1,2-Diarylcyclopentenes as selective cyclooxygenase-2 inhibitors and orally active anti-inflammatory agents.

A series of 1,2-diarylcyclopentene methyl sulfones and sulfonamides have been shown to be remarkably potent and selective cyclooxygenase-2 (COX-2) inhibitors. The methyl sulfone analogs 7 showed excellent COX-2 activity, with IC50s ranging from 0.003 (7f,n) to 0.87 (7o) microM. In addition, most analogs of 7 showed no activity (IC50 > 100 microM) against the COX-1 enzyme. Replacement of the methyl sulfone moiety with a sulfonamide group gave a slightly more potent (typically 2-5-fold) but less selective COX-2 inhibitor, mainly due to an increase (20- > 100-fold) in COX-1 activity. However, in vitro COX-1/COX-2 selectivity for the sulfonamides 8 could be increased in many cases by simply incorporating a substituent at the 3-position of the phenyl group. Furthermore, in vitro selectivity increased with the size and number of substituents, as demonstrated in the selectivity trend of 8k (8000) > 8j (1900) > 8i (500) > 8h (100). More importantly, the sulfonamide COX-2 inhibitors showed greatly enhanced oral activity in the rat model of established adjuvant-induced arthritis, with inhibition values of 79.0% (8a), 81.5% (8c), and 83.0% (8g) at 1 mg/kg. On the basis of its overall biological profile, sulfonamide 8c was evaluated as a potential clinical candidate, displaying an ED50 of 22 mpk in the rat carrageenan-induced paw edema model and an ED50 of 0.16 mpk in the rat established adjuvant-induced arthritis model with no indication of gastrointestinal toxicity in rats and mice at 200 mpk. In addition, a preparative-scale synthetic route to sulfonamide 8c has been developed.

Diarylspiro[2.4]heptenes as orally active, highly selective cyclooxygenase-2 inhibitors: synthesis and structure-activity relationships.

A novel series of 5,6-diarylspiro[2.4]hept-5-enes was shown to provide highly potent and selective cyclooxygenase-2 (COX-2) inhibitors. A study of structure-activity relationships in this series suggests that 3,4-disubstituted phenyl analogs are generally more selective than 4-substituted phenyl analogs and that replacement of the methyl sulfone group on the 6-phenyl ring with a sulfonamide moiety results in compounds with superior in vivo pharmacological properties, although with lower COX-2 selectivity. Several compounds have been shown to possess promising pharmacological properties in adjuvant-induced arthritis and edema analgesia models. The absence of gastrointestinal (GI) toxicity at 200 mpk of several selected compounds in rats and mice corresponds well with the weak potency for inhibition of COX-1 observed in the enzyme assay. Methyl sulfone 55 and sulfonamide 24 were shown to have superior in vivo pharmacological profiles, low GI toxicity, and good oral bioavailability and duration of action.

Synthesis and biological evaluation of the 1,5-diarylpyrazole class of cyclooxygenase-2 inhibitors: identification of 4-[5-(4-methylphenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl]benze nesulfonamide (SC-58635, celecoxib).

A series of sulfonamide-containing 1,5-diarylpyrazole derivatives were prepared and evaluated for their ability to block cyclooxygenase-2 (COX-2) in vitro and in vivo. Extensive structure-activity relationship (SAR) work was carried out within this series, and a number of potent and selective inhibitors of COX-2 were identified. Since an early structural lead (1f, SC-236) exhibited an unacceptably long plasma half-life, a number of pyrazole analogs containing potential metabolic sites were evaluated further in vivo in an effort to identify compounds with acceptable pharmacokinetic profiles. This work led to the identification of 1i (4-[5-(4-methylphenyl)-3-(trifluoromethyl)- H-pyrazol-1-yl]benzenesulfonamide, SC-58635, celecoxib), which is currently in phase III clinical trials for the treatment of rheumatoid arthritis and osteoarthritis.

Aryl sulfatase in ascospores of Neurospora crassa.

Neurospora crassa ascospores normally do not contain aryl sulfatase even when formed under conditions of sulfur limitation. However, when one of the parental strains is the nonrepressible mutant scon(c), the resulting (mixed) ascospores contain significant levels of aryl sulfatase even when formed under conditions of sulfur abundance.

Novel mutation causing derepression of several enzymes of sulfur metabolism in Neurospora crassa.

