5,15,[23][24][25]. Emerging small molecule targets includes phospholipases (PLA 2 ), microsomal prostaglandin E 2 synthase (mPGES-1) and inhibition of TNF-α. This review will focus on the recent drug discovery efforts toward developing novel small molecule ring templates as NO-NSAIDs, selective COX-2, dual COX/LOX, lipoprotein-PLA 2 (Lp-PLA 2 ), mPGES-1 and TNF-α inhibitors.

NO-NSAIDs
The concept of developing hybrid NO-NSAIDs was primarily conceived to decrease the gastrointestinal (GI) toxicities observed with traditional NSAID use. In the GI tract NO is known to exert its protective role by increasing the mucous secretion, mucosal blood flow and inhibition of neutrophil aggregation [24]. In addition, the recent controversy surrounding the cardiovascular side effects of selective COX-2 inhibitors, further supports the need to develop clinically useful NO-NSAIDs since NO is also known to exhibit beneficial effects on the cardiovascular system by inhibiting platelet aggregation and adhesion [11,26]. Accordingly, several studies focused on developing NO-NSAIDs based on the aspirin, naproxen and diclofenac ring templates. These agents contain organic nitrates or nitrosothiols as the NO-donor moiety [24,27,28].
Diazeniumdiolates (NONOates) represent a unique structural moiety that can be incorporated to develop NO-donating agents [29]. Recently, Knaus and coworkers described the design and synthesis of hybrid aspirin, ibuprofen and indomethacin derivatives coupled to diazeniumdiolates as novel NOdonating prodrugs.
The ever popular agent aspirin continues to be the focus of current research [34][35][36]. Recent studies have reported novel aspirin and aspirin derivatives possessing organic nitrate NO-donor moiety. In this regard, Gasco and coworkers prepared novel aspirin-like derivatives based on salicylic acid ring template possessing a nitrooxy-acyl NO-donor moiety (12) and aspirin derivatives possessing a (nitrooxyacyloxy)methyl ester NO-donor moiety (13). These agents exhibited effective oral antiinflammatory and vasodilatory properties with reduced GI toxicities [36].

Selective COX-2 Inhibitors
The adverse cardiovascular events associated with selective COX-2 inhibitors led to a dramatic decline in selective COX-2 inhibitor pipeline. In this regard, Tragara Pharmaceuticals is developing an orally active, selective COX-2 inhibitor (COX-2 IC 50 = 0.31 µM; COX-1 IC 50 = 2.2 µM) to treat different types of neoplasia such as tumors of lung, breast and pancreas [37]. Apricoxib has a 1,2diphenyl template attached to a central 5-membered pyrrole ring along with a para-SO 2 NH 2 COX-2 pharmacophore (14, Figure 4). In addition, a group of regioisomeric 1,5-diphenylpyrroles possessing a para-SO 2 Me COX-2 pharmacophore were reported as selective COX-2 inhibitors (15, COX-2 IC 50 = 2.1 µM; COX-1 IC 50 = 20.4 µM; 16 COX-2 IC 50 = 0.018 µM; COX-1 IC 50 > 100 µM) that exhibited effective oral anti-inflammatory and analgesic activities. However, these studies did not report the cardiovascular safety of these agents [38][39][40]. Furthermore, Pfizer scientists reported a group of benzopyran derivatives as selective COX-2 inhibitors [41]. These are structurally different from the diarylheterocyclic class of selective COX-2 inhibitors. A representative compound from this series SC-75416 (17) contains a COOH and a CF 3 substituent. It is noteworthy that this agent did not contain either a SO 2 NH 2 or a SO 2 Me COX-2 pharmacophore. Compound 17 exhibited effective oral anti-inflammatory activity and similar COX inhibition/selectivity (COX-2 IC 50 = 0.25 µM; COX-1 IC 50 = 49.6 µM) relative to celecoxib. In another study, Renard and coworkers designed novel nimesulide derivatives as selective COX-2 inhibitors [42]. The alkanesulfonamide (MeSO 2 NH) in nimesulide was replaced with a trifluoromethanesulfonamide (CF 3 SO 2 NH) moiety and the ether linkage was replaced with a secondary amine bridge. A representative agent 18 ( Figure 4) exhibited a good combination of oral antiinflammatory activity and in vitro COX-2 selectivity (COX-2 IC 50 = 0.12 µM; COX-1 IC 50 = 0.91 µM). In 2009, GlaxoSmithKline (GSK) scientists reported the development of a novel series of trifluoromethylpyrimidine based ring scaffolds (19, COX-2 IC 50 = 206 nM; COX-1 IC 50 = 62000 nM) as highly potent and selective COX-2 inhibitors [43,44]. Accordingly, several diverse classes of selective COX-2 inhibitors have been reported and a thorough discussion is beyond the scope of this review [45]. It should be noted that COX-1/COX-2 inhibition and selectivity data is highly variable based on the biochemical assay method used. In addition, in vivo antiinflammatory/analgesic activities and side effects (GI, renal and cardiovascular) of NSAIDs are highly dose dependent. These factors contribute to the difficulty associated in determining specific COX-1 and COX-2 selectivity ratios for future development.

