Discovery of bioavailable inhibitors of secretory phospholipase A2.

Substrate-mimetic inhibitors of sPLA2 with submicromolar in vitro potency were discovered by use of a novel dual substrate screening strategy. In vivo evaluation of selected inhibitors in the rat carrageenan paw edema model of inflammation, however, indicated that in vitro potency was not a good predictor of in vivo activity. Studies of the metabolic stability of early examples of these inhibitors suggested that the metabolic lability of these compounds was a major contributing factor to the observed weak in vivo activity. In an attempt to achieve improved in vivo activity, we prepared and tested compounds designed to overcome the observed metabolic instability. The design of the new compounds involved two types of changes in the inhibitor molecules. First, the C-2 ester moiety was replaced with an amide function so that direct cleavage by stomach acid and blood esterases at this site was minimized. Second, omega-oxidation of the decanamide moiety was eliminated by substitution of hydrogen with fluorine in this position. Compounds containing fluorine in the terminal positions of the alkyl chain retained sPLA2 inhibitory activity and also possessed improved in vitro metabolic stability and pharmacokinetic parameters relative to nonfluorinated inhibitors in this series. As exemplified by GW 4776, improvements in metabolic stability alone, however, were not sufficient to ensure oral activity. Thus, GW 4776 did not show oral activity in the carrageenan edema model and had only modest activity after i.v. dosing in the same model. In fact, the results for GW 9624 and GW 8219 suggested that factors in addition to potency of sPLA2 inhibition and metabolism affect the observed in vivo activity. Despite the fact that these two compounds varied only by a single oxygen-to-sulfur substitution, one was active whereas the other was not. One possible explanation for the observed variability is a compound-dependent difference in the rate of equilibration into tissue. This possibility is relevant as both the carrageenan paw edema model and the phorbol ester edema model involve a localized inflammation. No measurements were made to assess differences in the distribution of the different inhibitors between the blood and the localized site of inflammation. In summary, a series of bioavailable inhibitors of sPLA2 was prepared using an iterative approach that combined medicinal chemistry, in vitro and in vivo evaluation of biological activity, and metabolic and pharmacokinetic studies. Although some compounds in the series showed in vivo activity, the anti-inflammatory effect observed in animal models was modest and a decision was made to abandon sPLA2 as a molecular target for the development of anti-inflammatory agents.

Substrate specificity in short-chain phospholipid analogs at the active site of human synovial phospholipase A2.

The substrate specificity at the active site of recombinant human synovial fluid phospholipase A2 (hs-PLA2) was investigated by the preparation of a series of short-chain phospholipid analogs and measurement of their enzymatic hydrolysis at concentrations well below the critical micelle concentration. Substrates used in the study included 1,2-dihexanoylglycerophospholipids, 1,2-bis(alkanoylthio)glycerophospholipids, and 1-O-alkyl-2-(alkanoylthio)phospholipids. Turnover was observed for only a few of the 1,2-dihexanoylglycerophospholipids, and the rate of hydrolysis was very low, near the limit of detection of the assay. In contrast, selected 2-(alkanoylthio)-glycerophospholipids were hydrolyzed by hs-PLA2 at much higher rates at concentrations well below their critical micelle concentration (cmc). Thus, the 1,2-bis(hexanoylthio)glycerophosphatidylmethanol exhibits a k(cat)/K(M) = 1800 L mol-1 s-1. Over the calculated log P (cLogP) range of 3-9, cLogP and log(k(cat)/K(M) were linearly related for compounds with straight-chain sn-1 and sn-2 substituents. At comparable cLogP's, the sn-1 ethers and thioesters were hydrolyzed at comparable rates. A negative charge in the phosphate head group was required for enzyme activity. Unsaturation, aromaticity, and branching in the sn-2 substituent reduce turnover dramatically. The same structural modifications in the sn-1 substituent have less effect on turnover. Certain of these substrates, e.g., 1,2-bis(hexanoylthio)glycerophosphatidylmethanol, may be useful in assaying for active site inhibitors of PLA2. The structure--activity relationships established here for substrates should serve as a reference for the structure--activity relationships of substrate-based inhibitors.

6-Azasteroids: structure-activity relationships for inhibition of type 1 and 2 human 5 alpha-reductase and human adrenal 3 beta-hydroxy-delta 5-steroid dehydrogenase/3-keto-delta 5-steroid isomerase.

6-Azaandrost-4-en-3-ones were synthesized and tested versus human type 1 and 2 steroid 5 alpha-reductase (5AR) and human adrenal 3 beta-hydroxy-delta 5-steroid dehydrogenase/3-keto-delta 5-steroid isomerase (3BHSD) to explore the structure-activity relationship of this novel series in order to optimize potency versus both isozymes of 5AR and selectivity versus 3BHSD. Compounds with picomolar IC50's versus human type 2 5AR and low nanomolar Ki's versus human type 1 5AR with 100-fold selectivity versus 3BHSD were identified (70). Preliminary in vivo evaluation of some optimal compounds from this series in a chronic castrated rat model of 5AR inhibitor-induced prostate involution and dog pharmacokinetic measurements identified a series of 17 beta-[N-(diphenylmethyl)carbamoyl]-6-azaandrost-4-en-3-ones (compounds 54, 66, and 67) with good in vivo efficacy and half-life in the dog. Inhibitors with, at the minimum, low nanomolar potency toward both human 5AR's and selectivity versus 3BHSD may show advantages over previously known 5AR inhibitors in the treatment of disease states which depend upon dihydrotestosterone, such as benign prostatic hyperplasia.

The inotropic and beta blocking effects of a chimeric molecule that putatively inhibits both type III phosphodiesterase and beta adrenoceptors in anesthetized dogs.

The hemodynamic and beta adrenergic blocking effects of GI104313, a chimeric molecule containing a phosphodiesterase-inhibiting pyradazinone and a beta blocking phenoxpropanolamine, were examined in barbiturate-anesthetized, vagotomized dogs. The results of these studies were compared to those of indolidan, a known phosphodiesterase inhibitor, and xamoterol, a partial beta adrenoceptor agonist. The compounds were infused at six increasing dose rates in 10-min intervals. Isoproterenol (0.5 microgram/kg) was administered before each dose increment to determine beta adrenoceptor responsiveness. In a separate set of experiments, the hemodynamic effects of GI104313, indolidan and xamoterol were examined in the presence of complete beta blockade with atenolol. GI104313 elicited dose-dependent increases in heart rate, contractility (+dP/dt) and cardiac output and decreases in arterial blood pressure, left ventricular end diastolic pressure and systemic vascular resistance in unpretreated and atenolol-pretreated dogs. However, GI104313 was less potent hemodynamically in atenolol-pretreated animals. This was evidenced by a 4-fold dextral shift in the dose-response relation for several hemodynamic variables. In unpretreated dogs, GI104313 elicited potent dose-dependent blockade of the heart rate, diastolic blood pressure and +dP/dt responses to isoproterenol. Greater than 95% inhibition of isoproterenol response was attained at 1 mumol/kg GI104313 for all observed variables. Indolidan increased contractility and heart rate and decreased diastolic blood pressure in a dose-related fashion. Indolidan did not modify the stimulatory effects of isoproterenol. Atenolol had modest effects on indolidan's hemodynamic effect, only shifting its inotropic effect 2-fold. Xamoterol produced hemodynamic and beta blocking effects similar to GI104313.(ABSTRACT TRUNCATED AT 250 WORDS)