A group of enzymes of sulfur metabolism (arylsulfatase, cholinesulfatase, and a number of others) are normally repressed in Neurospora crassa by an abundant supply of a "favored" sulfur source such as methionine or inorganic sulfate. A mutant called scon(c) was isolated in which the formation of each of these enzymes is largely or completely nonrepressible. The structural genes for three of these enzymes have been mapped; scon(c) is not linked to any of them. It is also not linked to cys-3, another gene which is involved in control of the same group of enzymes. Two alleles of the structural gene for arylsulfatase [ars(+) and ars(UFC-220)] produce electrophoretically distinguishable forms of arylsulfatase. Heterokaryons with the constitution scon(c) ars(+) + scon(+)ars(UFC-220) were prepared. These heterokaryons produce both forms of arylsulfatase under conditions of sulfur limitation, but produce only the wild-type (ars(+)) form under conditions of sulfur abundance. When the alleles of ars and scon are in the opposite relationship, only the ars(UFC-220) form of arylsulfatase can be detected under conditions of sulfur abundance. Thus the effect of the scon(c) mutation seems to be limited to its own nucleus. The implications of these findings are discussed.

Absorption by the rhesus monkey of phenylalanine methyl ester and species differences in its metabolism by blood, plasma and intestinal mucosa.

Phenylalanine methyl ester (PM) is a decomposition product of the sweetening agent aspartame. The potential for absorption of PM was examined following intragastric and intraduodenal administration of 20-mg doses of [14C]PM to rhesus monkeys implanted with hepatic portal vein cannulae. Small amounts of unchanged PM (less than 0.1 micrograms/ml) were detected in portal and peripheral blood samples during the first 1-2 hours after administration, but none was detectable (less than 0.001 micrograms/ml) at later times. Comparison of the areas under the PM and total 14C blood concentration-time curves indicated that only 0.2% of the administered PM reached the portal blood unchanged, and 0.1% or less reached the peripheral blood unchanged. Blood and plasma from monkeys and humans hydrolyzed PM in vitro at very similar rates, but plasma PMase activity was much higher in the dog and rat than in the monkey or human. Hydrolysis of PM by intestinal mucosa homogenates was also faster for the rat and dog than for the monkey. The in vitro results suggest that absorption of intact PM in the human would be no greater than that found in the monkey.

Evidence for polymorphism in the canine metabolism of the cyclooxygenase 2 inhibitor, celecoxib.

The pharmacokinetics of celecoxib, a cyclooxygenase-2 inhibitor, was characterized in beagle dogs. Celecoxib is extensively metabolized by dogs to a hydroxymethyl metabolite with subsequent oxidization to the carboxylic acid analog. There are at least two populations of dogs, distinguished by their capacity to eliminate celecoxib from plasma at either a fast or a slow rate after i.v. administration. Within a population of 242 animals, 45.0% were of the EM phenotype, 53.5% were of the PM phenotype, and 1.65% could not be adequately characterized. The mean (+/-S.D.) plasma elimination half-life and clearance of celecoxib were 1.72 +/- 0.79 h and 18.2 +/- 6.4 ml/min/kg for EM dogs and 5.18 +/- 1.29 h and 7.15 +/- 1.41 ml/min/kg for PM dogs. Hepatic microsomes from EM dogs metabolized celecoxib at a higher rate than microsomes from PM dogs. The cDNA for canine cytochrome P-450 (CYP) enzymes, CYP2B11, CYP2C21, CYP2D15, and CYP3A12 were cloned and expressed in sf 9 insect cells. Three new variants of CYP2D15 as well as a novel variant of CYP3A12 were identified. Canine rCYP2D15 and its variants, but not CYP2B11, CYP2C21, and CYP3A12, readily metabolized celecoxib. Quinidine (a specific CYP2D inhibitor) prevented celecoxib metabolism in dog hepatic microsomes, providing evidence of a predominant role for the CYP2D subfamily in canine celecoxib metabolism. However, the lack of a correlation between celecoxib and bufuralol metabolism in hepatic EM or PM microsomes indicates that other CYP subfamilies besides CYP2D may contribute to the polymorphism in canine celecoxib metabolism.

Plasma protein binding of celecoxib in mice, rat, rabbit, dog and human.