Dual COX/LOX Inhibitors
Currently, LOXs are potential targets in the treatment of diseases such as asthma, atherosclerosis, cancer, and a variety of inflammatory conditions [14][15][16]. It was hypothesized that blocking the arachidonic acid (AA) metabolism via COX inhibition by either traditional NSAIDs or selective COX-2 inhibitors could lead to the generation of proinflammatory leukotrienes and lipoxins via the LOX pathway ( Figure 2  However, for this series of compounds, in vivo anti-inflammatory activities were not reported [46]. Furthermore, Lagunin and coworkers recently used structure-based virtual screening to identify suitable ring scaffolds as dual COX/LOX inhibitors [47]. This study revealed that a thiazolidinone ring scaffold could be used to develop novel anti-inflammatory agents. Compound 21 ( Figure 5) exhibited weak in vitro COX and soyabean LOX inhibitory potency (COX-2 IC 50 = 262 µM, COX-1 IC 50 = 125 µM; LOX IC 50 = 125.9 µM). In vivo 21 exhibited good anti-inflammatory activity (44.5% inhibition, dose = 0.01 mmol/kg) when administered through intraperitoneal route in animal models. However, oral activity was not reported.

Lp-PLA 2 Inhibitors
The phospholipase A 2 (PLA 2 ) enzyme catalyzes the release of fatty acids such as AA, a critical ratelimiting step, by acting on membrane phospholipids ( Figure 2). The released AA gets converted to various pro-inflammatory mediators such as prostaglandins, leukotrienes and platelet-activating factor (PAF) that are known to play a major role in regulating the vascular tone [53]. The PLA 2 is classified into three major subtypes: secretory (sPLA 2 ); cytosolic or Ca 2+ -activated (cPLA 2 ); and inducible or Ca 2+ -independent (iPLA 2 ). In this regard, Lp-PLA 2 also known as platelet-activating factor acetylhydrolase (PAF-AH) is a Ca 2+ -independent PLA 2 that is classified as group VIIA PLA 2 . Furthermore, recent studies have indicated that Lp-PLA 2 is closely involved in the onset and progression of atherosclerosis [53][54][55][56][57]. The enzyme Lp-PLA 2 or PAF-AH (EC 3.1.1.47) was first identified from plasma that was known to hydrolyze/inactivate PAF, a phospholipid mediator produced from macrophages, monocytes, platelets and neutrophils involved in inflammatory diseases including atherosclerosis [58,59]. In humans, Lp-PLA 2 is primarily produced from leukocytes and macrophages and is associated with circulating macrophages and low-density lipoproteins (LDL). It acts on polar phospholipids in oxidized LDL to form lysophosphatidylcholine and nonesterified phospholipids that are known to have proinflammatory properties by activating and recruiting macrophages/monocytes mediating plaque vulnerability, apoptosis, leading to onset and progression of atheroma [60,61]. These studies suggest that Lp-PLA 2 is a unique biomarker to predict long-term cardiovascular risk [62][63][64].
The drug discovery of novel small molecule PLA 2 inhibitors is an ongoing effort [65,66]. Several indole-based inhibitors of sPLA 2 have been developed to treat various inflammatory conditions such as pancreatitis, allergic rhinitis, rheumatoid arthritis, gout and atherosclerosis. For example, the indole derivative varespladib 24 (s-PLA 2 IC 50 = 15 nM, Figure 6) was developed as a treatment for rheumatoid arthritis and atherosclerosis [66][67][68]. A recent phase II trial showed that oral varespladib was able to reduce progression of atherosclerosis and associated cardiovascular events, without any evidence of adverse effects [67]. In this regard, a novel class of azetidinones represented by SB-222657 (25, Lp-PLA 2 IC 50 = 11.7 nM, Figure 6) were developed as active site directed Lp-PLA 2 inhibitors [69]. Further studies led to the discovery of a potent Lp-PLA 2 inhibitor SB-435495 possessing a pyrimidinone ring template (26, Lp-PLA 2 IC 50 = 0.06 nM, Figure 6) by GSK. Lead optimization resulted in the development of darapladib (27, Lp-PLA 2 IC 50 = 0.25 nM; Figure 6) as a clinical candidate and is the first agent developed as an Lp-PLA 2 inhibitor to treat atherosclerosis and associated cardiovascular diseases [70,71]. Phase II studies of oral darapladib therapy led to reduced Lp-PLA 2 activity in human atherosclerotic plaques in patients with stable coronary heart disease and reduced the levels of inflammatory mediator interleukin-6 [61,[72][73][74]. Currently, darapladib is undergoing phase III trials.
Although selective COX-2 inhibitors decrease the formation of proinflammatory PGE 2 , they exhibit cardiovascular side effects due their suppression of COX-2 derived vasodilatory prostacyclin (PGI 2 ) biosynthesis [11,77,78]. Therefore, novel orally active small molecule mPGES-1 inhibitors are considered as an alternative strategy to develop anti-inflammatory agents with superior safety profile. Several ring templates including fatty acid derivatives have been developed as mPGES-1 and dual mPGES-1/5-LOX inhibitors [78].

Conclusions
The story of treating fever, pain and inflammation continues to evolve. Small molecule NSAIDs have dominated the market for over a century. Advances in molecular biology, crystallography and rational drug design approaches have led to the successful identification of novel anti-inflammatory targets such as 5-LOX, COX-2, , Lp-PLA 2 , mPGES-1 and TNF-α, to mention a few. The risks involved in this endeavor, is clearly highlighted by the "coxib" controversy. In an era where new drug pipelines are drying-up and blockbuster agents are facing generic competition, the discovery of novel anti-inflammatory targets continues to propel the development of small molecule therapeutics to treat inflammatory conditions. It is evident that a rational drug discovery effort that combines HTS and fragment screening techniques can provide novel small molecule ring templates that can be optimized by medicinal chemistry methods, to exhibit suitable in vivo activity and optimal pharmacokinetic properties. In spite of the current increase in market share of biological therapeutics to treat inflammatory conditions, small molecule therapeutics continues to dominate the pharmaceutical landscape. The recent advances in deciphering the ability of small molecules to disrupt protein-protein interactions in vivo, provides an exciting opportunity to discover novel small molecule therapeutics to treat inflammation and a wide variety of disease states.