The plasma protein binding of celecoxib was determined for animals and humans using in vitro and ex vivo methods. Eight, healthy, human volunteers (three male, five female, 20-39 years) received celecoxib (600 mg) BID for 7 days, blood samples were collected and concentrations of bound and unbound celecoxib determined. The fraction of bound drug in the volunteers was constant (97.4 +/- 0.1%) at total celecoxib plasma concentrations ranging from 0.01 to 4.02 microg/mL. The ex vivo plasma protein binding of celecoxib in the animals was concentration-independent up to approximately 12, 8 and 10 microg/mL for mouse, rat and dog, respectively. The plasma protein binding of celecoxib after a single oral dose of 10 and 300 mg/kg to mice was 98.3 +/- 0.2%, of 1 and 400 mg/kg to rats was 98.3 +/- 0.2% and of 1 and 100 mg/kg to dogs was 98.5 +/- 0.1%. The percent binding of celecoxib to plasma proteins in vitro was slightly lower than those values determined ex vivo. The in vitro binding of celecoxib to plasma protein was constant over the concentrations of 0.1-10 microg/mL for all species, except rat.

Identification of N-beta-L-aspartyl-L-phenylalanine as a normal constituent of human plasma and urine.

A new beta-aspartyl dipeptide, N-beta-L-aspartyl-L-phenylalanine (beta-AP), has been isolated and identified in urine and plasma from normal human volunteers. beta-AP was isolated from urine samples by high performance liquid chromatography (HPLC). Its identity and stereochemistry were demonstrated by HPLC and gas chromatography/mass spectrometry (GC-MS). The mean urinary beta-AP concentration in the subjects was 0.63 +/- 0.14 microgram/mg creatinine when averaged over two consecutive days of urine collection. Daily beta-AP excretion, determined from two 24-h urine samples collected from five individuals, was 801 +/- 117 micrograms/d (2.7 mumol/d). No diurnal rhythm was evident within the 24-h collection periods. beta-AP was also isolated from human plasma by HPLC and identified by GC-MS. Plasma from subjects contained approximately 5 ng beta-AP/ml. Furthermore, beta-AP was formed when asparagine and phenylalanine were incubated with an enzyme extract from human kidney. Thus, at least some of the beta-AP present in humans, and presumably other beta-aspartyl dipeptides as well, appears to be synthesized endogenously.

SC-46275: a potent, long-acting gastric antisecretory prostaglandin with low oral bioavailability in the dog.

The synthetic prostaglandin, SC-46275, an omega chain cyclopentenyl analog of enisoprost, was studied to determine its gastric antisecretory potency and duration of action in meal-stimulated innervated (Pavlov) pouch dogs and its p.o. bioavailability in unoperated fasted dogs. SC-46275 exhibited potent antisecretory activity when administered directly into the gastric pouch, the ED50 being 0.01 micrograms/kg. It had a long duration of antisecretory action; significant (P less than or equal to .05) inhibition of total acid output was observed 16 hr after intrapouch administration of 0.03 micrograms/kg. At this dose p.o., neither SC-46275 nor its free acid metabolite was detected in plasma. These data indicate that SC-46275 has novel properties: it is a potent, long-acting gastric antisecretory agent which is not readily available systemically after p.o. administration. Thus, potential systemic side effects are expected to be absent or minimized at doses of SC-46275 which inhibit gastric acid secretion, and therefore it might be useful in peptic ulcer disease.

Pharmacokinetics, tissue distribution, metabolism, and excretion of celecoxib in rats.

The pharmacokinetics, tissue distribution, metabolism, and excretion of celecoxib, 4-[5-(4-methylphenyl)-3-(trifluoromethyl)-1H-pyrazol-1-yl] benzenesulfonamide, a cyclooxygenase-2 inhibitor, were investigated in rats. Celecoxib was metabolized extensively after i.v. administration of [(14)C]celecoxib, and elimination of unchanged compound was minor (less than 2%) in male and female rats. The only metabolism of celecoxib observed in rats was via a single oxidative pathway. The methyl group of celecoxib is first oxidized to a hydroxymethyl metabolite, followed by additional oxidation of the hydroxymethyl group to a carboxylic acid metabolite. Glucuronide conjugates of both the hydroxymethyl and carboxylic acid metabolites are formed. Total mean percent recovery of the radioactive dose was about 100% for both the male rat (9.6% in urine; 91.7% in feces) and the female rat (10.6% in urine; 91.3% in feces). After oral administration of [(14)C]celecoxib at doses of 20, 80, and 400 mg/kg, the majority of the radioactivity was excreted in the feces (88-94%) with the remainder of the dose excreted in the urine (7-10%). Both unchanged drug and the carboxylic acid metabolite of celecoxib were the major radioactive components excreted with the amount of celecoxib excreted in the feces increasing with dose. When administered orally, celecoxib was well distributed to the tissues examined with the highest concentrations of radioactivity found in the gastrointestinal tract. Maximal concentration of radioactivity was reached in most all tissues between 1 and 3 h postdose with the half-life paralleling that of plasma, with the exception of the gastrointestinal tract tissues.